Department of Commerce, Community, and Economic Development
Alaska Oil and Gas Conservation
Commission
HOME
EVENTS
DATA
Data List
Drilling
Production
Orders
Data Miner
Document Search
REPORTS
Reports and Charts
Pool Statistics
FORMS
LINKS
Links
Test Notification
Data Requests
Regulations
Industry Guidance Bulletins
How to Apply
ABOUT US
History
Staff
HELP
Loading...
The URL can be used to link to this page
Your browser does not support the video tag.
Home
My WebLink
About
CO 137
Conservation Order Cover Page XHVZE This page is required for administrative purposes in managing the scanning process. It marks the extent of scanning and identifies certain actions that have been taken. Please insure that it retains it's current location in this file. .,L~'~_ Conservation Order Category Identifier Organizing RESCAN [] Color items: [] Grayscale items: [] Poor Quality Originals: [] Other: NOTES: DIGITAL DATA [] Diskettes, No. [] Other, No/Type OVERSIZED (Scannable with lame plotter/scanner) [] Maps: [] Other items OVERSIZED (Not suitable for plotter/scanner, may work with 'log' scanner) [] Logs of various kinds [] Other BY: ,,~MARIA Scanning Preparation TOTAL PAGEZ~_~/ Production Scanning ~~?z..~..~?~"~ [.~ Stage I PAGE COUNT FROM SCANNED DOCUMENT: ~'~2.~/q ~.~ ~ PAGE COUNT MATCHES NUMBER IN SCANNING PREP--AR/~TIOrN' ,~r YES NO Stage 2 IF NO IN STAGE 1, PAGE(S) DISCREPANCIES WERE FOUND: ~ YES NO (SCANNING IS COMPLETE A~'I~IS POINT UNLESS SPECIAL ATTENTION IS REQUIRED ON AN INDIVIDUAL PAGE BASIS DUE TO QUALITY, GRAYSCALE OR COLOR IMAGES) General Notes or Comments about this Document: 5/21/03 ConservOrdCvrPg.wpd STATE OF ALASKA DEPARTMENT OF NATURAL RESOURCES DIVISION OF OIL AND GAS Alaska Oil and Gas Conservation Committee 3001 Porcupine Drive Anchorage, Alaska 99501 Re: The motion of the Alaska Oil and ) Gas Conservation Committee to hold ) a hearing pursuant to Title 11, ) Alaska Administrative Code, Sec- ) tion 22.540 to consider an amend- ) ment to Rule 4, Conservation Order ) 98-A, 98-B and 83-C to require ) bag-type blowout preventers while ) drilling the hole in which the sur-) face casing is set and to consider ) an amendment to Rule 3, Conserva- ) tion Order 98-A, 98-B and 83-C ) concerning changes in casing re- ) quirements in the permafrost por- ) tion of the hole. ) Conservation Order No. 137 Prudhoe Bay Field Prudhoe Oil Pool Prudhoe Bay Kuparuk River Oil Pool Prudhoe Bay Lisburne Oil Pool January 9, 1976 IT APPEARING THAT: 1. The Alaska Oil and Gas Conservation Commit'tee moved to hear testimony amending parts of the existing rules affecting blowout preventers and. casing requirements for all three of the established oil pools in the Prudhoe Bay Field. 2. Notice of public hearing was published in the Anchorage Daily News on October 25, 1975. 3. A public hearing was held on November 25, 1975 in the Municipal Chambers in the Z. J. LouSsac Library, 5th Avenue and F Street, Anchorage, Alaska at whiCh time interested parties were heard. 4. The hearing was continued until the close of business on December 10, 1975 to permit filing of additional statements. Additional statements were received. i CO 137 FINDINGS: · · , · ¸. Industry has performed numerous laboratory, field and engineering studies in an effort to more fully evaluate thaw subsidence and freezeback problems. Thawing of the permafrost causes sediment compaction and strain in the casing. The strain producing mechanism is complex and is directly related to the amount of thaw in the permafrost. Strains are due predominantly to an interaction of alternating layers of sand and silt and are dependent upon the relative thicknesses and the mechanical properties of these layers. The maximum expected tensile and compressive strains as a function of depth, lithology, and radius of thaw can be predicted. The maximum strains measured in the casing joints in a 5-spot thaw test well, representative of a Prudhoe Oil Pool producing oil well, were 0.13% in compression and 0.08% in tension. 6. The worst case strains calculated for a representative Prudhoe Oil Pool producing oil well are 0.7% compressive .and 0.5% tensile. These worst case strains were predicted by a mathematical model which was ~t~pe~--~-to~...ma~-~ field test results· 7. Casing is available which can withstand strains in oil well completions through the permafrost interval. . · 10. 11. Full scale tensile and compressive tests and finite element model studies of the API N-80 Buttress threaded casing show that a 13-3/8 inch, 72 pounds/foot, N-80 Buttress threaded connection can resist minimum ultimate compressive and tensile axial strains in the pipe body of 2.3% compressive and 3.4% tensile. This casing is now being utilized for development oil wells and has design factors of 3.3 in compression and 6.8 in tension. These design factors are arrived at by dividing 2.3% compressive strain.by the worst case strain of 0.7% and 3.4% tensile strain by the worst case strain of 0.5%. Axial strain data for other types of casing have not been made available to the Committee. The highest casing design factor in common use is 1.8. Early industry concerns that the upper 500 feet of sediments, or first gravels, contained excess ice and that thaWing would cause large permafrost and casing strains have been disproved. Recent information indicates no excess ice below 50 feet. Differential vertical strains in the thawed region will be minimized by shear failure in the surrounding frozen material. Maximum casing strains in the top 430 feet are less than half of the maximum strains measured below that point. The term annular blowout preventer appears to be a more precise term than bag-type preventer. 1 CO 137 12. 13. 14. 15. 16. 17. 18. Over 100 wells have been drilled through the base of the permafrost within the defined pool limits east of the Kuparuk River without encountering hydrocarbons down to 2700 feet. Geological conditions to the base of the permafrost are unknown within parts of defined pool areas. The first well drilled from any drill site may provide sufficient information to determine if blowout prevention equipment is required when drilling surface hole for subsequent wells from the same drill site. The blowout prevention equipment utilized until surface casing is set should function only as a diverter system to divert well flow away and downwind from the drilling rig. The diverter system piping design should be of sufficient size to minimize restriction to flow. The diverter system should be designed to prevent freeze-up when operating in arctic conditions. Since the Prudhoe Oil Pool overlies much of the Lisburne Oil Pool, the rule for blowout prevention equipment and practice should be identical for both pools. CONCLUSIONS:' · 1 . . · Thaw-induced strains are directly related to the amount of thaw in the permafrost and becOme significant for wells producina large volumes of warm fluids over long periods of time. The permafrost compaction-casing strain relationship requires new casing performance criteria that specify tensile and compressive strain capacities of the casing and couplings in contact with the perma- frost. Casing designs should provide minimum axial strain properties greater than the 'worst case strains, calculated to be 0.7% in compression and 0.5% in tension. A design factor of 1.8 applied to the worst case tensile and compressive strains appears adequate to protect the surface casing against damage from thaw subsidence. Thus the criteria for minimum axial post-yield strains would be 1.26% for compression (1.8 x 0.7%) and 0.9% for tension (1.8 X 0.5%). Application of thaw-induced strain criteria is only significant for casing strings playing~a role in the integrity of the well. The alternating axial strains and localized nature of the strains associated with the permafrost thaw are such that the effect on casing cemented to the permafrost would not be transmitted to the next inner string if the two strings were not cemented together through the permafrost interval. m CO 137 . e The casing programs for injection wells, gas wells, observation wells, low volume oil wells, and exploratory wells need not meet the same criteria for permafrost thaw protection as high volume oil wells. Annular blowout preventers and a flow diverter system of adequate size should be properly installed for use while drilling the hole in which the surface casing will be set. This equipment may not be required if drilling experience at a multiple drill site indicates that this equipment is unnecessary. NOW, THEREFORE, IT IS ORDERED: Rule 1. Conservation Order No. 98-A, Rule 3 and Rule 4, is amended to read as follows: Rule 3. Casing and Cementing Requirements (a) (b) (c) (d) (e) Casing and cementing programs shall provide adequate protection of all fresh waters and productive formations and protection from any pressure that may be encountered, including external freezeback within the permafrost. For proper anchorage and to prevent an uncontrolled flow, a conductor casing shall be set at least 75 feet below the surface and sufficient cement shall be used to fill the annulus behind the pipe to the surface. For proper anchOrage, to prevent uncontrolled flow and to protect the well from the effects of permafrost thaw, a string of surface casing shall be set at least 500 feet below the base of the permafrost section but not below 2,700 feet unless a greater depth is approved by the Committee upon a showing, that no potentially productive pay exists above the proposed casing setting depth, and sufficient cement shall be used to fill the annulus behind the pipe to the surface. The surface casing' shall have minimum post-yield strain properties of 0.9% in tension and 1.26% in compression. If the surface casing does not meet the strain requirements in (c) above, the integrity of the well shall be protected from the effects of perma- frost thaw by running an inner string of casing also set at least 500 feet below the base of the permafrost section and properly cemented except that the two casing strings shall not be bonded together within the permafrost section. This inner string of casing shall not be utilized as production casing. Other means for maintaining the integrity of the well from the effects of permafrost thaw may be approved by the'Committee upon application. m CO 1 37 (f) Production casing shall be landed through the completion zone and cement shall cover and extend to at least 500 feet above each hydrocarbon- bearing formation which is potentially productive. In the alternative, the casing string may be set and adequately cemented at an intermediate point and a liner landed through the completion zone. If such a liner is run, the casing and liner shall overlap by at least 100 feet and the annular space behind the liner shall be filled with cement to at least 100 feet above the casing shoe, or the top of the liner shall be squeezed with sufficient cement to provide at least 100 feet of cement between the liner and casing. Cement must cover and extend at least 500 feet above each hydrocarbon-bearing formation which is potentially productive. (g) Casing and l'iner, after being cemented, shall be satisfactorily tested to not less than 50% of minimum internal yield pressure or 1,500 pounds per square inch, whichever is less. (h)' No well shall be produced through the annulus between the tubing and the casing unless a cement sheath extends from the top of the pay to the shoe of the next shallower casing string. Rule 4. Blowout Prevention Equ.ipment and Practice. (a) (b) (c) The use of blowout prevention equipment shall be in accordance with good established practice and all equipment shall be in good operating condition at all.times. All blowout prevention'equipment shall be adequately protected to ensure reliable operation under the existing weather conditions. All blowout prevention equipment shall be checked for satisfactory operation during each trip. Before drilling below the conductor string, each well shall have installed at least one.remotely controlled annular type blowout preventer and flow diverter system. The annular preventer installed on the conductor casing shall be utilized to permit the diversion of hydrocarbons and other fluids. .This low pressure, high capacity diverter system shall be installed to provide at least"the equivalent of a 6-inch line with at least two lines .venting in different directions to insure downwind diversion and shall be designed to avoid freeze-up. These lines shall be equipped with full-opening butterfly type valves or other valves approved by the Committee. A schematic diagram, list of equipment, and operational pro- ' cedure for the diverter:system shall be submitted~with the application Permit to Drill or Deepen (Form 10-401) for approval. The above require- ments may be waived for subsequent wells drilled from a multiple drill site. Before drilling below the surface casing all wells shall have three remotely controlled blowout preventers, including one equipped with pipe rams, one with blind, rams and one annular type. The blowout pre- venters and associated equipment shall have 3000 psi working pressure and 6000 psi test pressure. N CO 137 (d) (e) The associated equipment shall include a drilling spool with minimum three-inch side outlets (if not on the blowout preventer body), a mini- mum three-inch choke manifold, or equivalent, and a fill-up line. The drilling string will contain full-opening valves above and immediately below the kelly during all circulating operations with the kelly. Two emergency valves with rotary subs for all connections in use will be conveniently located on the drilling floor. One valve will be an inside blowout preventer of the spring-loaded type. The second valve will be of the manually-operated ball type, or any other type which will perform the same function. All ram-type blowout preventers, kelly valves, emergency valves and choke manifolds shall be tested to required working pressure when installed or changed and at least once each week thereafter. Annular preventers shall be tested to 50% recommended working pressure when installed and once each week thereafter. Test results shall be recorded on written daily records kept at the well. Rule 2. Conservation Order No. 98-B, Rule 3 and Rule 4, is amended to read as follows: Rule 3. Casing and Cementing Requirements (a) (b) (c) (d) (e) Casing and cementing programs shall provide adequate protection of all fresh waters and productive formations and protection from any pressure that may be encountered, including external freezeback within the permafrost. For proper anchorage and to prevent an uncontrolled flow, a conductor casing shall be set at least 75 feet below the surface and sufficient cement shall be used to fill the annulus behind the pipe to the surface. For proper anchorage, to prevent uncontrolled flow and to protect the well from the effects of permafrost thaw, a string of surface casing shall be set at least 500 feet below the base of the permafrost section but not below 2,700 feet unless a greater depth is approved by the .Committee upon a showing that no potentially productive pay exists above the proposed casing ~setting depth, and sufficient cement shall be used to fill the annulus behind the pipe to the surface. The surface casing shall have minimum Post-yield strain properties of 0,9% in tension and 1.26% in compression. If the surface casing does not meet 'the strain requirements in (c) above, the integrity of the well .shall be protected from the effects of perma- frost thaw by running an inner string of casing also set at least 500 feet below the base of the permafrost section and properly cemented except that the two casing strings shall not be bonded together within the permafrost section. This inner string of casing shall not be utilized as production casing. Other means for maintaining the integrity of the well from the effects of permafrost thaw may be approved by the Committee upon application. I (~(' (~ CO 137 (f) (g) (h) Production casing shall be landed through the completion zone and cement shall cover and extend to at least 500 feet above each hydrocarbon- bearing formation which is potentially productive. In the alternative, the casing string may be set and adequately cemented at an intermediate point and a liner landed through the completion zone. If such a liner is run, the casing and liner shall overlap by at least 100 feet and the annular space behind the liner shall be filled with cement to at least 100 feet above the casing shoe, or the top of the liner shall be squeezed with sufficient cement to provide at least 100 feet of cement between the liner and casing. Cement must cover and extend at least 500 feet above each hydrocarbon-bearing formation which is potentially productive. Casing and liner, after being cemented, shall be satisfactorily tested to not less than 50% of minimum internal yield pressure or 1,500 pounds per square inch, whichever is less. No well shall be produced through the annulus between the tubing and the casing unless a cement sheath extends from the top of the pay to the shoe of the next shallower casing string. Rule 4. Blowout Prevention Eq.uipment and Practice. (a) (.b) (c) The use of blowout prevention equipment shall be in accordance with good established practice and all equipment shall be in good operating condition at all times. All blowout prevention equipment shall be adequately protected to ensure reliable operation under the existing weather conditions. All 'blowout prevention equipment shall be checked for satisfactory operation during each trip, Before drilling below the conductor string, each well shall have installed at least one remotely controlled annular type blowout preventer and flow diverter system. The annular preventer installed on the conductor casing shall be utilized to permit the diversion of hydrocarbons and other fluids. This low pressure, high capacity diverter system shall be installed to provide at least the equivalent of a 6-inch line with at least two lines venting in different directions to insure downwind diversion and shall be designed to avoid freeze-up. These lines shall be equipped with full-opening butterfly type 'valves or. other valves'approved by the Committee. A schematic diagram, list of equipment, and :operational pro- cedure for the diverter system shall be submitted with the application Permit to. Drill or Deepen (Form 10-401) for approval. The above require- ments may be Waived for subsequent wells drilled from a multiple drill site. Before drilling below the surface casing all wells shall have three remotely controlled'blowout preventers, including one equipped with pipe rams, one with blind rams and,one annular type. The blowout pre- venters and associated equipment shall have 3000 psi working pressure and 6000 psi test pressure. 7 CO 137 (d) (e) (f) Before drilling into the Prudhoe Oil Pool, the blowout preventers and associated equipment required in (c) above shall have 5000 psi working pressure rating and 10,000 psi test pressure rating. The associated equipment shall include a drilling spool with minimum three-inch side outlets (if not on the blowout preventer body), a mini- mum three-inch choke manifold, or equivalent, and a fill-up line. The drilling string will contain full-opening valves above and immediately below the kelly during all circulating operations with the kelly. Two emergency valves with rotary subs for all connections in use will be conveniently located on the drilling floor. One valve will be an inside blowout preventer of the spring-loaded type. The second valve will be of the manually-operated ball type, or any other type which will perform the same function. All ram-type blowout preventers, kelly valves, emergency valves and choke manifolds shall be tested to required working pressure when installed or changed and at least once each week thereafter. Annular preventers shall be tested to 50%.recommended working pressure when installed and once each week thereafter. Test results shall be recorded on written daily records kept. at the well. RUle 3. Conservation Order No. 83-C, Rule 3 and Rule 4, is amended as follows: Rule 3. Casing. and Cementing Requirements (a) ('b) (c) (d) Casing and cementing programs shall provide adequate protection of all fresh waters and productive formations and protection from any pressure that may be encountered, including external freezeback within the permafrost. For proper anchorage and to prevent an uncontrolled.flow, a conductor casing shall be set at least 75 feet below the surface and sufficient cement shall be used to fill the annulus behind the pipe to the surface. For proper anchorage, to prevent uncontrolled flow and to protect the well from the effects of permafrost thaw, a string of surface casing shall be set at least 500 feet below the base of the permafrost section but not below 2,700 feet unless a greater depth is approved by the Committee upon a showing that no potentially productive pay exists above the proposed casing setting depth, and sufficient cement shall be used .to fill the annulus behind the pipe to the surface. The surface casing shall have minimum post-yield strain properties of 0.9% in tension and 1.26% in compression. If the surface casing does not meet the strain requirements in (c) above, the integrity of the well shall be protected from the effects of perma- frost thaw by running an inner string of casing also set at least 500 feet below the base of the permafrost section and properly cemented except that the two casing strings shall not be bonded together within the 8 CO 137 (e) (f) (g) (h) permafrost section. This inner string of casing shall not be utilized as production casing. Other means for maintaining the integrity of the well from the effects of permafrost thaw may be approved by the Committee upon application. Production casing shall be landed through the completion zone and cement shall cover and extend to at least 500 feet above each hydrocarbon- bearing formation which is potentially productive. In the alternative, the casing string may be set and adequately cemented at an intermediate point and a liner landed through the completion zone. If such a liner is run, the casing and liner shall overlap by at least 100 feet and the annular space behind the liner shall be filled with cement to at least 100 feet above the casing shoe, or the top of the liner shall be squeezed with sufficient cement to provide at least 100 feet of cement between the liner and casing. Cement must cover and extend at least 500 feet above each hydrocarbon-bearing formation which is potentially productive. Casing and liner, after being cemented, shall be satisfactorily tested to not less than 50% of minimum internal yield pressure or 1,500 pounds per square inch, whichever is less. No well shall be produced through the annulus between the tubing and the casing unless a cement sheath extends from the top of the pay to the shoe of the next .shallower casing string. Rule 4. BlowoUt. Prevention Equipment and Practice. (a) (b) The use of blowout prevention equipment shall be in accordance with good,established practice and.all equipment shall be in good operating condition at all times. All blowout prevention equipment shall be adequately protected to ensure reliable operation under the existing weather conditions. All blowout prevention equipment shall be checked for satisfactory operation during each trip. Before drilling below the conductor string, each well shall have installed at least one remotely controlled annular type blowout preventer and flow diverter sYstem. The annular preventer installed on the conductor casing shall be utilized to permit the diversion of hydrocarbons and other fluids. This low pressure, high capacity diverter system shall be installed to provide at least the equivalent of a 6-inch. line with at least two lines venting in different directions to insure downwind diversion and shall be designed to avoid freeze-up. These lines shall be equipped with full-opening butterfly type valves or other valves approved by the Committee. A schematic di'agram, list of equipment, and operational pro- cedure for'the diverter system shall be submitted with the application Permit to Drill or Deepen (Form 10-401) for approval. The above require- ments may be waived for subsequent wells drilled from a multiple drill site. m CO 137 (c) (e) (f) Before drilling below the surface casing all wells shall have three remotely controlled blowout preventers, including one equipped with pipe rams, one with blind rams and one annular type. The blowout pre- venters and associated equipment shall have 3000 psi working pressure and 6000 psi test pressure. Before drilling into the Prudhoe Oil Pool and/or the Lisburne Oil Pool, the blowout preventers and associated equipment required in (c) above shall have 5000 psi working pressure rating and 10,000 psi test pressure rating. The associated equipment shall include a drilling spool with minimum three-inch side outlets (if not on the blowout preventer body), a mini- mum three-inch choke manifold, or equivalent, and a fill-up line. The drilling string will contain full-opening valves above and immediately below the kelly during all circulating operations with the kelly. Two emergency valves with rotary subs for all connections in use will be conveniently located on the drilling floor. One valve will be an inside blowout preventer of the spring-loaded type. The second valve will be of the manually-operated ball type, or any other type which will perform the same function. All ram-type blowout preventers, kelly valves, emergency valves and choke .manifolds shall be tested to required .working'pressure when' installed or changed and at. least once each week thereafter. Annular preventers shall be tested to 50% recommended .working pressure when installed and once each week thereafter. Test results shall be recorded on written daily records kept at the well. 10- CONSERVATION ORDER NO. 137 January 9, 1976 Done at Anchorage, Alaska, and dated January 9, 1976: Harry W. Kugler/Acting ExecUtive Secretary Alaska Oil and Gas Conservation Committee Concurrence: Thomas R. Marshall, ., Acting irman Alaska Oil and Gas Conservation Committee Hoyl~ H. ~mil ton ~ Member Alaska Oil and Gas Conservation Committee Conservation August 21, 1978 Re: ~,{aiver of flow diverter system and annular type blowout .preventer requirements per C.O. 137, Rule 4(b) for the "Q" drilling pad. ~,~r. P.R. Jud(] District Drilling Engineer Sohio Petroleum Company Pouch 6-612 Anchorage, Alaska 99502 Dear ~(r. J~dd: This is in response 'to your request, dated August 17, .1978, for a waiver of the requirements of Conservation Order No.. 1.37, Rule. 4(b) concerning Ge use of a diverter syste~n and annular preven- ters for future wells to be drilled from the '~'Q" drilling pads. This request for waiver is based on your report that no gas or ~,xcess pressures ~.,~ere encountered in the drillinq of the surface hole for th~ P.rudhoe Bay Unit ~Q-3 well, the first well drille~l from the "Q" drilling pad. This request for waiver is granted for all. wells drille~d from the "Q" drilling pad, provi~led no 'gas. or abnormal pressures are encountered .in the surface holes of: s~sequent wells° Yours very truly, \ ~arry W~. Ku~ler Executive Sec~ta~ Alaska Oil & Gas Conservation Co~ittee i{~'~K ~' be Conserva ti on J, Harch 24, 1977 Frank I~. District Drilling Engineer Atlantic Richfield Company P. O. Dox 360 Anct~orage, Alaska 995~)1 Re- llaiver of flow diverter system and annular type blowout~ preventer requiremenl~ per CO 137 Rule 4 (b), Well ~Io. DS 5-7 ~ Dear Hr, Broun: This is in response to your request dated !~arch 21, 1,°.77, for a waiver of t'l~e requirer:'~nt of Conservation Order. ii(}. 137, Rule .q (h) concerning the use of a diverter system when drilling the surface )lole of the captioned well on the grounds that no gas or abnor~al pressures have been encountered in the area developed from Drill Site 5 during the drilling to date. This request for waiver is granted for the subjec'~ well and any addi- tional wells drilled from this pad, provided no gas or ahnorn,)al pressures are encountered in subsequent drilling from [)rill Site 5. Sincerely, lhos. R. Marshall, Jr. Executive SeCretary Alaska 011 and Gas Conservation C(m'~lttec: March 2'4, 1977 G. D. Taylor Eanager Operations BP Alaska Inc. P.' O..Dox 4-1379 Anchorage, Alaska 99503 RE: Waiver of flow diverter system and annular type blowout preventer requirements per CO 137 Rule 4 (b), I,~'ell I'!o. Sag Delta No. 4 ~ /' (36-12-16) Dear Nr. Taylor: This is in response to your request for a ¥,,aiver of the requirement of Conservation Order tto, 137, Rule 4 (b) concerning the use of a diverter system when drilling the surface hole of the captioned well on the grounds that no gas or abnormal pressures have been encouatered in the drilling of Sag Delta ~3 at the same surface site. This request for waiver is granted for 'the subject well and any addi- tional wells drilled from this site, provided no gas or abnormal pressures are encountered in subsequent drilling. Sincerely, / // " ~ i/ Thos. R. fiarshall, Jr. Executtve Secretary Alaska Oil and Gas Conservation Conm~tttee l'RM:bJm Conservation I~arch 22, 1977 Frank M. Dro~m District Drilline Engineer Atlantic Richfield Company P. O. Box 360 Anchorage,. Alaska 99501 Re' Waiver of flow diverter system and annular type blo~,:out preventer requirements per CO 137 Rule 4 (b), l~ell Number DS-9 Dear Mr. Brown:' This is in response to your request for a waiver of the requirement of Conservation Order.No. 137, Rule 4 (b) concerning the use of a diverter system when drilling the surface hole of the captioned' well on the grounds that no gas or abnormal pressures have been encountered in the area developed from Drill Site 9 during the drilling to date. This request for waiver is granted for the subject well and any addi- tional wells drilled from this pad, provided no gas or abnormal pressures are encountered in subsequent drilling from Drill Site 9. Sincerely, Thos. R. Marshall, Jr. Executive Secretary Alaska Oil and Gas Conservation Conmlittee lP~.l:bJm AtlanticRichfieldCompany North American Producing Division North Ala~' District Post Office ~ox 360 Anchorage, Alaska 99510 Telephone 907 277 5637 March 21, 1977 Mr. Hoyle Hamilton State of Alaska Department of Natural Resources Division of Oil and Gas 3001 Porcupine Drive Anchorage, Alaska 99501 Subject' Waiver to Conservation Order No. 137 ..... .C'_..ENG I I ENG 2 ENG ! 3 E,",'C- 1 .., r, t ...r' i--I ......... ssc I Dear Mr. Hamilton' We are hereby requesting waiver to Conservation Order No. 137 dated January 9, 1976, Rule 4, Paragraph (b) for the remaining wells to be drilled from Drill Site 5, i.e., wells DS 5-7 through 5-18. At the present time, six surface holes have been drilled from this location and no hydrocarbon producing formations have been found. In our judgment, the annular type blowout preventer and flow diverter system is not required for drilling the remaining surface holes from this location. Very truly yours, Frank M. Brown District Drilling Engineer FMB/RAR/jw ticRichfieldCompany North Am~ri,c, an Producing Division North Ala~,-~}LDistriCt Post Offic~ ~'~ox 360 Anchorage, Alaska 99510 Telephone 907 277 5637 March 17, 1977 ' Mr. Hoyle Hamilton State of Alaska Department of Natural Resources Division of Oil and Gas 3001 Porcupine Drive Anchorage, Alaska 99501 Subject' Waiver to Conservation Order No. 137 Dear Mr. Hamilton' We are hereby requesting waiver to Conservation Order No. 137 dated January 9, 1976, Rule 4, Paragraph (b) for the remaining wells to be drilled from multiple Drill Site Nine, i.e. wells DS 9-12 through 9-20. At the present time, eleven surface holes have been drilled from this location and no hydrocarbon producing formations have been found. In our judgment, the annular type blowout preventer and flow diverter system is not required for drilling the remaining surface holes from this location. Very truly yours, Frank M. Brown District Drilling Engineer FMB/CSA/jw RECEIVED MAR 2 1977 Divtelon of Oil & I~ ~r~l~n Conservation March 8, 1977 P. R. Judd Oi strict Drill lng Engineer BP Alaska Inc. P. O. Box 4-1379 Anchorage, Alaska 99503 Re: Waiver of diverter requirement per CO 137 Rule 4 (b) and surface casing depth s.etttng 'requirement of Rule 3 (c) Well ~)umber )~-7 {2g-ll .13), Prudhoe Bay Oi 1 Field .Dear Mr. Judd: This is in response to your request for a waiver of the requirement of Conservation Order No. 137, Rule 4 {b) concerning the use of a diverter when drilling the surface hole of the..captioned well on the grounds that no gas. or abnormal pressures have been encountered in the area developed from ",l" pad. during the drilling t.o date.. Also waived is the requirement to set the surface casing above 2700 feet. on grounds that an ideal casing seat exists Slightly .below this depth. This request for waiver is granted for the subject well and any addi- tional wells, drilled from this pad, provided no gas or abnormal pressures are encountered in.subsequent drilling from I(:' pad. Thos. R. Mar.sba:Il, Jr. Executive Secretary Alaska Oil' and Gas' Conservati°n Committee AFFII AVlT OF PUBLICATION STATE OF ALASKA, ) THIRD JUDICIAL DISTRICT, ) ss. Vicki Cunningham being first duly sworn on oath deposes and says that ...... she .... is the .... L..e..g~.l.....c..]:.e.r.~ .... of the , Anchorage News, a daily news- paper. That said newspaper has been approved as a legal news- paper by the Third Judicial Court, Anchorage, Alaska, and it is now and has been published in the English language continually as a daily newspaper in Anchorage, Alaska, and it is now and during all of said time was printed in an office maintained at the aforesaid place of publication of said news- paper. That the annexed is a true copy of a Np.$.i.c.e....~f..~..q.b..Z..~.¢ as it was published in regular issues (and not in supplemental form) of said newspaper for. · period of ....0.a...e. ........ insertions, commencing on the ..... ..l.~.tb. day of ........ Nov. embe~.,19 ....'7.6, and ending on the .... ],~ ........day of both dates inclusive, and that such newspaper was regularly distributed to its subscribers dur- ing all of sa,id period. That the full amount of the fee charged for the foregoing publication is Hearing..... the sum of $ 21,00 which amount has been paid in full at the rate of 30 f per line; Mini- mum charge $10.00. me thistgtb., day of .... ~ovembe~- 19.~.6.. Notary Public in and for the State of Alaska, Third Division, Anchorage, Alaska MY CON~ISSION EXPIRES NOTICE OF PUBLIC HEARING STATE OFALASKA ' DEPARTMENT OF NATURAL RESOURCES Division,of, Oil ond Gas Conservation , 'Alaska. Oil and Gas Conservation Co.mmittee ., ConservatiOn' File No. 143 ',Re: Atlantic RiChfield, Company ~'udhOerBay Drill Site', . . , ,o.,-~ We l!l..' .. . ' ..... : an exCeptlonto Rule 3(c) o( Con: servation Order' No. 137. whl~:h' states.in pdrt "...surface casing shall be:"set 'at least 500' feet below-the base of the permafrost sect.Ipn;')'*," '.~ Thl~, well ~(~s drilled orig]'nall~ · a~. a permafrost observatldn .~ell and :compllcatlohs arose ~hlch .peEmlffed the surface c~slng tb b~ set only fo a de~th 379 fee. I;;,t~l.,aw'.fhe base of fha permat~os~Jlefore completion, the operatorY;proposes, to set a string of 9 $/8~' Protection caslnd In the'hole, l~fore hydrocarbon 'bearing 4or:matlpns will be erl- P~rtl'e~he.~;y be ag~r;e~ea It the r.e~l'l~te)l.er, de.r. I~ Issued are allo~ll)l~,l~ d'ak~Tf~bm' the'dal'e of thl,~l[~lJ~..tJ~n, ini.whlch, to file a PrO~t~.h_'.d,;r~lUeSt for hearl.ng. _P ~ e~'.~.f~i~.C~ [s. e00t P,urcu.me urlv~ A~na~..g6, :AI0sk'b 99501. if~t Is..tlmel¥ filed, i~h. e' matter will be I~'ve address at 11: ember 29, 1976 at ., protestants and otb- be heard, If no such timely filed, the. Com- mittee will consider the Issu- ance of the order without a hear- lng., . ' · : . · ,.. ..,.:..~ · . Thomas:R,.~brshalt; Jr., Executive Se~'etarv Alaska Oil and Gas Conservotio~ Committee 3001 ~Purcupli~eDHve; "Aritho'rage; A'lbsl~a' 99501 ........... i~/.:L~ ........................ ,9 ...... .?-7 ! L.I L/Ir.? LL!? /x-,.l l ,/ I)I']PAI{TMENT OF NATURAL RESOURCES / Division of Oil and Gas ' / 3001 Porcupine Drive, Anchorage, Alaska 99501 ! JAY S. HAM~OIIfD, GOVE~/VOR July 20, 1976 G. D. Taylor Manager Operations BP Alaska, Inc. P. O. Box 4-1379 Anchorage, Alaska 99509 Re: Casing Design in the Permafrost Interval Dear Mr. Taylor: Your letter of June 28, 1976, stated that five of your Prudhoe Bay wells had 13 3/8" surface casing strings through the permafrost which included 68 lb/ft, N-80 buttress ~casing. These wells are A-7, C-5, C-6, C-7 and C-8. Based on your studies you submit that 13 3/8", 68 lb/ft, N-80 buttress casing satisfies Rule 3(c) of Conservation. Orders No. 98A, No. 98B, and No. 83C. The above.Rule 3(c) states that surface casing through the permafrost shall have minimum post-yield strain properties of 0.9% in tension and 1.26% in compression. Arco's May 1975 report, "Prudhoe Bay Field Permafrost Casing. and Well Design for Thaw Subsidence Protection", describes results of full.- scale strain limit tests on 13 3/8", 72 lb/ft, modified N-80 (HN-80) buttress casing and'of a finite element model calibrated from the tests. Results of Arco's tests and studies indicated 13 3/8", 72 lb/ft, MN-80 buttress casing exceeded requirements of Rule 3(c). One of the prime reasons that controlled yield or modified MN-80 casing 'was specified in the Arco tests was to preclude the practice of manufac- turer's substitution for N-80, with higher steel grades which, due to physical properties, fail in Arctic applications. Therefore, approval of N-80 grade casing for surface casing through the permafrost requires assurance that other grades have not been substituted for N-80. ,Ok.LiT IO/v 1276 - 191 r° "1776-A 7i]IBUT'E FROM OUti STATE '1'0 OUR NA TION-197G" G. D. Taylor -2- H. H. Hamilton July 20, 1976 Your studies using Arco's finite element model demonstrated that failure strain in compression for 13 3/8", 68 lb/ft, MN-80 buttress casing was 4.2%. Results of finite model calculations shown in Table III-1 of the above mentioned Arco report alludes that failure strain in' tension for the same casing exceeds that of 13 3/8", 72 lb/ft, MN-80 .buttress casing. Our analysis supports this point. Based on these studies, 13 3/8", 68 lb/ft, MN-80 buttress casing adequately meets the requirements of Rule 3(c) of Conservation Orders No. 98A, No. 98B, and Ho. 83C. Sincerely, Hoyle H. Hamilton Acting Director ' MAILING ADDRESS: BP AL.ASKA INC · ANCHORAGE. ALASKA 99509 ~ 3111 -C- STREET Director Division of Oil and Gas Department of Natural Resources State of Alaska 3001 Porcupine Drive Anchorage, Alaska 99504 Dear Mr. gilbreth: FtLE, '~ ~~]. Subject: C~sing Design in the Pe~afrost Interval Zu the ~eZd ~u~es o~ ~auua=~ 9~ ~976, ~u~e 3 o~ Co~ser~at~o~ 0~de~ ~o. 98-A states that t~e sub,ace cas&~8 should ba~e ~u~ ~ost-~2eZd stra~ ~ro~ert~es o~ 0,9~ desc=~bes t~e tests o~ 72 ~b./¢t., ~3-3/8 ~c~ ~-80 ~uttress cas~8 a~d sbo~s ~o~ it adequately meets this standard.__ (~/-'~O A number of our wells are completed with 68 lb./ft., 13-.3/8 inch N-80 Buttress casing as the surface string. These wells are: A-7 (12-26-11-13), C-5 (41-25-11-13), C-6 (41-30-11-14), C-7 (31-24-11-13) and C-8 (33-18-11-14). In order to confirm that this 68 lb./ft, casing meets the requirement, we have carried out studies using the finite element program employed by ARCo and referred to in their report. The input data were the same except for the slightly reduced wall thickness. We did not model the tensile loading condition because we are confident that it will meet the requirement, based on the fact that ARCo's work demonstrated that a reduction in wall thickness from the 72 lb./ft, condition would increase the tensile strain capacity of the casing. For the 72 lb./ft, condition the tensile strain in the casing at failure was found to be 3.4%, compared with the minimum required figure of 0.9% and therefore comfortably meeting the standard. In compression, the analysis demonstrated that the failure strain was 4.2% for the 68 lb./ft, case compared with the requirement of 1.26%, We submit therefore that 13-3/8 inch, 68 lb./ft.¥Buttress casing adequately meets the Field Rules, Very truly yours, BP ALASKA INC. G. D. Taylor Manager Operations DBLW:vjy cc: Drilling Superintendent District Drilling Engineer File - Permafrost: Casing Strain (13-3/8" 68#) 2/5/76 Lonnie C. Smith Memo to File Re: CO # 137 and Drilling Permit Casing Requirements Hoyle Hamilton and Lonnie Smith talked with Dick Knowles of ARCO who confirmed that the casing used in the Casing Strain Tests and submitted as evidence for CO # 137 in the Prudhoe. Ba~ Field Permafrost Casing and Well Design for Thaw Subsidence Protection. May 1975 report was modified N-80 (MN-80) grade, although on various pages of the report the casing was referred to as "Normalized N-80 CYN-80 (controlled yield N-80) and N-80". As explained by Mr. Knowles and on pages 6 & 7 of the Mannesmann Tube Co., LTD. Seamless Casing Catalog. the "modified" N-80 carries the same minimum physical properties as r~gular grade N-80 but is restricted on the maximum yield strength due to, he special tempering process which allows development of a "fully tempered martensite grain structure" with "hardness controlled to Rockwell C 22 maximum". The fact is that this casing has now been approved by API and is designated as Grade L-80 in their publication 5AC, Ninth Edition, March, 1975 entitled Specification for Restricted Yield Strength Casinq and Tubing. These specs. require a maximum Rockwell C23 hardness. Casing design criteria are based on minimum strength properties but Arctic enviroment use also requires a restriction on the maximum hardness, which relates to grain structure and tempering. Casing manufactures sometimes have higher grades of casing, which fail to meet specifications for that grade and is downgraded and sold as a lesser grade for which it meets the specs., i.e. P-110 sold as N-80. This practice is incompatable with Arctic reqgirements which cannot tolerate brittleness of the harder P-110 chemistry and ~mpering. Thus the designation of MN-80 grade casing prevents substitutions by the manufacturers whereas designation of N-80 would not. MINUTES - CO 137 Oil and Gas Conservation Committee January 9, 1976 The Committee and all professional staff members met in the conference room of the Division of Oil and Gas from 10:30 AM until noon and 1:30 PM until 4:30 PM to review the draft of the appearances and the further appearing statements. The Committee considering this order was composed of Hoyle Hamilton, Harry Kugler and Tom Marshall. Selection of a design factor for permafrost casing was a principle. January 12, 1976 Committee reconvened at 9:00 AM to discuss the transcripts and regulations prepared by Lonnie Smith and John Miller. Mr. Gilbreth attended this Monday morning meeting. The entire proposed Rule 3 relating to permafrost casing was reviewed. The soft cover data book entitled, "Proposed Well Completions through the Permafrost Interval", submitted on December 10,1975 by BP Alaska, Inc. was referred to and studied by the staff.' The hearing record had been held open through DeCember 10, 1975, The BP publication was timely received and is made part of the official record. .J..anuary 13., 1976 Committee met from 9:00 AM to noon and from l:'O0 PM to 4:30 PM. Considerable time was spent rel.ating casing requirements'for one pool with that of super- imposed pools. January. 14, 1976 Committee met from 9:00 AM until 11:30 AM. Design factors suitable for casing strings here studied and~related to proposed rules. Marshall read. entire order and signed signature page. He had to go to Sitka on January 15. January 15, 1976 Committee met, reviewed the final draft and signed the order. BP ALASKA INC. 31111 - C - STREET · TELEPHONE {907) 279-0644 MAILING ADDRESS: P.O. BOX 4-1379 ANCHORAGE. ALASKA {)9509 DEV-PB-714/5 December 10, 1975 Mr. Thomas R. Marshall, Jr. Executive Secretary Alaska Oil & Gas Conservation Committee 3001 Porcupine Drive Anchorage, Alaska 99501 Re: Conservation Order File Number 137 There are .enclosed for filing three copies of a report entitled "Casing Damage Review- Prudhoe Bay Field- BP Alaska Inc." by Mr. P. R. Judd, District Drilling Engineer who, as you will recall, testified at the referenced hearing. Please include this report in the record of the referenced hearing in support of Mr. Judd's statements regarding external freeze back pressures. Very truly yours, ~hn A. Reeder Attorney blm Enclosures cc: Mr. P. R. Judd Mr. G. D. Taylor CASING DAMAGE REVIEW PRUDHOE BAY FIELD BP ALASKA INC. by P. R. Judd District Drilling Engineer Anchorage, Alaska December, 1975 Casing DaMage Review Prudhoe Bay Field BP Alaska Inc. BP Alaska Inc. has drilled in excess of sixty wells through the permafrost zone in the Prudhoe Bay area. Thirteen wells are known to have experienced casing damage as a result of collapse due to the freezing of fluids in casing/casing annuli. Table I lists these wells and shows the dates during which each well was drilled and the dates that workover operations were performed. Twelve of the thirteen wells have been re-entered to perform workover operations. The one remaining well, H-3 (23-22-11-13), is to be worked over at a later date. A summary of the casing strings found to be damaged is shown in Table II. This table also shows the depths to which damaged casing was removed during workover operations. Appendix A is a well-by-well history of the workover operations performed. Subsurface diagrams of'each well are given in Appendix B showing the status at the end of drilling, and in Appendix C showing the status after workover operations. One well, N-4 (23-07-11-13), was abandoned after unsuccessful attempts to remove a swaging tool which became stuck in the casing during a workover. Table III shows the type of completion planned for the twelve wells to be put on production. The groups shown correspond to those presented in testimony to the Alaska Oil and Gas Conservation Committee on November 25, 1975. During the workover of each well, damaged inner casing strings were removed using a combination of techniques including swaging, washing-over, cutting, and pulling. Whenever possible, casing having damaged intervals was removed to a depth below the base of the permafrost even though the casing wa~ L~oL necessarily damaged to that depth. In three wells, new casing has been installed with a casing patch to replace the damaged casing removed. New casing is to be patched into the remainder of the wells at a later date. In ten of the wells worked-over, the outer casing'string, which is adjacent to 'the permafrost, was exposed upon removal of the damaged inner string(s). Caliper surveys and/or full gauge tool runs confirmed that this outer casing in eight wells was not damaged. Apparently, the potential for casing collapse due to external freezeback is virtually eliminated in those wells with cement placed inside the outer casing. In those wells with another fluid inside the outer casing, the potential for collapse due to external freezing is apparently reduced if a satisfactory cement job was performed in the casing/permafrost annulus. In two wells, D-1 (23-23-11-13) and D-3 (23-14-11-13), the 20" outer casing was found to be damaged in the weaker 94# H-40 portion of the string, but not in the stronger 133# K-55 portion. In these two wells a cementing problem occurred which may explain the cause of the casing damage. During cementing, returns were lost and a subsequent top job was performed in the casing/permafrost annulus. As a result, mud was probably trapped between the cement jobs. The freezeback of the mud may have resulted in a pressure sufficient to collapse the casing. In our other wells with 20" 94# H-40 or 18-5/8" 96.5# K-55 outer casing without a cement backup inside the casing, no problems were encountered in cementing the casing/permafrost annulus. In conclusion, Casing damage (collapse) may occ~ur thorough the permafrost zone if due consideration is not given to the pressures that build up upon the freezing of fluids in annuli. Freezing of fluids in casing/casing annuli causes very high pressures, and the placement of non-freezing fluids in casing/casing annuli is a necessity to prevent casing damage. External freezeback pressures have been predicted by published data. Casing with sufficien~ collapse ratings and/or high-density non-freezing fluids placed inside the casing can resist external freezeback pressures. The pressures predicted by the published data do not appear to develop if a satisfactory cement job is performed in the casing/permafrost annulus. TABLE I Wells With Damaged Casing WELL B-3 (23-30-11-14) D-1 (23-23-11-13) D-3 '(23-14-11-13) D-5 (23-24-11-13) F-2 (22-11-11-13) F-4 (12-01-11-13) F-5 (14-36-12-13) F-6 (14-35-12-13) * H-3 (23-22-11-13) J-2 (23-10-11-13) · .. M-1 (43-01-11-12) ** N-4 (23-07-11-13) N-6 (34-05-11-13) DRILLING 5125170 - 7~7~70 1/4/70 - 3/18/70 4119170 - 6~8~70 7/29/70 - 9/8/70 10/25/70 - 12/12/70 2/2/71 - 3/20/71 3/21/71 - 4/28/71 5/1/71 - 6/5/71 7/29/71 - 8/19/71 5/3/69 - 5/15/69 9/24/69 - 11/6/69 7/21/69 - 8/30/69 1/24/71 - 2/25/71 3/2.9/71- 5/2/71 WORKOVER 9/9/70 - 11/27/70 11/3/74 - 12/3/74 12/16/74 - 1/15/75 1/16/75 - 2/2/75 6/16/72 - 7/3/72 7/4/72 - 7/17/72 7/17/72 - 7/29/72 7/29/72 - 8/7/72 1018172 - 10/17172 3~29~70- 5/31/70 8/2/71 - 8/28/71 5/15/74- 5/31/74 6/1/7.4 - 6/17/74 * Well H-3 has not yet been re-entered to remove damaged casing. ** Well N-4 was abandoned due to fish stuck in casing. TABLE II Summary of Casing Damage Due to Freezing of Fluids in Annuli Well B-3 D-1 D-3 D-5 F-2 F-4 F-5 Casing Thru Permafrost 10-3/4" 55.5# N-80 13-3/8" 68# K-55 20" 133# K-55 20" 94# H-40 10-3/4" 55.5# N-80 13-3/8" 68# K-55 20" 133# K-55 20" 94# H-40 10-3/4" 55.5# N-80 13-3/8" 68# K-55 20" 133# K-55 20" 94# H-40 10-3/4" 55.5# N-80 13-3/8" 68# K-55 20" 133# K-55 20" 94# H-40 13-3/8" 68# K-55 20" 133# K-55 13-3/8" 68# K-55 18-5/8" 96.5# K-55 13-3/8" 68# K-55 18-5/8" 96.5# K-55 Damaged Yes No Yes Yes No Yes Yes Yes No Yes Yes Yes No No Yes No Yes No Yes No Removed to Depth 995' 1991' 1945' 1993' 1994' 1991' 1975' 2083' 1934' No 1971' Notes (1) (2) (3) (2) (3) (2) Well Casing Thru Pe=mafrost Damagcd '~ F-6 13-3/8" 68# K-55 Yes 2033' 18-5/8" 96,5# K-55 No H-3 13-3/8" 68# K-55 Yes 18-5/8" 96,5# K-55 9-5/8" 47f! N-80 13-3/8" 68# N-80 13-3/8" 61~ N-80 20" 94# N-40 J-2 Removed to Depth Yes 2083' Yes 908' No No M-1 7" 29# N-80 Yes 1922' 9-5/8" 47# N-80 No - 13-3/8" 61# K-55 - - No N-4 13-3/8" 68# K-55 Yes 454' 18-5/8" 96,5# K-55 No N-6 13-3/8" 68f~ K-55 Yes 1976' 18-5/8" 96.5# K-55 Notes: (1) Well having 16" circulation string in 13-3/8" x 20" annulus. (2) Wells with 16" circulation string removed during workover, (3) Casing that has been patched back to replace that removed, (4) Well H-3 has not yet been re-entered to remove damaged casing, (5) Well N-4 was abandoned due to fish stuck in casing, .Notes (4) (3) (5) TABLE III Completion Type for Workover Wells Well Group B-3 D-1 D-3 D-5 F-2 ¥-4 F-5 F-6 H-3 N-6 APPENDIX A ,, .. __~eptemb er ~ 1970 · 8-9" . 10 - 16 .. 19 - 21 21 -. 22 23- 24 - 26' · · WELL HISTORY (Flow Testing) B-3 (23-30-11-1~) Moved rig to location and rigged up. , Attempted to enter well with 8½" bit. Encountered obstructiom at 220'. Six inch drill collar also stopped @ 220', but 4-3/4" drill collar passed through restriction. Ran 2-7/8" tubing. Held up at 575'. Ran 1040' of 1¼" EUE tubing with 2-3/16" collars to 1040'. Through- t~bing caliper log showed restrictions at 220' - 228' 576' 590' 789' - 806'. Swaged out obstruction at 220' with 8½" and 9-5/8" swages. Swaged obstruction @ 576' With 5-3/4", 8½" and then 9-5/8" swages. Swaged obstruction at 786' - 789' with 5-3/4", 8½" and 9-5/8" swages. Circulated hot brine. Ran retrievable packer and tested 10-3/4" (and 13-3/8") casing to 3000 psi. Ran Sperry-Sun gyro, CBL and Dialog free point indicator. Both CBL and FPI indicated 10-3/4" casing to be free above 2000'. Set EZSV bridge plug at 2620'. Tested 10-3/4" (and 13-3/8"), casing to 3000 psi. Circulated brine around 10-3/4" casing. Drilled'bridge plug. Ran in to baffle. Ran Gamma ray and cement bond logs. Perforated 10,016' - 10,026'. Ran in hole with test tools for DST #1. Set packer @ 9999.'. Opened tool and ~ecovered light oil. Pulled tools. Ran FIT test at 9920' and 9930'. Perforated 9940' - 9960". Set packer at 9960'. Opened tool and had communication between 9960' and 10,016'. Unseated packer, set EZSV squeeze packer at 9960'. Ran drill pipe and latched in. Established circulation behind cas- trig. Pumped 20-bbls. mud flush, 5 bbls. water; 65 sacks "G" with 1.3% FLAC mixed at 15.4 - 15.9 lbs/gal, followed by 10 bbls. water and 65 bbls. brine. Waited on cement. Tagged.cement at 9894' and · cleaned out to 9900'. · Set bridge plug at 9885'. Perforated 9685' - 9695'. Ran tools for DST #3. Set packer at 9594' after unplugging and resetting tools. Flowed to check for communication to gas cap. Flowed through 4½"' .and 2-7/8" burn lines. Shut in. Conditioned brine. Pulled DST tools · --~'~ B-3 (23-30-11-14) Sept e.mb.er ~ 1970 26- 30 October, 1970 1-5 5 - 18 23 - 24 .. ~November ~ 1970 -. 4 --' 7 December 18, 1970 Ran 'in open ended. Laid down pipe. Set 9-5/8" Baker production packer at 2474'. Ran 7" casing with seal assembly to 10-3/4" DV. Held up. Pulled casing. Drilled 10-3/4" DV with 9-5/8" bit. .Cleaned out through packer with 4-1/8" bit. Ran 7" casing, landed in packer at 2481'. Tested 7" pack-off and 10-3/4" - 7" annulus to 2000-psi. OK. Set 7" RTTS at 180'. Tested wellhead, casing, and drill pipe to 2000 psi. Started running 2-7/8" tubing. Ran 4~"/2-7/8" combination tubing string to 9186'. Tested Otis ball vaive and control lines satisfactorily. Tested 4½"/7" annulus to 3000 psi satisfactorily. Nippled up Xmas tree and pressure tested satisfactorily. Perforated 9695'-9715' 9730'-9'760' Flowed well Tools developed leak into 7" x 10-3/4" annulus. Removed tubing and 7" casing with some difficulty arising from broken control lines and unidentified junk behind production tools. Milled and retrieved Baker packer. Found leak'in 9-5/8" casing between 6645' and 6650' (probably in stage cement t°ol at 6649'). Set bridge plug at 6665'. Squeezed approxima[eiy 20 sacks "G:' through leak. Tested to 2000 psi satisfacto~ily. Drilled out temporary cement plugs. Conditioned hole to 9885'. Prepared to run ~ubing string. set 9-5/8" Baker retainer production packer at 9381'. Located 4~"2 tubing in packer. Nippled up Xmas tree. Tested tree to 5000 psi, 4½" tubing to 3000 psi. Satisfactory test. Opened side door sleeve and displaced tubing. Well would not kick off. Replaced Xmas tree due to leaking master valve. Reperforated 9685'-9715'. Flowed well on muir-rate test. .. Flowed well on multi-rate test. Ran bottom hble pressure buildup survey. Recovered bottom hole samples. Unable to reach perforating depth-due to fill, killed well. Installed BOPs and pulled production string. Added extra 2-7/8 tail pipe and ran to 9885'. Conditioned brine and cleaned hole. Pulled out and removed tail pipe. Reran 4½" production string with short tail pipe. Landed tubing. Checked hole with sinker bars to 9870'. Installed Xmas tree. B-3 (23-30-11-14) Novemb er ~ 19 70 8- 13 14- 17 18- 24 2.5 - 27 December 18, 1970 Washed in well and perforated 9685-9715 and 9730-9760 at 4 shots/ft. FloWed well to clean up and perforated 9775-9790, 9790-9810, 9795-9800. Flowed w'ell on multi-rate flow test. Ran pressure buildup survey. Lost gauges due to surge. Gauges fell into 9-5/8" casing. Made brief attempt to recover without success. Took bottom hole samples. Re- covered pressure gauges from casing. Killed well through sleeve with wireline plug in tubing tail-pipe. .. Installed BOPs. Otis lost Plug into casing while retrieving from tail-pipe.- T~ied to fish but unable to free fish or fishing string. Cut line. Pulled tubing. Set'DR plug with latch in Baker packer at 9381'. Sheared pins and attempted to spot cement plug through tdbing. cemented tubing due to damaged shear sub on bottom. POH. Ran open ended tubing and set 100 sacks` Class "G" from 9377' to 9157'. .. Ran Dialog FFI in 10-3/4" casing. Casing believed free to '2100'. Circulated heated brine. Ran mechanical cutter and attempted to cut at 2080' and 2085'. Unable to pull casing after each cut. Ran hydraulic cutter and cut at 2089' and 2085'. Unable to pull casing. Cut casing at 2020' and extended with Servco section, mill. Still unable to pull. Repeated Dialog FPI, casing free to 1090'. Cut 10-3/4" casing at 995' and recovered upper part of string. Jarred on fish with spear; no movement. Dressed casing stub with mill and s~t bit guide on i0-3/4" s Lub. Set EZSV packer at 2590' on drill pipe. Tested 10-3/4" and 13-3/8" to 1850 psi over 10.2 lbs/gal, brine column satisfactorily. Ran 12¼" bit and 13-3/8" casing scraper to 990'. Ran 13-3/8" RTTS and tested. 13-3/8" casing above packer to 2700 psi satisfactorily. Ran caliper log to 995' indicating no deformation of 13-3/8" casing. Displaced hole to diesel at 2588' installed wellhead and released rig at 0400 hours on 11/27/70. . · . , WELL HISTORY D-! BP 23-23-%1-13 The following is a summary of the work done on Well D-1. Filed in the well file are Tri-State's detailed daily reports on the swaging and fishing operations. Attached is a drawing showing casing damage. The drifts run in the 10-3/4" Easing were run-prior to the workover. Damage to the 13-3/8" and 20" shown on the drawing was taken from caliper logs. Also attached are downhole profiles . of Well D-1 before and aiter the workover. Ail measurements on the drawings and in this report are BKB unless noted otherwise. The patching procedure has also ~een enclosed for reference. Workdver Operations Nov. 4 to Nov. 7 1974 The rig was accepted on November 4, 1974, at 0600 hours. A 6" and 7-1/'2".swage was run in tandem, followed by 6", 7-1/2", and 9-1/2" swages in tandem. During the swaging, CaC12 was circulated to melt any ice and aid the swaging operation. After the 10-3/4" was swaged out and a cement plug was set from 2937' to 3008', a CC1 was run on the 10-3/4". A cut was made at 2000' in the 10-3/4". The casing was speared and 1200' was recovered. Two joints were split and the depths of the splits were 617' to 620' and ~755' to 781'. The 10-3/4" parted at a collar. The remainder of the 10-3/4" was speared and recovered to 2000' after the jars hit once. A CCI/caliper log was run in the 13-3/8" and the damage is in the attached drawings. Nov, 4 to Nov.'!4 1974 'On November 8~ the ].3-3/8" x ]6" x 20" ann,~,.,s w~s circ,~lated. The diesel recovered was badly cut with water and cement. A 12-1/4" swage was run to approximately 2000' after swaging at the intervals indicated by the caliper log. An RTTS packer was run and set at 1955'. The 5" drill pipe x 13-3/8" annulus would not hold pressure and returns came up the 13-3/8" x 16" annulus. The drill pipe was pressured up to 1000 psi for 10 minutes and the 10-3/8" x 13-3/8" lap held. A cut was made at 1960' in the 13-3/8", but after spearing the ca§lng at surface, it could not be pulled. A cut was made at 45', but could not be pulled. A cut was made at 41.5' and the stub was retrieved. The 13-3/8" spear.could not pass the 16" due to collapsed 16" at surface. A 16" spear was driven into the 16" and the 16" parted at 110' when the 16" was pulled. Several runs with 13-3/8" cutters and spears were made and the 13-3/8" was recovered to 862.5'. On a run with the 13-3/8" spear, it stopped at 843'. The 16'! had collapsed between pulling out of the hole with last fish and running in the hole with the 13-3/8" spear on the foll'owing fun. . A 15-1/8" swage was run and the 16" top was tagged at 201'. Light swaging was encountered from 250' to 255', 757' to 758', 841' to 855' and the swaging was very firm from 855' to 863'. The 13-3/8" was recovered to 1024' after some jarring, and the next run was with the 13-3/8" spear. It stopped at 872'. The 15-1/8" swage was run and the 16" was swaged out from 841' to 938'. The. 13-3/8" was speared and recovered to 1960'. A 13-3/8" swage was run inside the 16" and swaged at 259' to 288', 757' to 758', 841' to 863', 1316' to 1327'~ and 1337' to 1451'. The 16" was speared but could not be pulled. The 16" was cut at 1939' and 965' but the attempts to recover , . the fish produced nothing. The 1'6" was recovered to 859' by cutting and jarring. Nov. 15 to Noel' 18 On November 15, an 1~-1/2" bit was run and it stopped at 736'. 1974 The top of the 16" stub was at 859' Schlumberger ran the first of three caliper/CC1 log~ in the 20"'an'16''. A summary of the bad places in the 20" is shown in the attached drawings. The 16" caliper showed bad casings over the following intervals: 854' to 898', 910' to 920', 1120' to 1140', and 1282' to 1450'. 'The 20" was swaged out and the cutting and pulling of the 16" was resumed. The 20" had to be swaged out several times to let the 16" spear pass. A 16" fish was cut and pulled, leaving the top of the 16" at 1568', and there was a lot of drag while pulling the fish at 1300' to 1400'. · The 20" was suspect again. Nov. 19 to Nov. 24 On No~ember 19, a 16" spear was run to try and spear the 16" 1974 at 1568' but it stopped at 1446' in the 20". The second caliper log was run in the 20" and showed bad 20" from 1330' to 1448'. The cutting and spearing of the 16" was resumed and the 16" was recovered to 1700'.' After a cut was made at 1750', the spear was again run but it would not set properly. Inspection of the spear-showed that it was~jammed with cement and sand. This was the first indication of some cement inside the 20" and above the DV tool in the 13-3/8". Several runs with a bit ~ inside the 20" were necessary to clean the 20" and thus aid in fishing the 16". Nov. 25, 1974 On November 25, after a cut at 1952' on the 16", the 16" was'recovered. The 10-3/4" stub is at 2000', the 13-3/8" stub is at 1954' and the 16"-stub was at 1952'. A special blade mill, 15" x 17-1/2" on the O.D., and 13-1/2" I.D. was run to mill the 16" down below the 13-3/8" stub. The 16" stub was milled to 1972', after reaming cement from 1937' tor1952'. This operation exposed 18' of 13-3/8" stub for patching. Nov. 26 Go Dec. 3 1974 Patching Operation On November 26, the third caliper was run in the 20" and the 20" is damaged over the following intervals: 288' to 298~, 602' to 620', 830' to 910', and 1330' to 1448'. The log also shows some questionable 20" from 642' to 830' and from 1650' to 1720'. 'A washover beveling shoe was run to dress the ].3-3/8" stub for patching. The patching procedure is.attached and it is corrected to coincide with what was actually done; however, before the well was suspended, a free point was run by Dia-Log to determine if the 9-5/8" casing was free below the crossover. After considering the results of the free point, the decision was not to try to cut below and retrieve the 10-3/4"/9-5/8" crossover, and the well was suspended as per the patching program. This was the first time that this casing patch has been run and the problems encountered were minimal. The patch swallowed the stub approximately four feet and hung up momentarily. After rotating and picking up and setting down, it swallowed the stub, was set and tested without any further problems. A restriction in the area of the patch stopped a 12-1/8" mill at 1955' when the work on the 10-3/4" stub was being done. A check with Baash-Ross showed the I.D.'s of the patch to be sufficient. The 10-3/4" was cleaned up and the planned free point, etc., was accomplished, but to insure a full I.D. through the 13-3/8" stub in the future, it might be wise to run a tapered mill inside the 13-3/8" stub to clean the I.D. up and knock any burrs off before patching. Arctic-Pack Opera~ions The Arctic P~ck-mixing and pumping went verY weli. considering this was also a first for us. This was developed by ARCo and Baroid for a dual purpose (1) put a non-freezing fluid'between casing strings through the permafrost and (2) provide an insulator between the well bore and the permafrost. The fluid was heavy enough to displace the water in the 13-3/8" x 20" annulus and when the pre-flush came back there was a definite interface. From the samples of the returns that were taken, the water content was about 5%, which is the initial concentration. The 50 bbls. of pre-flush were put in the reserve pit and the rest of the returns were caught in a tank. The Arctic Pack was pumped at 2 bbl./min, and circulation was lost during the last 25 bbls. pumped with a surface pressure of 550 psi. Due.to the loss of returns,'Halliburton was shut down and 50 bbls. of the pre-mix were left in the residual tanks'. The rig pumps were uSed to displace the drill pipe. The 20" had some holes in it and'the loss of circulation came as no real surprise. After the analysis of the returns and confirming that the 13-3/8" x 20" annulus was full of Arctic Pack, the job was definitely a success Summary Discussion From an examination of the caliper logs, the 20" definitely collapsed due to freeze- ~back of the permafrost. The problems in the 13-3/8" and the 10-3/4" can be attributed to freeze-back within casing/casing annuli as noted by the close correlations from swaging and caliper data. From talking with the rig supervisors, the "diesel" in the 20" x 16" x 13~3/8" annuli was mud, cement and water cut too badly to even save. The 13-3/8" DV tool at 2254' was put in the string to wash excess cement out of the 13-3/8" x 16" x 20" annuli. The workover reports (Tri-State) state that cement jammed a spear as high as 1700' and that hard cement was reamed from 1782' until the 16~ was finally removed. The attempt to get the 10-3/4"/9-5/8" crossover out' was for the purpose of patching another pressure string back to surface. The cement job in the 9-5/8" x 13-3/8" ~nnulus is suspect and has only been tested to 1000psi. There seems to be several alternatives for an ultimate completion, but none are very attractive. (1) The c~o~ov~r co,~]Id be ~!led up and the 9-5/8" patched back to 2000' + w~th a !2' seal receptacle. Cement the 9-5/8" patch string for stability and have Thermo- Case II stung in and hung at surface~ The production string could be 7" tubing. (2) The 9-5/8" x 13-3/8" lap could be tested to 3500 psi and squeezed if it breaks down or just left as it is if it holds. The 10-3/4" x 7" Thermo-Case tubing could then be run. (3) Another possibility is to mill up the 10-3/4"/9-5/8" crossover and patch a 9-5/8" x 10-3/4" string back to surface. The completion would be ~8-5/8" x 5-1/2" Thermo-Case. Regardless of the ultimate completion, the 13-3/8" casing should be monitored to determine if damage occurs from freeze-back of the permafrost. Consideration should also be given to running a string of tubing to circulate a warm fluid followed by allowing the permafrost to re-freeze to see if damage to the 13-3/8" casing occurs.~ WORKOVER HISTORY BP WELL D-3 (23-14-11-13) The following is a summary of the highlights of the D-3 Workover. Drawings of Well D-3 before and after the workover are attached for your reference. The rig w~s accepted at 0800 hours on December 16, 1974. After the 10~3/4'' st. ring'waso.swaged out to 9-1/2", a caliper log was run. It showed bad pipe at 422' and from 672' to 690'' A cement plug was set at 3010' and the 10-3/4" and 13-3/8" strings were cut at 2000', 720' (just below a tight spot), 650' (just above a tight spot), and at 45' below ground level. These cups were made from the bottom up and by doing them continuously it saves changing bottom hole assemblies. The cutters had to be re-dressed but tool handling was cut down considerably. By making the cuts from the bottom up there was never any weight on the cutter knives to tear them up or hinder in the 'cutting job. The 16" had to be swaged out several times to get tool§ down to the 10-3/4"/13-3/8" stubs. The cement job in the 10-3/4"/13-3/8" annulus proved to be very poor as the 10-3/4" casing was speared and pulled out of the 13-3/8" casing on two occasions. Between 400' and 600' of diesel was found in the 20" x 16" x 13-3/8" annulus, the remainder being cement contaminated mud. The 16" string was recovered and a caliper' in the 20" showed damage from 396' to 470', 1152' to 1200' and 1222' to 1268'. The 18-1/4" swaging runs indicated collapsed pipe at 388', from 1178' to 1190' , and from 1231' to 1247'.~ It is interesting to note that the collapsed pipe is only in the 94# H-40 pipe and that the 133# K-55 pipe had no damage. There was quite a bit of gravel entry during the latter stages of the workover, probably from a hole in the 20" from 399' to 470~. The 50 sacks cement plug that was set at 3010' prior to cutting the 10-3/4" and 13-3/8" strings was never found. Apparently the plug strung out before it set up and there is a possiblity of junk further downhole on the cement plug at 10,313'. The final completion is aa shown in the drawing and the Arctic Pack was put behind the 13-3/8" patch string with no problem. The pre-mix was allowed to cool for several hours before the extra geltone was added. The returns were much thinner than on the D-1 patching job and our circulating pressure was about 250 psi as opposed to 550 psi on D-1. Circulation was maintained throughout the Arctic Pack job. The rig was released at 0200 hours on January 15, 1975. WORKOVER HISTORY BP WELL D-5 (23-24-11-13R) The following is a summary of ~the highlights of the D-5 Workover. Drawings of the well before and after the workover are attached for your reference. The rig was accepted at 0500 hours on January 16, 1975, for the D-5 WorkoVer. The 10-3/4" was swaged out to 9-1/2" with bad pipe from 597' to 607' and from 636' to 642'. A 50-sack cement plug was set on top of a Howco bridge plug at 2707'. The 10-3/4" casing was cut and pulled but very little diesel was found in the 13-3/8"/ 10-3/4" annulus. The 13-3/8" Caliper log and 12-1/4" swaging showed bad pipe in various places over the interval from 410' to 702', 1479' to 1486', and 1660' to 1670'. The lg" string did not interfere with the pull'ing of the 13-3/8" string. Cement was found on the 13-3/8" casing from 740' to 1450', but this cement plug in the 13-3/8" x 16" annulus was'anticipated from past well records. The 16" caliper log showed slight damage at 600'. The 16" casing had three thermometers strapped'on it at 2028', 1294' and 412' with wires to surface, and attemPts to cut and pull the 16" string were hampered by the wire jamming and sticking the 16". A wireline spear was made and run several times to retrieve the wire. The 20" did not interfere with the pulling of the 16" but a cement sheath on the O.D. of the 16" made it necessary to raise and hang the BOP stack so the cement could be chipped and/or washed off the 16" casing during pulling operations. After the 16" and the electrical wires were retrieved, a caliper log was'run in the 20" string. The 20" caliper log showed slight damage from 600' to 1800', and the well was left without patching a string of 13-3/8" back to surface. The rig was released at 2400 hours on February 2, 1975. ion 2A {6 - 20 ~ORKOVER H ~ STORY Re-entered well at 2359 hours, June 16, 1972. Tested Hydril to 2000 psi. Ran in with .2-7/8" tubing to 2501'. Displaced diesel fro= well with CaC12 brine' at 10.4 ppg. Spotted cement plug with 100 sacks Permafrost II cement with 15 lb./sack Pozzolan. Slurry weight a: 14.8 ppg. DamagEd crown while pulling out of hole..Made repairs to crown block, water' table and mast structure. :l - 30 Ran 8-1/2" and 9-1/2" swages in combination, swaged 13-3/8" casing from 313' to 1123', ran in hole cleanly to 1249'. RT to change swages. Ran 10-1/2" and 12-1/8" swages in combination, swaged casing from 312' to 1700', ran in cleanly to 2177' and tagged cement plug. Circulated hot brine to thaw annular fluids. Ran 11" bit cleanly to 2177', dressed plug to 2197'. Cut 13-3/8" casing at 47' and retrieved casing hanger and 29 feet of casing. Ran Dialog free point indicator which indicated casing relatively free to 500' Cut casing at 500' and then 157' but could not mov~ casing. . -.Washed over to 240' and cut casing at 238'. Jarred casing loose and zecovered. Drilled 6-3/4" diameter junk sump in cement plug. Contim. ue-.' washing over and cutting/pulling casing to 1770'. .July, 1972 {-3 Continued washing over and recovering casing to 2157' ( washover depth) --and 2083'(final cut). Ran 11-1/2" junk basket, took core from cemen= ])lug and recovered some junk. Ran 18-1/2" bit, then 18-1/2" casing ~craper to clean 20' casing. Ran 12" bit to 13-3/8" casing stub at 2083', unable to enter casing so displaced hole to diesel at that depth. _'Removed BOP,replaced wellhead and tested. Rig released at 1630 hrs. July 3, 1972. _ ! ~ec~on 2A ~ ,. J~ly, 19 72 1" 17 WORKOVER HISTORY Commenced operations at 2359 hours, July· 3, 1972. Tested Hydril to 500 psi. Ran Schlumberger CCL. Ran 4 arm Caliper, but held up at 351'. The 2-7/8" OD tubing also held up at 350'. Circulated to thaw casing at 350'. Picked up 8-1/2" swage and swaged from 239' to 827'. Ran in to 2000' and circulated out diesel with water. Ran in with 2-7/8" tubing, spotted cement plug from 2160' to 1942' with 100 sacks Permafrost II with 15 lb./sack Pozzolan. Ran in with 8-1/2" and 9-1/2" swages in tandem and swaged from 236' to 860' and ran in to 1500' cleanly. Ran 4 arm Caliper survey from 1940'. Ran 10-1/2" and 12-1/8" swages in tandem and swaged from 234' to 865'. RT with 11" bit to dress cement plug to 1955'. RT with 6-3/4" bit to drill 20' of rathole from 1955' · to 1975'. Cut 13-3/8" casing at 45' BRT and 9etrieved casing hanger and 28 ft. of 13-3/8" casing. Tested Hydril to 2000 psi. Washed over to 302' and cut and recovered casing at 281'. Continued washing over and retrieving · ~asing to 1950' (washover depth) and' 1934' (final cut). Ran 11" Globe junk basket, took core from cement plug and recovered some Junk. Ran 17-1/2" bit and 17-1/2" gauge ring to clean 18-5/8" casing. Ran 12-1/4" bit [one cone removed), entered fish and cleaned out cement to 2018'. Displaced hole to diesel. Removed BOP's and replaced and tested wellhead. Rig was released at 1015 hours, July 17, 1972. · . Workover History Well Number F-5 (14-36-12-13) · · 17 - 29 Commenced operations at 1530 hours, July 17,1972. Ran CCL to 2010' but 4 arm caliper log held up at 189'. Ran 2-7/8" tubing to 190'. Ran 6-3/4" and 9-1/2" swages in tandem, swaged at 191' and 847' then ran in cleanly to 2000'. Displaced diesel with 9.0 lb/gal. Ca~l2. Ran 4 arm caliper from 2000'. Ran 2-7/8" tubing and spotted cement plug with 100 sacks Permafrost II and 15 lb/sack Pazzolan from 2400' to 2226'. Ran 10-1/2" swage and swaged at 190' and 742'. RT for 10-1/2" an'd -12-1/8" swaged in tandem. Swaged at 190' and 747', ran in cleanly to 2226' and tagged cement plug. RT for 6-3/4" bit. Drillmd 51' of xat hole from 2226' to 2277'. . ~_Cut 13k3/8'' casing at 47' BRT and retrieved casing hanger and 27 ft. of casing. Continued washing over and recovering casing to 1991' o(washover depth). Stripped threads from 16" washpipe connector and left both connector and rotary shoe in hole. Ran inside cutter and cut '-:casing at 1971'. Recovered casing. Ran millhead and 11" Globe junk ..basket, cored cement plug and recovered some junk. Ran 17-1/2" bit and 17-1/2" gauge ring to clean 18-5/8" casing. Ran singlm joint Df ~ash pipe, tagged casing stub at 1951' and fish at 1983'. Displaced -hole to diesel at 1983'. Removed BOP's, replaced wellhead, and tested. · -iP~g released at O&00 hours: July 29, 1972. WORKOVER tIISTORY F-6 (14-35-12-13) July, 1972 Commenced operations at 1130 hours, July 29,1972. Ran CCL to 2300' but 4 arm caliper held up at 760'. Ran 2-7/8" tubing to 2550' and displaced d{esel to 9.0 lb/gal CaC12. Spotted cement plug with 100 sacks Permafrost II blended with 15 lb/sack Pozzolan from 2350' to 2179'. Ran 8-1/2" and 9-1/2" swages in tandem, swaged 13-3/8" casing from 284' to 957'. Ran in hole cleanly to 2179' (top of cement plug). RT to change swages. Ran 10-1/2" and 12-1/4" swages in tandem, swaged .casing from 284' to 957', then ran in hole cleanly to 2179'. Drilled '35' of 6-3/4" rat hole from 2179' to 2214' and dressed.top of plug to 2199'. Cut 13-3/8" casing at 47' BRT and recovered casing hanger and 28' of -~asing. Tested Hydril to 1000 psi. Washed ovAr to 273' and cut and . *recovered casing at 254'. · ¢ ~' August, 1972 . o · .. 'Continued washing over and recovering casing to 2053'(washover depth) --and 2033' (final cut). Ran 11" Globe junk basket and took core from _.cement plug and recovered some junk. Ran 17-1/2" bit with 17--1/2" gauge ring to clean 18-5/8" casing. Ran 1 joint wash pipe with 5" drill pipe stinger inside to 2053' and displaced hole to diesel. ",7'~.-..-Removed BOP's, replaced wellhead, and tested. , -'-Pig released at 1500 hours, August 7, 1972. WORKOVER. HISTORY (23-10-11-13) October, 1972 8 - 12 Commenced operations at 0500 hours, October 8,1972. Tested 20" Hydril to 1000 psi. Ran in with 5-1/2", 5-3/4", and 7-1/2" swages to'210. '. Swaged 9-5/8" casing from 210' to 775'. Ran in free to 2146'. Displaced diesel with CaC12. Ran 8-1/2" swage assembly and swaged from 234' to 693', and ran in free to 2146'. Fished out chain left from previous operations. Spotted cement plug with 50 sacks Permafrost II with 15 lb./sack Pozzolan. Slurry weight 14.6 ppg. Removed 20" Hydril and cut 9-5/8" casing at 31'. Ran Dialog free point from 200' to 1990'. Ran mechanical cutter., .jars to cut 9-5/8" casing at 1903', without success. Ran multiple string cutter and cut 9-5/8" casing at 1905'. Pulled and recovered casing. 13- 17 Ran 4-arm caliper in 13-3/8" casing. Ran Dialog free point indicator and CCL. Welded on 20" bradenhead and nippled up 20" Hydril. Ran 12-1/4" x 10-1/2" swaging assembly and swaged from 557' to 561'; felt slight drag at 681', and ran in free to 1079'. Ran 12-1/4" bit to top of 9-5/8" casing stub. Ran 8-1/2" bit and dressed cement plug to 2100'. Ran 6-3/4" bit and drilled 20" rathole in cement plug. Cut, pulled, and recovered 9-5/8" casing to 2050'. Cut, pulled, and recovered 13-3/8" casing to 908'. Ran 18-1/2" bit to top of 13-3/8" casing stub at 908'. Washed over 9-5/8" casing to 2083' with 11-3/4" washpipe. Circulated and displaced hole with diesel. Removed BOP, replaced well cap and tested to 1500 psi. Rig released at 1400 hours, October 17, 197.2.. l~'r'm No. P--4 P,..EV. 1.30-61 STATE OF ALASKA OIL AND GAS CONSERVATION COMMITTEE MONTHLY REPORT OF DRILLING AND WORKOVER OPERATIONS W~:LL w£LL OTHER NA.ME OF OPEP,.ATOR BI' Alaska Inc. ~. ADI~RE$S OF OPEP,.ATOR P. O. Box 4-CCC, AncJ~ora~e, Alaska 99503 4. LOCATION OF V,'ELL At surface: 458t WEL,' 1998' NSL, Sec. 1, TllM, R12E, UPM SUBMIT I~ DUPLICATE APl N D~.IERIC,kL CODE AP][ 50-029-20042 LEASE DESIGNATION AND SEK1AL NO. ADL-2 8260 r ~ ' IF INDIA]q. ALOTI]:,E O~ TRIBE NANIE B. L.~IT,F.~RA! OR LEASE N.%3AE Put River g. WELL NO BI' 01-11-12 10. Fiat.rs A_ND POOL. OR WILDCAT 'rudhoe Bay-Sadlerocl~it Pool 11. SEC.. T.. R.. Al.. (BOT'FO,M HOL~ O~JECrrVE) Il. PERMIT NO. 69-78 13. I~POt~T TOTAL DEPTH AT E~N'D OF MONTH, CH-~--N'GES IN HOLE SIZE, CASING A.ND CL_xlE.'CTING JOES INCLUDING DEPTH SET A...%"D VOL%~,I'ES USED. PERFOR~-TIONS. /-ESTS .~%~ ~SCL~. FSHLN-G JO~. JL~'K L~ HO~ ~ND SIDE-I~CKEDHO~ ~%~ ~Y O~K SIGNIFI~T ~NG~ ~ HO~ ~ITIONS- ~rch, 1970 CONFIDENTI~ 18- 29 ~ve-and rig ~. 29 - 31 ~t~le up ad ~f up for . . 1~. I hereby ccrt//y that the foregoing t u ci .~/rrect NOTE--Report on th~s form ~'?~quired for each calendar mont~, regardless of the status of operations, and must ~ filed in duplicatt with' ~e Division of Mines & Minerals by the 15th of the succeeding month, unless othe~ise dire~ed. ~ IN, tm Mo. P---4 REV, ~-a0-~ STATE OF ALASKA OIL AND GAS CONSERVATION CO/~AIT'I'EE MONTHLY REPORT OF DRILLING AND WORKOVER OPERATIONS OIL [~ OTHER ,~,, ,. ~ ~*" Appraisal W£LL ~UBMIT l~ DUPLICATE ~. NAMt. OF OPER.~TOR BP Alaska Inc. OF O PEF. ATOR Box 4-CCC, Anchorage, Alaska 99503 4. LOCA I'ION OF V~'F.,[.~ At Surface 458' WEt, 1998' ESL, Sec. 1, TllN, R12E, UPM APl NL~, XEHICAL CODE APr 50-029-20042 6. I,EASE DESiGNA%'ION AND SEEIAL NO. APL 28260 IF INDIAN', ALLOT ~E.E O~ TRIBE NAME L.~iT,F.~! OR LEASE NAME Put River g. WELL .~0 BP 01-11-12 10. FiEl.r) A_ND POOL. OR WILDCAT Prudhoe Bay - Sadlerochit Pool 11. SEC., T.. R.. .%1.. (DO'F]'OM HOLE o~ Sec. 1, TllN, P~12E, UPM 12. P]~IIT 69-78 13. REI'tH{T TOTAL DEPTH AT F--'qD OF MONTH, CHA-\'GES IN HOLE SIZE. CASING ~ C~i~%'TING JOBS INCLUDING DEP~ SET AN~ VOLL%~ USD, p~O~TIONS. ~TS ~%~ F~SUL~. F~NG JO~. J~K LN HO~ ~ND SIDE-~CKED HO~ ~[m ~y O~ SIGNIFIC~T ~N'G~ ~ HO~ ~N~IONS. I - 3 Completed rigging up. 3- 24 · .. The 7" casing was partially collapsed at the following intervals 110', 236', 371', 415', 500', and 631'. The damaged sections were ~!lled out and the ca~ing was inspected with iog~. 7he casing was removed down to 1922'. During the milling operations the 9-5/8" casing was damaged, at 110'. The 9-5/8" casing was cemented to the surface with 360 sacks of c~-~ent. Cleaned out and 7" casing run with a casing patch. After matching to the cut off casing at 1922' the casing wa~ c~ented to the surface with 350 sacks of cement. '24 - 30 Drilled bridge plug at 2500' and cleaned out to 3270'. hc,'~by certify tl~at the foregoing is true and correct mc.x~:n . . .~Operations MangerDA~~y 14, 1970 NOTE--Report on this form is required for each calendar month, regardless of the status ot operations, anO must ~ file0 in' duphcatt with the Division of Mines & Minerals by the 15th of the succeeding month, unless othe~ise directed. OIL AND GAS CONS£RVA~ION COMMITTEE fdONTHLY REPORT OF DRILLING AND WORKOVI!R OPERATIONS BP ALASKA INC. - -- p. At surface: 458' WEL, 1998' NSL Sec 1, TllN, R12E, UPM 50-029-20042 ADL 28260 ALOTT~iE O!~ 'IRIBE NAME UNIT,FAF~I 01~ L~ASE NPJ~-E Put River WELL NO BP 01' LDC/~q' Prudhoe Bay_--$adleroch±C 0il Pool T,. R.. M.. {~O~'OM }iO~ o~Iv~ Sec 1, TllN, ~2E, ~M 12. pEiP~%IIT NO. DEPTH REPORT TO'rAL DEPTH AT E2~D OF .MONTH. CHA_N'G¥~q IN HOLE SIZE. CASING A~N'D CIq-X!ENTING JOBS INCLUDING SET A.?,,-D VOLU~.IES usED. pEI::LFOP,_~'I-IONS. TESTS A.N"D lrLESULTS. FISHING J°B'`c;. JI"'rNK LN' HOLE A. ND SIDE.'EP,-ACKED HOLE A..N-D A2WY oTIEEH. SIGNIFiC~'"~T CI-LA_\'GES IN l-IOl..l~ CON'DITIONS- }__fay 1970_ CONFID~TIAL · 1 - 31 Perforated. and flcw tested following intervals: 9055-9060" 9000-9008' 8824-8884' (RTKB) · ' Released rig @ 2200 hours on May 31, 1970. '-~ . , . .-,/. -~,,- . DA'rE .1~ v ;'- / /:'"/% -- TITL~ Ii .' / k- Z.~,.f : SIGN'ED.------~~-r.' ' -- of operations, anO most be Iiled in duplicate NOT[--Rc-port on this form is required for each calen0ar month r ec~atClcss o! the status with the Division of Mines & Minerals by the 15th of the succeeding month, unless otherwise directed. Well No. BP N-4 (23-'07-11-13) Wate~ Test History _Au~us t, 1971 -- 2-5 10 - 13 14 - 20- 20 - 26 27:' 28 Completed rigging up at 1415 hours, August 2,1971. Nippled up and tested BOP. Displaced diesel with brine. RIH with 12-1/4" bit to bridge at 541'. Ran caliper log. Repaired casing. .Set.packer at 1219'. Established circulation behind 13-3/8" casing to surface. Pressure tested 18-5/8" annulus to 2250 psi. Recemented through drill pipe with 540 sacks Permafrost II, 40 barrels return. Squeezed down 18-5/8" x 13-3/8" annulus with 440 sacks Permafrost II.- Cleaned out liner to 10,350'. Repaired casing. Perforated 7" liner at 10,301' - 10,281' and 10,273' - 10,253'. Conducted water test. Killed well. Set 7'" bridge plug at 10,235'. TwiSted off in tubing connection while milling. Fish in hole to 9-5/8" x 13-3/8" cross-over. Ran fishing tools, jarred on ~fish. Following was left in the hole: one 8" drill collar, x-over sub, one tapered mill, one x-over sub, two 6-1/4" drill collars ,. o.ne bit s,_~b, ~nd one 8-1/2" bit. Top of fish at 2242' and bottom at 2313'. Displaced 13-3/8" casing with diesel at 2202'. Laid down drill pipe. Set Baker retrievable bridge plug at 1225'. Laid down 4-1/2" tubing. Removed BOP. Rig released at 1630 hours, August 28, 1971. Well No. BP N-4 (23-07-11-13) WORKOVER HISTORY '_May, 19 74 15 - 31 Commenced operations at 1200 hours, May 15, 1974. Tested BOP. .. Retrieved Model "C" bridge plug at 1225'.. Diesel displaced with CaC12 base mud down to 2240'. The collapeed portions of the 13-3/8" casing were swaged out with 12-1/4" swage down to 2240' An attempt was made to remove the fish wedged in the 13-3/8" x 9-5~8" casing crossover with partial recovery. The top of the fish is now at 2272' and consists of one crossover sub, one 12-1/4" tapered mill, one c'rossover sub, two 6" drill collars, one bit sub, one 8-1/2" bit. Spotted 50 sacks Canadien Blend Permaforst II cement plug 2100' -' 2045'. The 13-3/8" casing was washed over down to 469' and removed down to 454'. Spotted 180 sacks Canadien Blend Permafrost II cement plug from 1200' - 1000', displaced the top 500' to diesel, and re- installed the wellhead equipment. Rig was released at 1400 hours~ 'May 31., 1974. The well was temporarily suspended to re-evaluate workover techniques. ---: AUgUst; 1971 Removed wellhead, ran 4 joints of 4-1/2" casing to 112' belew . base flange elevation with 5' above ground level marked as required by State regulations. Cemented to surface with 200 sacks Permafrost cement on August 3, 1974. WELL ABANDONED. Well No. BP N-6 (34-05-11-13) WORKOVER HISTORY June, 1974 1 - 17 Commenced operations at 0800 hours, June 1, 1974. Rigged and tested BOP. Displaced diesel with 10.2 CaCi?~. Spotted 100 sacks Canadian Blend Permafrost cement plug at 2250'~ Removed 1976' of dam.aged 13-3/8" casing. Displaced CaC12 with diesel oil from 1996' to surface. Re-installed wellhead. Released rig at 0200 hours, June 17, 1974.. This well has been temporarily suspended due to the unavailability of a 13-3/8" casing patch. APPENDIX B · .II CASING -- SIZE P_199' 2286 2306' 20" 16" 103~~ 66~9' 10,266' WELL B-3 (23-30-11-14) COMPLETION DATE: dULY 7~ 1970 LEGEND REVISIONS: JDATE'. i~l~ /~9_7c AREA: DRILLING GENERAL JOB TITLE' PRUDHOE BAY DEVELOPMENT ,DRAWING TITLE: WELL B-3 (25-30-11-14) _ i1__ i i i i" --~. MUD DIESEL CEMENT B-UPDATE WELL NUMBER J. INITIAL J ISSUE F, .f~-¢~. _x..-..:.,;. A., ,Chanqlncl,, , , Dray/lng I~'umber __. AL A S K A INC, ,~.~_? i NONE I NOV. 2, 1970 ..,,.,.;,,. , ,-__ 5D-W-17 B i~.~ I~ I ..... [ ,,,' ....... DD :'500' 4O30 5103' 6810' 9 242' 9260' 10~089' CASING 20" I6" SIZE WELL D-I COMPLETION DATE' (23-23-i i-i:5) MARCH 18~ 1970 LE6END MUD DIESEL CEMENT - _ REVISIONS: B-UFDATE WELL NUMBER j DATE: J .I_N_,_T.!A_L , I S ~U.'' _ ~ /1'°70 .";'-:""~"-"'" I A Char, grog._nrawlng N:Jmber . . .,;,:.,.,,::,.,-.,,,¢ B P AL A e - ^ I"" 2;., ]i:;Af'£ .......... u~ ~l - . 1 ' i$CALE ' --' J NOV. 2~ 10'70 DRILLING GENERAL i ~-~"~ } I NONE :;,,' ' ,.r' ,,~,,,,,,~ ,,,~: ,~:~ ~-,(~-.~-,,-,~, i;.: l".'-l---J A 7SD-W- ~.4 ~ .! · · m' TVD' 80' 2200', 2:500 4~00' CASING SIZE 92:00' WELL D-3(25-14-11-15) COMPLETION BATE: dUNE 7~ 1970 ALL DEPTHS ARE APPROXIMATE LEGEND MUD DIESEL CEMENT REVISIONS: - B-UPDATE WELL I'JUMF~ER t DATE I I?.IT.I~A_L I ISSUE ......... I]/_11 / 1970 J ~,....~_:Z'.;~-,,,,¢,." IA. Ch~notqg Drr~wln.rl Hum, b..er ......... B P A L A S" ,', A I NC. ,,~ IAREA; DRILLING GENERAL FRUDItOE BAY DEVELOPMENT-':" - JOB TITLE: SC^LE NONE NOV. DRAWING N° A75 D -W-16 . , . · ~1970 '-~S'su[ B TVD 80' 2~00~ 2:300 4~00' CASING F 20" I6" ~o :~' SIZE 9200' WELL D-5 (23-24-11-13R) COMPLETIOr~ DATE: SEPT. 8~ 1970 ALL DEPTHS ARE APPROXIMATE REVISIONS: DATE: ii/ii / 1970 -'~ ~g-/,_,~ Ch.]nmn-.'] Drawlnq Number . *~. ~-t~_ _ ~;~_. .....~ ' LEGEND MUD DIESEL CEMENT ! BP ALASI(A INC. ~,,;,, ~Z I ~:'", .. 1_. J .... ANCHO/~A GF ' AREA: DRILLING GENERAL JOB TITLE' PRUDHOE BAY DEVELOPMENT )RAWING TITLE: WELL D-5(23-24-11-13R) COhfPLETI ON DIAGRAM WELL NO. F-2 (22-11-11-13) KELLY BUSHING ELEV. 56.74 FT. AMSL 30" CONDUCTOR Cemented to surface with 250 sacks PermafT. ost Tr. 2O" CASING Cemente~d to surface wi th 4700 sacks Permafrost II. 60 bbis. cement returns. 13 3/8~9 5/8" CAS I NG Cemented in three stages: FIRST STAGE' 1310 sacks Class 'G', lOt'~ Gel and l. 25~ CFR-2. Lost returns while pumping plug. SECOND STAGE' 622'sacks Class 'G', 10~ Gel and 1.25~ D-65. THIRD STAGE' 2350 sacks Fondu Fly Ash. 7" LINER Cemented with 500 sacks Class 1.0~ D--65 and 0.4~'~ D-13. Reversed out 50 bbls cement. 'G' wi th Diesel Above 2500 '"' V V V BASE FLANGE ELEV. l L 30" SHOE 38.09 FT. Af~ISL 95 FT. I 1/4" TUBING 1802 FT. 13 3/8" STAGE TOOL 2167 FT. 13 3/8~9 5/8" X-OVER 2250 FT. 20" SHOE 2321 FT. 9 5/8" STAGE TOOL 4582 FT. TOP OF LINER 8802 FT. TJ.W. PACKER 8809 FT. T.I.W. HAltGER 8822 FT. 9 5,/8" SHOE 9302 FT. MARKER JOINT (TOP) 9768 FT, LANDING COLLAR 10880 FT. 7' SHOE 10928 FT. T.D. 8 l/2~ HOLE 10940 FT. ALL DEPTHS INDICATED ARE DRILLERS DEPTHS BELOW KELLY BUSHING Sidetracked hole from 9336 ft. after stuck pipe at I05!~8 ft. FISH:8 1/2 inch bit, reamer, 26 ft. P,,onel Drill Collar. stabilizer, ]8 ft. l~onel Drill Collar. Stabilizer, 12 ft. x 30 ft. Drill Collars, 21 ft. x 30 ft. heavy wall drill pipe from 9441 ft. to 10,528 ft. Cement plug 130 sacks Class 'G' wi th O. 1"~ D-65 and 20 lbs. frac. sand/sack cement. F-2 (22-11-11-13) COMPLETION DIAGRAM WELL NO. F-4 (12-01-11-13) KELL;f BUSHING ELEV. 57.07 FT. A~SL BASE FLANGE ELEV. 39.17 FT. AMSL 0" CONDUCTOR Cemented to surface wi th -?-50 sacks Permafrost 1-[ L 30" SHOE 18 5/8" CASING ~emented to surface with 200 sacks Permafrost -"3 3/8" 9 5/8" CASING Diesel above 2500' .-~ ... I 1/4" TUBING AT 13 3/8" STAGE TOOL 13 3/8"9 5/8" X-OVER 18 5/8" SHOE · Cemented in three stages ~-TA G E 1 1230 sacks Class 'G' with 1Of/, .. Gel, 1.25~;'D-65 and 0.4f~'D-13 TAGE 2 1180 sacks Class 'G' with 10~ Gel, 1.25-~; D-65 and 0.4f,' D,13 TAGE 3 1775 sacks PermafrOst ~ " 60 bbls cement to surface.' · ~ 5/8" ST~Cr TOOL TOP OF LINER T.I.W. PACKER T. I. W. HAII'GER 9 5/8II ~HOE ~" LINER -~.60 sacks Class 'G', 1.257~ D-65 and 4.~ D-13 tteversed out 15 bbls of cement. MARKER JOINT (TOP) LANDING COLLAR 7" SHOE T.D. 8 I/2" HOLE ALL DEPTHS INDICATED ARE DRILLERS' DEPTHS BELOW KELLY BUSHING LEGEND -Jz',,?~,,,;'/¢] C em e n t , J Mud J Diesel 95 Ft. . 1065 Ft. 1984 Ft. 2022 Ft. 2004 Ft. 4773 Ft. 8786 Ft. 8774 Ft. 8786 F t. · 9312 Ft. 9908 Ft. 1O621 Ft. 1O665 Ft. 1O676 Ft. F-4 (12-01-11-13) COI,IPLET I ON DIAGRAM ~gELL NO. F- 5 (14-:~6-12-15) KELL~ BUSHING ELEV. BASE FLANGE ELEV. 56.11 FT. A~ISL F T. AI~SL --30" CONDUCTOR 300 Sacks Permafrost 18-5..;8" CASING Cemented to surface wi th 7710 sacks Permaf¥ost ~. ,--13 3/8" 9 5/8" CASING Cemented in three stages ~TAGE 1:' 1170 Sacks Class 'G' with 10~ Gel, 1.25~ D-65, 12. B-13.0 lbs./gal. _,STAGE 2: 714 Sacks Class 'G' with 10~ Gel, 1.25~ CFR-2, 12.8-13.0 lbs./gal.' · . ~TAGE 3 880 Sacks Permafrost ~ !4.8 lbs./gal. Reversed out i0 I~lS cement.' "LINER 250 Sacks Class 'G' with 1.25~; D-65 Rd .4c~ D-13, 16.1 lbs./gal. 30" SilO E 1 1/4" TUBING AT 13 3/8" STAGE TOOL 13 3/8" 9 5/8 X-OYER 18 5/8" SHOE 9 5/8" STAGE TOOL. TOP OF LINER 9 5/8" SHOE ~ARKER ]OllqT (TOP) LANDING COLLAR 7" SHOE T.D. 8 I/2" HOLE 85 FT.. 1630 FT. 2216 FT. 2295 FT. 2315 FT. 5153 FT. 9777 FT. 1O26O FT. 10606 FT. 11765 FT. 11811 FT. 11816 FT. m. ALL DEPTHS INDICATED ARE DRILLERS' · LEGEND [///////;~ C e m e n t I .' '-! DEPTHS BELOW KELLY BUSHING · . F-5 (14736-12-13) COMPLETION DIAGRAM WELL NO. -F-6 (14-35-12-13) KELLY BUSHING ELEV. BASE FLANGE ELEV. 55.61 · ;]6.6 ] Fl'. Al, IS L FI'. AM~L: 30" CONDUCTOR 300 sacks Permafrost :IT cement. 18 5."8" CASING Cemented to surface with 5200 sacks Permafrost 'Fl' ~ 14.8 lb/gal. _56 bbls. good cement to surface. · ~ 30" SH0E I 1/4" TUBING AT 13 3..8" STAGE TOOL 13 3."8'y9 5/8 X-OVER 18 5..'8" SHOE I00 FT. 1137 FT. 2216 FT. 2259 FT. 2306 FT. 13 3/8" 9. 5/8" CASING Cemented in three stages Stage 1.' , 1130 sacks Class 'G' wi th 10.'.; gel, 1. D-65 and 0.4,"~ D-13 ,~ 12.8 lb/gal. Reversed out 10 bbls. cement. Stage 2' 570 sacks ClaSs 'G' with lOf~ gel, 1.25f~ D-65 andO.4~; D-13 ,~ 12.8 lb/gal. Stage 3: 1550 sacks Permafrost 1-[ ~ 14.8 lb/gal. Good returns to surface. .. 7" LINER --Cemented wi th 400 sacks Class 'G' wi-th 1.255 ~3-65 and 0.4?-, D-13 .~ 15.8 lb:gal. Reversed out excess cement. 9 5/8" STAGE TOOL TOP OF LIHER 9 5 8" SHOE MARKER JOINT (TOP) LANDING COLLAR 7" SHOE T.D. 8 I 2" HOLE 47'4i FT. 9202 FT. 9703 FT. 10,438 FT. 10,642 FT. Iq684 FT. I0,700 FT. LEGEND K/x/'///'_ZJ Ce,ne n t ] ]D,esel ALL DEPTHS INDICATED ARE DRILLERS' DEPTHS BELOW KELLY BUSHING Sidetracked hole from 6102 ft. after twist of'f at 6470 ft. FISH' 12 I'4 in. bit. float sub, 2- 12 I/4 sq. drill collars, 2-7 3'4 in. Mortal D.C., 4- 7 3 4 in. Steel D.C.,' 4- 12 I/4 in. Stabilizers from 6287 ft. to 6470 ft. CEI,IENT PLUG' ]50 sacks Class 'G' cement ivi th 20; sand and I0':, TIC. F-6 (14-35-~12-13) COt~IPLET I ON D IAGRAI~I WELL NO. H-3 (23-22-11-13) KELLY BUSHING ELEV. BASE FLANGE ELEV. 63.89 ~4.34 FT. A~,~S L FT. AI~IS L ,~0" CONDUCTOR -T/.5 sacks Permafrost qr. 8 5/8 CASING Cemented to surface with 3700 sacks ~f Permafrost ']]L 5 bbls excess cement. ,3 3/8" .9 5/8" CASING Cemented in three stages: ',TAGE 3: .40 bbls. fresh water flush, 200 sacks ~lass 'G' 'with 56 bbls. WG-7, 20 bbls. iud flush with 207.~ salt and 1300 sacks Permafrost TI. Good returns. · .. ~TAGE 2: 'N hhl= frnch wotar flush ~N enrl~ Class 'G' 'with lOf~ Eel and 1.25f~ D~5. ~ood returns. . _;TAGE 1: 40 bbl's, fresh water, 1050 sacks Class G' with 10~ ~el and !.25~ CFR-2. _ost returns. -o 7" LINER ',emented wi th 20 bbls. mud flush with CW-7 and 376 sacks Class 'G' with 1 25~ D-65 and 0.4~: ~-13. No cement returns. J .... :.'.': ;.:.... ' j Diesel above ' i-:; }.i:'i';' . ::: ':':'i;'-; '.::.::.!:- ' .! , :.,:.f::~'~;: ..-: . .:..,.. :: -:;i-- ' I -:::.:- :.. -.. -' :::i:]' '- . ; :.:;.: .: . ': i':': ' .: -: F ~,' -::;_--: 30" SHOE I I/4' TUBING 13 3/8' STAGE TOOL 13 3/8' x 9 5/8 X-OVER 18 5/8" SHOE 9 5/8" ST~0E TOOL 'TOP OF LINER ~I.W. PACKER T.I.W, HAIIGER 9 5/8" SHOE MARKER JOINT (TOP) LANDING COLLAR 7" SHOE T.D. 8 1/2" HOLE 100 737 2239 2281 2327 ' '"IU I ! 8876 8915 8919 9396 10100 10513 10555 10565 Ft. Ft. Ft. Ft. Ft. Ft. Ft. Ft, Ft. Ft. Ft. Ft. Fi. LEGEND .~.//>'////~.,//J C em e n t ,.":'.'f!'::.~"','?'-..,':1 ~,:'...";;;,,, ;L",,J [~ ti d ALL DEPTHS INDICATED ARE DRILLERS' DEPTHS BELOW KELLY BUSHING H-3 (23-22-11-13) COMPLETI ON DIAGRAM WELL NO. BP ]-.2 (,23-111-11-13) THIXOTROPI C MUD RTKB BASE FLANGE 57.33 2D.O FT. AMSL FT. AMSL _30" CONDUCTOR PIPE Cet~ented to. surface with · 175 sacks of Fondu Fly Ash. · -. 20' CASING -Cemented with 2000 sacks of Fondu Fly Ash. · ..13 3 8" CASING · Cemented in two stages: FIRST STAGE: 3000 sacks Class 'G' and 0.75~ CFR-2. SECOND STAGE: 135 bbls mud and 150 sacks Class 'G' Di-esel Oil .. 30" SHOE 13 3/8" STAGE'TOOL 20' SHOE 9 5/8' BRIDGE PLUG 9 5/8' STACF T~OL 13 3 8' SHOE .. 70 FT. 885 FT. 931 FT. 2140 FT. 2201 FT. 4589 FT. 9 5'8' CASING Cemented in three stages: FIRST STAGE: 1370 sacks Class 'G' SECOND STAGE: 725 sacks Class 'G' and 10~ gel, and 14 CFR-2 THIRD STAGE: 140 bbls mud 9 5/8' STAGE TOOL Plugged back to 9 5 8' SHOE 6504 10365 10410 10420 FT. FT.. FT. FT. -6- CO~PL~'[iO~ 0 IAGRAM WELL NO, 20" ~OI',E)UCTOR PIPE CEI~IENTED t//I TH PER!b'~FROST CEI~EttT. 13 3. '8" CAS I NC~ · Cemenled _Lo surface with II60 sacks of Pe[maffost Cement. 9 5:(8" CASING Cemented in tw'o stages. Ist" 720 Sacks Class G cment. 2nd S rage 665 sacks Class G cement. 7" CASING" Cemented in tw.~ stages. 1st S ta'ge' 600 sacks of Class G:'IS-',, Salt. 2nd Stage '. 390 sacks of Class G.;IO', Gel.. ALL DEPTtlS ARE DRILLERS DEPTHS IH rT. RIKB.. RTKD TO BASE FLANGE 18.3 It. · BP 01-11-12 -6- RT~B BASE FLAltGE 41.23 It. AI.ISL 22.93 'l. AI,ISL 20 " SHOE. 94 lt. 13 3.1~' SHOE 928 It. · 5'8" STAGE TOOL 1611 ft. 5 8" SilO[ 40DB It. STAGE 100L 50G6 It. "SILO[ 9366 It. IOTAL LI[P111 9397 It. FI CUIIE //l. /¢' I "-/t cOMPLETION DIAGRAM WELL NO. N-4 (23-07-11-13) KELLY' BUSHINI~ ELEV' 62.19 FT. AMSL BASE FLANGE ELEV. 39.08 FT. AMSL 30" CONDUCTOR Cemented to surface with 500 sacks Permafrost l-r 18 5/8" CASi'HG Cemented wi th 5400 sacks Permafrost 3::[ . -- 13 3/8" 9 5/8" CASING Cemented in three stages STAGE 3 1970 sacks Permafrost STAGE 2. · .. -- 840 sacks Class 'G' wi th 10~ Gel and 1.25~ CFR-2 STAGE 1 1200 sacks Class 'G' wi th 10~ Gel and 1.25~ CFR-2. · o. . .:.'- ... 9iesel above 2500' : ..: : .: ... .. · · ':: - :i :' ' ' · : '.-'::-:::': -' i: :' 30" SHOE 93 Ft. 1 1/4" TUBIHG AT 1680 Ft. 13 3/8" STAGE TOOL 2189 Ft. 13 3/8" 9 5/8 X-OVER 2269 Ft. 18 5/8" SHOE 2322 Ft. 9 5/8"'STAGE TOOL 4942 Ft. TOP OF LINER 8830 Ft. 7" LINER Cemented with 300 sacks Class 'G', 1.25¢~ D-65, 0.4¢~ D-13. 9 5/8" SHOE 9295 Ft. MARKER JOINT (TOP) 9475 Ft. LANDING COLLAR 7'1 SHOE T.D. 8 1/2" HOLE 1O413 Ft. 1O455 Ft. I0456 Ft. LEGEND , c e m e n t -.C:!.:y.:..!- :'1 ~Lld , } Diesel ALL DEPTHS I~/DICATED ARE DRILLERS' DEPTHS BELOW KELLY BUSHING . 11-4 (23-07-11-13) TVD 80' 78:53' . 9286~ CASING SIZE 2d' 133/8'' 95/8I' 11 i WELL, N-6 (34-05-1 I-I 3 COMPLETION DATE= MAY2,1971 LEGEND -", ':,:- MUD ,,,~.~ DIESEL CEMENT ; APPENDIX C C~PLETION DI AGRAI4 WELL NO. B-3 (23-30-11-14) RTKB 59.33 BASE FLAIIGE 40.60 FT. AMSL FT. AI'4SL 30" COIIDUCTOR PIPE Cemented wi th -150 sacks Permafrost cement. 20" CASItlG Cemented wi th 4360 sacks Permafrost cement.+ O. 5f; CFR-2 ' . . · .- 13 3"8 CASIItG 1st Stage: Cemented with 1200 sacks Class G cement (4?; gel)+ O. 5~ CFR-2. : Circulated out 140 sacks cement. . 2nd Stage: Diesel Oil around 16 Casing.' 30" SHOE 10 3/4' CASII~G CUT 16" SHOE 20' SHOE., 10 3/4' STAGE TOOL 9 5/8" 10 3/4" l-OYE~ 13 3/8" STAGE TOOL EZSV PACKER 13 3/8" SHOE 80 FT. 995 FT. 2199 FT. 2286 FT. 2306 FT. -,,,~,FT. 22~8 FT. 2§9D FT. 43:43' FT. _9 5'8" '10 3 4" CASi!iG 1st Stage' --- Cemented with 1550 sacks Class G (4f~ Eel, .0.5; CFR-2) and 725 sacks Class G (18'?; sal t, 15 CFR-2) ._ 2nd Stage. Cemented with 1225 sacks Class G (!0~ gel) 3rd Stage- 500 sacks Class G (salt saturated) plus 125 sacks Class 'G' 9 5/8' STAGE TOOL CEMENT PLUG 100 SACKS BRIDGE PLUG BRIDGE PLUG 9 5/8" BAFFLE 9 5/8" FLOAT 9 5/8" SHOE 6649 9381 9882 996O 10135 10220 10266 FT. FT. FT. FT. FT. FT. FT. "~,LL DEPTHS 1~tl31r:ATED ARE DPILLER$ DEPT~IS II; FT. RTKE;. RTKB TO f~ASE FI. AIIGE [LE~'ATiOIt 18.73 rT. T.D. HOLE PERFORATED FROI,4 9685 FT. to 9715 FT. 9730 FT. to 9760 FT. 9775 FT. to 9810 FT. 10309 FT. B-3 (23-30~-I I-I 4) COMPLETION D I AGRAM WELL NO. BP D-I (2.3-23-I1-13) (,AFTER PATCHING BACK) · 30* CONDUCTOR PIPE 250 sacks Permafrost cement ARCTIC PACK 1910~ TO SURFACE 20' CASING Cemented in two stages'. F! RST STAGE: " 850 sacks Permafrost. Lost circulation after 650 sacks. SECOND STAGE: - 4050 sacks Permafrost (annulus) 13 '3/8" CASING Cementea in two stages: FIRST STAGE'. 1600 sacks. Class 'G' (4¢~ Gel, 0.5¢~ CFR-2) Circulated-out approx. 85 bbls. cement. SECOND STAGE: Diesel 9 5/8"/10 3/4" CASING 1300 sacks Class 'G' (10~ Gel, 1¢~ CFR-2) plus 1400 sacks Class 'G' (18;~ Salt, 1¢~ CFR-2). -Lost returns on I~st 250 bbls. RTKB 57.53 FT. AMSL BASE FLANGE 34.05 FT. AMSL DIESEL 100~ TO SURFACE 50/50 DIESEL/RESIDUAL 2500' TO 10'0 13 3/8"- 72//- HBO 30" SHOE 84 13 3/8" FOC 190 ! 13 3/8" STUB 1945 TOP OF BAASH-ROSS PATCH 1945 16'" STUB 1963 10 3/4" STUB 1991 13 3/8" DY 2254 20" SHOE 2447 9 5/8~/10 3/4' X-OVER 2427 CEMENT PLUG 2947 - 2999 13 3/8' SHOE 4030 9 5/8' STAGE TOOL 5103 CEI~tENT RETAINER 6810 50 SACK CEMENT PLUG CEMENT RETAINER 9242 BRIDGE PLUG 9620 9 5/8" BAFFLE 10004 9 5/8" FLOAT 10045 9 5/8' SHOE 10089 ALL DEPTHS INDICATED ARE DRILLERS DEPTHS IN FT. RTKB. RTKB TO BASE FLANGE ELEVATION LEGEND' ~ Cement ~ Mud ~ Diesel ~ Cement Contaminated Mud 23.48' T.D. HOLE 10130 Rev i sed: 12/4/74 After Workover ° D-1 (2.3-1 1-13) :C~PLETION DIAGRA~ WELL NO. D-3 (23-14-11-13). (AFTER WORKOVER) 30' CONDUCTOR PIPE Cemented to surface with 150 sacks of Permafrost. 20' CASING Cemented to su.r~ace'with 4800 sacks of Permafrost. Lost.returns after 2400 sacks. 13 3/8" CASING Hew string of 13 3/8" casing tied back to surlace with Baash - Ross casing patch. 20" - 13 9'8" annulus displaced to Arctic Pack, Top 100' to-diesel. 9' 5/8"/10 3/4" CASING 'Cemented in three stages: FIRST STAGE: 1400 sacks Class 'G' (10~ Gel) 650 sacks Class 'G' (salt . . SECOND STAGE: 1300 sacks Class 'G' (10~ Gel) Lost returns while bumping plug: THIRD STAGE: 480 sacks Class 'G' (salt) ALL DEPTHS-INDICATED ARE DRILLERS DEPTHS IN FT. RTKB. RTKB TO BASE FLANGE ELEVAT!ON 20.12 FT. RTKB 58.22 FT. AMSL BASE FLANGE 38.10 FT. AMSL ARCTIC PACK 1960..-100 FT. 30' SHOE 75 FT. HOWCO FOC 1900 FT. TOP OF 10'3/4" 1993 FT. 13 3/8" BAASH -ROSS CASING 1994 .FT. 13 3/8" DV PATCH 2236 FT. 20' SHOE 2289 FT. 10 3/4' DV 2297 FT. 9 5/8" 10 3/4' X-OVER 2379 FT. 75~ DIESEL/25¢~ RESIDUAL 2512 FT. 9.1'NaCl to 3290 FT. 13 3/8" SHOE 4430 FT. 9 5/8' DV 7097 FT. CEMENT PLUG-50 sacks Class'G'.10313 FT. BRIDGE PLUG 10320 FT. 9 5/8' FLOAT 10755 9 5/8' SHOE 10797 T.D. HOLE 10802 ET.' FT. FT. Perforated from 10,417 ft. to ]0,521 ft. Revised: 1/15/75 after y~orkover D-3 (23-14-I 1-13) COI~PLET ~ OH D I A GR AM WELt. NO. O-5 (23-24-11-13q? AFTER WORKOVER Diesel/Residual RTKB 57.92 BASE FLANGE 40.3'2 FT. AMSL FT. AMSL 30' CONDUCTOR PIPE Cemented to surface with 150 sacks of Fondu Fly Ash. 20" CASING Cemented to surface with 4300. sacks Fondu Fly Ash and 30 sacks Class 'G' 16" CASING Not Cemented .13 3/8' CASING Cemented wi th 1500 sacks of Class. ?G' (4.~ Gel, 0..5¢.~ D-65) 30" SHOE ! 1/4" TUBING 13 3/8" STUB I0 3/4" STUB 16" STUB 3 3~8' DV 20" SHOE 10 3/4' DV 9 5/8~10 3/4" X-OVER 40,~; DIESEL/60.*~ RESIDUAL TO 50 'SACK CEMENT PLUG BRIDGE PLUG 13 3/8" SHOE 70 FT. 1718 FT. 1975 FT. ' 1991 FT. 1997 2247 2317 2310 2437 2500 2599 2696 4193 FT. FT. FT. FT. FT. FT. FT. FT. FT. 9 5/8~10 3/4' CASING Cemented in two stages STAGE 1 640 sacks Class 'G' (10,~ Gel, I~ D-65, !~ D-13) 'G' 8~ D-13) 760 sacks Class 1., salt, 0.Bf~ Lost circulation during pumping down. STAGE 2 1200 sacks Class 'G"(IO~ Gel) Displaced 10 3/4'- 13 3/8" Annulus wi th -diesel from 2300~ ALL DEPTHS INDICATED ARE DRILLERS DEPTHS IN FT. RTKB RTKB TO BASE FLANGE ELEVATION 17.60 FT. NOTE: 16~13 3/~' Annulus plugged with contaminated cement. '9 5/8" DV PLUGGED BACK 9 5/8" FLOAT 9 5/8" SHOE T.D. HOLE NOTE' SYI~BOL~ Indicates Casing Col lapse. Ran 9 I/2" Gua~e to 372', Ran 8 1/2" Gauge to 542', Ran 2 7/8" Tubing to 544'. · November 12, 1972. · -- Revised after workover - 2/'4/75 7417 10950 · 10993 11038 11069 FT. FT. FT. FT. FT. D-5 (23-24-11-13R) COMPLETIOH DIAGRAM WELL ~10. F-2 (27-11-11-13). AFTER CASiI~G REMOVAL SUMi~ER 1972 KELLY BUSHING ELEV. BASE FLANGE EEEV. 56.74 FT.'~SL 38.09 FT. A~SL 3O" CONDUCTOR Cemented to surface wi th 250 sacks Permafrost qr. 2o' CASING Cemented to surface wi th 4700 sacks Permafrost ~. 60 bbls. cement returns. .13.3/8'~/9 5/8" CASING Cemented in three stages FIRST STAGE: -1310 sacks Class 'G', 10~ Gel and 1.25~ CFR-2. Lost. returns while pumping plug. SECOND STAGE: 622 sacks Class 'G', lO, Gel and 1.25~ D-65. Circulated out 46 bbls. good cement. THIRD STAGE: 2350 sacks Fondu Fly Ash. 30" SHOE 1 1/4' ~UBING TOP 13 3/8'~ CASING TOC 2~/13 3/8" ANNULUS 13 3/8" STAGE TOOL TOP CEMENT PLUG 13 3/8" x 9 5/8~ X-OVER 20' SHOE MUD BELOW BOTTOM 11 1/2' RATHOLE 9 5/8' STAGE TOOL 95 1802 2083 2157 2167 2197 2250 2321 2500 2218 4582 FT. FT. FT. FT. FT. FT. FT. FT. FT. FT. FT. TOP OF LINER T.I.~Y. PACKER T.I~, HAflGER 9 5/8' S OE 8802 8809 8822 9302 FT. FT, FT,. FT. 7" LINER :Cemented with 500 sacks Class 'G' with 1.0~ D-65 and 0.4~ D-13. Reversed out 50 bbls. cement. MARKER JOINT (TOP) 9768 FT. LANDING COLLAR 10880 FT. 7' SHOE T.D. 8 1/2" HOLE ALL DEPTHS INDICATED ARE DRILLERS DEPTHS BELOW KELLY BUSHING Sidetracked hole from 9336 ft. after stuck pipe at 10568 ft. FISH 8.1/2 inch bit, reamer, 26 ft. Monel Drill Collar, stabilizer, 18 ft. Monel Drill Collar. Stabilizer, 12 ft. x 30 ft. Drill Collars, 21 ft. x 30 ft. heavy wall drill pipe from 9441 ft. to 10,528 Ft. Cement plug 130 sacks Class 'G' with 0.1~ D-65 and 20 bls. frac. sand/sack cement. LEGEND ' Cement r---] Mud ~ Diesel Completion Fluid 10928 FT. 10940 FT. ~'-2 '(22-11-11-13 30" CONDUCTOR Cemented to surface.with 250 sacks Permafrost 5/8" CAS H Cemented to surface wi th 9200 sacks Permafrost 13 3/8'/9 5/8' CA'SING Cemented in three stages STAGE 1: 1230 sacks Class 'G' wi th 10,~ Gel, 1.25~ D-65 and O. 4~ D-13. STAGE 2: 1180 sacks Class 'G' with lOt~ Gel, I. 25~ D-65 and O. 4~ D-13. STAGE 3: 1775 sacks Permafrost t-I' 60 bbls. cement to surface. o .. COMPLETION DIAGRAM WELL ~10. F-4 ( 12-0 I-11-13) AFTER CASING REI~OVAL SUP.~MER 197 2 KELLY BUSI~iNG ELEV.57.07 FT, kMgL BASE FLANGE ELEV. 39.17 FT. ~SL 3O" SHOE 95 FT. I 1/4' TUBING 1065 FT. TOP 13 3/8" CASING 1934 FT: TOC 18 5/8:/13 3/8" ANNULUS 1950 FT. 13 3/8" STAGE TOOL 1984 FT. TOP CEI~ENT PLUG 2018 FT. 13 3/1~ x 9 5/8' X--OVER 2022 FT. 18.5/8" SHOE 2004 FT. MUD BELOW 2500 FT. 9 5/8" STAGE TOOL 4773 FT. TOP OF LINER 8786 FT. T.I.W. PACKER 877 5 FT. TJ.W. HAHGER 8786 FT, 9 5/8' SHOE 9312 FT. o. 7' LINER 460 sacks Class 'G', 1.25~ D-65 and 0.4~ D-13 Reversed out 15 bbls. of cement. MARKER JOINT (TOP) 9908 FT. LANDING COLLAR 10621 FT. · 7~ SHOE 10665 FT. T.D. 8 !/2" HOLE 10676 FT. ALL DEPTHS INDICATED ARE DRILLERS DEPTHS BELOW KELLY BUSHING LEGEND Cement ~ud Diesel Completion Fluid .: F-4 (21-01-1 1-13)' COMPLETI 0~! DI AGR/~! WELL NO. F-5 (14-30-12-13) AFTER CASING REI~]OVAL ~J~ER 1972 _ .- KELLY BUSHItlG ELEV. 56. II FT. AMSL BASE FLANGE ELEV. 37.1l FT. A~SL 30' CONDUCTOR 300 Sacks Permafrost ]:[. 18 5/8" CASING Cementbd to surface wi th 7710 sacks Permafrost 'Fl'. 13 3/8" x 9 5/8" CASING Cemented in three stages' STAGE 1' 1170 Sacks Class 'G' with 10~ Gel, 1.25~ D~'5, 12.8-13.0 lbs/gal. STAGE 2' 714 Sacks Class 'G' with 10~ Gel, 1.25~ CFR-2, 12.8-13.0 lbs/gal. STAGE 3: 880 Sacks Permafrost ]::[ 14.8 lbs/ gal. Reversed Out 10 bbls. cement. Dieset Above .: v Cement Ptug · . -. ..: _ . . '.:::.-'..~-~-i ::_:- :_ -._: ::: _- :??.-_ :':%: .- ,f:-. - :~-: 3O" SHOE 85 FT. 1 1/4" TUBIHG 1630 FT. TOP OF 13 3/8" CASIHG STUB 1971 FT. 'TOP OF FISH (16 3/4"W/0 Shoe and Coupling) 1983 FT. TOP OF CE&IENT 18 5/8°x 13 3/8" Annulus 1991 FT. 13'3/8" STAGE TOOL 2216 FT. TOP OF CEMENT PLUG 2256 FT, . BOTTOM OF 11 1/2" RAT HOLE 2282 FT. 13 3/8" ~( 9 5/8" X-OVER ' 2295 FT. 18 5/8" SHOE 23~5 FT. BOTTO~ OF CEMEIIT PLUG 240,0 FT. 9 5/8" STAGE TOOL 5153 FT. TOP OF LINER TI~V. PACKER T.I.~. HAilGER 9 5/8" SHOE 9777 FT. 9786 FL 9777 FL 10260 FT. 7" LIttER 250 Sacks Class 'G' with 1.25~ D-65 and 0.4,'.~ D-13, 16.1 lbs/gal. MARKER JOlllT (TOP) LANDING COLLAR 7' SHOE T.D. 8 1/2" HOLE ALL DEPTHS INDICATED ARE DRILLERS DEPTHS BELOW KELLY BUSHING' LEGEND ~ Cement ~ Mud ~ Diesel ~ Completion Fluid 10606 FT. 11785 FT. 11811 FT. 11816 FT. F-5 (14-36-12-13) -.. ' COMPLETION DIAGRAM WELL NO. F-6 (14-35-12-1.3) AFTER CASING RE~VAL SU~ER 1972 KELLY BUSHING ELEV. 55.61 FT. AMSL BASE FLANGE ELEV. 36.61 FT. AMSL .30" COt{DUCTOR 300 Sacks Permafrost 'IT cement. 18 5/8" CASINO cemented to surface with 5200 sacks Permafrost ~ at 14.8 lb/ gal. 56 bbls. good cement to surface. · 13 3/8" x 9 5/8' CASING Cemented. in three stages' STAGE 1: 1130 sacks Class 'G' with 10¢~ gel, '1.25¢~ D-65 and 0.4~ D-13 at 12.8 lb/ gal. Reversed out 10 bbls. cement. STAGE 2: 570 sacks Class 'G' with 10¢~ gel, 1.25~ D~5 and 0.4f~ D-13 at 12.8 lb/- gal. Rever-~ed e!lt 12 bhls. cement. STAGE 3: 1550 sacks Permafrost 'TI' at 14.8 lb/gal. Good returns to surface. Diesel Above Cement Plug : --~ ' : :-: · '" i:'i .:. 30' SHOE 100 FT. . 1 1/4' TUBING 1137 FT. 'TOP OF 13 3/8" CASING STUB 2033 FT. TOP OF CEhIENT18 5/8~ x13 3/8" ANNULUS 2053. FT TOP OF CB,tEt~T PLUG 2i99 FT. 13 3/8" STAGE TOOL 2216 FT. BOTTO~I OF 11 1/2"RAT HOLE 2234 FT. 13.3/8"x 9 5/8"X-OVER 2259 FT. 18 5/8' SHOE 2306 FT. BOTTOM OF CEI~IENT PLUG 2350 FT. 9 5/8" STAGE TOOL 4741 FT. ' Or, IT: T.I.W. HAtiGER ' 9256 FTo 9 5/8" SHOE 9703 FT. 7" LINER Cemented with 400 sacks Class 'G' with 1.25~ D-65 and 0.4% D,13 at 15.8 lb/gal. Reversed out excess cement. MARKER JOINT (TOP) 10438 FT. LANDING COLLAR 10642 FT. 7" SHOE 10684 FT. T.D. 8 1/2" HOLE 10700 FT. ALi' DEPTHS ItlDICATED ARE DRILLERS DEPTHS BELOW KELLY BUSHING LEGEND ~ Cement ~ ~lud ~ Diesel ~ Completion Fluid Sidetracked hole from 6102 ft. after twist off at 6470 ft. FISH 12 i/4 in. bit, float sub, 2- 12 1/4 sq. drill collars, 2-7 3/4 in. Monel D.C., 4- 7 3/4 in. Steel D.C., 4- 12 1/4 in. Stabilizers from 6287 ft. to 6470 ft. CB,IENT PLUG 150 sacks Class 'G' cement with 20¢~ sand and lOf~ TIC. C,3~]PLETIOH DIAGRAM . . ~:L.L NO. BP i-2 (,23-10-11-13) . ~RKOYER CO~PLETIOH 10-17-72 . - .30* CONDUCTOR PIPE Ce~nentaG. tff surface with 175 sacks of Fondu Fly Ash. 20' CASIHG Cemented wi th 2000 sacks of. Fondu Fly. Ash.' 1,j 3/8" CASING ,. Cemented in two stages: · FIRST STAGE: 3000 sacks Class 'G' 'and 0.75~ CFR-2. SECOND STAGE; 135 bbis mud and 150 sacks Class 'G' .9 5,/8" CASING Cemented in three stages; F I RST STAGE: 1370 sacks Class 'G- SECOND STAGE: 725 sacks Class .'G' 'and 10¢;, gel, and 1~- CFR-2 THI RD STAGE: 140 bbls mud THIS DRAWING WAS A1.4ENDED SEPT. 13, 1972 FOR W,"O OPERATIOHS. . . RTKB '57.33 FT. AhlSL BASE FLANGE 37.33 FT. A~ISL Diesel Oil. 30" SHOE 70 FT. Approx. Top of Cement (13 3/8" x 2O") 7oo FT. 13 3/8" STUB 908 FT. WASHED OVER 925 FT. 20" SHOE 931 FT. 9 5/8" STUB 2050 FT. 9, 5/8" WASHEDOVER '2083 FT. CMT PLUG 2100 FT. 9 5/8' BRIDGE PLUG 2146 FT. 9 5/8" STAGE TOOL 2201 FT. 13 3/8' SHOE.. 4589 FT. 9 5/(~" STAGE TOOL- 6504 FT. 2 Brine Plugged back to 10365 9 5/8" SHOE 10410 T.D. 1O42O FT. FT.' J-2 (23-10-11-13 COIZPLETIOI~ DI A,,,~A~ WELL NO. I~-1 (43-01-11-12) . . 20" CONDUCTOR PIPE .Cemented with 108 sacks Permafrost' cement. ' 13 3/8" CASING Cemented to surface with 1240 sacks Permafrost cement. 9 5/8' CASING Cemented in two stages' 1st STAGE' ??r~ s~cks r'l~ '(~' 'c~-mp. nt. 'Ho Returns 2nd STAGE: 685 sacks Class 'G' cement.' 30 sacks circulated out.' .: 7' CASING Cemented in two stages: 1st STAGE: 600 sacks of Class"G', 18~ Salt.- 2nd' STAGE: 390 sacks of Class 'G',' 10,~ Gel. ,WORKOYER 7" casing collapsed, replaced to 1922 Ft. and cemented with 350 sacks fondu fly ash cement. ALL DEPTHS ARE DRILLERS DEPTHS IN FT. RTKB. RTKB TO BASE' FLAHGE 18.3 FT. RTKB 41.23 BASE FLANGE 22.93 SHOE 13 3/8'. SHOE 9 5/8" STAGE TOOL Bowen ~atch in 7 inch Casing 9 5/8' SHOE 7" STAGE TOOL 2 7/8' BRIDGE PLUG 2 7/6' TU§ING SHOE 7~ PERMATRIEVE PLUG 7' BRIDGE PLUG 7' BRIDGE PLUG 7~ SHOE . TOTAL DEPTH PERFORATED FROI~I 8822 FT. to 8882 FT. 8966 FT. to 9008 FT. 9055 FT. to 9060 FT. fl-1 (-43=01-11-12) _ _ FT. ~I~SL FT. AMSL 50 FT. 928 FT. 1612 FT. 1922 FT. 4088 FT. 5O66 FT. 7714 FT.' 8710 FT.' 8710 FT. 8945 FT.' 9040 FT.' 9366 FT. 9397 FT. ' COMPLETION DIAGRA~ WELL NO. N-4 (23--07-;1-13) KELLY BUSHIHG ELEV. 62.18 I-'T. AI~SL 3O" CONDUCTOR Cemented to surface with 500 sacks Permafrost'Fl' _18 5/8' CASING Cemented. wi th 5400 sacks - Permafrost II. 60 bbls. cement to surface. 13 5/8" Cemented in three stages: STAGE 3: 1970 sacks Permafrost II. STAGE 2: 840 sacks Class-'G' with 105; Gel and 1.25',',~ cfr-2. 'STAGE l: 1200"sacks Class "G' wi th 10~ Ge ! and 1.25,~ CFr-2. , LINEr C~mented ~ith 300 sacks Class 'G', 1.25.~ D-65, 0.4S D-13. BASE FLAIIGE ELEV. Diesel above 2200' . .. ALL DEPTHS INDICATED ARE DrILLErS DEPTHS BELO:# KELLY BUSHING 39.08 FT. AMSL 30' SHOE 93 FT. 180 SK PERI~AFROST CEMENT PLUG 1000'- 1197 FT. TOP OF FISH 2272 FT. 1 1/4" TUBIltG 1880 FT. 13 3/8" STAGE TOOL 2188 FT. 13 3/8'~9 5/8" X-OVER 2269 FT. 18 5/8~ SHOE 2322 FT. 9. I ppg. CaCI/gel mud BOTTOf~t OF FISH 2345 FT. 9 5/8' STAGE TOOL 4942 FT. 10.'2 ppg;. HaCl/soda-ash brine TOP OF LINER 8830 FT. T. I. W. PACKEr 8837 FT, T. I. W. HAI~GEr 8830 FTo 9 5/8" SHOE 9295 FT. MARKER JOINT (TOP) 9475 FT. 7" BRIDGE PLUG 10235 FT. PERFORATI OltS 10252-10272 FT. 10280-10300 FT. LANDING COLLAr 10413 FT. 7" SHOE 10455 FT. T.D. 8 1/2" HOLE 10456 FT. NOTE: SEE DR¥1G. N-4 (FIG. I) FOR AFTER Y~'ORKOVER D1AGRAI~ WELL ABANDONED, RIG rELEASED 1400 HOURS, MAY 31, 1974. 23-07-11-13) coMPLETION D IAGRAI~I WELL NO. N--6 (34-05-11-13) KELLY BUSHING ELEV. 61,85 BASE FLANOE ELEV. 4,".49 · , · FT. A~,lSL AMSL --30" CONDUCTOR 250 Sacks Permafrost 18 5/8" CASi.NG 1450 Sacks Permafrost TT 60 bbls cement to surface. 3/8" 9 5/8" CASING ;emented in three stages ~tage 1: 160 Sacks Class 'G' with 10~; Gel, !. 25~,~ CFR-2. ~tage 2. _ _ 83 Sacks Class 'G' with lOf-,~ Gel, 1.25,'~ CFR-2. tage 3: _ ~BO Saci~s Permairost 7_I '":1 bbls cement to surface.' 30" SHOE 99 FT. 1 I/4" TUBING AT 743 FT. TOP OF 13 3/8" CASING 1976 FT. PERhlAFFIOST CE~.iENT PLUG 2186 - 2250 FT.' 13 3/8" STAGE TOOL 2246 13 3/8" x 9 5/8" X-O VER 2287 FT. 18 5/8" SHOE 2332 FT. 9 5/8" STAGE TOOL, 5039 FT. "LINER 5...50 Sacks Class 'G' Cement with 1.25f; -65 and O. 4f; D-13. TOP OF LINER 8927 FT. B.O.T. P A C K E R 8933 FT. B.O.T. HAItGER. 8936 FT. 9 5~.'8" SHOE 9466 FT. I~ARKER JOINT (TOP) 10699 FT. LANDING COLLAR 11153 FT. 7" SHOE 11200 FT. T.D. 8 1/2" HOLE 11213 FT. ALL DEPTttS INDICATED ARE DRILLERS' LEGEND ~ Cement t. ' '.' ~ ...... ]Brine DEPTHS BELOW KELLY · BUSHING · N-.6 ( ,34-05-1 i-i 3) AtlanticR. ichfieldCompany NorthAme~' q Producing Division North Alask"~, ,.dstrict Post Office Box 360 Anchorage, Alaska 99510 Telephone 907 277 5637 December 10, 1975 Mr. T. R. Marshall, Jr. Executive Secretary Alaska Oil and Gas Conservation Committee 3001 Porcupine Drive Anchorage, AK 99S01 Dear Mr. Marshall: Attached are reports and papers as you requested at the Prudhoe Bay Field Rules hearing on November 25, 1975. Included are S.P.E., A.P.I., and A.S.M.E. papers entitled: "A Study of Factors Influencing the Mechanical Properties of Deep Permafrost" by R. A. Ruedrich and T. K. Perkins, Atlantic Richfield Company. le . "Studies of Pressure Generated Upon Refreezing of Thawed Permafrost Around a Wellbore" by T. K. Perkins, J. A. Rochon and C. R. Knowles, Atlantic Richfield Company. 0 "The Mechanical Behavior of Synthetic Permafrost", by T. K. Perkins and R. A. Ruedrich, Atlantic Richfield Company. . "Solutions for Some Problems Resulting from Refreezing of Permafrost Around a Wellbore" by T. K. Perkins, G. R. Wooley and F. W. Ng, Atlantic Richfield Company. "Removal of Water Base Mud from Wellbores in Permafrost Zones", by F. W. Ng, Atlantic Richfield Company. "Precise Joint Length Determinations Using a Multiple Casing Collar Locator Tool", by R. A. Ruedrich and T. K. Perkins, Atlantic Richfield Company and D. £. O'Brien, Exxon Production Research Company. Also included are stress-strain data as described in the attached memo by C. R. Knowles dated December 9, 1975. If you have any additional questions, please contact C. R. Knowles or myself. Vary truly yojjrs, ~,.,.5. A. Rochon JAR/vaf Attachments cc: C. R. Knowles (w/o attachments) A Study of Factors Influencing the Mechanical Properties of Deep Permafrost R. A. Ruedrich, SPE-AIME, Atlantic Richfield Co. T. K. Perkins, SPE-AIME, Atlantic Richfield Co. Introduction The permafrost in the arctic regions has become a major new design factor for the practicing petroleum engineer. The drilling and completion practices of the petroleum industry must be modified in view of the mechanical and thermal behavior of the permafrost. At Prudhoe Bay, for example, completion procedures will be greatly influenced by the presence of nearly 2,000 ft of permafrost consisting of layers of gravel, sand, silt, and clay that are permanently frozen. Studies of the thermal behavior of wells completed in permafrost~-~ have indicated that some thawing around wellbores is generally to be expected during drilling and production operations. Other studies4,~ have shown that pressures will rise as saturated, thawed regions around a wellbore are refrozen. The level to which external pressures rise will be signifi- cantly influenced by the elastic and flow behavior of the surrounding frozen soil. There are a number of publications6-~'' dealing with the mechanical behavior of frozen soils. Much of the civil engineering litera- ture deals with the behavior of shallow permafrost at relatively Iow stress levels. Chamberlain et al.,~ how- ever, have reported behavior at very high stress levels that might be of interest for nuclear excavation or detection. In this paper we report a study of the factors influencing the mechanical behavior of .deep permafrost. We shall discuss the influence of these seven factors: (1) confining stress, (2) pore pressure, (3) mineralogy, (4) electrolyte concentration, (5) tem- perature, (6) strain, and (7) strain rate. These data can be incorporated into numerical4 or analytical5 models to predict stresses produced by re- freezing thawed zones in the permafrost. The Interaction of Ice and Soil Since permafrost is a composite material, we should expect it to exhibit mechanical characteristics similar to those of its constituents, ice and soil. A formal approach to ice/soil interaction is enlightening. Fig. 1 illustrates a cylindrical element of permafrost sub- jected to a total axial stress, ,~, and a total lateral stress, ,~t. Some fraction of the axial stress, *~8, is carried through the soil matrix; and the remainder, *ai, is carried through the ice phase. eat = ~8 + {r~i. .. (1) Similarly, the total lateral stress can be divided into that portion carried by the soil matrix, ~z,, and a por- tion carried through the ice phase, The uniaxial strength (i.e., when the sample is loaded along the axis but no lateral force is applied) of pure multicrystalline ice samples is knownTM to be a func- tion of temperature, strain rate, and strain. It has been reported in the literature,~ and confirmed in our labo-/ ratories up to lateral stresses of 1,000 psi, that f~ot ice under triaxial stress, the difference between axial and lateral stress is a function only of tempe~r~(~ure~. strain rate, and strain. The suppression of/f~eezirig In a detailed experimental investigation, seven factors a£]ecting the mecha~c~i"behavior of permafrost have been evaluated. That behavior is dictated by a combi~,/a.{ion of the resistance of ice to deformation and the elastic and flow behavior of the~:s°il Constituents. OCTOBER, 1974 1167 point by pressure has a small effect on strength, and that effect can be accounted for by correlating strength with temperature measured below the freezing point? ,~,,,~ = ~ ..... - *z .... (3) where o.,,,, = uniaxial strength of multicrystalline ice, a function only of temperature below the freezing point, strain rate, and strain, psi, ~r~,,, = lateral stress acting through multicrystal- line ice, psi, o-a,,, = axial stress acting through multicrystal- line ice, psi. For permafrost, the uniaxial strength must result from the resistance of ice to deformation in the pres- ence of the soil matrix. For a given lithology, the uniaxial strength of permafrost was shown°,~2 to be a function of temperature, strain rate, and strain, as would be expected for the ice component. The effect of hydrostatic pressure on freezing point and on water/soil interaction has been shown to be small at (-/'as O'at O'ai Fig. l~Stresses acting in a cylindrical permafrost sample. TRANSDUCER FOR RECORDING PRESSURE ~ PRESSURE SENSOR FOR / ' MAINTAINING CONSTANT 1 PRESSURE (ADJUSTABLE) ~'?1 SCREW T:..,~E PISTON PUMP ORL LLEO P F.qF;HRF vo.u . Pt. RMAFROST SAMPLE IN CUP E'~HYLENE GLYCOL BATEI, / TE'AP. CONTROLLED & AGITATED, IN FREEZER Fig. 2--Apparatus for determining unfrozen water content. pressures up to 1,000 psi. Uniaxial deformation or deformation of a triaxially confined sample would require essentially the same deformation history for the ice component, and we would therefore expect the relationship shown by Eq. 4. where = uniaxial strength of a permafrost sample, a function only of temperature below the freezing point, strain rate, strain, and lithology. Substitution of Eqs. 2 and 4 into Eq. 1 and rearrange- ment yields Eq. 5. (~a,- ~,,)= ~ + (~.~- ~,~) . . (5) The difference between total axial and total lateral stresses .during deformation of permafrost can thus be viewed as consisting of two components. The first component, which is time dependent, results from the resistance of ice to deformation in the presence of the soil matrix, and for the range of conditions of interest is essentially equal to the uniaxial strength of the permafrost sample. This contribution to strength will be a function of temperature below the freezing point, strain rate, strain, and electrolyte concentra- tion and will be influenced by the interaction of water with clay materials in the soil matrix. The second component, which is essentially inde- pendent of time, results from the resistance of soil to deformation under the constraint of the imposed boundary conditions. This contribution will be inde- pendent of temperature over the small range of tem- peratures of interest for permafrost, and will be in- sensitive to strain rate if loads are applied relatively slowly. This strength component will be a function of mineralogy and lithology, degree of compaction, and stress level and strain. In the following sections of this paper, permafrost and soil strengths and tests of Eq. 5 are described. Experimental Studies of Factors That Influence Deep Permafrost Behavior Consider first those factors that influence the resist- ance of ice to deformation. In natural permafrost, there are a number of variables that determine the fraction of water that will be frozen at a given tem- perature?-s° Two experimental approaches using calorimeters and dilatometers for investigating these relationships have been mentioned in the literature. We have chosen to use a dilatometer (Fig. 2) that will operate at pressures up to 1,000 psi. In using this equipment, a sample of thawed permafrost is frozen under pressure in a heavy-walled vessel and cooled to a low temperature. The temperature of the bath is then raised in steps while the pressure on the system is held constant with automatic devices. At each selected temperature, enough time is allowed for the system to come to equilibrium. The volume change accompanying a change of temperature is measured to within 0.0025 CC. The changes in volume of the vessel and of the fluid (determined from a previous calibration) are subtracted from the total change of 1168 JOURNAL OF PETROLEUM TECHNOLOGY volume to yield the volume change of the permafrost sample. The permafrost volume change is corrected slightly for thermal expansion of permafrost com- ponents to yield the change of volume due to phase change. After subsequent analysis of the total water content, volume changes resulting from the phase change are converted into pounds of unfrozen water per pound of soil solid as a function of temperature. Our studies indicated that the following factors affect the fraction of water in the ice phase: (1) temperature, (2) pressure, (3) electrolyte concentration, (4) min- eralogy, and (5) state of clay hydration. In preliminary studies, soil samples that had been dried, but then reconstituted with water, showed a smaller amount of unfrozen water than comparable samples that had never been dried. All subsequent studies w.ere conducted with preserved and fully hy- drated soil samples. Pressure suppresses the freezing point of water and also has a detectable effect on water/soil interaction as has been suggested in the literature? The effect of pressure is illustrated in Figs. 3 and 4. Analysis of a core taken through the permafrost at Prudhoe Bay has shown salt content of the forma- tion water varying from a few tenths of a percent to as high as 2.5 percent near the bottom of the permar frost. As water is converted into ice, the electrolyte remains in the liquid phase, thus concentrating and lowering the freezing point of the brine. Fig. 3 com- pares the electrolyte concentration effect for an "as received" sample containing 0.97 percent NaC1 with the behavior of a leached sample containing approxi- mately 0.12 percent NaC1. One of the more important factOrs influencing un- frozen water content is the amount of clay or fine material present. It is likely that water interacts with clay surfaces to partially orient the water molecules. There will also be a capillary pressure effect in fine- grain soils. 80 As the temperature changes, bOund water will not exhibit a volume change that would be ex- pected for unbound water. The water associated with soil constituents has been calculated from the dila- tometer data. Values for fully hydrated samples of near-surface Prudhoe Bay permafrost soil and a typi- cal drilling mud are shown on Fig. 4. Silt samples containing between 10 and 15 percent clay exhibited more bound water than sand and silty sand samples containing less than 5 percent clay. These values for associated water are of the same order of magnitude as values that w.ere previously reported in the litera- ture and that were obtained at low pressure. The combined effects of pressure, salinity, and lithOlogy on in-situ ice content have been estimated for Prudhoe Bay and are shown in Table 1. High Strain Rate Uniaxial Compression Tests Factors influencing the maximum uniaxial compres- sive strength of recompacted, saturated permafrost samples are (1) temperature, (2) strain rate, (3) salin- ity, (4) lithology, and (5) mineralogy. Three types of natural permafrost soils from Prud- hoe Bay have been selected for investigation. Tables 2 and 3 give particle size .distribution, mineralogy, and natural salinity. The procedure we used for pre- 0.45 0.40 0.35 0.30 0.25 LEGEND: ; 1000 PSI LEACHED } 0.12% NaC, 100 PSi LEACHED · 1000 PSI AS RECEIVED ~ 0.97% NaC~ · 100 PSi AS RECEIVED 0.12 0.10 - 0 08{ 0.06 0.04 -- __._....- 0.00 I I I I 10 15 20 25 30 TEMPERATURE, °F Fig. 3--Unfrozen water content in a Prudhoe Bay near-surface sand sample. 35 0.20 0.15 0.10 0.05 · 100 psi · A 1000 psi DRILLING MUD SILTY SAND AND SAND , I 35 · · 0 0 15 20 25 30 TEMPERATURE, °F Fig. 4---Unfrozen water content of drilling mud and natural Prudhoe Bay so!Is. (Soil samples were leached so that electrolyte concentration was negligibly small.) 40 OCTOBER, 1974 1169 TABLE 1--ESTIMATED IN-SITU ICE CONTENT AT PRUDHOE BAY In-Situ Average Ice Depth Ternperature Content (ft) (° F) (lb ice/cu ft soil) 200 15.1 27.0 400 17.5 25.6 600 18.8 17.2 800 20.6 10.6 1,000 22.3 10.5 1,200 24.1 14.5 1,400 25.7 19.8 1,600 27.2 20.7 1,800 28.9 18.0 1;900 29.6 6.5 paring samples has been described in Ref. 6. Maxi- mum uniaxial s. trengths are shown on Figs. 5, 6, and 7 for a temperature range from 8° to 28'F and for strain rates from 5 × 10-~ min.-~ to 5 × 10-'°' min.-~ The influence of salinity has been investigated further by preparing samples with interstitial water salinities of 0, .1, and 2¥2 percent NaC1. Maximum uniaxial strengths of natural sand and silty sand samples con- taining waters of these salinities are shown on Figs. 8 and 9 at different strain rates and temperatures. Uniaxiai Creep Tests When subjected to small shear stresses, ice within permafrost can creep at very low rates. Over many years, the cumulative ~deflection that can occur at low strain rates may be of engineering significance. The experimental study of creep presents practical diffi- culties because of the long periods of time required to gather data and because of the investment in equip- ment needed for running simultaneous experiments. The creep of recompacted natural soil samples has been studied using the technique described in Ref. 6,' The sequence of stress application has been modified, however, so that data can be obtained at lower strain rates in a practical amount of time. For synthetic sand-ice samples, the empirical relationship° of tem- perature, uniaxial stress, strain, and strain rate is given by Eq. 6. where (6) E ~ ~ -- stress, psi, · e = strain rate, minutes-% ~ = strain, ~.m ~,, = strain at which the minimum strain rate is achieved at essentially constant axial stress, and K~ and K, = empirical constants that are a function of temperature. To test the validity of this equation for natural soil samples, recompacted specimens were prepared from Prudhoe Bay soil and tested at various temperatures and at constant strain rates over the range 5 X 10-~ min.-' to 5 X 10-~ min.-'~. Values of the coefficients 10-1 - I I I Illll ' '1 I I Illll I I I Illi~- _ _ _ _ _ , _ -- 13 - - · 8o1: zs 18OF n 28OF I II I I I I 1111 10-2 10-4 10-5 .- ~ 10-3 Z er' 10 I00 1000 10,000 MAXIMUM UNIAXlAL STRENGTH. psi Fig. §~Maxirnurn uniaxial strength of recompacted, Prudhoe Bay sand. (Before freezing, pore space was saturated with water containing 0.97 percent Natl.) 1170 10-1 [ ' ' '[[" ' [ ' [['"l ~ ' ' '][['--- _ - 7_ , -- _ ! , · 8°F ~, 18OF u 28°F I I I III1 I I I IIII I I I II11 10-2 7 ._ E ~ 10-3 Z _ 1-- 10-4 10-5 10 100 1000 10,000 MAXIMUM UNIAXIAL STRENGTH, psi Fig. 6~Maximum uniaxial strength of recompacted, Prudhoe Bay silty sand. (Before freezing, pore space was saturated with water containing 0.15 percent NaCI.) JOURNAL OF PETROLEUM TECHNOLOGY' Ks and K~_ were determined from those data. Addi- tional permafrost samples were then placed in loading frames and were subjected to a uniaxial stress suffi- cient to strain the sample about 1 percent at strain rates of about 10-5 min.- ~. The data were fitted to Eq. 6 to evaluate Stain. The axial stress was then reduced by one-half. Additional deflections were measured over a period of several weeks so that the strain rate could be established with reasonable precision. Then the axial load was increased in a stepwise manner to obtain data at intermediate strain rates. Results for recompacted natural Prudhoe Bay sand samples are shown on Fig. 10. These samples behaved like the sand-ice cylinders reported previously. The straight- line relationship appears to be valid for strain rates as Iow as 3 >< 10-9 min.-~. Studies ct Soil Behavior Let us turn our attention now to the contribution of the soil matrix to permafrost strength. Eq. 5 suggests that this contribution can be measured with thawed Samples, Provided the soil is brought to the proper initial state and provided proper boundary conditions are imposed during deformation. We have studied the deformation behavior of thawed samples in a triaxial cell arranged for inde- pendent control of internal pore pressure. Recom~ pacted samples were frozen in cylindrical molds and then transferred to a precooled triaxial cell.. As sam- pies were thawed, the pore space was kept saturated by injection of fluid. Selected initial stress conditions were imposed, and then constant strain rate deforma- tions were conducted. Most deformation experiments 10-2 10-3 / o / ? / 'l I 0 25~F, 2:/,,% 0 13°F, 2:/~% · 25°F, 1% a 13~F, 1% O 25°F, 0% v I 3"F, 0% 10-4 , t , IIII/[ i I I I/Ill{ I 10 100 1000 10,000 MAXIMUM UNIAXIAL STRENGTH, Fig. 8~lnfluence of salt concentration on the uniaxial strength of a saturated, recompacted sample of natural Prudhoe Bay sand. 10-.2 , , , , , ,,J, 10-3 _ 10-4 10 /// // / o 25°F 2'/~% 0 13°F 2~/~% · 25°F 1% a 13°F 1% 0 25°F 0% V 13°F 0% I I I I IILl[ I I I I iiIIl I 100 1000 10,000 MAXIMUM UN[AXIAL STRENGTH, psi Fig. 9~lnfluence of salt concentration On the uniaxial strength of a saturated, recompacted sample of natural Prudhoe Bay silty sand. 10-1 10-2 7 <~ 10-3 · Z ~Y I I I II1~1 I I 111111 I I 11111'2 - _ . - -- -- - -- - - o fl°[ h 18°F ~,. o 28°F _ I I I I II1 I I I III I I I I III 10-4 10--5 10 100 1000 10,000 MAXIMUM UNIAXIAL STRENGTH, psi Fig. 7~Maximum uniaxial strength of recompacted, Prudhoe Bay silt. (Before freezing, the pore space was saturated with water containing 0.07 percent NaCI.) OCTOBER, 1974 10-4 - - I TEMP. ~ rain : · 18° ,045 _ ~ · 18° .025 · 8° .0075" - i 8~ .035 10-5 7 : lO-~ ,_ E z ~ 10-7 10-8 10-9 10 100 1000 10,000 2 Fig. 10~Creep behavior of re¢ompacted Prudhoe Bay near-surface sand. 1171 were run without removing or adding pore liquid, thus giving what is usually called undrained deformation. Internal pressures were recorded continuously with a pressure transducer of negligible displacement. The influence of the following factors has been considered (1) strain rate, (2) type of soil, (3) level of compaction, (4) stress level, (5) relationship between lateral and axial stress, and (6) amount of pore volume drainage. Vesic and Clough3~ have shown that under high stress, unconsolidated sands will cease to compact after a few hours. Clays, on the other hand, may continue to compact for longer periods of time? All the natural soils tested during this study were very sandy and highly permeable. Preliminary deforma- tions were run at several strain rates. Mechanical behavior was found to be essentially independent of strain rate over the range of low strain rates investi- gated. Subsequent studies were conducted at a con- venient strain rate of 0.00166 min.-~ Four types of soil have been studied--three natural soils described in Tables 2 and 3 and a 40- to 200-mesh pure quartz sand similar to that previ- ously studied by several investigators. Behavior of Pure Quartz Sand Saturated quartz sand samples were hydrostatically loaded while the pore pressure was kept about 10 to 15 psi lower than the external stress. The cores were then axially shortened. Data for a typical, undrained deformation are shown on Fig. 11. The difference between axial and lateral stresses was small during the first few percent strain, and the internal pressure was relatively constant. After a few percent strain, how- ever, the sand became dilatent and the internal pore pressure diminished. As the effective normal stress increased, the ability to support shear stress increased. As high shear stresses developed and strain continued, the pore pressure and the undrained shear strength of the sample approached relatively constant values. Undrained deformation behavior at various stress levels is shown on Fig. 12. The shear strength contri- bution of ice for similarly prepared samples was' reported in Ref. 6. The algebraic addition of strength contributions suggested by Eq. 5 has been tested by adding the uniaxial strength of frozen sand-ice cylin- ders to the strength values shown on Fig. 12. Fig. 13 TABLE 2--PARTICLE SIZE DISTRIBUTION OF NATURAL PRUDHOE BAY SOIL SAMPLES Description Percent Retained on U.S. Sieve Mesh Size of Samples. 20 40 80 120 200 < 200 Sand '25.6 24 42.2 2.0 1.7 4.4 Silty sand 3 5.4 39.$ 15.8 10.8 25.2 Silt 0.2 ~. 8 5.4 8.6 75.8 T 2000 1600 1200 800 400 INITIAL CONDITIONS: AXIAL STRESS : 1000 ps~ LATERAL PRESSURE : 1OO0 psi INTERNAL PRESSURE = 985 psi I I I 1 I I I I 1, I I I I I I 1000 ._ ~ a0o ~ 600 ~ 400 ~ 200 I 0 I , I I I I I I 0 .O2 .04 .06 .08 , ,10 .12.14 .16 AXIAL STRAIN Fig. II--Deformation behavi.or of thawed undrained 40- to 200-mesh sand. .18 shows that the calculated total resistance to deforma- tion agrees well with experimental values reported previously.° Behavior of Thawed Natural Soils Under Stress The effect of the initial state of stress in natural soils has been investigated. As an example of the stress range of interest, consider the in-situ stresses at Prudhoe Bay. Density logs and cores have shown that the average vertical stress is about 0.89 psi/ft. The pore fluid pressure gradient just below the permafrost is about 0.45 psi/ft, and we assume that this was the pore pressure gradient at the time of permafrost freez- ing. The total horizontal stress gradient is estimated at 0.65 psi/ft, which is typical of unconsolidated granular soils. The average total stress gradient is thus estimated to be [0.89 + 2(0;65)]/3 = 0.73 psi/ft. 'The ratio of pore pressure to average total stress is estimated to be about (0.45/0.73) ~ 0.6. Samples of natural soils were prepared by the fol- lowing procedure. 1. Recompacted samples were saturated and frozen in cylindrical molds with no confining stress. 2. Frozen samples were loaded into the triaxial cell and thawed. Liqui.d was injected at a slight pressure TABLE 3~MINERALOGY AND SALINITY OF NATURAL PRUDHOE BAY SOIL SAMPLES Salinity (percent NaCI in Weight Percent of Constituents Description of Samples Sand Silty Sand Silt formation Mixed water) Quartz Calcite Dolomite Kaolinite Illite Montmorillonite Layer Chlorite Total 0.97 99 0 0 0 0 0 0 0 99 0.15 75 15 I 0 4.4 0.7 0.8 0 99 0.07 62 23 2 0.9 3.9 0.3 3.1 2.4 98 1172 JOURNAL OF PETROLEUM TECHNOLOGY 2000 ..... 1600 1200 800 400 0 0 I I I I I { I 1 1500 ~--- ~oo_o - ~ .......... ??? ........ ,,~.~,,~."~" TOTAL LATERAL STRESS. ,~' ,02 ,04 .06 ,08 .lO ,12 ,14 .16 .18 AXIAL STRAIN Fig. 12--Behavior of thawed, undrained 40- to 200-mesh sand. Initial net compaction stress ~, 10 psi. to keep the pore space saturated during thawing. 3. External stresses were increased hydrostatically while internal pressures were maintained about 10 to 15 psi below the external stress. 4. While the external stress was maintained at a desired value, the internal pressure was cycled be- tween 60 and 99 percent of the external stress. 5. In some cases, the axial and radial stresses were varied to subject the soil to shear. During this treat- ment the average external stress was held constant and the pore pressure was held constant at 60 percent of the average external stress. 6. After the initialization treatment, the core was shortened without addition or removal of pore liquid. 3200 2800 2400 2000 1600 1200 800 400 EXPERIMENTAL DATA FROM REFERENCE 5 ~~/ ~ ~ I STRAIN RATE = 6.67 x 10-4 MIN.-1 ,~. ."; ......... .02 .04 .06 .08 .10 .12 .14 AXIAL 3200 '~ 2800 · m 2400 ~ 2000 ~ 1600 ~ 1200 < 800 4O0 m mi I m I m I " I I I ' I ~ I CALCULATED BEHAVIOR USING EQUATION5. ~ --~ ~ THAWED BEHAVIOR FROM FIGURE 12. ~'"'*-/"~ -- :--:--:-. ......... .~ ~ ~ 1500 psi TOTAL LATERAL STRESS .~.~'/' ~ ~ lOOO ........ .... ??o ......... /,,~ - -loo .... :. .. · I I I I , I I I I I I I I I ,02 .04 ,06 .08 .10 .12 ,14 AXIAL STRAIN Fig. 13~Comparison of experimental and calculated behavior of sand-ice samples under triaxial conditions. Samples frozen with zero confining stress. OCTOBER, 1974 1173 Preliminary experiments have shown that soil behavior is quite sensitive to the level of compaction. Most of the compaction effect is a:'.'b.~eved when the first cycle of compaction stress is applied. A slight additional strengthening was detected upon applica- tion of several additional cycles of compaction; we have arbitrarily subjected samples to 20 cycles of compaction. No additional effect was noted when samples were held under compacting stress for 24 hours before commencing the deformation test. Samples subjected to shear stabilization were some- what stiffer during an initial elastic deformation, but the ultimate shear strength was unaffected. Data for several initial stress levels and for three types of soil are shown on Fig. 14. Initial stress level is an impor- tant variable, but all these natural soil samples exhibited about the same behavior. In a separate set of experiments we studied the effect of varying the relationship between lateral and axial stresses. After subjecting the soil samples to the desired initial conditions, three types of experiments were run: 1. The lateral stress was held constant while the core was shortened. 2. The lateral stress was decreased by an amount equal to the increase in axial stress during shortening. 3. The lateral stress .was decreased one-half the amount of increase of the axial stress during shorten- ing. This experiment was thus conducted at constant octahedral normal stress. For fixed initial conditions, during shear failure of these soil samples the relationship between strain and (~r,~ - ~r~) was the same for each of the three experi- mental procedures described. In these undrained tests with natural soil, the behavior of internal pore pressure was different from that exhibited by pure quartz sand, illustrated on Fig. 11. During deformation of natural soils, internal pressures rose above their initial values, probably because of slipping or crushing of soil particles. Fig. 15 shows typical shear strengths and internal pres- sures for a natural silty-sand sample, a strong quartz- 600 500 400 300 200 0 0 I I I I' ' I' I '1 ''1 ...... CONSTANT LATERAL STRESS = ~ ~ psi J 150! 1000 ~/~--~ ~,.~ ~.~ ~ ..... . ~ . ' . ." 500 ~ . · · .' ,...' . .' . ......... · ........ ,... · ....... . · . .... ~ I , I I I I II I 2 4 6 8 I0 12 14 16 18 PERCENT AXIAL STRAIN Fig. 14--Shear strengths of natural Prudhoe Bay sands, silty sands, and silts. (During compaction, internal pressure was cycled 20 times between 99 and 60 percent of external stress. Undrained deformations were conducted with initial pore pressure equal to 60 percent of lateral stress.) sand sample, and a sample consisting of 95 percent sand and 5 percent clay. The increase in pore pres- sure exhibited by the natural soil and clay-sand mix- ture led to low net stresses and therefore Iow shear strengths. Since total radial pressure, total axial pressure, and pore pressure were independently measured, net soil stresses could be calculated and plotted in the form of Molar envelopes, as illustrated on Fig. 16. As the shear stress was initially increased, behavior was elastic. For the natural soil sample, when the net shear stress approached about 200 psi, particle failure or slippage occurred, bringing the net stress onto the envelope of shear failure. Similar experiments with strong quartz sand showed elastic behavior, which merged into shear failure as the envelope of shear failure was reached at much higher stress levels. During subsequent strain the Mohr envelope could be determined over a broad range by varying the lateral stress level. In the few experiments involving them, the samples prepared from 95 percent sand and 5 percent clay behaved much like the natural soil samples. In a separate set of experiments with natural soils, the pore pressure was held constant during deforma- tion by removing pore liquid. During shear failure, the apparent angle of internal friction increased with strain, reaching 22° at 5 percent strain and increas- ing to 28° at 10 percent strain. The difference between axial and lateral stresses was increased above those values observed for undrained deformations, 1200 c~ 1000 ~ a: 800 x c~ 600 _.1 ~ .-J 400 0 ~' 200 o 50o 400 300 200 100 I I I l I I I I I I I -'' NATURAL SILTY SAND o, ,,I I I I I I 0 2 4 6 8 I0 12 14 PERCENT AXIAL STRAIN Fig. 15--Effect of mineralogy on shear strength. All samples precompacted with 20 cycles. Total lateral stress held constant at 500 psi. 1174 JOURNAL OF PETROLEUM TECHNOLOGY thus confirming that the amount of pore drainage is an important variable. Elastic Behavior The initial behavior of well compacted soil samples appeared to be elastic. Experiments at constant lat- eral stress permit the direct calculation of Young's modulus. ~xa~ (1 + K3~), . . (7) where E= p.-- Aaa -- Ks -- Ks - Ks - Young's modulus, psi, Poisson's ratio, change in axial stress, psi, axial strain, 0 if Aa, = 0, 1 if ZXa: = Aa~/2, and 2 if Aa: = - Aa~. Moduli measured at constant lateral stress were less than 3 × 105 psi but were a function of initial mean net stress, as has been reported by other inves- tigators.~ Measured values of Young's moduli are shown on Fig. 17. The volumetric compressibility of ices3 is about 7.65 >< 10-7 psi-~ and that of soila4 constituents is about 6 >< 10-8 psi-~. The elastic modulus, Poisson's ratio, and average volumetric compressibility are re- lated by Eq. 8. 3(1 - 2~) (8) Substituting measured values of the modulus and calculated average compressibilities into Eq. 8 shows that Poisson's ratio for undrained deformations will be essentially 0.5. If fluid is expelled from the sample during deformation, then one should observe the lower values of Poisson's ratio often mentioned for rocks and soils. Deformations of thawed, undrained samples conducted with variable lateral stress have verified the consistency of Eq. 7 when using a value of Poisson's ratio of 0.5. If a free gas phase is present, lower values of Poisson's ratio would be expected. Behavior of Frozen Samples Under Triaxial Conditions The triaxial cell was operated in a different manner so that samples could be frozen under stress. After thawed samples were subjected to the desired initial conditions they were, while under stress, frozen from the bottom. During freezing the internal pressure was maintained at 60 percent of the external stress by removing excess liquid at the top through a heated tube. Experimental results for several conditions are shown on Fig. 18; the behavior calculated with Eq. 5 agrees well with the experimental data. Conclusions 1. The strength of permafrost can be partitioned 600 ', , , I , ' ' I , ' , I ' ' ' I , , ' ' a00 0 /~l I I ~ I ~ I I I I I [ ~ ~ I [ ~ I I I 0 200 400 600 800 1000 .... Fig. 16--Mohr envelopes for various thawed soils. 106 ~ 105 z I I I * ® PRUDHOE SAND 'X' PRUDHOE SILTY SAND ~ PRUDHOE SILT '~' PRUDHOE SOIL BEHAVIOR AFTER SHEAR STABILIZATION 104 I, i i i i i i i 00 1000 INITIAL MEAN NET STRESS, psi Fig, 17--Young's modulus as a function of initial mean net stress. Samples compacted under initial mean net stress with 20 cycles before commencing axial shortening. lOOO 800 600 400 200 ,,"~----~)'OTAL LATERAL STRESS -- OTAL LATERAL STRESS -- T 500 ps, _ ~ EXPERIMENTAL BEHAVIOR -- CALCULATED WITH ~'A UNIAXIA~ ~lHAWED SI~TY SANO A~ 500 ps~ EQUATION 5 L [] UNIAXIA~ '[HAWED SI~TY SAN(} A~ I000 psi 1 2 3 4 5 6 7 8 PERCENT AXIAL STRAIN Fig. 18--Comparison of yield behavior of frozen silty sand with summation of uniaxial and thawed behavior, OCTOBER, 1974 1175 into a resistance to deformation of ice in the presence of natural soil constituents, which is time dependent, and a resistance to deformation of soil constituents under confining stress, which is essentially indepen- dent of time. 2. The fraction of water in the ice phase is in- fluenced by the temperature, pressure, electrolyte concentration, mineralogy, and state of clay hydration. 3. The shear strength of the ice in permafrost is influenced by the same variables mentioned above and is also a function of the strain rate. 4. For natural Prudhoe Bay soils, the linear rela- tionship between the logarithm of the maximum uniaxial permafrost strength and the logarithm of the strain rate is valid to strain rates as low as 3 X 10-0 min. - ~. 5. The contribution of the soil matrix to the perma- frost shear strength can be measured with thawed samples, provided the soil is brought to the proper initial state and provided the proper boundary condi- tions are imposed during deformation. 6. Soil strength is insensitive to strain rate but is significantly influenced by type of soil, compaction and stress level, and condition of pore volume drainage. 7. The initial mechanical behavior of compacted samples is elastic. 8. The presence of clays or weak constituents in a natural soil limits the net shear stress that can be sustained by the natural soil during deformation. Nomenclature c = volumetric compressibility, psi- ~ E = Young's modulus, psi K~ = empirical coefficient, Eq, 6 K2 = empirical coefficient, Eq. 6 Ks = coefficient with discrete values 0, 1, 2, Eq. 7 , = strain ~,ln = strain at which a minimum strain rate is achieved at essentially constant axial stress · , = strain rate, min.-~ tx = stress, psi , = Poisson's ratio ~r,~ -- axial stress acting through the ice phase, psi ,~,~ = axial stress acting through multicrystal- line ice, psi ~r~ = net axial stress acting through the soil, psi ,,~ = total axial stress, psi ~r~, -- lateral stress acting through the ice phase, psi ~r~ = lateral stress acting through multicrystal- line ice, psi Original manuscript received in Society of Petroleum Engineers office Aug. 6, 1973. Revised manuscript received June 21, 1974. Paper (SPE 4587) was first presented at the SPE-AIME 48th Annual Fall Meeting, held in Las Vegas, Ney., Sept. 30-Oct. 3, 1973. © Copyright 1974 American Institute of Mining, Metal- lurgical, and Petroleum Engineers, Inc. This paper will be printed in Transactions volume 257, which will cover 1974. {rz~ -- net lateral stress acting through the soil, psi ,z~ -- total lateral stress, psi tr~ = uniaxial strength of multicrystalline ice, psi ~r,,~, -- uniaxial strength of a permafrost sample, psi References 1. Howell, E. P., Perkins, T. K., and Seth, M. S.: "Cal- culating Temperatures for Permafrost Completions," Pet. Eng. (April 1973) 69. 2. Couch, E. J., and Keller, H. H.: "Permafrost Thawing Around Producing Oil Wells," ~. Cdn. Pet. Tech. (1970) 9, 107. 3. Eickmeicr, J. R., Ersoy, D., and Ramey, H. J., Jr.: "Well- bore Temperature and Heat Losses During Production or Injection Operations," d. Cdn. Pet. Tech. (1970) 9, 115. 4. Perkins, T. K., Rochon, ~. A., and Knowles, C. R.: "Studies of Pressures Generated Upon Refreezing of Thawed Permafrost Around a Wellborc," J. Pet. Tech. (Oct. 1974) 1159-1166; Trans., AIME 257. 5. Goodman, M. A., and Wood, D. B.: "A Mechanical Model for Permafrost Freezeback Pressure Behavior," paper SPE 4589 presented at the SPE~AIME 48th Annual Fall Meeting, Las Vegas, Ney., Sept. 30-Oct. 3, 1973. 6. Perkins, T. K., and Ruedrich, R. A.: "The Mechanical Behavior of Synthetic Permafrost," Soc. Pet. Eng. J. (Aug. 1973) 211-220; Trans., AIME, 255. 7. Ladany, Branko: "An Engineering Theory of Creep of Frozen Soils," Cdn. Geotech. J. (1972) 9, 63. 8. Kiirfurst, P. J., and King, M. S.: "Static and Dynamic Elastic Properties of Two Sandstones at Permafrost Temperatures," J. Pet. Tech. (April 1972) 495-504; Trans., AIME, 253. 9. Akili, Waddah: "Stress-Strain Behavior of Frozen Fine- Grained Soils," National Research Council-National Academy of Sciences Highway Research Record 360 (1971) 1-8. 10. Zakharov, V. A.: ".Experimental Investigations of the Mechanical Characteristics of Frozen Ground at Differ- ent Rates of Loading," Osnoyaniya, Fundamenty i Mek- hanika Gruntov (Sept.-Oct. 1970) No. 5, 10-11. l l. Andersland, O. B., and A1Nouri, Ilham: "Time- Dependent Strength Behavior of Frozen Soils," J. Soil Mech. and Found. Div., ASCE (July 1970) 1249. 12. Goughnour, R. R., and Andersland, O. B,: "Mechani- cal Properties of a Sand-Ice System," J. Soil Mech. and Found. Div., ASCE (July 1968) 923. 13. Andersland, O. B., and Akili, Waddah: "Stress Effect on Creep Rates of a Frozen Clay Soil," Geotechnique (1967) 17, 27-39. 14. Lovell, C. W., Jr.: "Temperature Effects on Phase Composition and Strength of Partially-Frozen Soil," National Research Council-National Academy of Sci- ences Highway Research Bd. Bull. 168 (1947) 74-95. 15. Chamberlain, E., Groves, C., and Perham, R.: "The Mechanical Behavior of Frozen Earth Materials Under High-Pressure Triaxial Test Conditions," Geotechnique (1972) 22, No. 3, 469-483. 16. Ice and Snow Properties, Processes, and .4pplications, W. D. Kingery, Ed., M.I.T. Press, Cambridge, Mass. (1963) 4. 17. Anderson, D. M., and Tice, A. R.: "Low Temperature Phases of Interfacial Water in Clay-Water Systems," CRREL, Research Report 290 (Oct. 1970). 18. Anderson, D. M.: "The Interface Between Ice and Silicate Surfaces," CRREL Research Report 219 (March 1967). 19. Low, P. F., Anderson, D. M., and Hoekstra, P.: "Some Thermodynamic Relationships for Soils at or Below the Freezing Point ~ 1: Freezing Point Depression and Heat Capacity," CRREL Research Report 222 (Dec. 1966). 20. Anderson, D. M., and Hoekstra, P.: "Migration and Crystallization of Interlamellar Water During Freezing and Thawing of Wyoming Bentonite," CRREL Re- 1176 JOURNAL OF PETROLEUM TECHNOLOGY search Report 192 (Dec., 1965). 21. Dillon, H. B., and Andersland, O. B.: "Predicting Un- frozen Water Contents in Frozen Soils," Cdn. Geotech. J. (May 1966) III, No. 2, 53-60. 22. Antoniou, A. A.: "Phase Transformations of Water in Porous Glass," J. Phys. Chem. (1964) 68, No. 10, 2754-2764. 23. Nerseova, Z. A., and Tsytovich, N. A.: "Unfrozen Water in Frozen Soils," NAS-NRC Publ. 1287, Proc., 1st Intl. Conf. on Permafrost, Moscow (1963). 24. Litvan, G. C.: "The Freezing of Water and Xenon in Porous Vycor Glass," PhD dissertation, U. of Toronto (1962). 25. Williams, P. J.: "Specific Heats and Unfrozen Water Content of Frozen Soils," Proc., 1st Canadian Con- ference on Permafrost, Ottawa, April 17-18, 1962; NRC Tech. Memo 76, 109-126. 26. Buehrer, T. F., and Aldrich, D. G., Jr.: "Studies in Soil Structure, VI. Water Bound by Individual Soil Constituents as Influenced by Puddling," Tech. Bull. 110, U. of Arizona Agricultural Experiment Station, Tucson (June 1946). 27. Buehrer, T. F., and Rose, M. S.: "Students in Soil Structure, V. Bound Water in Normal and Puddled Soils," Tech. Bull. lO0, U. of Arizona Agricultural Experiment Station, Tucson (June 20, 1943). 28. Bouyoucos, G. J., and McCool, M. M.: "The Freezing Point Method as a New Means of Measuring the Concentration of the Soil Solution Directly in the Soil," Tech. Bull. 24, Michigan Agricultural College Experiment Station, East Lansing (Dec. 1915). 29. Martynov, G. A.: "The Calorimetric Method of De- termining the Quantity of Unfrozen Water in Frozen Soil," Canada National Research Council Technical Translation 1088, 143-148. 30. Morioka, P., Kobayashi, J., and Higuchi, I.: "Freezing of the Capillary Liquid Condensing in Fine Pores," J. Colloid and Interface Science (Jan. 1973) 42, No. 1, 156-164. 31. Vesic, A. S., and Clough, G. W.: "Behavior of Granular Materials Under High Stresses," J. Soil Mech. and Found. Div., ASCE (May 1968) 661. 32. Mitchell, J. K., Campella, R. G., and Singh, Awtar: "Soil Creep as a Rate Process," J. Soil Mech. and Found. Div., ASCE (Jan. 1968) 231. 33. Gow, A. J., and Williamson, T. C.: "Linear Com- pressibility of Ice," J. Geophys. Research (Nov. 1972) No. 32, 77, 6348-6352. 34. Cleary, J. M.: "Hydraulic Fracturing Theory, Part III, Elastic Properties of Sandstone," Circular 281, Illinois State Geological Survey, Urbana (1959). OCTOBER, 1974 11'/7 Studies of Pressures Generated Upon Refreezing of Thawed Perrnafrost Around a Wellbore T. K. Perkins, SPE-AIME, Atlantic Richfield Co. J. A. Rochon, SPE-AIME, Atlantic Richfield Co. C. R. Knowles, SPE-AIME, Atlantic Richfield Co. Introduction Since the discovery of oil at Pmdhoe Bay in 1968, much new technology has been developed for dealing with oil production in arctic regions. Part of this new technology has been the development of successful methods for drilling and completing wells through permafrost. Several previously published papers~-3 have dealt with the thermal aspects of drilling or producing warm oil through frozen soil, and in those papers it has been pointed out that some thawing around the wellbore is generally expected. If a well is allowed to refreeze, casing damage may result from two mechanisms. First, fluids confined within pipes and allowed to freeze may cause pressure to rise, dam.- aging casing2 Second, refreezing of thawed perma- frost or fluids external to the casing may develop pressures high enough to cause casing damage,s, 0 This paper considers only pressures resulting from refreez- ing of permafrost and fluids external to the casing. As ice melts, its volume decreases by about 9 per- cent. If the volume .decreases within the earth, there is a tendency for pressure within the liquid phase to be low. This tendency is offset by movement of fluid into the thawed region. We envision five possible sources of liquid that could flow into thawed regions: (1) drilling fluid filtrate, (2) water from below the permafrost, (3) water from near the surface, (4) brine moving laterally through permeable material into the thawed region, and (5) liquid within the thawed re- gion that becomes rearranged by gravity flow to re- saturate deep thawed regions. If a thawed region is saturated with water and then allowed to refreeze, the system volume increases as water is converted into ice. Since permafrost refreezes most rapidly near the surface, excess water may be trapped when deeper thawed.regions refreeze. Pres- sures rise, thus forcing liquid water to flow away through permeable material; if this is not possible, pressures rise until the excess volume can be accom- modated in some other manner. To insure the use of casing with the proper collapse strength, a number of studies of refreezing pressures have been under- taken. 1. Large-scale field tests have been conducted in full-size wellbores penetrating the permafrost at Prud- hoe Bay. 2. The mechanical properties of permafrost have been studied. 3. Pressures generated during refreezing have been measured in laboratory models. 4. Theoretical methods for calculating stresses and pressures have been developed. In the remainder of this paper we shall report the results of those studies. We shall show a favorable comparison between experimentally measured pres- sures and calculated pressures. Finally, we shall give calculated pressures for some field cases of interest. Field Tests Two full-size wellbores were drilled and completed through the permafrost at Prudhoe Bay. Thermistors Studies of heat transfer and of the mechanical properties of permafrost and an analysis of stresses around refreezing wellbores have been combined to develop a computer program that calculates refreezing pressures. Calculated values agree well with values determined in a large-scale field test. OCTOBER, 1974 1159 TABLE 1--DRILLING FLUID PROPERTIES Property Drill Site 4-6 Well Drill Site 1-6 Well Type of fluid Water-base mud Oil-base mud Weight 9.5 lb/gal 9.9 lb/gal Viscosity 250-sec Marsh funnel 72-cp plastic 32-cp plastic Yield point 26 lb/100 sq ft 40 lb/100 sq ft Gels 20 lb/100 sq ft initial 21 lb/100 sq ft initial 127 lb/100 sq ft 27 lb/100 sq ft 10 minutes 10 minutes Filtrate 3.2 equivalent parts per million (epm) calcium ion Alkalinity 0,15 P alkalinity of the filtrate (Pf) Water 98 percent 9 percent Oil 78 percent Solids 2 percent 13 percent and pressure transducers were attached to cables clamped external to the casing. Hot fluid was circu- lated through tubing in the wellbore to simulate the remainder of the drilling operation and later to simu- late a short production period. During the thawing and gubsequent refreezing cycles, temperatures and pressures were recorded at the surface. Description of the Wells The Drill Site 4-6 well was drilled with a fresh-water mud having the properties shown in Table 1. Con- ductor pipe (20 in.) was set at 109 RKB. A 17½-in. hole was drilled to 2,700 ft, and 13~-in. 72-1b N-80 Modified Buttress threaded pipe was run. Cables were attached to the casing in a manner that would place the bottom thermistor and pressure transducer at the desired depth. The three cables, one containing thermistors and two with pressure transducers, were banded to the casing with aA-in.-wide stainless steel straps. A conventional 4-bow centralizer was run on every collar crossed by the cable. The casing was cemented with the shoe at 2,191 ft. The first stage of cement was 700 sacks of Perma- frost II. The top of this stage was defined by cement bond log at 1,580 ft. The second stage was cemented with 400 sacks through a DV tool at 486 ft. The top of the cement was estimated to be at 400 ft. After cementing, freezable fluid in the 20 in. X 13% in. annulus was displaced with an oil-base casing pack. This displacement was achieved through a pair of 23~-in. risers--external to the 20-in. casing M welded into the casing at the bottom. A caliper sur- vey and final position of down-hole instrumentation is shown on Fig. 1. Drill Site 4-6 was completed with open-ended 3½-in. 9.2-1b N-80 Buttress threaded tubing hung at 2,044 ft. A second well, Drill Site 1-6, was drilled with oil- base mud having the typical properties shown in Table 1. A 20-in. conductor was set at 107 RKB, and the hole was drilled to 2,750 ft with 18½-in. bits. The casing string and cable attaching methods were similar to those of Drill Site 4-6. The casing shoe was landed at 2,698 ft, with the final position of down-hole in- strumentation as shown on Fig. 2. The 133~-in. casing was cemented with 1,350 sacks of Permafrost II cement in one stage. The top of the cement was esti- mated to be at 1,780 ft. The well was completed with perforated tubing hung at 2,544 ft. Instrumentation The use of thermistors to measure temperatures in permafrost has been reported previously in the litera- ture.7-* The temperature sensors used in this study were semiconductor thermistors having a range from - 38° to 237°F. The accuracy of the readout system was ± 0.5°F, with a reproducibility of 0.1 percent of the scale reading. The field testing being described here extended down-hole instrumentation to include measurement of pressure as well as temperature in the annulus be- tween open hole and casing. Pressures were measured with low- displacement, strain- gauge - actuated trans- ducers with an operating temperature range of 0° to 200°F. Accuracy was within ± 1 percent of full-scale output, including effects of nonlinearity, hysteresis, and repeatability. Circulation System The circulation systems for the two wells (Fig. 3) con- sisted of a direct-fired oil heater, circulating pumps, high-low temperature controls, fluid-rate meters, and temperature and pressure recorders for the circulating fluid discharge and intake. The nonfreezing circulat- ing fluid consisted of a 50-percent mixture of glycol and water. The capacities of the heaters were 4 ua '~ ua ua w ~ 13 3/8" ~ o ~ CASING 96 57 196 157 296 257 396 357 546 5O7 571 532 596 557 621 582 646 607 746 707 771 732 796 757 821 782 896 857 1046 1007 1296 1257 lO 11 12, 13- 14- 15- 16 THERMISTOR ONLY 0IT SIZE {lRVz } HOLE CALIPER 20" 30/~ 40" 109~ Fig. 1--Caliper log and down-hole instrument location for Drill Site 4-6. TOP OF CMT, IDV ~ 486; 1160 JOURNAL OF PETROLEUM TECHNOLOGY MMBtu/hr on Drill Site (D. S.) 4-6 and 2 MMBtu/hr on Drill Site (D. S.) 1-6; however, approximately 1 to 1½ MMBtu/hr of heat was actually required, well below the capacities of the heaters. The 3 ½-in. tubing hung below the base of the permafrost was used as the circulating string. The base of the permafrost at these drillsites is approximately 1,900 ft. Description of the Heating Cycles This study has included tWo complete thaw-and- frcezeback cycles for each well. The first heating cycle introduced an amount of heat similar to that expected during the operation of drilling and completing a development well in the Prudhoe Bay field. The sec- ond heat cycle was similar to what might be exPected during a short production period for a normal pro- ducing well. The first heat cycle on D. S. 4-6 consisted of 7 days of circulating at 12.4 bbl/min with 113°F inlet temperature and 110°F return temperature. This was the only circulating cycle in which the drilling rig pumps were used. All subsequent circulating was done with the permanent circulating system previously described. The second heat cycle on D. S. 4-6 lasted 39 days, with a'circulating rate of 2 bbl/min at 165° to 175°F inlet temperature and 130° to 140°F return temperature. The first heat cycle on D. S. 1-6 was accomplished by reverse circulation because the tubing perforations had become plugged with drilling mud that had been left in the 133A-in. casing when the sYstem was first changed over to the glycol-water mixture. The well was circulated for 7 days at 3' bbl/min with inlet and return temperatures of 130° to 150°F and 115° to 140°F, respectively. The second heat cycle on D. S. 1-6 consisted of forward circulation at 3 bbl/min for 11 days with an inlet temperature of 150° to 160°F and a discharge temperature of 125° to 135°F. Temperature History Wellbore temperature profiles at various times after heating are shown for both cycles of D. S. 4-6 and D. S. 1-6 on Figs. 4 and 5. As indicated on the figures, the temperatures throughout the entire monitored sec- tion of the well dropped to a freezing level within a few days after circulation had been stopped. The depth to which "freezeback" ocCurred is indicated by a break in the temperature profile. The freezeback depth progressed down the hole as water in the forma- tion and, in well D. S. 4-6, the drilling mud in the an- nulus outside the wellbore became completely frozen. The freezing point of water within the formation de- pends on salinity, pressure, lithology, and mineralogy. During the first part of the test, temperatures were obtained from the externally mounted thermistors. Later in the test many of the thermistors failed, and additional temperature surveys were obtained in both wells with a precision, wireline temperature instru- ment run inside the 133A-in. casing. Experimentally measured freezeback times were in excellent agree- ment with values calculated with computer programs.~ Pressure History The pressure history for the two wells is shown on 129 234 337 547 652 758 862 968 1078 1128 30 108 203 308 410 507 612 710 810 909 1006 1056 11 1109 12 1206 BIT SIZE (18 ~#) 13 3/8~ [ HOLE CALIPER CASING}I 20# 30# 40t ] 107t 50// I Fig. 2--Caliper log and down-hole instrument location for Drill Site 1-6'. DRILL SITE 1 - 6 w TEMPERATURE RECORDER (T~n, T out) . PRESSURE RECORDER (P in, P out) F LUID RATE METER ' 13 3/8" CASING @ 2698~ -------..~. WO CENTRIFUGAL PUMPS ITH 20 HP 3500 RPM* LECTRIC MOTORS 3 1/2" TUBING L~ 2544~ ------.--- 2 MMBTU/HR DIRECT FIRED OIL HEATER 't DRILL SITE 4 - 6 wTEMPERATURE RECORDER (T in, T out) PRESSURE RECORDER (P in, P out) FLUID RATE METER WO PD SCREW PUMPS ITH 15 HP 1750 RPM ELECTRIC MOTORS 4 MMBTU/HR DIRECT FIRED OIL HEATER 13 3/8" CASING @ 3 1/2" TUBING Ill I*1 @ 2044t ~_~ Fig. 3--Circulation systems. OCTOBER, 1974 ! 161 Figs. 6 and 7. Because the close groupings of several of the transducers gave very similar pressures, not all the transducer pressures are shown. Initially, the transducers indicated a pressure gradient equal to the hydrostatic gradient of the fluid outside the 133~-in. casing. However, the external pressures during tile heating cycles fell below a fluid gradient pressure, indicating a drop in the fluid level in tile annulus. We confirmed the drop in fluid level by refilling the annulus. The nonlinear behavior of the pressures with depth later during the heating cycles indicated bar- tiers to vertical communication outside the casing of both wells. Such behavior is ~nore pronounced during the second heating cycle because of additional slough- ing that has taken place during the first cycle. For example, as Fig. 6 shows, during the second heat cycle on D. S. 4-6 all the deeper transducers indicated pres- Sures of less than 200 psi, with the exception of the transducer at 571 ft. Similar conditions are shown during the second heat cycle in D. S. 1-6 where the deeper transducer pressures dropped below 200 psi. The 'shallower transducer pressures in both wells re- mained at or near a fluid gradient because the annulus was filled periodically with water at the surface during the heating cycles. Immediately after each heat Cycle was discontinued, pressures began to increase. The pressures in Well D. S. 4-6 increased at each transducer level as the temperature .dropped to the freezing level and con- tinued to increase until the temperature 'fell below the freezing point at that particular transducer level. The independence of each transducer in Well D. S. 4-6 is base mud left in the annulus sequentially froze across each of the transducers, preventing additional vertical transfer of the pressure in the annulus. The pressure behavior differed in Well D. S. 1-6 be- cause oil-base mud had been left in the annulus during the completion of this well. The oil-base mud, which does not freeze or solidify, allows the vertical transfer of pressure in the annulus. This explains groups of transducers reaching a maximum pressure at tile same time, independent of the temperature level at each of the transducers. This is illustrated on Fig. 7 in Cycle 1, where maximum pressures were reached simul- taneously at transducer depths of 30, 129, and 337 ft. Also, a group of transducers at 547, 652, and 968 ft reached maximum pressures simultaneously. This would indicate vertical barriers caused by sloughing in the annulus at 450 ft 4- and 1,000 ft ±. The same group of transducers is also apparent during the sec- ond cycle. The locations of these two barriers and additional minor barriers were indicated by bonding on a cement bond log run before and after the second heating Cycle. Caliper Surveys and Cement Bond Logs The tubing Was pulled after each circulating period to allow a 4-arm caliper to be run in the 133~-in. casing. The caliper surveys run at the end of both freezeback periods on both wells indicated no casing collapse or change of any kind in the internal diam- eter of' the casing. The cement bond log (CBL) run on D. S. 1-6 indicated "bonding" from 370 to 460 ft, m , ,,~ 4,~us~ro~e,., on w;'-,,~.'~% which sao.-s]-' that the maximum ...... ,: ........ , .............. pressure at each transducer depth occurred as the .... ~ .... , temperature dropped below freezing regardless of the .... ~ . ,,,' , . ...... ~~, '.,., ~96's~1,STOPPEDj ~~ClgCUtAIlOn pressure level of the adjacent transducers. The water- . iooo g ioo · 571' ' 571' .......... Ol, [ -'~ ' '1 " [ ' [I I I---*TT~ ....... F--~ ' s4s' ." ,. 400 15 ~5 DAY* AF TlR ClRCULATIO i ~ DA SI DAYSSTOPP[D ~ / ?0 DAYS ~rT[R ,/ '" ~'' '" ~ ', ~ CIRCULATION ~ ............ ',, : Fig. ~External pressures measured on Drill .... '", Si~e 4-6; all depths RKD. / t* ~o ~ ~o a~ s0 ~ ~0 ~ ~0 ~ ~o 1400 ~ BEGAN ClRCULAIION 9.16 Fig. ~Drill Site 4-6, temperature vs depth, g STOPPED CIRCULATION 9.27 , . .,~ 1128F y ~ ., .. ~ '/~ ~'~'~ ..... '. Q" 1078 o . I I I I -- I / · I I I I I / lO00 ~ ~ I~t THAW CYCL[ 6 2 ~ 7 4~O YS0 S ( c CULATIO ,, '- ..........,, ; ..... '/,o74,' ...... ~ eoo ' ............... [4/ / /' C nCULAT on I ,oo 400~ ~~ i",': i':;- 0/~ i , , i ~1 ~ ~ , ~ , ~ i ~ ~ i ~ ~ i ~ ~ ~oo JUN~ JULY AUG, $EPI. OCT, NOV, DEC.] JAN, ~[~, ~-- I I I ~q?~ 1973 t5 20 25 30 35 40 tS 20 25 30 35 40 ............ ' ............. " Fig. 7~External pressures measured on Fig. 5~Dfill site 1-6, temperature vs depth. Drill Site 1-6; all depths RKB. 1162 JOURNAL OF PETROLEUM TECHNOLOGY 600 to 615 ft, 630 to 650 ft, 1,020 to 1,030 ft, 1,085 to 1,095 ft, 1,160 to 1,190 ft, and 1,230 to 1,250 ft. The bonding was much more pronounced on the CBL run after the frcezeback compared with one run im- mediately after the heat cycle. The indicated "bond- ins" verifies that the formation sloughed in against the pipe in the cement-free annulus. Only one CBL was run in D. S. 4-6; this was near the end of the second freezeback pcriod~ The CBL indicated good bonding from the surface to 940 ft and essentially no bonding below that point in the monitored interval. The 940-ft depth was the depth to which the well was frozen back at the time the survey was run. Laboratory and Theoretical Studies Laboratory 'and theoretical studies have consisted of the following: (1) studies of refreezing pressures in physical models, (2) studies of the mechanical proper- ties of permafrost, (3) estimates of earth stresses, (4) analysis of streSses around a refreezing well,~gnd (5) development of a computer program to calculate re- freezing pressures. Pressures Generated in Physical Models Pressures generated during refreezing in physical models have been measured in order to develop con- cepts, and data that coUld be used in testing a com- pUter simulation program. Fig. 8 illustrates th~ ap- paratus. To prepare the models, a wellbore filled with oil was. suspended in a cylindrical mold that was then filled With sand and. water. The model was frozen from the bottom under no confining Stress,' and excess water was permitted to escape at the top. The frozen model was removed from the mold and fitted with metal plates at top and bottom and coVered With a rubber sleeve. The entire assembly was then placed in a pre- cooled pressure vessel. The wellbore consisted of con- centric tubes and a wire-wound heater at the bottom. The heater was perforated at one level to permit fluid flow between the model and the interior of the well- bore. Several methods for resaturating the thawed region were studied. A reliable method was to circu- late 34°F water at a high rate during the time that the wellbore heater was on. The external pressure and the pressure at the perforation were set equal so that no compacting stress was imposed during the thaw phase. Thermocouples embedded in the model per- mitted monitoring of the thawed radius. When the desired thaw radius had been created, the wellbore heater was turned off, water was circulated out of the wellbore with carbon tetrachloride, the wellbore valves were shut, and the desired external pressure was ap- plied. During refreezing, pressures were monitored in the central tUbe and annulus of the wellbore with ex- ternally mounted pressure transducers. Models of two sizes were studied. Small models 4½ in. in diameter were prepared with 1-in. thaw radii and were studied at temperatures from 18° to 27°F and external pres- sures of 100 to 1,700 psi. Larger models 12 in. in diameter were prepared with 2-in. thaw radii and studied at temperatures from 19° to 22°F and with external pressure from 500 to 1,500 psi. Fig. 9 illus- trates two tYpical runs in the large model: After th~tw- ins, internal pressures responded immediately upon the application of high external pressures. During re- freezing, internal pressures rose, with some evidence of slight pressure breakbacks. A thermoCouple in the perforations confirmed that maximum pressure was reached at the moment of comPlete refreezing. The compressibility of material in the thawed region is small, and these models, together with a theoretical analysis, have shown that pressureS will rise until the frozen soil can be pushed back at a rate nearly equal to the rate at which excess volume is being created by the refreezing process. In these small models there was no outer elastic region to limit outward mo.ye, ment of frozen .material. Thus, continUed creep after refreezing permitted stresses to fall and eventually approach the external stress. Fig. 9 also shows be- havior calculated With a computer program (described later in the paper) based on the mechanical and thermal behavior of the sand-ice mixture. --~~ PRESSURE TRANSDUCERS 12~-------P,PE FORTNERMOOOU'LE ! l; ! :I~ Top METAL PLATE PRESSURE VESSEL~ RUBBER SLEEVE D BOTIOM METAL PLAT[~ Fig. 8~Laboratory apparatus for studying thawing and refreezing around a wellbore. 2500 2000 1500 PRESSURE APPLIED TO THE OUTER ,,~ BOUNDARY OF THE MODEL ~r ~-~.-EXTERNAL PRESSURE = / I~ 1500 PSI 'CALCULATED 500 - - HEATER Oi 50 100 150 200 250 ELAPSED TIME, MINUTES Fig, 9--Pressures induced during refreezing in a 12-in.- diameter laboratory model. Temperature of the outer boundary of the model was held constant at 21° to 22°F. Initial thaw radius, 2 in. 300 OCTOBER, 1974 ! 163 RADIUS Fig. lO---Stress 'regions mar may be present during the refreezing process. W W W 32 ,., ..-, .;,: ,', { i ';,V.,'_~ THAWED REGION FROZEN REGIONI RADIUS THE FROZEN REGION IS DIVIDED INTO ANNULAR ELEMENTS EACH OF WHICH IS SUBJECTED TO A TOTAL RADIAL STRESS Sr AND A TOTAL TANGENTIAL STRESS Se. Fig. 11--Sketch of an idealized, radially symmetrical wellbore system. 3000 I I I TOTAL HORIZONTAL STRESS = 1065 psi IN SITU PERMAFROST TEMPERATURE = 27,2"F MiXIMUM THAW RADIUS 'T,, 2000 ~ ¢~ LIMIT OF ELASTIC BEHAVIOR I l 1 ~ RADIAL STRESS lO00 ~ MUD PERMAF OST ' TANGENTIAL STRESS INTERFACE o[ 10 20 30 RADIUS, FEET Fig. 12--A typical stress distribution calculated for the time at which refreezing is complete. Mechanical Behavior of Frozen Soils Studies of the mechanical behavior of frozen soils have been reported elsewhere.9, ~o Briefly, it has been determined that the resistance of a frozen soil to de- formation results from the resistance of the ice to deformation in the presence of the porous medium pIus the elastic and flow resistance of the soil matrix. The resistance of ice to deformation is a function of temperature, strain, strain rate, and electrolyte con- centration, and is influenced by the interaction of water with soil constituents. The soil behavior is in- fluenced by the type of soil, compaction and stress level, and amount of pore volume drainage. For com- pacted samples, initial behavior is elastic, but shear strength is limited by the strength and internal fric- tion of soil constituents. Estimates of Earth Stresses Density logs and cores from Prudhoe Bay permafrost have shown that the average vertical stress is about 0.89 psi/ft. The pore fluid pressure gradient just below the permafrost is' about 0.45 psi/ft. A total horizontal stress gradient of 0.65 psi/ft, which is typical of unconsolidated granular' soils, has been estimated. Except for a few hundred feet near the surface where alternate deposition and freezing may have occurred, we believe that most of the perma- frost was deposited, compacted under overburden stress, and frozen with pore pressures about equal to a head of water. There is no evidence of overcom- paction by glaciers. Analysis of Stresses Around a Refreezing Well and Development of a Computer Program Fig. 10 shows several stress regions that xnay be pres- ent during the refreezing process. Farthest from the wellbore is a region never thawed and characterized by elastic behavior. If refreezing pressures are suffi- ciently high, there will also be a region never thawed but brought to a point of nonelastic behavior. Closer to the wellbore will be the thawed region, which may be partially or completely refrozen. Fluid pressures within the thawed region will be high enough to force the frozen soil back at a rate nearly equal to the rate at which excess volume is being created by the re- freezing process. The net soil stress can be defined as the difference between the total stress and the fluid pressure. Since the fluid pressure in the thawed re- gion is essentially equal to the total lateral stress, the previously thawed region that is refrozen will have been refrozen under Iow net confining stress. Its me- chanical properties will therefore be different from those in regions never thawed. A computer program that is valid for a radially symmetrical system has been developed for calculat- ing stresses and pressures around a refreezing well. Thermal behavior is insensitive to stress or pressure level created during refreezing, and consequentlY the temperature distribution at any stage of refreezing can be calculated directly by numerical methods2 Fig. 11 illustrates 'an idealized temperature distribu- tion around a refreezing well. For purposes of sinm- lating stress, the frozen region is divided into annular 1164 JOURNAL OF PETROLEUM TECHNOLOGY 0091 ~- x, ~ ON¥ 1 8]qDAD ~ _ S~]3nosN~I N]]3AI]9 N OIIO¥1:t ~ 1NI -- I~ ~ ~ 39DAD V I gnDA3 I '9-1 elis lipa Joj saJns$oJd pe~elnoleo pue peJnseom wnw!xelAl--lrI '~!=l ~$d '~NIZ33UJ3~I ~)NIU[IO 031¥~3N39 3~t~$8]~td 03~trIS¥]~1 0091 00~! 00gl 0001 009 009 00¢ 00g 0 OOt~l i i i 0081 · · '9-$' a~!S IlPO Jo; soJnsso~d po3elnoleO pue pa~nseom wnw!xelAl--£I 0001 OOt~l OOgl 000l 008 009 00~ OOJ I I I · · · 000I 009 _ 009 -- 00~ -- 00~ ,~'tumsXs I~oppu!IXo e :[0t mnFq!I!nbo ~Io uo!lenb, oql s.t d!qsuoB~io:[ ]s:[§ oq.l, · X[snoouullnm!s pogs.q~s oq lsnm ll~ql o~ pul~ s~ ~II.tA[oA -U! sd.tqsuo.BelO:~ o~1 o:m oxoq& 'sso:qs Ie.quo~uel I~1Ol oql ,o£ puc 'ssoxls Im. pex Imm oql o~e ozoql luomo[o qoeo xo2:I 'sluomo[o I~O!l~moql~m 0 4O0 BOO 1200 1600 I, YEARS lO F~EEZEBACK TO THE DEPTH SHOWN YEARS OF PRODUCTION BEFORE REFREEZING COMMENCE,~ 800 1600 2400 3200 4000 EXTERNAL PRESSURE, psi Fig. 15--Casing strength (casing full of 9.6 lb/gal fluid) and estimated maximum freezeback pressures. of two weights of casing filled with 9.6-bbl/gal fluid, assuming no additional strength resulting from the surrounding cement. Conclusions 1. Mechanisms are available in the earth for sup- plying some excess water to thawed regions around a wellbore. 2. Upon refreezing, pressures will rise until the frozen soil can be pushed back at a rate essentially equal to the rate at which excess volume is being created by the refreezing process. 3. Pressures calculated with a mathematical model based on thermal and mechanical behavior of Prud- hoe Bay permafrost are in good agreement with values measured in a large-scale field test. 4. External refreezing pressures estimated for nor- mal operating conditions are high, but they are in a range that can be tolerated if the proper casing is selected. Original manuscript received in Society of Petroleum Engineers office Aug. 6, 1973. Revised manuscript received July 10, 1974. Paper (SPE 4588) was first presented at the SPE-AIME 48th Annual Fall Meeting, held in Las Vegas, Nev., Sept. 30-Oct. 3, 1973. © Copyright 1974 American Institute of Mining, Metal- lurgical, and Petroleum Engineers, Inc. This paper will be printed in Transactions volume 257, which will cover 1974. Nomenclature r '-- radius, ft & = total radial stress, psi S0 -- total tangential stress, psi *~0 '- net radial stress acting through the soil, psi ~,p -- uniaxial compressive strength of a per- mafrost sample, psi ~r0, - net tangential stress acting through the soil, psi References 1. Howell, E. P., Perkins, T. K., and Seth, M. S.: "Cal- culating Temperatures for Permafrost Completions," Pet. Eng. (April 1973) 69. 2. Couch, E. I., and Keller, H. H.: "Permafrost Thawing Around Producing Oil Wells," J. Cdn. Pet. Tech. (1970) 9, 107. 3. Eickmeier, J. R., Ersoy, D., and Ramey, H. J., Jr.: "Wellbore Temperature and Heat Losses During Pro- duction or Injection Operations," J. Cdn. Pet. Tech. (1970) 9, 115. 4. Bleakley, W. B.: "North Slope Operators Tackle Pro- duction Problems," Oil and Gas J. (Oct. 25, 197I) 69, 89-92. 5. Shalavin, A. M., and Klyushin, G. P.: "Method for Determining the Pressure on the Casing Pipes With Freezing of the Flushing Fluid in the Well," [Metodika opredeleniya velichiny davleniya na obsadnye truby pri zamerzanii promyvovhnoi zhidkosti v skavzhine] Burenie (1972) No. 9, 24-26. 6. Goodman, M. A., and Wood, D. B.: "A Mechanical Model for Permafrost Freezeback Pressure Behavior," paper SPE 4589 presented at the SPE-AIME 48th An- nual Fall Meeting, Las Vegas, Ney., Sept. 30-Oct. 3, 1973. 7. Brewer, Max C.: "Some Results of Geothermal In- vestigations of Permafrost in Northern Alaska," Trans., AGU 39, (Feb. 1958) No. 1, 19-26. 8. Klujucec, N. M., and Telford, A. S.: "Well Tempera- ture Monitoring With Thermistor Cables Through Per- mafrost,'' J. Cdn. Pet, Tech. (July-Sept. 1972) 11, No. 3, 33-37. 9. Perkins, T. K., and Ruedrich, R. A.: "Mechanical Behavior of Synthetic Permafrost," Soc. Pet. Eng. I. (Aug. 1973) 211-220; Trans., AIME, 255. 10. Ruedrich, R. A., and Perkins, T. K.: "A Study of Factors Influencing the Mechanical Properties of Deep Permafrost," J. Pet. Tech. (Oct. 1974) 1167-1177; Trans., AIME, 257. l l. Timoshenko, S., and Goodier, J. N.: Theory o! Elasticity, McGraw-Hill Book Co., Inc., New York (1951). 1166 JOURNAL OF PETROLEUM TECHNOLOGY The Mechanical Behavior of Synthetic Permafrost T. K. PERKINS MEMBER SPE-AIME R. A. RUEDRICH ATLANTIC RICHFIELD CO. PIANO, TEX. ABSTRACT Discoveries of oil in Arctic regions have led to several engineering problems that are relatively new to the petroleum industry. An understanding of some of the new problems associated with construction of surface facilities as well as with the drilling and completion of wells requires an understanding of the mechanical properties of p e rmafros t. Synthetic permafrost samples have been prepared /rom quartz sand as well as/rom natural soils talaen /rom Prudhoe Bay permafrost cores recovered from depths as great as 1,753 ft. All samples have been recompacted and frozen under a condition of zero confining stress. Samples prepared in this way should exhibit behavior similar to that of shallow permafrost. Samples have been tested in uniaxial compression at constant strain rates as well as with constant axial stress. At constant temperature and low strain rates, the log of the maximum shear strength will plot as a straight line. vs the log of the strain rate. For sand-ice samples at high strain rate, another mode of failure was evident that led to a maximum shear strength independent of strain rate. Under triaxial conditions, the maximum shear strength of sand- ice samples was generally increased with increasing stress level. In uniaxial tension, the tensile strength of sand-ice samples was found to be a function o/temperature and strain rate. Elastic response o/ these samples was obscured by the more dominant flow behavior at low strain rates. Only at very high strain rates was an elastic response clearly discernible. Young's modulus measured after 10 to 15 percent plastic strain increases with increasing stress level. INTRODUCTION Within the last few years significant oil discoveries have been made in Arctic regions. There is much speculation that additional oil will be found in regions that are characterized by quite low ambient and soil temperatures. The drilling of Paper (SPE 4057) was presented at SPE-AIME 47th Annual Fall Meeting, held in San Antonio, Tex., Oct. 8-11, 1972. (~) Copyright 1973 American Institute of Mining, Metallurgical, and Petroleum Engineers, Inc. 1References given at end of paper. This paper will be published in Transactions volume 255, which will cover 1973. wells and production of oil under these environmental conditions poses new problems not traditionally faced by the petroleum industry, but which presumably will be of increasing concern within the next few years. One new engineering challenge is that of dealing with permafrost, soil which has been continuously frozen for a number of years. Already at Prudhoe Bay a number of wells have been drilled through about 2,000 ft of permafrost. As an example of permafrost influence, measurements have shown that, when thawed permafrost around a well refreezes, significant pressures can be generated. In order to understand this phenomenon, it will be necessary to understand the mechanical behavior of permafrost. In addition, surface facilities have been constructed where there is a thin, active region (which thaws during summer months) underlain by permafrost. An understanding of permafrost mechanical behavior will aid in the design of foundations for surface facilities. There are a number of variables that can influence the mechanical behavior of frozen soils such as minerology, percent of ice saturatiOn, presence of excess ice, salt content, etc. In this paper we will describe a laboratory study of relatively fine-grained granular materials with pore st>aces saturated with ice. The results presented here may not be applicable to frozen clays or gravels, where pore spaces are undersaturated or where a large amount of excess ice is present. Since permafrost is composed of ice and soil, its behavior will naturally reflect that of its constituents. The rate of yield or flow of ice is known to be a function of temperature, shear stress and strain, but is independent of hydrostatic pressure level. Soil, on the other hand, exhibits yield behavior that is independent of temperature over the small range of permafrost temperatures of interest. For sandy soil, yield behavior is relatively independent of strain rate, but is significantly influenced by strain and stress level. Under stress, a dominant characteristic of shallow permafrost is that of yield Or flow. Its rate of flow will be a function of all the variables mentioned above. Over-all deformation results from a combination of elastic and flow behavior. The paper describes a method of preparing samples that we believe are similar to shallow permafrost, as well as a study of the elastic and flow behavior of the samples under compressive AUGUST, 1973 211 and tensile conditions. BEHAVIOR OF ICE AND SOIL A brief review of the behavior of ice and soil can give much insight into the complex behavior of permafrost. When water freezes, its volume increases about 9 percent, giving ice of a density2 of about 0.917 gm/cc. The density will increase as the temperature falls. 13 The freezing point of fresh water is 32°F, but the freezing point will be depressed if the ice is under pressure. Inconsistent values have been reported for Young's modulus of ice when measured by compression or bending.2 These scattered values presumably result from inelastic creep or flow of ice under these relatively static conditions. Under conditions of very high strain rates or low shear stress such as when testing by sonic methods, much more consistent values of Young's modulus are reported.3 Although depending on grain orientation, temperature, etc., Young's modulus is of the order of 1.4 × 106 psi. Poisson's ratio has been reported2 to be about 0.33 to 0.38. Compres- sibility is then calculated12 from the relation for isotropic materials to be about 6 to 8 × 10-7 psi-1. Ice compressibility, however, may depend on trapped air and might also depend on stress level. That is, for polycrystalline ice, initial rearrangement of ice crystals may lead to greater compressibility when loads are first applied. Several investigators have reported values for the apparent viscosity of ice.2 These values are quite inconsistent. It is clear that ice will readily flow, but it is equally clear now that its rheology is non-Newtonian.4'11 Experiments have shown that ice shear-strength is quite temperature-dependent and that a few percent strain are required for the ice to develop its full strength.11 The maximum shear strength is a function of strain rate. Fig. 1, which is based on the data assembled by Dillon and Andersland, 10 summarizes creep data on polycrystalline ice and suggests that different modes of failure can occur at different strain rates. Let us turn our attention now to soil behavior. The strength of soil constituents is essentially independent of temperature over the small range of permafrost temperatures of interest. Vesi~ and Clough14 have shown that, under high stress, unconsolidated sand will cease to compact after a few hours. Hence at low strain rates, yield behavior of sand is essentially independent of strain rate. If the soil contains mostly clay, then creep over extended periods of time might be anticipated.15 It is well known that the shear strength of an unconsolidated soil is a function of stress level; and it should also be remembered that substantial strain is required to develop the full shear strength. If ice fills the pore spaces of a granular material, then much of the deformation will presumably occur under what is commonly called an undrained condition; that is, where the pore volume is held essentially constant. On the basis of our permafrost experiments as well as undrained deformation experiments with sand packs, we believe that an initial hydrostatic loading of permafrost will bring some stress to bear on the soil matrix because of initial compaction of ice or compression of.trapped air. Initial deformation results in crushing of soil particles, which tends to transfer stress back into the ice phase. With continued strain, the sand becomes dilatentll and stress tends to be transferred from the ice back to the soil matrix. SAMPLE PREPARATION Synthetic permafrost samples were prepared in cylindrical metal molds 1 in. in diameter by 4 in. long for uniaxial compression and 7/8 in. in diameter by 3 in. long for triaxial compression. The molds were filled with distilled water, and then soil was added while tamping to give a dense pack. The tops of the molds were insulated and the entire assemblies were placed in a cold chest. Freezing occurred from the bottom, thus allowing excess water to be expelled at the top during the freezing process. The top surfaces were trimmed perpen- dicular to the cylindrical axis, and the split molds were then disassembled for removal of the cores. Much of this study has been conducted with samples prepared from quartz sand. Such material has been studied previously and results are available in the literature11'16 for ready comparison. On the other hand, frozen natural soils may exhibit behavior somewhat different from that of pure quartz sand. Accordingly, several samples have been prepared from natural soils taken from permafrost cores recovered at Prudboe Bay from depths as great as 1,753 ft. These samples contained naturally occurring salts. In addition, a FIG. I0-1 10-2 10-3 10_4 _~- 10-5 10-6 10_7 10-8 10-;2 10-1 I00 I01 102 03 OCTAH~DRAL SHEAR STRESS, PSt 1 m CORRELATION OF CREEP DATA FOR POLYCRYSTALLINE ICE. 212 SOCIETY OF PETROLEUM ENGINEERS JOURNAL few samples have been prepared from mixtures of quartz sand and low-yield clay. These latter samples seem to behave more nearly like the natural soil samples in some respects. Tables 1 and 2 give the properties of the frozen samples and soil material. Since all samples have been recompacted and frozen under zero confining pressure, the possible effects of soil compaction and initial soil stress level have not been investigated. We believe, however, that samples prepared in this way will exhibit behavior similar to shallow permafrost. COMPRESSIVE YIELD BEHAVIOR HIGH STRAIN-RATE UNIAXIAL TESTS High strain rates were achieved with a load machine that could be operated at various constant crosshead speeds. The cylindrical samples were placed in a constant temperature environmental chamber mounted in the test machine. After allowing the sample to reach uniform temperature, the crosshead was advanced to give strain rates from 0.25 min-1 to 5 x 10-s rain-1. Applied forces were measured with a load cell. Typical behavior at low strain rates is illustrated on Fig. 2. The induced axial stress is clearly a function of strain rate, temperature, and axial strain. For a given experiment, the stress rose rapidly, reached a peak at about 5- to 7-percent strain (for sand-ice cores), FIG. 2 -- TYPICAL STRESS-STRAIN CURVES FOR SAND-ICE CORES. and then fell gradually with further strain. Fig. 3 shows the ratio of axial stress-to-maximum axial stress vs axial strain for sand-ice cores. Eq. 1 is an empirical equation that describes typical behavior as maximum strength is being developed at low strain rates. The typical value of ep for sand-ice cores is 0.055. 2 - - , · ~- ep .... (1) After achieving maximum axial strength, some samples exhibited a significant decrease in axial stress as strain was increased by several percent. TABLE 1 I PROPERTIES OF SYNTHETIC PERMAFROST SAMPLES *Density of Percent Retained on U.S. Sieve Mesh Size Frozen Less Sample Samples Than Number Description of Sample (Ib/cu ft) 20 40 80 120 200 200 Ep S-1 Penn Sand 129 0 0 83.5 8.~5 7.8 0.7 0.055 NP-I* Natural permafrost soil from 120 47.76 34.47 14.94 1.14 0.974 0.812 0.04 899 to 902 ft, Prudhoe Bay NP-2 Natural permafrost soil from 106 29.508 37.7 26.64 2.992 1.639 1.516 0.035 957 to 960 ft, Prudhoe Bay NP-3 Natural permafrost soil from 110 0 2.475 75.612 8.99 5.576 7.348 0.05 1,418 to 1,420 ft, Prudhoe Bay NP-4 Natural permafrost soil from 113 0 16.397 62.979 5.612 9.353 5.658 0.025 1,751 to 1,753 ft, Prudhoe Bay SC-1 15 percent Iow yield clay 113 0.025 85 percent Penn Sand SC-2 30 percent Iow yield clay 102 0.015 70 percent Penn Sand *Soil sample NP-1, as received, contained gravel as large as 1 in. in diameter. Fifty-seven percent by weight would not pass througha ~A-in. U.S. sieve screen and was removed. The remainder, having the particle-size dis- tribution shown above, was studied. All other natural soil samples were studied in their "as received" condition. Sample Number S-1 NP-1 NP-2 NP-3 NP-4 TABLE 2 -- APPROXIMATE ANALYSIS OF MINERAL CONTENT AS DETERMINED BY X. RAY Description of Sample Weight Approximate Content of Remaining Constituents Percent Percent Percent Percent Quartz Chlorite Gypsum Illite Penn Sand 100 Natural permafrost soil from 899 97 to 902 ft, Prudhoe Bay Natural permafrost soil from 957 97 to 960 ft, Prudhoe Bay Natural permafrost soil from 89 1,418 to 1,420 ft, Prudhoe Bay Natural permafrost soil from 79 1,751 to 1,753 ft, Prudhoe Bay Percent Percent Percent Kaolinite Montmorillinite Muscovite 6 2 1 5 5 3 2 AUGUST, 1973 213 Most samples, however, exhibited only a 10- to 20-percent decrease in axial stress with an additional 10-percent strain. Fig. 4 is a similar plot showing typical behaviors of natural soil-ice samples. On the average, these samples achieved maximum axial strength with a somewhat smaller axial strain. Typical behavior is also represented by Eq. 1; values of ~,, are given in Table 1 b' ' The relationship between temperature, strain rate, and maximum axial strength is shown on Fig. 5 for sand-ice cores. At strain rates less than about 2 x 10-3 min-1 the yield process appears to be governed by ice flow. These samples exhibit behavior similar to that which has been reported for ice cores. At constant temperature, maximum stress plotted vs strain rate on log-log paper will yield a straight line. Such a plot is characteristic of rate processes that have been postulated for ice flow. The maximum strength can be represented mathematically by empirical Eq. 2. 8 ......... (2) Crmox = 4~p · At strain rates greater than about 0.01 min-1 a different mode of yield behavior is evident. Maximum strength is a function of temperature, but is independent of strain rate. This behavior is also similar to that which has been reported for ice and which is shown on Fig. 1. Samples that were yielded in this manner generally remained competent. Visual inspection showed a multitude of slip lines that were readily apparent on the outer surface of the cylinders. Fig. 6 shows typical stress-strain curves obtained at high strain rates. These curves indicate an elastic response with additional strain resulting from plastic flow. Total strain is thus represented mathematically by Eq. 3. ! e t = + ¢o dt ....... (3) O 1,0, 0,9 -- 0.8 -- 0,7 -- -~o,4 0,3 0.1 I . I 0 .OI ,02 ,03 ,04,0§ ,06 ,07 ,OB ,09 ,10 ,l! ,12 TOTAL AXIAL STRAIN FIG. 3 -- STRAIN REQUIRED TO DEVELOP MAXIMUM UNIAXIAL STRENGTH FOR SAND - ICE CORES; TEMPERATURES FROM 8° TO 30°F; STRAIN RATE FROM 5 x 10-s TO 5 x 10-a MIN-1. Combining Eqs. 1, 2 and 3 gives Eq. 4. _ -- ~ < ep . . . (4) Fig. 7 indicates that data from high-strain-rate experiments are consistent with this equation. It would appear that early deformation results from elastic deformation and ice flow, but that a sec. ond mode of shear failure limits the maximum strength of the samples. --- 07 ~ uPI ~ 0,~ FZa. 4 -- STRAIN REQUIRED TO DEVELOP MAXIMUM UNIAXIAL STRENGTH FOR NATURAL SOIL-ICE SAMPLES AT LOW STRAIN RATES, TEMPERATURES FROM 8° TO 28°F. ~~ I I1111111 I l ll~lI , ~,~l1 0,! i 10_:> 10-4 ~ 10-5 10 100 103 104 MAXIMUM AXIAL STRESS, PSI FIG. 5 ~ MAXIMUM UNIAXIAL COMPRESSIVE STRENGTH OF SAND-ICE CORES. 214 SOCIETY OF PETROLEUM ENGINEERS JOURNAL Fig. 8 is a typical plot of the maximum uniaxial compressive strength for the natural soil-ice samples. For the same temperature and strain rate, the natural soil samples invariably yielded with lower peak stress than did sand-ice cores. Log-log plots of peak stress vs strain rate again yielded straight lines. The second mode of failure, wherein maximum strength was independent of strain rate, was not evident in samples NP-2, NP-3 and NP-4. There was evidence of this phenomenon occurring in sample NP-1. We have .speculated that the slightly samples may result from salt content, presence of clay, or high moisture content. These samples were of low density when prepared by the described procedure. Numerous authors17 have reported unfrozen moisture in soils containing fine-grain silt and clay at temperatures well below 32°F. This concept has been tested by preparing synthetic samples containing quartz sand and Iow-yield clay. A soil mixture containing 85 percent quartz sand and 15 percent Iow-yield clay gave Ep = 0.02, whereas 70 percent quartz sand with 30 percent clay gave different behavior exhibited by the natural soil Ep = 0.015. Fig. 9 shows maximum uniaxial ~ 0~ -1:3 ~ 0 0 0 -- 28OOF ~ O .o-' o-'-' c-.o .... °'-o ! / '0 .o' o'' '.~ ''O CURVE STRAIN RATE' TEMPERATURE I~ I '*0, ,0'* ^,*" ~x M IN-1 oF -- / ? o-°-o-o.o,,'~ -o '-g'- - 00-'~-'-~~ - 2400[-- 2oo01-- '/- , '~ ~ ~ ~ ~ ~, ,oo5 1~ - ~ I ~ e, .005 27.5 ~ i ,~ .25 17 ~: 1600 O--O~ O-~ O~ ~o~ ~ ~ ~ o - ~00 o? I I I I I I I I I I I I I I - 0 .01 .02 .03 .04 .05 .06 .07 .08 .09 .10 .11 .12 .13 .14 .15 TOTAL AXIAL STRAIN FIG. 6 -- STRESS-STRAIN CURVES FOR SAND-ICE CORES DEFORMED AT HIGH STRAIN RATES. 0.B 0.7 o o O O 0~0~-O--O~o~ CURVE 0 [] STRAIN RATE MIN.-1 .0O5 .05 .25 .005 .05 .25 .005 .05 .25 .25 TEMPERATURE _ 11.5 12 11 18 18,5 17 27,5 27 28 28 0,3 0.2¸ 0.1 0 o .01 ,02 .03 .04 ,05 ,06 .07 ,08 .09 .10 ,11 ,12 .13 .14 o TOTAL AXIAL STRAIN FIG. 7~STRAIN REQUIRED TO DEVELOP MAXIMUM UNIAXIAL STRENGTH FOR SAND-ICE CORES DEFORMED AT HIGH STRAIN RATES. AUGUST, 1973 .215 compressive strengths vs strain rate for clay-sand- ice samples at 18°F. Maximum strengths were essentially equivalent to those of natural soil sample NP-3. The gross shear mode of failure was not evident at high shear rates. VERY LOW STRAIN-RATE UNIAXIAL TESTS Many problems DE engineering interest involve the behavior of permafrost over many years of time. 10-2 10-3 -- -- _- _ ~ - -- -- -- _ -- - / - -- -- -- o ~ x 27 TO 28°F -- -- -- o 18 TO 19'F - '" 8 TO _ _ 10-5 10 x~ I IIllllll I I I 11111I I I I11111 100 103 04 MAXIMUM AXIALSTRESS, PSI FIG. 8 -- MAXIMUM UNIAXIAL STRENGTH OF NATURAL SOIL-ICE SAMPLE NPo3. o,~ __-- I ---- ~ / - · -- x '' ._ / ._ I 7 lO-~ _-_ / · _ ,_ ,_ ~- 10-3 --~ - _ _ _ x-15% CLAY 10-4 ~ I 85% SAND -- [ - _ x~ O~ 30% CLAY _ -- - 70% SAND _ -- _ -- _ io-5 i i Ill,III I I [llllll I I I1~111 10 I00 103 104 MAXIMUM AXIAL STRESS, PSI FIG. 9 -- MAXIMUM UNIAXIAL STRENGTH OF CLAY- SAND-ICE CORES AT 18°F. Very low rate yield behavior is therefore of real interest. Ladany~-8 reviews mathematical theories of creep, including much of the work reported in Russian journals. From an expetimental viewpoint, very low rate creep experiments present severe practical difficulties. Long periods of time are required to obtain data, and considerable investment in equipment is required if multiple experiments are run concurrently. We have investigated low strain-rate creep by using constant axial force equipment. Several load frames such as that illustrated on Fig. 10 were constructed. Frozen sand-ice samples 7/8 in. in diameter by 3 in. long were installed and covered with a loose fitting rubber sleeve to prevent sublimation. The entire assembly was placed in a cold chest. The hydraulic cylinder was filled with a nonfreezing liquid and pressured with a gas-filled cylinder. Friction of the O-ring was determined by calibration. Strain of the core was measured with dual potentiometers. These potentiometers were powered with a very stable power supply and voltages were read to five significant figures with a digital voltmeter. Calibration against a micrometer yielded an accuracy of about _+ 2 x 10-4 in. For sand-ice cores, it is necessary to keep the strain rate below about 0.001 rain-1 to avoid shear failure. High strain rates can be achieved with rather low stress until the cores have strained a measurable amount. On the basis of experience gained with constant strain-rate experiments, a starting procedure was developed that would maintain the instantaneous strain rate less than HYDRAULIC CYLINDER ).RING SEAL STEEL LOADING FRAME ,YNTHETIC PERMAFROST CORE EAR POTENTIOMETERS FIG. 10 -- HYDRAULICALLY LOADED PERMAFROST CORE HOLDER. 216 SOCIETY OF PETROLEUM ENGINEERS JOUIINAI, 0.001 min-1. This starting procedure involved the gradual increase of stress over a period of time of several hours. When the desired maximum pressure was achieved, the gas supply valve was closed. Pressures and displacements were recorded periodically thereafter by an electronic data acquisition system. The gas-loading system was designed such that the axial force remained nearly constant even though the core continued to yield. At essentially constant axial stress, the strain rate diminished with continued axial strain until a minimum was reached after about 5 to 9 percent strain. The strain rate then increased with further strain. Fig. 11 shows typical plots of creep strain vs time. The strain at which a minimum creep rate is achieved is essentially equivalent to the strain at which maximum strength is achieved during constant strain-rate deformation. Rearranging Eq. ! yields Eq. 5. O'max : 2 ' · ~- em ' (.5) Fig. 12 shows that creep data, when plotted in accordance with Eq. 5, are in agreement with the flow relationship developed from constant strain-rate experiments. HIGH STRAIN-RATE TRIAXIAL TESTS Tests of samples under triaxial compressive conditions were conducted in a triaxial cell. Only sand-ice samples have been tested under triaxial conditions. Cylinders for these experiments were 7/8 in. in diameter by 3 in. long. The cylinders were loaded into the precooled cell, the annular space was filled with a cold, nonfreezing liquid, and the whole assembly was placed in the constant temperature chamber mou~nted in the load machine. After reaching uniform temperature, the axial stress and radial pressure were raised slowly, keeping the sample in a nearly hydrostatic condition at all times. When the maximum desired radial pressure was reached, it was thereafter maintained constant with a constant pressure system. The crosshead was then advanced at a uniform rate while 07I 1 [ IIIIIII ; I IlflllI I I SllillI I 06 · PSI 1050 095 2 le ~75 055 .2 3 o~ ol - io-~ 1o-~ ol 1 TIM[, DAYS FIG. 11 ~ CREEP VS TIME. I f I 1111t lo0 monitoring the axial stress with a load cell. Fig. 13 shows a typical plot of the axial stress minus the radial pressure vs the axial strain. The flow behavior of ice is reported9 to be independent of hydrostatic stress level. It is well known, however, that granular materials are strengthened at high hydrostatic stress levels. Behavior of these sand-ice samples indicates that some stress is being carried through the sand matrix upon initial hydrostatic loading. There is little effect of radial pressure upon the stress difference if the radial pressure is between 500 and 1,670 psi and if axial strain is less than about 4 percent. Above about 4 percent strain, the sand became dilatent, thus transferring more of the stress to the sand matrix and increasing the axial-radial stress difference. At large strains, the peak stress difference is increased with increasing radial stress level. The axial strain, at which the peak stress difference is achieved, is increased at higher stress levels, probably because of greater crushing of sand particles. After reaching its. peak, the stress difference decreased slightly with additional axial strain. TENSILE YIELD BEHAVIOR Sand-ice samples were tested in uniaxial tension using the sample holders sketched on Fig. 14. During sample preparation, a split cylindrical mold was spaced between the end pieces and clamped to give a watertight container. The mold was filled with water and sand and frozen from the bottom to permit expulsion of excess water. After freezing, the split cylindrical mold was removed from the 10-2 ~ 10-3 -- < ¢n 10-4 10-5 10-6 10 100 103 104 · _.~_e 2 1_(1_ .m) FIG. 12 -- COMPARISON OF CREEP DATA WITH CONSTANT STRAIN RATE DATA. AUGUST, 1973 217 3200 2800 2400 2000 1600 1200 800 400 I ~ I TEMPERATURE = 18'F STRAIN RATE = 6.67 x 10-4 ' I . ---_.._. LEGEND: 1670 PSI RADIAL PRESSURE 1000 ...... .......... 500 ...... ......... 100 ...... 0 .02 .04 .06 .08 .10 .12 .14 AXIALSTRAIN FIG. 13 -- SAND-ICE CORE BEHAVIOR UNDER TRIAXIAL CONDITIONS. center, exposing a test section 2 in. long and 1-19/32 in. in diameter. A linear transducer was attached to measure axial strain, and the entire assembly was mounted into the constant temperature chamber contained within the load machine. The assembly was pinned to the load cell with an intervening universal joint to prevent eccentric loading. After reaching constant temperature, the crosshead was retracted at a constant rate. Maximum tensile strengths of sand-ice samples are shown on Fig. 15. The maximum strength was found to be a function of strain rate at low strain rates, but appeared to be essentially independent of strain rate at the two highest values measured. The maximum strength was found to be a function of temperature, with the strength falling more LOW THERMAL CONDUCTIVITY PLASTIC ROD SYNTHETIC PERMAFROST CORE 2 INCH LONG CYLINDRICAL SECTION EXPOSED LINEAR TRANSDUCER INTERNAL GROOVES RESIST CORE SLIPPAGE ~g 3£= ~..~ END CAPS FIG. 14 m SKETCH OF TENSILE CORE HOLDER. rapidly as the temperature approached the melting point. For low strain rates where we believe ice flow plays a dominant role, the ratio of induced stress to peak stress is plotted vs axial strain on Fig. 16. The peak stress was reached at 2 to 5 percent strain, after which the strength decreased with further strain. Fig. 17 depicts typical stress-strain behavior. At high strain rates, there is a detectable elastic response with superposed flow as axial stresses get large. Thus Eq. 3 is again expected to apply. At high strain rates, there was ultimately total separation'of cores along a plane of tensile failure as shown on Fig. 17. At low strain rates, the 1o3 1oo STRAIN RATE a __ 0.1 MIN.-1 0,01 ., -- 0,001 ', -- 0.0001 " 10 I I I I I t 4 fl 12 16 20 24 28 32 TEMPERATURE, °F FIG. 15 -- UNIAXIAL TENSILE STRENGTH OF SAND- ICE CORES. 218 SOCIETY OF PETROLEUM ENGINEERS JOUllNAL experiments were terminated after significant strain even though axial stresses were still large. ELASTIC BEHAVIOR The concept of permafrost elasticity is quite meaningful when applied to processes that occur at high strain rate or when shear stresses are small, such as during the transmission of sonic vibrations. The concept of elasticity should be used with caution, however, when dealing with processes that occur slowly. For these recompacted samples, which were initially frozen under zero confining stress, the dominant deformation behavior exhibited at low strain rates is that of flow. During this investigation, elastic response was only revealed clearly at the highest strain rates studied. The Young's modulus could be estimated from three types of experiments: 0.8 o.4 I 0.~ o[ 0 · o~ · · · ° o Da ax a I I a a STRAIN RATE O,1 MIN,-I O,O1 '. O,OO1 ,, O,OOO1', .01 ,02 ,03 ,04 ,05 ,06 ,07 TOTAL AXIAL STRAIN FIG. 16 -- STRAIN REQUIRED TO ACHIEVE MAXIMUM TENSILE STRENGTH AT LOW STRAIN RATES; TEMPERATURES 18° TO 28°F. ooo 800 700 6OO ~soo ~oo 2oo 1oo o I I I CURVEI STRAIN RATE T EM P,EFRATUR E I MiN,_i * o ,Ol 17,2 o .OOl lB o I ,01 23 t, I .OOl 1~.5 ~ ,OOOl o ,Ol 28 o ,00l 24 o ,001 28 o ,OOOl 27,5 ~- INDiCATESCOMPL£TE TENSILE FAILURE I [ ] .06 ,07 ,08 FIG. 17 -- STRESS-STRAIN CURVES FOR SAND-ICE CORES IN UNIAXIAL TENSION. 1. Application of uniaxial compressive stresses at high strain rates. 2. Application of uniaxial tensile stresses at high strain rates. 3. Removal of uniaxial or triaxial compressive stresses after having plastically shortened the samples at various low strain rates. The initial Young's modulus under high strain rate, uniaxial compression is estimated from Fig. 6 to be approximately 2 to 3 × l0S psi. The initial Young's modulus under high strain-rate tension is estimated from Fig. 17 to be of about the same range. After plastically'straining samples 10 to 15 percent in the triaxial cell, the crosshead was withdrawn at the maximum rate until the axial stress approached the radial stress. Stress-strain behavior was nonlinear with the tangent modulus diminishing as the axial stress approached the radial stress. A secant modulus has been estimated and is shown on Fig. 18. The secant modulus appears to be relatively independent of temperature, but is noticeably increased at high stress levels as is typical of granular materials. CONCLUSIONS 1. Under conditions of low strain rate, elastic response of synthetic permafrost that is initially frozen under zero confining stress is obscured by its more dominant flow behavior. 2. At constant strain rate either in tension or compression, several percent strain will occur before maximum strength is achieved. 3. At a fixed temperature and for low strain rate uniaxial compression, the log of the maximum strength will plot as a straight line vs the log of the strain rate. 4. For sand-ice samples tested in uniaxial compression at high strain rates, the maximum strength is independent of the strain rate but is a function of temperature. 5. For a given temperature and at a given low strain rate, the shear strength of sand-ice samples, which are initially frozen under zero confining stress, increases as the radial stress increases. 6. Uniaxial tensile behavior of sand-ice samples is also characterized by substantial flow at low IO7 '' ; : ', I' '11I 106 ~ 105 x 18°F o 28°F ~o4 i iiiiilll I I IIItlll I IIIIIIl[ i i ii,l~tl I lllllll o,1 1 1o 100 1o3 104 RADIAL PRESSURE, PSL FIG. 18 -- YOUNG'S MODULUS OF SAND-ICE CORES AFTER 10- TO 15-PERCENT PLASTIC STRAIN. AUGUST, 1973 219 strain rates. Tensile strengths are a function of temperature and strain rate. 7. At high strain rates, elastic response of sand-ice samples can be discerned. Stress-strain behavior after 10 to 15 .percent of plastic strain is nonlinear. Secant values of Young's modulus increase with an increasing stress level. NOMENCLATURE A = coefficient in empirical Eq. 2, a function of temperature B = exponent in empirical Eq. 2, a function of temperature E -- Young's modulus, psi e -- axial strain em -- value of axial strain at which a minimum strain rate is achieved at essentially constant axial stress ep -- value of axial strain at which peak strength is achieved during constant strain rate deformation et = total axial strain t = time :p = plastic strain rate q = axial stress, psi amax = maximum axial stress, psi REFERENCES 1. Dorsey, N. Ernst: Properties o/ Ordinary Water- Substance, Reinhold Publishing Corp., New York (1940) 603. 2. Review o/ the Properties of Snow and Ice, Mantis, Homer T., editor'. Snow, Ice and Permafrost Research Establishment, U.S. Army, Corps of Engineers, U. of Minnesota Institute of Technology Engineering Experiment Station (July, 1951) Report 4. 3. Boyle, R. W. and Sproule, D. C.: "Velocity of Longitudinal Vibrations in Solid Rods with Special Reference to the Elasticity of Ice," Cdn. ]. Research (1931) Vol. 5, 601. 4. Glen, J. W.: "Experiments on the Deformation of Ice," J. o! Glaciology (1952-1956)Vol. 2, 111. 5. Nye, J. F.: "The Flow Law of Ice from Measurements in Glacier Tunnels, Laboratory Experiments and the Jungfraufirn Borehole Experiment," Proc., Royal Soc. of London (1953) Series A, Mathematical and Physical Sciences, Vol. 219, No. 1139, 477. 6. Steinemann, Samuel: "Results of Preliminary Exper- imen'ts on the Plasticity of Ice Crystals," J. of Glaciology (1954) Vol. 2, No. 16, 404. 7. Glen, J. W.: "The Creep of Polycrystalline Ice," Proc., Royal Soc. of London (1955) Series A, Mathematical and PhysicaI Sciences, Vol. 228, No. 1175, 519. 8. Butkovich, T. R. and Landauer, J. K.: "The Flow Law for Ice," U. S. Army Snow, Ice and Permafrost Research Establishment, Corps of Engineers, Wilmette, Ill. (Aug., 1959) Research Report 56. 9. Glen, J. W.: "The Rheology of Ice," Ice and Snow Properties, Processes, and Applications, W. D. Kingery, editor, MIT Press, Cambridge, Mass. (1963) 3. 10. Dillon, H. B. and Andersland, O. B.: "Deformation Rates of Polycrystalline Ice," Physics o/ Snow and /ce, Hirobumi Oura, editor, The Institute of Low Temperature Science, Hokkaido U., Sapporo, Japan (1967) 313, 11. Goughnour, R. R. and Andersland, O. B.: "Mechanical Properties of a Sand-Ice System," J. Soil Mech. and Foundations Div., Proc., ASCE (July, 1968) 923. 12. Gold, L. W.: "Some Observations on the Dependence of Strain on Stress for Ice," Cdn. J. Physics (19S8) Vol. 36, 126S. 13. Pounder, E. R.: The Physics of Ice, Pergamon Press, Oxford (1967). 14. Vesid, A. S. and Clough, G. W.: "Behavior of Granular Materials Under High Stresses," J. Soil Mech. and Foundations Div., Proc., ASCE (May, 1968) 661. 15. Mitchell, Jo K., Campanella, R. G. and Singh, Awtar: "Soil Creep as a Rate Process," J. Soil Mech. and Foundations Div., Proc., ASCE (Jan., 1968) 231. 16. Andersland, O. B. and Alnouri, Ilham: "Time- Dependent Strength Behavior of Frozen Soils," J. Soil blecb, and Foundations Div., Proc., ASCE (July, 1970) 1249. 17. Scott, Ro F.: "The Freezing Process and Mechanics of Frozen Ground," Cold Regions Science amd Engineering Monograpb, U. $. Army, Corps of Engineers, CRREL, Hanover, N. H. (Oct., 1969) 11-DI. 18. Ladanyi, Branko: "An Engineering Theory of Creep of Frozen Soils," Cdn. Geotecb. J. (1972) Vol. 9, 63. 220 SOCIETY OF PETROLEUM ENGINEERS JOURNAL SOCIETY OF PETROLEUM ENGI~[EERS OF AI~ 6200 North Central Expressway Dallas, Texas 75206 THIS iS A PREPRINT --- SUBJECT TO CORRECTiOi[ Preci se Jo int Length Determi nat i on Mu l t i p I e Oas i ng Col I ar Locator PAPER S P E Using Tool 5087 A R. A. Ruedrich and T. K. Perkin~, Atlantic Richfield Co., and D. E. O'Brien, Exxon Production Research Co., Members SPE-AIME ©Copyright 1974 American Institute of Mining, Metallurgical, and Petroleum Engineers, Inc. This paper was prepared for the 49th Annual Fall Meeting of the Society of Petroleum Engineers of AIMS, to be held in Houston, Texas~ Oct. 6-9, 1974. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment of where and by whom the paper is presented. Publication elsewhere after publication in the JOURNAL OF PETROLEUM TECHNOLOGY or the SOCIETY OF PETROLEUM ENGINEERS.JOURNAL is usually granted upon request to the Editor of the appropriate journal provided agreement to give proper credit is made. Discussion of this paper is invited. Three copies of any discussion should be sent to the Society of Petroleum Engineers office. Such discussions may be presented at the above meeting and, with the paper~ may be considered for publication in one of the two SPE magazines. ABSTRACT Numerous examples of changes in casing length resulting from fluid withdrawal, subsidence, etc., are repOrted in the petroleum industry literature. Logging tools have previously been described which measure changes in casing length in situ. This paper describes the design and application of a new tool that provides reproducible measurements with a precision of a few hundredths of an inch, tenfold better than methods previously reported. The new tool consists of four magnetic casing collar locators arranged with a particu- lar spacing in the logging tool. Laboratory studies have determined the influence on the collar locator signal of the following factors: (1) stress level in the casing, (2) temperature, (3) logging velocity, (~) changes in radial position of the sonde in the casing, (5).effect of recorder speed, (6) effect of cable cross- talk, (7) effect of magnetic fields, and (8) changes in typeof collar. The laboratory studies also gave insight into desirable recording and interpretation techniques. Digital recording permits automatic computer processing of data using cross- correlation methods. A field tool was built and tested in 10 joints of 5~-~-in. casing to verify the tool's capability to detect small changes in joint length that were produced by loading the casing string. The joint length was determined for each joint by analysis of both analog records and digital records. The casing string was stretched by picking up an additional 63, O00 lb. An analysis of the logs run on the stressed string shows the expected elongation of the individual joints without a change in the stan' dard deviation of the measured joint length. INTRODUCTION The precise measurement of casing joint length in situ has been the goal of several research efforts in the petroleum industry. Precise length determination is needed to evaluate the effects of stress changes that might occur in the vicinity of the well over a long period of time. A loading process of common interest is that produced by pore pressur reduction that increases the stress in the rock or soil matrix. Strains, which result from compaction of the matrix, can be transmitted to the casing and become a factor in casing design. Several devices employing magnetic casing collar locatorS have been developed previously to obtain casing strain data. Allen and his associates1 have described a "collar counting" technique that employed a single magnetic collar locator. Subsequently they2 employed a tool References and illustrations at end of paper. P? ~ISE JOINT LENGTH DETERMINATION USI5 ~JLTIPLE CASING COLLAR LOCATOR TOOL SPE 5087 that had approximately a full joint length spacing between two collar locators. The two sensors in this tool produced signals concur- rently and reduced the dependence on logging velocity information. A third tool built by Allen utilized a displacement wheel on the tool to record actual tool displacement as a function of time. Davies and Boorman,3 in a field test to investigate the effect of thawing permafrost, used a triple collar locator. The upper pair of magnetic collar locators were 42.22 ft apart (approximately one casing joint). The magnetic center of the third locator was 24 in. below the magnetic center of the second locator. These closely spaced sensors would pass the same collar within a few seconds, thus providing a measure of tool velocity without the down-hole displacement wheel employed by Allen. Allen found the precision of his tools to be on the order of 0.04 ft utilizing various discrete point analysis procedures. The work of Davies and Boorman suggests that the precision of the triple collar locator was also of the order of a few hundredths of a foot. In an effort to improve the precision of in-situ casing length measurement, an experi- mental evaluation of magnetic collar locators was undertaken to determine their capabilities and reliability. Subsequently, a new multiple collar locator tool was designed, built, and field tested. The potential variables of design or operation of a multiple collar locator tool that were studied in the laboratorywere (1) mechanical stress changes in the casing, (2) thermal changes in the casing, (3) logging velocity, (4) changes in radial position of the sonde in the casing, (5) effect of recorder speed, (6) effect of cable cross-talk, (7) effect of magnetic fields, and (8) changes in type of collar. LABORATORY EVALUATION OF MAGNETIC CASING COLLAR LOCATORS The magnetic casing collar locators 'consisted of magnets and a coil of wire. This arrangement would produce an induced voltage if the tool was pulled past a discontinuity in the pipe wall, such as a collar. The ideal response of a particular locator and collar combination is shown in Fig. 1. Experimental equipment employed in the study of variables is illustrated in Fig. 2. A double-acting hydraulic cylinder was incorpor- ated into each leg of a load frame so that 100,000 lb of tensile or compressive force could be produced. A force of this magnitude will produce a stress of nearly 20,000 psi in the 5~-~-in., 17-1b/ft casing. The casing string consisted of two standard API buttress thread collars and three pup joints. The pup joint lengths were 30, 36, and 52 in., respectively, from top to bottom. The over-all assembled length of the casing string was 120 in. The casing string was threaded into 5~-in. NPT flanges, which were bolted to the load frame. The basic string was modified for collar design investigations. These included the utilization of long steel collars and long nonmagnetic collars as well as magnetically discontinuous collars on which magnetic fields of varying strengths were imposed. Dial gauge assemblies were installed on the string to independently determine the change of length of the 36-in. pup joint and its associated collars. The string was wrapped with heating tapes and insulation to permit elevated temperature experiments. As shown in Fig. 2, a draw works was fabricated to operate the casing collar locator in the casing string. The logging sonde was run on a 16-gauge three-conductor cable. There was no differential stretch in this heavy cable during the course of a logging run. The cable passed through a series of three sheaves en route to the cable spool from the top of the casing. The first sheave was on a variable position mounting above the string so that by locking the sheave at different orientations above the open casing, the cable's point of exit could be specified with respect to radial position. The draw works utilized a spool with a lO-ft circumference. The rotation of the spool allowed a traverse of the casing string to be completed in a single revolution of the spool, No overlapping of the cable or change in tool velocity was permitted. A 30-rpm reversible electrical motor and worm gear assembly were used to drive the spool. The logging speed was varied by changing this worm gear assembly. A pair of microswitches limited the travel of the logging sonde to the length of the string. A third microswitch and a series of cams on the spool were used to verify that the cable was being taken up at a uniform rate. This microswitch was closed as each cam rotated past the microswitch. The cable displacement between microswitch closures was measured, and thus incremental velocities were computed from these distances and the time increments between closures. In the logging region of interest, the incremental velocities were reproducible to within less than 0.1 percent. The cams that produce microswitch closures prior to logging the first collar and after logging the second collar were left on the spool. These served as displacement markers on the strip chart record- ing of the collar locator response. An example trace is shown in Fig. 3. The markers provided an automatic length calibration for the log since the markers were a known distance apart. A simulated cable and an actual logging cable were used in the recording circuit of the experimental se6up to determine signal attenua- tion and cross-talk between conductors. The simulation was an RC network that approximated sPE 5087 the effect of 18,000 ft of conventional logging cable. Additional cable influence data were collected with a 6,~O0-ft length of new seven- conductor armored cable be%ween the tool and the recorder. The cross-talk in %he cable was simulated by superimposing vol%ages on nearby cables using a vol%age generator. The generator was tuned %o produce a 2-hertz signal with an amplitude of 0.5 volts, which approximates the amplitude and dominant frequency of the casing collar locator signal when operated at a velocity of ?50 f%/hr. ANALYSIS OF LABORATORY DATA R. ~ ~UEDRIOH, T. K. PERKINS and D. E. ~ fEN The experimental logging program carried out with the original casing configuration is shown in Table 1. The reproducibility of the AZ ~4 A5 logging process was established for each experimental condition by running a set o£ five replicates. A trace, similar to that illus- trated in Fig. 3, was produced by each logging run in the casing. In order to determine the distance between collars, each collar locator signal must be located on the trace. Six specific points on each response were selected as reference points. They were the two positive peaks (numbered 1 and 6 on Fig. 5), the negative peaks (3 and &), and the zero crossings (2 and 5). A signal loca- tion is based on a single point or on a combin- ation of points. The from the (B - 1 = (Bz - = (B3 - = (B4 - = (B5 - = (B6 - distance between collars is determined recorder trace by using Eqs. 1. A.~). A "~ AZ)' A A3)' A A4)' A As)' A A6) · A Al+ 6 AZ+ 5 A 3+4 Al+ 4 A3+ 6 A1 + 3 + 4 + 6 A1 + 2 + 3 + 4+ 5 + 6 where Ai = the distance between collars using reference point i A : L/(BL - AL) AL = the calibrated distance of log travel between displacement marks BL - AL = the measured chart distance between displacement marks Ai = location of signal produced by collar A based on reference point i Bi = location of signal produced by collar B based on reference point i i = a reference point or a combination of two, four, or six reference points. The reference points could be read to the nearest hundredth of an inch on the chart paper. For the first case in Table 1, in which 2.5 in. of sonde travel equaled i in. of chart travel, : (A~ + A6)/Z : (Az + A5)/Z = (A3 + A4)/Z : (A1 + A4)/Z : (A3 + A6)/Z = (A1 + A3 + A4+ A6)/4 = (A1 + AZ + A3 + a4+ A5 + A6)/6 _~ the sonde position was, therefore, determined to the nearest 0.025 in. The mean joint length and the standard deviation of %he individual measurements from the mean length for each set of runs were computed for each of the above possibilities. The mean joint lengths are enumerated in Table 2, and the standard devia- tions are shown in Table 3. As would be expected, the standard deviation of the estimated joint length declines as more discrete reference points are combined to establish the location of a given collar locator response. When the zero crossing points were used in combination with the four peaks, no noticeable reduction in standard deviation was produced. Furthermore, base line drift could make the determination of the zero crossing difficult or PR~qlSE JOINT LENGTH DETERMINATION USINg- A .ULTIPLE CASING COLLAR LOCATOR TOOL SPE 5087 meaningless. Thus, zero crossings were deleted in subsequent analyses. When axial stress is applied to the casing or if its temRerature is changed, a change in joint length is expected. After accounting for these expected effects, the logging measurements indicated that neigher mechanical stress nor thermal expansion produces any unexpected changes in the response of the magnetic casing collar locator in the casing. No effect due to the simulated cable or the real cable could be detected. Since the reference points could only be determined within about 0.01 in. on the recorder trace, a reduction in recorder speed led to a greater ratio of tool travel to chart length and thus to a greater uncertainty in calculated joint length. A similar effect was produced by increasing logging speed without a commensurate increase in recorder speed. An increase in logging velocity did Proportionally increase the amplitude of signal without side effect. The sheave position did not influence the response provided the tool was free to travel near the wall of the casing. Five different positions were utilized in Runs 41 through 45. The variability of these data was substantial. A closer inspection of the five runs showed that the runs made with a centered cable indicated a different joint length from those runs for which the cable was near the edge of the casing. The centered runs were different since the magnetic force, which held the tool near the wall, was overcome by the cable tension to produce a tilting of the tool in the upper portion of the casing. This cosine effect is only of consequence if a lateral pull is applied to an end of the tool and if the tool length is not significantly greater than the diameter of the casing. The actual waveform produced by the casing collar locator was not distorted by any of the factors mentioned above. When the collar design was grossly changed, the character of the signal did change. A steel collar, which provides a 2&-in. separation between the joints that thread into the collar, produced a signal similar to that shown in Fig, 4. The second trace in Fig. & is that produced by the locator when passing through a 24-in. long nonmagnetic collar. These signals offer no particular advantage in signal interpretation. The long collars pro- vide a major disadvantage in that they are a nonstandard item. The second trace in Fig. & does show an amplification of two peaks relative to the other two peaks. The casing joint ends in the nonmagnetic collar were observed to have definite magnetic poles of opposite polarity that are velieved to have caused signal dis- tortion. Since some collar locator field data contained similar distortions of the signal, an attempt was made to produce distorted waveforms in the laboratory. A conventional collar was parted and a copper spacer was inserted into the cut when the string was installed in the load frame. Different strength magnetic fields were imposed across the nonmagnetic copper spacer. Typical waveforms produced by variable magnetic coupling are shown in Fig. 5. The analysis of the logs produced by passing the sonde between the parted collar and a normal collar were completed using the available peaks in the dis- torted waveform. The. mean lengths for the four distorted waveform cases are given in Table ~, and the standard deviations of the individual measurements are shown in Table 5. The mean values of the estimated joint length varied several tenths of an inch from what was believed to be the best.estimate of the length (the four peak analysis on the 5c signal). The standard deviations were generally a factor of two to three greater than for the runs in the normal collars. The individual peaks in the collar locator signal did not appear to migrate axially as the signal lost its characteristic four peaks. THE MULTIPLE CASING COLLAR LOCATOR A new multiple casing collar locator (MCCL) to measure the length of a joint of conventional casing was designed and built. The measuring element is a pair of Schlumberger magnetic collar locators separated by a spacer of approximately the length of the casing joint. The two locators will respond nearly simulta- neously to the two collars at the ends of a joint of casing. The new MCCL tool contains two pairs of collar locators as shown in Fig. 6. This arrangement allows two measurements of casing joint length to be recorded for each logging run. At the same time, the instanta- neous velocity of the tool can be estimated by determining the time for each pair of closely spaced locators to pass a given collar. The length of the casing joint is calculated with Eq. 2. il-3 = L1-3 [1 + ~ (T- 74)]+ 61_3 , (2) where Jl-3 = length of the measured casing joint LI_3 = length of the ~ECL tool between magnetic centers of the first and third collar locators mea- sured at 74OF ~ = average coefficient of thermal expansion between collar loca- tors 51_3 = distance logged between the first and third waveform locations determined by cross-correlation (positive if the third waveform occurs first, negative if the first waveform is the initial response) T = temperature of the tool at the SPE ~0~ time of joint length measurement A similar procedure can be used to calcu- late casing joint length from responses of the second and fourth collar locators. The long spacer between the second and third locators is made of Invar steel to mini- mize the coefficient of thermal expansion of the tool. The coefficient of thermal expansion for the Invar spacer is 0.000606 in./°F, where- as the coefficient for a tool with a mild carbon steel spacer would be 0.0035 in./°F. Thus, un- certainties in tool temperature will have a relatively minor effect on calculated casing joint lengths. The data produced by the tool can be analyzed by picking reference points from an analog recording as well as by employing cross- correlation techniques on digitized data. The data from all four locators are digitized as the log is run. The sampling process is triggered by cable displacement, each locator signal being digitized once per 0.066 in. of cable recovery. A convolution filter designed to produce a sharp cutoff without ringing is applied to these data records. The filtered data from the four locators are then cross-correlated with a single waveform produced by a locator in a collar of the same design. The cross-correlation coeffi- cients in the neighborhood of the maximum coefficient were fitted to a quadratic equation. The maximum value of the quadratic equation for a particular signal is located on the log. The location of the maximum value is thereby deter- mined more precisely than to the nearest digitization point. Only a few percent of the maximum cross-correlation coefficients are less than 0.9. Collar positions determined from maximum cross-correlation coefficients of less than O.8 are arbitrarily discarded in further analyses. These four locations are used to pro- duce 51_3 and 52_~ as defined in Eq. 2. FIELD TEST OF ~HE NEW MCCL TOOL A performance test of the MCCL tool and ancillary equipment, under simulated field conditions, was implemented to resolve the following questions. 1. Could accurate, reproducible collar signals, such as those generated during lab testing, be obtained in a simulated test well? 2. Were surface film and digital recording devices and drive systems functioning optimally? 3. Would the locator tool detect casing strains in the field of the order of magnitude desired? This performance test would employ logging rig-up methods and operative procedures as close as possible to those norma]_ly encountered in the field. A special test string was run in EPR Co.'s test well at Friendswood, Tex. Design of this subsurface string is shown in Fig. 7. The maj~or components include a bottom section of four joints of 9~-in. OD drill collars, 900 ft of 5-in. drill collars, a bumper sub, a speci- ally built cross-over sub that holds a wireline- retrievable plug, and an upper &80-ft section of 5~-in., 17-1b/ft test casing with standard API buttress collars. These 12 joints of casing were cut to a length of 39 ft, 7 in. with a tolerance of ±0.01 ft and included triangular scribe lines to facilitate making up the joints to very close tolerances. To determine the strain measurement capability of the Schlumberger tool, a two-stage test was conducted. The first portion of the test, the base unstressed case, had the 5~-in. casing hanging under its own weight and filled with fluid. In the second portion of the test, the casing supported an increased tensile load. These two conditions were achieved by having the string designed with the series of drill collars below the casing separated from the casing by the bumper sub. The entire assembly was first landed on bottom and the bumper sub barely disengaged to allow the casing to hang under its ownweight for the first portion of the test. The second portion of the test utilized an hydraulic jacking technique. At round level, the existing casinghead flange Fig. 7) was modified to accept special hydrau- lic jack assemblies. An upper flange unit was supported on the jacks and included a bowl and slips to hold the string in position for jacking The 5~-~in. casing had a mark scribed on it to permit monitoring of movement as load was applied. The hydraulic system included an attached gauge calibrated in pounds force to directly measure the stress level. A precision cathetometer at the surface tracked the displace- ment. Using the jacking assembly, the drill collar load was applied to the casing to imple- ment the last part of the test. The weight of the unfilled casing was approximately 8,500 lbs. Upon jacking, an additional 63,000-1b load was applied as measured by the integral gauge. This corresponds closely to the buoyant weight of the drill collar string as recorded by the weight indicator on the rig floor. For the logging runs, the Schlumberger wireline unit was placed approximately 120 ft from the center line of the well. A rigid lower sheave assembly on the rig floor was used with an upper sheave mounted approximately 100 ft higher. Six logging runs with the casing in the unstressed base condition and five logging runs with the casing in the stressed condition were performed to determine the MCCL's signal PRr 'SE JOINT LENGTH DETERMINATION USINC ~LTIPLE CASING COLLAR LOCATOR TOOL s?w, reproducibility and the tool's ability to detect stress changes of approximately 13,000 psi. Ail logging runs with the pipe in both the stressed and unstressed states went smoothly. An on-site analysis of the traces generated showed that smooth, reproducible signals had been generated. On-site inspection of temperature logs run prior to each set of MCCL logs revealed that the temperature gradient in the well was invariant during the course of the logging process. ANALYSIS OF FIELD TEST DATA Field test data were recorded in both analog and digital formats. The analog traces were recorded on conventional photographic paper as shown in Fig. 8. These traces were inspected immediately to verify log quality and later were analyzed by picking reference points as pre- viously described. A displacement marker on the analog record produces one spike for each inch of cable takeup. Casing joint lengths were calculated using Eq. 2, where 51_3 = (inches of chart between reference points for collar loca- tore one and three). A, and A = (inches of cable takeup per inch of chart determined from displacement marker spikes in the neighborhood of the signal of interest). Computed statistics for the free-hanging pipe are shown in Table 6. The logs taken when the casing string was subjected to additional tension were handled in a similar way; statis- tics are given in Table 7. The average stand- ard deviation of the individual measurement for both the stressed and free-hanging case was 0.0151 in. The measured length changes produced by the loading are shown in Table 10. The average elongation was 0.193 in. The digitally recorded data were analyzed with the cross-correlation procedure described previously. Over 95 percent of the cross- correlation coefficients were greater than 0.92. Correlation coefficients less than 0.92 were generally caused by starting the digital recorder too late or stopping too soon, such that an incomplete record was obtained. The digital analysis did indicate a slight but con- sistent difference in the pairs of joint length measurements. The distance between magnetic centers of the first and third locators was determined to be 0.O1 in. greater than the dis- tance between magnetic centers of the second and fourth locators. This small difference in tool length was incorporated into the analysis so that consistent values of casing length could be measured with either set of collar locators. Results of the analysis of digital data are shown in Tables 8 and 9 for the free-handing and stressed cases. The average standard deviation of the individual measurement for both sets of data was 0.0072 in. Hence, the digital tech- nique, which Uses roughly 50 samples from each waveform, produces a 50 percent reduction in standard deviation relative to the analysis that utilizes the two positive peaks of the analog record. Table 10 shows the length changes based on digital analysis. The average elongation was 0.19~ in. This agrees excel= lently not only with the value of 0.193 in. determined from manually processed analog data, but also with the calculated elastic elongation of 0.206 in. The calculation of elastic elonga- tion is based on an estimated total applied force measured by the hydraulic system and nominal casing dimensions. CONCLUSION AND SUMMARY The sensitivity of magnetic collar locators to a number of variables normally encountered in a logging operation was investigated. The response of the locators in a laboratory environment indicated no unexplained effects of mechanical stress, thermal expansion, logging velocity, radial position of the sonde in the casing, collar design, or cable cross-talk. The only factor that produced a noticeable influence on the signal was the degree of magnetic coupling in the collar. A multiple casing collar locator system has been developed that can detect in-situ casing strain with a precision of essentially a hundredth of an inch. This quality of perform- ance was attained via a digital recording of the signal and applying a cross-correlation tech- nique to the data. This performance is produced by using only five or six replicate logs. ACKNOWI.E~GMENTS The authors would like to acknowledge the contribution of personnel of the Schlumberger Co. in the development of the MCCL tool and recording equipment. REFERENCES 1. Allen, D. R.: "Collar and Radioactive Bullet Logging for Subsidence Monitoring," SPWLA Tenth Annual Logging Symposium, May 25-28, 1969, p. G. 2. Allen, D. R.: "Physical Changes of Reser~ voir Properties Caused 'by Subsidence and Repressuring Operations," J. Pet. Tech. (Jan. 1968) 20, 23-29. 3. Davies, B. E. and Boorman, R. D.: "Field Investigation of Effect of Thawing Perma- frost Around Wellbores at Prudhoe Bay," paper SPE ~591 presented at SPE-AIME ~Sth Annual Fall Meeting, Las Vegas, Sept. 30- Oct. 3, 1973. Run No. 1-5 6-10 11-i5 Cable None Simulation 18, 000' TABLE 1 - SUM~t&RY OF LABORATORY LOGGING PROGRAM Noise on Conductors cps Stress Temp. psi OF 0 75 I! Sheave Position South !! Ve loc ity ft./hr. 750 Chart Speed !~/ sec. 1 16-ZO 21-25 26 -30 31-35 0 + 7780* " " " " II 0 II II II II " - 12725* " " " " 11 0 II II II II 36-40 " " " 155 41 -45 " " " 75 46-50 " " " " 56-60 " " " " 11 Variable South Il Il 900 750 II II I! 0.5 61-65 " .... " 66-70 " " " " 71-75 " " " " West East North ]. , II II 162-166 6,400' " " " South " " 167-171 " Z " " " " " 172-176 " 0 " " " " " + Compression - Tension TABLE 2 - Y~m_AN JOINT LE~{GTH DET~RMINAT!O~S FROM LABORATORY STUDIES (INCHES) 1Kuns h1 AZ 23 A4 A5 26 A1 + 6 A3 + 4 AZ + 5 A1 + 4 A3 + 6 51+3+4+6 51+2+3+4+5+6 1-5 36.86 36.91 36.95 36.8Y 36.91 36.89 36.88 36.90 36.91 36.87 36.93 6-10 36.87 36.84 36.85 36.84 36.88 36.84 36.93 36.85 36.86 36.85 36.89 11-15 36.88 36.87 36.85 36.8l 36.87 36.93 36.90 36.83 36.87 36.85 36.89 36.90 36.90 36.87 36.87 36.87 36.87 16-20 36.87 36.84 36.87 36.83 36,88 36.92 36.89 36.84 36.86 36.85 36.89 21-25 36.87 36.87 36.87 36.88 36,92 36.95 36.91 36.87 36,90 36,87 36.91 26-30 36.87 36.91 36.88 36.88 36.95 36.99 36,93 36.88 36.93 36°88 36.94 31-35 36.87 36.88 36.88 36.87 36.91 36.94 36.91 36,87 36.89 36.87 36.91 36.87 36.86 36.89 36.89 36.91 36.91 36.89 36.89 36-40 36.91 36.91 36.87 36.85 36.91 36.98 36.94 36.86 36.9~ 36.88 36.92 41-45 Not Analyzed. See discussion in text, 46-50 36.81 36.83 36.85 36.8g 36,93 36,97 36.89 36.83 36.88 36.82 36.91 36,90 36.90 36.86 36.87 56-60 36.79 36.77 36.90 36.88 36.83 36.88 36.83 36.89 36.80 36.84 36.89 61-65 36.69 36.75 36.82 36.82 36,82 36.86 36.78 36°82 36.78 36.75 36.84 66-70 36,77 36.76 36.73 36.86 36.84 36.80 36.79 36.77 36.75 71-75 36.$4 36.79 36.86 36.'76 36.85 36.85 36.85 36.81 36.82 36.80 36.86 36.83 36. 83 162-166 36.73 36.83 36.83 36.76 167-171 36.82 36.99 36,96 36.66 36.69 36.62 36.72 36.81 38.84 36.74 36.79 36.76 36.78 36.79 172-176 36.75 36.81 36.85 36.78 36.80 TABLE 3 - STANDARD DEVIATION OF INDIVIDUAL JOINT LENGTH DETERMINED IN LABORATORY STUDIES (INCHES) Runs A1 AZ A3 A4 A5 A6 A1 + 6 A3 + 4 AZ + 5 A1 + 4 A3 + 6 A1+3+4+6 1+Z+3+4+5+6 1-5 .031 .040 .040 .032 .019 .013 .021 ,026 ,022 .033 .011 .017 6-10 .022 .012 .039 .018 .043 .035 .028 .022 ,023 .019 .017 .016 11-15 .035 .042 .021 .025 .011 ·033 .024 ,017 ,019 .025 .021 ,016 · 020 · 017 · 015 16-20 .033 .039 .020 .030 .055 .046 .029 .024 .015 .017 .024 .017 21-25 .053 .026 .030 .043 .035 .041 .042 ,018 .029 .OZl .029 .024 26-30 .012 .019 :034 ,025 .025 .038 .018 .027 .017 ,013 .015 .009 31-35 .030 .024 .024 .031 .015 .017 .017 .023 .011 .015 .007 .008 · 01Z · 025 .009 · 009 36-40 .023 ·014 .043 .024 ·006 ·030 ·023 ,022 .006 .011 .020 .010 41-45 Not Analyzed. See discussion in text. 46-50 .039 .038 .058 .038 .062 ,061 .048 ,035 ,047 .033 ,044 .034 · O06 · 038 56-60 .042 .048 .049 .075 .071 .070 .030 .050 .050- ·035 ·039 .035 61-65 ,026 ,051 .026 ·047 ·034 .044 .028 ,033 .042 .037 .028 ' .028 66-70 ,040 ,065 .051 .050 · 038 . 032 · 050 71-75 ·042 .035 .086 ,030 ·020 .047 ,041 ,036 .021 .023 .033 .014 · OO8 162'166 ' .030' .0i0 .016 .015 167-171 .039 .031 .049 .034 .OZZ .055 ·029 .038 .025 .033 .046 .030 172-176 .041 .044 .023 .OZ4 · 014 · 025 .022 Figure 5 Designation TABLE 4 - MEAN JOINT LENGTHS FOR DISTORTED SIGNALS DETERMINED IN LABORATORY STUDIES (INCHES) Al A3 A4 A6 A1 + 6 A3 + 4 Al + 4 A3 + 6 A1 + 3 + 4 + 6 a 37.04 36.91 36.97 b 37.03 36.97 36.99 c 36.79 36.99 36.70 36.75 36.77 36.84 36.74 36.87 d 36.75 36.97 36.67 36.82 36.7! 36.81 Mean 36.77 37.00 36.69 36.88 36.77 36.83 36.73 36.94 36.81 Figure 5 De signation TABLE 5 - STANDARD DEVIATION OF INDIVIDUAL JOINT LENGTHS FOR DISTORTED SIGNALS (INCHES) A1 A3 A4 A6 A1 + 6 A3 + 4 A1 + 4 A3 + 6 A1 + 3+ 4+ 6 a .047 .045 .027 b ·036 .070 .031 c .029 ·053 .034 .100 .059 .025 .0Z0 ·054 d .040 .055 .058 .055 .039 · 034 TABLE 6 - FREE HANGING CASING JOINT LENGTH MEASUREMENT USING MCCL ANALOG RECORDING Joint No. 2 3 4 5 . 6 7 8 9 10 Difference in Joint Length from Spacing of Collar Locators 1 and 3 Difference in Joint Length from Spacing of Collar Locators 2 and 4 Mean Length* Mean Length** Standard Deviation (inche s) (inc he s) ( inche s) Standard Deviation (inches). 0076 0100 0057 0200 0173 0087 0113 0129 0185 0251 01894 - 0626 - 0226 0768 - 2703 - 0599 0469 - 0355 .0194 .0933 -.00494 -.0837 -.0349 .0413 -.2783 -.1112 -.0573 -.0663 .0055 .0571 0149 0065 012 O104 0139 0138 0071 0122 0075 0048 ,:,Distance between magnetic centers of collar locators 1 and 3 measured at 74°F = 475.01 inches. **Distance betwee, n magnetic centers of collar locators Z and 4 measured at 74°F = 475 inches. TABLE 7 - STRESSED CASING JOINT LENGTH MEASUREMENT USING MCCL ANALOG RECORDING Difference in Joint Length from Spacing of Collar Locators 1 and 3 Mean Length* Joint No. (inche s) 1 .Z10 Z .155 3 .200 4 ·251 5 -. 063 6 .113 7 .160 8 .145 9 . Z04 10 .259 Standard Deviation (inches) Difference in Joint Length from Spacing of Collar Locators 2 and 4 Mean Length** (inches) Standard Deviation (inches) 008 . Z07 .0Z0 017 .131 .016 0Z0 .178 .007 012 .206 .016 008 - . 069 . 0ZZ 020 .084 .012 · 026 . 133 . 018 · 018 . 114 . 010 · OZ1 .195 .017 · 017 . Z21 .01Z *Distance between magnetic centers of collar locators 1 and 3 measured at 74°F = 475. 01 inches· **Distance between magnetic centers of collar locators Z and 4 measured at 74°F = 475 inches. TABLE 8 - FREE HANGING CASING JOINT LENGTH MEASUREMENT USING MCCL DIGITAL RECORDING AND PROCESSING Difference in Joint Length from Spacing of Collar Locators 1 and 3 Difference in Joint Length from Spacing of Collar Locators Z and 4 Mean Length* Standard Deviation Mean Length** Standard Deviation Joint No. (inches) (inches) (inches) (inches) 1 . 018 . 003 . 005 . 003 Z - . 058 .004 -. 06'5 .006 3 - . 031 . 004 - . 046 . 003 4 .064 .004 .055 .002 5 - . 256 . 009 - . 262 · 009 6 - . 078 . 004 - · 098 . 017 7 - . 055 . 005 - . 06Z . OOZ 8 - . 042 .010 -. 053 . 007 9 . 0ZZ . 007 . 01Z . 007 l0 ,. 077 . 006 . 060 . 005 *D,istance between magnetic centers of collar locators 1 and 3 measured at 74°F = 475· 01 inches· **Distance between magnetic centers of collar locators Z and 4 measured at 74°F = 475 inches· TABLE 9 - STRESSED CASING LENGTH MEASUREMENT USING MCCL DIGITAL RECORDING AND PROCESSING Difference in Joint Length from Spacing of Collar Locators 1 and 3 Difference in Joint Length from Spacing of Collar Locators 2 and 4 Mean Length* Standard Deviation Mean Length** Standard Deviation Joint No. (inches) (inches) (inches) (inches) 1 .ZlZ .005 .199 2 . 146 . 003 . 148 3 .182 . 002 . 172 4 . 259 . 005 . 241 5 - . 048 .005 -. 054 6 . 100 . 002 · 081 7 .150 .003 . 131 8 . 156 . 013 . 140 9 .187 .010 . 183 10 .253 .010 .244 OOZ OOZ 005 002 014 018 004 010 004 007 *Distance between magnetic centers of collar locators 1 and 3 measured at 74°F = 475.01 inches. **Distance between magnetic centers of collar locators Z and 4 measured at 74°F = 475 inches. TABLE 10 - CHANGE OF JOINT LENGTH DUE TO INCREASED AXIAL STRESS Analog Analysis Digital Analysis Joint No. (inches) (inches) 1 . Z01 .194 Z .216 .208 3 .218 .Z15 4 .169 .191 5 .208 .208 6 .185 .180 7 .197 .199 8 .180 .195 9 .188 .168 10 .165 .180 Average .193 .194 TIME Fig. 1 - Typical collar Iocatop pesponse fop a coupling, CABLE HYDRAULIC CYLINDER DIAL GAUGE (3) COLLAR ~ 5t/2// CASING COLLAR LOGGING TOOL DISPLACEMENT MARKER CABLE DRUM LIMIT SWITCH (2 Fig. 2 - Assembly to test collar Iocator. AL COLLAR A COLLAR B / 1 6 1 6 2 4 4 Fig. 3 - Example trace (shortened). LLAR NON MAGNETIC COLLAR Fig. 4 - Waveforms produced by extra long col lars. EXTERNAL MAGNETIC FIELD ,~.os.D ~o ~-¥ ,,~ co..^. D EXTERNAL MAGNETIC FIELD IMPOSED TO COUNTER POLARITY IN COLLAR INCREASED EXTERNAL MAGNETIC FIELD STRENGTH IMPOSED TO COUNTER POLARITY IN COLLAR Fig. 5 - Waveforms for parted collar. LOCATOR NO. I LOCATOR NO. 2 LOCATOR NO. 3 LOCATOR NO. 4 LI..I NO. i RESPONSE NO. 2 NO. 3 LOCATOR Fig. 6 - Multiple casing collar Iocator. NO. 4 BOWL AND SLIPS.....~__ FORCE GAUGE HYDRAULIC PUMP PPER FLANGE ASSEMBLY HYDRAULIC .,lACKS WER FLANGE ASSEMBLY CASING HEAD FLANGE ~ 57~~/TEST CASING --CROSSOVER SUB --BUMPER SUB --5n DRI'LL COLLARS ~ 91/2' DRILL COLLARS Fig, 7 - Field test configuration. LOCATOR NOS. DISPLACEMENT MARKER Fig, 8 - Typical MCCL analog record. Paper No. 75-Pet-36 be given lo A$/~E, th~ $3.00 PER COPY $1.00 TO ASME MEMBERS dit should Division, and ' ' · ' F. W. Ng Atlantic Richfield Company, Anchorage, Alaska Removal of Water Base Mud From Well Bores in Permafrost Zones In completing an Arctic oil or gas well, it is sometimes desirable to remove water-base drilling mud from a casing-casing or casing-tubing annulus to minimize the possibility of casing damage resulting from freezing and thus expansion of the fluid. The proposed field procedure involves an initial water wash step, where the mud is washed from the annulus by water, followed by the pack placement step where the water is displaced by an oil-base gelled and weighted casing pack. The casing pack fills the entire annulus and also acts as a thermal insulator by suppressing convection currents with its high .gel strength. The study was performed to examine the fluid displacement mechanics in- volved in the field operations and to determine the methods of operation which should be adopted. The operations were studied in 10-ft long model annuli of various sizes in the laboratory and then in a 900-ft test hole in the back.yard. Findings from these ,~tud- ies were then applied to the displacement procedures in a number of field operations at Prudhoe Bay, Alaska. Data gathered from these field operations were analyzed, and they compared satisfactorily with predictions from laboratory data. It was found that, for the water-wash step, two system volumes of water was sufficient to remove virtually ali the drilling mud from the annulus. For the pack placement step, two system volumes of pack was sufficient to remove virtually all the water. Both displacements should be carried out with the water in turbulent flow. A system of commercially available electronic in- struments was developed for monitoring water contamination of the pack returns. It provides instant indication of excessive contamination in case of displacement failure due to equipment breakdown or other unforeseen circumstances, so that remedial ac- tions may be taken immediately. IntroductiOn In completing a well in the Arctic region, it is sometimes desir- able to remove water-base fluids from the permafrost section, which usually extends from the surface down to a depth of about 2000 ft. When cooled by the permafrost, the pressures generated by the expansion of such fluids upon freezing are sometimes capa- ble of damaging the casings in a well. Fig. 1 shows a possible Arctic completion, in which the permafrost section of the 9% X 13% in. casing-casing annulus and the 51/2 x 9% in. tubing-casing annulus may be filled with fluids which would not freeze at permafrost in situ temperatures. The purpose of the present study is to dete/- mine, by experiments, the most efficient combination of existing field procedures and existing special fluids for the removal of the water-based fluids and their replacement by nonfreezing fluids. The problem of main interest in the field concerns the 9% X 13% in. casing-casing annulus. The field procedures studied involve an initial water-wash step, where the mud is washed from the annulus by water, followed by the pack placement step where the water is displaced by an oil-base gelled and weighted casing pack. The cas- ing pack is mixed at 60°F or above, according to the following for- mula (for 100 bbls at 9.5 lb/gal): Contributed by the Drilling and Production Committee of the Petroleum Division for presentation at the Petroleum Mechanical Engineering Confer- ence, Tulsa, Okla., September 21-25, 1975, of THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS. Manuscript received at ASME Head- quarters June 25, 1975. Paper No. 75-Pet-38. Copies will be available until June, 1975. Diesel 3450 gal EZ Mul (an emulsifier) 150 gal Saltwater (1 lb/gal NaC1 solution) 200 gal Geltone (oil wettable bentonite) 2500 lb Barite (weight material) 10,300 lb Discussion on this paper will be accepted at ASME Headquarters until October 27, 1975 20" 94t I-{40 az 100' Permafrost bottom at ZOO0' FO cementer at 2400' FO cementer at 2650' 13 3/8" 72~ MNS0 at 2700' Cement Arctic Pak Oil Base Fluid e.g, Diesel Water based drilling mud _ S l/Z" 17// NS0 tubing 9 5/8" 47// SOO at 10,000' Fig. I An Arctic completion This mixture, referred to here as the "premix," is cooled down to 45°F or lower before it is pumped into place. While being pumped, an extra 25 lb/bbl of geltone is added to the premix, forming what will be referred to as the "gelled pack." The Present Study The general approach of the present study was as follows: I The displacements to be studied were: Ca) displacement of water-base drilling mud by water; Cb) displacement of water by the casing pack. 2 Model studies were carried out in the laboratory using annuli of various sizes in order to study the influence of various operating parameters. 3 Displacements were then performed in a scaled-up model in the back yard to investigate any size effect and to produce data for extrapo]atlon to field operations. 4 Field operations were monitored to evaluate the findings of the study. Conceptual Studies. A number of displacements under ]ami- nar flow conditions were carried out in a 5 ft long 2 X 1.6 in. model concentric annulus. Values for various parameters are specified in Table 1. From the results shown in Fig. 2, it can be seen that, even with the water in laminar flow at a Reynolds number Rew of about 300, 93 percent of the mud and 98 percent of the water was dis- placed by circulating 2.5 annulus volumes. It was then postulated that a more efficient displacement pro- cess may be obtained by displacing under turbulent flow condi- 0.5 1.0 1.5 2.0 2.5 3.0 t~V / ¥,yI Fig. 2 Displacements In 5-ft model Water dilptaced by Arctic Pak Water ba,ed mud dieptaced by Rater tions, where the turbulent eddies would produCe a better cleaning action at the walls of the annulus, and the more square velocity profile would minimize the channeling of one fluid through the other. Due to its small volume it would not have been practical to evaluate turbulent flow displacements in this 5-ft model. Thus, a larger model was built to conduct these tests. Theoretical Considerations. Laminar and turbulent flow cri- teria have been developed by Lohrenz and Kurata [l.]~ for concen- tric annuli, which at radius ratios above 0.3 and 0.5, respectively, fl)r laminar and turbulent flows may be approximated by flow be- tween parallel plates [2]. Reynolds number for flow in an annulus is defined by Lohrenz and Kurata as Re- where D~ = a (Do - Dc) a is a function of radius ratio and is 0.816 for parallel plates. With Reynolds number so defined, these authors showed that turbulent flow in an annulus occurs at Reynolds numbers above 2000-3000 for both water and air. At normal displacement rate in the field of' 3 bbl/min in 9% X 13% in. casing-casing annulus, the Reynolds numbers for water Re~ is 16,000. Due to the limited pump rate available in the laboratory, no attempt was made to duplicate the field values of Re~ in the laboratory model, but flow regimes, i.e., whether turbulent or laminar flow, were matched. Displacement efficiency in any model was obtained by forward difference inte- gration of effluent concentration with respect to the volume pumped. Backward and central difference integration were also computed, but these results were within I percent of the forward difference integration. Laboratory Model Studies Description of the Models. Displacement tests in the labora- l Numbers in brackets designate References at end of paper. ,, Nomenclature Unless otherwise specified, the following symbols are used throughout this report: C = concentration, percent D = diameter, in. A = change ~ = displacement efficiency, volume recov- ere&volume originally in place H = gap size, in. K -- radius ration, Dc/Do ! ,i L = length of system ge = effective viscosity P = pressure, lb/in3 PV = plastic viscosity, centipoise p = density, slug/ft:~ Re = Reynolds nt, mber U = velocity of fluid parallel to walls, ft/s V = volume, barrels YP = yield point, lb/100 ft2 Subscripts a = average an = annulus E = equivalent e = effective i = inner pipe o = outer pipe sys = system = inner pipe plus annulus tv = water wbm = water-base mud 2 Transactions of the ASME Table I Model parameters for conceptual study }!adc, 1 Si.'.c 2" x 1,6" x 5'2" x 1.6" x 5' PV centipoise 50 51 YP lb./100 ft.2 22 41 l0 second ~.el lb./103 £t.2 3 21 10 ninute gel lb./100 £t.2 d 26 9en>it¥ lb./aal. 11 .S 0.7 Flow llatc gal./m~n. 0.$S 0.3S tory were carried out using a model with a nominal length of 10 ft, which was designed to accommodate nominal 2 or 3 in. OD and nominal 1, 15/s, 2, or 21/~ in. ID. The layout of the facilities was as shown in Fig. 3. To measure the composition of the effluent in the- water-water based mud displacements, a mixing orifice and a resis- tivity cell were used, as shown in Fig. 3. For the pack-water dis- placements, the total effluent from each run was collected in about 30-40 batches of about 500 cc each. The contents of each beaker were then stirred thoroughly and a sample was distilled to measure relative oil-water contents. The volume and composition of the contents of these beakers were then used to calculate the displace- ment efficiency at each sampling point. The Water-Wash Step. Displacement of the water based mud by water (the water-wash step) was studied in the laboratory model using six different size annuli, as summarized in Table 2. The average properties of the mud used in these experiments were within the range of the following field specifications: Plastic viscosity, centipoise 26-32 Yield point, lb/100 ft2 9-14 0/10 min gel, lb/100 ft2 2-4/5-8 Weight, lb/gal 11.0-11.3 Fig. 4 shows a plot of typical displacement efficiency and mud concentration Cwbm in the effluent for runs in Group I from Table 2. From these data, the following conclusions may be drawn: I Comparing Group I with Group V and Group II with Group VI in Table 2, it is apparent that for given values of Rew and radius ratio K, the gap size of an annulus has a Significant effect on dis- placement efficiency. The average volume pumped at 95 percent recovery of mud was changed from 1.3 annulus volumes to 1.6 in the first case and 1.75-2.3 in the other. 2 For a given annulus geometry, variation of Reynolds number within the turbulent range from about 4000 to 10,000 has no signif- icant effect on displacement efficiency. Cwbm 4O-- 20-- Annulu~ 3. Gap O. 70" Run ~ RCwbm 266 288 3 Cwbm 0 A ~ 1 [ I °u} o V o, vov ~{ v , 1,0 Z.O 3.n AV Fig. 4 Displacement efficiency and effluent concentration for water-wash in laboratory model, Group I 3 Comparing Group III with Group V shows that, if gap si=e is essentially constant, radius ratio does not significantly affect dis- placement efficiency. This is to be expected since, for K larger than 0.3, all annuli can be approximated by parallel plates with the same gap size. 4 Cornpared to the laminar flow displacements in the concep- tual studies, these turbulent flow displacements are more efficient processes. Virtually complete removal of the mud is obtained with 2.5-3.0 volumes of water pumped. The Pack Placement Step. In the field, with temperatures below 45°F, the theology of an unyielded gelled pack is not very different from that of the premix at room temperature. For the present study, it would be impractical to carry out the laboratory and back yard displacements at temperatures below 45°F; only the premix was used in these studies which were carried out in an av- erage room temperature of about 72°F. The premix was prepared. according to the same formula as that used in field operations. Plots of efficiency and effluent concentration for a typical pack-. water displacement in the laboratory model are shown in Fig. 5. The test runs are summarized in Table 3. The effluent from these displacements was collected in a large number of small containers as described before; thus, in order to prevent splashing and loss of fluids, Re~, had to be limited to about 4000. From these data, the following conclusions may be drawn: I Within the range of values examined, the behavior of the dis- placement process is not significantly affected by Reynolds num- ber or annulus size. Note that of the gap sizes used here, one (0.584 '.~Pr cs s .£fluent Storag,: l~ipe -~- ~ valve Gear ~alvc pump Drain Fig. 3 Laboratory model test arrangement 100 a k Annulus 3. 068" x 1.900" Size 80 · Gap Size 0,584 / Run ii 26 Z7 60 & Rew 2200 3120 ' Re 67 1 18 40 C & O w (%) 20 O o ~ I 1.0 2.0 3.0 AV/Van Fig. 5 Displacement efficiency and water concentration for water-pack displacement in laboratory model, runs 26 and 27 Journal of Engineering for Industry 3 Table 2 Summary of water-mud displacements in the laboratory model Annulus Gap Radius Reynolds Number Group G Size Size for tVa:er for Hud Run Nos. DO x Di (DO . Di)/2 RatioK Rew R~bm Average Volume Pu~pcd 0 nan · g55 ! 5.068" x 0.701" 0.54 4,700 to 170 to (1-3) 1.660" 10,250 290 II 3.06S" x 0.$50" 0.75 3,000 to IlO to l.?S (7-8) 2.375" 7,000 250 III 3.068" x 0.584" 0.62 2,750 to 100 tO l.SO (g-Il) 1,900 6,000 220 1V 3.068" x 1.009" 0.34 3,700 to 80 tO 1.4S {12-14) 1.0SO" 9,000 190 V 2.067" X 0,S08" O.S1 4,600 to 105 to 1.60 (15-18) 1.OSO" 10,IS0 305 Vl 2.067" x 0.204 0.80 3,850 to 150 to 2.30 (19-21) 1.660" 9,200 250 Table 3 Summary of casing pack-water displacements in the laboratory model A~nulus Gap Radius Reynolds Number Size Size Ratio tVatet Pack k£f'icicncy Oo x Oi (DO - Dj)/2 K Rew Re° Break-through 26 5.068" X 0. 584" 0.62 2200 67 96.4', l.g00" 27 3.068" x 0.584" 0.62 3120 118 95.8% l.gO0" 28 3. 068" x 1. oog" 0.34 4363 78 95.3% I.OSO" 29 3,068" X l. 009" 0.34 2980 37 95..1% 1.050' Back 5.500" x 0.565" 0.67 16000 @ 0.8 bbl/min yard 2. Field 12.347" 1.$61" 0.779 16000 0 5.0 bbl/min in.) corresponds to the gap size used in the back yard model (0.563 in.), while the other (1.009 in.) approaches the field value (1.361. in.). 2 The highly gelled emulsion at the water-pack interface ob- served in some of the early field operations did not appear in any of the laboratory displacements. Fig. 6 Back-yard displacement model OCT TYPE TftSA .~ 4 l/z,, ~ z )Is" l'nG~ 2 3/8" - ~I{D EUE 1o' PUP GROUND 4, 7# J55 4 I/Z" CSG. 9.5# 355 9.75' to. 10.75' 5 JT$. ? yrs. 7 JTS. 1 ! JT. PLBS MULE SHOE 4 I/Z" COLL~R 8RD 4 1/2" SCI[. 40 LAP WELDED TO 4" CSO. B&W TYPE KKI?. CENTRALIZER. Lll & RII, z EACH. SLIP-ON TYPE, NOS, 3513300 & 3523400 Z7 JTS. z 3/8" 4.7# 365 ATLAS BRADFORD GST TBG. 10' MULE SHO~ YrS. Z 3/8" 'Jb5 4.7~ AB-GST TBG. ANNULUS 5.69 bbl (6.4 bht/1. 000 ft. ) TUBING 3.44 bdt (3.9 bbI/I,O00 ft.) 3 Practically complete removal of the water is produced in all cases with 1.5 volumes of pack pumped. Scaled-up Back Yard Model Studies Description of the Model. A 900 ft deep test hole with 4 in. casing was completed with 2% in. OD, 4.7 lb/ft tubing as shown in Fig. 6. A triplex pump, on which a mechanical stroke counter was installed, was used to displace all the fluids. Effluent concentra- tion was not continuously monitored, but large numbers (30-40 for each run) of quart-size samples were collected at predetermined Fig. 7 Displacement efficiency and effluent concentration for water-mud displacement in back-yard model, run BY8 100 ~ 80 -- 60 Cwb~ 4O (%) 2O 0 O o~OOOOOO0oO Annulus 3. 500" x 2. 375" Gap 0. 563" Run # BY8 Rew Tbg. 26,633 Ann. 16,230 Rewbm 941 336 ll · Cwbm 0 Forward Displacement J I . J 2.0 3.0 4 Transactions of the ASME Table 4 Summary of water-wash in the back-yard model Rew Tubing Annulus 26,633 16,230 64,365 39,223 28,512 17,375 65,475 39,899 Run Direction of No. Circulation BY8 Down Tubing BY9 Down Tubing BY10 Down Annulu~ BY11 Down Annulus ~o. of Volumes Pumping Pumped at ~=~e 95% Tubin~ReWbmAnnulus bbl./min. 0.98 941 336 0.816 1.00 3,054 1,158 1.972 0.98 941 336 0.874 1,00 3,054 1,158 2.006 Annulus Size 2-3/8" x 3-1/2" Gap Size 0.563" Length of Model 884' Table 5 Summary of pack pJacement In the back-yard model Displacement Run Effective Rew Rea Pump Rate Efficiency at NO._~.,Length A~.lus___~) ~Annulus) bbl./min. ~rcakthroumh BY2 884' 15,912 58 0.8 92% 8Y7 884' 15,912 58 0.8 92~ .BY3 1768' 15,912 S8 0.8 95~ BY4 884' 32,620 181 1.6 95% Annulus Size 2-3/8" x 3-1/2" Gap Size 0.563 intervals. The appropriate stroke count and sample composition were then 'used to calculate displacement efficiency. The Water-Wash Step. The water base mud used in these dis- placements was mixed to produce properties within the range of field specifications as shown before. Effluent samples were ana- lyzed by specific gravity, which varies linearly with percentage mud content. Two forward and two reverse (down the annulus) displacements were carried out. They were summarized in Table 4, and, in particular, displacement efficiency and effluent concentra- tion plots for a typical run are shown in Fig. 7. From these data, the following conclusions may be drawn: I For a given direction of displacement, efficiency was not very sensitive to pumping rate in the range examined, although these displacements were all much more efficient than those in the labo- ratory model. 2 Reverse displacement shortens the mixing zone and was ini- tially a more efficient process, but the net result after 1.3 volumes pumped was about the same. 3 At most 1.5 system volumes pumped in all cases produced virtually complete removal of the water based mud. 4 The scaling up in model sizes between the laboratory model and back-yard model apparently resulted in a more efficient dis- placement process in the latter system. Thus, it may be expected that the data for v and Cwbm from the field would resemble that of the back-yard system. The Pack Placement Step. Four pack-water displacements were carried out in the back yard. Fig. 8 shows displacement effi- ciency and effluent concentration for a typical run. These runs are summarized in Table 5. Except for run number BY3, wiper plugs were used to separate the water and the casing pack in the tubing, so that the fluid to fluid displacement took place only in the annu- lus. The casing pack used was mixed according to the formula de- scribed before the field operations. All the displacements were car- tied out at fluid temperatures of 80-95°F, and no extra geltone was added to the premix. Samples were later analyzed by distilla- tion. From these data, the following conclusions may be drawn: 1 Length of the system was essentially doubled by omiSSion Of the wiper plug, and flow rate was doubled between runs BY2 and BY4. There were no major differences in efficiency and concentra- tion plots for these displacements. In particular, run BY7 repli- cated BY2 almost exactly. 100~ 0 · · · · · S · · · Gap 0. 563 100 Rew 15,912 80 R.co 58 :°I / C Cw0 60 40~ / O No wiper plugs used. Cwbrn / (~°) / / ~o 0 , I °ooooo.4 , l 0 1,0 Z,0 3.0 0 ~V/V Fig. 8 Displacement efficiency and effluent concentration for pack-water displacement in back-yard model, run BY3 Fig. 9 Comparison of field data and laboratory data for water-wash step Journal of Engineering for Industry 5 100 8O 60 Cwbm 40 El A ~ Laboratory Model ·~ ~ Back Yard Model Field _ o/ o , I ~ I, ~ 0 1.0 Z.0 AY/V Rew 3, 120 16,600 Z3.000 Fig. 10 Comparison of field data and laboratory data for pack placement step 2 On a fractional volume basis, there are no major differences between the pack-water displacements in the laboratory model and the back-yard model. Since these model lengths changed by a factor of about 90, and Rew changed by an order of magnitude, it may be concluded that the pack placement operation in the field, where the system length would be about triple that of the back yard, would follow the typical curve shown in Fig. 8. 3 As in the laboratory displacements, no gelled emulsion was observed at the pack-water interface, even though these displace- ments were carried out at higher temperatures. 4 As shown in Fig. 8, the longest 10-90 percent mixing zone is about 12 percent system volume long. Thus, if the gelled pack in the field is preceded by this percentage volume of spacer of pure premix, excessive water contamination of the gelled pack can be prevented. Field Data Field data were gathered by monitoring the displacement opera- tions at a number of wells at Prudhoe Bay. As shown in Figs. 9 and 10, there is close agreement between these field data and the labo- ratory results, both in the water-wash step and the pack placement step. According to the conclusions from the back yard pack placement tests, the 90-10 percent water mixing zone length would stay at about 1.2 percent of system volume. Thus, this percentage volume of spacer of thin premix pack without the extra 25 lb/bbl geltone was used. This volume of spacer was in fact found to be sufficient to prevent excessive water contamination of the gelled pack. The gelled and viscous emulsion of water and pack at the interface of the two fluids, observed during some of the early field displace- ments, did not appear. The maximum water contamination of the pack returns behind the spacer was only 8 percent. The Net Oil Analyzer probe was tested at these wells for its ability to monitor water contamination in the pack returns. It was found to be a very stable and reliable tool for this purpose, and may be used on such displacement processes to insure an instant indication of the quali- ty of the process. The Net Oil Analyzer Probe The Net Oil Analyzer Probe has been used in many production facilities to monitor water content in produced crude oil. Fig. 11 shows the probe and a schematic diagram of the measuring circuit used in monitoring the field displacements. Depending on the composition and, hence, the dielectric constant of the fluid filling the unit, the probe forms a capacitor of varying capacitance, caus- ing an oscillator to generate outputs of varying frequency. By call- NOA Probe ~no inverter To 110 v 60 Hz supply Fig. 11 Schematic o! NOA probe measuring circuit brating output frequency against samples of casing pack of known water contamination, the probe can be used to monitor the el- fluent during pack p]acmnent. The reading is instantaneous and can be taken as often as desired. Readout instruments for the probe have been made available in a compact portable package for routine use in the field. Conclusions 1 There is good agreement between extrapolation of laboratory and back-yard model data and field data on displacement efficien- cy and effluent concentration, both for the water-wash step and the pack placement step. 2 Two system volumes of water pumped at 4 or more bbl/min is recommended for the water-wash step. It will remove virtually all the drilling mud from the system. 3 2.0 system volumes of casing pack, pumped ¢!~ 3 or more bbl/min for ~he pack placement step is recommended. It will re- move virtually all the water left after the water-wash step. 4 A minimum of 20 barrel spacer of ungelled premix can be used between the water and the gelled pack to prevent contamina- tion of the latter fluid. Without this spacer, a highly gelled and vis- cous emulsion is formed by the two fluids at their interface. The emulsion may have a high water content and may lodge itself in the annulus. 5 The Net Oil Analyzer Probe was tested in the laboratory and in the field, and was found to be a reliable tool for monitoring water contamination of the pack returns. Readout instruments for the probe have been made available in a compact portable package for routine use in the field. Acknowledgments The author wishes to express his gratitude toward the following colleagues of Atlantic Richfield Company for their assistance and advice on this project: Mr. C. R. Knowles and members of the Drilling Department of the North Alaska District for their help and cooperation, Mr. L. II. Miles of the Chemical Engineering Department for his assistance with the field studies, and Dr. R. A. Ruedrich for the use of some . of his data from the conceptual study. References 1 Lohrenz, J., and Kurata, F., "A Friction Factor Plot for Smooth Circu- 6 Transactions of the ASME lar Conduits, Concentric Annuli and Parallel Plates," Ind. & Eng. Chem., Vol. 52, No. 8, 1960, pp. 703-706. 2 Brighton, T. A., and Jones, T. B., "Fully Developed Turbulent Flow in Annuli," Journal of Basic Engineering, TRANS. ASME, Series D, Dec. 1964, pp. 835-844. Journal of Engineering for Industry Printed in U. S. A. 7' AtlanticRichfieldCompany North Amerii~' ' Producing Divisio~ North AlaskA, ,strict Post Office Box 360 Anchorage. Alaska 99510 Telephor~e 907 277 5637 December 9, 1975 Mr. O. K. Gilbreth State of Alaska Department of Natural Resources Division of Oil and Gas 3001 Porcupine Drive Anchorage, Alaska 99501 Dear Mr. 6ilbreth: On November 25, 1975, Mr. J. A. Rochon and I presented the Atlantic Richfield Company's report entitled "Prudhoe Bay Field Permafrost Casing and Well Design for Thaw Subsidence Protection" at the Prudhoe Bay Field Rules hearing. At that time, the question arose as to the strain characteristics of casing strings in the Prudhoe Bay Field other than the 13-3/8" 72# Mod. N-80 Buttress covered in that report. It was requested that we supply you with data to aid in your evaluation of this point. The strain limits of a casing string, as shown in the Figure 11-2 from above report, are dependent on three items. These are the type of steel in the pipe wall, the thread form inthe connection and the interaction between body and the connection. As noted in the report, the evaluation of the thread form requires both mathematical analysis and carefully instrumented full scale joint tests. Due to the difficulty of this type analysis and the variety of available joints we have not undertaken to evaluate other thread forms used or proposed for use in the Prudhoe Bay field. The other variable in evaluating the strain limits of a casing string is the physical properties of the body of the pipe. To aid you in this respect, we have obtained load versus strain data for other grades of oil field tubular goods from the Metallurgical .Engineering department of U. S. Steel Corporation. These data cover J-55 and H-40 materials. Mr. O. K. Gilbreth Page 2 December 10, 1975 We did not obtain a data sheet for K-55. We discussed this grade with Mr. Sheldon H. Reynolds of U. S. Steel in some detail. The load versus strain diagram for K-55 would have been very similar to the J-55 diagram and therefore they elected not to send it. These steels differ only in their minimum, ultimate strength. Had the test specimen for J-55 been loaded to the failure point, it would have parted at or above its minimum ultimate strength of 75,000 psi fiber stress. However, K-55 steel will con- tinue to bear a load up to or above its minimum ultimate strength of 95,000 psi. If plotted, the K-55 data would overlay the J-55 data and continue at the same slope to the 95,000 psi point. Simply stated, K-55 is ductile over a broader loading range than J-55. In practice most 55,000 psi minimum yield tubular goods sold today are K-55 material which meets and exceeds the older specification of the J-55 grade. Attached are the load versus strain diagrams as they were received from the U. S. Steel metallurgical laboratory. They are poor quality copies and to use the data in terms of fiber stress, the load on the sample must be divided by the specimen's cross-sectional area. We have prepared stress versus strain diagrams from the U. S. Steel data that are more easily related to the figures presented in our Thaw Subsi- dence Report. These stress versus strain diagrams project the load versus strain data on to minimum ultimate strength. We have also.included a diagram that extends the J-55 data as described by U. S. Steel's Mr. Sheldon Reynolds. This grade of casing has been used extensively at Prudhoe Bay. As is shown in the stress versus strain diagram, the pipe body for this grade of casing will comfortably strain beyond the limits imposed by the "worst case" thaw subsidence strains predicted in our work on the subject. ~ether a casing string made with this grade of steel will withstand these strains is then a function of the thread form used in the casing connection and the connection's interaction with the pipe body. Apart from the thaw subsidence loading, any casing string exposed to the permafrost must be designed to withstand external freezeback pressures. These pressures were defined by data from the DS 1-6 and DS 4-6 wells at Prudhoe Bay. This study has been presented in several papers and is a part of the attached paper entitled "Solutions for Some Problems Resulting From Refreezing of Permafrost Around a Wellbore". Figure 4 in this paper shows the predicted pressures which will be exerted on a casing string from external freezeback. The thaw subsidence strains and the external freezeback pressures are not additive loads. The thaw strains occur and are relieved in a hot well. This does not decrease the strength of the casing as exhibited by the stress versus strain diagrams. The external freezeback pressures occur in a cold well · Mr. O. K. Gilbreth Page 3 December 10, 1975 and all casing collapse values are calculated for minimum yield strengths of the casing. In short, these two forces do not form a compound load and the action of one of these forces does not detrimentally effect the pipe's response to the other force. If you have any further questions on this subject, please contact us and we will be glad to discuss this or other items at your convenience. Sincerely, C. R. Knowles District Drilling Engineer CRK:alh cc: J. A. Rochon L. N. Bell R. F. Fei 120 100 6O 4O ZO -- . . ~ ~ , - , ,. ; ~ , _ , : ....... ~ , . , : ,_ ........ ' ..... t : . : ' : ' - ~' ' : ~ ~ , { i , , , , : . ~ , , , ~ ........... ~,.,~ , ~ ..... :,~:./ ,,, : : - : ..... ,. :,, ~ ,,~: ..... ; ..... ~~-7:-~:~ :-':?--:-~ZF:-~=~~ :: :'~-t-: ~LE~L_:?~-I (75,000 psi Min Ult~ate Yield) 0 .S 1.0 1. S 2.0 2.5 3.0 3.$ 4.0 Strain 120 100 ~ 80 i60 Strain % 120 100 ~ 80 i6O 4O 2O ._~. ; . :. , . : :... ; .... . ~ ' : : '.'. ~ ; ~;. ''' ' , : ~. --- .............. 'X];_; ~ . ~ ..... ;~ .... ~:. :-;-: ;~ : ;--i-;--~-z; ~-~ .... + r:' : ~4~;-;-:; { (55,000 psi Min Yield) ,--~ .................. ~ z ........ ~ ..... m~ ~-~;---~-+ ~-~+--I f ~ =-~--' , I 0 .$ l.O 1.S 2.0 2.5 3.0 3.5 4.0 Strain % .~1~ £xteniila Laid ¥~ s 2~SO Lbi. 20r ~ Lbl ]~O00L Lbs 5000 Lba B~['O ~ Lbe I1~ LLe .II~ [xtNlm LNd Yte'l~t = 8240 Lbo. ,Am t .t3gS Sq, Ytlid .$.trlljth - S.9,070 &i BP ALASKA INC. 31 1 1 - C - STREET · TELEPHONE {907) 279-0644 MAILING ADDRESS: P.O. BOX 4-1379 ANCHORAGE, ALASKA 99509 November 26, 1975 Mr. Thcmas R. Marshall, Jr. Executive Secretary Alaska Oil & Gas Conservation Co~mittee 3001 Porcupine Drive Anchorage, Alaska 99501 Re: Conservation OrderFile ~137 Mr. Judd, who testified at the hearing on November 25, pointed out to me an error in his prepared testimony. Specifically, on page 2, line 4 the word "strains" should be deleted and replaced bytheword "strings". I would appreciate yourmaking a notation in the record to reflect thiso Very trulT....;yours, Q~2~der Attorney blm cc: P. R. Judd FIELD RULES ~ ~tRING NOVEMBER 25, 1975 ~ CONSERVATION ORDERS #98-A, #98-B, and #83-C · OF ~q~2 PRUDHOE BAY FIELD POOL RULES The Alaska Oil and Gas Conservation Committee in their Conservation Orders #98-A, #98-B, and #83-C established Rule #3 defining casing and cementing requirements. Sections of Rule #3 deal with casing design requirements in the permafrost. The testimony we present here deals with Rule #3 and in particular, the casing design required to prevent failure due to strains imposed on the casing by formation subsidence caused by thawing of the permafrost. Rule #3-B states that "wells shall be protected from. damage caused by a permafrost thawing by the use of refigeration and/or insulation or by use of slip joints". ~ ~]~e Con~nittee also stated that "other methods may likewise be safe and so~d and continued surveillance by t]~e Committee of all teCbMiques will be necessary to insure maximum safety in the future". E×~erience gained since the original hearing will be presented herein to justify another safe completion. The passage of time and the progress that has been made in obtaining d~ata and developing new concepts makes it possible to provide additional information which we trust will assist the Committee in reviewing the existing Field Rules and adopting new rules. The importance of the statement mad~ by the Co~nittee concerning the continued surveillance of ne~v tec]miques is warrm~ted by the data being presented today. Also as time passes, additional completion tec]miques may be developed. ?]~erefore, we suggest a field rule which will allow approval of new completions as they are developed and presented to the Committee. ~]~e following wording could provide the Committee with this flexibility: "Development well completions shall be design~ed to provide adequate' protection from permafrost thaw loading. Such well completions shall' a~ be ce Be ga_p._a.p..l_e of wi th>~.an_d_..in_g., t.h~_eff_e_.ct_s of thaw to protect the integrity of the well; or Employ a me~.an_.s of limiting thaw to protect the integrity~ of the well; or Employ other means acceptable to the Co~nittee." The testimony being presented today is contained in a report submitted herewith entitled "Pru~hoe Bay Field Permafrost Casing and Well Design for Thaw Subsidence Protection" dated May, 1975. ~fhis report presents a well completion using 13-3/8", 72#/foot, N-80 Buttress casing through the permafrost. The report clearly demonstrates that this casing provides a safe completion in the permafrost, and therefore, we would like to request approval of this design by the Committee. Prudhoe Bay Field Permafrost Casing and Well D~.~ign for ~aw Subsidence Protection/Report. S~nmary Since the initiaI development of the Prudhoe Bay Field, A.R. Co., BP, and Exxon have performed numerous laboratory, field, and engineering studies in an effort to more fully evaluate thaw subsidence problems. In 1969, BP Alaska cored the permafrost at Prudhoe -and had a variety of laboratory studies performed on the cores. In 1970, BP Alaska started two single well subsidence tests. In 1973, Atlantic Richfield and Ex~xon started a field test to simulate thaw that might be expected after about 15 to 20 years of production. In conjunction with the field test, A.R. Co. and Exxon independently developed mathematical models to predict the permafrost thaw subsidence · behavior. These models were then used to extend the field tests and e×~olore permafrost property variations. In addition to these measurements of permafrost behavior, there have been several tests and studies to determine the strain capacity of casing loaded by permafrost subsidence. A series of axial tensile and compressive strain limit tests have been run on samples of 13-3/8" OD, 72#/foot, N-80 Buttress casing. In addition, a finite element model has been developed for the 13-3/8", N- 80 Buttress threaded casing connection that provides a general method of predicting the minimun ultimate tensile and compressive strain capacities of'this casing connection. The report is a s~nmary of the most important conclusions from these efforts. The report includes a section covering the PrudJ~oe Bay Field five spot thaw subsidence test, the permafrost thaw strain predictions, allowable casing strain limits, permafrost geology, and pe~vaafrost thaw predictions. Several conclusions were reached in this report and are as follows: Thawing of the permafrost causes permafrost compaction and strain in the casing. 2~ Casing is available which can withstand strains in oil well completions through the permafrost interval. 3. This permafrost compaction-casing strain relationship requires a new casing performance criteria that utilizes tensile and compressive strain capacities of the casing and coupling in contact with the permafrost. 4. The strain producing mechanism is complex ,and is directly related to the amount of thaw in the permafrost. Strains are due predominately · to an interaction of alternating layers of sand and silts and are dependen.t upon the relative thickn~esses and mechanical properites of these layers. 5. We can predict the maximum expected tensile or compressive strains 'as a function of depth and radius of tha~v. 6. The maximLm~ strains measured in any casing joint in the five spot thaw test after 16 months, which represents a producing period of approximately 15 years for a mid-structure Prudhoe well as proposed herein, are .13% in compression and .08% in tension. ?." The worst case stains calculated for our proposed well designs are 0.7% compression and 0.5% tensile. These worst case strains are predicted by our mathematical model. 8. Full scale tensile and compressive tests and finite element studies of AP1 N-80 Buttress threaded casing show that a 13-3/8", 72#/foot, N-80 AP1 Buttress threaded connection can resist minimum ultimate compressive and tensile axiel strains in the pipe body of 2.3% compressive and 3.4% tensile. · 9. Using these worst case strains due to the permafrost tha~ and the minim~n ultimate strain capacities in the pipe body, our proposed well design and completion program give design factors of 3.3 in compression and 6.8 in tcnsion. Our current well design sho~ on the first slide (Figure II in the report) utilizes gelled oil in the 9-5/8" by 13-3/8" armulus and includes 13-3/8", 72#/foot, N-80 AP1 Buttress casing. The thaw test strains were 0.13% in compression and .08% in tension. Combining these values with the casing strain limits of 2.3% in compression · and 3.4% in tension results in design factors of 17.7 in compression and 42.5 in tension. To support these conclusions, I would like to review the principal items contained in the submitted report, namely the field test, the mathematical models, and studies of casing strain capacity. Field Test A five spot thaw subsidence test was started in Aug~mt, 1973. 'rhe test utilized five, 2200' cased holes in a closely spaced five spot p~tten~ to prodUce a thaw representing approxhnately i5 years of production from a Prudhoe Bay Field mid-structure oil well. ~]e purpose of the test was to measure the strains imposed on casing in the pern~frost caused by subsidence of the permafrost during thaw. The test was desi~ed to obtain the maximum data and yet provide sufficient time for well completions prior to field startup. The test was located near the center of the field and adjacent to the site of BP's extensive pe~nafrost coring program, thus the core data could be used in the subsidence test evaluatio~i, l~]e second slide (Figure Appendix I-1) is a map of the test site showing .the arrangement of the five thaw wells and the placement of the temperature m~d subsidence monitoring holes. The five tha~ wells were used to circulate a hot glycol-water solution to introduce heat to the test area. The temperature and'subsidence monitoring holes, as sho~] on the exhibit, were used to monitor the temperature to 400' and to measure the shallow subsidence from 15' to 250'. -6- rI~e test measurements used to evaluate the strains in the casing during the thaw test were as follows' · 1. ~e most important measurements in the subsidence test were the casing j gint length measurements taken with the Schlumberge~~ ~ultiple Casing Collar Locator tool. ~is tool provides casing length measurements of sufficient precision that we can compute length changes with a standard deviation of .01" in a 40' casing joint. This is equivalent to a strain of ~0. 002%. (Reference - SPE Paper 5087, 1974 Fall meeting, Houston, Texas, "Precise Joint Length Dete~nination Using a Multiple Collar Locator Tool.) " Slide #3 (Figure Appendix 1-12) shows the effects of thaw on casing stra~ by a series of plots of the MCCL measurements versus dept]~ at wrious time periods during the thaw test. Also sho~ on this ex]~ibit is the base of the first gravels at approx~nately 400' and the base of the permafrost at approximately 1850'. Slide #4 (Figure Appendix 1-13) shows the casing strain measurements after 16 months ~ of thaw with a g~mna-ray log and an intrepretation of the lithology of this well. Note that the peak measured compressive strain is · only .13% and that it occurs ~ a joint that is adjacent to a massive silt. Also note that the peak measured tensile strain is only .08% and that it occurs in a joint that is adjacent to a thick . sand. 2. Surface and near surface subsidence movements were calculated from periodic measurements of the elevations of the subsidence monitoring rods and wires. The total movement after 16 months of thaw indicated only a few tenths of a 'foot subsidence existed do~m to 250' and that the movements diminished with depth. -7- 3. A measurement of the relative movement of the soil and casing was made by placing radioactive bullets in the formation m~d using a · gamma-ray collar locator tool to record the relative positions versus time. Although there was some scatter of the data due to the precision of the measurements, we conclude that there is little, if any, relative movement between the pipe and the formation. 4. Temperature surveys were run in the three, 400' temperature observation wells throughout the test. The temperatures measured in these wells correlated very closely with the thermal modelS. 5. Pore pressures in the thawed region were measured with pressure transducers attached to the outside of the casing. Included in the report is a plot of the estimated pore pressure profile after 16 months of thaw together with the r~ge of pressure measureme~ts. ~]~e five spot thaw pattern was analyzed with an X-Y two d'imensional thaw simulation model. ~]~e progress of the thaw at 1000' level is shown on the next two slides, '(Figures Appendix 1-6 and 1-10). Slide #5 shows the calculated thaw area a~ 1000' depth for each of the five spot test wells after 3.9 months of thaw. Slide #6 shows the calculated that area at 1000' after 18.5 months of thaw. As can be seen on Slide #6, the thaw of the five wells has merged into one single thaw area simulating one producing well. The equivalent single well thaw radius for the five spot pattern versus depth is shown in Slide #7 (Figure Appendix 1-11). Ail the data from the thaw test indicated a low level of strain in 'the casing, and as stated previously, the maximin measured compressive strain was 0.13% and the maximum measured tensile strain was 0.08%. M[~thematical Strain Model Casing strains resulting from the large thaw radius have been measured with good precision in the five spot thaw test. In order to extend casing .strain predictions across the field, mathematical models were developed. These models were then used to calculate strains for different lithologic distributions. The problem of computing the expected strains at a given location in the pe~nafrost is complex. There are at least three in~ortant casing loading mechanisms that influence the e×~ected subsidence strains. 1. An inward lateral movement of surrounc]ing frozen permafrost. 2. An upward elastic flexing of soil below the thawed ~egion. 3." A tendancy for downward movement of thawed material generally throughout the thawed region caused by the mass of the thawed material, the'gravitational field, and. the drop in pore pressure. ~e overall subsidence loading mechanism is 'illustrated by the sketch shown in Slide #8 (Figure Appendix II-1). l~ile the permafrost ~is thawing, the volume resulting from the conversion of ice to water ~ the pore space causes a reduction in the pore pressure in the thawed region. The decrease in pore pressure in the thawed region reduces the total horizontal aa~d vertical stresses, and in turn causes the permafrost at the frozen boundary to move laterally toward the wellbore and compress the thawed soil. The second ~d third loading mec]mnisms are the upward flexing of the soil below the thawed region at the base of the permafrost and the tendancy for a dmmward movement of the thawed material due to gravitational forces and the drop in pore pressure. ~e maximum e×~ected permafrost and casing strains result from tl~e interplay of downward acting fox~ces within the thawed region and from the horizontal compaction of the permafrost in' intervals where there are alter~ating layers of stiff sand and compressible silts. /J~e axial tensile and compressive strains produced by this interaction are caused by the difference in compressibilities of the layers as sho~m in the Slide #9 (Figure Appendix 1I-2). As can be seen on this slide, as the boundary of the thawed permafrost moves toward the wellbore, it laterally compacts both layers of sand and silt. Since the sand is less compactible tha~ the silt, lateral compaction causes the sand to e×~and vertically into the silt thereby generating tensile or lengthening strains in the sand and adjacent casing, and compressive strains in the silts and adjacent casing. The magnitude of the tensile and compressive strains is highly dependent on the layer thickness, their mechanical properties, the thaw radius, and the depth of the interval. Another important thaw subsidence mechm~ism is the reaction of the pressure boLuxdary at the base of the permafrost. ~e base of the permafrost is deflected upward because of the difference of pore px~essures between the thawed region and. the formation below the pern~afrost. The u~ward movement at the base of the permafrost causes axial compreSsive strains in the permafrost and casing above the base and tensile strains in the unfrozen formation below the permafrost. ~e third mechanisn~ predominates in the first gravels which includes the upper 400' to 500' of the pex~nafrost. Vex~y little vertical strain in the permafrost or casing have been measured or was e×pected in this interval because of two factors. First, the gravel appears to be underconsolidated. ~q~is tu~derconso].idation leads to a low soil shear strength in the frozen material. Shear failure in the frozen pe~anafrost ! essentially at the thaw boundary, permits the thmved soil and p{pe to move do~n together. This limits the differential vertical strain in this region. Second, permeability is apparently high J~ these gravels and measurements show that the pore pressure is essentially not reduced below a hydrostatic head of water during thaw. ~]~is factor also tends to limit thaw strain. ]]~e discussion of the pe~m~afrost geology to be presented later will help clarify the differences present in the top 400' to 500" of the permafrost. ~]~e n~nerical subsidence model was developed from a basic understsa~ding of the subsidence mechanism and measurements of mechanical properties of the permafrost .materials. The q~.~ality of the fit between the calculated perfonnsmce and the test measurement is shown on several figures in the report for various months of thaw. Shown on Slide #10 (Figure Appendix II-9) is a comparison of the computed strains versus 'the thaw subsidence test strains after 16 months of thaw. qg~e dete~'~ination of .maximum strain across the field' was hm~dled ~y a worst case approach using the expected thaw radius versus depth and conservative values of pore pressure loading and mechanical properties. ~he model was used to compute the highest co~ressive and tensile strains that would result from any possible combination or relative thickness of sands and silt layers. Since these studies include all of the expected lithologic co;nbinations in Prudhoe, the results are applicable across the PrucU~Oe Bay Field. 3]~e maxJ~ncun possible compressive strains occur in highly compactible silt layers. As sand layers become very thick, the silt layers are essentially isolated from one ~mother and consequently undergo maximum strain. Now consider thin sand layers located in thick silt sections. · Because of the greater stiffness of the sand, tensile strains are induced below the sand and increased compressive strains are induced in the silt above the sand layer. As sand layers become widely spaced, they become essentially isolated from one another and computed strains reach a maximum. As the thickness of the isolated sand layer increases, its resistance to deflection and thus its effect on contiguous silts increases and approaches a maximum effect. Shm~ in the report are the envelopes of maxim~ tensile and compressive strains drawn to include all maximum worst case values. At shallow depths, the largest values of strain are computed for isolated layers of sand, in thick silt sections. Near t~e bottom of the permafrost, the largest values of compressive strain are confuted for the thin silt layers in thick sa~d sections. ~ne final pore pressures in the thawed region of the producing well will probably be higher titan the pressures measured in the highly accele'rated thaw test. However, for the worst case approach, we have assumed a~ even more conservative pore pressure profile than fom~d in the five spot test. Slide #11 (Figure Appendix I I~3) shows the pore pressure.profiles in the thawed ~ermafrost for the five ~spot thaw test and the pressure gradient used in. the worst case calculations. Near the top of the pe~mafrost, the field data shows the pressure gradient is essentia]~ly a full hydrostatic ]~ead of water. This smne gradient was assumed for the worst case. ~e pressure gradient measured in the silts below the shallow gravels was also used for the worst case because these press~ures are considered to be the minim~n pressures that can be expected during the extended life -12- of a real producing well. Belo~v this trend, a zero pore pressure profile · was exten~.!,~,d to a point near the base of the permafrost. A limiting pressure approach was made for the pressure below the zero profile to the hydrostatic pressure at the base of the permafrost. ]]~e earliest industry concerns for the subsidence of tJ~e upper 500', or first gravels, were based on the concem~ that this interval might contain excess ice and that thawing would cause large permafrost and casing strains. Since these early studies, there has been considerable evidence developed that the first gravels do not contain excess ice below the top 50'. Differential vertical strain in the thawed region will be minimized by shear failure in the surrounding frozen material. · Sma. ll vertical strain is certainly sho~m in results of the five spot .thaw subsidence test. Shovm i~ the report, are max~]~n strains in the top 430' which are less than half of the strains measured below that point. The m~.mxim~n calculated thaw subsidence casing strain for worst case conditions for the well design presented earlier after '20 years of tha~ was calculated by the model to be 0.7% compressive m~d 0.5% tensil, e. Casing Strain Capacities In addition to predicting the max~n~n casing strain that will be imposed by thmving the permafrost, we must determine the strain which our proposed casing can withstand. We thus undertook laboratory m~d mathematical studies to dete~nine this strain Iimit. (Reference - SPE Paper #5598, Fall, 1975, "Casing Strain Test of 13-3/8" N-80 Buttress Casing". ) Full scale strain limit tests were conducted to establish tensile and compressive strain lJ~nits of 13-3/8", 72#/foot, N-80 Buttress ! casing. Six tests were performed in all including three tension and three compress, ion tests. AxiaI and hoop strains ~'ere measured at several places on the pipe body exten~al to the coupling, inside the surface of the pipe body, opposite the threads, and on the outside of the coupling. The data at a point 1" from the coupling indicates that the strain at this location at the time of failure of the connection was between 1.9% and 3.4% on the compressive strain tests. 11~e strain 1" from the coupling at the time of the com~ection failure was between 2.75% and. 4.05% in the tension tests. Slide #'12 (Figure Table Ili-2) s~narizes the data for the three compression and tension tests on the 13-3/8", 72#/foot, N-80 Buttress casing. l~e minim~n ultimate casing strain, capacities for 13-3/8", 72#/foot, N-80 Buttress threaded connections were calculated with a finite element model. The model was calibrated and the strain .limit failure criteria were established using the results of the full scale Strain limit tests. Failure in compression is indicated in the model when the first tooth on the pin loses its sealing force, which represents a potential for separation between this tooth and the collar. Failure in tension is d~fined in the model by the magn~itude of the average strain in the pipe body at the root of the last loaded tooth on the pin. ~l~e ultimate strain capacities of permafrost casing are defined as the average axial strains in the pipe body that are required to fail the coupling. The model predicts a 3.3% ultimate strain capacity for a simple compressive load on norn~lized N-80 casing with normal thread nmkeup, The addition of pressure on a no~analized. N-80 casing has a beneficial effect on the calculated compressive strain capacity. For combined loads of 1800 psi internal pressure, a thermal load represented · by 100°F increase in temperature on normalized N-80 casing with nominal AP1 wall thick~ess, the model indicates a compressive strain limit of , 3.9%. A calculation was also performed using properties of quenched and tempered N-80 casing. ~]e results indicated that quenched and tempere~-' .......... ~ N-80 developes even higher ultimate strains than normalized N-_~.~._~'~'~]e~ results for quenched and tempered pipe show that failure has not occured with a 7.2% compressive strain in the pipe body: The model was also used to evaluate the sensitivity of the strain limit to makeup interference between t]~e pin and the colla.r. IVith 40% less than the normal interference, the ultimate compressive strain .capacity is reduced to 2.3%. With more than the normal makeup, the ultimate conpressive strain capacity is increased. The ulthnate strain capacity for tensile loads is higher than for compressive loads. The ultimate tensile strain capacity for normalized N-80 casing under normal conditions is 3.7%. The addition of 1800 psi internal pressure decreases the strain l~nit to 3.4%. Reducing the makeup interference apparently has little effect on the tensile strain capacity. A calculation using 40% less than normal makeup interference i~dicates a tensile strain limit of 3.7% which is the same as the limit for a normal makeup condition. Slide #13 (Table III-1) sun~narizes the finite element model calculations. Shm~n in the report are ultimate strain predictions for com~ection failure of 13-3/8", 72#/foot, N-80 Buttress casing and a comparison of the measured m~d the computed strains in compression m~d in tension. From these studies, we conclude that the minimum ultimate strain capacities of 13-3/8", 72#/foot, N-80 Buttress casing are 2.3% con~ressive and 3.4% tensile. · Pemnafrost '~aw Predictions 2~e~m~al models were developed early by the petrole~n industry to predict pemnafrost thaw. Several of these thermal models have been described in technical papers previously published. Shown in the report are several calculated thaw radii versus depth cu~es used in the model studies. All subsidence predictions have been based on the thaw predictions for a mid-structure oil well after 20 years of production. SL~ary Slide #14 (Figure II-2) st~nmarizes all the data from our model predictions. Shown on this exhibit are the ma×i~nt~n expected con[~ressive ~d tensile casing strains based on our model studies for the worst case analysis. Also sho~ are the minim~n ultimate compressive and tensile. strain capacities of 13-3/8", 72#/foot, N-80 Buttress casing. As discussed earlier, usi2~g the worst case approach, our proposed well design and completion program give design ~actors of 3.3 in compression a~d 6.8 in tension ~d using the actual measured strains from the thaw subsidence test, give design factors of 17.7 in compression and 42.5 in tension. STATEMENT OF P. R. JUDD STATE OF ALASKA, 0~L & GAS CONSERVATICN CO~II~I~E NOVEMBER 25, 1975 P. R. JUDD Qualifications for Expert Witness State of Alaska - Oil and Gas Conservation Ccnmtittee In December 1963, I graduated frcm the University of Washington in Seattle with a Bachelor of Science degree in Civil Engineering. At that time, I was employed by Shell Oil Cc~pany of Los Angeles. After receiving training and initial work assignmants in Ventura, California and Houston, Texas I was transferred here to Anchorage in November 1964 to a position as Drilling Engineer. During my five year assist in Anchorage,. I was involved with engineering and planning for drilling in. Cook Inlet, both offshore and onshore and on the North Slope. In November 1969 I was transferred to Bakersfield, California as Senior Drilling Engineer involved in drilling engineering in the San Joaquin Valley in California and in .the State of Utah. In AuguSt 1970 I was 'transferred to Houston, .Texas where I was involved with a wider range of drilling activities, both offshore and onshore, including wells to below 23,000 feet. In 1971 I returned to Anchorage and joined BP Alaska in October 1971. With BP I have been involved with drilling on the North Slope, primarily in the Prudhoe Bay field. My assists have included both engineering and direct supervision in the field. In September 1974 I received my present position as ~District Drilling Engineer. I am in charge of BP Alaska's Drilling Engineering section and am responsible for engineering and technical support for drilling and well completion operations. I also assume the Drilling Superintendent's duties in his absence. I am a m~mt~r of the SOCiety of Petroleum Engineers. MR. CH~IRMB2W ~ MEMBERS OF THE CC~v~tI~i'itlE: This testimony is in regard to Rule 3, Section (b) of Conservation Order Numbers 98-A, 98-B, and 83-C. As stated by Mr. Reeder,. BP Alaska Inc. supports the conclusion of Atlantic Richfield Company's report, "Prudhoe Bay Field Permafrost Casing and Well Design for Thaw Subsidence Protection", dated May 1975, which has been sulmtitted to you in their test- imony today. We have been kept fully informed throughout the planning and execution of the field test and assOCiated~ studies sunnmrized in that report. We have satisfied ourselves that the concepts are. sound and that the data used fell within reasonable bounds. This work has resulted in a far better understanding of the effects of thaw subsidence and of casing design required to withstand thaw induced strains. The report has shown that permafrost thaw can result in significant casing strains and that "worst case" strains can be prc~ticted for purposes of design. We should l~ke to emphasize "worst case" in that -the predicted casing strains that have been generated frcm the computer model studies repres~]t the effect of a thaw radius resulting from a c~nbination of severe, or "worst case", conditions, such as continuous sustained fluid production over a period of 20 years with the worst ccmbination of lithological conditions in the soil. The resulting strain predictions, then, form a basis for conservative criteria for use in designing casing and completion programs having more than adequate safety for withstanding ~haw induced strains. It is well to keep in mind that the maximum strains observed in the ARCO field test which, after all, is 'the most likely situation we will observe during pro- duction, are significantly below yield. ARCO's report also shows casing strings are conn~rcial!y available which can withstand the worst~ case design strains with an adeqUate design factor -2- without failure of the connections. Specifically, the report shows that a 13-3/8" diameter, 72 lb/ft, N-80 grade casing with API Buttress thread connections can easily resist the Predicted thaw induced strains. We' agree with this conclusion and also feel that other casing strains may also be suitable. That is, other combinations of size, weight, grade, . and/or connection thread may be suitable for well design configurations· which differ from ARCO's current well design. with a view tovrard extending this knowledge, we have carried out laboratory testing of casing like that already set in two exiSting wells, to determine the post yield strain characteristics of that casing. Pre- liminary results indicate that this particular casing may not have satis- factory post yield strain performance. Based on this experience, we would suggest laboratory testing or mathematical analysis should be required to show 'that where~ necessary, other casing strings could withstand thaw induced strains. The current Prudhoe Bay Field Rules permit only three specific types of well design with respect to permafrost thawing. Rule 3, Section (b) of Conservation Order Nos. 98-A, 98-B, and 83,C states, "Wells shall be protected frc~ 'damage caused by permafrost thawing by the use of refrig- eration and/Or insulation or by the USe of slip joint casing". Based upon testimony presented today, we recC~mend broadening the rule to reflect new knowledge gained to date and to allow any 'future knowledge to be incorp- orated in possible new casing and well ccmpletion designs. Any new rule should also take intO account the fact that thaw indUCed strains are directly related to the amount of thaw in the permafrost and only become significant for wells producing large volumes of warm fluids over long periods of time. That is, injection wells, gas wells,, observation wells, low volume oil wells, and exploratory wells need not necessarily be de- signed to meet the same criteria for permafrost thawing as previously discussed. As a more specific example, time limitations on oil production would limit the amount of thaw and could be a method of r~ducing the assoc- iated strains to acceptable levels in wells for which other methods of designing for thaw induced strains are impractical. Similarly, reduction of the amount of thaw through the use of insulation has already been recognized as an acceptable procedure, Any new rule should also consider that application of thaw induced sltrain criteria is only significant for casing strings playing a role in the integrity of the well. As an example, predicted failure of the outer- most casing connections in the permafrost zone due to thaw induced strains is acceptable if the integrity of the well is not affected. The alter- nating axial and localized nature of the strains associated with the permafrost thaw are such that the effect on casing cemented to the permafrost would not be transmitted to the next inner casing if the two strings are not cemented together through the permafrost, interval. We support the proposed field 'rule as stated by Atlantic Richfield C~pany, which is repeated for your consideration. "Develo~nent well cc~pletions shall be designed, to provide adequate protection from the effects of pern~frost thaw lOading. Such well completions shall: a] Be 'capable of withstanding the effects of thaw to protect the integrity of the well; or b) c) Employ a means of limiting thaw to protect the integrity of the well; or Employ other means acceptable to the Ccmmittee." We believe that the current practice of the Division of Oil & Gas in requiring notification of the plan for ccmpletion of a particular well provides adequate well-by-well procedure for monitoring proposed designs to ensure they cc~ply with this broadened rule. Assuming Cc~ttee approval, our current plans for the cc~pletion of the initial oil wells to m~et the startup of the Trans,Alaska Pipeline may be grouped into four categories: Group 1) Those wells Which will employ an outer casing string through the' permafrost, which can withstand thaw induced strains. Group 2) Those wells which will employ high-grade insulated tubing to limit the amount of thaw. Group 3) Those wells which have an outer casing string installed through the permafrost for which the ability to withstand thaw induced strains has not been demonstrated. In these wells, however, any strains in the outer casing are not transmitted to the inner~ casing strings. The integrity of the well is maintained even though t/~e outer casing connections may not be able to withstand w~rst case thaw induced 'strains. Group 4) This final category contains a few wells ~lich were drilled early in the development phase, fOr which cc~pletion plans have not yet been finalized. We are considering proposing a limited production life which would limit the extent of thaw to a safe level. We have no testimony to offer regarding other sections of Rule 3. I thank you for your attention. Do you have any questions for me? STATIK~]'~ OF R. B. VICKERY STATE OF ALASKA.- OIL AND GA,~'.~ CC NSE~Af;~I ION COMMI'i'I'EE NO~/}?iMBER 25, 1975 QUALIFICATIONS FOR EXt.>ERT WI'INESS - STA'fE OF ~%I. JkSi{A NOVEMBER 25, 1975 I jo.h~ed BP in 1973 and an responsible for Exploration ~-~d Devel%m'~omt drilling operations for BP Alaska Inc. as Drill~]g Superh]t~dent. I graduated fr~n the Colorado School of Mines in 1962 wittt'a degree Petroleum Engineering. After graduation, I was employed as a Drilling/ Production Engineer for Pan American Petrole~n Corporation in Wyoming and Montana. I jo~]ed Vickery Drill/rog C~npany, Inc., a fzmtily midwestern drilling' con~ract firm. I adv,]ced fr~n tool pusher 'to drilling engineer, to Drilling Superintendent, to Vice-Presiden't, O[×~rahions over an 8 year period. I was responsible for a nine rig contract drilling operation working in natural gas storage exploration -.and developnmnt. subsequently fonne~d and was President of Refuge Exploration Co. Inc., a shall midwestern oil and gas exploratory operating comp~my. BP Alaska Inc. has camde fully consider:cd tJ'~e prolxpsal to a~ne~d Rule 4, Conservation Order 98-A, 98-B, ~u~d 83-C of t~e Pz~ac~oe Bay Field rules to require a bag type blowout preventer while drilling 'tJ'~e hole in which~ the surface casing is set. WitJ~in ~e trait area note th~ 100 wells have been drilled tJ~rough the base of the pernefrost. No hydrocarbons have been encountered above our surface casing setting depth of 2700' east of the Kuparu~i River. The area enccm~passed by our current Sadlerochit develolm~t, n~ely our present 10 development drilling pads, lies east of tine Kup-aruk River. We ~nm~e convinced in these areas, based on evideu~ce we have gather~:t, it is safe to drill through the Pem~frost witl~out the necessity of Divc~ters. Any zn~le change should maintain currently accepted safe drillJmg practices where the absence of hydrocarbons over the Pemnafrost intervals is proven. Should M~is rule be anmnded requiring ~nn mmular preventer to be installed on the conductor pipe it should be designated as a Diverter System. This Safety System should be designed to have its pr.imaz.7 fm-~ction to divert well flow aWay from the drilling rig. BP 'Alaska Inc. supports ~is proposed rule change in areas of ur~own or questionable geologic, information at t]~e base of tt~e Pernmfrost; providing t]~e surface oontrol syst~n is primarily designewt as a Diverter, .with surface outlets of sufficient size to nu[nJ.mize restriction-to flow. F~-ther, t~e su~.-face diverting lines should be designed to minJm~i, ze freeze up wlnen operating in arctic conditions. We suggest the follc~vh~g wording whicl~ could a[~ply to Conservation Orders 98-A, 98-B, ~u~¢t 83-C. "The first well on ~y drilling pad shall have a ra~tely tuu~ul~~ t)532 Diverter installed prior to drilling below the conductor cash~g. This surface control systcn's primary fm~ction is t'.o divert well flc~ away fran the drilling o[>aration should pressure be encore%feted while drilling the surface hole. This system~ is to have surface outlets of. sufficient size to minimize the restriction to flow and shall be designed to avoid freeze up." STATEMENT OF R. H. CLARKE STATE OF ALASKA - OIL AND GAS CONSERVATION COMMIYTEE 'NOVEMBER 25, 1975 Note: References to authorities may be found in the Atlantic Richfield Company Report "Prudhoe Bay Field Casing and Well Design for Thaw Subsidence Protection", dated May, 1975. R. H. CLARKE Qualifications for Exper~ Witness - State of Alaska November 25, 1.975 I am a geologist in the Production Planning Del~oartment, BP Alaska I'nc., San Francisco. I reCeived a BSC degree in Geology from the University of Bristol, England in 1963. In ].966, I obtained a .Ph.D degree from the same, university. After seven Years' offshore explOratory research, con- cerned mainly with Tertiary and Quaternary deposits, I joined British Petroleum Co. in 1970. After two and a half years of exploration geology in the Nort'h Sea, I was transferred -to A.D.M.A. Co. in Abu Dhabi where I was mainly concerned with production and development geology. In April, !974, I was transferred 'to BP Alaska' Inc. and have since been working on various aspects of the deVelopmen-h of 'the Prudhoe Bay Field. November 25, 1975 Tuesday, 9'00 AM [000] MR. MARSHALL' Good Morning. This is a meeting of the Oil and Gas Conservation Committee. My name is Tom Marshall and I am Acting Chairman. Our regular chairman, Easy Gilbreth, is temporarily out of pocket due to a leg injury. On my right is Harry Kugler, the Acting Executive Secretary, who is acting in my old job. On my left is Hoyle Hamilton, our Chief Petroleum Engineer. Harry, by the way, is a Petroleum Geologist. Backing us up in the rear are Lonnie Smith and John Miller, Petroleum Engineers, and Barbara Morgan working the dictating equipment. I'd like to also introduce Jeff Lowenfels with our District Attorney's office here in Anchorage. There are two principle items on the agenda this morning and believe me, they are very distinct items. One, the first we will handle, deals with a statewide proposed rule change and this part of the hearing is really a re- hearing, because we went through this whole thing back in 1970, every aspect , of it, but somehow or other, and we are not sure how, in Juneau, through a clerical error, this proposed rule change did not appear in the Alaska Administrative code the last time it was published, so our legal eaqles in Juneau advised us to be supersafe about the thin~ and toeing the mark on everything that we should bring it up for public hearing again and so we are doing this. This is why this whole hearing took thirty days notice. This is part of the Alaska Administrative Procedures Act to give thirty days notice. The second item we will consider is Conservation Order No. 137 which deals with proposed possible changes in the Prudhoe Bay Pool rules. This is of course'something entirely different and deals with existing pool rules that have been in effect for sometime and we are considering changes in the blowout prevention requirements and in the casing requirements of the Prudhoe Bay Pool rules. A little bit about housekeeping matters - there is a No Smoking ordinance in this public library. The signs are not showing; they were here the other day. This is a public meeting and by law smoking is prohibited. We will have a rest break in approximately one hour and we will play that by ear depending on what stage our testimony is at that time. I would ask people who come forth as witnesses to speak directly into the microphone because it gives us all a much better transcript that way. Expert witnesses will be asked of their qualifications and will be sworn. This is a matter we will take up again after we dispatch the first item on the agenda. At this time I would like Mr. Kugler to read the public notice of the proposed statewide regulation change. MR. KUGLER' "Notice is hereby given that the Division of Oil and Gas of the Department of Natural Resources, under authority vested by AS 31.05.030, proposes to amend a regulation in Title 11 of the Alaska Administrative Code as follows' (1) 11 AAC 22.030 (2) is amended by adding the following sentence- 'Within permafrost, no fluids with a freezing point above the minimum permafrost temperature shall be left in the annulus between any two strings of pipe after completion or suspension of a well without committee approval.' This proposed amendment was carried through the hearing and adoption process in 1970 and has been enforced since that time, however it has been found that through clerical error, it was not published in the Alaska Administrative Code Register. Notice is also given that any person interested may present oral or written statements or arguments relevant to the action proposed at a hearing to be held in the Municipal Chambers of the Z. J. Loussac Library, Fifth Avenue and F Street, Anchorage, Alaska at 9-00 A.M. on November 25, 1975. -2- The Division of Oil and Gas of the Department of Natural Resources, may as a result of the hearing adopt the above proposal substantially as above set out without further notice." This is signed by O. K. Gilbreth, Jr., Director of the Division of Oil and Gas, and is published October 25, 1975. MR. MARSHALL: Thank you, Harry. As I mentioned, this particular item has already been through public hearing and we are just for the record reopening it. At this time I would like to ask, are there any comments, objections, or statements on this particular agenda item? Hearing none, we will go on to the next order of business, Conservation Order No. 137. Mr. Kugler, I would like you to read that notice please. MR. KUGLER: "The Alaska Oil and Gas Conservation Committee will hold a hearing on its own motion pursuant to Title 11, Alaska Administrative Code Section 22.540 in the Municipal Chambers of the Z. J. Loussac Library, Fifth Avenue and F Street, Anchorage, Alaska at 9:00 AM on November 25, 1975. The Committee will seek testimony on the following matters: 1) To consider an amendment to Rule 4, Conservation Order g8-A, 98-B, and 83-C of the Prudhoe Bay Field pool rules to require a bag type blowout preventer while drilling the hole in which the surface casing is set. The present regulations pertain to requirements for blowout prevention equipment only while drilling below the surface hole. 2) To consider an amendment to Rule 3, Conservation Order 98-A, 98-B, and 83-C. Sections a, b and d in particular will be considered but the entire rule will be open for possible changes. Experience gained since 1970 appears to warrant changes in casing requirements in the permafrost portion of the hole." This is signed by Thomas R. Marshall, Jr., Executive Secretary, Alaska Oil and Gas Conservation Committee, and it also was published October 25, 1975. MR. MARSHALL: Thank you, Harry. Our order of handling the business will be as follows. We will give the companies or individuals with expert testimony -3- a chance to testify first. We will follow that with Questioning of those testifying. Then we will open up the meeting for questions or statements from the public or any interested or involved individuals and possibly we will leave the record of the hearing open for possibly 10 or 15 days depending on any problems that may loom as requiring additional information. Without further adieu, I would like to ask if there is anyone who would like to present testimony at this time. MR. REEDER: Mr. Chairman, John Reeder with BP. We have testimony on Rules 4 and Rule 3. We, if possible, would like to present our testimony on Rule 4 at this time. MR. MARSHALL: Very fine. We would appreciate it if you would sit at the witness table. We will get better reception on our recording equipment. MR. REEDER: Mr. Chairman, I have some prepared testimony of Mr. R. B. Vickery, who will be our witness. MR. MARSHALL: Would at this time, to save doing it all over again, would anybody who knows that they are going to be an expert witness please come to the front and stand because we will swear you all in at once. Harry, will you please swear in the witnesses. MR. KUGLER: Please raise your right hand. In the matter now at hearing, do you solemnly swear to tell the truth, the whole truth and nothing but the truth so help you God? RESPONSE: I do. -4- MR. MARSHALL: John, have your witness' qualifications been given the Committee before? MR. REEDER: Mr. Chairman, I don't believe so. As a matter of fact, none of the persons that are speaking for BP today have been qualified before the Committee as expert witnesses and they have all prepared a statement and I would ask the Committee at this time, after the statement is read, to accept them as expert witnesses. MR. MARSHALL' Very good. ~178] MR. VICKERY: Mr. Chairman, ladies and gentlemen of the committee, my name is R. B. Vickery. I joined BP in 1973 and am responsible for Exploration and Development drilling operations for BP Alaska Inc. as Drilling Superintendent. I graduated from the Colorado School of Mines in 1962 with a degree in Petroleum Engineering. After graduation, I was employed as a Drilling/Production Engineer for Pan American Petroleum Corporation in Wyoming and Montana. I joined Vickery Drilling Company, Inc., a family .owned midwestern drilling contract firm. I advanced from tool pusher to drilling engineer, to Drilling Superintendent, to Vice-President, Operations over an 8 year period. I was responsible for a nine rig contract drilling operation working in natural gas storage exploration and development, I subsequently formed and was President of Refuge Exploration Co., Inc., a small midwestern oil and gas exploratory operating company. MR. REEDER: Would the committee~accept Mr. Vickery as an expert witness? MR..MARSHALL: Hearing no objections, we accept Mr. Vickery. -5- MR. VICKERY: BP Alaska, Inc. has carefully considered the proposal to amend Rule 4, Conservation Order 98-A, 98-B and 83-C of the Prudhoe Bay Field pool rules to require a bag type blowout preventer while drilling the hole in which the surface casing is set. Within the unit area more than 100 wells have been drilled through the base of the permafrost. No hydrocarbons have been encountered above our surface casing setting depth of 2700' east of the Kuparuk River. The area encompassed by our current Sadlerochit development, namely our present l0 development drilling pads, lies east of the Kuparuk River. We are convinced in these areas, based on evidence we have gathered, it is safe to drill through the permafrost without the necessity of diverters. Any rule change should maintain currently accepted safe drilling practices where the absence of hydrocarbons over the permafrost intervals is proven. Should this rule be amended requiring an annular preventer to be installed on the conductor pipe it should be designated as a Diverter System. This Safety System should be designed to have its primary function to divert well flow away from the drilling rig. BP Alaska, Inc. supports this proposed rule change in areas of unknown or questionable geologic information at the base of the permafrost; providing the surface control system is primarily designed as a diverter, with surface outlets of sufficient size to minimize restriction to flow. Further, the surface diverting lines should be designed to minimize freeze up when operating in arctic conditions. -6- We suggest the following wording which could apply to Conservation Orders 98-A, 98-B, and 83-C: "The first well on any drilling pad shall have a remotely controlled annular type Diverter installed prior to drilling below the conductor casing. This surface control system's primary function is to divert well flow away from the drilling operation should pressure be encoun- tered while drilling the surface hole. This system is to have surface outlets of sufficient size to minimize the restriction to flow and shall be designed to avoid freeze up." Gentlemen, that is the end of my testimony and I am open for any questions that you might have. MR. MARSHALL: Thank you, Mr. Vickery. Hoyle, you had some questions for Mr. Vickery? We will have some questions possibly at a later time, so if this concludes your testimony why make yourself comfortable but we will very likely call you back, Mr. Vickery. Thank you very much. MR. REED: Mr. Chairman, at this point we would yield to Atlantic Richfield Company. We have no further comments on Rule 4. MR. MARSHALL: Thank you very much gentlemen. MR. SCOTT: Mr. Chairman, gentlemen, for the record, I'm John Scott, the attorney for Atlantic Richfield Company. We wish to present testimony on the proposed amendments to Rule 3. Our first witness will be Mr. J. A. Rochon, or Mr. Jerry Rochon, our senior operations engineer for our North Alaska district. Mr. Rochon has not been previously qualified as an expert before the committee, so he is prepared to do that today. Our second witness. -7- will be Mr. R. Knowles, Jr., or Dick Knowles, our District Drilling Engineer for North Alaska and he also will need to be qualified. Also sworn in but having no direct testimony but to be available for ouestioning is Dr. C. K. Perkins, director of Production Research at our company's North American Pro- ducing Divisions' Research facility in Plano, Texas. Dr. Perkins has previously qualified as an expert in the Field Rule hearings, in fact I think, in the first Field Rule hearing and with that we will proceed with the Qualifica- tions statement of Mr. Rochon. MR. MARSHALL' Very good. MR. ROCHON' Chairman, Committee, I graduated from the University of Cali- fornia in 1953 with a degree in Petroleum Engineering. After graduation I worked for the Richfield Oil Corporation which later merged with the Atlantic Refining Company to become the Atlantic Richfield Company. I worked for a year as an engineer trainee and then assumed responsibilities in the fields of drilling, production, reservoir engineering. During the period from 1964. to 1966 I was the Senior Engineer working on Thermal Simulation in the Bakersfield, California office. In 1966 I came to Alaska and until 1970 worked as a Senior Operations Engineer on Atlantic Richfield's Cook Inlet operations. Since 1970 I have worked as a Senior Operations Engineer on drilling and production engineering duties. During the period 1970 to the present I have been the Senior Operations Engineer working on the permafrost freezeback tests and the thaw subsidence tests. I am a registered Petroleum Engineer in the State of Alaska and the State of California. MR. MARSHALL- Hearing no objections, we accept your qualifications Mr. Rochon. You may proceed. -8- [288] MR. ROCHON: I have some prepared testimony I would like to hand out at this time. I would also like to set up the projector at this time to show some slides. The Alaska Oil and Gas Conservation Committee in their Conservation Orders No. 98-A, 98-B and 83-C established Rule #3 defining casing and cementing requirements. Sections of Rule #3 deal with casing design requirements in the permafrost. The testimony we present here deals with Rule #3 and in parti- cular, the casing design required to prevent failure due to strains imposed on the casing by .formation subsidence caused by the thawing of the permafrost. Rule #3-B states that "wells shall be protected from damage caused by a perma- frost thawing by the use of insulation and/or refrigeration or by'use of slip joints". The Committee also stated that "other methods may likewise be safe and sound and continued surveillance by the Committee of all techniques will be necessary to i.nsure maximum safety in the future". Experience gained since the original hearing will be presented herein to justify another safe compl eti on. The passage of time and the progress that has been made in obtaining data and developing new concepts makes it possible to provide additional informa- tion which we trust will assist the Committee in reviewing the existing Field Rules and adopting new rules. The importance of the statement made by the Committee concerning the continued surveillance of new techniques is warranted by the data being presented here today. Also as time passes, addi- tional completion techniques may be developed. Therefore, we suggest a field rule which will allow approval of new completions as they are developed and presented to the Committee. The following wording could provide the Committee with this flexibility: "Development well completions shall be designed to provide adequate protection from permafrost thaw loading. Such well completions shall: -9- a. Be capable of withstanding the effects of thaw to protect the integrity of the well; or b. Employ a means of limiting thaw to protect the integrity of the well; or c. Employ other means acceptable to the Committee." The testimony being presented today is contained in a report submitted here- with entitled "Prudhoe Bay Field Permafrost Casing and Well Design for Thaw Subsidence Protection" dated May, 1975. At this time, I'd like to give you copies of this report, submit them as our exhibits for the entire presentation. MR. MARSHALL: Very good. We will designate these as Exhibit #1, Atlantic Richfield. Thank you. MR. ROCHON: This report presents a well completion using 13-3/8", 72#/foot, N-80 Buttress casing through the permafrost. The report clearly demonstrates that this casing provides a safe completion in the permafrost, and therefore, we would like to request approval of this design by the Committee. Now the rest of the presentation will be a summary of the report that you have in front of you, the slides that are being presented here are selected slides from the report identiCal to the ones that are in the report and just for brevity we have not included all of the slides or all of the data that is in the report. Since the initial development of the Prudhoe Bay Field, ARCO, BP and Exxon have performed numerous laboratory, field, and engineering studies in an effort to more fully evaluate thaw subsidence problems. In 1969, BP Alaska cored the permafrost at Prudhoe and had a variety of laboratory studies -10- performed on the cores. In 1970, BP Alaska started two single well thaw subsidence tests. In 1973, Atlantic Richfield and Exxon started a field test to simulate thaw that might be expected after about 15 to 20 years of producti on. In conjunction with the field test, ARCO and Exxon independently developed mathematical models to predict the permafrost thaw subsidence behavior. These models were then used to extend the field tests and explore permafrost property variations. In addition to these measurements of permafrost behavior, there have been several tests and studies to determine the strain capacity of casing loaded by permafrost subsidence. A series of axial, tensile, and compressive strain limit tests have been run on samples of 13-3/8" , 72#/foot, N-80 Buttress casing. In addition, a finite element model has been developed for the 13-3/8", N-80 Buttress threaded casing that provides a general method of predicting the minimum ultimate tensile and compressive strain capacities of this casing connection. The report is a summary of the most important conclusions from these efforts. The report includes a section covering the Prudhoe Bay Field five spot thaw subsidence test, the permafrost thaw strain predictions, allowable casing strain limits, permafrost geology, and permafrost thaw predictions. Several conclusions were reached in this report and I would like to read these at this time. 1. Thawing of the permafrost causes permafrost compaction and strain in the casing. 2. Casing is available which can withstand strains in oil well completions through the permafrost interval. -11- 3. This permafrost compaction-casing strain relationship requires a new casing performance criteria that utilizes tensile and compressive strain capacities of the casing coupling in contact with the permafrost. 4. The strain producing mechanism is complex and is directly related to the amount of thaw in the permafrost. Strains are due predominantly to an interaction of alternating layers of sands and silt and are dependent upon the relative thicknesses and the mechanical properities of these layers. 5. We can predict the maximum expected tensile or compressive strains as a function of depth and radius of thaw. 6. The maximum strains measured in any casing joint in the five spot thaw test that we have run after 16 months, which represents a producing period of approximately 15 years for a mid-structure Prudhoe well as proposed herein, are .13% in compression and .08% in tension. 7. The worst case strains calculated for our proposed well designs are 0.7% compression and 0.5% tensile. These worst case strains are predicted by our mathematical model. Later we'd like to go into more detail about what we mean by worst case strains. 8. Full scale tensile and compressive tests and finite element studies of the API N-80 Buttress threaded casing show that a 13-3/8", 72#/foot, N-80 Buttress threaded connection can resist minimum ultimate compressive and tensile axial strains in the pipe body of 2.3% compressive and 3.4% tensile. 9. Using these worst case strains due to the permafrost thaw and the mini- mum ultimate strain capacities in the pipe body, our proposed well design and completion program give design factor of 3.3 in compression and 6.8 in tension. I'd like to start with the slides now, so if we can have the lights cut down... -12- [Fiaure II in the report] [399] Shown on this slide is our proposed casing design well completion which incorporates a gelled oil in the 9-5/8" by 13-3/8" annulus through the permafrost and includes the 13-3/8", 72#/foot, N-80 API Buttress casing. One thing that I would like to mention at this time is that the thaw test strains were only .13% in compression and .08% in tension. Combining these values with the casing strain limits of 2.3% in compression and 3.4% in tension results in design factors of 17.7 in compression and 42.5 in tension. To support these conclusions, I would like to review the principal items contained in the submitted report, namely the field test, the mathematical models, and the studies of the casing strain capacity. A five spot thaw subsidence test was started in August of 1973. The test utilized five, 2200' cased holes in a closely spaced five spot pattern to pro-. duce a thaw representing approximately 15 years of production from a single Prudhoe Bay Field mid-structure oil well. The purpose of the test was to measure the strains i~posed on casing in the permafrost caused by subsistence of the permafrost during thaw. The test was designed to obtain the maximum amount of data and yet provide sufficient time for well completions prior to the Prudhoe Bay field startup. The test was located near the center of the field and adjacent to the site of BP's extensively cored permafrost program, thus the core data could be used in the subsidence test evaluation. The next slide I have here (Figure Appendix I-l).is a map of the five spot test. The 5 double circles shown there are the five thaw subsidence wells in which we circulated hot glycol-water to introduce heat to the thaw pattern. The squares, triangles, diamonds, circles are surface shallow subsidence monitoring wires and rods that measure the subsidence -13- of the ground from 15' to a depth of 250' and there were three temperature observation wells which are shown as the circles with the crosses in the center off to the right side. The test measurements used to evaluate the strains in the casing during the thaw test were as follows' 1. The most important measurements in the subsidence test were the casing joint length measurements taken with the Schlumberger Multiple Casing Collar Locator tool. This tool provides casing length measurements of sufficient precision that we can compute length changes with a standard deviation of .01" in a 40' casing joint which is much more accurate than we can measure in casing on the ground. This is equivalent to a strain of +_0.002%. There has been a paper presented at the 1974 Fall SPE meeting covering the development of this tool. It is titled, "Precise Joint Length Determination Using a Multiple Collar Locator Tool". The next slide (Figure Appendix 1-12) shows the effects of the strain on the casing. This is one particular well, the center well, measured by the Multiple Collar Locator Tool. From the left to the right are a sequence of logs run during different time periods in the thaw test. At the top of the scale is strain + and - . Left is - which is compressive strain, to the right is + in tensile strain. As can be seen the strains build up to the last log run that is shown here of sixteen months. You can see alternating strains of compression and tension which we will discuss later. The base of the permafrost is shown on this slide at 1850'. The base of the first gravels which will be discussed more fully later is shown at about 400'. This slide takes the data from the sixteen month curve on the previous slides and plots it on the right side of this curve. (Figure ApPendix 1-13). -14- Again we have the strain in %, + or - on the curve to the riQht. On the curve to the left is a gamma ray curve used to evaluate the lithology in the permafrost. The center curve is just a schematic of the data from the gamma ray showing sands and silts, silts to the right and sands to the left. As you can note on this slide, at about 900' is the maximum compressive strain that we saw and as I mentioned that's .013% and it is opposite a massive silt. The maximum tensile strain is at about 1000' and it is .08% and it is opposite the massive sand. 2. Surface and near surface subsidence movements were calculated from periodic measurements of the elevations of the subsidence monitoring wires and rods. The total movement after 16 months of thaw indicated only a few tenths of a foot subsidence existed down to 250' and that the movements diminished with depth. 3. A measurement of the relative movement of the soil and casing was made by placing radioactive bullets in the formation and using a gamma-ray collar locator tool to record the relative position of the gamma-ray bullets in the formation with the casing collars. Although there was some scatter of the data due to the precision of the measurements, we conclude that there is little, if any, relative movement between the pipe and the formation. 4. Temperature surveys were run in the three, 400' temperature observation wells throughout the test. The temperatures measured in these wells correlated very closely with our thermal models. 5. Pore pressures in the thawed region were measured with pressure trans- ducers attached to the outside of the casing. Included in the report is a plot of the estimated pore pressure profile after 16 months and this slide will be presented at a later time. -15- The five spot thaw pattern was analyzed with an X-Y two dimensional thaw simulation model. The progress of the thaw at 1000' level is shown on the next two slides (Figures Appendix I-6 and 1-10). This slide represents the thaw pattern after over 3 1/2 months at a depth of 1000'. As you can see, there are five individual thaw radiuses around each of these wells which are spaced about 22' apart. As time progressed, and after 16 months this thaw pattern expanded into one single thaw radius and this would be what we would simulate a producing well. It's not exactly a circle, but very close to it. This slide takes the data as presented on the previous slide and will measure it at many .depths and have plotted a profile of the thaw at various time periods during the thaw subsidence test versus depth. All the data from the thaw test indicated a low level of strain in the casing, and as stated previously, the maximum measured compressive strain was 0.13% in compression and 0.08% in tension. Casing strains resulting from the large thaw radius have been measured with good precision in this field test. In order to extend casing strain predictions across the field, mathematical models were developed. These models were then used to calculate strains for different lithologic distributions. The problem of computing the expected strains at a given location in the permafrost is complex. There are three important casing loading mechanisms that influence the expected subsidence strains. 1. An inward lateral movement of surrounding frozen permafrost. 2. An upward elastic flexing of the soil below the thawed region. 3. A tendency for downward movement of thawed material generally throughout the thawed region caused by the mass of the thawed material, the gravitational field, and the drop in pore pressure. -16- The overall subsidence loading mechanism is illustrated by this next slide (Figure Appendix II-l). As we said, the area within the thawed region has a drop in pore pressure due to the melting of the ice and a reduction in the volume of ice to water. This reduction in pore pressure causes an inward movement of the frozen front shown on the right side and left side of this thawed area, causing compressive forces within the permafrost. Secondly there is an upward movement of the soil at the base of the permafrost due to this difference in pore pressure. At the top near the surface you can see the downward sagging which is exaggerated on this slide, movement of the thawed area due to gravitational forces and the downward movement of the thawed material. The maximum expected permafrost and casing strains result from the inter- play of downward acting forces within the thawed region and from the horizontal compaction of the permafrost in the intervals where there are alternating layers of stiff sand and compressible silts. The axial tensile and compres- sive strains produced by this interaction are caused by the differences in compressibilities of the layers of sandl~and silts. This next slide (Figure Appendix II-2), as can be seen the boundary of the thawed permafrost moves toward the wellbore, it laterally compacts both the layers of the sand and the silt. Since the sand is less compactible than the silt, lateral compaction causes the sand to expand vertically into the silt thereby generating tensile stresses in the sands and in the adjacent casing. The compressive strains in the silt...~and adjacent casing are caused by the compaction of the silts. The magnitude of the tensile and compressive strains is highly dependent upon the layer of thickness, their mechanical properties, the thaw radius, and the depth of the interval. -17- Another important thaw subsidence mechanism is the reaction of the pres- sure boundary at the base of the permafrost. The base of the permafrost is deflected upward because of the difference of pore pressures between the thawed region and the formation below the base of the permafrost. The upward movement at the base of the permafrost causes axial compressive strains in the permafrost and casing above the base and tensile strains in the unfrozen forma- tion below the permafrost. The third mechanism predominates in the first gravels which includes the upper 400' to 500' of the permafrost. Very little vertical strain in the perma- frost or casing has been measured or was expected in this interval for two reasons. First, the gravel appears to be underconsolidated. This undercon- solidation leads to a low soil shear strength in the frozen material. Shear failure in the frozen permafrost essentially at the thawed boundary permits the thawed soil and the pipe to move down together. This limits the differen- tial vertical strain in this region. Second, permeability is apparently high in these gravels and measurements show that the pore pressure is essentially not reduced below a hydrostatic head of water during the thaw. This factor also tends to limit the thaw strain. The discussion of the permafrost geology to be presented later will help clarify the difference~' present in the top 400' to 500' of the permafrost. The numerical subsidence model was developed from a basic understanding of the subsidence mechanisms and measurements of mechanical properties of the permafrost materials. The quality of the fit between the calculated perfor- mance and the test measurement is shown on several figures in the report for [Figure Appendix II-9] various periods during the thaw test. Shown on the next slide is a comparison of the computed strains versus the thaw subsidence test strains after 16 months of thaw. The stars indicate the major strains from the permafrost thaw test, -18- the circles indicate the strains calculated by our mathematical model. As you can see, they are very close. Again at the base of the permafrost, below the base you can see the tensile forces due to the upward flexing at' the base. You can see the compressive forces just above the base. The alternating strains as we have discussed are also shown on the slide. The determination of maximum strains across the field was handled by a worst case approach using expected thaw radius versus depth and conservative values of pore pressure loading and mechanical properties. I'd like to clarify this worst case a little more at this time. We really don't think this is the worst case that exists in Prudhoe Bay field, but what we calcu- lated is the worst case that we thought would ever exist in the Prudhoe Bay field with conditions that we couldn't really see or that we couldn't anti- cipate. Therefore we made all the parameters that were put into the model study the maximum limits to determine what the maximum strain could be. Since these studies included all of the expected lithologic combinations in Prudhoe, the results are applicable across the Prudhoe Bay Field. The maximum possible compressive strains occur in highly compactible silt layers. As sand layers become very thick, the silt layers are essentially isolated from one another and consequently undergo maximum strain. Now consider thin sand 'layers located in thick silt sections. Because of the greater stiffness of the sand, tensile strains are induced below the sand and increased compressive strains are induced in the silt above the sand layer. As the sand layers become widely spaced, they become essentially isolated from one another and computed strains reach a maximum. As the thickness of the isolated sand layers increases, its resistance to deflection -19- and thus its effect on contiguous silts increases and approaches a maximum effect. Shown in the report are the envelopes of maximum tensile and com- pressive strain drawn to include all maximum worst case values. These will be shown on a slide at a later time. At shallow depths, the largest values of strain are computed for isolated layers of sand, in thick silt sections. Near the bottom of the permafrost, the largest values of compressive strain are computed for the thin silt layers in thick sand sections. The final pore pressures in the thawed region of the producing well will probably be higher than the pressures measured in the highlY accelerated thaw test. However, for the worst case approach, we have again assumed an even more conservative pore pressure profile than found in the five spot test. On this graph (Figure Appendix II-3) we have shown depth versus the pore pressure and we have several curves shown here. The line to the far right is a hydrostatic curve, the .horizontal lines shown from the top to the bottom are the range of the pressures that were measured in the thaw test from the pressure transducers. In the top 400' we have used the dotted line that is shown here, a hydrostatic gradiant, and as you can see it is less than the pressures that were measured through that interval. The solid line from that point down to about 900' is the pressures that were used in the model study and again looking at the horizontal lines representing the measured pressures, we are below those values again. From 900' down to about 1650' we have used a zero pore pressure and again you can see that the measured pressures in the thaw test were higher than this value. From 1650' to the base of the permafrost we have used a limiting compaction approach for pressures and again these pressures are less than were actually measured in the thaw test. -20- The earliest industry concerns for the subsidence of the upper 500', or first gravels, were based on the concern that this interval might contain excess ice and that thawing would cause large permafrost and casing strains. Since these early studies, there has been considerable evidence developed that the first gravels do not contain excess ice below about 50'. Differential verti- cal strain in the thawed region will be minimized by shear failure in the surrounding frozen material. Small vertical strain is certainly shown in results of the five spot thaw subsidence test. Shown in the report are maximum strains in the top 430' which are less than half of the strains mea- sured below that point. The maximum calculated thaw subsidence casing strain for again as I say worst case conditions for the well design presented earlier after 20 years of thaw was calculated by the model to be 0.7% in compression and 0.5% in tension. Now at this time I'd like to turn the testimony over to Dick Knowles who will cover the casing strain capacities. JOHN SCOTT' Dick Knowles will begin with his qualifications statement. [63.7] DICK KNOWLES' Mr. Marshall and members of the Committee, my name is Charles Richard Knowles, Jr. I received a Bachelor of Science degree in Petroleum Science from Marietta College in 1963 and a Master of Science degree in Petroleum Engineering from the University of Texas in 1966. My thesis work was on the prediction of flow patterns that evolve in two phase horizontal flow of hydrocarbons. This research has been the basis for the design of large diameter flow lines in the Middle East and is presently being used to size production equipment and flow lines in the eastern half of the Prudhoe -21 - Bay field. After graduating from the University of Texas in 1966 I worked as Drilling Production Engineer for Pan American Petroleum Corporation in Louisiana before becoming employed with the Drilling Reseach section of Atlantic Richfield Dallas Research Center. I came to Alaska in February of 1970 as Senior Drilling Engineer. Since September 31, 1974 I have held the position of District Drilling Engineer with the primary responsibility for the supervision of the drilling, engineering and operations planning for the North Alaska District. My work on the Arctic slope has led to several patents in various fields, including downhole drilling equipment, surface instrumenta- tion as well as others. I have developed downhole instrumentation for outside the casing strains which has yielded data that has significantly changed the drilling completion practices in Alaska and the Canadian Arctic. I am a member of Phi Epsilon Tau Engineering Society and a registered professional engineer in the State of Texas. MR. MARSHALL: Hearing no objections, we accept your qualifications, Dick. MR. KNOWLES: Thank you, Tom. My part of the testimony today will deal with the more mechanical aspects of our study, in particular the model studies and the physical experimental work that we did to determine the strain capa- bilities of the casing and the interrelation of this with the thaw test data that you have seen earlier. In addition to predicting the maximum casing strain that will be imposed by the thawing the permafrost, we must determine the Strain which our proposed casing can withstand. We undertook laboratory and mathematical studies to determine this limit. This work has been published in SPE paper #5598 presented at this last fall conference of the Society of Petroleum Engineers, titled,"Casing Strain Tests of 13-3/8"~ N-80 Buttress Casing',. -22- Full scale strain limit tests were conducted to establish the tensile and compressive strain limits for 13-3/8", 72#/foot, modified N-80 Buttress casing. We did six tests. Three of the tests were compression and three were tension. The axial and hoop strain limits were measured at several places on the pipe body external to the coupling, inside the surface of the pipe body, opposite the thread', and on the outside of the collar itself. Data at a point 1" from the coupling indicates that the strain at this location at the time of failure of the connection was between 1.9 and 3.4% on the compressive strain tests. The strain 1" from the coupling at the time of the connection failure was between 2.75% and 4.05% in the tension tests. If we might have the lights off, we will put up a slide that is a table showing these test results. This is a table showing the summary of the results for our compression and tension tests for this particular type of casing. And again, we are talking about 13-3/8", 72#/foot, normalized N-80 Buttress casing. As you can see the test numbers on the left indicate that the first three are compression, the second three are tension. The next column shows the load in pounds at the point of failure. The next column shows the pipe strain with the indication of 1" from the coupling and this was at the point at which the pipe body started to strain. The first two are the point at which the. pipe body starts to strain, the following two columns are the load and the per cent of strain 1" from the coupling at the point of failure. That is the mode of failure. The minimum ultimate casing strain capabilities of this 13-3/8", 72#/foot, modified N-80 Buttress threaded connections was calculated using a finite element -23- model. The mathematical model was calibrated and the strain limit failure criteria were established using the results of these full scale tests. Failure in compression is indicated in the model when the first tooth on the pin loses its sealing force, which represents a potential leak. Failure in tension is defined in the model as the magnitude of the average strain in the pipe body at the root of the last loaded tooth on the pin. The ultimate strain capabilities of permafrost casing are defined as the average axial strains in the pipe body that are required to fail the coupling. The mathematical model predicts a 3.3% ultimate strain capability for a single compressive load on normalized N-80 casing with normal thread makeup. Since we are talking about casing that is going to see additional forces other than the test, we took into account the effects of pressure and temperature on this casing strain. The addition of pressure is a beneficial effect on the calculated compressive strain capacity. For combined loads of 1800 psi internal pressure and a thermal load representing a IO0°F increase in temper- ature on the casing strain, the model indicates a compressive strain limit of 3.9~o.°~ A calculation was also performed using properties of quenched and tempered modified N-80 casing. The results indicated that quenched and tem- pered N-80 develops even higher ultimate strains than normalized N-80. The results of quenched and tempered pipe show that failure has not occured with a 7.2% compressive strain in the pipe body. The model was also used to evaluate the sensitivity of the strain to the limit of makeup or interference between the pin and the collar for this size Buttress connection. With 40% less than the normal makeup or normal inter- ference in the pipe thread, the ultimate compressive strain capacity is re- duced to 2.3% strain. With more than the normal makeup, the ultimate compressive strain capacity is increased beyond 3.3%. [730] -24- The ultimate strain capacity for tensile loads is higher than for compres- sive loads. The ultimate tensile strain capacity for normalized N-80 casin~ under normal conditions is 3.7%. The addition of 1800 psi internal pressure decreases the strain limit to 3.4%. Reducing the makeup interference apparently has little effect on the tensile strain capabilities of this joint. A calcu- lation using 40% less than normal makeup interference indicates a tensile strain limit of 3.7% which is the original 3.9% which is very close to the normal makeup condition. The next slide (#13, Table III-l) summarizes the finite element calculations and this is again very close to the format you saw before with the first three tests being the compression tests. We are again looking at this particular specimen, the position of the calculated strain where on the slide before we were looking at the position of the actual strain. Shown in the report which we have given here a few moments ago are the ultimate strain predictions for connection failure of 13-3/8", 72#/foot, N-80 Buttress casing and a comparison of the measured and the computed strains in compression and tension. From these studies, we conclude that the minimum ultimate strain capabilities of 13-3/8", 72#/foot, N-80 Buttress casing are 2.3% compression and 3.4% for tension. Now these are not the maximum strains we calculated from the model or measured in our test. These are again a worst case prediction based on the sensitivity of the parameters of the effect of strain carrying capabilities of the connection. This slide is in essence a summation of all the data we've been talking about this morning. Plotted along the bottom are per cent strains working from a neutral point there at zero, to the left is compression, to the right is tension. [#14, Figure II-2] -25- The vertical axis is depth. The dotted line up the middle is of course neutral strain with no load on the casing. The next line out on both the tensile and compressive side show the maximum compressive and tensile strains that were measured in the test. The sensitivity study done during the test indicated that there was a worst case that may be possible using the parameters that were determined in the test. This is shown as the next set of lines out and on the right we are showing maximum tensile strain for a worst case analy- sis. On the left we are showing the maximum compressive strains for the worst case analysis. Then far outside of those we see the minimum ultimate strain capabilities of the casing and again, this is the worst case calculation. So we have in essence brought in the strain capabilities of the casino by driving them towards the center using the worst case analysis techniques and we have expanded the tensile and compressive strains outward using the worst case analysis, and we still see a comfortable design factor between the two. As we have discussed earlier this morning using this worst case approach our proposed well design completion program using design factors of 3.3 in compression and 6.8 in tension, using the actual strain measured in the Prudhoe Bay test, we find that these design factors are 17.7 in compression and 42.5 in tension. This is the end of my testimony and if there are any questions, I would like to answer them. MR. MARSHALL: Thank you very much, gentlemen. That was a very good testimony and very explicit slides. You have addressed yourself, the Atlantic Richfield group, to Item No. 2 on our agenda, that is amendments to Rule No. 3, and at this time it would probably be appropriate if we would direct questions to Item No. 2, Rule No. 3. Do you have any Questions, Hoyle? -26- MR. HAMILTON: Yes, I have a few questions. Mr. Rochon, getting back to your field test, subsidence thaw test, clarify for me the worst case again. After you were doing your field test, you used your mathematical model to simulate the test, then you got pretty good agreement as shown from your slide. Then you selected the type of the lithology that would give you your worst case, and you calculated the maximum strain that would be under this type of litho- logy. Is that the way you developed your . . .? MR. ROCHON: That's right. We developed a model and it was calibrated against the field test to see how accurate it was and we modeled very closely the results that were obtained in the field test. Then to say that we have a test in one place in the field, how do we extrapolate that across the entire field? We were a little concerned about the exact correlation across the field and the geology which we are going to cover a little later, therefore to be sure that we covered every possible lithology that might be encountered, we did a sensitivity test and this is really what the maximum worst case is, is a sensitivity test and looking at a very thin silt between massive sands near the base of the permafrost, that was the worst case at that location. So those are the worst strains the we calculated at that location. Up shallow massive sands, or excuse me, massive silts and small sands were the maximum strains at that location. So those were the maximum strains that we have put in our worse case at that location. MR. HAMILTON: What type of subsidence did you get at the surface around your well bore? MR. ROCHON: There was some fairly, well, the average subsidence in the area between the five wells was about two feet, a little over two feet after the -27- 16 to 18 months of thaw. About 20, 25' from the center it was down to less than a foot. So I would say the maximum was about 2 1/2' around the center well. MR. HAMILTON: Did you also measure the change in elevation of your well head, had that changed? MR. ROCHON: Yes, the well heads initially went up because of the thermal effects, then as the thaw went on it went down slightly, which ties in with the strain predictions we have made and then from what we saw in the test itself, the strains in the upper joints, and the surfaces went down about .2 to .3'. The top of the casing went down .2 to .3 of a foot, two or three inches. MR. HAMILTON: Mr. Knowles, in your testimony you indicated that your test you made on your 13 3/8", 72#/foot, N-80 , the strain capacity was by far exceeding any strain you saw in your thaw subsistence test. Have you done any tests on any other grades and weights of casing to indicate that they might also be acceptable? MR. KNOWLES: No, Mr. Hamilton, we have looked very explicitly at this particu- lar weight and grade connnection. To give you some background into what this entails, to give you a feeling for the magnitude of this test, there are only three places in the Continental U.S. that are capable of full-strain, full-scale strain tests for 13 3/8". We used the facilities at the University of California at Berkeley. The mathematical analysis that was used to confirm these experimental results was developed over about an 8 month period so that the whole strain analysis is not something that is necessarily undertaken like other reference material. This is the only joint of connection that we lOoked at, the only weight and grade of pipe that we looked at. There are other joints, there are other weights, there are possibly other grades of pipe that may also satisfy this that we could not look at. -28- MR. MARSHALL: We've been in session a little over an hour now and let's take a 15 minute break and meet back here at 25 after 10. We have a little getting together of our heads for some additional questions. PART I I MR. MARSHALL: We will reconvene our hearing on Conservation Order No. 137. Mr. Scott, you have further testimony at this time. We've decided we would like to hold our questions until the BP people have testified on some of the materials in your exhibit #1. MR. SCOTT: I understand BP has a witness to testify on this point, so I will yield to Mr. Reeder. MR. MARSHALL: We will have mechanical type questions that will follow BP's testimony. MR. REEDER: Chairman, we have two witnesses to speak on aspects of the Atlantic Richfield presentation and I have copies of their testimony here. Before they begin I have two items I Would like to cover with the Committee. The first is a letter dated June 30, 1975 to the Director of the Division of Oil and Gas, Mr. Gilbreth, which I would like to put in the record as an exhibit. I will read the letter. "Dear Mr. Gilbreth, Atlantic Richfield Company has submitted a report to you entitled,'Prudhoe Bay Permafrost Casing and Well Design for Thaw Subsidence Protection' which summarizes the recent investigation into the problem of permafrost thaw subsidence. BP Alaska has been involved in the planning and execution of these investigations and supports the conclusions." I believe the Committee has the original and I have copies here. -29- MR. MARSHALL: Mr. Reeder, hearing no objection from you we will submit this into the record as British Petroleum Exhibit #1. MR. REEDER: All right. Mr. Chairman, I have a further point and this is somewhat out of context. It is in relation to Rule 3. It's more a matter of information for the Committee and specifically with respect to Subpart A of Rule 3. You will recall that this subsection reads as follows: "The casing and cementing program shall provide adequate protection from all fresh waters and productive formations and protection from any pressure that may be encountered." For the general information of the Committee, BP Alaska is presently drilling a well located near its Gathering Center #2 and is coring the Lower Tertiary and Upper cretaceous sands and we will.be testing the sand and we will report to the Division of Oil and Gas information it may learn with respect to water bearing formations in these sands. I think this is the first time this has been actually tested. I believe we are now in a position to present our first witness commen%ing on the Atlantic Richfield report and that is Mr. R. H' Clarke, who is a geologist with British Petroleum. [893] MR. CLARKE: Mr. Chairman, members of the Committee, my name is Richard Hedy Clark. I am a geologist in the Production Planning Department, BP Alaska Inc., in San Francisco. I received a BSC degree in Geology from the University of Bristol, England in 1963. In 1966, I obtained a Ph.D degree from the same university. -30- After seven years' offshore exploratory research, concerned mainly with Tertiary and Quaternary deposits of the Continental shelf around the British Isles, I joined British Petroleum Co. in 1970. After two and a half years of exploration geology in the North Sea, I was transferred to A.D.M.A. Co. in Abu Dhabi where I was mainly concerned with production and development geology. In April, 1974, I was transferred to BP Alaska Inc. and have since been working on various aspects of the development of the Prudhoe Bay Field. MR. REEDER' Will the Committee accept Mr. Clarke as an expert witness? MR. MARSHALL' Hearing no objection, we do. MR. CLARKE' Gentlemen, the geological testimony I am presenting is included in the Appendix of the ARCO Exhibit No. 1. My .testimony summarizes two aspects of the geology of the permafrost in Prudhoe Bay, the Stratigraphy and Age of the Sequence and Geological Factors Revelant to the Consolidation and Ice Content of the Permafrost. I will begin by considering the strati- graphy and age of permafrost sediments which is the first section of the Appendix. The permafrost zone at Prudhoe Bay includes an upper "First Gravels" interval which is 200 to 500' thick and which overlies a thick sequence of sands and silts, the Sagavanirktok Formation. As illustrated by the gamma ray log correlation diagrams, which are included as Figures IV-1 and IV-2, and covering Mat IV-3 which were prepared by Mr. Eason with ARCO, the SaRavanirktok Formation contains several correlateable silt horizons which dip gently to the northeast across the field area. The first gravels contained no correlateable silt intervals and form a homogenous, northeasterly thickening clastic wedge. Both intervals of the permafrost are continuous across the Prudhoe Bay Field area. The base of the permafrost was determined from temperature measurements and by log evaluation and lies in the Sagavanirktok Formation, as shown on the cross sections. -31 - Palynological examination of the BP permafrost cores by both ARCO and BP have shown that the First Gravels were deposited in an Arctic climate similar to or slightly to or slightly warmer than that of the present day. The under- lying formation (Sagavanirktok) contains a rich pollen flora which can be correlated with early and middle Miocene floras found elsewhere in Alaska. The climate at that time was cool and humid. The stratigraphic sequence at Prudhoe Bay has been related to the late Cenozoic history of the Brooks Range to the south. I refer you to Professor Porter's paper in the American Journal of Science. 'Sagavanirktok Formation represents the products of the erosion of the Brook Range during the Tertiary, which virtually ceased with the onset of Plio- Pleistocene climatic cooling. The First Gravels record a relatively recent phase of glacial erosion of the Brooks Range, particularly during the Wisconsin and Recent or Holocene I should say..Superficial deposits at Prudhoe Bay closely resemble those in the First Gravels beneath. Because no correlateable marine intercalations have been found in the First Gravels it is probable that this interval is nearly all Wisconsin to Recent or Holocene in age. My second summary section is entitled: Permafrost Consolidation and Ice Content. The Sagavanirktok Formation, including the permafrost interval below the first gravels, was deposited in a humid, cool, late-Tertiary environment long before the development of permafrost on the North Slope. The normal consolidation profile which the formation acquired during deposition was not significantly affected by the refrigeration and permafrost development which occurred in the upper part during the Pleistocene. In fact there is some evidence that cyclic permafrost growth and degradation during the Pleistocene may have contributed to the stiffness of the soils in the deeper layers of the existing permafrost interval. -32- The First Gravels are the products of stream deposition in an environment resembling that of the present day. The pollen evidence suggests that most of the deposition took place in slightly warmer conditions, probably when the various valley glacier systems recognized by Porter were decaying and releasing large volumes of water and glacial detritus. Although ice structures have developed both in the near-surface interval (0 to 50 feet) at the present time, and probably did so on older surfaces within the First Gravels interval, the burial of these older surfaces was accompanied by the destruction of the excess ice structures such as icebergs. This may have happened in at least two ways: (1) Firstly, the larger thaw lakes are associated with substantial local thaw of the permafrost beneath, especially of the decaying frost polygon structures which contribute to their development. (2) Secondly, the streams which bring in the sands and gravels to bury existing land surfaces also represent a major source of heat input for the near-surface permafrost. A thaw bulb develops below and adjacent to each active stream of sufficient dimensions to thaw the previous generation of near-surface ice structures. Evidence of this process is available from the modern Sagavanirktok River. I refer you to Sherman's paper in the proceedings of the Second International Conference on Permafrost. Because the First Gravels may have remained frozen after burial to depths greater than those influenced by the thaw beneath streams and lakes, the deeper gravels may be somewhat underconsolidated. -33- To summarize: The First Gravels contain free ice structures in the near-surface interval (about 0 to 50 feet), but the deeper part of the First Gravels are free of excess ice although probably somewhat underconsolidated. The deeper permafrost is more or less normally consolidated and free of excess ice. The sequence of events, and the stratigraphy of the permafrost are outlined on Table IV-1. And on this table I have four columns: Formation, Age, Flora/Climate, and Permafrost. The uppermost formation, the First Gravels are almost certainly late-Pleistocene to Recent or Holocene in age. The evidence from the cored well is that an arctic tundra climate prevailed during the deposition of these gravels, only moss, fern, pollen and spores were recovered with occasional reworked tree pollen. The climate and the permafrost presumably were similar to that of the present day. There was then a large gap in the sequence and I have recognized this gap by correlation of the sequence of Prudhoe Bay with the Pt. Barrow sequence and I believe this gap represents most of the Pl io-Pleistocene time. We have no record but there is external evidence that the climate was probably frigid with several episodes of marine submergence and it was during this interval the permafrost must have developed in the Sagavanirktok formation. The lower formation in the permafrost interval is the Sagavanirktok Formation which is dated as Tertiary to Mid-Miocene and contains a flora indicating a cool temperate forest association and the absence of permafrost. That's my testimony. MR. REEDER: Are there any questions of the Committee to Mr. Clarke at this time? MR. MARSHALL: Thank you, Mr. Clarke. -34- MR. REEDER: Mr. Chairman, we have one further witness, Mr. P. R. Judd, who is District Drilling Engineer for BP Alaska here in Anchorage. MR. JUDD: Mr. Chairman, members of the Committee, in December of 1963 I graduated from the University of Washington in Seattle with a Bachelor's of Science degree in Civil Engineering. At that time, I was employed by Shell Oil Company in Los Angeles. After receiving training and initial work assignments in Ventura, California and Houston, Texas, I was transferred here to Anchorage in November 1964 to a position as Drilling Engineer. During my five year assignment in Anchorage, I was involved with engineering and plannina for drilling in Cook Inlet, both offshore and onshore and on the North Slope. In November 1969 I was transferred to Bakersfield, California as Senior Drilling Engineer involved in drilling engineering in the San Joaquin Valley in California and in the State of Utah. In August 1970 I was transferred to Houston, Texas where I was involved with a wider range of drilling activities, both offshore and onshore, including wells to below 23,000 feet. In 1971 I returned to Anchorage and joined BP Alaska in October of 1971. With BP I have been involved with drilling on the North Slope, primarily in the Prudhoe Bay field. My assignments have included both engineering and direct supervision in the field. In September 1974, I received my present position as District Drilling Engineer. I am in charge of BP Alaska's Drilling Engineering section and am responsible for engineering and technical support for drilling and well completion operations. I also assume the -35- Drilling Superintendent's duties in his absence. I am a member of the Society of Petroleum Engineers. MR. MARSHALL: Hearing no objection, we accept your qualifications, Mr. Judd. MR. JUDD: Mr. Chairman and members of the Committee, my testimony is in regard to Rule No. 3, Section B of Conservation Orders No. 98-A, 98-B and 83-C. As already stated by Mr. Reeder, BP Alaska Inc. supports the conclusions of the Atlantic Richfield Company's report entitled, "Prudhoe Bay Field Permafrost Casing and Well Design for Thaw.Subsidence Protection", dated May, 1975 which has been submitted to you in their testimony today. We have been kept fully informed throughout the planning and execution of the field test and associated studies summarized in that report. We have satisfied ourselves that the concepts are sound and that the data used fell within reasonable bounds. This work has resulted in a far better understanding of the effects of thaw subsi- dence and casing design required to withstand thaw induced strains. The report has shown that permafrost thaw can result in significant casing strains and that "worst case" strains can be predicted for purposes of design. We should like to emphasize "worst case" in that the predicted casing strains that have been generated from the computer model studies represent the effect of a thaw radius resulting from a combination of severe, or "worst case" conditions, such as continuous sustained fluid production over a period of 20 years with the worst combination of lithological conditions in the soil. The resulting strain predictions, then, form a basis for conservative criteria for use in designing casing and completion programs having more than adequate safety for -36- withstanding thaw induced strains. It is well to keep in mind that the maximum strains observed in the ARCO field test which, after all, is the most likely situation we will observe during production, are significantly below yield. ARCO's report also shows casing strings are commercially available which can withstand the worst case design strains with an adequate design factor without failure of the connections. Specifically, the report shows that a 13-3/8" diameter, 72#/ft, N-80 grade casing with API BUttress thread con- nections can easily resist the predicted thaw induced strains. We agree with this conclusion and also feel that other casing strains may also be suitable. That is, other combinations of size, weight, grade, and/or connection thread may be suitable for well design configurations which differ from ARCO's current well design. With a view toward extending this knowledge, we have carried out labora- tory testing of casing like that already set in two of our existing wells, to determine the post yield strain characteristics of that casing. Preliminary results indicate that this particular casing may not have satisfactory post yield strain performance. Based on this experience, we would s.uggest that laboratory testing or mathematical analysis should be' required to show that where necessary other casing strings could withstand thaw induced strains. The current Prudhoe Bay Field Rules permit only three specific types of well design with respect to permafrost thawing. Rule No. 3, Section (b) of Conservation Order Nos. 98-A, 98-B, and 83-C states, "Wells shall be protected from damage caused by permafrost thawing by the use of refrigeration and/or insulation or by the use of slip joint casing". Based upon testimony pre- sented today, we recommend broadening the rule to reflect new knowledge -37- gained to date and to allow any future knowledge to be incorporated in possible new casing and well completion designs. Any new rule should also take into account the fact that thaw induced strains are directly related to the amount of thaw in the permafrost and only become significant for wells producing large volumes of warm fluids over long periods of time. That is to say, injection wells, gas wells, observation wells, low volume oil wells, and exploratory wells need not necessarily be designed to meet the same criteria for permafrost thawing as previously discussed. As a more specific example, time limitations on oil production would limit the amount of thaw and could be a method of reducing the associated strains to acceptable levels in wells for which other methods of designing for thaw induced strains are impractical. Similarly, reduction of the amount of thaw through the use of insulation has already been recognized as an acceptable procedure. Any new rule should also consider that application of thaw induced strain criteria is only significant for casing strings playing a role in the integrity of the well. As an example, predicted failure of the outermost casing connec- tions in the permafrost zone due to worst case thaw induced strains is accept- able if the integrity of the well is not affected. The alternating axial and localized nature of the strains associated with the permafrost thaw are such that the effect on casing cemented to the permafrost would not be transmitted to the next inner string if the two strings are not cemented together through the permafrost interval. We support the proposed field rule as stated by Atlantic Richfield Company earlier which is repeated here for your consideration: "Development well completions shall be designed to provide adequate protection from the effects of permafrost thaw loading. Such well completions shall: -38- a) Be capable of withstanding the effects of thaw to protect the integrity of the well; or b) Employ a means of limiting thaw to protect the integrity of the well; or c) Employ other means acceptable to the Committee." We believe that the current practice of the Division of Oil & Gas in requiring notification of the plan for completion of a particular well pro- vides adequate well-by-well procedure for monitoring proposed designs to ensure that they comply with this broadened rule. Assuming Committee approval, our current, plans for the completion of the initial oil wells to meet the startup of the Trans-Alaska Pipeline may be grouped into four categories as follows: Group 1) Those wells which will employ an outer casing string through the permafrost, which can withstand thaw induced strains. Group 2) Those wells which will employ high-grade insulated~ tubing to limit the amount of thaw. Group 3) Those wells which have an outer casing string installed through the permafrost for which the ability to with- stand thaw induced strains has not yet been demonstrated. In these wells, however, any strains in the outer casing are not transmitted to the inner casing strings. The integrity of the well is maintained even though the outer casing connections may not be able to withstand worst case thaw induced strains. Group 4) This final category contains a few wells which were drilled early in the development phase, for which com- pletion plans have not yet been finalized. We are -39- considering proposing a limited production life which would limit the extent of thaw to a safe level. We have no testimony to offer regarding other sections of Rule 3. I thank you for your attention. Do you have any questions for me? MR. MARSHALL: Thank you. We do have some questions that apply possibly to your testimony and possibly that of Atlantic Richfield Company and it would be good if possibly we could move up another chair or two to the table and get Jerry and Dick back up here, because I think our questions might overlap over both testimonies. Just to be sure, were there any other witnesses which you wish to bring forth on this question before we ask our questions? MR. REEDER: No, Mr. Chairman. MR. MARSHALL: Fine, thanks. Hoyle, . . MR. HAMILTON: Mr. Judd, I believe in your testimony you mentioned that you had done some testing on some other grades and weights of casing other than this N-80 which we discussed earlier. I wonder if you could possibly make the results of this testing available to the Committee. MR. JUDD: Mr. Hamilton, I prefer not to divulge the results of that test at this time for two reasons: (1) The record is in a preliminary stage; (2) The casing utilized a proprietary thread and we cannot really disclose the name of the casing until we have'a chance to talk to the manufacturer. MR. HAMILTON: I see. We've talked so far about internal freezing problems in casing when we brought up the statewide rule change we wanted to get made and have approved today. We've also had considerable testimony on thaw subsi- dence forces, strains primarily, and I wonder if any of you have any testimony to bring forth on freeze back problems that might be considered? -40- MR. KNOWLES: Yes, Hoyle, I might comment to that. There has been a large volume of work done on all phases of the permafrost, part of which deal with the external freezeback studies. We have published several of these papers which we feel cover the subject very broadly and these papers while they were not prepared for testimony are certainly matters of public record. May I suggest two things: (1) That your present Rule 3, section (a) on casing and cementing requirements is very adequate to deal with this and I read: "A cementing and casing program shall provide adequate protection of all fresh waters and productive formations and protection from an_9_y- pressure that may be encountered," which is very adequate to deal with it and at a later date, very soon, I would like to supply you with copies of the publications that deal with the e~ternal freezeback portions of our studies. MR. HAMILTON: Fine, we'd be glad to hold this hearing open for a period of time that would give you enough time to gather that information. MR. JUDD: Mr. Chairman, we are aware of these studies that have been con- ducted in the Prudhoe Bay and published papers on this subject and we agree that considerable pressure can be generated by refreezing the water in the permafrost. Based upon our field experience, however, we believe that possibly other factors come into play so that these pressures may not reach the levels predicted by the published work, in fact they may not reach levels which would tend to endanger the casing. We do routinely at this time design the surface casing to withstand those pressures to 55 (inaudible). MR. MARSHALL: Thank you. Any further questions, Hoyle? MR. HAMILTON: Do you have any material or evidence that you plan on submitting to the Committee along these lines, Mr. Judd? MR. JUDD: Wi th regard to? MR. HAMILTON: Freezeback pressures. -41- MR. JUDD: No. We have had some experience in looking over wells for another reason which was already covered earlier and that is on the subject of internal freezing of fluids between casing and casinghead. We are in the process of preparing a report on that subject but I am not in a position to talk about that now. MR. MARSHALL: Hoyle has some additional questions. MR. HAMILTON: We seem to have reached a point here in time where we are looking at new types of completions at Prudhoe Bay compared to what we were looking at the time that the Field Rules were written in 1970 which at that time' discussed and actually outlined in the Field Rules the use of refrigeration and slip joint casing and that was actually put into practice in the field on a number of wells up there. I'd like to get some comments from both BP and ARCO regarding how they eliminated that method of completion. Was it impractible and the wells that do have those type of completions now, do you consider them safe to produce or are you planning on reworking those wells over again? [1161] MR. KNOWLES: Hoyle, I'd like to speak to that subject for our wells. The design criteria that we used for the slip joint wells was based on the fact that at that time current knowledge expected large loads from the upper gravel sections. Since that time as Mr. Clarke pointed out we have had the opportunity to fully evaluate these upper gravel sections. We are now fully convinced that we will not see the magnitude of forces that were first considered in 1969 and 1970. For this reason, the slip joint 20" casing string is not necessary as a deterent to strain induced loads, to the inner integrity of the well. There was anOther reason at that time for the 20" string. You recall when we first started drilling at Prudhoe the surface gravels were a tremendous problem. Simply being able to make and maintain holes long enough to make another con- nection. Since that time we have looked at that problem and that is no longer a problem. I believe this is true in all the drilling in Prudhoe. The gravel string or the permafrost string as it was called at that time is not a mechanical -42- necessity any more. We can drill to 2700 feet in 3 days without having to have a string to shut off the upper gravel. As to its integrity in the present well program, we have looked at this problem and Mr. Judd alluded to this in some of his prior statements. The 20" string itself consists of K55 and H40 in the pipe body. Both of these grades of oil field tubulars have strain capabilities that if we were considering the pipe body alone with strain enough to relieve the forces induced by thaw subsidence. There are loads that would be imposed on the casing string that the 20" Vetco joint could not withstand in tension. We have looked at this. We have studied the problem. There are, before these loads can be imposed on the 13 3/8 string, they must be transmitted through cement if there is cement. We feel that the fact that the 20" joint may reach parting, load in tension, that this load would not be detrimental to the integrity of the well for two reasons: (1) The cement itself would go into a tension mode; in other words, if you had the outer joint and inner joint and they were cemented together, before the outer joint could transmit any force to the inner joint, it must be transmitted through the cement. And the (inaudible) to the cement would fail in tension. Cement has very low strength in tension and were these strains induced to the inner string our present study feels that that inner string is fully capable of withstanding loads. So for these two reasons we do not feel that they are now a detriment at all. MR. HAMILTON: Would you agree then with BP's prior testimony that with two strings inside of a string of pipe that might fail in tension or compressing strain is adequate safety for producing? MR. KNOWLES: Yes, I believe if I understand that their position is that they require as a matter of integrity for the well that you have two near pressure bearing strings inside of what is referred to as surface pipe, and any pipe outside of that has nothing to do with the pressure integrity of the well nor structural integrity. Yes, I would agree with that. -43- MR. JUDD' Mr. Chairman, this is Dick Judd. MR. MARSHALL- Yes, Mr. Judd. MR. JUDD- You asked about these earlier kinds of wells, like the refrigeration type. You asked for comments regarding why those have been discontinued. I think the word used was eliminated. I wouldn't say that they were exactly eliminated the refrigeration type, but we have discontinued using that type of well which you know is involving the circulation of a refrigerant to reduce the amount of thaw. We have discontinued this because we feel it is impractical solution, really other solutions are better or less costly. The refrigeration type involves, as I understood it, this came about before my time, Quite a bit of maintenance work and the cost involved in the drilling and casing program and in later production would probably be too high. MR. MARSHALL' Mr. Judd, pardon me for interrupting you, I think you are going to have to speak just a little louder, we have some extraneous noise here and the people in the back row might miss some of this. Hoyle, do you have any further questions? MR. HAMILTON' I may have here in a moment. Let me collect my thoughts. MR. KUGLER' I have one question for Mr. Judd. You remarked here that certain wells such as injection wells, gas wells, observation wells, low volume oil wells and exploratory wells might not. necessarily need to be designed. When we talk about low volume oil wells, is this a decision that you addressed with your company say, or with the Oil and Gas Conservation Committee? MR. JUDD' I probably should restate low volume oil wells to, and I think I clarified it later on in my testimony, talk about limited time, production on a limited time period. MR. KUGLER' In' other words, a well that has, we are talking about wells that have already been drilled and no future wells? .... MR. JUDD' Well, I made the statement in a general sort of way, but as a matter of fact, we do have some wells in the Group 4 that I discussed, in -44- which we probably will come to you at the time when we propose to complete the well with a proposal to limit the production life of those wells to a given perod of time to reduce the amount of thaw. MR. HAMILTON: Mr. Judd, in those wells that you plan on using insulation material, do you have any recommendations on what the minimum requirements might be for insulation material as far as U factor, btu's per hour, degrees Fahrenheit and so forth? MR. JUDD.' Mr. Hamilton, I'm really not prepared to give those figures today because I had understood that the insulation was already an acceptable means and I really don't have the numbers at hand. MR. HAMILTON: In one of our pool rules, namely Rule 83 C covering the Lis- bourne Pool, apparently when we wrote those pool rules back in 1970 there- abouts we omitted any discussion of pressure rating of the BOP's to be used and we would like to get those written into those pool rules and we propose using the BOP test pressures similar to the Saddlerochit and I'd like to hear any discussion along those lines, anyone having any objections. MR. KNOWLES: Yes, Hoyle, I might speak to that. We feel that's probably entirely proper because for the area of interest that we are discussing, the area bounded in the description of Prudhoe Bay Field and the Lisbourne Field, those townships and sections involved, I believe that it would be entirely necessary to have that type of preventor on location because we would have to penetrate the Saddlerochit before you could reach the Lis- bourne and I think that would be entirely proper. There may be cases later in the life of the field for instance if bottom hole pressures decrease in Saddlerochit due to pressure production decreases that certain workover opera- tions, for instance there may be exceptions to this present set of circumstances and I believe that in the forseeable future any drilling into the Lisbourne ought to be entirely proper. -45- MR. JUDD: Mr. Hamilton, Mr. Judd again. My comment on that may be very similar to Mr. Knowles. More specifically the present rule for 98B which has to do with the Saddlerochit pool requires 5000 lbs working pressure. If you are looking for a number, in my opinion that would be adequate for 98A also. MR. HAMILTON: I don't see Mr. Vickery here, I had a question of him. MR. MARSHALL: I think, just to be safe you should come up here. We do have trouble picking up voices away some twenty feet. MR. HAMILTON: Mr. Vickery, in your testimony this morning you outlined more or less a recommendation for establishing some, or putting some type of di- vertor on the conductor in Prudhoe and there's no mention here of any pressure test on that divertor or of that nature. Do you have anything in mind of a pressure test that you would like to recommend? MR. VICKERY: I wouldn't recommend a pressure test. It would have to be ~- very low because you are anchoring it on a 80' conductor pipe and without giving it any great deal of thought, I'd say 250, 300 pounds maximum. MR. HAMILTON: I just wanted to get your thoughts on that. Does anyone else have any? MR. JUDD: Mr. Hamilton, I don't have a thought on that particular topic, but I wanted to correct something I said a minute ago regarding in my opionion 5000 lbs. working pressure would be adequate and I said 98 A, I meant to say 83 C which is the Lisbourne. I wanted to make sure . . that that was understood. MR. HAMILTON: Fine. MR. MARSHALL: Do you have any further questions, Harry? MR. KUGLER: No. MR. MARSHALL: We're going to take a minute or two to sort out some possible questions. MR. HAMILTON: Mr. Vickery, is there any particular size that you would recommend in order to handle any diversion that might be necessary in drilling a shallow well? -46- MR. VICKERY: I can only speak on what we would do and what we would do at BP Alaska and their drilling operations would probably go to two 8 or 10 inch lines. The purpose of those lines would be to divert that flow away from the drilling rig and have the opportunity to get windwise any flow that came away to try to keep it from flowing back on the drilling rig, but that's only would be...there are a number of ways to handle a diverter system and our only concern is that we consider it as a diverter and not to shut the well in. MR. HAMILTON: Thank you. I'm kind of jumping around here if you will excuse me, Mr. Rochon, back to your field thaw subsidence testing, I guess I missed what rate was that equivalent to, your producing rate that was the equivalent to that you were going to test on? MR.ROCHON: We looked at several producing rates for different structural wells and came up with the well that would produce the maximum amount of heat through the permafrost and it was what would be a mid structure oil well and rates essentially would be in the range of 1500 barrels a day. MR. HAMILTON: I see, and what type of material did you have or did you assume in your tests or run in your tests, was it some type of gel pac or fluid pac in there? MR. ROCHON: For the test or the calculations? MR. HAMILTON: For both I should say. MR. ROCHON: In a field test we had 5 1/2" casing cemented to the permafrost and then we circulated the glycol water down the tubing and up the annulus so we had no pac at all in the test wells. In our calculations for the model we assumed a gelled oil base but in the gel pac, in the 13 3/8" by 9 5/8" annulus. MR. KNOWLES: Mr. Hamilton, I'd like to speak on that a little bit. The test was designed when we combined the fact that we had cemented the casing -47- directly to the permafrost, that combined with the chosen temperature and flow rate are what simulated the 20 year life in the structure well or in that instance because it was a simulation and we wanted to put the same amount of heat out into the formation and did not use the arctic factor of gelled oil in the thaw test. MR. HAMILTON: Mr. Clarke, in your work that you've done on the permafrost, I think in Figure 4-1 you have an exhibit showing the base of the permafrost, what methods did you use to select the base of the permafrost? MR. CLARKE: Well, there are several methods, the best one is probably a temperature log run in the hole a considerable period after the well has been completed and we have quite a few wells on which this is being done and these provide a fairly solid basis for mapping the distribution of the permafrost. There are other methods, as you probably know, the ice in the permafrost interval produces much higher resitivities than thawed water and this can be seen on electric logs. It also produces a much faster transit time on the sonic log and if you look at these two logs together you can generally get a fairly close approximation to the lowest appearance of ice in the permafrost interval which approximates the base of the permafrost. MR. MARSHALL: Mr. Clarke, just one question on that particular item, how much time do you allow to lapse before you make your temperature tests? MR. CLARKE: I can't answer that directly, I haven't done any of the work, but I refer you to Mr. Howitt's paper in the World Petroleum. MR. ROCHON: We have done some work on temperature surveys and we've run them on periods say if we were drilling a Saddlerochit well, we'd run 6 months a year, two years, til they stabilize and for observation holes we run those at various periods too. We like to wait normally about a year before we've run temperature surveys and I think BP has done the same thing, has run several to see when they have stabilized, I'm sure they have stabilized data. MR. MARSHALL: Thank you very much. -48- MR. HAMILTON' Mr. Judd, in the wells where you are planning or already have run thermo case tubular goods, do you have any pressure rating tests made on that type tubular good that you could make available to us? MR. JUDD- We have pressure tested those tubulars to 5000 psi. MR. HAMILTON- 5000 psi. MR. MILLER' Is there any reason, Mr. Judd, why the ratings of the casing strings either inside or outside using your thermal case would be less than the manufacturer's rating of this pipe? Could we assume when we look at these designs that the manufacturer's rating say if you are using N-80, 72#/ft., 13 3/8" as an example would be any different when it's related to the thermal casing on the manufacturer's ratings? MR. JUDD' As regards to the integrity of the well or regards the integrity of the string, no, because, well more specifically we have a 10 3/4" OD thermal case with a 7" inner shell. While it. is true that the end seal or the diaphram which connects the 7" to the 10 3/4 has a pressure rating of 5000 psi, I believe that is less than either one of those two pipe bodies, however as regards the pressure integrity of the string which is maintained by the connection in the 10 3/4 inch pipe, the pressure rating would be a published rating of that connection. MR. MILLER' Thank you. [1358] MR. HAMILTON- Mr. Smith brought up a point I'm going to bring out here, and correct me if I'm wrong, but haven't all the casing failures resulting today been a result of freeze back problems in Prudhoe Bay? MR. KNOWLES' Yes, HOyle, I might speak to that, all of the casing damage -~-' .... that has brought on the various workovers in the field that we have seen to date have been the result of internal freezeback, where there was a freeze- able fluid left between two casing strings, that is all the damage that Atlantic Richfield has observed. MR. HAMILTON' Another thing I'd like to point out here now that we've essentially adopted the Statewide rule requiring nonfreezind fluids or -49- fluid with a freezing point lower than the minimum permafrost temperature be left between two casing strings, this puts the burden on the operator to make sure in placing that fluid in there that he gets it in good placement in there, that regulation doesn't necessarily cure anything unless the operator is prudent in putting it properly in place, I just wanted to bring that out. MR. JUDD: Mr. Hamilton, this is Dick Judd again. I might say also that on all the workover operations that we have performed which I have referred to earlier, were indeed for the reason of removing collapsed casing caused by this freezing of fluids between casing joints and all our existing wells do not have any freezing fluids in them. MR. KNOWLES: Hoyle, I might add something there. I noticed that you voiced the concern that just because we have a regulation doesn't mean it's going to get done. In some of my prior comments concerning some of our studies we are going to send a copy of which to you we will include in that a study we have done on the displacement analysis of the removal of drilling muds, whatever, from case well bores and this was the subject of a considerable research project and we will make those results available to you, an SPEA paper. MR. HAMILTON: Fine. We appreciate that. MR. MARSHALL: I'd like to state at this time that this is of course a public hearing and any Person in the audience is free to give testimony or voice questions. Is there any one in the audience who is so compelled? Yes, sir. MR. KRUGER: Exxon would like to give a brief supportive statement. MR. MARSHALL: Very good. Would you please stand close to the mike someplace where it is convenient, and identify yourselves please. MR. KRUGER: I'm Tom Kruger, an attorney with Exxon in Los Angeles and here to make a brief statement for our company is Mr. Clifton C. Stewart, the Western -50- Division Production Engineer for Exxon Company, which is a division of Exxon Corporation and his office is also in Los Anaeles. MR. MARSHALL: Thank you. MR. STEWART: Mr. Chairman and the Committee, Exxon has been closely involved in the permafrost studies which have been described earlier this morning by Mr. Rochon and Mr. Knowles of Atlantic Richfield and we have also developed an independent permafrost strain model through our affiliate, Exxon Production Research Company. We've arrived at and support the same conclusions testified to by Atlantic Richfield and the full scale field test at Prudhoe Bay and analytical strain models, the casing strain work and other studies give evidence of the thorough basis for these conclusions which were presented. Further, we believe that the strain predictions are conservative so that we can be con- fident of having bounded the actual permafrost performance that we can expect to encounter. As a result, it's likely that there's a further margin of safety beyond the specific numbers that were presented here this morning. Regarding the Conservation Orders Nos. 98-A, 98-B and 83-C for the Prudhoe Bay Field, we support the suggested field rules proposed by Mr. Rochon and repeated by Mr. Judd. Thank you. MR. MARSHALL: Thank you. Is there anyone else who would care to make a statement? Before we close the hearing I'd like to say that tentatively we think it's proper to keep this record open until the close of business on December 10. Now that's fifteen days off. Would this cramp anybody who has intentions of submitting any further information or data? All right, then the hearing record will be kept open until the close of business which is 4:30 for the State on December 10, 1975, and with this we will close the Oil and Gas Conserva- tion hearing on Conservation Order No. 137. -51 - For the matter of the record, the sworn witnesses at this hearing were: R. B. Vickery, BP Petroleum Jerry Rochon, Atlantic Richfield C. R. Knowles, Atlantic Richfield P.R. Judd, BP Alaska R. H. Clarke, BP Alaska The sixth sworn witness did not present testimony. [1420] PERMAFROST GEoLoGY R. H. Clarke November 25, 1975 1. Stratigraphy and Age of Permafrost Sediments The permafrost zone ~%t Prudhoe Bay includes an upper "First Gravels" interval, 2'~0 to 500 feet thick, wl~icl'~ overlies a thick sequence of sa~ds and sJ.[ts, the SagavanJrktok Formation (Howitt, ref. 8; Barker, ref. 6; Clarke, ref. 7). As illustrated by the gamma 'ray ].og correlatio~ d.i. agrams, Figures IV-1 and IV-2, prepared by J. E. Eason with A.R.Co., the Sagavanirktok Formation contains several correlateable silt horizons which dip gently to the northeast. The First Gravels contain no correlateable sil't intervals and form a homogenous, northeasqerly thict~ening clastic wedge. Both intervals of the. permafrost are continuous across the Prudhoe Bay field area. The 'base of the permafrost was determined from temperature measurements and by log evaluation, and lies in the Sagavanirktok Formation, as shown on the cross sections. Palynological examination'of the BP permafrost cores by Arco9, and BP8, have shown that the First Gravels were deposited in an Arctic climate sin, ilar to or sligl]tly warmer than that of the present day. Thc underiLying Formation contains a rich pollen flora which c~.~.n.be correlated with early to middle Miocene floras found elsewhere in Alaska. The climate at IV-1 that time was cool and humid. The stratigraphic sequence at Prudhoe Bay has beel~ relat,~'d I;o the late Ce~ozoic ]~is'tory of the Brooks Range to l:he South (Porter, ref. 10). The Sagavanirktok Format].on re~re.~ents the prod,~cts of the erosion of the Brooks Range during the Tertiary, which virtually ceased with the onset of Plio-Ple;~stocene climatic cooling. Tlne Fi.rst Gravels record a rel~tively recent phase of glacial erosion of the Brooks Range, particularly d~.[ri~lg the Wisconsin and Recent. Superfic~ial depo~i;i'ts at Prudho~ Bay closely resemble those in the First Gravels beneath. Because no correlateable marine inter~'~alations have been found in the First Gravels it is ~')robable that 'this interval is nearly all wisconsin to Recent in age~ 2. Permafrost Consolidation and Ice Content The Sagavanirktok Formation, including the }?ermafrost interval below the First Gravels, was ~!leposited in a humid, cool., l'ate- .Tertiary enVironment long bef~'ore the development of permafrost on 'tile North Slope. The normal consolidation profile which tlne Formation acquitted during deposition was not significantly affected by the refrigeration and permafros't~ development which occurred .in the uppe.lf part during the PleisUocene. In fact 'there is some evidence that. cyclic permafrost growth and 'degradation during the Pleistocene may have' contributed to IV- 2 the stif.fness of the soils .~n the deeper la[~'ers of the existing permafrost i. nterv~L1. The First Gravels are the l'~roducts of s trea,u deposition in an environment resembling tha~-, of the preseut day. Tt~e pollen evidence suggests t]~at most of the dcpo.~ition took place in slightly wa]?mer conditions, probably ~vhen the various valley glacier system, s recognJ, zed by porter10 were decaying and releasing large volumes of water and glacial detritus. Although ice structu~'-es have ¢]eveloped both in the near-surface interval (0-50 ft, a[-~pr.oximately) at the present 'time, and ~robably did so on o..I.der surfaces witt]in the First Gravels, 'the burial of these older surfaces is ac'com[~anied'by the destruction of the excess ice structures. This may happen in at least two ways:- (1) The larger thaw lakes are associated wi. th substantial local thaw of the permafrost beneath, especially of the decaying frost 'polygon structures which contribute · to their development. (2) The streams which bri~,g in the Sands and gravels to bury existing land surfaces also represent a major source of heat J.nput for the near-surface permafrost. A thaw bulb develops below and adjacen~ to each active , stream of sufficient dimensions to 'thaw the previous IV-3 generation of near-surface ice structures. Evidence of this process is av~ilnble from the modern Sagavanirktok River (Sherman, ref. !1). Because the First Gravels may have remained fr~)zen after burial to depths greater titan tl~ose influenced by the thaw beneath streams and ],akes, the deeper gravels may be somewhat underconsolidated. To summarize: the FJ.rst Gravels contain free J. ce structures in the near-surface interw.~l (about 0-50 ft), but the deeper part of the First Gravels are free of excess ice though probably somewhat underconsolJdated. The deeper permafrost is more or less normally consolidated and free of ex. cess ice. The sequence of even.hs, ant[ tJ.t.e stra'tigraph%z of the permafrost are outlined on Table IV-1. IV-4 TABLE IV-1 PERMAFI~.OS'[' GEOLOGY AT PRUDItOE BAY. - FORMATION AGE FLORA/CLIMATE First Gravels Late-Pleistocene Mosses and ferns to Recent with occasional kree po].len, probably rewor].~ed. .Arctic, tundra c lima'te. Hiatus Plio- Pleistocene ~]o record, .probably frigic~ with several episodes of marine sub- mergence. Sagavanirktok' Tertiary Formation to Mid- Miocene Cool temperate fo r e s t association. PERMAFROST Similar to present day Development of permafrost in Sagavanirktok Fm. perma frost (Additional) REFE:-~NCES 8. Howitt F. "PermafrosI: Geology at Prudhoe Bay", World Petroleum, September (1971'), 28-38. 9. Peterson E.T., Hedlund R.W., Tabbert R.L., ~:~nd Bennett U.E., "Palynological exami,.tation of BP 12-10-14 porm~tfrost cores and BP 27-11-14 cutt~.ngs," (1970) Arco Repo,.~t. 10. Porter S.C. "Late Pl~istoc~.'~ne Glacial Chronology o'f , ' " Amer g'our Science North-Central Brooks Range A.].aska , . . . (1964) 262, 446-460. 11. Sherman R.G. "A groundwater:: supply for an o.~.], camp near Prudhoe Bay, Arc'tic Alaska," Proc. 2nd. Int. Conf. Permafrost (1973) 469-472. vni-xELATiC~'~S FOP,rc, u'~-"'DHO''= ~-AY AR~.=A i i-- 2oo --600' I000' i800' ': ' (,~; '1' I.'-:. 1', TI P, R'," GO I? t'? E I_ AT I 0 lq S FOR {iO O' '-- l 0,( !C,~' · .... ILl I!.l I l>.1 I~I t B L ~ ()~ I ,., c, 0 0 ...... H 0 I? I' C) II '1"A I, '- I'~= I G ~000 ' V [..t f,' TI t: I'~ L I" 4 - : 00' ,---'7: I . \ / ~' :~'°~ss-, ' · ~'~ ~ ' :...~ ,- ~, -. rE, ~:£R: I ~ -. , i : , ' ~ %,, ~ · · - -i-~:,..v ;:,-=,, ! ........ -,::. '...,-.' "--'" ' " -" :---: + ~ :5--- ~ '.._u : ., :, . \ / ~- 't f ,,~--------=__~ ~ :_':..--. -.: . . ._~....-.~ ! >:,,'---~ ,~': '-- ---- ~ - '"~-- - ~ ...."-"---- 'u--~- , -' ' i ~ SC. CA. .:s ~ ;. ' , . ", ": t ": i i _,, __ : i ."'.:-.-L,', _.,/: . _. : ........ ' ,'.-~- · ' '~ ° ' ' ~ ' · ! " "' "~ "-'"' ° . ' i ~ I , · , , ! ~.. ~ ............. i :.-'~ ...... "N~ I.. ' /-,' I ....... I -~'- .. · .-4> t / '' " , . .___tx_k'-~ ',~---~,: . . _ .' []2-0~1 B STATE of ALASKA DEPARTMENT OF NATURAL RESOURCES OFFICE OF THE COMMISSIONER TO: Thomas R. Marshall Chief Petroleum Geologist FROM: William C. Fackler Deputy Commis sione~ ~-~'~ DATE : SUBJECT: November 14, 1975 Oil and Gas Conservat Committee Designatio! I DIR : --'j'-: -d76~-~: ~-$~ ~ ~ .......... ', 1 EN..?.- --'l- ~ ENG --I' .: ~-'"'G" ... ~ '-i ,-.-, ~-',.~'~" : i'0n ?, i~ v D ..-,r T "' t-- -)~'~ - ............ ! SEC The injury to Mr. O. K. Gilbreth, Director, makes it necessary to designate an alternate for him to serve on the Oil and Gas Conservation Committee at.the hearing scheduled for November 25, 0 1975, in Anchorage. Accordingly, Mr. Harry Kugler is designated as the alternate to act in the place of Mr. Gilbreth as provided for in 11 AAC 22.510 and AS 31.05. 030. CONFER: AFFIDAVIT OF PUBLIC;A'I"ION STATE OF ALASKA, ) THIRD JUDICIAL DISTRICT, ) ss. being first duly sworn on oath deposes and says that..~:L~. ....... is the_.~:-e-~!L~---C-':~-~-~-~ ...... of the Anchorage News, a daily news- paper. That said newspaper has been approved as a legal news- paper by the Third Judicial Court, Anchorage, Alaska, and it is now and has been published in the English language continually as a daily newspaper in Anchorage, Alaska, and it is now and during all of said time was printed in an office maintained at the aforesaid place of publication of said news- paper. That the annexed is a true copy of a. as it was published in regular issues (and not in supplemental form) of said newspaper for. a period of ..... >.,~._ _,;. ...... insertions, commencing on the ..... ~?day of _..~¢.~.:_~gTo___e.~ ....... ,19 .'~_, and ending on the ........ :~"'.=! .... day of of ._.g~.9.'.b..e.,.T_ .......... , 19___~_, both dates inclusive, and that such newspaper was regularly distributed to its subscribers dur- ing all of said period. That the full amount of the fee charged for the foregoing publication is , c'.,',O which the sum of $~.~ amount has been paid in full at the rate of 30 f per line; Mini- mum charge $]0.00. Subscrib>~d and ~rZc~bj~,rn o before . :. ?t.~-~ , . O~'.k, ober me this __~.~:Taay or .... :Y__:'.:_;._':_:_., ' ' .o~c~°~'U~'LiE ' ' STATE OF ALASKA DEPARTMENT OF NATURAL RESOURCES DIVISION OF OIL ~ AND AlaskaOii andGas · Conservation Committee Consem, ation Order File No. The Alaska Oil and Gas Conserva ti°n Committee will hold a hearing on its owa motion to Title 11 Alaska Administrative Code Section'22.4~0 in the MuniciPal Chambers of the Z. J. 'Lonssac LlbralT, Fifth Avenue and Street, Anchorage, Alaska at 9:00 AMi i'on November 25, 1975. The Committee[ Will seek testin~ony On the following'~ I matters . .. 1) To consider an amendmen~ to jRifle 4, Conservation Orde~ 95-A, and8~.C of the Pmdhoe Bay Field pOolj rules to require a bag type blOWout/ preventer while drilling the hole Which the.sUrface fnsing is Set. Thet preSent regulations pertain to require-I ments for blowout preVention equiP-~ merit only while drilling below thei: surface hol~. ' ' . ~) To c~sider an 'amendment t~ Rule ;3, Conservation Order 98.A, ,and'83.C. Sections a, b, and d in par- ticularl~ will be considered.bUt th~ etttire rule will be open for possiblej changes. EXperience gained since 1911~ al~ears to warrant changes in casi~l requirements in the perm&frost por-J tiOn of the hole. Thomas R, Marshall, Jr. Executive Secretaryl Alaska Oil ~nd Ga~ C°nsefvation Commit tee ~1 Porcupine DHve Anchorage, Alaska 99501, PUBLISH: October 25, t9/5~.._~ Notary Public in and for the State of Alaska, Third Division, Anchorage, Alaska MY COMJAISSION EXPIRES NOTICE OF PUBLIC HEARING STATE OF ALASKA DEPARTMENT OF NATURAL RESOURCES DIVISION OF OIL AND GAS Alaska Oil and Gas Conservation Committee Conservation Order File No. 137 The Alaska Oil and Gas Conservation Committee will hold a hearing on its own motion pursuant to Title 11, Alaska Administrative Code Section 22.540 in the Municipal Chambers of the Z. J. Loussac Library, Fifth Avenue and F Street, Anchorage, Alaska at 9:00 AM on November 25, 1975. The Committee will seek testimony on the following matters: 1) To consider an amendment to Rule 4, Conservation Order 98-A, 98-B, and 83-C of the' Prudhoe Bay Field pool rules to require a bag type blowout preventer while drilling the hole in which the surface casing is set. The present regulations pertain to re- quirements for blowout prevention equipment only while drilling below the surface hole. 2) To consider an amendment to Rule 3, Conservation Order 98-A, 98-B, and 83-C. Sections a, b, and d in particularly will be considered but the entire rule will be open for possible changes. Experience gained since 197Q appears to warrant changes in casin~ requirements in the permafrost portion of the hole. Thomas R. ~!arshall, Jr. Executive Secretary Alaska Oil and Gas Conservation Committee 3001 Porcupine Drive Anchorage, Alaska 99501 Publish' October 25, 1975 BP ALASKA INC. 3111 - C ~ STREET · TELEPHONE (907) 279-0~44 MAILING ADDRESS: P.O. BOX 21. o 1379 ANCHORAGE. ALASKA 99509 June 30, 1975 Mr. O. K. Gilbreth, Director State. of Alaska Department of Natural Resources Division of' Oil & Gas 3001 Porcupine Drive Anchorage, Alaska '99502 Dear Mr. Gilbreth: p~~ i'~ Well Design for Thaw Subsidence Protection Atlantic Richfield Company has sukmitted a report to you entitled "Prudhoe Bay Permafrost Casing and Well Design for Thaw Subsidence Protection", which summarizes the recent investigations into the problem of ~rost thaw subsidence. BP Alaska has been involved in the planning and execution of these investigations and supports the conclusions. Very truly yours, BP AIJL~KA INC. F. J. Venn Prudhoe Bay Fields Manager PRUDHOE BAY FIELD PEI~ROST CASING AND 17ELL DESIGN FOR THAW SUBSIDENCE PROTECTION T. K. Perkins J. A. Rochon R. A. Ruedrich F. J. Schuh G. R. Wooley Atlantic Richfield Company North American Producing Division May 1975 INDEX I. Introduction II. Conclusions III. Discussion Appendix I. Prudhoe Bay Five Spot Thaw Subsidence Test Appendix II, Permafrost Thaw Strain Predictions Appendix III. Allowable Casing Strain Limits Appendix IV. Permafrost Geology Appendix V. Permafrost.~haw Predictions I. INTRODUCTION Shortly after the discovery of the Prudhoe Bay Fie].d the industry became concerned with the possibility of casing and well damage as a result of permafrost thaw subsidence. At the 1969 Field Rules Hearing, industry requested, and later received, permission from the state to use slip joints, refrigeration, and/or insulation to protect the producing wells at Prudhoe Bay.' Since that time, the industry has performed numerous laboratory, · field., and engineering studies in an effort to more fully ew~luate thaw subsidence problems. In 1969, BP Alaska cored the permafrost at Prudhoe and had a variety of laboratory studies performed on the cores. In ].970, BP Alaska started two single well subsidence tests. These wells were thawed to radii of 10' to 15' by circulating hot oil for one year. These tests simulated thaw that a conventional oil well would, produce after approximately' two to five years of production. In 1973, Atlantic Richfield Company-Exxon Company started thawing th.e perma- frost in a closely spaced 5-spot well pattern. Thaw in this test pattern was desiN~ed to simulate that which might be expected after about 15 to 20 years of production. In addition to these measurements of permafrost behavior, there have been several tests and studies to determine the strain capacity of casing that will be loaded by permafrost subsidence. In January, 1975, ~xxon ran a series of axial tensile and compressive strain limit tests on samples of 13-3/8" OD 72#/ft CYN-80 Buttress casing. In addition to the e×~perimental measurements, there are several major on-going research and engineering studies. Both A.R. Co. and F~xxon have developed mathematical models to predict permafrost thaw subsidence behavior. During the early part of the test, these models were compare8 with the test results and concepts of model development were modified. These models were then used to extend the Field Test and e×~plore property variations. Associated studies of the permafrost make-up and the~.Tnal history have been prepared by BP Alaska. and A.R. Co. for use as input to the permafrost sub- sidence models. In addition A.R. Co. and a consultant, Prototype Development Associates (Santa Ana, California), have developed a finite element model of the ].3-3/8" N-80 Buttress threaded casing connection that provides a general method of predicting the minim~n ultimate tensile and compressive strain capacities of this casing connection. This report is a s~mnary of the most importm~t conclusions from these efforts. The report is divided into two main parts. This first part inc].udes the Introduction, the Conclusions, and the Discussion. The second part is an Appendix with sections covering the PrudJ~oe Bay 5-spot Thaw Subsidence Test, Permafrost Thaw Strain Predictions, Allowable Casing Strain Limits, Permafrost Geology, and Permafrost Thaw Predictions. 1-2 II. CONCLUSIONS The studies of the thaw subsidence at Pru~oe Bay and the associated work allow the following conclusions: Conclusions (1) Thawing of permafrost causes permafrost compaction and strain in the casing. (2) Casing is available which can withstand strains in oil well completions through the permafrost interval. (3) This permafrost compaction-casing strain relationship requires a new casing performance criteria that utilizes tensile and compresSive strain capacities of the casing and couplings in contact with the permafrost. (4) The strain producing mechanism is complex and is directly related to the amount of thaw in the permafrost. Strains are due 'predominantly to an interaction of alternating.layers of sand and silts and. are dependent upon'the relative thicknesses and mechanical properties of these layers. (5) We can predict the maximum expected tensile or compressive strains as a function of depth'and radius of thaw. (6) The maximum strains measured in any' casing joint in the 5-spot thaw test after 16 months, which represents a producing period of approximately 15 years from a mid-structure Prudhoe well as currently being completed, are 0.13% in co~ression and 0.08% in. tension. (7) The worst case strains predicted for our proposed well design are 0.7% con~ression and 0.5% tensile. These worst case strains are predictedby our mathematical model which was developed in conjunction with the 5-spot thaw test. II-1 (8) Full scale tensile and compressive tests m~d the finite element studies of AP1 N-80 Buttress threaded casing show that a 13-3/8" ?2#/ft N-80 AP1 Buttress threaded connection can resist minimum ultimate compressive and tensile axial strain in the pipe body of 2.3% compressive and 3.4% tensile. (9) Using these worst case strains due to the pe~nafrost thaw and the minimum ultimate strain capacities in the pipe body, our current well design and completion progrmn give design factors of 3.3 in compression and 6.8 in tension. Our current well design shown in Figure II-1 utilizes gelled oil in the 9-5/8" by 13-3/8" annulus and includes 13-3/8" 72#/ft N-80 AP1 Buttress casing. Figure 1I-2 shows the maximum expected casing strains and minimum ultimate casing strain capacities for our current well completions. 1I-2 CURRENT PERMAFROST PERMAFROST THAW CASING TEST FIGURE II- & WELL DESIGN "2.0" CONOUd_.TOIZ 5ET e 8,0' CEMENT ,, F'E~~AFE05 ID C,.~A ?_.7'OO' ~ ELLE-D ~l L I~ALL VALVE C.,A$1 N5 4OO 800 - LU ", 1200-- LU 1600 - 2000 - -3.5 MAXIMUM COMPRESSIVE CASING STRAIN ENVELOPE m SILT TO SAND COMPACTIBILITY mo, OF 3.4 I I I I i I .~----- BASE OF PERMAFROST I DESIGN FACTOR I -2.0 -1.5 -1.0 ~ -0.5 MAXIMUM TENSILE CASING STRAIN ENVELOPE FOR WORST CASE ANALYSIS MAXIMUM TENSILE CASING STRAIN FOR FIXED RATIO OF SILT TO SAND COMPACTIBILITY OF 3.4 DESIGN FACTOR 0.5 I 2.0 2.5 3.0 T~I ell c FIGURE 11-2' __ I . III. DISCUSSION Three methods were approved for completing wells through the permafrost at the 1969 Field Rules Hearing. Although all were mechanically feasible, they all imposed added costs to a well over and above a conventional completion. Several also had the added disadvantage of complicating well completions and workovers. For this reason, studies as mentioned earlier were Undertaken to better determine and define the characteristics of the permafrost and the impact of its thawing on oil wells drilled through the permafrost. The logical approach which was followed initially by BP Alaska was selected again by Atlantic Richfield and a thaw subsidence test was undertaken. Mathematical models were developed so that the results of this test could be extended to the complete field. ~ Both A.R.Co.'s Research and Development Group and ~Exxon Production Research Company undertook the task of developing these mathematical models. Meanwhile, a thaw subsidence test was initiated on the North Slope. ~e results from this test were continually fed into the research groups tO provide the concepts required for mathematical development. Additionally, numerous laboratory studies were ~mdertaken to more completely understand the mechanical behavior of the permafrost., A second aspect of the study involved destructive tests on short sections of 13-3/8" 72#/ft N-80 Buttress casing and the development of a mathematical model to determine loading limits for casing com~ections. III-1 Brief summaries of each of these efforts are presented below and a more detailed writeup is provided in the Appendix. Field Test A 5-spot thaw subsidence test was started in August 1973. This test was designed to achieve a thaw of approximately 35' radius, which is equivalent to the thaw that would occur around a con- ventionally completed mid-structure oil well during approximately 15 years of production. The test was designed to obtain the maximum amount of data and provide sufficient time for well completions prior to field startup. The test was located in Section 11, T10N, R14E, in the center of the Prudhoe Bay Field which is near the site of BP's well which was extensively cored in the permafrost. The test is comprised of five, 2200' cased wells spaced 23' apart in a 5-spot pattern. In addition to the five thaw wells, monitor holes were completed at 15', 30', 100', 150', and 250' depths to measure the near surface ground subsidence. Heat was introduced to the thaw pattern by circulating a 190°F glycol water mixture in the five thaw wells. ~e most important data derived from the thaw test was casing strains obtained by the use of a multiple casing collar locator tool. This tool has the precision to measure the lengthening or shortening of the 40' casing joints to ±.01 inches. (This correspondes to *0.002% strain accuracy). ~e log measurements indicated that after 16 months of thaw (approximately 15 years of production from a mid-structure we~l) the maximum strains 111-2 measured in the thaw wells were 0.13% in compression and 0.07% in tension. Other significant data was obtained from cement bond logs and caliper logs run periodically in the five thaw wells. The surveys indicate that there has been no measurable radial deformation of the 5-1/2" casing in any of the five wells. Ninety-one radioactive bullets were fired into the formation on 21' spacing in the center thaw well prior to running the 5-1/2" casing. A conventional gamma-ray collar locator tool has been used to record the relative position of the bullets to the casing collars to determine the relative movement of the soil and the casing. Although there has been some scatter of the data due to the inaccuracy of the measurements, we conclude that there has been little, if any, relative movement between the pipe and the formation. Mathematical Strain Model Casing strains resulting from a large thaw radius have been measured with good precision in the 5-spot thaw test. In order to extend casing strain predictions across the field, mathematical models were' developed by both A.R. Co. and Exxon Production Research. The field data were used when developing modeling concepts. The numerical model was then used to calculate strains for different lithologic distributions. Computing the expected strains in the permafrost is quite complex. However, from a simplified point of view there are three important casing loading mechanisms that influence the expected subsidence strains: 111-3 1. An im~ard lateral movement of surrounding frozen pe~nafrost. 2. An upward elastic flexing of soil below the thawed region. 3. A relatively small downward movement of thawed material caused by the mass of the thawed material, the gravitational field, and a drop in pore pressure. The magnitude of the tensile and compressive strains is dependent on layer thickness, the soil mechanical properties (especially the ratio of the compressibilities of the adjacent layers), the thawed radius and the depth of the interval. All of these factors have been taken into account in the development of the model and the moSel correlates well with the data from the 5-spot thaw test. The determination of ma×'i.mum strain across the Field was handled by a worst case approach using the expected, thaw radius versus depth and conservative values of pore pressure loading ~nd mechanical properties. The model was used to compute the highest compressive and tensile strains that would result from any possible combination or relative th'ickness of sand '-and silt layers. Since these studies include all of the expected ].ithologic combinations in Prudhoe, the results are applicable across the Prudhoe Bay Field. The maximum calculated t]~a.w subsidence casing strains for the current well design after 20 years of thaw was calculated by the model to be 0.7% compressive and 0.5% tensile. 111-4 Casing .Stra~ Capacities In addition to predicting the maximin casing strain that will be imposed by thawing the pemnafrost, we must dete~nine the strain which our proposed casing can withstand. We thus undertook laboratory and mathematical studies to determine this strain limit. Strain measurements taken during full scale axial load tests clearly demonstrate that the Buttress connection on 13-3/8'' 72#/ft N-80 casing can transmit loads which exceed the yield strength of the casing. A finite element model of the Buttress connection and the full scale strain limit tests were used to determine the stress-strain distribution in the connection during post yield loading. These analytical and e×~perimental results were used to develop failure criteria for both axial compressive and tensile loads. Casing strain capacities were calculated based on these criteria and the effects of make-up, and pressure and temperature loads. The minimum ultimate casing strain 'c~pacities of the 13-3/8" 72#/ft N-80 Buttress casing are 2.3% compressive and 3.4% tensile. These capacities provide more thm~ ample design factors of 3..3 in compression and 6.8 in tension when compared with the predicted maximum thaw strains of 0.7%~ in compression and 0.5% in tension. Field Lithology As discussed earlier, one of the significant parameters in predicting strain is the lithologY of the permafrost interval. 111-5 Thus, geologic studies on the permafrost interval have been an integral part of our work since the discovery of the Pru~oe Bay Field. These studies have indicated that lateral continuity exists in the permafrost strata in the Prudhoe Bay Field area. The base of the permafrost in this area varies in depth from approximately 1600' to 2000'. Geological studies by A.R. Co. and BP Alaska have contributed significantly to the understanding of the thaw test m~d extending its results across the Prudhoe Bay Field. ~ese studies conclude that the First Gravels may be undercon~acted, but appear to be void of excess ice, ice lenses, or wedges below approximately 50' of depth. Below the FirstGravels, the soils are more co~acted and beco~e stiffer with depth. Thermal Model A necessary tool. in our studies is the thenna.1 model used to predict permafrost thaw resulting from production. Thermal models were developed early by the Petroleum Indnstry to predict permafrost thaw. Several of these thermal models have been described in teC]mical papers previously published (Reference 2 through 5). See Figures V-1 through V-3 for calculated thaw radii. ]111-6 . APt)'ENDIX I - PRUDIIOE BAY FIVE SPGF TIIAIq SUBSII)F~CI~ 212ST 1. Test Program The 5-spot thaw subsidence test holes were drilled and the heating and monitoring equipment was installed during June and July of 1973. A 22 month thaw period will be completed in June 1975. The test utilized five 2200' cased holes in a closely spaced S-spot pattern to produce a thaw of approximately 35' radius which represents approximately 15 years of production from a Prudhoe Bay Field mid-structure oil. well. The test was desigmed to obtain the maximLm~ data and provide sufficient time for well completions prior to field startup. The test was located near the center of the field on BP's Central Pad 3 in Secti.on 11, T10N, R14E. Since this was also the site of BP's extensive permafrost coring program, the core data could, also ? be used in the subsidence test evaluation. Figure I-1 'i.s a map of the test site sho~ing the arrm~.gement of the five thaw wells and the placement of the temperature and subsidence monitoring holes. Figures I-2 through 1-5 are sketches of thaw wells, the subsidence monitoring rod wells, th.e subsidence monitoring wire wells, and the temperature monitoring wells. 2. Test Results The S-spot thaw pattern was analyzed with m~ X-Y two dimensional thaw simulation model. The progress of thaw at. the 1000' level. is shown in Figures 1-6 through 1-10 at five different stages of thaw. The equivalent single well thaw radius for the 5-spot pattern versus depth is shown in Figure I-il. Until the later stages of the test where the five patterns have merged into one, the equivalent single well tlmw radius should be used with caution. I-1 The most important measurements in the subsidence test were the casing joint length measurements taken with the Schlumberger Multiple Casing Collar Locator tool. This tool provides casing length measurements of sufficient precision that we can compute length changes with a standard deviation of .01" in a 40' casing joint. (Reference 1) (Equivalent to a strain of ±0.002%) The effect of thaw on casing strain is shown by the series of plots of the MCCL measurements versus thaw in Figure 1-12. Figure 1-13 shows the casing strain measurements after 16 months of thaw with the gamma-ray log and an interpretation of the lithology in this well. Note that the peak measured compressive strain is only .13% and that it occurs in a joint that is adjacent to a massive silt. Pore pressures in the thawed region were measured with pressure transducers attached to the outside of the casing. Figure 1-14 is a plot of the estimated pore pressure profile after 16 months of thaw, together with the latest pressure measurements. Surface and near-surface subsidence movements were calculated from periodic measurements of the elevations of the subsidence monitoring rods and wires. The total movement after 16 months thaw is summarized in Figure 1-15. Note that the near-surface n~vements were only a few · tenths of a foot and that they diminish with depth. 1-2 Subsidence of the wellheads with thaw time is shown in Figure 1-16. The initial increase in elevation is the result of a thermal expansion of the well coupled with a release of the near-surface cement to formation bond after thaw was begun. A measurement of the relative movement of the soil and casing was made by placing radioactive bullets in the formation and using a gamma-ray collar locator tool to record their relative positions· versus time. Figure 1-17 is a summary of the log measured movements versus thaw time. Although there is some scatter due to the precision of the measurements, we conclude that there is little, if any, relative movement between the pipe and the formation. Figure 1-18 is a plot of the open hole caliper measurements taken on the five thaw wells. Note the consistent magnitude of the hole enlargement of the First Gravels versus the remainder of the hole. Figure 1-19 is a plot of the series of cement bond logs run on the center well. These logs show good bonding in the area of alternating strain and poorer bonding in the First Gravels. The temperature surveys run in the three 400' Temperature Observation Wells are shown on Figure 1-20. Temperature profiles are shown for the period from August 1973, prior to heating, to December 1974. The temperatures measured in these wells correlated very closely with our thermalmodels. The apparent abnormal temperatures in well "A" at 275' and in well "B" at 290' are due to the close proximity of the observation wells to thaw wells at these respective depths, due 1-3 to a drift in the well courses during drilling of the temperature observation wells. The test measurements and model studies confirm the three subsiSence mechanisms discussed in Section III. The pressure measurements, the surface and near-surface s~sidence monitors, and model studies all confirm low net vertical soil strains and give no evidence of excess ice in the First Gravels. The mechanism for generating alternating layers of axial compressive strains in the silts and tensile strain in the sands is clearly shown in Figure 1-13. The reaction at the base of the permafrost is also evident in each of the casing strain measurements. 1-4 WELL c-2 G-I SW WELL H-I C-I A-4 H-IA © CENTRAL WELL B-I E-I F-I F-2 PERMAFROST NE WELL E-2 sE WELL THAW TEST · THREE FIGURE I-1 APPROXIMATE 18 MONTH THAW ZONE AT I000' 400' TEMPERATURE HOLES F-6 LEGEND [] six 15' MONITOR RODS ~ SIX 50' MONITOR RODS ~ THREE I00' MONITOR WIRES 0 FOUR 150' MONITOR WII::~S 0 TWO 250' MONITOR WIRES ~ THREE 400' TEMERATURE HOLES SUBSIDENCE MONITOR & TEMPERATURE HOLE PATTERN PEF~'I,: ~-ROST Tht,~." 'TEST FIGURE I-2 I~ ~/s" CSA 75' P CEdE. MI l- 57z" CsA 2~oo' I NISTRUN'IENI'I' CABLE W,,/TFIEIZt','IISTOR$ Tiaa NI SDUCE i~S I ?_'/,~" WIOL F_. 2-]/8" I~DTTRES$ TUBING FLUNG @ ZOO0' P~F4AF-zROST T~AW T~ST MON.I ITQI:;R ~OD COIv!,PLFr_.TIOI, d ~' ~' OF 20D ~N" NOLE 4-Yz" BUTTRESSTL)BI~G .SET ¢. 30' I M f~ HOLES ..,3%" × 4' i~AT NOLE ............ O,D. STEEL C E t,.,'~ LF___N 1'- 5" STEEL WASFIED FIGURE I, PERMAFROST TI-lAW TEST MOIxlITO~ WII~F___ COP1PLET ON ?8" I NVAI? .. ~,P,~" HOLE // FIGURE I-' MOLE too' Ik.~ 5 I...4OLES 150' It,,l ~ MOLES 250' I M Z HOLES 7__"X 2P:2.' WEIGHTED ALIC~40~ PE~'h~,FPOS ]- T~4A¥,~/ TEC~T TE~PC~A'TLJR'E OBSE~MA, TIOI~ WELL COI'~PLETION FIGURE I. , HOLE ........ C ElVlENT Z~/8'' [~IJT'TRESS TUBING 5ET C~ q-O0' Ihd ~ NC)LE..'6 FIGURE !-6 CALCULATED THAW AREA AT 1000 FT. DEPTH FOR FIVE SPOT TEST AFTER 3.9 MONTHS OF THAW 4.*****.1 1 1 LEGEND: EACH SYMBOL REPRESENTS A 2 FT. x 2 FT. AREA *INDICATES THIS MATHEMATICAL ELEMENT HAS COMPLETELY THAWED. NUMERALS INDICATE THAT A MATHEMATICAL ELEMENT IS ONLY PARTIALLY THAWED. % THAWED : TEN TIMES NUMERAL SHOWN CALCULATED THAW AREA AT 1000 FT. DEPTH FOR FIVE SPOT TEST AFTER 7.4 MONTHS OF THAW 3'****? _ ************************ FIGURE I-7 LEGEND: EACH SYMBOL REPRESENTS A 2 FT. x 2 FT. AREA *INDICATES THIS MATHEMATICAL ELEMENT HAS COMPLETELY THAWED. NUMERALS INDICATE THAT A MATHEMATICAL ELEMENT IS ONLY PARTIALLY THAWED. % THAWED - TEN TIMES NUMERAL SHOWN CALCULATED THAW AREA AT 1000 FT. DEPTH FOR FIVE SPOT TEST AFTER 9.6 MONTHS OF THAW 1221 1732 ***************************** *********************** ************************* ***************************** 15 ~ * 2 12321 FIGURE I LEGEND: EACH SYMBOL REPRESENTS A 2 FT. x 2 FT. AREA * INDICATES THIS MATHEMATICAL ELEMENT HAS COMPLETELY THAWED ,.' NUMERALS INDICATE TI-IAT A MATHEMATICAL ELEMENT IS ONLY PARTIALLY THAWED. % THAWED - TEN TIMES NUMERAL SHOWN CALCULATED THAW AREA AT 1000 FT. DEPTH FOR FIVE SPOT TEST AFTER 12 MONTHS OF THAW 1 4 c ~ ~ '/ ~1 3 ~ c ~ ~ · 7 ? ************************* 2 7 * * * vr ~' ] ~ ~ a * * 7 2 11 FIGURE l-9 LEGEND: EACH SYMBOL REPRESENTS A 2 FT. x 2 FT. AREA *INDICATES THIS MATHEMATICAL ELEMENT HAS COMPLETELY THAWEI NUMERALS INDICATE THAT A MATHEMATICAL ELEMENT IS ONLY PARTIALLY THAWED. % THAWED = TEN TIMES NUMERAL SHOWN CALCULATED THAW AREA AT 1000 FT. DEPTH FOR FIVE SPOT TEST AFTER 18.5 MONTHS OF THAW FIGURE I-1(: . EACH SYMBOL REPRESENTS A 2 FT. x 2 FT. AREA *INDICATES THIS MATHEMATICAL ELEMENT HAS COMPLETELY THAWED NUMERALS INDICATE THAT A MATHEMATICAL ELEMENT IS ONLY PARTIALLY THAWED. % THAWED : TEN TIMES NUMERAL SHOWN EQUIVALENT SINGLE WELL THAW RADIUS FOR THE FIVE SPOT SUBSIDENCE TEST I 0.5 1.7 3.7 7 12 16~ ~ MONTHS OF THAW · FIGURE 1-11 600 1200 1800 0 10 20 30 40 EQUIVALENT THAW I 50 60 RADIUS, FEET 8O FIGURE 1-12 CASING STRAIN MEASUREMENTS - CENTER THAW WELL CASING STRAIN - % -.05 0 4-.05 '~.05 0 +.05 -.10 0 +,10 -.10 O. +,10 -.10 0 +.10 ,I > , ' j ._/._ ) B/~iSE,01:;FI R51 ( /EL~ _~, --J =mtm,m-- -~mm= m-- -- .1 m m~___mm-- mm I mm m----"mm m--mmmmmmmk.-- 400-m -- ',aoo ~i . , I 1600 ........ :>000 · 5 MONTHS 4 MONTHS 9.5 MONTHS 14 MONTHS 16 MONTHS OF THAW OF THAW OF THAW OF THAW OF THAW -- FIGURE 1-13 COMPARISON OF GAMMA RAY LOG, LITHOLOGY ~ 16 MONTH THAW CASING STRAIN FOR THE 5 SPOT THAW SUBSIDENCE TEST FOR THE CENTER THAW WELL GAMMA RAY LITHOLOGY CASING STRAIN % APl UNITS O 50 IOOr _ O! O! -- .~ 5 O '~ .25 ...... 2 ¢) f)' 2 (-] 0 ' 4 ()() ,1 (')() ()1:' AVl I ~; I()()() I ()()() i 2 () ()' I ? O 1400' 1400' -- 1600' 1600' ---1800' 1800' 200 O' 20'00' BASE OF PERMAFROST FIGURE I-1, 400 800 1200 1600 2000 PORE PRESSURE PROFILES IN THE THAWED PERMAFROST FOR 5 SPOT TEST RANGE OF MEASURED PORE PRESSURES BASE OF FIRSI GRAVELS PORE PRESSURE PROFILE FOR SIMULATION OF COMPACTION AFTER 16 MONTHS THAW PROBABLE PRESSURE GRADIENT IN ICE PHASE OF FROZEN PERMAFROST 0.45 PSI/FT. BASE OF PERMAFROST I 400 J \ 800 1200 PRESSURE , PSI FIGURE I-1: MAXIMUM SUBSIDENCE vs DISTANCE FROM CENTER THAW WELL AFTER 18 MONTHS OF THAW .4 .6 -.8 .2 .4 .2 .2 .2 DISTANCE FROM CENTER THAW WELL FIGURE 1-16 WELLHEAD ELEVATIONS vs TIME ., ,,,j J + ! -- -- --- }._ - , -4- © 1 '""~ .,,,..~ w , i -- ° 1 A S 0 N D J F M A M J d A S 0 N D .3 .3 .3 .I 0 -.3 o i-n m z rtl m .-~ 57:3' I W _J 0 RELATIVE MOVEMENT OF SELECTED RADIOACTIVE BULLETS TO CASING COLLARS vs TIME BELOW BASE OF PERMAFROST 1:349' 1768' 1957' +.5 -.5 +.0 ~- 0 0 m z 'i-.5 ! 0 m m +.5 0 -.5 +.5 ..5 FIGURE 1-17 LEGEND ACTUAL READING ., ,, ACCURACY OF TOOL FIGURE 1-18 PERMAFROST THAW TEST - OPEN HOLE CALIPER SURVEYS I0" 40" I0" 40" I0" 40" I0" 40" I0" 40" -- 12-1/4" BIT SIZE 12-1/4" BIT SIZE 12-1/4" BIT SIZE IL~I/4" BIT SIZE 12-1/4" BIT SIZE L -- -- 200 --==~=.. ~BASE F 'FIRS'F GRAV~ LS 400 -' ~-~ ~-- ~ '-~ ~ ~ ~ ~ ----~'~ " ~-- ~~-= ....... - . I000 - _o,..... 2000 ........ NORTHWEST WELL NORTHEAST WELL SOUTHWEST WELL SOUTHEAST WELL CENTER WELL CEMENT BOND LOGS - CENTER THAW WELL FIGURE 1-19 0 MV 50MV 0 M9 ' 50MV OMV 50MV OMV 50 MV ~ Increas~ Bondin~ ~ Increas~ed Bondin(j ~ ~ ~ ~ ~ _,~Increased Bondincj ~ Increased Bonding ~oo--§~ ~' ~ ~-~ ~ 22~s~- o~ ~,~sr ~: ~VE s ~~ 4oo-~ ~ .... --- -------- --- - - 600-~w 800 ,~ - I OOO- I~00 ,, 1400- ,~oo ~~- - ' ------- -,------=,BASE~----~OF----., ERMAF ROST .--~ , BEFORE THAW AFTER ~ MONTHS OF AFTER ~.5 MONTHS AFTER 16 MONTHS THAW OF THAW OF THAW FIGURE !-2~ PERMAFROST THAW TEST TEMPERATURE PROFILES FOR TEMPERATURE OBSERVATION WELLS TEMPERATURE --OF 14 18 22 26 3,0 ;54 ;58 42 46 50 54 58 / ~2 I I I ~ \ ~,~ ~ ~, t2:/12 ~ ,. %~ / 8/14-~ ,' / {~- ' ~.- ..... ~ ........ TEMPERATURE OBSERVATION WELL A (E- II II I II II 9/9 11/6 i/2~/74 ~ ~ % ~ '~ ........... TEMPERATURE OBSERVATION WELL B (E-4 I I IIII ~ I I II i ii i i I, , ..... ~t/~b ~ TEMPERATURE OBSERVATION WELL C (E-6) 5/15 ~1 I .I II I I IIII I II I III I I00 200 500 400 I00 2O0 $00 400: I00 300 400 APt'ENDIX II - PEI,h\iAFROST THAW STIb\IN PREDICTIONS 1. Permafrost Thaw Strain Mechanism The maximum expected subsidence strains included in this report were computed with A.R. Co. 's numerical mode].. The problem of computing the ex~pected strains at a given location in the permafrost is quite complex. There are at ]_east three important casing loading mechanisms that influence the expected subsidence strains. 1. An inward lateral movement of surromtding frozen permafrost. 2. An upward elastic flexing o£ soil below the thawed region. 3. A tendency for do~m.ward movement of thawed material generally throughout the thawed region caused by the mass of the thawed mater'[al, the gravitational field, .and the drop in pore pressure. The overall, subsidence loading mechanism is complex but is il].ustrated by the sketches in Figures II-1 and 11-2. While the permafrost is thawing, the shrinkage in volume resulting from the conversion of ice to water i.n the pore spaces causes a reduction in the pore pressure in the thawed region. Figure 11-3 shows our test measurements and. interpretation of the pore pressure profile in th.e 5-spot subsidence test thaw zones after 16 months of thaw. This decrease in the pore pressure in the thawed region, tends to reduce the total horizontal an.d vertical stresses. This reduction in horizontal stress in turn causes the permafrost at the frozen boundat~y to move II-1 laterally toward the wellbore and compress the thawed permafrost. The significance of this mechanism is more easily appreciated if the thawed region is dra~m to scale as is sho~n in Figure 11-4. The maximum expected permafrost and casing strains result from the interplay of downward acting forces within the thawed region and from the horizontal compaction of the permafrost in intervals where there are alternating layers of stiff sand and compressible silts. The axial tensile and compressive strains produced by this interaction are caused by the difference in the compressibilities of layers as shown in the sketch in Figure 11-2. As the boundary of the thawed permafrost moves toward.~he wellbore, it laterally compacts both the layers of sand and silt. Since the sand is less compressible than the silt, the lateral compaction, causes the sand to expand vertically into the silt thereby generating tensile (lengthening) strains in the sand and adjacent Casing and compressive strains in the silts and adjacent casing. Super- imposing the effect of gravity can cause tensile strains in compactible layers located just below relatively thick and stiff layers of sand. Furthermore, large compressive strains are induced in compactible layers just above relatively thick and stiff layers of sand. See Figure II-5. The magnitude of these tensile and compressive strains is highly dependent on the layer thickness, their mechanical properties (especially the ratio of the compressibilities of the adjacent layers), the thaw radius, and the depth of the interval. 11-2 The second most important thaw subsidence ~nechanism is the reaction of the pressure boundary at the base of the pemnafrost. The base of the permafrost is deflected upward because of the difference in the total vertical stresses in the thawed region and the formations below the permafrost. As is indicated in Figure II-1, the upward movement at the base of the permafrost .causes axial compressive strains in the pe~nafrost and casing above the base and tensile strains in the unfrozed formations below the permafrost. The magnitude of these strains, relative to the alternating tensile and. compressive strains due to the lithology in the subs'idence ,. ', test, is shown by the measured casing strains in the 5-spot test included in. Figure 1-12. The third mechanism predominates in the upper 400' to 500' of the permafrost lin. own as the "First Gravels". The casing strain measurements and computer simulations of this interval indicate that it reacts differen, tly than the fomations below this depth. This difference results in lower permafrost and casing strains than will occur in. the formations below this depth. Very little vertica.l strain in the permafrost or' casing has been measured or is expected in this interval because of two factors. First, th.e gravel, appears to be underconsolidated. .This underconsolidation leads to a low soil shear strength in the frozen material. Compaction occurs principally from a. lateral direction. Shear failure in the frozen permafrost, essential, ly at the thaw boundary, pemit the thawed soil and pipe to move do~m together. This limits the differential vertical strain in this region. Second, the permeability is apparently high in these gravels and measurements show that the pore pressure is essenti, ally not reduced below a hydrostatic 11-3 head of water during thaw. This factor also tends to limit thaw strain. These differences are consistent with the results of geologic studies of the permafrost at Prudhoe. q~ey are also consistent with the difference noted while drilling the permafrost and is best shown by the hole caliper logs included in Figure 1-18. Note that the ~per gravels are washed out to a ~nuch greater extent than any of the zones below that point. Geologic studies of the depositional cycle indicate that the upper gravels at Prudhoe were cyclically deposited, frozen, thawed, and refrozen. This cycle causes the excess surface ice formations to be melted and the soils compacted, to a net vertical stress equivalent to about 50' of burial or less. 2. Maximum Expected Casing Strains The numerical subsidence model was developed, from a basic ~mder- sta~ding of the subsidence mechanism and.~ measurements of mechanical pr°perites of the pe~nnafrost materials. The quality of the fit between the calculated perfonna.nce and test measurements is shown. by the plots of calculated m~d measured casing strains for 5 months, 7 months, a.~d ll months of thaw in Figures 11-6, 11-7, and. I1'-8. The results were calculated using mechanical properties for the permafrost wh"lch are consistent with the measured casing strain data taken after' 16 months of thaw. The model was calibrated by selecting formation compressibilities that would provide the best fit between the calculated and. meas~red strain data after 16 months of thaw in the experiment. The quality of this fit is shown in Figure 11-9. Figure II-10 compares the formation compressibilities of Px~dhoe permafrost soils measured in the lab with the compressibility values calculated from the calibration of the subsidence model for the test site. This plot also includes curves of the compressibilities used. in our worst caso stud'i.es with the sub- sidence model. The two worst'case curves encompass 99% of th.e data. I1-4 The maximum expected strains for any layer thickness or location in the field were co~uted with the subsidence model using con- servative data and worst case constraints for the critical loading mechanisms. With this definition, the model was used to calculate the maximum possible casing strains as a function of layer thickness and depth for the thaw pattern of the proposed well design. The maximum possible compressive strains occur in highly compactible silt layers. Consider first thin silt layers which alternate with sand.layers. Figure II-11 illustrates maximom computed strains for silt layers five feet thick which alternate with sand layers 25', 50', and 75' thick. As sand layers become very thick, the silt layers are essentially isolated from one another and consequently undergo maximum strain. Figure I1-12 is a plot showing computed maximum compressive strains for essentially isolated silt layers of various thicknesses. Now consider thin sand layers located in thick silt sections. Because of the greater stiffness of the sand, tensile strains are induced below the sand and increased compressive strains are induced in silts above the sand layers, As sand layers become widely spaced, they become essentially isolated from one another, and computed strains reach a maximum. Figure 11-13 illustrates computed behavior for widely spaced sand layers. As the thickness of an isolated sand layer increases, its resistence to deflection (and thus its effect on contiguous silts) increases and approaches a maximum effect. Figure 11-14 shows similar calculations for essentially II-5 isolated sand layers located at different depths. The envelopes of maximin tensile and compressive strain have been dra;m to include all maximum values. At shallow depths, the largest values of strain are computed for isolated layers of sand in thick silt sections. Near the bottom of the permafrost the largest values of compressive strain are computed for thin silt layers in thick san~d section. The most i~nportant factors to be considered in strain ca].culations are the mechanical properties of the pexm~afrost and the final pore pressure in the thawed pe~.~nafrost. ]~e key mechanical properties for this calculation are the compressibilities of the thawed permafrost. Figure II-10 shows: (t) Compaction data which were determi.ned in the laboratory using soil samples from Prudhoe Bay. (2) Va].ues of compaction coefficients which were used in the numerical model to match the field test data. (3) Maximum credible limits of compaction coefficients which were used during calculation of worst case strain. The maximum limit selected for ML (silt) litl'~.ology is also greater than compaction coefficients which were e×3~erimentally determined for a number of pure clay samp].es. The conservative nature of this selection is better shown in Figure 11-15 which compares the ratio of the compressibilities of silt to sand as a function of depth for the worst case calcul~tion along with the ratio of average compressibilities determined in laborato~3~ experiments. Ba. se~. on statistical ana.].ysis of these data., w'e compute that the probabi].ity of 1I-6 finding adjacent layers with higher compressibility ratios to be less than .1%. The final pore pressures in the thawed region of a producing well will probably be higher than the pressures measured in the highly accelerated thaw test. t-lowever, for the worst case approach we have assm~ed an even more conservative pore pressure profile than found in the 5-spot test. The selected gradient is shown in Figure 11-3 along with thc profil, e used in calibrating the model with the 5-spot test. Near the top of the permafrost the field data. shows that essentially a full hydrostatic head of water is maintained in the shallow gravels even during the accelerated test. This stone gradient was assumed. :for the worst case. The pressure gradient measured in the silts below the shallow gravel was also used for the worst case because these pressures are considered to be the minimum pressures that can be expected duri..ng tl~e extended life of a real producing well. Below this trend a zero pore pressure profile was extended to a point near the base of the pe~nafrost. The selected gradient near the base of the Permafrost that connects the assumed zero profile with the no~nal hydrostatic pressures below the permafrost was determined as the bom'tding case. This pressure gradient was based on the concept that the minimum pore pressure and the maximmt compressive strain that can' occur near the base of the permafrost are limited by the amom~t of compaction that is necessary to account for the shrinkage of the ice during thaw. ~is limiting pressure varies gradually with depth because the percentage of in situ frozen water near the base of the pet~nafrost changes rather slowly with depth. 11-7 3. Subsidence Strains in the Top 500' of Permafrost The earliest industry concem~s for subsidence of the upper 500' or First Gravels were based on the concern that this interval might contain excess ice, and that tha~ving would cause large permafrost and casing strains. This concern was in part justified because of the evidence that the upper gravels were deposited during the last ice age and might have been alternately deposited and frozen. This would lead to tu~derconsolidation of this interval and the possibility of the inclusion of ice lenses and. wedges at depth. Since these early studies, there has been considerable evidence developed that the First Gravels do not contain excess ice below the top 50'. Underconsolidation will lead. to relatively low net soil stresses and shear strengths of the frozen permafrost. This results in a slightly different sul~sidence mechanism than occurs in a normally consolidated interval. C~npaction occurs principally in a lateral direction. Differential vertical strain in the thawed region will be minimized by shear failure in the surrounding frozen material. 5hall vertical strain is certainly shown in results of the 5-spot subsidence test. As shown in. Figure 1-].2, the maximum strains in the top 430' were less than half of the strains measured below that point. Although no other site has been thawed to the extent of the 5-spot subsidence test, there have been a number of other wells that thawed the permafrost without noting ~ny adverse effects in the upper gravels. These, of course, include th.e two BP Alaska subsidence test holes II-8 thawed to about 10' to 15', the A.R. Co. topping plant st~ply and injection wells that included the Sag River No. 1, and the Drill Site 1-1 and 1-3 wells, and lastly, the Drill Site 1-6 and 4-6 surface holes that were thawed several times during the exten~al freezeback tests. The 5-spot test also included measurements of the surface and near-surface subsidence movements. Figure 1-15 is a s~nmary of the measurements taken after 16 months of thaw. Note that all of the movements are small and become negligible below 150' ". The model predictions for the top 500' of permafrost are most influenced by the change in pore pressure upon thaw and the initial horizontal and vertical, stresses in the frozen region. The pore pressure measurements in the 5-spot test indicate that the thawed gra:vels have remained saturated and that the pore pressures are equival, ent to a full colun~ of water .from the surface to the base of the gravels. Calibration of the model for the 5-spot test requires the asstm~ption of low net vertical stresses in the Fi.rst Gravels in order to match the test measurements. This assumption is consistent with the depositional theory that the Fi.rst Gravels were cycl. ically deposited, frozen, thawed, and. refrozen which prevents further consolidation with increasing depth of burial. 11-9 CONCEPTUAL SKETCH FIGURE I1-1 WELLBORE SUBSIDENCE SURFACE FROZEN REGION SHEAR BOUNDARY MOVES IN THAWED REGION REGION BEHAVING REGION POSSIBLY ELASTICALLY FAILING IN SHEAR ELASTIC FLEXING UPWARDS A CONCEPTUAL SKETCH SHOWING HOW LAYERS OF DIFFERING LITHOLOGY CAN LEAD TO ALTERNATING COMPRESSION AND TENSION IN CASING FIGURE 11-2 SILT LAYER SAND LAYER SILT LAYER / THAW FRONT COMPRE '~ 0 CASING I'~ LOW TENSION TENSION , ********************************************************* COMPACTIBILITY HIGH COMPACTIBILITY FIGURE PORE FOR 5 PRESSURE SPOT TEST PROFILES IN THE THAWED PERMAFROST AND WORST CASE PRESSURE GRADIENT 4oo 8OO 1200 RANGE OF MEASURED PORE PRESSURES BASE OF FIRST GRAVELS PORE PRESSURE PROFILE FOR SIMULATION OF COMPACTION AFTER 16 MONTHS THAW GRADIENT USED IN WORST CASE DESIGN PROBABLE PRESSURE GRADIENT IN ICE PHASE OF FROZEN PERMAFROST 0.45PSI/FT. 1600 (LIMITING COMPACTION) BASE OF PERMAFROST 2000 400 800 PRESSURE, PSI I 1200 11-3 FIGURE 11-4 BASE A TYPICAL THAW ZONE AFTER TWENTY YEARS SURFACE LENGTH OF ONE JOINT OF CASING THAWED REGION DRAWN TO SCALE OF PERMAFROST FIGURE 11-5 A CONCEPTUAL SKETCH SHOWING COMPRESSIVE STRAIN ABOVE, AND TENSILE STRAIN BELOW AN ISOLATED STIFF SAND LAYER THAW FRONT RELATIVELY STIFF SAND LAYER CASING COMPRESSION DISTORTION OF A MASSIVE SI THAT WOULD BE EXPECTED IN THE ABSENCE OF A SAND LAY CASING TENSION FIGURE 11-6 COMPARISON OF MEASURED AND COMPUTED CASING DEFORMATIONS FOR THAW SUBSIDENCE FIELD TEST AFTER 5 MONTHS THAW 200 -- 400 -- 600. -- 800 --' u. 1000' -- 1200; 1400' 1600~ 1800' 2000 COARPRESSIVE '~ TENSILE -.20 -.15 -.10 -.05 MEASURED STRAIN COMPUTED STRAIN BASE OF FIRST GRAVELS BASE OF PERMAFROST I I .10 .15 .20 STRAIN,% FIGURE COMPARISON OF CASING DEFORMATIONS AFTER COMPRESSIVE 200 400 600 800 1200 1400 1600 1800 2000 MEASURED AND COMPUTED FOR THAW SUBSIDENCE FIELD TEST 7 MONTHS THAW TENSILE I I I I -.20 -.15 -.10 -.05 0 .05 MEASURED STRAIN COMPUTED STRAIN STRAIN, % BASE OF FIRST GRAVELS BASE OF PERMAFROST I I I .10 .15 .20 FIGURE 11-8 200 400 600 800 :z: 1000 1200 1400 1600 1800' 2000 COMPARISON OF MEASURED AND COMPUTED CASING DEFORMATIONS FOR THAW SUBSIDENCE FIELD TEST AFTER 11 MONTHS THAW COMPRESSIVE 'A' TENSILE I I I I -.20 -.15 -.10 -.05 ~' MEASURED STRAIN · COMPUTED STRAIN BASE OF FIRST GRAVELS I 0 .05 .10 .15 .2~0 STRAIN ,% FIGURE 11-9 200 400 600 800 1000 1200 1400 1600 1800 2OOO COMPARISON OF MEASURED AND COMPUTED CASING DEFORMATIONS FOR THAW SUBSIDENCE FIELD TEST AFTER 16 MONTHS THAW I I I Cfi ~ I 1 I I COMPRESSIVE '~ t TENSILE -.20 -.15 -10 -05 ,~' MEASURED STRAIN · COMPUTED STRAIN BASE OF FIRST GRAVELS ~,,,-~ BASE OF PERMAFROST 0 .05 .10 .15 .20 STRAIN, % 10 3x10 -2 -3 -3 10 3x10-4 _ -5' -5 10 10 lOJ4 3x10 FIGURE I1-1 COMPARISON OF EXPERIMENTALLY DETERMINED COMPACTION COEFFICIENTS WITH VALUES REQUIRED FOR NUMERICAL MODEL TO PREDICT TEST WELL BEHAVIOR i e SAN D (SP) LABORATORY A SILTY SAND DATA DSANDY SILT (SM) OSILT (ML) vCOMPACTIBILITIES REQUIRED FOR MODEL TO PREDICT TEST WELL BEHAVIOR IO0 300 1000 3000 DEPTH, FEET FIGURE I1-1' 400 800 12OO 1600 2000 TYPICAL COMPRESSIVE STRAINS IN SILT LAYERS IMBEDDED IN THICK SAND SECTIONS MAX.COMPRESSION ENVELOPE SILT LAYER THICKNESS -' 5 FEET RATIO OF SAND TO SILT LAYER THICKNESS a 15 b IO c 5 ~ 4 I I I -.5 -.4 -.3 -.2 -.1 COMPRESSIVE STRAIN , % '1 .1 .2 I I(,tllil II MAXIMIIM ~ ()MI'I~I ',',IVI I',()IAII I) ',111 I A YI I~ I ()1~ '~II~AIN IN AN ',1 I1( II I) I)1 I'111', .4 .3 .2 .1 I I 1 5 10 15 DEPTH - 600 ~1100 -~ 1400 LAYER THICKNESS, FEET 400 800 -r- 1200 1600 2000 FIGURF 11-13 TYPICAL STRAIN CALCULATIONS TO DETERMINE WORST CASE STRAIN DISTRIBUTIONS IN PERMAFROST SECTION CONSISTING OF A SERIES OF THICK SILT SECTIONS AND THIN SAND SECTIONS I I I I ! I ~ I ] I ! I I ~ " .-" MAX TENSION ENVELOPE / ' / JF~'5...-"-" /t,,CKNESS(Ft.) stra,n,n - ~ /,/] .. ......... / SILT I SAND SAND SILT o~ ~~~1 '5. "-?C~..-z~.. - C VE - -.7 -.6 -.5 -.4 -.3 -.2 -.1 0 .1 .2 .3 .4 .5 .6 400 8OO -r- 1200 1600 2OOO FIGURE 11-14 TYPICAL STRAIN CALCULATIONS TO DETERMINE WORST CASE STRAIN DISTRIBUTIONS IN PERMAFROST SECTION CONSISTING OF A SERIES OF THICK SILT SECTIONS AND THIN SAND SECTIONS ~ - ~~ / / .AX, TENSION ENVELOPE~r£ /'/ " ' .e ~ CALCULATED ~ /~.~' /THICKNESS (FT.) STRAIN IN ...... " /' SILT SAND SAND SILT - · '" ' i 320 40 i~ ® MAX. COMPRESSION ENVELOPE ~ ~' ~ ' - ~ Lo /" /~ ~._~ '~'"',,480 20 [] · I I i : i / I I i -.7 -.6 -.5 -.4 -.3 -,2 - ! 0 .! ,2 .3 .4 .5 .6 STt~AIN FIGURE z 12 10 100 RELATIVE OF COMPACTION COEFFICIENTS SAND AND SILT SOILS I RATIO OF THE LABORATORY AVERAGE OF DETERMINATIONS 300 DEPTH , FEET 1 1000 2000 APPENDIX III - AI,LOWABLE C~tSING STFt&IN LIMITS 1. Minimum Ultimate Strain Capacities The minimt~n ult~nate casing strain capacities were calculated with a finite element model of the 13-3/8" 72#/ft N-80 Buttress threaded connection. The results of these calculations are sun~tarized in Table III-1 and are show~ in Figure II1-1. The finite element mode], was calibrated, and the strain limit failure criteria was established using the results of full scale stra. i~ limit tests of ].3-3/8" Buttress casing co~mections. Failure in compression :i.s indicated, in th.e model when. the first tooth on the pin loses it sealing force, which represents a potential for separation between this tooth and the collar. Failure in tension is defined, in th.e model by the magnitude of the average strain in the pipe body at the root of the last loaded tooth on the pin. Failure occurred in the test when th.e calculated strain reached 6.7% at the critical location. The ultimate strain capacities of pex~na, frost casing are defined. as the average axial, strains i.n the pipe body that are requi'red to fail the coup].ing. The model predicts a 3.3% ultimate strain caps. city for a simple compressive load on Normalized N-80 casing with measured casing wa.l], thicl~ess and normal th'read make-up. ~f~e sddition of III-1 pressure on a Normalized N-80 casing has a beneficial effect on the calculated compressive strain capacity. For combined loads of 1.800 psi intenna/ pressure, a thmm~al load represented by a 100°F increase in temperature on Normalized N-80 casing with the nominal APl wall thickmess, the model indicates a compressive strain limit of 3.9%. A calculation was also performed using a constituative relation that approximates the properties of Quenched and Tempered N-80. The results indicate that Quenched and Tempered N-80 develops even higher ultimate strains than Normalized N-80. The results included in Table III-1 for Quenched and Tempered pipe show that failure has not occurred with a 7.2% compressive strain in the pipe body. The model was also used. to evaluate the sensitivity of the strain limit to make-up interference between the pin and the collar, l.~ith 40% less than the normal interference, the ultimate compressive strain capacity is reduced to 2.3%, with more than the normal make-up the ultimate compressive strain capacity is increased. The ultimate strain capacity for simple tensile loads is higher than for compressive loads. For simple t~nsile loads on Normalized N-80 casing with measured casing wall. thickness and normal make-up interference, the ultimate strain capacity is 3.7%. The addition of an 1800 psi internal pressure and employing a nominal casing wall thickness decreases the strain limit to 3.4%. Reducing the make-up interference apparently has little effect on the tensile strain capacity. A calculation using 40% less than normal make-up interference indicates ..111,2 a tensile strain limit of 3.7% which is the stone as the lhnit for a normal make-up condition. From these studies, we conclude that the minim~n ultimate strain capacities of 13-3/8" 72#/ft N-80 Buttress casing are 2.3% compressive and 3.4% tensile. 2. Casing Strain Limit Tests A series of full scale, well instr%~nented strain limit ~tests were conducted to aid in establishing the tensile and compressive strain limits of 13-3/8" 72#/ft N-80 Buttress casing. A total of three tension and three compression tests were performed. The results of these tests are sm~narized in Table III-2 an.d are compsred with the finite element model predictions in Figures 111-2 and Ill-3. Axial. and hoop strains were measured at several places on the pipe body external to the coupling, inside the surface of the pipe body opposite the threads, and on the outside of the coupling. Strains were also calculated at the same positions in the finite element model. Figures 111-2 and Ill-3 are comparisons of the measured and calculated axial strains at two places on the coupling and at a position one inch away from the coupling on the pipe body. Figure 111-2 compares the calculated compressive strains with the results of three compression tests. Failure of the connection is predicted with a pipe body strain of 2.1% at a point one inch from the coupling. The data indicates that the strain at this location is between 1.9% and 3.4% at failure in Test #1 and greater than 1.9% in Test #2. In Test #3, creep was measured at each load 111-3 and the strain increased above the predicted value. Figure 1II-3 compares the model predictions with the results of the three tension tests. Failure is predicted with a pipe body strain of 3.1% at a point one inch from the coupling. The strain at failure one inch from the coupling is between 2.75% m~d 2.9% in Test #4., and between 3.6% and 4..05% in Test #5 where the threaded connection failed. Test #6 shows a failure strain greater than 2.6%. In Test #4, the ultimate was not reached because the test fixture failed. Figures 1II-4. m~d III-5 are photographs of the 13-3/8" test specimens. .Note that all of the compression san~les show at least one localized buckle in the pipe immediately above the coupling. ]Zis same distortion is also indicated by the finite element studies. Also the tension failures clearly occured in the threaded region of the pipe body near the last tooth. Following the tensile m~d compressive tests the remaining threaded connections, which had not failed, were pressure tested with water. All of these threaded connections were found to be pressure tight ~nder tests ranging from 500 to 1000 psi. 111-4 FI?~-iTE ELEMENT MODEL CALCULATIONS 13 3/8", 7Z Lb/Et, Normalized, N-80, Buttress Casing TABLE I!!-1 Case No. De s c ription (1) t/fall(Z) A~xial Pipe Strain(3) Axial Pipe Strain Thickness Axial Load 1" from Coupling Away from Coupling (Inches) (106 Lbs) (%) (%) Compression failure test comparison Compression failure with P = 1800 psi, AT = 100°E 0.493 -Z~ 14 -Z.07 0.5 14 -Z. 14 -2. 16 -3.27 -3.88 Compression failure ~vith 40% less than normal makeup 0.514 -Z. 10 - 1.38 -Z. 29 Below failure compression quenched 8~ tempered N-80 Tension failure test comparison Tension failure with P = 1800 psi 0.493 -2.05 -4.40 0.493 2.31 3.08 0' 514 2.46 2.98 -7. 19 3.68 3.39 Tension failure with 40% less than normal makeup 0.514 2.33 3.10 3.68 (1) Ali cases are normalized N-80 material, normal makeup, and no internal pressure or temperature change unless specified otherwise. (Z) 0. 493" is the wall thickness of the test specimens, 0. 514" is nominal thickness. (3) small tensile strain due to makeup is not included. In compression the strain at this location includes a component due to flexure in the pipe, so the axial deflection is greater than is indicated by the strain gage. SUMMARY OF FULL SCALE COMPRESSION AND TENSION TESTS 13 3/8", 72 Lb/Ft, Normalized, N-80 Buttress Casing TABLE !11-2 Test No. Pipe Yield Load Type Load 106 Lbs. Pipe Strain 1" from Pipe Strain 1" Coupling Load fr om Coupling % 106 Lbs. % Connection Failure (2) Failure Mode(1) Compression Compression Compression 1.78 0.26 2. 10 1.9 - 3.4 1.76 0.25 2.10 >1.9(3) 1.76 O. 26 2.25 >3. 1(3) Tension 1.74 0.30 2.25 2.7 - 2.9 Thread Jump Thread Jo_rn p Fracture at Bulge in Pipe Fracture in the Weld at the Test Fixture Tension 1.74 0.30 2.35 3.6 - 4. 1 Tension 1.78 0.30 2.25 >2.6(3) Fracture at Threads in Pipe Fracture at Threads in Pipe (1) All failures occurred in the connection made up at the mill. Some strains at failure are estimated from residual strain data taken after failure. (3) Ultimate strain not available because of strain gage failure. ULTIMATE STRAIN PREDICTIONS FOR CONNECTION FAILURE 13 3/8", 72 LB/FT , NORMALIZED , N-80 , BUTTRESS CASING FIGURE II1-1 120 m 100 o x 40 20 0 B o-COMPRESSION FAILURE A-TENSION FAILURE I B-40% BELOW NOMINAL MAKE UP N-NOMINAL MAKE UP P-1800 PSI INTERNAL PRESSURE I T-100°F TEMPERATURE INCREASEI NN B,N N P,T Z 4.0 AVIAI DIDI: C:TDAIkl 07' FIGURE III- 2 COMPARISON OF MEASURED AND COMPUTED STRAINS IN COMPRESSION 13 3/8'~ 72 LB/FT, NORMALIZED, N-80, BUTTRESS CASING 2.5 -- 2.0 -- 1.5- 1.0- 0.5- ~ 120 - 100 - u., 80 ~-. 60 X 4O > 20 / B A I / / LOCATIONS o - TEST #1 o - TEST #2 ~ ~ / /~- TEST #3 (CREEP ,/Z AT EACH LOAD) ! / x COMPUTED ~ ~ DENOTES FAILURE /,~ i I i 1.0 2.0 BEYOND DATA POINT 3.0 J AVG. AXIAL STRAIN, % COMPARISON OF MEASURED AND COMPUTED STRAINS IN TENSION 13 3/8", 72 LB/FT, NORMALIZED, N-80, BUTTRESS CASING FIGURE 111-3: 2.5 2.0 1.5 1.0 0.5 C 120~ B A C 0 .---.~.~ ~ I",.., J I",.,- 100 '~ ~' ~o"~' ~'°'~'°'-° ~ I / ~ -c~ o j,,,_ o-TEST #4 (CREEP AT j ~ ~ EACH LOAD) '~ a--TEST #5 80 ~--TEST #6 I O O ~ ~/ ---× co~.u~o / I 11 D~NO~ES FA,LU.E / 60 ,,z~ BEYOND DATA POINT/ 1 COCAZ~ONS 1 20 -o I . 4.0 _L L L -J- 1.0 2.0 ~_: 3.0 AVG. AXIAL STRAIN, % 13 s/8 FAILED inch, COMPRESSION 72 #/ft CYN-80 FIGURE SPECIMENS Buttress Casing III- 13 FAIL~2D inch, 72 TENSION SPECIMENS #/ft CYN-80 Buttress FIGUFIE 111-5 Casing APPENDIX IV - PEI'ClAFROST GEOLOGY 1. StratigraFhy and Age of Permafrost Sediments ~e permafrost zone at Pru~oe Bay includes an upper "First Gravels" interval, 200 to 500 feet thick, which overlies a thick sequence of sands and silts, the Sagavanirktok Fm~nation (Hm~itt, ref. 8; Barker, ref. 6; Clarke, ref. 7). As illustrated by the gamma ray log cor- relation diagrams, Figures IV-1 and IV-2, prepared by J. E. Eason with A.R.Co., the Sagavanirktok Formation contains several cor- relateable silt horizons which dip gently to the northeast. The First Gravels contain no correlateable silt intervals and form a homogenous, northeasterly thickening clastic wedge. Both intervals of the permafrost are continuous across the Prudhoe Bay field area. The base of the permafrost was determined from temperature measurements and by log evaluation, and lies in the Sagavanirktok Formation, as shown on the cross sections. Palynological examination of the BP permafrost cores by A.R. Co.9, and BP8, have shown that the First Gravels were deposited in an Arctic climate similar to or Slightly warmer than that of the present day. The underlying Formation contains a rich pollen flora M~ich can be correlated with early to middle Miocene floras found elsewhere in Alaska. The climate at that time was cool and humid. The stratigraphic sequence at Prudhoe Bay has been related to the late Cenozoic history of the Brooks Range to the South (Porter, ref. 10). The Sagavanirktok Formation represents the products' of the erosion of the Brooks Range during the Tertiary, M~ich virtually ceased with the onset of Plio-Pleistocene climatic cooling. The First Gravels record a relatively recent phase IV-1 of g]acial erosion of the Brooks Rm~ge, particu]arly during the ;, . Wisconsin and--ReCent. Superficial deposits at Plvdhoe Bay closely resemble those in the First Gravels beneath. Because no corre].ateable marine intercalations have been fmmd in the First Gravels it is probable that this interval is nearly ali Wisconsin to Recent in age. 2. Permafrost Consolidation and Ice Content ~I]~e Sagavanirktok Formatio]], including the permafrost interval below the First Grave]s, was deposited in a h~mtid., cool, ].ate-Tertiary envirorm~ent long before the development of permafrost on the North Slope. The normal consolidation profile which the Formation acquired during deposition was not signif:i, cantly affected by the refrigeration and permafrost development which occurred in the upper part during the Pleistocene. In .fact there is some evidence tha.t cyclic permafrost growth and degradation during the Pleistocene may have contributed to the stiffness of the soils in the deeper layers of the existing permafrost interval. The First Grave].s are the products of stream deposition in an envi.ro~nent resemb].ing that of the present day. The pollen evidence ,, suggests that most of the deposition took place in slight].y wa.truer conditions, probably when the various valley glacier systems recognized by Porter10 were decaying and releasing large voltages of water and glacial detritus. Although ice structures have developed both in the near-surface interval (0-50 ft, approximately) at the present time, and probably did so on oider surfaces wi. thin the First Gravels, the burial of these o].der surfaces is accompanied by the destruction of the excess ice structures. 'this may happen in at least two ways' -o IV-2 (1) The larger thaw lakes are associated with substantial local thaw of the permafrost beneath, especially of the decaying frost polygon structures which contribute to their development. (2) The streams which bring in the sands m~d gravels to bury existing land surfaces also represent a major source of heat input for the near-surface permafrost. A thmv bulb develops b~low and adjacent to each active stream of sufficient dimensions to thaw the previous generation of near-surface ice structures. Evidence of this process is available from the n~odern Sagavanirktok River (Sherman, ref. ll). Because the First Gravels may have remained frozen after burial to depths greater than those influenced by the tha~ beneath, stre~ns and. lakes, the deeper gravels may be somewhat rmd.erconsolidated. To smm~arize: the First Gravels contain free ice structures in the near-surface interval (about 0-50 ft), but the deeper part of the First Gravels are free of excess ice though probably somewhat und.er- consolidated. The deeper permafrost is more or less no~nally consolidated and free of excess ice. The sequence of events, and the stratigraphy of the permafrost are outlined on Table IV-1. IV-3 ~4 -IP~H i -Phit I · '-' "--" ! INPEX MAP OF PERMAFRO.ST cROss sECTIONS'" " "<~ ~ W ~1, FIGURE IV-3! E,~e,.- $~t .~ / Sn-CS Mob-Phll , $OCAL ConI-Sn-CS $o 36 d el ol / I / I #! BP "WELL t Cont-Sn-CS 33 I 34 3,5 Cont-Sn, -CS ' i SOCAL .. 2~ Pio¢l ,,' ~.~ OI 30 36 .3, 2-2 33 Put R No. 19-10-15 34 Ham et ai I - St. No. I uni( . ,2¸ ' Pennzc~,"~ OI g)e/,'~ 4'p ,..ir/,//. i, ,-PA TABLE IV-1 PERMAFROST GEOLOGY AT PRUDHOE BAY FORMATION First Gravels Hiatus Sagavanirktok Formation AGE Late-Pleistocene to Recent ~ ~. ~q~/~ ~ Plio-Pleistocene Tertiary to Mid-Miocene FLORA/CL~iATE Mosses and ferns with occasional tree pollen, probably reworked. Arctic, tundra climate. No record, probably frigid with several episodes of marine sub- mergence. Cool temperature forest association. PERMAFROST Similar to present day Development of permafrost in Sagavanirktok Formation No permafrost FIGURE IV-1 QUATERNARY & UPPER TERTIARY CORRELATIONS FOR PRUDHOE BAY AREA BP M-I (I-1-12) 200'--- 600' · I000'--- ~ O' :D 140 I 1800' BP BP BP BP AJR.Co. Exxon BP A.R. Co. Exxon Cent er PermafrostWell J-I D-I B-2 ThawWell 19-10-15A Lake St ~-- I (9-11-13) (23-11-13) (32-11-14)(27-1-14) (11-10-14) (19-10-15) (24-10-15) OF OF PERMAFROST 200 600' I000' '1400' 1800' SCALE' HORIZONTAL -- I"= 16,000' VERTICAL -- i"= 400' FIGURE IV-2 QUATERNARY TERTIARY CORRELATIONS PRUDHOE BAY AREA FOR ZOO" 600' I000' 1400 1800' B 2200' SOCAL 31 -25 (;)5-10-14) A. R. Co. ThowWell DS 2- 2 (11- 0-14)(56-11-14) OF -- EXXON P~. ~ ~ o,,(~o_~ _.~,~) s ~ ,: I SCALE' HORIZONTAL- i"= 16~000' VERTICAL I" - = 4 00' APPENDIX V - PE~FROST ]I-~W PREDICTIONS Several papers (Ref. 2-5) have been published that describe the thermal models for predicting permafrost thaw. Figures V-1 and V-2 are thaw predictions for two typical Prudhoe Bay oil wells using the current well design. Fi~ure V-1 shows the progress of thaw versus time for a mid-structure oil we~ that is expected to flow for the first 20 years of its producing life. Figure V-2 is the thaw prediction for an oil well completed lower on the structure. This well is expected to flow for 8 years and then be gas-lifted for the remainder of its producing life. All subsidence predictions have been based on the thaw prediction for a mid-structure completion after 20 years of production. Although the thaw should still be increasing after 20 years of production, we believe that our forecasts of the well completions, producing rates, artificial lift performance, and reservoir performance are no longer credible beyond the 20 year forecasts. Figure V-3 shows thaw calculations for an assumed 50 year producing history for a typical Prudhoe Bay completion. Note that for this case, the increase in tha~ beyond the 20 year curve is minimal and that at 50 years the~thaw front has retreated because the well has begun to refreeze as flow rates diminish late in the life of the well. V-1 CALCULATED THAW RADII FOR CURRENT WELL .FIGURE V-1 DESIGN U,.I 0 6OO 1200 1800 1860 MID STRUCTURE PRUDHOE BAY OIL WELL 0 10 20 30 40 50 60 70 80 90 TII A I&l n A I'~1 I I C' rCCT CALCULATED THAW RADII FOR CURRENT WELL FIGURE V-2 DESIGN 0 600 1200 18OO 1860 O LOW STRUCTURE PRUDHOE BAY OIL WELL 1 2 5 10 ,. 20 YEARS - 1 10 20 30 40 50 60 70 80 90 THAW RADIUS, FEET CALCULATED THAW RADII FOR CURRENT WELL FIGURE DESIGN V-3 6OO 1200 10 T-- F 1' 50 YEARS 20 PRUDHOE BAY OIL WELL 1800 1860 L 2O 30 40 50 THAW RADIUS, FEET 6O 7O 8O REFERENCES 1. Ruedrich, R. A., Perkins, T. K., and O'Brien, D. E., "Precise Joint Length Determination Using A ~ltiple Collar Locator Tool," SPE Paper 5087, presented at the 40th Annual Fall Meeting of SPE, Houston, Texas, October 6-9, 1974. 2. Couch, E. J. m~d Keeler, H. H., "Permafrost Thawing Around Producing Oil Wells," J. Canadian Pet. Tech. (1970), 9, 107. 3. Eicl~neier, J. R., Ersoy, D., and Raney, H. J., Jr., "Wellbore Temperature and lleat Losses During Production of Injection Operations,'' J. Canadian Pet. Tech. (1970), 9, 115. 4. Howell, E. P., Perkins, T. K., and Seth, M. S., "Calculating Tanperatures for Permafrost Co~'l~letions," Pet. Eng. (April 1973) 69. 5. Merriam, R., Wechsier, A., Boorman, R., and Davies, B., "Insulated Hot Oil Producing Wells In Permafrost," J. Pet. Tech., ~arch 1975), P. 357. 6. Barker, P. A., "Quaternary of the Prudhoe Field.," 1972, A.R. Co. "Report, Anchorage, Al. aska. 7. Clarke, R. H., "Quaternary History and the FirSt Gravels of the Prudhoe Bay Area.," March 1975, BP Alaska Inc. Report, San Francisco. 8. Howitt, F. "Permafrost Geology at Prudhoe Bay", World Petroleum, September (1971)~ 28-38. 9. Peterson, E. T., Hedlund R. W., Tabbert, R. L., and Bennett, U. E., "Pal~n~ological Ex-.ama~iation of BP 12-10-14 Permafrost Cores and BP 27-11-14 Cuttings," (1970) A.R. Co. Report. 10. Porter, S. C., "Late Pleistocene Glacial Chronology of North-Central Brooks Range, Alaska", Amer. Jour. Science (1964) 262, 446-460. 11. Sherman, R. G., "A Groundwater Supply for an Oil Camp Near Prudhoe Bay, Arctic Alaska," Proc. 2nd. Int. Conf. Permafrost (1973)4.69-472. PREPRINT--SUBJECT TO CORRECTION For release- April 7, 1975 Paper No. 364-A NOTICE TO EDITORS: Permission is hereby granted to reprint this paper on or after April 7, 1975, provided that the auspices under which it was presented be conspicuously acknowledged, the author's name and affiliation be stated, the original title be used, and that the paper be printed in full. Any devia- tion from this policy shall be approved by the author of the paper. If reprinted in installments, the foregoing conditions apply to each installment. SOLUTIONS FOR SOME PROBLEMS RESULTING FROM REFREEZING OF PERMAFROST AROUND A WELLBORE by T. K. Perkins, G. R. Wooley, and F. W. Ng Atlantic Richfield Company Dallas, Texas --oo0oo-- For presentation at the 1975 Annual Meeting Division of Production American Petroleum Institute Fairmont Hotel Dallas, Texas April 7-9, 1975 --oo0oo-- (The statements and opinions expressed herein are those of the author and should not be construed as an official action or opinion of the Institute.) --oo0oo-- Division of Production American Petroleum Institute Dallas, Texas SOLUTIONS FOR SOME PROBLEMS RESULTING FROM RE~REEZING OF PERMAFROST AROUND A WELLBORE T. K. Perkins, G. R. Wooley, F. W. Ng* ABSTRACT New well completion technology has been developed to deal with problems unique to arctic areas. Some of these ~problems result from the interaction of permafrost with the Wellbore. During drilling, completion, and production, some thawing of permafrost around the wellbore occurs. If wells are not produced immediately after completion, or if produc- tion is interrupted at a later time, the thawed region a~ound the wellbore will begin to re£reeze. Refreezing of the permafrost can lead to two types of problems. First, as water in the thawed region is converted to ice, its volume increases. This results in high fluid pressures which are imposed on the outer casing. Full scale field tests, extensive laboratory measurements of the mechani- cal and thermal behavior of permafrost, and theoretical and computer studies have led to an understanding of these pres- su~es. External refreezing pressures calculated for normal ~' operating conditions are in a range that can be tolerated if the proper casing is selected. A second potential problem is the refreezing of fluid within the wellbore system itself. If freezable fluids are left in annuli during the completion process, the possibility of high internal pressure exists due to the volume increase when freezing occurs. Because of the low pipe expansibility, pressures can rise to values which will cause collapse of inner strings or burst of outer strings. There are several practical approaches to avoiding this potential problem. This paper describes a unique displacement process which has been used to replace freezable fluids with nonfreezable, therm- ally insulating casing fluid. INTRODUCT ION The discovery of oil in arctic areas has led to signifi- cant changes in drilling and well completion practices. The harsh winter environment has necessitated substantial changes in surface operations. In a less obvious way, subfreezing temperatures have also created subsurface problems of consider- e:-At'lantic Richfield Company, Dallas, Texas A-1 able significance. The subsurface problems considered in this paper result from the interaction of permafrost with the wellbore. Permafrost of various thicknesses occurs in arctic regions where commercially significant hydrocarbon reserves may be found. Many wells have now been drilled and completed on the Alaskan North Slope where approximately 2,000 feet of permafrost are encountered. Even thicker permafrost has been reported in Russian technical literature. During the drilling and well completion operation, some thawing of the permafrost is caused by the drilling mud which brings heat up from the earth below the permafrost horizon. Drilling can typically result in a thawed region of a few feet in radius. Produc- tion of hot oil for a long perisd2o~ time can lead to thaw radii of several tens of feet, ~' '~ the exact value depend- ing on the operating conditions and the degree of wellbore insulation provided in the permafrost region. If the wells are not produced immediately after completion, or if produc- tion is interrupted at a later time, the thawed region around the wellbore will begin to refreeze. Refreezing of the permafrost can lead to two types of problems. First, as water in the thawed region is converted to ice, its volume increases. The increasing volume leads to an increase in fluid pressure which is imposed on the outer casing. The wellbore must be designed to withstand these "freezeback" pressures. A second potential problem is the refreezing of fluid within the wellbore system itself. If freezable fluids are left in'the annuli during the completion process, the possi- bility of high internal pressures exists due to the volume increase accompanying freezing. Because of low pipe expansi- bility, pressures can rise to values which will cause collapse of inner strings. This paper describes some practical solutions to these two potential problems. EXTERNAL FREEZEBACK Consider first external freezeback pressures. As the permafrost region around a wellbore is heated, ice in the soil pore space thaws; and its volume is thereby reduced approximately 9 percent. A pressure decline resulting from the volume decrease tends to be limited by gas expansion or influx of fluid such as drilling fluid filtrate, water from a thawed region near the surface or below the permafrost, laterally flowing brine, and gravity flow from thawed sections A-2 above. Permafrost refreezes most rapidly near the surface; hence, excess water may be trapped when deeper thawed regions refreeze. Upon refreezing, the water in the pore space expands, thus tending to increase the pore pressure. The increased pressure may simply force fluid to flow to another region. However, if this is not possible, the pressure con- tinues to rise until the soil or well casing deflects or yields sufficiently to accommodate the excess volume. Freezeback pressures h~.ve been studied in full scale field tests at Prudhoe Bay.~ Two wells were drilled and completed through permafrost to investigate thawing and refreezing phenomena. Thermistors and pressure transducers were attached to cables ~hich were external to the casing. Thawing was accomplished by circulating hot fluid in the wellbore. During the thawing and subsequent refreezing cycles, temperatures and pressures were recorded at the surf ac e. The Drill Site 4-6 was spudded February 14, 1972, and 20 inch conductor pipe was set at 109 feet RKB. A 17 1/2 inch hole was drilled to 2700 feet with fresh water mud. Casing, 13 3/8 inch 72 lb N-80 modified buttress, with external instrument cables was z~n to a depth of 2191 feet. The first stage of cement was 700 sacks of Permafrost II with the top identified at 1580 feet with a cement bond log. The second stage was cemented with 400 sacks through a DV tool at 486 feet. The 20 inch by 13 3/8 inch annulus was displaced with an oil-base casing pack, and the well was completed with open ended 3 1/2 inch 9.2 lb N-80 buttress threaded tubing hung at 2044 feet. The second well, Drill Site 1-6, was spudded May 27, 1972. An 18 1/2 inch hole was drilled with oil-base mud to 27~0 feet. The well was completed with 20 inch conductor set at 107 feet, 13 3/8 inch landed at 2698 feet, and 3 1/2 inch tubing perforated at the bottom and hung at 2~44 feet. The bottom of the 13 3/8 inch casing was cemented with 1350 sacks of Permafrost II cement in one stage. The fluid circulation system for the two wells consisted of direct-fired oil heaters, eirculation pumps, temperature controls, flow meters, and temperature and pressure recorders. The circulating fluid was a 50 percent mixture of ethylene glycol and water. The heat flux of I to I 1/2 MM BTU/hour for thawing was produced by fluid flow rates of 2 to 3 BPM. Several thaw-and-freezeback cycles have been completed ~in each well. A summary of thaw cycle data is presented in Table I. The first cycles in both wells were designed to simulate heat transfer to the permafrost during the drilling and completion of a development well in the Prudhoe Bay Field. The second cycles represent heat transfer to the per- mafrost during short production periods for a normal produc- ing well. Circulation in one of these wells for one month would transfer heat to the permafrost equivalent to many months of production since the production wells will be com- pleted with materials in the wellbore that provide greater insulation. Later cycles were made in the wells to investi- gate the effects of multiple thaw-and-refreezing and of changing the thaw radius. Temperatures external to the casing were recorded each six hours during both the th~w and refreezing parts of each cycle. Several additional temperature ~sttrveys were run in each well using a precision wireline instrument. Within a few days after circulation was stopped, the temperatures throughout the monitored sections of the wells dropped to the freezing point. The depth at which total freezeback had occurred was inddcated when the temperature would start dropping below the freezing point, which depends on water salinity, pressure, lithology, and mineralogy. This freeze- back depth progressed down the hole as water in the formation and in D.S. 4-6 the mud outside the casing, became completely frozen. Experimentally measured freezeback times agreed well with calculated~ values. Pressures external to the 13 3/8 inch casing were record- ed each six hours during both the thaw and refreezing parts of each cycle. The pressure history for the two wells is shown on Figure 1. Initially, the transducers indicated a pressure gradient equal to the hydrostatic gradient of the fluid outside the 13 3/8 inch casing. However, during the heating cycles the pressure measurements indicated a drop in the fluid level in the casing-borehole annulus. The drop in fluid level was confirmed by refilling the annulus. The non- linear behavior of the pressures with depth later during the heating cycles indicated barriers to vertical communication outside the casing of both wells. This is more pronounced during the second heating cycle due to additional sloughing that took place during the first cycle. The nonlinear pres- sure variation with depth is illustrated on Figure l, where du~ing the second heat cycle on D.S. 4-6 all of the deeper transducers indicated pressures of less than 200 psi with the exception of the transducer at ~71 feet. Similar condi- tions are shown during the second heat cycle in D.S. 1-6 where the deeper transducer pressures dropped below 200 psi. The shallower transducer pressures in both wells remained at or near a fluid gradient due to the annulus being filled periodically with water at the surface during the heating cycles. Immediately following the discontinuation of each heat cycle, pressures began to increase. The pressures in well D.S. ~-6 increased at each transducer level as the tempera- ture dropped to the freezing level and continued to increase until the temperature fell below the freezing point at that particular transducer level. The independence of each trans- ducer in well D.S. 4-6 is illustrated on Figure i where the maximum pressure at each transducer depth occurred as the temperature dropped below freezing regardless of the pressure level of the adjacent transducers. ~The water-base mud left in the annulus froze across each of the transducers, prevent- ing a vertical transfer of the pressure in the annulus. The pressure behavior differed in well D.S. 1-6 because of the presence of the oil-base mud left in the annulus du~- lng the completion of this well. The oil-base mud, which does not freeze or solidify, allows the vertical transfer of pressure in the annulus. This explains groups of transducers reaching a maximum pressure at the same time, independent of the temperature level at each of the transducers. Th~ is illustrated on Figure i in cycle No. i where maximum pres- sues were reached at transducer depths of 30, 129, and 33? feet, simultaneously. Also, a group of transducers at 6~2, and 968 feet reached maximum pressu~es simultaneously. This ~ould indicate vsrtical barriers in the annulus at 4~0 feet L and 1000 feet .. The same grouping of transducers is also apparent du~ing the second cycle. The location of these two barriers, plus additional minor barriers, was indicated by bending on a Cement Bond Log run prior to and following the second heating cycle. In addition~t~ the large scale field experiments, exten- sive theoretical~' and computer studies, and laboratory in- vestigations of peEnafrost properties have been undertaken to more completely understand the freezeback phenomena. The theoretical understanding begins with a computer simulation of theEnal behavior since it is the thawing or refreezing of permafrost that leads to its influence on the well. The mechanical behavior of tha~ed8and frozen soils has been dete~nnined experimentally.J, Using the results of the experimental studies and the theoretical equation relating stress and strain, a computer program has been developed~ that calculates refreezing pressures. Pressures calculated with the computer program agrees with the field data which are shown on Figure 1. Figure 2 shows, for example, the maximum freezeback pressures as a function of'depth for the various thaw cycles of the D.S. 4-6 well. These data indica- te that multiple cycles do not strengthen permafrost. Figure 3 shows similar behavior for the D.S. 1-6 well. Heat loss from this well was lower than for the D.S. 4-6 well because of the low thermal conductivity of the oil-base fluid outside the 13 3/8 inch casing. Maximum pressures observed for this well were generally lower than for the D.S. ~-6 well because the thaw radius was smaller and because the oil base fluid outside the casing could not freeze and create excess volume. The refreezing pressure gradient appeared to be lower in the region above ~00 feet where large washouts were logged. Pressure transducers in the interval from ~47 to 988 feet reached maximum pressure simultaneously and were related by the hydrostatic gradient of the oil-base fluid surrounding the pressure tranducers. Figure 3 also shows calculated behavior including the effect of non-freezing fluid at the wellbore. This analysis indicates that fluid communication through the annulus of unfrozen oil-base mud causes shallow freezeback pressures to be larger than they otherwise would be, and deeper freezeback pressures are less than they would be if the outer casing annulus was cemented or filled with a freezable liquid. The previously described computer program has been used to calculate freezeback pressures for field operating condi- tions. In a typical field case, the degree to which water can flow into the thawed region to maintain saturation, or the degree to which liquid can flow out of the thawed region to relieve pressure, is not known with certainty. In order to calculate a maximum likely pressure, we have assumed that the thawed region will remain saturated du~ing heating and that no fluid can escape from the thawed region during refreez- lng. The thaw radius depends on the exact well completion and on operating conditions. To illustrate behavior that is expected, we have modeled a well such as D.S. 4-6 which is completely cemented to the surface and which is insulated with a gelled oil fluid in the 13 3/8 inch by 9 5/8 inch annulus as well as in the 9 ~/8 inch by ~ 1/2 inch annulus. The well is assumed to be produced at moderately high flow rates for various lengths of time and then shut in to allow complete freezeback to the wellbore. Calculated freezeback pressures versus depth for refreezing after one year of production and for refreezing after 20 years of production are shown on Figure 4. Estimated times to refreeze at various depths are also shown. These calculations indicate that some- what greater pressures are expected for large thaw radii provided a mechanism is available to resaturate the thawed zone. On the other hand, extremely long periods of time would be required to refreeze and thus generate the pressures. Figure 4 also shows the collapse pressure of two weights of casing when filled with 9.6 ppg fluid and assuming no addi- tional strength resulting from the surrounding cement. A-$ A parameter study has helped to identify important variables, in the course of this study freezeback pressures were compared with that calculated for the following case: (Porosity) (Ice Saturation) - 35% Initial Thaw Radius : 5.3 ft. Depth -- 1000 ft. Earth Stress = 6~5 psi Permafrost Equilibrium Temperature = 22.3°F Freezing Temperature - 31.0°F Final Freezeback Radius = 0.557 ft. (13 3/8 in. casing) Freezeback Time = 200 days The effect of parameter variation is shown on Table II. The greatest changes in pressure occur when changing the maximum thaw radius and the final freezeback radius. Of course, it is also clear that depth is a significant variable. INTERNAL FREEZING Consider now the possibility of having water-base fluids in portions of the cased wellbore within the permafrost region. As permafrost 'refreezes the water-base fluid also refreezes and expands in volume. Pressure will rise until the excess volume can be accommodated by fluid compressibility or expan- sion of the pipe. If large volumes of fluid are frozen, the pressure may increase to a value which will cause casing damage. Many solutions to this potential problem have been proposed such as 7- ~ displacing the freezable liquid with non-freezable fluids or cement, removing undesirable fluids by swabbing and gravity drainage of the annulus, or by dril- ling the well with an oil-base drilling mud. One of the most attractive solutions is to displace the freezable liquid with 'an oil-base fluid which gels when it is heated by pro- duced fluids in the wellbore. Such gelled oils can be for- mulated to have low thermal conductivities and also be opaque to radiant heat transfer. Since gelling suppresses convective heat transfer, the gelled fluid serves as a theEnal insulator. Displacement of water-base mud with large excess volumes of gelled oil has beendescribed previously.~ More recently, a modified displacement process has been used which improves the efficiency of the displacement operation and reduces cost. Figure 5 shows a typical well completion. As explained in the previous section of the paper, when casing of adequate strength is employed, it will be satisfactory from a freezeback A-7 point of view to cement annuli external to the 20 inch and 13 3/8 inch casings. The annulus between 9 ~/8 inch casing and 13 3/8 inch casing can be displaced through a stage cementing tool located well below the bottom of the perma- frost. The water-base drilling mud which is being displaced will generally have a modest plastic viscosity and gel strength. Oil-base displacing fluids have been formulated which will irreversibly develop gel strengths sufficiently high to prevent convection as they are heated by produced fluids in the wellbore. During initial placement, the fluid is quite pumpable, yet it exhibits a relatively high plastic viscosity and gel strength even when injected at a low tem- perature. Because of the high viscosities and gel strengths of these two fluids, a direct displacement is usually in laminar flow. High displacement efficiency of the drilling mud requires that a considerable excess volume of gelled oil be injected. The more recently developed process includes as inter- mediate wash step using fresh water. Because of the lower viscosity, high pump rates will produce turbulent flow in the displacing water. If water is pumped down the drill pipe so as to enter the casing annulus and displace mud upwards, the viscosity ratio and gravity ratio in the annulus are unfavorable. Nevertheless, displacements can be relatively efficient. Turbulent eddies erode bypassed mud, transfer it laterally into the moving fluid stream, and thus lead to good displacement efficiencies which are characteristic of turbulent conditions. Displacement behavior during this step has been studied in small laboratory models (approximately ~ and 10 feet long) in an inteEnediate depth well (900 feet) and in full scale displacements at Prudhoe Bay. Figure 6 illustrates, for example, the types of displacement behavior observed. The data indicate that no more than two system (drill pipe plus annulus) volumes should be required to achieve good displacement of the water-base drilling mud. Following the water wash step, gelled oil is injected down the drill pipe to displace the water. If desired, a wiper plug between oil and water can be used. As the weighted gel enters the annulus, both gravity and viscous forces will be favorable. Further, the displaced fluid will have a low viscosity and no gel strength. Studies in laboratory models and in actual field displacements indicate highly efficient displacements. Figure 7 shows a comparison of typical results. Again less than two system volumes appears to be adequate to achieve a near perfect displacement. Use of a device to directly indicate water contamination of the gel in the effluent stream has proven to be very help- ful. A co~mnercially available "Net 0il Analyzer" has been A-8 successfully used at Prudhoe Bay. It consists of a Teflon coated metal probe located centrally in a length of Teflon coated pipe, both of ~hich are wired to an electronic oscil- lator circuit mounted integrally with the unit. The electri- cal capacitance of the device depends on the composition of the fluid in the pipe. This causes the oscillator to generate outputs of various frequencies. By calibrating output fre- quency against samples of gelled oil containing known water contamination, the probe can be used to monitor the effluent during gel placement. The reading is instantaneous and can be taken as often as desired. For field displacements, a comparison of probe readings with retort analysis of effluent samples has Shown excellent agreement. C 0 NCLUS I0 NS 1. Field experiments have shown that within the earth, there are mechanisms available to maintain positive pore pressures in thawing regions around a wellbore. 2. Upon freezing, pressures will rise until excess fluid can be dissipated, or until the frozen soil can be pushed back at a rate nearly equal to the rate at which excess volume is being created by the refreezing process. 3. Pressttres calculated with a mathematical model which is based on theEnal and mechanical behavior of Prudhoe Bay peEnafrost are in good agreement with values measttred in a large scale field test. 4. Pressures estimated for noEnal operating conditions are in a range that can be tolerated if the proper casing is selected. ~. In the completion of a typical arctic oil or gas well, water-base fluids which are inside casing opposite the permafrost section must be removed in order to prevent casing damage resulting from freezing of the fluid. 6. If freezable fluids are displaced by gelled oil, freezing problems will be eliminated while at the same time the gelled oil serves as a theEnal insulator. 7. An improved displacement process consists of an intermediate water wash step to displace drilling mud before injection of gelled oil. 8. Two system volumes of water are sufficient to remove virtually all the water-base drilling mud from the system. 9. Displacement of water with weighted gelled oil is highly efficient, and less than two system volumes has given near perfect displacements. 10. Use of readily available electronic instruments per~its continuous monitoring of the gelled oil effluent stream. A-10 REFERENCES 1. Couch, E. J., and H. H. Keller: "Pez~nafrost Thawing "Journal of Canadian Petroleum Around Producing 0il Wells, Te.c~nology, 1970, volume 9, P. 107. 2. .Eickmeir, J. R., D. Ersoy, and H. J. Ramey, Jr.: "Wellbo~e Temperature and Heat Losses During Production or Injection Operations," Journal of Canadian Petroleum Technology, 1970, volume 9,-p. 11~,- ' ...... 3. Howell, E. P., T. K. Perkins, and M. S. Seth: "Calculating Temperatures for Permafrost Completions," P_e~trole.um Engineer, April 1973, volume 45, P. 69. ~. Perkins, T. K., J. A. Rochon, and C. R. Knowles: "Studies of Pressures Generated Upon Refreezing of Thawed "Journal of Petroleum Technology, Permafrost Around a Wellbore, AIME T~r.ans.actions, October, 197~, volume 257, P~' 'i1~9."' 5. Goodman, M. A., and D. B. Wood: "A Mechanical Model "SPE Paper No for Permafrost Freezeback Pressure Behavior, . 4589 presented at 48th Annual Fall Meeting, Las Vegas, Sept. 30-0ct. 3, 1973. 6. Shalavin, A. M., and G. P. Klyushin: "Method for Determining the Pressure on the Casing Pipes with Freezing of the Flushing Fluid in the Well," (Metodika opredeleniya velichiny davleniya na obsadn~e truby pti zamerzanii promy- vovhnoi zhidkosti v skvazhine) Burenie, No. 9 (1972) pp. 2~-26. 7. Perkins, T. K., and R. A. Ruedrich: "Mechanical "SPE Journal, AIME Trans- Behavior of Synthetic Permafrost, actions, August, 1973, volume 255, p. 211. 8. Ruedrich, R. A., and T. K. Perkins: "A Study of Factors Influencing the Mechanical Properties of Deep Perma- frost,'' Journal of Petroleum Technology, ..... AIME Transac~ion.s, October, '1'974,' volume 2~7~ ~' 1167. 9. Bleakley, W. B.: "North Slope Operators Tackle "0il and Gas Journal, Oct 25, 1971, Production Problems, ........ volume 69, pp. 89-92. - 10. Flumerfelt, R. W.: "An Analytical Study of Laminar "SPE Paper No 4612, 48th Annual Non-Newtonian Displacement, . SPE Fall Meeting, Las Vegas, Nevada, 1973. A-11 ll. Graham, H. L., "Rheology-balanced Cementing " The Oil and Gas Journal, volume Improves Primary Success, ~ 70, Dec. 18, 1972, p. 12. Clark, C. R., and L. G. Carter: "Mud Displacement " Journal of Petroleum T~chnology, with Cement Slurries, ~- . _ . volume 25, July, 1973, P. 77~'. '~ " ' 13. Garvin, T., and K. A. Slagle, "Scale-Model Displace- ment Studies to Predict Flow Behavior During Cementing," Journal of Petroleum Technology, volume 23, Sept 1971, p 1081 -- ~ - --~- ~- ,~ .... ~ - u · ,, __ i · · · A-12 TABLE I A SUMMARY OF THAW CYCLES Cycle No. D.S. 4-6 D.S. 1-6 Date at Beginning of Thaw Cycle 2/22/72 6/28/72 3/17/73 6/4/73 7/27/73 Time at Beginning of Thaw Cycle Measured from Beginning of Fir st Thaw Cycle days 0 126 389 468 521 Average Input Fluid Temp. oF 113 160 140 130 135 Circulation Rate BPM 12.4 (Rig Pumps) 2 Circulation Time days 7 32 6/9/72 9/17/77. 3/7/73 201 373 138 150 125 3 (Reverse) 10.5 9O TABLE II PARAMETER VARIATION RESULTS Initial Thaw Radius ft 1 2 5 5.3 10 20 Final Freeze Radius ft 0. 557 0.75 1.0 1.5 2.0 (Porosity) (Ice Saturation) 0.10 0.20 O. 30 0.35 O. 40 Freezeback Time days 50 100 200 300 500 Pressure psi 1218 1388 1567 1582 1672 1678 Pressure psi 1582 1517 1467 1378 1312 Pressure psi 1418 1508 1562 1582 1602 Pressure psi 1587 1584 1582 1581 1580 Computed Freezeback Pressure ComPuted Freezeback Pressure for Base Case 0.77 0.88 0.99 1.0 1.05 1.06 Co. mputed Freezeback Pressure] Computed Freezeback Pressure for Base Case 1.0 0.96 0.93 0.87 0.83 Computed Freezeback Pressure COmputed ~'reezeback Pressure for Base Case 0.90 0.95 0.99 1.0 1.01 cComputed Freezeback Pressure omputed Freezeback Pressure for Base Case 1.00 1.00 1.0 1.00 1.00 gRILL SITE ~-6 EXTERNRL PRESSURE TIME (DRYS) ' 1 ii~l DEPTH (l#l! 1078 · 11~8 .~.~ .,~;;;~, ~ .... ~~-~~~ I/-- ...... ",.""-~'~ ~~- " ~~~ ,~ ~ ,.'" ~ . _ '~ -- ~ . ~ .... ~ ~.~ ~ .... _~~~ ' ' ,~ ' o.o ~.o ~.o m.o ii~,o IM.O l~,o IIO.O ~o ~o ~o [~.o ~.o ~,o ~.o ~.o [~.o Isto.o ~.o TIHE (DRYS) ORILL SITE 1-6 EXTERNRL RRESSURE FIGURE 1 EXTERNAL PRESSURES MEASURED AT D.S. 4-6 AND D.S. 1-6 DURING THAWING AND REFREEZING 20O 4OO 600 800 1000 1200 ID o MEASURED o CYCLE 1 0 CYCLE 2 D CYCLE 3 · CYCLE 4 LINES SHOW CALCULATED BEHAVIOR CALCULATED "'~ ,"~""~' '" CYCLE 5 ........... 200 400 600 800 PRESSURE, psi FIGURE 2 NAXIMUM FREEZEBACK PRESSURE DRILL SITE ,3-6 I I 1000 1200 1400 1600 200 400 "' 600 i, a_ 800 1000 1200 1400 MEASUREMENTS: o CYCLE 1 O CYCLE 2 ~ CYCLE 3 --- CALCULATION INCLUDING EFFECT OF NON-FREEZING FLUID AT THE WELLBORE 200 400 600 800 PRESSURE, psi FIGURE 3 MAXIMUM FREEZEBACK PRESSURES DRILL SITE 1-6 A-16 ,, J . 1000 1200 ,~°~ 100 D _J m ~ 80 ~; ~ 60 ~ ~ LEGEND' ~ 4o ~ n "' o I I I I 0 0.5 1.0 1.5 2.0 2.5 VOLUME OF WATER PUMPED IN ] VOLUME OF DRILL PIPE AND ANNULUS LABORATORY MODELS 900 FT. DEEP WELL FULL SCALE WELL AT PRUDHOE BAY I 3,0 FIGURE 6 DISPLACEMENT EFFICIENCY DURING THE WATER WASH STEP REYNOLDS NUMBER OF WATER 2700 TO 1],000 ]6,000 TO 40,000 26,000 O.5 1.0 1.5 2.0 - 100 - 80 60 40 -- 20 ~J Z Z z_ LEGEND · "' LABORATORY MODEL 900 FT, DEEP WELL FULL SCALE WELL AT PRUDHOE BAY VOLUME OF GELLED OIL PUMPED IN 1 J VOLUME OF DRILL PIPE AND ANNULUS FIGURE 7 DISPLACEMENT EFFICIENCY FOR WATER BEING DISPLACED BY GELLED OIL REYNOLDS NUMBER OF WATER 3,120 16,600 23,000 A-18 4OO 8OO 1200 1600 2OO0 2400 ~ ~. 10 ~e '~20 eYEARS TO FREEZEBACK TO THE DEPTH SHOWN 800 1600 ~e \~e ! 2O YEARS OF PRODUCTION BEFORE REFREEZING COMMENCES I ,1 2400 3200 4000 EXTERNAL PRESSURE, psi FIGURE 4 CASING STRENGTH (CASING FULL OF 9.6 PPG FLUID) AND ESTIMATED MAXIMUM FREEZEBACK PRESSURES PERMAFROST BOTTOM @ 2000 FT. 20" 94# H 40 @ 100 FT. FO CEMENTER @ 2400 FT. FO CEMENTER @ 2650 FT. ]3 3/8" 72# MN80 @ 2700 FT. LEGEND: CEMENT GELLED OIL WATER-BASE DRILLING MUD 51/2'' 17# N 80 TUBING SET AT PAY ZONE 9 5/8" 47# S0095 @ 10,000 FT. FIGURE 5 TYPICAL ARCTIC COMPLETION A-17 NOTICE OF PUBLIC HEARING STATE OF ALASKA DEPARTMENT OF NATURAL RESOURCES DIVISION OF OIL AND GAS Alaska Oil and Gas Conservation Committee Conservation Order File No. 137 The Alaska Oil and Gas Conservation Committee will hold a hearing on its own motion pursuant to Title 11, Alaska Administrative Code Section 22.540 in the Municipal Chambers of the Z. J. Loussac Library, Fifth Avenue and F Street, Anchorage, Alaska at 9:00 AM on November 25, 1975. The Committee will seek testimony on the following matters: l) To consider an amendment to Rule 4, Conservation Order 98-A, 98-B, and 83-C of the Prudhoe Bay Field pool rules to require a bag type blowout preventer while drilling the hole in which the surface casing is set. The present regulations pertain to re- quirements for blowout prevention equipment only while drilling below the surface hole. 2) To consider an amendment to Rule 3, Conservation Order 98-A, 98-B, and 83-C. Sections a, b, and d in particularly will be considered but the entire rule will be open for possible changes. Experience gained since 1970 appears to warrant changes in casing requirements in the permafrost portion of the hole. Thomas R. Marshall, Jr. Executive Secretary Alaska Oil and Gas Conservation Committee 3001 Porcupine Drive Anchorage, Alaska 99501 Publish: October 25, 1975
