18 DECEMBER 2013 SPE DALLAS SECTION GEORGE E. KING, P.E.

Transcription

1 FACTS ON ENVIRONMENTAL RISK IN FRACTURING & WELL CONSTRUCTION: WHAT DO THE NUMBERS SAY? 18 DECEMBER 2013 SPE DALLAS SECTION GEORGE E. KING, P.E.

2 Table 2 Fracturing Risk Events (From SPE ) Event Event Description Event Event Description 1 Spill transport of fresh or low salt 10 Frac ruptures surface casing water 2 Spill 15 gal biocide 11 Cooling pulls tbg string out of packer. 3 Spill 50 lb. dry additives 12 Frac opens mud channel, well < 2000 ft. 4 Spill 150 gal from truck wreck 13 Frac opens mud channel, well > 2000 ft. 5 Spill 2500 gal from refueler wreck 14 Frac intersects another well in pay zone. 6 Spill frac tank of water, no additives 15 Frac intersects properly abandoned wellbore. 7 Spill frac tank of water with food grade polymer only 16 Frac intersects improperly abandoned wellbore. 8 Spill 10 gal. diesel during refueling 17 Frac to surface or groundwater through the rock, well > 2000 ft. deep. 9 Spill 100 bbls of produced water 18 Frac produces earthquake that can be felt at surface. 19 Frac intersects a natural seep. 20 Frac emissions in excess of limits. 21 Normal frac operations no problems

3 Table 2 Fracturing Risk Events Event Event Description Event Event Description 1 Spill transport of fresh or low salt water 10 Frac ruptures surface casing 2 Spill 15 gal biocide 11 Cooling pulls tbg string out of packer. 3 Spill 50 lb. dry additives 12 Frac opens mud channel, well < 2000 ft. 4 Spill 150 gal from truck wreck 13 Frac opens mud channel, well > 2000 ft. 5 Spill 2500 gal from refueler wreck 14 Frac intersects another well in pay zone. 6 Spill frac tank of water, no additives 15 Frac intersects properly abandoned wellbore. 7 Spill frac tank of water with food grade polymer 16 Frac intersects improperly abandoned wellbore. only 8 Spill 10 gal. diesel during refueling 17 Frac to surface or groundwater through the rock, well > 2000 ft. deep. 9 Spill 100 bbls of produced water 18 Frac produces earthquake that can be felt at surface. 19 Frac intersects a natural seep. 20 Frac produces emissions in excess of limits. 21 Normal frac operations no problems

4 FRAC HEIGHT GROWTH IN ~FOUR THOUSAND JOBS NOT EVEN CLOSE TO WATER (Reprinted from the July 2010 issue of The American Oil & Gas Reporter with permission from Pinnacle, A Halliburton Service) Microseismic signal from top of fracs in relationship to bottom of fresh water. >3800 fracs tracked in 4 shales Table 5 Fracture Height-Growth Limits in Four Major U.S. Shale Plays (Fisher, 2011) Shale Number of fracs with microseismic data Primary Pay Zone Depth Range Typical Water Depth and (Deepest ) Typical Distance Between Top of Fracture and Deepest Water Closest Approach of Top of Frac in Shallowest Pay to Deepest Water Barnett to (TX) 8000 (1200 ) Eagle Ford (TX) 13,000 (400 ) Marcellus (PA) to (1000) Woodford (OK) , (600) Separation is 1 to 2 km. No breach of fresh water. The top-most microseismic signals are most likely stress transfer and do not represent fracture growth.

5 Slide 5 Well Barrier Failure vs. Well Integrity Failure Barriers are containment elements - Can individually isolate design loads. If one barrier fails, the next barrier accepts the load. Multiple, nested barriers = redundant barrier system. No Leak Path Well integrity failure = if all barriers fail in series = leak path Is formed Barrier Number Press Durability Casing + Cmnt 2 to 7 Very High Very High Hanger + Seal 2 to 4 High High replace easy. Pipe body 1 to 3 High Very High w/ corros. maintenance Packer, Plug 1 to 2 Mod. Mod. Safety Valve 1 Mod. to High Mod. Valves, Spools in Prod. Tree 4 to 20+, tandem? Mod. to High Mod- easy repair SPE , Barrier vs. Well Failure, King

6 Completed Well - How Many Barriers are Typical? SPE , Barrier vs. Well Failure, King

7 How Much Cement is Needed for Isolation? Every inch of cement is NOT required to be perfect. Slide 7 Quality of cement is more important than the volume. Over 10,000 psi can be held with less than 50 ft of cement, but 200 to 300 ft is routinely used. Isolation can only be measured with a pressure test. Bond logs are not always best tool ~10% channels missed. Instances of false negatives. SPE , Barrier vs. Well Failure, King

8 Risk = Frequency of Occurrence vs. Impact Slide 8 Risk exists in every action. What is operationally safe? Occurrence & impact create a threat level that we can understand & accept or reject based on what we believe: hopefully on assessment of facts. SPE , Barrier vs. Well Failure, King

9 Slide 9 Well Failures & Improvements in Development Eras Time Era Operation Norms Era Potential For Pollution 1830 to 1916 Cable Tool drilling, no cement, wells vented High 1916 to 1970 Cementing isolation steadily improving. Moderate 1930 s Rotary drilling replace cable tool, BOPs Moderate & Lower 1952 Fracs reduce # wells. Better pipe & cement Lower from Frac aspects 1960 Gas tight couplings and joint make up Moderate 1970 Cement improving, Horizontal Wells introduced Lower 1988 Multi-frac, horizontal wells, pad drilling Lower reducing environmental land footprint 90% 2005 Well integrity assessment, premium couplings, adding barriers & cementing full strings. Lower after 2008 to 2010 (STRONGER Reg Review) 2008 Chemical toxicity & endocrine disruptors Lowest yet, most states sharply reduced. Real time well integrity needs caught up with design and studied - early warning & avoidance. inspection requirements. SPE , Barrier vs. Well Failure, King

10 Well Study Review >650,000 wells Failure Factors Recognized: Type of Well Maintenance Culture Era of Construction Geographical Location Age of Well Specifics of Design & Construction Usage Change Single barrier compromised by tubing leaks. SPE , Barrier vs. Well Failure, King

11 Table 10 - Distribution of Barrier Failures and Well Integrity Failures Showing Improvement by Era (all land wells). > 650,000 wells Area / Number of Barrier Fail Freq. Well Integrity Leaks to Number of construction Range (containment) Failure Range GW by Wells failures (containment lost) sampling Ohio / 64,830 TX / 253,090 TX / 16,000 horizontal multi-frac MN / 671 Alberta / 316, fail initial cement test. 39 failed in production. 10 fail initial cement test. 56 failed in production. No reported failures added barrier Salt creep crush casing Total vent flow data % in 34,000 wells 0.1% in older wells worst case. 0.02% all wells. 0.02% for older era wells 0.004% for newer wells No failure reported No failure data or pollution reports Data Sources 0.06% for all wells Detailed not Kell, 2011 available 0.005% to 0.01% for producers 0.03% to 0.07% for injectors No well associated pollution 5.5% Unknown None reported No separation data 4.6% taken as available worst case. No data mostly gas escape Kell, 2011 TGPC data 1997 to 2011 Kell, 2011 Clegg, 1971 Watson & Bachu, 2009

12 Table 10 - Distribution of Barrier Failures and Well Integrity Failures Showing Improvement by Era (all land wells). > 650,000 wells Area / Number of Barrier Fail Freq. Well Integrity Leaks to Number of construction Range (containment) Failure Range GW by Wells failures (containment lost) sampling Ohio / 64,830 TX / 253,090 TX / 16,000 horizontal multi-frac MN / 671 Alberta / 316, fail initial cement test. 39 failed in production. 10 fail initial cement test. 56 failed in production. No reported failures added barrier Salt creep crush casing Total vent flow data % in 34,000 wells 0.1% in older wells worst case. 0.02% all wells. 0.02% for older era wells 0.004% for newer wells No failure reported No failure data or pollution reports Data Sources 0.06% for all wells Detailed not Kell, 2011 available 0.005% to 0.01% for producers 0.03% to 0.07% for injectors No well associated pollution 5.5% Unknown None reported No separation data 4.6% taken as available worst case. No data mostly gas escape Kell, 2011 TGPC data 1997 to 2011 Kell, 2011 Clegg, 1971 Watson & Bachu, 2009

13 Table 10 - Distribution of Barrier Failures and Well Integrity Failures Showing Improvement by Era (all land wells). > 650,000 wells Area / Number of Barrier Fail Freq. Well Integrity Leaks to Number of construction Range (containment) Failure Range GW by Wells failures (containment lost) sampling Ohio / 64,830 TX / 253,090 TX / 16,000 horizontal multi-frac MN / 671 Alberta / 316, fail initial cement test. 39 failed in production. 10 fail initial cement test. 56 failed in production. No reported failures added barrier Salt creep crush casing Total vent flow data % in 34,000 wells 0.1% in older wells worst case. 0.02% all wells. 0.02% for older era wells 0.004% for newer wells No failure reported No failure data or pollution reports Data Sources 0.06% for all wells Detailed not Kell, 2011 available 0.005% to 0.01% for producers 0.03% to 0.07% for injectors No well associated GW pollution 5.5% Unknown None reported No separation data 4.6% taken as available worst case. No data mostly gas escape Kell, 2011 TGPC data 1997 to 2011 Kell, 2011 Clegg, 1971 Watson & Bachu, 2009

14 BARRIER AND INTEGRITY FAILURES: >330,000 US WELLS Barrier or Integrity Fail Texas Horizontal Integrity Failure Texas Newer Wells Barrier Failure Texas Newer Wells Integrity Failure Texas Old Wells Barrier Failure Texas Old Wells Integrity Failure Ohio Newer Wells Barrier Failure Ohio Newer Wells Integrity Failure Ohio Old Wells Barrier Failure Ohio Old Wells Older well data often skewed by lack of barrier & integrity differentiation % 0.020% 0.040% 0.060% 0.080% 0.100% Things That Keep Real Integrity Failures Very Low 1. Pressure inside the wells is lower than outside in hydrostatic of water table. 2. Modern wells are built with multiple barriers. 3. Cement reinforces and protects the casing. 4. Regulations are tighter now than 3 years ago. 5. Multi-Fractured horizontal wells replace 5 to 10 vertical wells in shale. Less pollution potential with fewer water table penetrations. What Proves it? rankings of proven groundwater pollutants.

15 Area Proven Another Way - % of Produced Fluids Leaked From Production Leaks and Spills US Gulf of Mexico US Gulf of Mexico Number of Wells 11,498 (3542 active Type of Wells Platform based wells 4,099 Shoe test failures required repair Barrier Failure Freq. Range (w/ containment) 30% overall first annulus SCP 50% of cases. 90% of strings w/ SCP have less than 1000 psi. 10% are more serious form of SCP (Wojtanowicz, 2012) 12% to 18% require cement repair to continue drilling Norway 406 offshore 18% 0 GOM /Trinidad Matagorda Island 623 All Sources in SPE , Barrier vs. Well Failure, King, 2013 Integrity Failure (leak path in or out) Slide % to 0.05% of wells leaked % to % based on produced oil spilled 1980 thru (all repaired before resuming drilling) 2,120 Sand Control 0.5 to 1% 0% subterranean ~0.0001% via surface erosion potential 17 Compaction failures; casing shear & sand fail Sumatera 175 without maintenance 80% to 100% - the high number is due to high pressure and formation compaction. 43% 1 to 4% Wells routinely shut-in and repaired prior to restart.

16 Area Proven Another Way - % of Produced Fluids Leaked From Production Leaks and Spills US Gulf of Mexico US Gulf of Mexico Number of Wells 11,498 (3542 active Type of Wells Platform based wells 4,099 Shoe test failures required repair Barrier Failure Freq. Range (w/ containment) 30% overall first annulus SCP 50% of cases. 90% of strings w/ SCP have less than 1000 psi. 10% are more serious form of SCP (Wojtanowicz, 2012) 12% to 18% require cement repair to continue drilling Norway 406 offshore 18% 0 GOM /Trinidad Matagorda Island 623 All Sources in SPE , Barrier vs. Well Failure, King, 2013 Integrity Failure (leak path in or out) Slide % to 0.05% of wells leaked % to % based on produced oil spilled 1980 thru (all repaired before resuming drilling) 2,120 Sand Control 0.5 to 1% 0% subterranean ~0.0001% via surface erosion potential 17 Compaction failures; casing shear & sand fail Sumatera 175 without maintenance 80% to 100% - the high number is due to high pressure and formation compaction. 43% 1 to 4% Wells routinely shut-in and repaired prior to restart.

17 So What are Actual Groundwater Pollutants? Slide 17 UST Gas & Diesel Septic Systems Landfills Spills Fertilizer Large Industrial Facilities Hazardous Waste Sites Animal Feedlots Pesticides Surface Impoundments Storage Tanks surface Urban Runoff Salt Water Intrusion Mine Drainage Agriculture Chem. Facilities Pipelines & Sewer Shallow Inj. Wells Salt Storage & Road Salting Land application of Waste Irrigation Practices EPA, 2000 SPE , Barrier vs. Well Failure, King

18 What are Groundwater Pollutants Today & Where do Oil & Gas Wells Fit in this Picture? Slide 18 Used Texas as a Study Case. Over a million penetrations through the 29 major & minor aquifers in Texas. Texas is #2 in total Groundwater withdrawals with ~ 80% going to Agriculture & Municipalities. If the water was really polluted by O&G wells, we d see it quickly in Municipal & Ag. SPE , Barrier vs. Well Failure, King

19 Last 12 years of Pollution Reports in Texas Top 20 Listed - TCEQ & TGPC Database Slide 19 Number of New Reports Per Year SPE , Barrier vs. Well Failure, King

20 Allocation of Texas TCEQ Pollution Claim Frequency Producing Wells are less than 1% of total for most years. SPE , Barrier vs. Well Failure, King

21 Gas migration >>200+ yrs. old, highly regional, many causes, 1000 s of seeps. SPE , Barrier vs. Well Failure, King

22 Methane Seepage from Soils Oil & Gas Seeps are indicators of oil & gas beneath the surface Many natural seep flows diminished as wells were drilled & produced. SPE , Barrier vs. Well Failure, King

23 Comparing Spills and Seeps Lakeview Gusher, CA Onshore, 1910 Santa Barbara Blowout, CA, 1969 Tanker Grounding, MA, 1976 Tanker Grounding, AK, 1989 Tanker Grounding, TX, 1990 Sabotage, Kuwait, 1992 Tanker Grounding, LA, 2000 Pipelines Ruptured by Hurricanes, Barge Collision, LA, 2008 Tanker Collision, TX, 2010 Pipeline Corrosion, MI, 2010 Macondo Blowout, GOM, 2010 Natural Seeps, Coal Point, CA, Yearly Natural Seeps, GOM, Yearly Single Estimate High Value Range 1,000 10,000 BARRELS 100,000 1,000,000 10,000,000

24 WATER PRODUCED AFTER A FRACTURE TREATMENT

25 EFFECTIVE REGULATIONS? - OR - DEVELOPMENT PREVENTERS? Effective & Development Encouragers Standards reduce real risks Consistent operations by all operators Encourages produced water reuse instead of disposal. Provides a level field for well development. Requires reporting of all fracs to public available site ( Development Preventers Prevents Any Development Every well a target for environmental impact and endless public challenge. Creates unachievable goals or excessive cost to comply. Let s just wait until it s perfect.. Continues to make companies prove a negative regardless of a individual performance record.

26 QUESTIONS I WOULD ASK. 1. Will all chemicals, water volumes and water source be listed on the 2. Is minimum distance between designed top of fracture & deepest fresh water greater than 1500 ft? 3. Will green completions or the regional specific equivalent, or better alternative, be used? 4. Can multi-well pads be used in the development phase instead of single well pads? 5. Is it possible to limit busy development activity to a certain season, time or remote area?

27 QUESTIONS I WOULD ASK. 6. Will site investigation find abandoned wells, seeps, mines or potential problem area near the well site and path? 7. Is methane migration a potential problem? (Gas migration is strongly linked to specific geologic regions.) 8. Will fresh water wells be tested, where permitted, within at least ¼ mile of the proposed well site prior to drilling? 9. How can truck traffic and road damage be reduced? 10. How is frac water supply sustainability approached? 11. How will recycled water for fracturing be stored? 12. Does the company have a list of chemicals they will not use? (may not share them, but can describe types of chemicals they will not pump).

28 WHAT ARE SOME OF THE REAL RISKS? Transport spills same frequency as other chemical transport options (rail, barge, truck). Wrecks and road damage. Fracturing old wells w/ questionable casing & cement. Improperly sited salt water disposal wells. Fracturing in shallow wells (<2000 ft). Not using modern technology. Ignoring well monitoring and maintenance. General ignorance on both sides of the issue.

29 A FEW CONCLUSIONS Slide Risk of GW pollution from producing well is low and Barrier failure rates and well failure rates vary widely. 2. Failure of wells of a specific time era are artifacts of that era; not reflective of wells completed today. 3. Methane gas migration from deep drilling doesn t appear connected to current O&G production. 4. Improperly plugged old wells & water wells may be conduits for methane migration. 5. Groundwater composition changes generally unrelated to oil & gas ops. See SPE , and for more info