EPRI Global Climate Change Research Seminar Natural Gas Supply Francis O Sullivan, Ph.D. May 25 th, 2011 1
A review of 2009 U.S. primary energy consumption by source and sector reveals the broad systemic importance of natural gas Supply Sources Demand Sectors 94.6 Quads 35.3 Petroleum 72% 22% 1% 5% 3% 94% 3% 27.0 Transport 23.4 Natural Gas 19.7 Coal 7.7 Renewables 7% <1% 93% 26% 12% 9% 53% 3% 32% 30% 35% 40% 76% 18% 48% 11% 41% 7% 11% 1% 17% 7% 1% 22% 18.8 10.6 38.3 Industrial Residential & Commercial Electric Power Nuclear 8.3 100% Source: EIA 2
The global natural gas resource 3
There is a lot of gas in the world, ~16,000 Tcf but these resources are highly concentrated Breakdown of the total global remaining recoverable gas resources by EPPA region Tcf of Gas Middle East Russia United States Africa Central Asia Canada Latin America W. Europe Dynamic Asia Brazil Rest of E. Asia Oceania Global gas consumption in 2009 was ~107 Tcf Proved Reserves Reserve Growth Unconventional Resources** Yet-to-Find Resource (Mean) P90 YTF P10 YTF Working Draft - Last Modified 5/4/2010 6:23:07 PM Printed China Mexico India 0 1,000 2,000 3,000 4,000 5,000 6,000 Source: MIT Gas Supply Team analysis 4
Much of this resource can be developed at low cost but long distance transportation costs will significantly increase its delivered cost Global breakeven gas price $/MMBtu* 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 Example LNG value chain costs incurred during gas delivery $/MMBtu** Liquefaction $2.15 Shipping $1.25 Regasification $0.70 Total $4.10 P90 Mean P10 Volumetric uncertainty around mean of 16,200 Tcf 2.00 0.00 P90 12,500 0 4,000 8,000 12,000 16,000 20,000 P10 20,600 * Cost curves based on 2007 cost bases. North America cost represent wellhead breakeven costs. All curves for regions outside North America represent breakeven costs at export point. Cost curves calculated using 10% real discount rate ** Assumes two 4MMT LNG trains with ~6,000 mile one-way delivery run Source: MIT Gas Supply Team analysis, ICF Hydrocarbon Supply Model, Jensen and Associates 5
Early analysis suggested that there may be very extensive global unconventional resources Breakdown of global unconventional GIIP by region and type Tcf of gas 9,000 8,000 7,000 CBM Shale Tight 6,000 5,000 4,000 3,000 2,000 1,000 North America Latin America Western Europe Central and Eastern Europe Former Soviet Union Middle East and North Africa Sub- Saharan Africa Central Asia and China Pacific (OECD) Other Asia Pacific South Asia Source: An assessment of world hydrocarbon resources, H-H Rogner, Annu. Rev. Energy Environ., 1997 6
A recent global assessment of shale gas suggests at least 6,000 Tcf of recoverable resources, with over 1,200 Tcf located in China Breakdown of global recoverable shale gas resources by region Tcf of gas 1500 Study only assessed 31 countries Future work expected to increase the resource estimate substantially 1200 900 600 1404 1042 1225 300 396 624 0 Australia Asia Africa Europe South America Top two shale gas resource holders by region China 1,275 Tcf India 63 Tcf South Africa 485 Tcf Libya 290 Tcf Poland 187 Tcf France 180 Tcf Argentina 774 Tcf Brazil 226 Tcf Source: World Shale Gas Resources: An Initial Assessment of 14 Regions Outside the United States, ARI 2011 7
Gas dynamics in the United States 8
The past 5 years have seen a dramatic increase in both proved reserves and more importantly in the technically recoverable resource All due to shale Illustration of US gas production, reserve and resource dynamics from 90 to 10 Tcf of gas Illustration of the change in recoverable resource assessment by gas type over the past decade 1 Tcf of gas 4000 3500 3000 2500 Resources Proved Reserves Cumulative production +80% 2500 2000 1500 Shale Resources Non-Shale Resources Proved Reserves 616 827 2000 1500 1000 500 0 1000 500 0 1218 1290 1219 1219 184 204 245 273 NPC 2003 PGC 2006 PGC 2008 EIA 2010 1. EIA 2010 assessment based on 2008 PGC assessment with updated estimates of technically recoverable shale gas volumes Source: NPC data, PGC data, EIA data 9
The emergence of shale gas as a recoverable resource is illustrated in the production dynamics Shale gas is driving U.S. production growth Breakdown of US gas production by type 1 Tcf of gas 25 Comparison of production by type 00 vs. 09 1 Percent of total production 20.5 Tcf 22.1 Tcf 7% 8% 20 15 10 5 22% 16% 55% 14% 25% 11% 41% 0 2000 2002 2004 2006 2008 2010 2000 2009 In 2000, 14.6 Tcf of conventional gas was produced, representing 71% of total US marketable production By 2009, conventional production fell to 11.6 Tcf, or just over 50% of total production In the same period shale gas output grew from almost nothing to over 3.5 Tcf, and is now the only production segment that is growing Coal Bed Methane Shale Tight Associated Conventional 1. United States production figures represent marketable production, and so exclude gas produced in Alaska, which is subsequently reinjected Source: HPDI commercial production database 10
United States shale rock deposits include some very large deposits near major gas consuming centers in the Northeast Map of United States shale deposits Source: EIA 11
The focus on shale gas has led to large increases in mean resource estimates; however, these mean estimates are accompanied by wide error bars Comparison of mean estimates of shale gas resources in the United States Tcf of Gas Breakdown of the PGC 2009 shale gas resource estimates by major U.S. shale play* Tcf of Gas Min Mean Max Recent focus on assessing the shale gas potential in the U.S. has resulted in dramatic increases in resource estimates ICF 09 PGC 09 ICF 08 Fort Worth Basin: Barnett Shale Arkoma Basin: Fayetteville/Woodford E. TX & LA Basin: Haynesville Shale Appalachian Basin: Marcellus/Ohio/Utica Shale 25 59 100 70 110 146 60 112 182 92 227 549 EIA 07 Anadarko/Permian Basins: Barnett/Woodford Shales 3 6 16 NPC 03 2002 2003 2004 2005 2006 2007 2008 2009 2010 Year Other Basins: 51 100 224 Total Mean Estimate: 616 301** 1217** * Mean volumes represent the most likely estimates reported by the PGC and can be aggregated by arithmetic addition to yield an aggregated mean estimate of shale gas resources in the United States. The per basin min and max numbers reported here assume perfect statistical correlation within basins ** US min and max totals are for illustrative purposes only, and are calculated by direct addition of volumes, not statistical aggregation Source: Various commercial and institutional resource assessments 12
The US gas supply curves reveal large volumes of relatively cheap gas Remarkably, shale gas is the most important source of low cost resource United States breakeven gas price $/MMBtu* 40.00 35.00 30.00 25.00 20.00 Breakdown of United States breakeven gas price by resource type $/MMBtu* 40.00 35.00 30.00 25.00 20.00 15.00 15.00 10.00 5.00 0.00 P90 Mean P10 10.00 5.00 0.00 Conventional Shale Tight CBM 0 500 1,000 1,500 2,000 2,500 3,000 0 200 400 600 800 1,000 * Cost curves calculated using 2007 cost bases. U.S. costs represent wellhead breakeven costs. Cost curves calculated assuming 10% real discount rate Source: MIT Gas Supply Team analysis, ICF Hydrocarbon Supply Model, Data strictly for illustrative purposes only 13
IP rate variability is a major issue with shales In the Barnett, IP rates vary by 3X between P20 and P80 performance, while other shales display even greater variability 2010 Probability distribution of initial production rates for Barnett wells Mcf per day (30 day average) 2010 Cumulative probability distribution of initial production rates for Barnett wells Mcf per day (30 day average) 3,030 Mcf/day < P20 Probability density Cumulative probability P50-P80 P20-P50 > P80 1,020 Mcf/day Initial production rate Mcf/day Initial production rate Mcf/day Source: MIT Gas Supply Team analysis, HPDI commercial production database 14
Shale gas technological and environmental considerations 15
Shale is extremely tight, and so to produce gas from it economically, it is necessary to create as much reservoir contact as possible horizontal wells help to achieve this Shales tend to be thin, on the order of 100 s of feet in depth, but they are areally extensive, often extending over 1000 s of acres Consider a shale 7,500 feet underground: A vertical well will only provide ~100 of contact, while a horizontal well could provide 5000 or more of reservoir contact Horizontal wells also enable pad drilling where multiple horizontal well bores are drilled from one surface location Much more attractive economics and much less surface disturbance Source: MIT Gas Supply Team 16
Hydraulic fracturing has evolved rapidly in recent years A move to Open Hole Multi Stage fracking has enabled more frac stages in less time Typical frac site Pumpers, water, sand and additive tankers along with control vehicles Wellhead rigged for fracing This is the goat head Elements required to carry out hydraulic fracturing Horse power 8-10 2,500 HP pumpers required for typical frac job Pumpers must be pressure rated to 15,000 psi Each pumper is typically rated to 15 bblpm at operating pressures 5 M gallons of water required for a typical 10 stage frac job 2000 MT of sand required for a typical 10 stage frac job Cemented liner, plug and perforate multistage fracturing The original approach Original approach to multistage fracturing in shale plays Time consuming when number of stages increases as it requires multiple wireline trips Perforations can damage formation and inhibit well productivity Open hole multistage hydraulic fracturing The state-of-the-art State-of-the-art fracturing techniques for gas shales Very fast No need to open well for entire duration of multistage job No formation damage Maximizes well productivity Source: MIT Gas Supply Team 17
The rapid expansion of drilling and fracking activities in shale plays has led to some significant environmental concerns Some key environmental concerns include: Water: Freshwater aquifers could become contaminated by fluids used for fracking Surface water sources could become contaminated by fluids used for fracking Post use treatment and disposal of fracking fluids could be hindered by a lack of appropriate facilities Sourcing adequate volumes of water for fracking operations could strain overall water availability at a local level Surface: Drilling activities may lead to well blowouts which could endanger both life and property Intensive shale drilling will result in significant surface disturbance and habitat interference Drilling and fracking activities lead to significantly increased traffic in areas lacking appropriate road infrastructure Drilling and fracking operations will result in significant noise and air pollution Source: MIT Gas Supply Team 18
Shale gas production is facing a wide range of surface issues relating to water sourcing, transport and disposal Some Challenges: Shale well drilling and completion can requires 5 million of more gallons of water Large volumes of injected frac water return to the surface 20-70% flowback depending on situation Flowback water can be heavily polluted and needs to be treated Recovered water needs to be completely contained on site and properly disposed of to avoid pollution Permitted water treatment facilities not capable of handling frac fluid and drilling waste Underground injection capacity not adequate in some plays PA in particular High transportation costs to haul frac water to treatment facilities Source: MIT Gas Supply Team Illustration of shale well site and fluid containment pond There is a strong economic incentive for operators to find solutions: On-site water treatment facilities and closed loop water reuse systems are being developed More efficient frac procedures are being deployed to reduce fluid injection volumes Operators are making more use of centralized production operations (pad well development) to reduce the need for water hauling 19
The growth in shale gas production and the size of frac jobs has meant that annual shale gas related water demand is now at 20 billion gallons per year Annual well additions in each of the major U.S. shale plays from 2005 2010 # of wells Annual water requirements for drilling and fracking in the major U.S. shale plays from 2005 2010 Millions of gallons of water 5000 Marcellus 20000 Marcellus 4500 4000 Woodford Haynesville Fayetteville 18000 16000 Woodford Haynesville Fayetteville 3500 Barnett 14000 Barnett 3000 12000 2500 10000 2000 8000 1500 6000 1000 4000 500 2000 0 2005 2006 2007 2008 2009 2010 0 2005 2006 2007 2008 2009 2010 Source: MIT Gas Supply Team, HPDI commercial production database 20
However, it appears that water consumption for shale gas activities still represents a small portion of the total water usage in the major shale plays 2008 water use by type in the major shale gas plays* Percent of total, Billions of gallons per year 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 466 1,340 88 3,570 Barnett Fayetteville Haynesville Marcellus Shale gas Livestock Irrigation Industrial/Mining Public supply * Modern Shale Gas: A Primer, United States Department of Energy, April 2009 Source: MIT Gas Supply Team 21
Comments & Questions? 22
Supplemental materials 23
Along with intra-play variation, there is huge inter-play variation among the big shale plays The Haynesville is very different to the Barnett Difference between a typical Haynesville and Branett decline curve over the first 10 years Mcf/day 10,000 Haynesville wells have 30-day average IP of ~10,000 Mcf/day, compared to ~1,800 in 8,000 the Barnett Haynesville wells experience 80-85% first 6,000 year decline Haynesville wells produce up to 25% of 4,000 EUR in year 1 2,000 0 0 24 48 72 96 120 Decline onset sensitivity to output in Haynesville assuming EUR of 112 Tcf Bcf/day 12 10 6 5 4 3 2 1 0 8 6 4 2 0 10 Bcf/day 5 Bcf/day 2010 2020 2030 2040 2050 2060 Decline in output upon cessation of drilling in the Haynesville shale Bcf/day Haynesville output is very sensitive to drilling activity 2010 2020 2030 2040 2050 2060 Source: MIT Gas Supply Team analysis, HPDI commercial production database 24
The analysis of well performance across shale plays is illustrating the variation in per-well economics that exists between and within the plays Cumulative Probability 1 0.8 0.6 0.4 0.2 Cumulative probability of peak production rates of Fayetteville wells drilled in 2009 Mcf/day (30-day average) P80 P50 P20 Overall Core 0 1,000 2,000 3,000 4,000 5,000 6,000 Variation in breakeven gas price for the 09 Fayetteville shale well population $/MMBtu Overall Core P20 P50 P80 $3.74 $5.37 $8.77 $3.41 $4.65 $6.47 Cumulative Probability 1 0.8 0.6 0.4 0.2 Cumulative probability of peak production rates of Woodford wells drilled in 2009 Mcf/day (30-day average) P80 P50 P20 Overall Core 0 2,000 4,000 6,000 8,000 10,000 12,000 Variation in breakeven gas price for the 09 Woodford shale well population $/MMBtu P20 P50 P80 Overall Core $4.71 $8.04 $20.12 $3.18 $3.73 $6.60 Source: MIT Gas Supply analysis HPDI production database 25
Concern exists about direct contamination of fresh water aquifers due to fluid migration from a frac zone analysis would suggest this is unlikely once wells were correctly completed Illustration of the scale of separation between freshwater aquifers and shale deposits Depths to freshwater aquifers and producing layers in major shale plays* Basin Depth to shale (ft) Depth to aquifer (ft) 100 s ft to bottom of aquifer Barnett 6,500 8,500 1,200 Fayetteville 1450 6,700 500 1000 s ft to shale layer Marcellus Woodford 4,000 8,500 6,000 11,000 850 400 Haynesville 10,500 13,500 400 Shale gas resources are separated from freshwater aquifers by 1,000s of feet of alternating layers of siltstones, shales, sandstones * Modern Shale Gas: A Primer, United States Department of Energy, April 2009 Source: MIT Gas Supply Team 26
There is a strong operational incentive to eliminate any fluid leakage since containment of the fluids in the shale is critical to the success of the frac job Illustration of multiple well casing used to isolate produced fluids from the aquifer in a gas well Extensive regulation exists at State level regarding the protection of groundwater during oil and gas operations Current well design requirements demand extensive hydraulic isolation At depths coincident with the aquifer, groundwater will be separated from produced gas and fluids by at lease three layers of steel and three layers of cement The integrity of the isolation measures is tested Probabilistic analysis for injection wells suggests the likelihood of a well leaking given properly installed casing is less than 1 in 1 million* Injection wells are consistently operated at high pressure, while production well pressure declines, further reducing the probability of leakage * Michie & Associates. 1988. Oil and Gas Water Injection Well Corrosion. Prepared for the American Petroleum Institute.1988 Source: MIT gas supply team 27
Although frac fluids are almost entirely comprised of water and sand, a range of other chemicals are also present Illustration of the composition of a typical fracing fluid* % by volume 99.51% Water & Sand 0.49% Chemical Additives 0.6% 0.5% 0.4% 0.3% 0.2% 0.1% 0 KCl Gelling Agent Biocide Scale Inhibitor PH Adjustment Breaker Crosslinker Iron Control Acid Friction Reducer Corrosion Inhibitor Frac fluid composition varies from play to play due to the underlying geology; however, the vast majority of fluids are >98% sand and water Legitimate concerns have been raised about what additives are used in frac fluids, which the operators need to address in a more transparent manner * Modern Shale Gas: A Primer, United States Department of Energy, April 2009 Source: MIT gas supply team 28