resource and production potential. It integrates engineering, engineering, It integrates production potential. and resource (BEG) Haynesville’s of Geology the of Economic Bureau the by astudy summarizes article This conditions. economic right the under promising remains play the this, Despite . northwest and Texas east in Haynesville slowed development have of the prices gas natural Low Austin of TexasUniversity Bureau of Economic Geology W. Tinker Scott Potter Eric Tad Patzek Grote Carl Horvath Susan Smye Katie Gülen Gürcan Male Frank Ikonnikova Svetlana Browning John production recovery before decline final Haynesville gradual forecasts Study TYPICAL HAYNESVILLE PAY ZONECROSSSECTION* *GR =gammaraylog,NPHIneutronporosity, DPHI=densityporosity. Gray shadedareasonDPHIblacklineareporosityvalues.Payzoneishighlightedinblue. 0 . -0.1 0.3 300 0 GR

10,600 10,500 10,400 10,300 10,200 10,100 10,000 Depth, ft . -0.1 0.3

NPHI DPHI

300 0 GR

11,800 11,700 11,600 11,500 11,400 11,300 12,000 Depth, ft . -0.1 0.3 -0.1 0.3 NPHI DPHI TECHNOLOGY PRODUCTION DRILLING & 300 0 GR

11,900 11,800 11,700 11,600 11,500 11,400 11,300 Depth, ft . -0.1 0.3 -0.1 0.3 NPHI DPHI contributor to US natural gas production for at least 30 years. years. 30 production for gas at least to US natural contributor a significant to be continue will formation the assumption, to 1.7 decline 2045. by bcfd permanent its before starting 2020s, early the slowly recover in to 5bcfd will Production a 2012 2015. in of4 bcfd aboutto 6 bcfd roughly plateau 2064. through producing 2045 and through drilled to be wells, new and productionexisting by Haynesville of tcf cumulative 46 pricing and other conservative parameters, we estimate we estimate parameters, conservative other and pricing prices. gas natural on dependent are of these All production peak. play’s the and activity, drilling future recovery, potential ultimate estimated total-field resources, gas recoverable technically parameters. economic and technical several on based production scenarios and drilling of forecasting capable model asimulation geology, into economics and Even using a constant $4/MMbtu Henry Hub price Hub price Henry $4/MMbtu aconstant using Even from declined production has annual The Haynesville’s In a base-case scenario using $4/MMbtu Henry Hub Henry $4/MMbtu using scenario abase-case In for remaining the cases base determines The article 300 0 GR

12,400 12,300 12,200 12,100 12,000 11,900 Depth, ft . -0.1 0.3 -0.1 0.3

NPHI DPHI 300 0 GR

12,200 12,100 12,000 11,900 11,800 11,700 Depth, ft . -0.1 0.3 -0.1 0.3 NPHI DPHI FIG. 1

151207OGJdti-01 TECHNOLOGY

This article identifies areas where HAYNESVILLE SHALE NET POROSITY, THICKNESS* future drilling is likely to occur, when Fig. 2 and under what economic conditions Harrison drilling will occur, what the drilling and production profile will look like, Gregg Caddo and the economic reserve additions Bossier that will result. Bienville

Rusk Study parameters, methods Panola The study underlying this article used Red River production data from all individual De Soto Porosity thickness, ft Haynesville wells drilled 2008-12, starting with the production history 2 4 6 8 10 12 14 16 of all wells and then determining what remains to be drilled under Shelby Natchitoches various economic, geologic, and Nacogdoches Louisiana technologic scenarios. The result is a -02 San Augustine comprehensive view of the field. The study assesses production Angelina potential in six geographic tiers and 0 20 miles Sabine Texas estimates future production scenarios 151207OGJdti * Corrected for clay content and calibrated to gas- lled core porosity. Porosity is net from density logs. according to these tiers. Source: Bureau of Economic Geology Well economics vary across the basin because of productivity and cost differences caused by geology HAYNESVILLE ESTIMATED ORIGINAL FREE GAS IN PLACE and other factors. The article accounts Fig. 3 for these variances, as well as for distributions around natural gas price, Harrison drilling cost, economic limit of each Gregg well, advances in technology, and Caddo Bossier many other geologic, engineering, and Bienville economic parameters. Including these variables allows determination of how Panola much gas can be extracted from future Rusk Red River wells under different economic and Bcf/sq mile block De Soto technical conditions. 82 93 The study includes a method of 36 67 101 110 122 139 160 estimating ultimate production for each well based on the physics of Shelby Natchitoches the system, rather than using just Nacogdoches the mathematical decline curve. This Louisiana method has successfully predicted San Augustine shale-well production declines in Angelina other basins and was used by BEG in Sabine 151207OGJdti-03 20 miles Texas previous studies of the Barnett and 0 Fayetteville (OGJ, Aug. 5, 2013; Source: Bureau of Economic Geology Jan. 6, 2014).1 To tier productivity and analyze decline curves, the study looked at all 2,527 wells drilled actual 2012 drilling and production results. through 2012 and analyzed declines for 2,131 wells with at The Haynesville play is particularly sensitive to price least 12 months’ production history. variations. Breakeven costs are near or above Henry Hub gas The production outlook model covers field development prices of 2011-12 because of higher well costs in this deep, from 2013-45, then extends production through 2064, high-pressure reservoir. allowing the 2045 wells to deplete. While wells developed in Many high-quality well sites have been drilled only to 2012 provide incomplete results for a stable decline analysis, the extent necessary to meet acreage commitments. These they allow benchmarking predictions for that year against locations will become economically attractive when a TECHNOLOGY

HAYNESVILLE WELL DATA, SIMULATION PREDICTION chosen to maximize spatial coverage Fig. 4 of the field. Major stratigraphic tops 1.0 were picked in all wells. The pay zone was selected primarily on the basis of 0.8 gamma ray values, which are higher than in underlying carbonate units, and NPhi and DPhi log curves. 0.6 A gas effect appeared in the porosity log responses (Fig. 1), shown 0.4 by higher apparent DPhi (as the bulk Recovery factor Simulation density is reduced by the presence of Actual well RFs 0.2 gas) and lower apparent NPhi (as the hydrogen content is lower for gas than for oil or water). 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 151207OGJdti-04 Structure, net pay-zone thickness Square root scaled time (H), and density porosity (DPhi) maps were produced, and Phi and H maps were combined to create a net porosity- thickness (Phi-H) map (Fig. 2).3

HAYNESVILLE 20-YEAR GAS PRODUCTIVITY Fig. 5 Porosity-thickness mapping initial- ly showed an area of high Phi-H in the Harrison northern part of the play that did not correspond with good well productiv- Gregg Caddo ity. The poor performance of wells in Bossier this area is related to lithology. Higher

Bienville clay content from ancestral river dilu- Rusk Panola tion to the north affects production, making more dif- Red River ficult.4 Proppant becomes embedded DeDe SotoSoto Bcf recovery/4,800-ft well* more easily and connectivity is lost more readily than in harder, more cal- 2.0 3.0 4.0 5.5 7.0 15.0 cite-rich rock.5 High clay content is reflected in the Shelby Natchitoches Nacogdoches logs as a greater separation between DPhi and NPhi (Fig. 1).6 Clay-volume Louisiana calculations based on NPhi-DPhi San Augustine separation have been used in the Angelina Haynesville shale.7 DPhi log curves 151207OGJdti-05 0 20 miles Sabine Texas modulated on the basis of reservoir

*Each sq mile colored based on estimated average productivity of 4,800-ft horizontal well. quality produced a Phi-H map that Source: Bureau of Economic Geology accurately represents areas of good porosity and thickness. Separation sustained increase in gas prices or a decline in drilling and between NPhi and DPhi log curves completion costs leads to resumed development of the field. determined reservoir quality. Net pay-zone calculations excluded areas too clay-rich Geologic characterization (>5% separation between NPhi and DPhi) to correspond to The BEG study encompasses the extent of all previous drilling good production. The resultant net porosity-thickness map within the known geologic boundaries of the field. A total of provided a good correlation to well productivity, a key driver 5,212 sq miles was included, though only 2,068 sq miles had in predicting future production from undeveloped areas of been tested by drilling through 2011. the field. Researchers performed a log-based assessment of key The study divided the play into square-mile grids, or blocks, geological parameters influencing production.2 Digitized and calculated original free gas in place (OGIPfree), excluding logs were deep enough to reach the base of the Haynesville, adsorbed gas, for the entire play on a block-by-block basis, including gamma ray, density porosity (DPhi), and neutron using a conventional volumetric approach (Fig. 3). porosity (NPhi) logs. When calculating OGIPfree, density-log porosity was The study’s data set consisted of logs from 115 wells adjusted to equivalent gas-filled core porosity by applying a TECHNOLOGY used it to forecast production. it to forecast used then and properties, reservoir from estimated interference time-to- their had interference interfracture yet experienced (Fig. 4). not had interference that Wells to interfracture time the by divided of root time square the production against plotted cumulative volume, then reservoir stimulated the within of gas value fitted the by production well for each decline. predicted the in resulting pattern, well-fracture the within reached are conditions interfracture-boundary until log time production versus of cumulative increase to production. contributing from gas absorbed prevented pressures reservoir High production. affected interference interfracture followed, as decline exponential An completions. and properties on reservoir depending lifes, of well 1to 2years first over the time total of root square to the proportional inversely was decline on linear-transient flow in the reservoir. the in flow on linear-transient 2012. through drilled wells of 10.3 EUR an for 2,527 tcf predicted the analysis Decline life. 25-year maximum assumed an and area; drainage well’s the within interference interfracture from deterioration of late-life effects the declines; base-well included study OGIP Total reservoirs. of similar typical be to assumed were properties Gas depth. of reservoir afunction were lations for this step. for this basis the provided sources additional from coredata and shales Barnett and Fayetteville the in BEG’s experience of factor log =0.60field-wide ×core. study normalized EURs as if all wells had been drilled to drilled been had wells all if as EURs normalized study The length. with decreasing slightly length of lateral unit per EUR incremental average but the length, well lateral with increasing wells of Haynesville EUR BEG found the Reservoir-quality tiers (EURs). recoveries ultimate expected individual their determining 2012, through 2,527 drilled all wells of decline the analyzed The study Production-decline analysis free as another important driver of OGIP driver important another as ofend 2011. pressure BEG identified the by one well at by least penetrated were that blocks 227 underlying tcf the south from 10,000 ft in the north. north. the 10,000 from in south ft the to 14,000 deepens in ft formation the as increases Pressure about 0.95 psi/ft. of agradient with pressure, high mally . The Haynesville exhibits abnor exhibits . TheHaynesville In addition to this theoretical model, BEG divided BEG model, divided theoretical to this addition In astraight-line solution yields flow linear A theoretical The study used a well-production decline method based based method decline awell-production used The study Temperature and pressure calcu free was estimated at 489 tcf, with tcf, with at 489 estimated was 8

9 Key input variables in the the in variables input Key - - - 1 Per-well production DRILLED, POTENTIALWELLS/GASPRICES Wells, no 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0 ir1Te ir3Te ir5Tier 6 Tier 5 Tier 4 Tier 3 Tier 2 Tier 1 Cumulative drillingthrough2012 Potential, partlydrilledblocks Potential, undrilledblocks relatively low productivity. relatively have thins, reservoir the where flanks, The field block. same the in other to each next exist wells poorer-performing and wells better-performing cases, some In blocks. performing poorer- with interspersed blocks better-performing with heterogeneity, reservoir considerable shows map The tier 5). (Fig. map productivity-tier afull-field play, of the yielding features geological the following and interpolation mathematical using tiers assigned were well, existing any by penetrated not blocks, Theundrilled tiers. productivity six in blocks the Actual drainage areas are not ideal rectangles and the the and rectangles not ideal are areas drainage Actual (Bg). factor expansion depth. of well function using: calculated are volumetrics Reservoir well. each by drained reservoir volume of the the quantify to volumetrics reservoir with values We EUR combined Reservoir drainage block. the penetrating segments well on the based ft, EUR/ average aweighted assigned block was 1-sq mile Each wells. for all surveys directional using paths, well-drilling time. at the practices drilling common reflecting ft, oflateral 4,800 a uniform Tiering reveals areas of higher and lower productivity. and of higher areas reveals Tiering ranking allowed values productivity average The resulting A rectangle represents the volume-drainage represents the A rectangle area. • • • • along mapped were values EUR/ft Length-normalized

Ty Re Pa Ca y-zone thickness. y-zone thickness. pical gas properties, from which we derive the gas- the derive we which from properties, gas pical servoir pressure and temperature for each well as a as well for each temperature and pressure servoir librated gas-filled porosity values. values. porosity gas-filled librated Henry Hubpricing,$/MMbtu 2013-45 potentialdrilling; 10 6 4 3 TECHNOLOGY 10

Fig. 6

151207OGJdti-06 TECHNOLOGY

estimated the amount of drained and HAYNESVILLE TYPE CURVE, ESTIMATED ULTIMATE RECOVERY* Fig. 7 7 undrained acreage for each 1-sq mile EUR, bcf block of the reservoir. Assuming that 6 Tier 1- 7.9 the acreage left undrained by the wells Tier 2- 5.9 in each block was known, we then 5 Tier 3- 4.5 Tier 4- 3.3 created an inventory of future feasible 4 Tier 5- 2.4 drilling sites based on expected RFs, Tier 6- 1.4 EUR, and estimated OGIPfree by tier

MMcfd 3 for every location (Fig. 6). This study estimates a remaining technically 2 recoverable resource (TRR) of 177 Tcf. In this projection, the higher-productivity 1 tiers are more developed and the lower- 151207OGJdti-07 0 productivity tiers remain uneconomic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 at almost any foreseeable gas price. Years *Average 4,800-ft well. Well economics The study looked at average EUR/well/ tier, assuming a 20-year well life (Fig. HAYNESVILLE BREAKEVEN GAS PRICE, INTERNAL RATE OF RETURN Fig. 8 7). More than 80% of EUR is recovered 17 16 in the first 5 years, except for Tier 1, 15 IRR where an average well recovers 78% 14 20% of EUR during that period. Most wells 13 10% will be nearly depleted by year 10, 12 0% 11 with more than 90% of EUR produced. 10 The average EUR for all wells is likely 9 8 to be lower because of attrition and 7 economic limits. 6 The study’s production model 5 includes historical attrition rates, which Henry Hub, 2011; $/MMbtu 4 3 increase as rock-quality tier decreases. 2 151207OGJdti-08 BEG applied the average well profile 1 in each tier to estimate average well 0 Tier 1 Tier 2 Tier 3 Tier 4 Tier 5 Tier 6 economics. Input from operators in the Haynesville validated a representative set of well-economic parameters (Table 1). A comprehensive well cash-flow combination of hydraulic and natural fractures can model determined the internal rate of return (IRR) for an cause gas to flow from outside this idealized drainage average well in each tier (Fig. 8). area. Rectangles, however, provide an acceptable shape somewhat consistent with microseismic results, as well as Production outlook a means of accounting for the drained volume. The study modeled the pace of future Haynesville It was unclear initially if wells drained a large volume development using the productivity-tier map, inventory with a small recovery factor (RF) or a small volume with a of future well locations available in each tier, and an large RF to achieve their calculated EUR. understanding of the economics of an average new well in To estimate RF, BEG observed closely spaced wells that, each tier. based on changes in the original well-decline pattern as An activity-based model predicts new drilling based on nearby wells were added, appeared to interfere with each available-location inventory and well economics. The pace other. Many closely spaced wells exhibited some degree of activity is adjusted annually in the model, driven by the of interference, indicating drainage areas about equal to economics of the average well in a given tier. The model current well spacing in closely spaced blocks. distinguishes six productivity tiers based on economic We next developed 3D well-simulations of specific, incentives to drill. The historical pace of drilling is used to closely spaced blocks and adjusted recovery factor until help scale the model’s reaction to future prices. the RF, drainage area, and predicted well-declines matched BEG’s model tracks the number of wells/year drilled in actual well performance. each tier and totals the production effect using average well Using drainage-area calculations for every well, BEG profiles by tier. TECHNOLOGY TECHNOLOGY

New drilling production is next HAYNESVILLE PRODUCTION OUTLOOK, PRICE SCENARIOS Fig. 9 layered on top of extrapolated production decline of all existing 11 Henry Hub pricing, $/MMbtu wells. The model accounts for 10 observed drilling inertia to predict 4 how the pace of drilling will increase 9 4, $8 million CAPEX or decrease, based on reservoir 3 quality, as a function of price (change 8 6 10 in IRR) and the size of the remaining 7 well inventory.. AEO 2014 price forecast The model can restrict the 6 developable area, simulating, 5 for example, surface limitations Production, bcfd Production, or spacing inefficiencies such as 4 leasing obstacles, among many other such adjustments. The result is an 3 outlook of future completions from 2 the field through 2045, and a full- field EUR through 2064 for any set 1 of assumed parameters. 151207OGJdti-09 The model’s key assumptions 0 2008 2011 2014 2017 2020 2023 2026 2029 2032 2035 2038 2041 2044 include average well declines, effects of late-life deterioration on decline, effects of attrition, and a maximum 20-year life for all wells. HAYNESVILLE WELL ECONOMICS Table 1 The base-case scenario allows development of a maximum Well shut-in economic limit 0.05 MMcfd of 80% of the acreage in currently producing blocks, but Basis, Henry Hub –$0.07/MMbtu Royalty rate 25% only 40% of the acreage in undeveloped blocks. It also Severance tax rate 4%, 2 years exempt Marginal tax rate 40.2% (32.2% federal, 8% sets minimum activity levels in each tier, reflecting past state) performance in low price periods, and incorporates several Inflation rate 2.5% Drilling cost, 12,000-ft total vertical depth $10.5 million, 20% tangible other assumptions (Table 2). Related capital expenditures 13% Expense/well/year $82,500, +13% overhead The model generates a production outlook (Fig. 9). Gathering, compression, treatment $0.49/Mcf With $4/MMbtu natural gas at Henry Hub, Haynesville Lease cost/acre $3,000 Spacing, acres 80 production peaked in 2012 and declined rapidly as annual Depletion $0.03-0.40/Mcf well count decreased in response to lower prices in 2013. Abandonment cost $140,000 Production recovered somewhat as prices rebounded above $3.5 in 2012-13, encouraging drilling in the better locations, before declining steadily as the number of high- quality drilling sites dwindled. The better locations in Tiers HAYNESVILLE PRODUCTION ASSUMPTIONS Table 2 1 through 3 are developed, and the lower tiers do not justify Henry Hub gas price $4.00/MMbtu Development ceiling, partly drained acreage 80% development at prevailing prices. Development ceiling, undrilled acreage 40% Annual technology improvement 0.39% Our study forecasts full-field cumulative production in Annual well-cost improvement 0.24% Well shut-in economic limit 0.05 MMcfd the Haynesville of 46 tcf, including the 10.3 tcf from 2,527 Minimum completions/year 30, Tiers 1-3; 10, Tiers wells drilled through 2012. The production outlook and 4-6 resulting EUR are sensitive to natural gas price (Fig. 9). A substantial portion of the reservoir has breakeven gas prices of $4–6/MMbtu. Higher prices will extend both the production buildup and the subsequent plateau period. At Reducing capital expenditure to $8 million/well drops $6 Henry Hub pricing, the Haynesville would produce 56.9 breakeven prices, leading to more drilling sooner at $4/ tcf with more Tier 2 and 3 locations drilled. At $10 Henry MMbtu. In this scenario, field EUR reaches 51.9 tcf from Hub, full-field EUR reaches 72.3 tcf, boosted by significant 9,674 wells, about 6 tcf and 1,275 wells more than in the drilling in Tiers 4 and 5. base case. Peak production occurs in 2020 at about 7 bcfd Drilling and completion costs have declined in the and is sustained for 3 years. Ultimately, the better locations Haynesville. Even before the recent reduction in service fees in Tiers 1-3 are developed and the lower tiers do not justify following the oil price collapse, Haynesville capital expenditures development at prevailing prices. sank as low as $8 million/well (OGJ, Feb. 3, 2014). TECHNOLOGY

References Frank Male ([email protected]) is a postdoctoral fellow at 1. Patzek, T.W., Male, F., and Marder, M., “Gas production in BEG. He holds a BS in physics and a BA in political science from the obeys a simple scaling theory,” Proceedings of Kansas State University, and a PhD in physics from the University the National Academy of Sciences, Vol. 110, No. 49, Dec. 3, 2013, of Texas, Austin. As an undergraduate he was a research intern pp. 19731-19736. at the Max Planck Institute for Dynamics and Self-Organization in 2. Fu, Q., Horvath, S., Potter, E. C., and Roberts, F., “Log- Goettingen, Germany. He is an SPE member. derived thickness and porosity of the Barnett Shale, Fort Worth basin, Texas: Implications for assessment of gas shale resources,” Gürcan Gülen ([email protected]) is a research scientist AAPG Bulletin, Vol. 99, No. 1, January 2015, pp. 119-141, and senior energy economist at BEG’s Center for Energy Economics. 3. Ver Hoeve, M., Meyer, C., Preusser, J., and Makowitz, A., He holds a PhD in economics from Boston College, and a BA in “Basin-wide delineation of gas shale “sweet spots” using density economics from Bosphorus University, Istanbul. He is a member and neutron logs: Implications for qualitative and quantitative of USAEE, SPE, the American Economic Association, and the Gulf assessment of gas shale resources,” AAPG Hedberg Conference, Coast Power Association. Austin, Tex., Dec. 5–10, 2010. 4. Hammes, U., Hamlin, H., Ewing, S.,, and Thomas, E., Katie Smye ([email protected]) is a research associate at “Geologic analysis of the Upper Haynesville Shale in east BEG, joining as a postdoctoral fellow in 2013. She holds a PhD in Texas and west Louisiana,” AAPG Bulletin, Vol. 95, No. 10, October earth sciences from the University of Cambridge, and BS degrees in 2011, pp. 1643-1666, geology and chemistry from the University of Oklahoma, Norman. 5. Thompson, J.W., Fan, L., Grant, D., Martin, R.B., Kanneganti, K.T., and Lindsay, G.J., “An Overview of Horizontal- Susan Horvath ([email protected]) is an associate with Well Completions in the Haynesville Shale,” Journal of Canadian Goldman Sach’s global natural resources group. She was previously Petroleum Technology, June 2011, pp. 22-34, a research scientist associate at BEG and a senior GIS analyst for 6. Bhuyan, K. and Passy, Q.R., “Clay estimation from GR and Rosetta Resources and the Arkansas Geologic Survey. She holds neutron-density porosity logs,” SPWLA Logging Symposium, Tulsa, a BS and an MS in geographic information systems from Eastern Okla., June 19-22, 1994. Michigan University, Ypsilanti. She is a member of AAPG, the 7. Eastwood, R. and Hammes, U., “Log model development Houston Geological Society and ESRI’s Petroleum Users Group. for Bossier and Haynesville shales,” SPWLA Logging Symposium, Colorado Springs, Colo., May 14-18, 2011. Carl Grote ([email protected]) is an analyst with Robert W. 8. Stoneburner, R. K., “The Haynesville Shale: What We Have Baird & Co.’s energy investment banking group. He was previously a Learned in the First Two Years,” SIPES Quarterly, Vol. 46, No. 3, graduate research assistant at BEG, and holds an MS in energy and February 2010. earth resources from the University of Texas, Austin. He also holds a 9. Male, F., Islam, A., Patzek, T., Ikonnikova, S., Browning, J., BA in economics from Washington and Lee University, Lexington, Va. and Marder, M., “Analysis of gas production from hydraulically fractured wells in the Haynesville Shale using scaling methods,” Tadeusz Patzek ([email protected]) is a professor of The Journal of Unconventional Oil and Gas Resources, Vol. 10, chemical and petroleum engineering and director of the upstream June 2015, pp. 11-17. petroleum engineering research center at the King Abdullah 10. Ikonnikova, S., Browning, J., Horvath, S., and Tinker, University of Science & Technology in Saudi Arabia. Patzek holds S.W., “Well recovery, drainage area, and future drill-well inventory: an MS and PhD in chemical engineering from the Silesian Technical empirical study of the Barnett play,” SPE Reservoir University in Poland, and is a full professor there. Patzek is a Evaluation & Engineering, Vol. 17, No. 4, November 2014, pp. distinguished member of SPE. 484–496. Eric Potter ([email protected]) is program director for energy The authors research at BEG after working for 25 years with Marathon Oil Co. as John Browning ([email protected]) is a senior research fellow at an exploration geologist and geoscience technology manager. He the Bureau of Economic Geology (BEG). He retired from ExxonMobil holds a BA in earth science from Dartmouth College, Hanover, NH, after 33 years. He holds a BS in mechanical engineering from the and an MS in geology from Oregon State University, Corvalis. University of Tennessee, Knoxville. Scott W. Tinker ([email protected]) is director of the Svetlana Ikonnikova ([email protected]) is a Bureau of Economic Geology at the University of Texas, Austin, the research associate and energy economist at BEG. She holds an State Geologist of Texas, and is acting associate dean of research in MS in applied mathematics and physics from Moscow Institute of the Jackson School of Geosciences at the UT. Tinker holds a PhD Physics and Technology and a PhD in economics and management in geological sciences from the University of Colorado, Boulder, from Humboldt University, Berlin. She is a member of the United an MS in geological sciences from the University of Michigan, Ann States Association for Energy Economics (USAEE), and an associated Arbor, and a BS in geology and business administration from Trinity member of AAPG and SPE. University, San Antonio, Tex.

Eprinted and posted with permission to University of Texas at Austin from Oil & Gas Journal December © 2015 PennWell Corporation