FRACTURING FLOWBACK: CONTROLS, ANALYSIS & BENEFITS SPE GCS WESTSIDE STUDY GROUP JANUARY 15, 2015 GEORGE E. KING, P.E. I ll review several presentations from the SPE Workshop on fracturing flowback, 6-7 November 2013, San Antonio
First it is Produced Fluid not a waste Returning Fluid Composition: FLOWBACK WHAT IS IT? Frac base water and frac additives Waters from one or more formations A variety of salts and ions some stable, some not. Isotopes that can range from benign to low dose radioactive Solids of silica and many other minerals Hydrocarbon gas and liquids Other gases Consistency highly variable Early time gaining salinity Late time less saline? 2
WATER MANAGEMENT: QUANTITIES FLOWED BACK IN SHALE RESERVOIRS (RANGES) Basin or Area % Frac Water Recovered Typical Frac Volume Used (Gal.) Typical Chemical % in Frac Barnett 30 to 50% 4 to 5 mm 0.2% <0.05% Devonian 40 to 50% 4 to 5 mm 0.2% <0.1% Eagle Ford 5 to 10% 4 to 5 mm 0.3 to 0.4% (Hybrid Frac) Fayetteville 30 to 60% 3 to 4 mm 0.2% <0.05% Haynesville 5 to 15% 4 to 6mm 0.3% (Hybrid Frac) Horn River 30 to 50% 10 to 12mm (salt water Supply wells) <0.1% (Apache) <0.05% Woodford 30 to 50% 4 to 5 mm 0.2% <0.05% Chemical % in Flowback (Gross Est.) <0.2% (polymer dominated) <0.1% (polymer dominated) Sources: SPE 133456, SPE 152596, communication with operators in these basins. Also SPE papers on produced water treating.
WHAT CAN FLOWBACK CONTROL & ANALYSIS DO FOR YOU? 1. Help you decide if flowback control will work in your reservoir. 2. Indicate whether the frac is planar type or complex. 3. Tell you where some of the frac has gone. 4. Tell you a bit of history & predict some of the future. 5. Tell you what type of control and how much drawdown control to use. i. The type of drawdown control to use (rate, pressure drop, time ii. iii. or other factor). Can drawdown control really increase EUR, hydrocarbon liquid production, gas rate and decrease water recovery? Is the amount of water recovery related to production improvements? 4
MESSAGE IN A BOTTLE (SPE 168892) Produced water following hydraulic fracture stimulation frequently contains unique messages (data) from the stimulated formations. Changes in chemistry of water reflect the architecture of the producing stimulated network. Processes of water mixing; solid dissolution ion diffusion from matrix water to fracture water and the effect of area-to-volume ratio in leaching of ions from walls of the fracture to the injected water describe where the frac water went. Is rapid flow easier from a planar fracture than a complex fracture? 5 Bearinger, D.: Message in a Bottle, (Nexen) SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
BASICS OF LEACHING AND MIXING Dissolutions favors the most soluble materials (NaCl). Mixing of connate waters generates faster rise in return water salinity than leaching of minerals by frac water. Connate water in pores, natural fractures & fissures already at equilibrium and mixing of frac water with connate waters dominates ion transfer (some leaching possible by frac water). Fresh rock break surfaces are not wetted and leaching dominates ion transfer. Contacted area-to-volume ratio is the major control: If frac opens natural fractures (complex fracturing), water is usually higher salinity than if fresh rock is opened (planar fractures). Frac water returning from freshly fractured formation will have different ion ratios than fracture waters returning from natural fractures. 6
CONTACT AREA Planar frac (traditional biwing, single frac) Area to volume ratio ~ 50:1 Complex or network frac (opened natural fractures) Area-to-volume ratio > 1000:1 (maybe 10,000:1?) 7 What happens when the crack is forced to open wide for first time in a million years?
BASICS OF LEACHING & MIXING - ORGANICS Organic composition in recovered water may also have traces or water miscible hydrocarbons (PAH and/or branched chain organics) that are relatively rare in shales. Shales are essentially a large reactor and the ultra-low permeability assures the pores and fractures have not been swept and washed as have migrating or trapped hydrocarbon fluids in conventional reservoirs. These hydrocarbons may be the first indicator of what may be in the reservoir whether it is produced or not. What you produce may not match what remains. 8
ION RATIOS IN THE PRODUCED WATER Dissolution favors NaCl and KCl. Sodium & potassium ions appeared to move faster by diffusion than Ca & Mg ions. Critical for ion transfer & ion ratio judgment. Formation clay & permeability may affect cation ratios. (could still be some cation exchange occurring) Slope of TDS rise related to connate water salinity and contact area of the fractures & the formation. (e.g., planar vs. complex) Ion concentration & ion ratios impacted by area-to-volume ratio of fracs. Lower AVR in planar fracs & higher AVR in complex fracs. AVR of natural fractures generated higher TDS & changes in Ca/Na ratios. Waters in producing wells were diluted by condensation from gas, but dilution has no effect on ion ratio. After load fluid recovered, (and gas breaks through) salinity may drop as formation water production decreases & condensed water increases. 9 Bearinger, D.: Message in a Bottle, SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
SPECIFICS FOR SHALE RESERVOIRS Lack of mobile water in most shales prevents sampling of insitu water. In shales, connate brines become more concentrated as water is vaporized into gas stream Bennion, 2002. Lack of matrix permeability favors diffusion as dominant process for ion leaching & transfer. Opening natural fractures forces water in and changes the ion balance. Ion concentration & ion ratios are impacted by area-to-volume ratio (AVR) of fractures. Lower AVR in planar fracs and higher AVR in complex natural fractures. AVR of the natural fractures generates higher TDS in initial returns. Changes in Ca/Na ratios may identify difference between diffusion dominated and mixing dominated areas. 10
VARIANCE IN PRODUCED WATER TDS Time of sampling is critical. Frac base water recovered first, then salinity increases as water recovered from complex (natural fractures). Water reaches a plateau characteristic of stable water flow from early production. Salinity declines after frac flowback is exhausted, connate water decreases and condensed water increases with gas production increase. Bearinger, Doug: Message in a Bottle, (Nexen) SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. 11
TRACERS WHY THE VARIANCE? Number of factors influencing amount of tracer recovered Area of surface to volume of water in contact w/ that area. Reduction in tracer > reduction in water recovered. Loss to matrix or influenced by dilution. TDS in flowback is a function of shale and geographic area In Eagle Ford - ratio of Ca/Na took a sharp rise for a very small rise in TDS. New influx? Higher TDS waters? Changing Ca/Na ratio suggest changing flow source/process. 12
FLOWBACK TO GAUGE COMMUNICATION - TRACERS Each stage was marked with a different tracer. The shows in offset wells ranged from none to three tracers types. Stage 1 impacts large area around frac path. Growth of stages 2, 3 and 4 were distinctly pushed away from the first stage. Evidence of stress from fracturing and utility in fracture spacing calculations? Woodroof, R.: Lessons Learned from Chemical Tracer Datasets, SPE Workshop Hydraulic Fracturing Flowback, 13 6-7 November 2013, San Antonio, TX. & SPE 168892
CAN TRACERS BE TRUSTED AS FLOW INDICATORS? Production log & tracer response shows interesting results: 50% of gas from toe largest development of tracer range. 28% from heel no overlap with other fracs free to grow. Woodroof, R.: Lessons Learned from Chemical Tracer Datasets, SPE Workshop Hydraulic Fracturing Flowback, 14 6-7 November 2013, San Antonio, TX. & SPE 168892
What is this telling us? If we get the highest gas recovery with lowest water recovery, should we be trying to get less early water recovery? It may help with fracture type estimation & early is a key. 15 Ezulike, O.: Flowback Analysis for Determining Load Recovery and Its Effects on Early-Time Hydrocarbon Production Rate, (U Alberta) SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
HOW ONE AUTHOR APPLIED THE OBSERVATION Other factors will impact a decision on fracture type. Insitu Stresses Presence/density of natural fractures Formation perms Hydrocarbon liquid Accuracy of sampling & analysis. Large amount of data to neutralize the wide variations in U.C. formations. Ezulike, O.: Flowback Analysis for Determining Load Recovery and Its Effects on Early-Time Hydrocarbon Production Rate, 16 (U Alberta) SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
FLOWBACK VOLUME VS. GAS PRODUCTION 17 Ezulike, O.: Flowback Analysis for Determining Load Recovery and Its Effects on Early-Time Hydrocarbon Production Rate, (U Alberta) SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
EVIE HIGHER GAS PROD. W/ LESS WATER RECOVERY 18 Bearinger, Doug: Message in a Bottle, (Nexen) SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX.
SLOWBACK - CONTROLLED FLUID RECOVERY Rate: from 200 gallons per hour to 75 bbls/hr. Choke size: initial choke settings of 8/64 to 12/64 and increasing choke with time or pressure drop. Pressure drop: 25 to <50 psi per day is typical. Can set choke to a predetermined level and not moved for duration of test (excluding shut-ins or maintenance). Well rates, cumulative production and pressure are monitored and compared to direct offsets on the pad. The test is concluded once the well has declined to line pressure (other considerations may shorten the test). 19
SLOWBACK WILL IT ACTUALLY HELP ME? Info from well tests in Marcellus of WV & PA. Noble s conclusion is that early time pressure drawdown management increases the expected ultimate recovery. They believe propped pressure drawdown sustains contributions from micropores and micro-cracks, most relatively close to the wellbore. They also see benefits of minimizing fines and proppant migration. 20 Arden Rinze and Matthew Englert of Noble Energy: Pressure Drawdown Management in the Marcellus Shale, SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
HOW CLEAR ARE THE BENEFITS? The claim is a 5% increase in gross reserves. Southern Slowback well comparable to best well on the pad. Cumulative Gas Cumulative Oil 21
THOUGHTS ON SLOWBACK It works better in some places than others. Where? Haynesville, Eagle Ford, wet Marcellus? More dependent on specific formation fabric? 22 Vincent, M.: Five Things You Didn t Want to Know about Hydraulic Fracture, Proc. Of International Conference for Effective and Sustainable Hydraulic Fracturing, ISRM, Brisbane, Australia, May 2013.
HOW ARE WELLS CONNECTED AND FOR HOW LONG Inter-well communication can be characterized through the use of readily available completions and production data. Inter-well communication in frac & flowback is healthy Expect decreases It is important to characterize inter-well communication to manage production and potentially maximize production. Litchfield, T. & Lehmann, J. spoke about Inter-well Interference during stimulation, flowback and production history. 23 SPE Workshop Hydraulic Fracturing Flowback, 6-7 November 2013, San Antonio, TX. & SPE 168892
24 CONCLUSIONS 1. It will take trials to decide if flowback control will work in your reservoir. 2. It can indicate whether frac is planar or complex. 3. It can indicate where some of the frac has gone and when flowback changes. 4. As unpropped fracture connection closes, will job design change? Depends on when and impact. Spacing changed. Perhaps new proppant needed? Refracturing as a trial? 5. It will take trials to find what type of control and how much drawdown control to use. It is probably formation specific. i. Type of drawdown control (rate, pressure drop, choke, time or ii. iii. other). Can drawdown control really increase EUR, hydrocarbon liquid production, gas rate and decrease water recovery? Depends on the fluid and even that depends on conditions. Is the amount of water recovery related to production improvements? Jury is still out on that one.