CoreBased Evaluation of Cyclic Solvent Injection HuffnPuff in
Core-Based Evaluation of Cyclic Solvent Injection (Huff-n-Puff) in Tight Rocks: Experimental & Modeling C. Song 1; H. Hamdi 1; C. R. Clarkson 1; A. Ghanizadeh 1 1 Tight Oil Consortium, University of Calgary, Canada
Outline q Motivations q Objectives q Theory q Experiments q Key Results q Discussions q Future Work q Conclusions Slide 1
Motivations q Limited recovery (˂5 -10%) of tight oil resources q Value of core-based huff-n-puff (HNP) evaluations Ø Constraining simulation models for EOR applications Ø Providing critical inputs for evaluation of field-scale pilots Ø Understanding fundamental controlling mechanisms q Limitation of current core-based HNP experiments Ø Primarily focused on high-permeability core samples Ø Unrealistic experimental configurations § Flow through the rock matrix (intact core plugs) § Flow around the rock matrix (intact core plugs) Slide 2
Motivations q Flow-around-Matrix approach: Current methodology Hawthorne et al. 2017 (SPE 185072) Slide 3
Motivations q Flow-through-Fracture: New methodology Fractured Duvernay core plug Clarkson et al. 2017 (NRCan Proposal) Fractured Montney core plug Slide 4
Objectives q To quantify oil recovery using cyclic gas injection with fractured core plugs q To identify fundamental recovery mechanisms in tight oil formations q To examine operational controls on oil recovery Ø Injection pressure & time Ø Soaking time Ø Production pressure & time To be able to understand underlying physics: ü Start simple with single-component gases (CO 2, CH 4, C 2 H 6), and then, ü Go complex with multi-component gases (produced gas) Slide 5
Theory: Huff-n-puff (HNP) Process As-received Oil saturated Injection Soaking q Controlling processes Ø Ø Swelling effect Viscosity reduction Interfacial tension reduction Light components extraction Production q Controlling process Ø Solution gas drive Slide 6
Samples: Rock q Rock Sample Ø Sample shape: Fractured core plug Ø Sample condition: As-received Ø Formation: Duvernay (Alberta, Canada) Ø Dominant mineralogy: Quartz/Illite-rich Ø TOC content (wt. %): 4. 26 wt% Ø Helium porosity (%): 3. 3 Ø Fracture (oil) permeability: 78 md (900 psi effective stress) Ø Slip-corrected gas (N 2) matrix permeability: 0. 000124 md (900 psi effective stress) Slide 7
Samples: Fluids q Fluid Samples Ø Liquid: Duvernay formation (dead) oil (dewaxed) § Density: 0. 82 g/cm 3 § Viscosity o 2. 043 c. P (ambient condition) o 2. 130 c. P (high-pressure condition; 1000 psig) § Compressibility: 6. 47 × 10 -6 psi-1 Ø Gas: CO 2 (99. 99%) Slide 8
Experimental Setup q Core Plugs § Length: 0. 1 - 3” q Paxial & radial § ˂ 10, 000 psi § Diameter: 1, 1. 5” q Pfluid § ˂ 5, 000 psi § Any liquid • brine • oil • etc Slide 9
Experimental Procedures q Gas (CO 2) HNP End Measure oil permeability (kfracture; stress-dependent) Oil production ≤ 1. 0% of OOIP Gas (CO 2) injection ` Saturate core plug with liquid (oil) Soaking Oil production Change (in this work): injection time, soaking time, production time Slide 10
Experimental Design q 3 D displacement model (core plug) Novelty: Lab-derived inputs were used for constraining the simulation ü PVT model tuned to Duvernay formation (dead) oil psi (70 atm) Simulated Minimum Miscibility Pressure (MMP): ~1028 üinjection pressure (selected): Gas/Liquid relative permeability data (Montney) 1300 psi – conducted in 6 cycles ü Stress-dependent fracture/matrix permeability (Duvernay) Slide 11
Key Results: Liquid Fracture Perm Fracture Liquid (Oil) Permeability (md) 300 Gas perm 250 200 150 Effective stress loading 100 Liquid Permeability_0. 5 cc/min Liquid Flow Rate_Effetive Stress Loading Liquid Permeability_1. 0 cc/min Liquid Flow Rate_Effective Stress Loading Liquid Permeability_1. 5 cc/min Liquid Flow Rate_Effective Stress Loading Liquid Permeability_0. 5 cc/min Liquid Flow Rate_ Effective Stress Unloading Liquid Permeability_1. 0 cc/min Liquid Flow Rate_Effective Stress Unloading Liquid Permeability_1. 5 cc/min Liquid Flow Rate_Effective Stress Unloading Gas Permeabilty_Mean Pore Pressure=15. 2 psi Liquid perm 50 Effective stress 0 unloading 0 1000 2000 3000 4000 5000 Effective Stress (psi) Slide 12
Key Results: CO 2 HNP Injection pressure: 1300 psi 1 hr injection 1 hr soaking 4 hrs production 0. 5 hrs injection 1 hr soaking 1 hr production Production pressure: atmosphere pressure Slide 13
Key Results: CO 2 HNP q Pressure vs. Oil production: Oil sample #1 Oil sample #2 Slide 14
Key Results: CO 2 HNP q Pressure vs. Gas production: Gas sample #1 Gas sample #2 Slide 15
Discussion: Advantages q More representative experimental conditions Ø Flow through fracture as opposed to flow around matrix o More representative of the subsurface condition Ø Possibility of applying axial/radial load during tests o Not a possibility for flow-around-matrix approach q Time/Labor-inexpensive: 6 cycles of HNP in 28 hours Slide 16
Discussion: History Match q Strategies to history match the experimental results Ø Work with uncertain diffusion coefficients Ø Modify the fracture width to allow more gas to the system Ø Increase kmatrix to its higher end measured in the lab Ø To honor the other laboratory measured data, we have limited the adjusting parameters to only a few Slide 17
Discussion: Experiment vs. Simulation Oil Recovery Factor (%) 50 45 40 35 30 25 20 15 Simulation results 10 Experimental results 5 0 0 5 10 15 20 25 30 Time (hr) Slide 18
Future Work q CH 4 HNP tests with variable operational conditions Ø Injection pressure & time Ø Soaking time Ø Production pressure & time q Rich gas (70% CH 4+30% C 2 H 6) HNP tests Ø Duvernay Ø Montney Ø Permian Basin q Laboratory PVT analysis of crude oil to further constrain the simulated PVT model q Further tuning of the displacement model to match both oil pressure and gas production data Slide 19
Conclusions q 6 cycles of CO 2 HNP tests on Duvernay core samples are completed with varied operational parameters q A recovery factor of 48. 5% is obtained during the CO 2 HNP process q Most oil is produced during the first two cycles, and oil recovery tends to decrease thereafter q Larger pressure drop is obtained during soaking periods of the first two cycles leading to a higher oil recovery q Oil recovery is well-matched by core-based simulation Slide 20
Acknowledgements q Sponsors of Tight Oil Consortium q NSERC* * Natural Sciences and Engineering Research Council of Canada Slide 2121
Q & A: Discussion Slide 2222
- Slides: 23