Model calibration for forecasting CO2-foam enhanced oil recovery field pilot performance in a carbonate reservoir

M. Sharma, Z. P. Alcorn, S. B. Fredriksen, A. U. Rognmo, M. A. Fernø, S. M. Skjæveland and A. Graue
Petroleum Geoscience, 26, 141-149, 15 November 2019, https://doi.org/10.1144/petgeo2019-093

Abstract

Application of foam has been found to mitigate challenges associated with field-scale CO2 floods for enhanced oil recovery (EOR) by providing in-depth mobility control. The field pilots that have been run so far have shown varying results, inferred mainly from inter-well tracer studies and production data analysis. A research collaboration has been set up to advance the technology of using foam as a mobility control agent for CO2 EOR, with focus on integrated reservoir modelling to assist technology transfer to a high-cost environment. A heterogeneous carbonate reservoir onshore in west Texas, USA has been selected for the field trial. The reservoir has been waterflooded for more than 50 years, and a significant part of it has been on continuous CO2 injection for the last 5 years. An inverted five-spot pattern, which had rapid CO2 breakthrough in adjacent producers and is currently recycling significant amounts of CO2, has been selected for the study. The pilot is planned for 2 years with surfactant alternating gas (SAG) injection in the first year, followed by CO2 injection in the next year.

A reservoir model was created by integrating available static and dynamic information. Since the measurement of static information and production performance is usually imprecise, even the most carefully constructed models do not exactly represent reality. In this paper, we present a workflow that was used to calibrate the reservoir model to historical data for practical forecasting, which takes into account a wide range of uncertainties caused by the inaccessibility of information. Laboratory studies were performed with reservoir cores, fluids and selected surfactant to obtain the base values of foam model parameters. As an output, distributions for key performance indicators such as cumulative oil production and CO2 retention were generated for the proposed pilot to guide further decision-making.

Supplementary material: The details of simulation set-up and results for history matching and prediction are available at https://doi.org/10.6084/m9.figshare.c.4704374

This article is part of the Energy Geoscience Series available at https://www.lyellcollection.org/cc/energy-geoscience-series

The miscible and immiscible CO2 flood technology for enhanced oil recovery (EOR) has evolved over the last four decades in terms of laboratory studies, numerical modelling and field operations. A large number of commercial CO2 floods have been operated since the first commercial CO2 injection in the SACROC Unit in 1972 (Merchant 2010). Some of these projects are still operational, with CO2 injection reaching over 80% HCPV (hydrocarbon pore volume), and tertiary recovery ranging between 15 and 20% of oil-initially-in-place (OIIP). Based on the knowledge gained from field-scale projects, it has been realized that a substantial volume of reservoir remains unswept during CO2 floods, leading to lower oil recovery compared to recovery at core-scale. This is mainly because of the low density and viscosity of CO2, and reservoir heterogeneity, resulting in viscous fingering, gravity segregation and poor sweep (Jarrell et al. 1990). Several technologies have been tested to improve CO2 flood performance, including near-well polymer or gel treatment, cementing for zonal isolation, alternating CO2 injection with water, and smart completions with inflow control valves (Sharma et al. 2016), with limited to moderate success for in-depth mobility control.

The laboratory studies and field pilots conducted so far confirm the viability of foam for CO2 mobility control away from the injector (Heller 1994). Also, it offers a better control during field operations because of the dynamic state of foam, allowing its effect to be completely reversed if required. Previous field trials with foam (Heller et al. 1985; Holm & Garrison 1988; Jonas et al. 1990; Chou et al. 1992; Hoefner & Evans 1995; Harpole & Hallenbeck 1996; Sanders et al. 2012; Mukherjee et al. 2016) have demonstrated the benefits of this technology to variable extent, with a few meeting all planned objectives and making a good return on investment.

A research collaboration has been set up between industry and universities to design and execute a field pilot with foam in a heterogeneous field onshore Texas, to better understand the large-scale displacement process using integrated reservoir modelling. The secondary objective is to demonstrate the application of foam in improving CO2 storage efficiency through selective isolation of high-permeability regions, and mobilization of water from pore space to allow more room for CO2 than in conventional injection schemes. The field selected for the pilot study is located onshore in the Permian Basin, west Texas. It came on line in the early 1940s and produced 12% of OIIP until the late 1960s. Waterflood began in early 1970s with wells on an 80 acre pattern. The field was developed throughout the early 1980s with infill drilling to establish 40 acre peripheral waterflood patterns. However, with a low primary plus secondary recovery of only 22% of OIIP by the late 1980s the operator had realized the need to reduce the pattern size. An infill programme was run to develop a field on a 20 acre five-spot pattern. Infill drilling yielded excellent results with an increase in oil production rate from 400 to 1200 stbd. However, a steep decline in production and high residual oil saturations in the reservoir rock after waterflood indicated the potential for tertiary oil recovery. Tertiary CO2 injection for EOR started in the eastern part of the field in October 2013 to target any remaining oil and further expanded to other patterns, which has resulted in an increase in oil production rate from 250 stbd in October 2013 to 800 stbd in March 2018 (Fig. 1). The peripheral producers of most of the patterns, however, have already experienced CO2 breakthrough, with breakthrough occurring as early as within 4 months from the start of CO2 injection in nearby injectors. The reservoir has a poor volumetric sweep due to reservoir heterogeneity and the unfavourable mobility of CO2, which makes it a good candidate to test foam for improving sweep and reducing CO2 recycling.

Fig. 1.
Fig. 1.

Field historical production. ROZ, residual oil zone.

After discussions with the field operator, an inverted five-spot pattern around well 1 (Fig. 2) was selected for injection for the field trial, which has representative geology, good well injectivity, short CO2 travel time and a high gas/oil ratio. The reservoir exhibits a large vertical heterogeneity, with a Dykstra–Parsons permeability variation coefficient of 0.79 and a Lorenz coefficient of 0.84. Because of short inter-well distances of around 700 ft, the reservoir is expected to respond to foam in a much shorter time interval.

Fig. 2.
Fig. 2.

Field layout and location of selected pilot area. The surfactant will be injected in well 1.

This paper presents a workflow for history-matching previous waterflood and CO2 injection to obtain a revised estimate of uncertainty about the static characteristics and dynamic behaviour of the field. Laboratory coreflood experiments performed with reservoir core and selected surfactant under representative conditions were used to characterize foam behaviour. The uncertainties in modelling foam behaviour were added to the uncertainties in the reservoir model to obtain a reliable forecast for the pilot performance in terms of the distributions for key performance indicators (KPIs) such as cumulative oil production and CO2 retention.

Methodology

History-matching approach

To assist the numerical modelling work for pilot design, a three-dimensional model was set up for a sector which included the wells in the pilot area, and peripheral water and CO2 injectors. The reservoir model was then tuned to available production data in two stages: Phase I for historical water injection from January 1971 until September 2013, and Phase II for historical CO2 injection from October 2013 until March 2018. Wells 1–11 (Fig. 2) were included in the sector model, which had 28 layers and around 70 000 active cells with areal dimensions of 50 ft. The plan is to inject surfactant in well 1 and observe the response at wells 3 and 5. Since the well status has changed during these phases, a prefix of ‘P’, ‘WI’ or ‘GI’ has been used in this paper to denote the state of the well as producer, water injector or CO2 injector, respectively.

It was realized that large uncertainties exist in modelling inputs because of the absence of seismic information, limited core availability, gaps in production data and limited information on well operations or workovers. Since the main objective of reservoir modelling was to create a reliable forecast that reflects all the available information, the geological model was history matched to reduce the range in uncertainties before using it for the pilot design. The integration of dynamic data requires the solution of flow equations several times in an iterative fashion. The workflow that was used for assisted history matching is outlined in Figure 3, and includes: setting up an uncertainty matrix and objective function, model validation, selecting starting points for assisted history matching, and completing the history-matching phase using evolutionary strategy. The workflow transforms uncertainty about input information into an ensemble of predictions that describes the uncertainty in production. As we proceeded, initial views about uncertainties were revised to get a reasonable fit to observations using more than one set of values for model parameters. The history-matching steps and detailed results are presented in the Supplementary material.

Fig. 3.
Fig. 3.

History matching and forecasting workflow.

Prediction approach

Reservoir management plan

The current reservoir pressure is 600–800 psi higher than hydrostatic pressure (2300 psi). The injectors operated at a flowing bottom-hole pressure constraint of 3950 psi, close to fracture pressure, until March 2018. The reservoir management plan going forward is to inject at a lower rate, at levels that are almost half that of historical injection rates. For the sector model, it is equivalent to a reduction in injection rate from 0.1 PV/year (pore volume per year) to 0.05 PV/year. A disposal well, completed in a separate deeper reservoir, is being used to depressurize the reservoir before the start of the pilot. The gas injectors, after conversion from either a producer or a water injector, have been on continuous CO2 injection in the past. The revised plan is based on water alternating gas (WAG) injection, with water injection for 1 year followed by CO2 injection for 6 months. Based on discussions with the operator, it has been agreed to implement surfactant alternating gas (SAG) injection in the chosen injector. Optimization of slug sizes for the WAG and foam pilot was not considered for this part of the study.

Modelling foam rheology

The foam rheology was modelled using an implicit-texture approach (Cheng et al. 2000; Alvarez et al. 2001), which uses an empirical relationship to capture the effect of surfactant concentration, water saturation, oil saturation and shear thinning due to flow velocity on foam mobility. In this approach, the gas permeability in the presence of foam (krgf) is modified by multiplying the gas relative permeability without foam (krgnf) at a specific water saturation with a mobility reduction factor (MRF):krgf=krgnfMRF (1)The water permeability in the presence of foam remains unchanged. We studied the effect of water saturation, shear rate, surfactant concentration, oil saturation (Farajzadeh et al. 2012) and permeability on the mobility reduction factor in numerical modelling, given by the expression:MRF=11+fmmobFwaterFshearFsurfFoil (2)where fmmob refers to the maximum gas mobility reduction factor that can be achieved for foam. Fwater, Fshear, Fsurf and Foil, with the expressions available in the Supplementary material, capture the effect of water saturation, shear rate, surfactant concentration and oil saturation dependence, with all lying in the range of 0–1. The capillary number Nca represents the relative effect of viscous and capillary forces.

A non-ionic water-soluble surfactant (Surfonic L24-22) was selected for the field pilot based on surfactant screening studies for the reservoir (Nguyen et al. 2015). Surfonic L24-22 is a linear alcohol ethoxylate produced by the addition of ethylene oxide (EO) to linear, primary alcohols. It is a 22 mol ethoxylate of linear, primary 12–14 carbon number alcohol. The prediction steps and detailed results are presented in Appendix.

Results and Discussion

Waterflood match

Thirty years of waterflood data were available for wells (Fig. 4) in the pilot area that was used to calibrate the geological model. Only the 15 most significant uncertainty parameters (UPs), which were found to influence the mismatch in the cumulative oil production the most, were considered from a comprehensive list for history matching. Figure 5a and b show the cumulative oil production and water cut for the entire sector, respectively, for cases selected to update the range for the 15 UPs after history match. Similarly, Figure 6 compares the cumulative oil production for the five producers (P-1, P-2, P-3, P-4 and P-5) for the selected cases with the wells on the liquid production rate control.

Fig. 4.
Fig. 4.

Historical well performance during waterflood: (a) cumulative water injection; and (b) cumulative oil production.

Fig. 5.
Fig. 5.

Simulation results at the sector level for cases selected to update (posterior) uncertainty parameter ranges after running the evolutionary strategy (ES): (a) cumulative oil production; and (b) water cut.

Fig. 6.
Fig. 6.

Cumulative oil production for producers P-1, P-2, P-3, P-4 and P-5, for cases selected to update (posterior) uncertainty parameter ranges after running the ES.

CO2 injection match

CO2 flood data were available for 4.5 years (Figs 7 and 8) for wells in the pilot area which was used to further calibrate the reservoir model. All studies for the CO2 match were performed by fixing the 15 UPs from the waterflood match at their mean values. Based on the results from the sensitivity analysis, 43 (out of 70) UPs were carried forward in the history matching. Figure 9a–d compares the observed and simulation responses for cumulative oil production, water cut and gas/oil ratio for the entire sector after the history match. The relative error in cumulative oil production was reduced to less than 10% after the history match. Figures 10 and 11 compare the observed and simulation responses after history matching for the cumulative oil production, water cut and gas/oil ratio for P-3 and P-5, respectively.

Fig. 7.
Fig. 7.

Historical well performance during CO2 injection: (a) cumulative CO2 injection; and (b) tubing-head pressure.

Fig. 8.
Fig. 8.

Historical well performance during CO2 injection: (a) cumulative oil production; and (b) gas production rate.

Fig. 9.
Fig. 9.

Simulation results at the sector level for cases selected to update (posterior) uncertainty parameter ranges: (a) cumulative liquid production, showing that producers do not switch from the assigned liquid rate control; (b) cumulative oil production; (c) water cut; and (d) gas/oil ratio.

Fig. 10.
Fig. 10.

Simulation results for P-3 for cases selected to update (posterior) uncertainty parameter ranges after running the ES: (a) cumulative oil production; (b) water cut; and (c) gas/oil ratio.

Fig. 11.
Fig. 11.

Simulation results for P-5 for cases selected to update (posterior) uncertainty parameter ranges after running the ES: (a) cumulative oil production; (b) water cut; and (c) gas/oil ratio.

As shown in Figure 12, four injection profiles have been recorded for GI-1 at a 1 year interval since the start of CO2 injection in October 2013. The profiles were recorded using a radioactive tracer logging tool in slug tracking mode, where a ‘shot’ of tracer was injected into the wellbore. A gamma-ray detector was used to record the location of the slug and amplitude variation with time by successive upward and downward passes or ‘drags’. For each drag, the area under the trace and above the common baseline of the traces was used to quantify the fraction of injection fluid still in or near the wellbore. Even though the technique is economic for multi-well surveys and broadly identifies zones of injection, it has a poor vertical resolution. The profiles were therefore not used directly for model calibration. The fractions of injected CO2 entering into each of the two pay zones were calculated instead, and were used to test the predictive capability of the history-matched models. Figure 13 shows the simulation results for the fraction of injected CO2 entering into the main producing zone (MPZ) after calibration, which reduces from 100% at the beginning of the CO2 injection to around 70% after 2 years of injection. Well 1 was a producer during the waterflood phase with completions in the MPZ, and the formation in the residual oil zone (ROZ) was perforated when it was converted into a CO2 injector. Because of selective depletion in the MPZ around well 1 due to production, CO2 preferentially flowed into the MPZ until local equilibration between the zones was achieved. Since the last injection profile for well 1 is 2 years old, a new survey is planned before the start of the pilot which will act as a baseline for future profiles during the pilot phase.

Fig. 12.
Fig. 12.

Historical injection profiles for GI-1 recorded every year since the start of CO2 injection in October 2013, showing the fraction of CO2 entering into the main producing zone (MPZ) and residual oil zone (ROZ).

Fig. 13.
Fig. 13.

The fraction of CO2 injected into GI-1 entering into the main producing zone (MPZ) for cases selected to update the posterior uncertainty parameter ranges after CO2 injection match.

An inter-well tracer study was initiated on 9 January 2018, with injection of a passive non-radioactive gas tracer (Khan et al. 2016) in GI-1, to characterize the communication between GI-1 and the surrounding producers P-2, P-3, P-4 and P-5. The first set of data received from the field until March 2018 showed tracer breakthrough in P-3 and P-5 in 17 days. P-3 and P-5 were sampled twice a week, and have produced 16.5 and 6% during this period. No tracer has been observed in P-2 and P-4. The tracer study is still in progress, with a reduced sampling frequency of one sample every 2 weeks. The tracer response was not a part of the objective function for the current history-match cycle, and was only used to validate the quality of the matched models. Figure 14 compares the simulation results for the cumulative tracer production for both the wells as a fraction of the total amount of tracer injected. Even though the models predict lower tracer production, mainly because of the presence of high-resolution features in the reservoir which cannot be captured in the current model, their performance was found to be acceptable.

Fig. 14.
Fig. 14.

Cumulative tracer production, as a fraction of injected volume, for cases selected to update the posterior uncertainty parameter ranges: (a) P-3; and (b) P-5.

Foam pilot performance prediction

Twelve UPs were considered for modelling foam behaviour, which were combined with the 58 UPs from the history-matching phase to generate 100 Latin hypercube (LHC) cases in two scenarios. As shown in Figure 15a, the first scenario is based on the operators' current plan to implement WAG into the pilot area with 1 year of water injection followed by 6 months of CO2 injection. The second scenario (Fig. 15b) considers 12 cycles of SAG starting on 1 November 2018, followed by continuous CO2 injection. The SAG strategy, which is based on discussions with the operator and cross-sectional numerical modelling, includes 10 days of surfactant injection and 20 days of CO2 injection. Water and CO2 injection will be limited to 300 stbd and 1000 Mscfd, respectively, which are around half of the maximum rates that can be injected into well 1 at the maximum allowable flowing bottom-hole pressure. These injection schemes result in a similar injection of around 1.1 MMbbl under reservoir conditions for both scenarios. For each case, simulations were run using the sampled values of the UPs for all three phases: waterflood, CO2 injection and forecast for 3 years. The initial pressure, saturation and composition for each case for the second and third phases were based on the values extracted from the last step of the previous phase.

Fig. 15.
Fig. 15.

Injection scheme for: (a) the base-case scenario with WAG; and (b) the pilot with 12 SAG cycles followed by CO2 injection.

Figure 16a shows the cumulative probability distribution for incremental oil forecasted using the LHC cases for Scenario 2 with respect to Scenario 1 after 1, 2 and 3 years of the start of the pilot. Because of injection at a low rate (0.05 PV/year), it is expected that the incremental volume of oil produced with foam will reach a significant level only after the SAG cycles have been completed. In addition to incremental oil production, CO2 retention was evaluated which is defined as:CO2retention=CO2injected−CO2producedCO2injected (3)

Fig. 16.
Fig. 16.

Cumulative distribution for KPIs: (a) incremental oil; and (b) the increase in the CO2 retention factor.

In the above expression, the produced CO2 volume corresponds to the recycled CO2 volume assuming no CO2 loss at the surface, while the injected CO2 volume corresponds to the sum of the purchased and recycled CO2 volumes. Figure 16b shows the cumulative probability distribution for the CO2 retention factor for both scenarios after 2 and 3 years of the start of the pilot. The selective foam generation in high-permeability regions allows the diversion of CO2 into low-permeability regions, resulting in a higher CO2 retention. Table 1 lists the P90, P50 and P10 values for incremental oil, and the increase in CO2 retention with foam.

View this table:
Table 1.

Confidence intervals for KPIs based on simulation cases

Conclusions

A research programme has been initiated to advance the technology of using foam as a mobility control agent for CO2 EOR and storage as part of a carbon capture, utilization and storage (CCUS) value chain, with the focus on integrated reservoir modelling to assist technology transfer to a high-risk and high-cost environment. A heterogeneous carbonate reservoir onshore in west Texas, USA has been selected for the field trial.

In this paper, we discussed a workflow to create reliable forecasts of future field behaviour which honours all the available information, including core and petrophysical data, production data, and laboratory data. It relies on applying engineering judgement while using a reservoir model to obtain an understanding about the uncertainty in the production forecasts, as accurately as possible. Inputs from all stakeholders were used to understand the uncertainties in the initial numerical model, which were revised while calibrating the model performance to match historical production data. The predictive capability of the calibrated models was tested for injection profiles and inter-well tracer test data, which were excluded during history matching. Even though the input data were sparse and had uncertainties, the calibrated models were found to be fit-for-purpose. The workflow was then used to generate a spread of forecasts to predict the foam pilot performance. A baseline injection profile and fall-off test is planned before the start of the pilot, which will be further used to improve the model quality and guide decision- making during field operation.

Abbreviations

MRF
Mobility reduction factor
OIIP
Oil Initially In-Place
PV
Pore volume
PVT
Pressure Volume Temperature
ROZ
Residual oil zone
SAG
Surfactant alternating gas
SCAL
Special core analysis
STB/D
Stock tank barrels per day
UP
Uncertainty parameter
WAG
Water alternating gas
Wt %
Weight percent

Acknowledgements

The authors acknowledge the industry partners, TOTAL E&P, Shell E&P and Equinor ASA, and the field operator, Tabula Rasa, for support. The authors acknowledge the industry partners of the National IOR Centre of Norway for support. The authors also acknowledge NOTUR for their support to perform this work on the Abel Cluster, University of Oslo.

Funding

The authors acknowledge the Research Council of Norway CLIMIT programme for support under grant No. 249742 (‘CO2 Storage from Lab to On-Shore Field Pilots Using CO2-Foam for Mobility Control in CCUS’) to the Petroleum and Process Technology Research Group at the University of Bergen.

Author contributions

MS: Conceptualization (Lead), Methodology (Lead), Writing – Original Draft (Lead); ZPA: Conceptualization (Supporting), Writing – Review & Editing (Supporting); SBF: Conceptualization (Supporting), Writing – Review & Editing (Supporting); AUR: Conceptualization (Supporting), Writing – Review & Editing (Supporting); MAF: Conceptualization (Supporting), Writing – Review & Editing (Supporting); SMS: Conceptualization (Supporting), Writing – Review & Editing (Supporting); AG: Conceptualization (Supporting), Writing – Review & Editing (Supporting).

References

  1. ↵
    1. Alvarez, J.M.,
    2. Rivas, H. &
    3. Rossen, W.R.
    2001. A unified model for steady-state foam behavior at high and low foam qualities. SPE Journal, 6, 325–333, https://doi.org/10.2118/74141-PA
    OpenUrl
  2. ↵
    1. Cheng, L.,
    2. Reme, A.B.,
    3. Shan, D.,
    4. Coombe, D.A. &
    5. Rossen, W.R.
    2000. Simulating foam processes at high and low foam qualities. Paper SPE-59287 presented at the SPE/DOE Symposium on Improved Oil Recovery, 3–5 April 2000, Tulsa, Oklahoma, USA.
  3. ↵
    1. Chou, S.I.,
    2. Vasicek, S.L.,
    3. Pisio, D.L.,
    4. Jasek, D.E. &
    5. Goodgame, J.A.
    1992. CO2 foam field trial at North Ward-Estes. Paper SPE-24643 presented at the SPE Annual Technical Conference and Exhibition, 4–7 October 1992, Washington, DC, USA.
  4. ↵
    1. Farajzadeh, R.,
    2. Andrianov, A.,
    3. Krastev, R.,
    4. Hirasaki, G.J. &
    5. Rossen, W.R.
    2012. Foam–oil interaction in porous media: Implications for foam assisted enhanced oil recovery. Advances in Colloid and Interface Science, 183–184, 1–13, https://doi.org/10.1016/j.cis.2012.07.002
    OpenUrl
  5. ↵
    1. Harpole, K.J. &
    2. Hallenbeck, L.D.
    1996. East Vacuum Grayburg San Andres Unit CO2 flood ten year performance review: Evolution of a reservoir management strategy and results of WAG optimization. Paper SPE-36710 presented at the SPE Annual Technical Conference and Exhibition, 6–9 October 1996, Denver, Colorado, USA.
  6. ↵
    1. Heller, J.P.
    1994. CO2 foams in enhanced oil recovery. In: Schramm, L.L. (ed.) Foams: Fundamentals and Applications in the Petroleum Industry. Advances in Chemistry, 242. American Chemical Society, Washington, DC, 201–234.
    OpenUrl
  7. ↵
    1. Heller, J.P.,
    2. Boone, D.A. &
    3. Watts, R.J.
    1985. Field test of CO2 mobility control at Rock Creek. Paper SPE-14395 presented at the SPE Annual Technical Conference and Exhibition, 22–26 September 1985, Las Vegas, Nevada, USA.
  8. ↵
    1. Hoefner, M.L. &
    2. Evans, E.M.
    1995. CO2 foam: Results from four developmental field trials. SPE Reservoir Engineering, 10, 273–281, https://doi.org/10.2118/27787-PA
    OpenUrl
  9. ↵
    1. Holm, L.W. &
    2. Garrison, W.H.
    1988. CO2 diversion with foam in an immiscible CO2 field project. SPE Reservoir Engineering, 3, 112–118, https://doi.org/10.2118/14963-PA
    OpenUrl
  10. ↵
    1. Jarrell, P.M.,
    2. Fox, C.E.,
    3. Stein, M.H. &
    4. Webb, S.L.
    1990. Practical Aspects of CO2 Flooding. SPE Monograph Series, 22.
  11. ↵
    1. Jonas, T.M.,
    2. Chou, S.I. &
    3. Vasicek, S.L.
    1990. Evaluation of a CO2 foam field trial: Rangely Weber Sand Unit. Paper SPE-20468 presented at the SPE Annual Technical Conference and Exhibition, 23–26 September 1990, New Orleans, Louisiana, USA.
  12. ↵
    1. Khan, M.N.,
    2. Iwama, H.,
    3. Al-Neaimi, A.,
    4. Al-Shehhi, O. &
    5. Chatterjee, M.
    2016. A to Z of gas tracers – A decade of learning and experience. Paper SPE-183413 presented at the SPE Abu Dhabi International Petroleum Exhibition & Conference, 7–10 November 2016, Abu Dhabi, UAE.
  13. ↵
    1. Merchant, D.H.
    2010. Life beyond 80: A look at conventional WAG recovery beyond 80% HCPV injection in CO2 tertiary floods. Paper SPE-139516 presented at the SPE International Conference on CO2 Capture, Storage, and Utilization, 10–12 November 2011, New Orleans, Louisiana, USA.
  14. ↵
    1. Mukherjee, J.,
    2. Nguyen, Q.P.,
    3. Scherlin, J.,
    4. Vanderwal, P.G. &
    5. Rozowski, P.
    2016. CO2 foam pilot in Salt Creek Field, Natrona County, WY: Phase III: Analysis of pilot performance. Paper SPE-179635 presented at the SPE Improved Oil Recovery Symposium, 11–13 April 2016, Tulsa, Oklahoma, USA.
  15. ↵
    1. Nguyen, Q.P.,
    2. Hirasaki, G.J. &
    3. Johnston, K.P.
    2015. Novel CO2 Foam Concepts and Injection Schemes for Improving CO2 Sweep Efficiency in Sandstone and Carbonate Hydrocarbon Formations. Technical Report. United States Department of Energy/National Energy Technology Laboratory, Pittsburgh, PA, https://doi.org/10.2172/1178538
  16. ↵
    1. Sanders, A.,
    2. Jones, R.M.,
    3. Linroth, M.A. &
    4. Nguyen, Q.P.
    2012. Implementation of a CO2 foam pilot study in the SACROC Field: Performance evaluation. Paper SPE-160016 presented at the SPE Annual Technical Conference and Exhibition, 8–10 October 2012, San Antonio, Texas, USA.
  17. ↵
    1. Sharma, M.,
    2. Taware, S.V. &
    3. Datta-Gupta, A.
    2016. Optimizing CO2 floods using rate control with smart wells under geologic uncertainty. Paper SPE-183191 presented at the SPE Abu Dhabi International Petroleum Exhibition & Conference, 7–10 November 2016, Abu Dhabi, UAE.
View Abstract
Back to top
Alerts
Citation tools
Permissions
View PDF
Email to
Print
Download PPT