Summary

This document is the Final Report for the U.S. Department of Energy under contract No. DE-FG26-97BC15047, a three-year grant titled: “Improved Efficiency of Miscible CO₂ Floods and Enhanced Prospects for CO₂ Flooding Heterogeneous Reservoirs.” The research improves our knowledge and understanding of CO₂ flooding and includes work in the areas of fluid and matrix interactions, conformance control/sweep efficiency, and reservoir simulation for improved oil recovery predictions. The specific areas covered in this report are:

1. Phase behavior: This included the effects of multiphase flow as a natural CO₂ mobility control system. The effect of multiphase flow on injectivity. Slim-tube, fat-tube, swelling, compositional analysis, and flow tests were used to examine phase behavior with the effects of flow behavior in core with and without fractures. Experiments to illustrate the actual process of waterflooding and CO₂ gravity drainage in a naturally fractured reservoir and the effects of permeability, of initial water saturation, and of injection schemes. Results validate the premises that CO₂ will recover oil from a matrix such as found in reservoirs in the Spraberry Trend Area of west Texas.
2. Foam for mobility control: Foaming agents were identified that possess selective mobility reduction (SMR), decreasing relative mobility more in the high permeability regions than in the low permeability regions. The effects of oil, heterogeneity, capillary contact, rock type, sacrificial agents, and mixed surfactant systems on CO₂–foam systems are presented. Mixed surfactant and sacrificial agent systems reduced the required good foaming agent; improved economics, improved mobility control, lowed surfactant concentrations, decreased adsorption requirements, improved SMR, increased injectivity, and increased oil recovery.
3. Brief field test review.
4. History match: A field CO₂ foam pilot test was successfully modeled using PRRC’s modified DOE pseudo-miscible model MASTER. A novel approach for performing history matching using small clusters of PCs and paralleling the history matching process is presented.
5. Field studies: The Wellman study reviewed the history of the field CO₂ flood and methods to improve CO₂ efficiency. The Teague-Blinebry study presents the possibility of a CO₂ flood in this low permeability (<1 md) reservoir.
6. Injectivity: A review that summarizes the literature that covers hypothesis and theories as to the causes and expectations of injectivity behavior in various CO₂ flooded reservoirs. A WAG injectivity forum reviewed a number of field cases of reduced water and CO₂ injectivity. Results of laboratory tests to data that attempt to identify and understand the causes of greater than expected reduced injectivity.

The U.S. Department of Energy has continued its support of improved oil recovery (IOR) research by carbon dioxide (CO₂) flooding at New Mexico Petroleum Recovery Research Center under the three-year grant: “Improved Efficiency of Miscible CO₂ Floods and Enhanced Prospects for CO₂ Flooding Heterogeneous Reservoirs.” This is the final report under contract No. DE-FG26-97BC15047. The research proceeded the areas of fluid and matrix interactions (understanding the problems), conformance control/sweep efficiency (solving the problems), and reservoir simulation for improved oil recovery (predicting results). All areas originate from research on oil recovery by high-pressure CO₂.

Phase behavior is an integral part of understanding and predicting behavior of CO₂ in a reservoir. The effect of multiphase flow creates a natural CO₂ mobility control agent that can significantly effect injectivity. Slim-tube, fat-tube, swelling, compositional analysis, and flow tests were used to examine phase behavior with the effects of flow behavior using consolidated core with and without fractures. Traditionally, CO₂ injection has been considered an inefficient method for IOR from natural-fractured reservoirs. Experiments in this work illustrate the actual process of waterflooding and CO₂ gravity drainage in a naturally fractured reservoir. The results demonstrate that CO₂ gravity drainage could significantly increase oil recovery after a waterflood. The effects of permeability, of initial water saturation, and of injection schemes were examined. Produced oil samples indicated that lighter components were, extracted and produced from tight matrix blocks early in the test. Results validate the premises that CO₂ will recover oil from a tight, unconfined matrix such as found in reservoirs in the Spraberry Trend Area of west Texas.

To aid in CO₂ mobility control in water alternating with gas injection systems (WAG), surfactant was added to the aqueous phase, creating foam that increases the apparent viscosity of the system. Foaming agents were identified that possess selective mobility reduction (SMR), decreasing relative mobility more in the high permeability regions than in the low permeability regions. The effects of oil, heterogeneity, capillary contact, rock type, sacrificial agents, and mixed surfactant systems on CO₂–foam systems are considered. CO₂-foam improved flood economics by significantly delaying breakthrough and improving oil recovery in relatively homogeneous single cores and heterogeneous parallel-connected composite core having two very different permeability regions, both with and without capillary contact. Mixed surfactant and sacrificial agent systems proved to reduce the required “good” (and more expensive) foaming agent. Mixed surfactant test results demonstrated synergistic effects that included improved mobility control, lower required good surfactant concentrations, lower adsorption requirements, better SMR, higher injectivity, and increased oil recovery. Surprisingly the low-cost, poor foaming lignosulfonates have shown great potential to reduce chemical cost by 75%, increase recovery efficiency, and increase injectivity in a mixed surfactant system.

A review of several field tests of CO₂ foam show that the foam diverted the gas stream, increased oil production, and decreased CO₂ production at offending wells. The pilot test in the East Vacuum Grayburg-San Andres Unit (EVGSAU) has been used to test foam models developed from our laboratory tests. DOE’s pseudo-miscible model MASTER was modified to simulate foam flooding. The simulated results of the foam test simulation are consistent with the field pilot results and are adequate for field scale CO₂-foam simulation. This report also summarizes a novel approach for performing history matching referred to as MASTER Web that provides an easy, effective way for paralleling the history matching process, at a fraction of the cost of using high-performance machines. Experiments on a small cluster of PCs, have demonstrated that a significant practical speedup (i.e., in terms of wall clock time) is achieved through using MASTER Web to perform the history match, thanks to adaptable parallel computing. The results are encouraging and indicate the tremendous potential of our proposed novel approach for oil companies.

Two field studies were performed. The Wellman study reviewed the history of the field CO₂ flood, the possibility of reducing CO₂ injection pressure to reduce required purchased CO₂, and the efficiency of CO₂ mobilizing reserves in the zone below the oil-water contact. The Teague-Blinebry study reviews previous work, contains water and CO₂ flood laboratory tests, defines reservoir pay for the field, and reviews the possibility of a CO₂ flood in this low permeability (<1 md) reservoir.

Based on the fluid flow properties of CO₂, one would expect that gas injectivity would be greater than the waterflood brine injectivity. However, in practice this behavior is often not the case. In addition, water injectivity is often lower than the waterflood brine injectivity. What is perplexing is that some reservoirs lose and others increase injectivity after the first slug of CO₂ is injected. These phenomena may occur on a local or field scale. A number of researchers have studied and proposed reasons for these phenomena over the past 20 years. This review summarizes the literature that covers hypothesis and theories as to the causes and expectations of injectivity behavior in various CO₂ and gas flooded reservoirs. A dozen field cases were presented in a 1999 WAG injectivity forum. All documented injectivity reduction for water cycle and many for CO₂. The attempted solutions were shown to be uneconomical, inconsistent, or negative. As a result of these discussions laboratory tests are in progress to attempt to identify and understand at least some of the phenomena.

As a result of the injectivity forum and the WAG injectivity literature review a laboratory study was initiated. Two reservoir rock types, dolomite and limestone, were examined. Both systems exhibit dissolution. In the dolomite core, dissolution of anhydrites occurred during the brine flood while dolomite dissolution occurred during the WAG cycles. In the limestone core, calcite dissolved during the WAG process. Recrystallization was detected in the longer limestone core, resulting in decreased downstream permeability. In neither test system was there evidence that oil contamination caused permanent permeability reduction. In each case the system returned to pre-oil conditions after CO₂ was injected into the system.