Summary A pilot test of microemulsion flooding was conducted in a single five-spot pattern in the Chateaurenard field in France. The test had to accommodate a 40-mPa·s (40-cp) oil viscosity and a regional pressure gradient across the pattern. A very clear oil bank was observed, resulting in a substantial increase in oil production. Introduction Société Nationale Elf Aquitaine (Production) is conducting a microemulsion/polymer flooding pilot test in its Chateaurenard field in the southern part of the Paris basin. This project involves the participation of the Inst. Franç du Pétrole, and it has been partially funded by the Commission of the European Communities. The primary objectives of the pilot are to (1) demonstrate the efficiency of the process under field conditions, (2) develop the technology involved in such an operation, and (3) assess the economics of fieldwide extension. The permeability of the unconsolidated sand reservoir is very high, and the connate water is fresh. Both of these factors are favorable for micellar flooding application. The oil is relatively viscous [40 mPa·s (40 cp)]. Even though the oil saturation is far from residual, the well effluents have a high water cut due to poor areal efficiency of oil displacement by water. A microemulsion flooding scheme, with its readily achievable mobility control, appeared to offer a feasible recovery mechanism. However, the injection of viscous fluids limits the injection rate and requires a higher concentration of chemicals. The Chateaurenard oil viscosity is above the limit of 20 to 30 mPa·s (20 to 30 cp) mentioned in some publications.1,2 The well arrangement is a 10´103-m2 (2.5-acre) inverted five-spot pattern with one observation well. The pilot area was determined through coring, logging, and interference testing. These data indicated an initial pressure gradient across the pattern. Three separate injection phases compose the process: microemulsion, polyme, and chase water. Microemulsion was formulated using field crude and water. Injection was initiated in Feb. 1978 and followed by polymer concentration started in April 1979 and led to chase water in Nov. 1979. The polymer buffer has been overdimensioned somewhat to test specifically the efficiency of the microemulsion, without interference from poor mobility control following it. A significant response was observed from the producing wells 2 months after the start of operations. In this paper we present pilot design, performance, and an interpretation of the results. Field Description The Chateaurenard field, outlined in Fig. 1, is part of the Neocomian (Lower Cretaceous) oil reservoirs found in the southern part of the Paris basin. They are located 100 km (62 miles) south-southeast of Paris; the oil-bearing zones extend over an area of 20 km2 (7.7 sq miles).

Early results of a successful polymer project started in 1985 in the Courtenay sand of the Chateaurenard field located south of Paris, France, were presented at the fall 1988 SPE meeting. We update these data and show that subsequent performance has confirmed the highly favorable oil recovery behavior of this project. The objective of this paper is to report the results of a detailed compositional simulation study conducted to understand this favorable result. Both alternative reservoir and process characteristics were investigated in these three-dimensional simulations. In particular, both a classical layered reservoir description and a geostatistical reservoir description were investigated. The latter represents a new approach to the interpretation of polymer flooding and played an important role in our study since the attenuation of reservoir heterogeneities is known to be an important attribute of polymer flooding. We also evaluated the importance of permeability reduction and adsorption of the polymer on process performance. After obtaining a good match of the production of oil, water and polymer, we then investigated the importance of design factors on the oil recovery performance. The actual polymer flood consisted of injecting a large polymer slug followed by five smaller slugs of decreasing polymer concentration to decrease the viscosity contrast with chase water, a classical polymer flooding design method called polymer grading. We found that this grading really made very little difference in the simulated oil recovery indicating that fingering was of little importance under these conditions. This was because the oil recovery - was dominated by other factors such as polymer adsorption, fractional flow, and heterogeneity. Remarkably, this appears to be the first time that a study of this type with an accurate compositional simulator has been reported for polymer flooding. The combination of better reservoir characterization methodology and accurate process simulation of polymer flooding should aid in the successful exploitation of this technology in the future. This project clearly shows the high potential of polymer flooding under the appropriate reservoir conditions and with good design and implementation methods. Introduction A polymer flood pilot test in the Courtenay sand of the Chateaurenard field was started in 1985. This pilot test is characterized by high oil recovery (1.5 bbl oil per lbm polymer) and high retention of polymer in the reservoir. The performance of this pilot was simulated using a compositional chemical simulator called UTCHEM developed at The University of Texas. Actual field data and available physical property data were used as much as possible to evaluate simulation input parameters. UTCHEM is a three-dimensional, compositional, multicomponent, multiphase, finite-difference simulator. A third-order correct in space finite-difference method with a flux limiter is used to approximate the partial differential equations. This method reduces both numerical dispersion and grid orientation to very low levels without any problems with numerical oscillations even at very high cell Peclet numbers. This enables us to accurately model the flow of polymer solution in permeable media including in particular the full dispersion tensor. UTCHEM includes models for polymer adsorption, non-Newtonian viscosity, permeability reduction and inaccessible pore volume. P. 781^

The Courtenay polymer flooding test is being completed, using a larger than usual slug size ; a special attention was given to monitoring the concentration and quality of the effluents. The test is implemented on a ten acres inverted five spot in the center of the field of Courtenay, which is part of the Chateaurenard field. A rather large slug of partially hydrolized polyacrylamide was injected, consisting of 42% pore volume of a 1000 ppm active polymer concentration solution followed by a 39% pore volume solution at a tapered concentration. By matching the oil cuts and polymer concentrations in the effluents of the producing wells we estimate that the producing wells we estimate that the adsorption of the polymer on the rock was 35 mug/g of rock. Concentrations and main characteristics of produced polymer at the wellhead were produced polymer at the wellhead were determined using a sampling and separation methodology which prevents chemical and mechanical degradation. An increase in polymer molecular weight versus time has polymer molecular weight versus time has been observed, this polymer is not harmed by flow in the reservoir. Introduction Three tests have been implemented in the Chateaurenard field before the Courtenay test :A large polymer pilot (ref-1), on a 7 spot configuration, with one injector in the center and 6 producers at a distance of 400 m. A horizontal well, which extends 341 m inside the reservoir, was drilled in the northern edge of the pattern in 1985. The pore volume is about 800 000 M. Altogether a total of 280 000 m of polymer solution has been injected, polymer solution has been injected, followed by 410 000 ma of chase water. Total oil production is 86 000 M, of which 29 500 M is considered as tertiary oil.A research microemulsion pilot, which was run on a small 100 m spacing inverted five spot (ref. 2 and 3). The injection of a microemulsion composed of high molecular weight sulfonates and C4-C5 alcohols started in 1978. It was displaced by a viscous polymer solution. Definite oil banks were observed on 3 of the 4 producing wells. We estimate that 70% of producing wells. We estimate that 70% of the oil in place in the swept areas has been produced.An industrial microemulsion pilot (ref. 4 and 5). The pilot, which is nearing completion, is an extension on a large spacing, with an economically designed microemulsion, of the former pilot. The pilot pattern is composed of four adjacent inverted five spots, distance between producers is 280 m (913 ft)- The volume of the micellar system was 3.5% of the enclosed pore volume (224 000 m)- 40% PV of a partially hydrolized PV of a partially hydrolized polyacrylamide solution concentrated at polyacrylamide solution concentrated at 1700 ppm was injected followed by the same volume with a tapered concentration and then chase water. The total injected volume was 247 000 M or 110% pore volume. P. 395

No abstract

The polymer injection in the Courtenay field has reached a mature st?ge as a large part of the polymer slug has been injected. The total production of oil confirms the original expectations and the operation is profitable. The Courtenay field is an elongated layer at a depth of 600 m. The oil is viscous (40cP), water is fresh and the temperature is 30"C, making the conditions favorable for a polymer injection. The area swept by polymer encompasses 640000 ma. It consists of three contiguous zones with a total of 4 injectors and 18 producers. Injection of a 900 ppm solution of polyacrylamide diluted in field water started in July 1989 at a total rate of 380 m3/day. Up to now, 640000 ms (100 % pore volume) o{ solution have been injected, polymer injection was terminated in November 1993; since then, water is injected at a rate of 420 m3/day. Compared to primary production, a strong increase in oil production has been observed. Definite tertiary oil banks breakthroughs have been observed at 14 producing wells. The operation has been simulated with the Ecorepol model, a 3 dimensional 3 phases simulator with polymer and tracers option.The total oil production is well represented on a global basis but the actual quantity of polymer produced in the effluent water is lagging behind the value expected by simulation.