Infill Drilling in a Steamflood Operation: Kern River Field

Abstract

Summary. An infill well program involving 574 wells in previously steamflooded idle reservoirs proved to be economical and increased recovery in some areas from 50 to 58%. These wells were recompleted to zones under active steam drive and showed similar recovery increases in addition to acceleration of reserves. These increases in recovery were attained without additional fuel. Introduction The Kern River field is a large, shallow, heavy-oil deposit located 5 miles [8 km] northeast of Bakersfield, CA. The productive formation is a sequence of sands called the productive formation is a sequence of sands called the Kern River series. The structure is a simple homocline, dipping southwest at 5 deg. [0.09 rad]. The updip sands pinch out, and downdip is bounded by an oil/water contact. These fine-to-very-coarse sands, averaging 60 ft [18.3 m] in thickness, are separated by silt and clay interbeds. The unconsolidated sands have high permeabilities of 1 to 5 darcies and porosities of 28 to 33 %. The average oil viscosity and reservoir temperature are 4,000 cp and 85 deg. F [4 Pa.s and 29 deg. C], although viscosities vary between sands from 2,000 to 40,000 cp [2 to 40 Pa.s]. Texaco Inc.'s Kern River field steamflood involved more than 3,640 producing wells and 1,875 injection wells before infill drilling. The field had been developed on 2 1/2-acre [10 100-m2] five-spot patterns. Up to seven sands are swept, one at a time, usually beginning in the lowest sand. These sands range in depth from 300 ft [91.4 m] in the northeast section of the field to 1,200 ft [366 m] in the southwest. The reported Kern River steamflood oil recovery ranges from 42 to 73% of presteam oil in place (OIP), with an average of 50% a generally place (OIP), with an average of 50% a generally accepted recovery efficiency. We have completed steamfloods in 1,500 pattern sands. Some of the techniques for improving steamflood recovery were summarized by Bursell. They include infill wells, partial or limited perforation of producers, variable steam quality and rate, injection diverting, crossflooding, and injector-to-producer conversion. Before infill wells were drilled, numerical simulation work indicated several attractive benefits from infill wells:recovery increases from 50 to 60% of OIP even with well spacing as small as 0.625 acre [2530 m2],possible similar recovery increase from the 1,500 completed and idle pattern sands, andaccelerated recovery and fuel pattern sands, andaccelerated recovery and fuel reduction. These results have been proved. The following review of four projects will compare theoretical to actual results. The 574-infill-well program was unusual in that most of the wells were first completed in idle sands and were subsequently perforated in upper active steamfloods and the well spacing was reduced to only 0.625 acre [2530 m2] per well. The patterns went from 1.25- to 0.625-acre [5060- to 25 30-m2] well spacing by conversion of a 2 1/2 -acre [10 100-m2] five-spot into a 2 1/2-acre [10 100-m2] nine-spot pattern. 1970 Canfield Project The 1970 Canfield project covers 148 acres [599×103 m2] in the updip portion of the Kern River field, as shown in Fig. 1. A cross section through the project (Fig. 2) shows the steamflood intervals. Table 1 shows reservoir properties. Steam injection in the first displacement sand, Sand R1, began in July 1970. Injection was ended at the steamflood economic limit in Sept. 1977. Fig. 3 shows project performance history. Cumulative steamflood oil recovery performance history. Cumulative steamflood oil recovery from the project was 147,000 STB [23 400 stock-tank m3] per pattern, or 57% of the predisplacement OIP, as shown in Table 2. In Feb, 1978, the project was recompleted upward to Sand R. Conduction heating from the Sand R1 displacement raised Sand R temperature from 85 to 165 and 92 deg. F [29 to 74 and 33 deg. C] at the base and top of Sand R, respectively (Fig. 4). Previous work 4 reported the beneficial effects of this preheating. Steam injection in Sand R was terminated in the 55 patterns during the period from Oct. 1982 to June 1984. Again, injection was ended in these patterns at the steamflood economic limit. Steamflood oil recovery was 97,200 STB [15 450 stock-tank m3] per pattern, or 49%, as shown in Table 2. During this same 1982–84 period, recompletion and injection into Sand K (above Sand R) was started. Concurrently, 80 infill wells were drilled and initially completed in the idle Sand R1. As shown in Fig. 5, these wells were drilled at midpoints of the pattern boundaries. The average production rates for these infill wells are shown in Fig. 6. SPERE P. 243

Reservoir Characterization, Production Characteristics, and Research Needs for Fluvial/Alluvial Reservoirs in the United States

“…The object of Mill drilling is to recover the oil that would otherwise be bypassed by the injected steam. Texaco was the first operator to practice infill driJJing in a stearnflood (Restine et al 1987). Although in.fill drilling accelerated production and recovered oil fkom the so-called blind spots (pockets of bypassed oil), some steam injection practitioners think it does not increase the ultimate recovery Decisions to drill inlill wells have historically been based on accelerated economics, not on inmemental recovery (BIevins 1990).…”

Research needs for strandplain/barrier island reservoirs in the United States

“…The object of infill drilling is to recover the oil that would otherwise be bypassed by the injected steam. Texaco was the first operator to practice infill drilling in a steamflood (Restine et al .,. 1987).…”