Analysis of the Low-Tension Pilot at Big Muddy Field, Wyoming

Abstract

Summary. Conoco Inc, initiated a low-tension pilot test in 1973 at the Big Muddy field east of Casper, WY. The process mobilized an oil bank ahead of the slug, reached a peak oil cut of 20%, and recovered 36% of the residual oil saturation (ROS). Tracers were injected in the preflush and postflush to determine swept volumes and distributions of flow. Tracer responses showed that more than 95% of the flow to the center well of the five-spot came from the two northern injection wells. The overall performance of the pilot was analyzed with a numerical simulator. pilot was analyzed with a numerical simulator. As a result of the pilot and the supporting research, a nearby 90-acre [36.4-ha] commercial demonstration low-tension flood was initiated in 1980 and was operated as an EOR project until July 1985, although oil production still continues. Introduction The pilot test was conducted at the Big Muddy field between 1973 and 1978. Objectives of the pilot wereto determine the oil recovery for a known low-tension slug size and reservoir swept volume;to calibrate a numerical simulator for application to future projects; andto develop process design technology for projects; andto develop process design technology for field-scale projects. Oil recovery by the low-tension process was analyzed with the help of a reservoir simulator. Tracers were injected during the preflush and postflush periods to determine flow distributions and preflush and postflush periods to determine flow distributions and swept volumes. Calculation techniques and highlights of the design and operations are presented. Finally, the results of a single-well surfactant test (SWST) conducted in the Big Muddy field are compared with the pilot results. Design and Operation Highlights Reservoir Description and Operating Problems. The Big Muddy Wall Creek reservoir is an anticlinal structure with edgewater encroachment. The pilot area is located near the anticlinal crest in the center of what was the best waterflood area. Four injection wells and a center producing well were drilled inside four existing wells. The five-spot was about 1 acre [0.4 ha], and the total pilot area was 5.5 acres [2.2 ha], as shown in Fig. 1. Formation data were obtained from a comprehensive logging and coring program that included oriented cores and a sonic borehole televiewer log. These data were complemented with pressure-transient testing and a single-well tracer test (SWTT) to pressure-transient testing and a single-well tracer test (SWTT) to determine ROS. 3 These data were used to achieve a balanced, confined five-spot operation. 4 Preflush and postflush tracers were injected to help determine the distribution of flow and swept volumes. Well J30, the center well of the five-spot, produced 25 % of the total injection rate. Although the total producing rate from Well J30 remained nearly constant, injection well rates were adjusted at the end of the preflush as shown in Table 1. The adjustment was based on tracer data that were later determined to be erroneous as a result of microbial degradation of the ethanol tracer. The adjustment, rather than balancing the five-spot, actually resulted in a less-balanced pattern. Because of the likelihood of microbial degradation of the ethanol tracer in the preflush and concern that polysaccharide and dilute isobutyl alcohol (IBA) would also act as substrates, efforts for microbial control were intensified at the start of the low-tension slug injection. After supply water was chlorinated, caustic was added to increase the pH to a range of 12.5 to 13.0. For the remaining project life, microbial cultures were routinely obtained from project life, microbial cultures were routinely obtained from production and injection sites. Bacteria were not observed in any of production and injection sites. Bacteria were not observed in any of the cultures on the injection side. Ethanol and C were injected as tracers in Well 13 1. Ethanol tracer degraded during the post-flush, however, as evidenced by a smaller fraction of the ethanol post-flush, however, as evidenced by a smaller fraction of the ethanol tracer that was recovered relative to the C tracer. Average injection or producing rates were only about 100 B/D [15.9 m3 /d], or about half of the desired rate without stimulation. Therefore, the injection wells and center producing well were fractured and propped with 2,000 lbm [907 kg] of sand. The small sand volume was designed to produce propped fractures of 40 ft [12.2 m], or about 15% of the well-to-well distance. Greater fracture lengths carried the risk of unacceptable loss of oil recovery if the fracture direction and the pattern orientation were misoriented more than 15 [0.26 rad].4 An unusual feature of the Wall Creek formation is that the parting pressure determined from step-rate injectivity tests was about 75 pressure determined from step-rate injectivity tests was about 75 to 150 psi [517 to 1034 kPa] less than hydrostatic pressure. Downhole pressure sensors similar to tank-level bubblers were used to monitor and to control maximum injection pressure. The bubbler depths were set slightly below a depth corresponding to the interpreted parting pressure. For pilot interpretation, this system was limited because there were no pressure measurements when the fluid levels were below the bubblers. As previously stated, the rates of the injection wells were changed before surfactant injection was begun. While all four injection rates were initially about 200 B/D [31.8 m /d], the rates of Wells J31 and W79 were increased 50 % to 300 B/D [47.7 m3/d], and Wells J14 and S27 were decreased 50% to 100 B/D [15.9 m3/d]. At the same time, the rates of the offset producers were also increased to maintain lower reservoir pressures. During interpretation of pilot performance, it was pressures. During interpretation of pilot performance, it was concluded that two wells, Wells W79 and J31, were unintentionally pressure parted after the rate changes. Therefore, because of pressure parted after the rate changes. Therefore, because of operating conditions that likely lowered the parting pressure below the operating pressure and because of the rather drastic rate changes, the fractures in the two wells were probably extended. As discussed later with the SWST, when the rates of off-set producers were raised to increase drawdown, we observed that pressure parting could occur as low as 960 to 980 psi [6620 to 6760 kPa]. During the development of a reservoir description for simulating the pilot performance, it was independently concluded that the fractures of only performance, it was independently concluded that the fractures of only these two wells were extended. Low-Tension Process. The selection of chemicals and the chemical composition of five sequential slugs of the low-tension process were previously reported. The injection plant design provided for blending the injected slugs to specified concentrations and viscosities within tolerances of 0.1% concentration and 0.5 cp [0.5 × 10 Pa s]. Precoat filtration was used with all injected fluids to minimize formation plugging. SPEFE P. 315