Experiments have been performed in a linear near-adiabatic system for the purpose of extending data on reverse and forward combustion from atmospheric pressure to 1,000 psig.Results obtained from reverse combustion appear to conform qualitatively with the existing description of the process when reaction kinetics are suitably modified to account for the increase in pressure, z. e., increasing the pressure decreases the peak temperature and increases the combustion zone velocity.Forward combustion appears to be a fuel dominated process wherein peak temperature and combustion zone velocity are not very sensitive to changes in pressure. The moderate effects of pressure that do exist at low flux virtually disappear at high flux providing all oxygen is consumed. With this provision, increasing the pressure decreases the frontal velocity and increases the peak temperature.Results are shown graphically which demonstrate the effects of pressure on peak temperature, rate of advance, oil recovery, air/oil ratio, carbon oxides produced and temperature distributions. Introduction Numerous field tests of oil recovery using the technique of underground combustion are now in progress. Operating pressures used are always substantially in excess of atmospheric pressure. Nevertheless, the literature contains no laboratory data pertaining to the effects of pressure except for those of Martin, Alexander and Dew. Unfortunately, these data appear to reflect heat losses which may have obscured some of the effects of increased pressure.The underground combustion processes are exceedingly complex, and general concepts relating to them have necessarily been described in relatively simple terms. Particularly with regard to forward combustion, the interaction between multiphase fluid flow, heat transport and various rate mechanisms is so involved that to date a rigorous treatment is not available.In this paper a linear near-adiabatic system is described which was found capable of operating at pressures up to 1,000 psig and temperatures of at least 1,100 F. Using this equipment, experiments on both forward and reverse combustion were accomplished at elevated pressures and the data presented as a function of air flux or temperature.It will be assumed that the reader is familiar with the papers of Wilson, Wygal, Reed, Gergins and Henderson on forward combustion and that of Reed, Reed and Tracht on reverse combustion, so that it is not necessary to repeat the details of the two processes. EQUIPMENT AND PROCEDURE It was decided to study the effects of pressure over the interval 0–1,000 psig using a near-adiabatic system. Since satisfactory operation of such a system requires very thin tubing walls, it proved necessary to insert the entire combustion tube and heater assembly into a high pressure jacket in such a way that the internal tube pressure could be counterbalanced by pressurizing the annulus. This meant that the numerous heater and thermocouple wires had to pass through gas-tight seals that required considerable effort to maintain effective.An unexpected difficulty arose in connection with the nature of the annulus pressuring gas. It turned out that the composition of this gas critically affected the functioning of the adiabatic control system.Operation of the equipment at elevated pressures was not particularly difficult. The problems that did arise were a direct consequence of the large numbers of connections and seals involved. EQUIPMENT The equipment consisted of the combustion tube which fitted within the pressure jacket. SPEJ P. 127^

Terwilliger, P.L., SPE-AIME, Gulf Research and Development Co., Clay, R.R., Gulf Research and Development Co., Wilson Jr., L.A., SPE-AIME, Gulf Research and Development Co., Gonzalez-Gerth, Enrique, SPE-AIME, Gulf Research and Development Co. In 1964 Mene Grande Oil Co. began a fireflood in a sand reservoir of the Miga field in Eastern Venezuela. It was expected that only 5 percent of the 13 degrees API gravity oil would be recovered by primary means; the fireflood has recovered more than twice that amount. No serious operating problems have been encountered. problems have been encountered. Introduction In 1964 Mene Grande Oil Co. started a fireflood in the P2-3 sand reservoir in the Miga field of Eastern Venezuela. The project has continued since that time. The original oil in place was estimated at 23.2 million bbl and 1.2 million bbl, or 5 percent, was expected to be produced by primary depletion. To date, an additional 2.6 million bbl, or more than twice the primary production, have been recovered by the use of the fireflood process. The air injection rate has averaged about 10 MMcf/D over the 9-year life. The average air/oil ratio (AOR) has been 11,000 cu ft/bbl. No serious operating problems have been encountered during the fireflood. The loosely consolidated sand is controlled through use of pressure-gravel-packed liners. Corrosion has not been a problem. No water injection has been used for producing well cooling, although a lighter oil is used for down-the-hole blending to increase the producing rates and facilitate the surface handling of the oil. Past performance and sweep pattern studies indicate that fireflooding could result in the production of 50 percent of the original oil in place, whereas the ultimate percent of the original oil in place, whereas the ultimate primary recovery would be only 5 percent. Experience both primary recovery would be only 5 percent. Experience both in this reservoir and other similar ones had shown that gas drive and waterflooding were completely ineffectual. Reservoir Description The project was performed in the P2-3 sand, MG-517 reservoir, of the Miga field located in Eastern Venezuela. A structure-isopach map of the project reservoir appears in Fig. 1. Fig. 2 includes a summary of the reservoir properties. This reservoir is one of several in the P2-3 properties. This reservoir is one of several in the P2-3 channel sand, which is found in scattered locations throughout both Miga and the neighboring Oleos fields. The updip seal is a combination of faulting and sand thinning. Lateral limits are considered to be the 10-ft isopach, as indicated in Fig. 1. The downdip limit is formed by a fault and the original water-oil contact. Reservoir volume is estimated at 18,600 acre-ft. Well MG-525 was high-GOR when completed, suggesting a small initial gas cap. At the time this well was selected for air injection, the south boundary of the reservoir was believed to be a fault located just south of the well. The revised interpretation of the reservoir indicates this is not the case; the reservoir actually extends much further south, as indicated in Fig. 1. Depth of the reservoir ranges from 4,000 to 4,350 ft below ground level, with dip to the north of about 2 degrees. The sand is loosely consolidated, with a porosity of 22.6 percent and an estimated average porosity of 22.6 percent and an estimated average permeability of 5 darcies. Maximum sand thickness is permeability of 5 darcies. Maximum sand thickness is about 25 ft. Connate water saturation is about 22 percent; stock-tank oil originally in place was 23.2 percent; stock-tank oil originally in place was 23.2 million bbl. P. 9