The Effect of Pressure on Oxidation Kinetics of Tar from a Tarmat Reservoir

The oxidation kinetics of a tar with physical and chemical characteristics similar to those of a reservoir tar were studied employing a variable-temperature oxidation reactor. Mixed with clean, loose sand, the tar showed oxidation behavior typical of heavy crudes with LTO and HTO peaks in oxygen consumption. Higher pressures caused larger LTO-oxygen consumption, lower HTO-oxygen consumption, lower HTO-peak temperatures, higher apparent H/C ratio of fuel, and lower HTO activation energy. All these effects are attributed to suppression of light-end evaporation at low temperatures. Compared with clean sand, natural crushed-core material promoted LTO of the tar but did not alter HTO parameters significantly. With HTO-peak temperatures and activation energies above 500 oC and 100 kJ/mol, respectively, the tar is not expected to provide for sustained in-situ combustion in the reservoir. INTRODUCTION Tarmat is a loosely defined term ascribed to a layer of very viscous tar or bitumen that exists around the periphery of some oil reservoirs, usually at the oil/water contact (Tripathy, 1988). Reservoirs displaying various extents of this feature are termed tarmat reservoirs. Such reservoirs are encountered throughout the world and, particularly, in the Middle East where a typical case is the Arab-D reservoir of the super-giant Ghawwar field (Osman, 1985). The tar deposit in the Uthmaniya area only of this carbonate reservoir is estimated at 400 million m. The immediate consequence of the tarmat is deprivation of the reservoir’s oil zone of adjacent aquifer support. In severe cases where the tarmat totally surrounds the oil zone, the reservoir behaves like a finite lens with rapid drop in pressure accompanied by an alarming increase in production gas/oil ratio during primary depletion (Osman, 1985). In other cases where the tar has some mobility, the pressure differential across the tarmat could build up to a level that might cause the tar seal to beak down rather abruptly allowing severe water coning into nearby wells. Methods for removal or dissipation of the tarmat have been investigated in the laboratory but not field tested yet. One study involved water injection below the tarmat (Abu-Khamsin et al., 1993). This simple technique proved ineffective as extremely high pressure gradients were required to breach the tar layer, and significant amounts of oil were bypassed by water fingering through the tarmat leading to poor oil recovery. Another study experimented with injecting various solvents, all driven by hot water (Okasha et al., 1998). The results showed efficient displacement of tar at an optimum slug size for each solvent; and maximum oil recovery was achieved when the optimum solvent slug was injected in portions alternating with hot water. The economics of the process, however, are yet to be evaluated by a field test. It is expected, though, that well-bore heat losses would render such thermally-assisted, miscible displacement technique rather ineffective in the deep reservoirs of the Middle East. This leaves in-situ combustion with its high heat efficiency and low cost as a viable, alternative solution to this problem. In-situ combustion has been applied successfully for a number of heavy oil reservoirs. Description, mechanisms and requirements of the process are well documented in the literature (Butler, 1991) and need not be repeated here. A crucial factor that influences performance of in-situ combustion is availability and reactivity of the fuel. Insufficient fuel deposition ahead of the combustion front or a slow rate of high-temperature oxidation (HTO) within the front deprive the front of the heating rate necessary to sustain the process leading to premature extinction. A summary of the effects of various process variables on the amount and nature of the fuel was provided (Abu-Khamsin et al., 1988). While medium and heavy crudes usually satisfy the fuel requirement (Alexander et al., 1962), and tar is expected to conform to this rule, the rate of HTO of the fuel as dictated by oxidation kinetics becomes the controlling factor (Shahani and Hansel, 1987). Under ideal conditions, the HTO reaction proceeds in the following manner: Fuel + O2 CO + CO2 + H2O (1) The rate of this gas-solid reaction is modeled by a simple kinetic equation: RHTO = kHTO PO2 Cf (2) Where RHTO is the rate of HTO of fuel (g per cm of rock per s), kHTO is HTO reaction rate constant (consistent units), PO2 is partial pressure of oxygen (Pa), Cf is fuel concentration (g per cm of rock), m is reaction order with respect to oxygen, and n is reaction order with respect to fuel. As governed by Arrhenius law, the HTO reaction rate constant is expressed as: kHTO = A exp(-EHTO /RT) (3) Where A is a frequency factor (consistent units), EHTO is HTO reaction’s activation energy (kJ/mol), R is the universal gas constant, and T is temperature (oK). As Equs. 2 and 3 indicate, values of four kinetic parameters (A, EHTO, m and n) are required before the rate of HTO can be estimated and, thus, feasibility of combustion can be assessed. A survey of published kinetic data as well as a report on new data for mostly heavy crudes was provided (Fassihi et al., 1984, p. 399). In that data, order n varied slightly around unity but order m showed wide variation. Wider variation was found with A, but most investigators presume it to depend upon the specific surface area of the porous medium. EHTO is largely a property of the fuel or its parent oil (Fassihi et al., 1984, p. 399) with more carbonaceous fuels showing higher energies. Given sufficient fuel, combustion can proceed smoothly at normal combustion temperatures if EHTO is low to medium in magnitude. If this energy were high, a large A, a higher oxygen pressure or a higher combustion temperature would be needed to carry the process forward. Since heavier crudes tend to show higher EHTO, based on this parameter only tars would make poor candidates for in-situ combustion. HTO reaction kinetics are studied employing variable-temperature oxidation reactors operated under a wide range of conditions. In such reactors, progress of various oxidation reactions is followed by monitoring the reactor’s temperature and composition of its effluent gas. A graphical technique to extract HTO kinetic parameters from the reactor’s data was described (Fassihi et al., 1984, p. 408). Such approach has been adopted in this study whose purpose was to investigate HTO kinetics of a tar similar in characteristics to the tar deposit found in the Arab-D reservoir of Ghawwar field. The results would help provide a preliminary assessment of the feasibility of in-situ combustion as a tar displacement method. EXPERIMENTAL WORK Materials Due to the scarcity of reservoir tar samples, the tar used in this study was prepared by distilling Ghawwar Arab-D crude under nitrogen atmosphere to yield a residue of physical and chemical properties similar to the reservoir tar. Such approach is justified by results of mass spectroscopy and other tests, which showed that the natural tar and Ghawwar oil are chemically similar and genetically related, and that the tar had formed from the oil by gas de-asphalting over geologic time (Riley et al., 1977). Table 1 lists physical and chemical properties of the prepared tar as well as average properties of 6 tar samples obtained from 3 wells in Uthmaniyah area of Ghawwar. Table 1: Physical and Chemical Properties of Tars. Property Tar Source Prepared Reservoir Density @ 15 C, g/ml 0.922 0.976 Density @ 104 C, g/ml 0.873 0.919 Viscosity @ 15 C, cP 10322 12400 Viscosity @ 104 C & 2000 psia, cP 23 Sulfur content, wt% 3.25 3.75 Nitrogen content, wt% 0.25 Carbon content, wt% 81.8 82.2 Hydrogen content, wt% 11.8 11.5 Atomic Hydrogen/Carbon ratio 1.73 1.68 Saturates, wt% 32.8 30.0 Aromatics, wt% 32.4 34.8 Resins, wt% 9.0 9.6 Asphaltenes, wt% 25.8 25.6 * Average of 6 samples from 3 wells in Uthmaniyah, Ghawwar. Two types of porous media were used: clean dune sand and crushed ArabD core material both thoroughly washed with distilled water and cleaned with toluene. Sieve analyses of both media is listed in Table 2. Setup Figure 1 shows a schematic of the apparatus, which consisted of gas supply and control, oxidation reactor and heating oven, reactor temperature measurement, a combustion analyzer for effluent gas analysis, and a data acquisition system. The oxidation reactor (Fig. 2) was made of a thick-walled, stainless steel pipe measuring 12 cm (length) x 3.18 cm (OD) x 2.54 cm (ID). The pipe was sealed at both ends by flanges and copper O-rings. A 1.6-mm OD, stainless steel tube fitted through the top flange and reaching down to 2 cm above the reactor’s bottom flange served as a thermowell, which housed a thermocouple. Table 2: Sieve Analyses of Porous Media. Screen Mesh Size Wt% Retained Dune Sand Crushed Core 30 0 0 40 0.39 0.88 50 60.10 23.72 60 20.11 14.84 100 6.62 18.93 140 5.11 11.34 200 6.94 12.78 Fines 0.73 17.51 100.00 100.00 * Cores from Arab-D reservoir, Ghawwar. Procedure A typical run began by packing a mixture of about 2 g of tar and 25 g of porous medium into the reactor. About 20 g of clean sand were placed on top of the mixture to help preheat and distribute air flow. The reactor was then assembled, pressure tested with nitrogen, placed into the oven, and connected to the flow system. After pressurizing with nitrogen to the desired level and attaining an initial temperature of 100 oC, the oven’s heating program was started at a rate of about 70 oC/h. Simultaneously, air flow through the reactor was initiated and maintained at 760 std. ml/min. During a run, the tar-sand mixture’s temperature and combustion analyzer readings were recorded continuously. A run was terminated