If all gas in reservoirs was methane and all oil was decane and heavier, the PVT properties of the reservoir fluids would be quite simple because the quantities of oil and gas obtained from a mixture of the two would be almost independent of the temperatures, the pressures, and the type of the gas liberation process by which the two are separated. Low volatility crudes approach this behavior, which is approximately indicated by reservoir temperatures below 150°F, solution gas-oil ratios below 500 SCF/STB, and stock-tank gravities below 35 °API. Because the propane, butane, and pentane content of these fluids is low, the volatility is low.
For the conditions listed previously, but not too far above the approximate limits of low volatility fluids, satisfactory PVT data for material balance use are obtained by combining separator tests at appropriate temperatures and pressures with the flash and differential tests according to the procedure discussed in the previous section. Although this procedure is satisfactory for fluids of moderate volatility, it becomes less satisfactory as the volatility increases; more complicated, extensive, and precise laboratory tests are necessary to provide PVT data that are realistic in the application, particularly to reservoirs of the depletion type.
With present-day deeper drilling, many reservoirs of higher volatility are being discovered that include the gas-condensate reservoirs. The volatility is higher because of the higher reservoir temperatures at depth, approaching 500°F in some cases, and also because of the composition of the fluids, which are high in propane through decane. The volatile oil reservoir is recognized as a type intermediate in volatility between the moderately volatile reservoir and the gas-condensate reservoir. Jacoby and Berry have approximately defined the volatile type of reservoir as one containing relatively large proportions of ethane through decane at a reservoir temperature near or above 250°F, with a high formation volume factor and stock-tank oil gravity above 45 °API.9 The fluid of the Elk City Field, Oklahoma, is an example. The reservoir fluid at the initial pressure of 4364 psia and reservoir temperature of 180°F had a formation volume factor of 2.624 bbl/STB and a solution gas-oil ratio of 2821 SCF/STB, both relative to production through a single separator operating at 50 psig and 60°F. The stock-tank gravity was 51.4 °API for these separator conditions. Cook, Spencer, and Bobrowski described the Elk City Field and a technique for predicting recovery by depletion drive performance.10 Reudelhuber and Hinds and Jacoby and Berry also described somewhat similar laboratory techniques and prediction methods for the depletion drive performance of these volatile oil reservoirs.11,9 The methods are similar to those used for gas-condensate reservoirs.
A typical laboratory method of estimating the recovery from volatile reservoirs is as follows. Samples of primary separator gas and liquid are obtained and analyzed for composition. With these compositions and a knowledge of separator gas and oil flow rates, the reservoir fluid composition can be calculated. Also, by recombining the separator fluids in the appropriate ratio, a reservoir fluid sample can be obtained. This reservoir fluid sample is placed in a PVT cell and brought to reservoir temperature and pressure. At this point, several tests are conducted. A constant composition expansion is performed to determine relative volume data. These data are the flash liberation volume data listed in Table 7.3. On a separate reservoir sample, a constant volume expansion is performed while the volumes and compositions of the produced phases are monitored. The produced phases are passed through a separator system that simulates the surface facilities. By expanding the original reservoir fluid from the initial reservoir pressure down to an abandonment pressure, the actual production process from the reservoir is simulated. Using the data from the laboratory expansion, the field production can be estimated with a procedure similar to the one used in Example 5.3 to predict performance from a gas-condensate reservoir.
Leave a Reply