PVT studies provide window into the future

Better data will lead to better decisions

T.Guidry
A.Elfaki
T.Meche
R. Huval
R.Coleman
D.Leger
J.Leblanc
Core Laboratories

Reservoir fluid pressure-volume-temp-erature (PVT) studies are nothing new to reservoir engineering. In fact, PVT labs have been around for 50 years or more. They provide a window into the future behavior of hydrocarbon systems by physically simulating reservoir depletion and surface recovery.

PVT reports have evolved with the growth of computing and word processing methods, but PVT experiments themselves and the theory behind them have basically remained the same for 50 years, although improvements in accuracy have followed improvements in technology. Extracting properties such as formation volume factor (FVF), viscosity, and solution gas-oil ratios is relatively easy, even in large reports, but many of these properties are inter-related and must all follow the laws of thermodynamics.

Material balance techniques

PVT tests are physical simulations of reservoir depletion, usually by flashing initial reservoir fluid to ambient conditions, yielding residual oil and a volume of evolved gas. Different fluid types are simulated by different depletion "paths." However all types yield atmospheric liquid and varying flash gas volumes at the test's conclusion.

When a sample is charged to a PVT vessel, the system becomes a mini-reservoir of sorts and can be treated as such. Just as engineers perform material balance calculations on reservoirs to determine reserves and recoveries, similar material balances can be calculated on PVT experiments.

With the obvious advantage of already knowing what is in place initially, the percent error can be calculated. The foundation for the basic material balance is "what goes in must come out." A detailed comparison of the mass of the initial fluid should equal, within reason, the total mass of the combined flash products. An error of +1% is acceptable for the mass and molar balances of a PVT study. Knowing actual experimental masses is unnecessary, as PVT reports provide all data necessary to calculate these balances, starting with a few logical assumptions.

Fundamental assumption

Again, the comparison is the mass of the single-phase fluid at the start of the test with the resultant combined mass of the produced and/or residual products at the conclusion of the test. To start the calculations, the fundamental is a residual oil volume of 1 bbl after the completion of the experiment. This assumption makes the other volumetric and mass calculations, using the reported data, relatively simple.

By definition, the formation volume factor is the volume of oil at reservoir pressure and temperature relative to a residual barrel at ambient conditions. Therefore, the assumption of 1 bbl of residual oil defines the volume of oil initially as the unit formation volume factor at the pressure of reference. The difference in the volume and mass of these two values is the gas evolved during the depletion of the fluid from the respective saturation pressure to atmospheric pressure.

With the assumption of 1 bbl of residual oil, the volume/mass of liberated gas becomes the solution gas-oil ratio reported at a respective pressure. In the case of 6,326 psi, assuming 1 bbl of residual oil defines the volume of oil at 6,326 psi as 1.329 bbl and the liberated gas volume/mass as 795 scf. At any pressure reported, the mass of 1 bbl of oil combined with the mass of the solution gas shown are equivalent to the mass of the initial volume (unit FVF), calculated using the oil density at that respective pressure. Intuitively, the initial oil of 1.329 bbl is equivalent to the combined solution gas (795 scf) and residual oil (1 bbl) from a mass and molar standpoint. Most PVT reports show "cumulative" solution gas gravity, which is a weighted average of the entire evolved gas volume. If this value is not reported, then it needs to be calculated using the summation of the incremental gas gravities and liberated gas volumes as weights.

Molar balance

A molar balance should also be calculated. The mass of a fluid divided by its molecular weight yields the moles of fluid. A molar balance combines the mass balance with a QC of the associated compositional analyses performed on the samples, since the molecular weights are normally derived from compositional analyses, such as chromatographic analysis.

It is possible for the mass balance to match and the molar balance to have a significant discrepancy (5% or more). This is usually caused when unrepresentative molecular weights are used, which are normally calculated with chromatographic data. If the reservoir fluid mole weight is calculated correctly and the associated residual liquid and cumulative evolved gas mole weights are accurate, then the molar balance will be in agreement with the mass balance.

Compositional balance

A more detailed compositional balance can also be calculated. The reservoir fluid composition is normally calculated by a mathematical recombination of the flash products from a single stage flash test of the initial reservoir fluid. Residual gas and liquid compositions are mathematically mixed together based on the molar gas-oil ratio observed from the flash test.

Individual differential liberation gases can be combined with the residual liquid to re-calculate the reservoir fluid. This reservoir fluid can be directly compared to the "reported" composition from the single stage flash mentioned above. This yields two distinct methods and measurements of the reservoir fluid composition. If these two compositions show significant variations, problems exist in either the compositions themselves, the experimental volumes measured, or the balances discussed earlier.