This paper is the result of experience gained in developing a sour retrograde condensate gas/volatile oil reservoir in the Brazeau River area of Alberta, Canada. It was not possible to obtain a satisfactory PVT description and therefore reliable reservoir engineering calculations could not be made. Research in the literature, and offsetting government PVT data, identified that these problems were not unique to this reservoir. Examples are given of the variations in PVT properties from two other reservoirs.
A significant effort was expended in identifying why these problems occurred, what impact they would have and what steps could be taken to correct them. Relevant issues include how high H2S fluids are generated (Geochemistry), obtaining stabilized samples (compositional reservoir simulation), problems with obtaining PVT data (physical laboratory procedures) as well as field implementation (testing program, screening samples, and checking PVT results). The implications to reservoir development are also discussed.
Home Oil Company Limited (subsequently a subsidiary of Anderson Exploration Ltd.) made a number of discoveries in the Brazeau River area of Alberta, which is located roughly 100 miles south west of Edmonton Alberta. In particular, a pool was discovered which early testing indicated contained a sour (approximately 29 mole percent H2S) retrograde condensate gas. Liquid dropout volumes from a constant volume depletion (CVD) experiment on this fluid were very high at over 40 percent of pore space.
The retrograde condensate system with very high liquid dropouts suggested that a gas cycling scheme would likely be an economic development strategy. A major logistical problem in the Brazeau River area was obtaining sufficient gas processing capacity. Recently, a number of projects have been implemented in Alberta in which sour gas has been reinjected back into the reservoir. Such a scheme could solve the plant capacity problem and, since sulphur prices are depressed, recovering H2S was not a priority.
Most cycling schemes are now modelled with a compositional simulator. This technique typically relies on cubic Equations of State (EOS) to describe the pressure-volume-temperature (PVT) behaviour of the in-situ wet gas and the interaction with injected gases. EOS calculations are not accurate enough to be used directly on a predictive basis. Input parameters must be "tuned" to match experimental data. Results are therefore heavily dependent on the amount and accuracy of PVT laboratory data.
The first sample obtained was derived from a bottom hole wireline tool (MDT) of 450 cc. The sample was sent out for PVT analysis, which included: a constant composition expansion (CCE), with liquid volumes recorded, a constant volume depletion (CVD) test and gas phase compositions from the CVD test. Key results are outlined below:
The dewpoint was measured at 27,642 kPa, at a temperature of 85 C.
The maximum liquid dropout from the Constant Composition Expansion was 36.65 percent at a pressure of 22,802 kPa.
The maximum liquid dropout during the Constant Volume Depletion was 41.64 percent at a pressure of 17,326 kPa.
These results were used as the basis for EOS tuning, which is described in the following.
The initial plan was to get a characterization with a larger number of components and then simplify the system until the characterization degraded below acceptable levels. Initially a 10 component model was used. P. 691