A solvent enhanced steam drive pilot in the Peace River area in Canada was executed over a period of two years on an inverted 5-spot pattern to evaluate bitumen uplift and solvent recovery. Data quality and uncertainty management were used to assure conclusive results for the pilot; in particular, this paper focuses on how sampling played a key role in providing conclusive results. Especial focus is provided to the design, performance, execution, and learnings from the use of the automatic proportional samplers which was particularly challenging, considering their novel use on heavy oil production containing abrasive material such as sand and H2S. Moreover, to determine solvent production, several newly developed algorithms were tested to split the solvent and bitumen compositions, since they overlap over a wide range of components. Results from the pilot show that significant errors and misinterpretations can occur while evaluating solvent recovery whenever the assumptions under the solvent-bitumen split algorithms are not checked against frequent sampling data. The pilot also provided best practices for the utilization of proportional auto samplers in heavy oil production in the presence of abrasive material inherent to in-situ production, and the sampling of gas streams with heavy condensate for the accounting of solvent production through casing vent gas.

You can access this article if you purchase or spend a download.