Abstract
A successful cyclic steam project has been implemented by Chevron USA Production Company in the Antelope Shale Formation of the Cymric 1Y Field, California. Recovery is expected to exceed 50% OOIP from this low permeability, heavy oil diatomite reservoir. The expected high recovery will be achieved through aggressive infill drilling, optimization of current steam injection operations, and favorable interaction of steam with induced fractures.
The patented cyclic steam process induces fractures by injecting steam at pressures exceeding the reservoir fracture gradient. Small, controlled injection/production cycles are applied to maximize steam efficiency. Optimizing cyclic operations in over 400 wells, which are, on average, cycled more than once per month, is a formidable task.
To follow is a description of two very different approaches for predicting and optimizing the performance of the cyclic steam process: (1) use of a very detailed, multiple-pattern simulation model, and (2) use of Design of Experiment techniques (DOE) coupled with much simpler, single-well simulation models.
While the detailed simulation model was able to capture complex operational history and provided a good history match, its use was limited due to long run times. Consequently, sensitivities to various key uncertainties are time consuming to obtain, limiting the flexibility of this approach. The DOE approach allows us to study more reservoirs and operational parameters in a relatively short time. Results from both approaches were combined to analyze and forecast field performance, and for aiding reservoir development decisions. Our combination method represents an optimal approach for cyclic thermal recovery process evaluation in heavy oil diatomite reservoirs.
