Methods for economic optimization of reservoirs
MetadataShow full item record
Operators can improve a reservoir’s value by optimizing it in a more holistic manner, or over its entire life cycle. This thesis developed approaches to life cycle optimization, with emphasis on accessible technical and economic modeling techniques for production. The challenges of life cycle optimization are properly scheduling the times at which the operator should switch from one recovery phase to the next, along with determining other field design parameters such as well spacing and injection pressures for waterflooding and enhanced oil recovery processes. To deliver the most value, the operator needs to produce from a reservoir the greatest quantity of oil, at a relatively low cost, reasonably soon, and ideally at a time when the oil price is high. This is quite a tall order, as these goals are often in conflict. This thesis extended existing research regarding lifecycle optimization, first modeling production from a reservoir using an exponential decline model and assuming the oil price’s behavior can be approximated with mean-reverting processes. Implications of operating and capital costs potentially being correlated with the oil price were also examined. Finally, a mean-reverting price model that forecasts the mean oil price as increasing and described by a logistic model was proposed to accommodate both recent price forecasts and economic reality. As exponential decline models are more appropriate for characterizing existing production history rather than making a priori predictions, a geologic-parameter-based model was developed using a tank model for primary recovery and a model based on Koval theory and parameterizing a reservoir in terms of flow capacity and storage capacity for waterflooding and CO2 flooding. This model was adapted from existing theory to account for situations where a waterflood has incompletely swept a reservoir at the start of CO2 flooding. Analytical expressions were also derived for estimating injection rates into a formation parameterized by flow capacity and storage capacity. The geologic-parameter-based model was combined with economic assumptions and optimized using a genetic algorithm. This optimization suggested an operator should switch from primary recovery to a CO2 flood with a large WAG ratio relatively early in the reservoir’s life.