81 巻 (2016) 6 号 p. 458-468
Under the circumstances of low oil price, not only oil companies but also universities have continued a variety of efforts to save labor and cost. The investigation on the optimization using computer programs is one of such efforts. This paper introduces some of the computer programs applying the optimization algorithm that have been developed in the laboratory of the author.
First example is the program to optimize facies distributions. This program has three functions of estimating facies distributions using multi-point statistics, optimizing the facies distributions thus estimated by the genetic algorithm and improving the computational time in selecting superior specimens in the genetic algorithm applying the multi-dimensional scaling and k-means clustering.
Second example is the program to estimate the true fluid composition based on the composition of fluid samples acquired in inappropriate conditions such as disturbed 2-phase flow condition and gas leakage condition, applying the iterative Latin hyper cube sampling method as an optimization method.
Third one is the program to estimate the oil relative permeability in the 3-phase condition. Since it is time consuming and hence costly to measure the relative permeability in the steady-state 3-phase condition, this program enables the estimation of the relative permeability as functions of oil, gas and water saturations, by reproducing the results of unsteady-state 3-phase core flooding experiments through optimization with Gauss-Newton method.
The last program optimizes the CO2 injection rate and the location of wells so that the economics of a CCS-EOR project can be maximized, taking the oil increment by CO2 injection and the CO2 tax/credit by storing CO2 in a reservoir into consideration simultaneously.
Although the above programs may be still in the primitive stage, the simulation runs executed for validating the performances of these programs showed the promise of the optimization programs for conserving labor and time.