Controlling potential risks in Petroleum Industry is one of the most difficult tasks for geologists, reservoir engineers and others. Evaluating oil and gas reserves becomes riskier and less accurate for fields which have complex geologic structure and high heterogeneity. In these cases, correlation relationship between dependent reservoir parameters will be represented better and more exactly with non-linear regression models.
The "box" method is introduced in this research as an improved technique in handling these more accurate non-linear regression models compared to conventional method – Spearman's rank order correlation method – which is available in many commercial Monte Carlo simulators. The conventional method is just correct in uniform and homogeneous reservoirs whereas the "box" method is better for complex and heterogeneous reservoirs.
To get the most accurate estimation of oil and gas reserves, a computer program using the above improved technique along with Monte Carlo simulation method was written. Some randomly generated data that were similar to data from a specific heterogeneous reservoir were also considered. And exhaustive simulation runs were performed to compare the results from written computer program to those from a commercial Monte Carlo simulation package. It was seen that their outputs were different due to different methods used.
Regarding such exactly estimated amount of oil and gas from written computer program, project designing team planned field development more accurately and efficiently. This very important factor led to a successful production project.