An Improved Regional Segmentation for Probability Perturbation Method
H. Khani, H. Hamdi, L. Nghiem, Z. Chen and M. Costa Sousa
Event name: 79th EAGE Conference and Exhibition 2017
Session: Uncertainty and History Matching - From Exploration to Production
Publication date: 12 June 2017
Info: Extended abstract, PDF ( 496.91Kb )
Price: € 20
Conventional history matching methods do not consider seismic and geologic continuity data. Caers (2002) introduced a novel history matching method named Probability Perturbation Method (PPM) by extending the multiple-point geostatistics framework to production data; The method’s key point is to perturb the underlying probabilities used to generate properties and not the properties directly. In single region PPM, one perturbation parameter is used for the entire reservoir. However, in multi-parameter perturbation, different amounts of perturbation are applied to different parts of reservoir In our method, a weight factor is assigned to each point (well location) in a way that the volume of each generated region is proportional to the rate of well located inside the region. In other words, volume divided by rate is equal for all regions. Therefore, the question is how to find the weight factors. A set of equations is formed and the solution is found by an iterative method. In each time step, the weight factors and consequently regions could be updated based on well rates. The Voronoi diagram has already been used for defining regions, however the novelty of this work is that defined Voronoi regions are proportional to rate and update dynamically without flow simulation.