Numerical Well testing using Unstructured PEBI Grid

Hui Deng, Zhangxing Chen, Xia Bao and Ruihe Wang


History matching is challenging for low permeability gas reservoir because of significant differences between the static properties in geostatistical model and in-situ properties measured from well testing. Literatures have documented that reduction of in-situ permeability due to overburden pressure can be in two orders of magnitude. Numerical well testing provides a way of tuning the static model with dynamic well test information. However, traditional single well testing model using Cartesian LGR is not ideal for predicting the pressure transient behavior. Furthermore, stand-along well testing conditioned model cannot be fully coupled into the full-field model to honor the flow regime.

This paper presents a methodology of using PEBI grid in simulation model to match the well test data for a low permeability gas reservoir in Canadian Foothills. PEBI-LGR grid is created around the well and vertical hydraulic fracture is implemented to match the post fracture pressure build up. History match parameters include static properties and hydraulic fracture properties such as fracture half length and fracture permeability. Finally, well testing conditioned effective properties will be stochastically populated into simulation model for the full-field history matching. In conclusion, PEBI gridding technique links well testing with reservoir simulation and provides the most efficient workflow in modeling the unconventional gas reservoir with multi-stage hydraulic fractures.