This paper presents a modelling workflow to efficiently history match naturally fractured reservoirs using Pressure Transient Analysis (PTA) with application to a complex naturally fractured sandstone reservoir.

Incorporated in the modelling workflow is the use of deconvolution to streamline the history matching process. To ensure this workflow is flexible to mature fields with interference between wells, a work around to achieve a similar output to multi-well deconvolution is presented.

The modeling workflow was successfully applied to a case study resulting in reduction of computer processing time by around 60%. The product of this workflow was not just a single dual-porosity model which best matched the historical data-set, but a suite of models which sample the range of possibilities. From the multiple history matched models, the low (P90), mid (P50) and high case (P10) models were selected statistically.

Fracture porosity is shown to be the input parameter with largest impact on the pertinent simulation outputs in the case study such as time to water breakthrough, pressure match and recoverables. As such, time should be spent gathering and processing data to narrow the uncertainty range. However, care should be taken not to reduce the range prematurely as fracture porosity estimation is a source of wide uncertainty.

Additionally, the case study highlights that horizontal anisotropy can have a significant impact on the field GIIP during the history matching process and consequently impact the reserves in a fractured field. In the case study, the introduction of horizontal anisotropy not only improved the history match but also increased the field GIIP post-history match by around 30%. The value in studying anisotropy within a fractured field and its inclusion into modeling is clearly demonstrated.

This paper brings together various innovative data interpretation approaches and modeling techniques and presents them in an applied workflow. The workflow shows the practicing engineer how they can best interpret and utilize the available data-sets through the application of PTA. The result is not only a better history match, but also a more efficient process which provides a statistical representation of the likely possibilities of the naturally fractured system.

You can access this article if you purchase or spend a download.