Application of geostatistical reservoir description for maximizing waterflood infill drilling recovery from La Cira Field, Colombia

Abstract

One of the prospective ways to increase the oil production is to maximize the oil recovery from mature oil fields. In this study we apply an integrated approach that combines geostatistical reservoir description and reservoir simulation to evaluate infill drilling scenarios to maximize the secondary oil recovery from a Colombian oil field. A geostatistical simulation technique called Sequential Gaussian Simulation (SGS) is used to generate fine-scale realizations of permeability for a waterflood pattern of La Cira field in Colombia, South America. This fine scale permeability field is upscaled to yield a coarser grid and input into a 3D simulation model. A selected fine-scale permeability realization of 100 layers was upscaled to a coarse 6-layer model using a power averaging approach. Two upscaled models were constructed: one using a power average with combined harmonic and arithmetic average with power exponent of 0.5, and the other using an arithmetic average with a power exponent of I.O. Three-dimensional numerical simulations were performed for history matching of 53 years of production performance of the waterflood pattern LC-1210. Three simulation cases were matched: 1.Using smooth interpolation for generating permeability distributions. 2. Using the geostatistical upscaled permeability model with a power average of 0.5. 3.Using the geostatistical upscale permeability field with arithmetic average. permeability the best history match was obtained with the arithmetic-average model. An evaluation through numerical simulation of multiple permeability realizations showed very low uncertainty in the geostatistical reservoir description. With a selected stochastic permeability realization, a simulation model was set up to assess the impact of the infill drilling on oil recovery in the study unit. Seven simulation cases were performed, the first two runs assuming that no infill wells were drilled and the remaining consisting of different infill drilling schemes. The 'base" case (under current injection conditions) represented an additional oil recovery of 4.5% in 20 years (1993-2013). The best infill scenario consisted of drilling one production well in the west side of the waterflood pattern LC-1 2 1 0. This case predicts an incremental oil recovery of 10. 5% and ultimate oil recovery of 35%.