Predicting Connected Hydrocarbon Volume Uncertainty Ranges: From Static Fault Model to Dynamic Simulator

Abstract

Abstract In this paper we present two methods for calculating uncertainty ranges for connected hydrocarbon volumes as applied to the Sunrise gas-condensate field, Timor Sea.A "quick look" approach which uses a simplified, exploration scale fault data set.A more detailed fault interpretation which is integrated with 3-D geological and dynamic reservoir modelling techniques. Despite the differences in sophistication, the two methodologies generate similar results in terms of absolute connected hydrocarbon volume ranges. In both studies the fractal relationships between the fault throws, lengths and frequency of occurrence were used to determine the minimum throw of faults fully interpreted on the seismic. This structural interpretation was then augmented by the probabilistic infill of sub-seismic faults with throws greater than the critical juxtaposition-sealing throw of 20 m. The resulting maps were then used to estimate a range of potentially connected gas volumes. The large number of across fault sand-sand juxtapositions within the areally extensive field suggests that as a most likely case the reservoir is 100% connected due to effective communication across faults or around fault tips. Sensitivities were analysed around this most likely case. As a worst case scenario all faults were assumed to be totally sealing due to the potential effects of shale gouge or cataclasis. The quick look study, which took no account of any sedimentological connectivity issues, predicted a worst case connectivity of 77% whilst the detailed 3-D dynamic simulator study predicted a connectivity of 72%. Although the areal distribution of connected hydrocarbons was different for the two approaches, the similarity of the final connected volume figures suggests that the quick look methodology provides a useful technique for the rapid estimation of connected hydrocarbon volume ranges in fields with limited data content and perceived simple sandbody architectures. Introduction The Sunrise gas-condensate field [1] [2] is located 450 km northwest of Darwin on the edge of the Australian continental shelf and 50 km from the adjacent Timor Trench (Fig. 1). In-place estimated volumes range from (P90 to P10) 10.4 to 19.2 Tcf gas and 474 to 914 million bbls condensate. The field has an areal closure of 920 square kilometres (km) with a maximum vertical closure of 180 metres. The main reservoir interval is the marginal marine, Bathonian Upper Plover Formation. The field was discovered in 1974, however, sovereignty disputes and a perceived lack of gas markets halted further appraisal until 1995. The recent appraisal campaign has improved expectations and confidence in reservoir performance and allowed the pursuit of gas markets. The data set now includes six wells that have intersected the gas-bearing reservoir, a 1×2 to 3 km 2-D seismic grid, 3 wells fully cored over the reservoir interval (300 metres of core), eleven production tests and abundant high resolution wireline log data. Extensive geological, geophysical, petrophysical and reservoir engineering analyses of these data have determined the range of subsurface uncertainties [1] [2] [3]. The resource is currently at a stage where it can be commercially exploited. Domestic gas or LNG (shored-based LNG, near Darwin, or Floating LNG technology) markets are potential customers. Due to the large capital expenditure required for any of the potential field developments, the identification of key subsurface uncertainties (show-stoppers and value enhancers) has been a critical goal of these analyses.