A Geologically Based Markov Chain Model for Simulating Tritium Transport With Uncertain Conditions in a Nuclear-Stimulated Natural Gas Reservoir

作者：Ye M^*; Cooper C A; Chapman J B; Gillespie D; Zhang Y

来源：SPE Reservoir Evaluation and Engineering, 2009, 12(6): 974-984.

DOI：10.2118/114920-pa

Nuclear-stimulation technology, which used subsurface nuclear detonation to increase permeability of tight natural gas reservoirs, was evaluated in the late 1960s and early 1970s. The Rulison site, located in the Piceance basin, Colorado, is one of three sites in the US where the technology was tested. An increase in exploration and production for natural gas in the basin has led to a need to quantify the extent of radionuclide (mainly tritium) migration after the detonation and potential migration under likely production scenarios. To meet this need, a numerical model was developed to simulate gas flow and tritium transport toward a hypothetical production well. A crucial problem in the model development is that limited on-site data are too sparse to quantify uncertainty of subsurface properties. This problem is partly resolved by using indirect data and information, such as parameter measurements from a nearby site and geological information regarding lithofacies geometry. In particular, a geologically based Markov chain model was developed to simulate spatial distribution of the sandstone lithofacies.
This paper presents an application of the numerical model for simulating tritium transport from the nuclear chimney toward the production well at a likely location producing at a rate typical for the basin. The results show that under the circumstances considered in this paper, tritium will not reach the production well with a confidence level of 95%. The results also show that the lithofacies structure is more critical in controlling tritium transport than parameters of the sandstone and hydraulically fractured sandstone. The parameters become important only when the connectivity of sandstone lenses exists to support tritium transport from the chimney to the production well. The developed modeling framework can be updated as additional subsurface data are collected. The framework can be used to support establishment of drilling restrictions that protect public health and the environment for different production well scenarios.

• 出版日期2009-12