Probabilistic Distribution Model for Release Rates of H_2S in Sour Gas Wells of Sichuan and Chongqing Areas

陈胜,吴庆善,江田汉,席学军,郭再富,王建光
DOI: https://doi.org/10.3321/j.issn:0253-2697.2009.05.026
IF: 1.468
2009-01-01
Acta Petrologica Sinica
Abstract:The probabilistic distribution model for the release rates of H2S in the sour gas wells of Sichuan and Chongqing areas was developed using the complementary cumulative distribution function.The model parameters such as the scaling exponent and the lower bound to the power-law behavior were estimated by the maximum likelihood method and the Kolmogorov-Smirnov statistic method.And the latter was also applied to goodness-of-fit tests with the synthetic data generated by the Monte-Carlo procedure from the hypothesized distribution.The release rates of H2S in the sour gas wells of Sichuan and Chongqing areas follow a non-gaussian and heavy tailed distribution.The cumulative probability distribution of level 1 and level 2 sour gas wells is much higher than that of the gaussian distribution.The distribution probability closely follows the power-law behavior with the scaling exponent of 1.5395 when the H2S release rate is over 0.0067m3/s,and its tail corresponds to the probabilities of level 1 and level 2 sour gas wells with high hazard.Those sour gas wells mainly located in the northeastern and northwestern areas of Sichuan Gas Field,Chongqing Gas Field and Puguang Gas Field.This model can describe reasonably the probabilistic distribution of the H2S release rates in the sour gas wells of the Sichuan and Chongqing areas.
What problem does this paper attempt to address?