Abstract: The uncertainties of a gas reservoir located in the north part of Shaanxi Province were studied with stochastic modeling. The responding gas reserves of the reservoir were calculated with three groups of data obtained from two hundreds and fifty four wells, one hundred and forty six wells, thirty wells by using fifty, one hundred and two hundreds of random seeds, respectively. Probabilistic distributions of these reserves were analyzed. Three probabilistic reserves (P90, P50, P10) for the three groups of data were obtained respectively by using one hundred of random seeds. The probabilities of three probabilistic reserves larger than the real reserves are 90 percent, 50 percent and 10 percent respectively. The calculated results illustrate that as the number of wells increase, the mean values of gas reserves continuously increase, and their mean variances decrease, which results in the reduction of the uncertainties of reservoir. The top strata surface and net thickness were obtained by Kriging method. The spatial distributions of porosity, permeability and gas saturation were obtained by simulation of Gaussian field.

Key words: natural gas reservoir, stochastic modeling, physical parameter, probabilistic reserves

摘要： 应用储层随机建模方法,研究了陕北地区气田的一个储层,分析了它的不确定性.分别利用了由254口、146口和30口井组成的3组数据,利用50、100、200个随机种子,计算了相应的天然气储量,分析了储量的概率分布.利用100个随机种子,分别求得了这3组数据概率储量P90,P50和P10,大于实际储量的概率分别是90%、50%和10%.计算结果表明,随着井数的增加,天然气储量分布的均值会不断地增加,而它们的均方差则减小,从而减小不确定性.利用克里金估计方法求取了地层的层面深度和有效厚度,利用高斯场模拟方法求取了孔隙度、渗透率、含气饱和度的空间分布.

关键词: 天然气储层, 随机建模, 物性参数, 概率储量