高级检索

大宁—吉县区块煤层气产出动态的深度效应及产能主控因素分析

Depth effects on coalbed methane production performance and analysis of key productivity controlling factors in the Daning—Jixian Block

  • 摘要: 深部煤层气与浅部煤层气生产特征具有显著差异,煤层埋深作为贯穿储层物性、流体赋存与渗流规律的核心因子,其对气水产出的调控作用尤为重要,深入分析深、浅部煤层气生产特征差异的影响因素及产出动态的深度效应对后续煤层气开发阶段的产能预测与生产具有重要意义。为明确大宁—吉县区块深浅部煤层气动态的深度效应,基于煤层埋深、含气性、储层压力、镜质体反射率、构造特征、水化学特征及压裂改造强度等地质与工程因素开展研究。结果表明:大宁—吉县区块深部煤储层较浅部煤储层具有“高含气量、高吸附饱和度、高储层压力和低渗透率”的特征,深、浅部煤层气井的气水产出动态存在显著差异。采用支持向量机(SVM)回归模型和数据增强策略,构建了融合地质与工程多因素的煤层气产能预测模型,实现了对浅部与深部煤层气井稳产气量的高精度预测。模型在浅部样本中测试集R2达0.91,深部样本R2达0.90,表现出良好的拟合能力与泛化性能。基于地质与工程多参数的综合分析及机器学习SHAP归因分析,明确了深、浅部煤层气井生产特征的主控因素及其贡献序列。浅部煤层气产能主要受工程改造强度主导,主控因素贡献度依次为:压裂液总量、矿化度、断裂强度系数、含气量、埋深、储层压力和镜质体反射率,表明在浅部,人工压裂改造是突破储层物性限制、提升产能的核心手段。而深部煤层气产能则转而由储层地质与流体性质主导,主控因素贡献度依次为:矿化度、压裂液总量、含气量、埋深、储层压力、镜质体反射率和断裂强度系数,凸显了成藏流体属性与深部封闭性水文地质条件对维持高产稳产的关键控制作用。上述研究成果揭示了大宁—吉县区块煤层气产出动态的深度效应与产能主控因素,对研究区和邻近地区煤层气生产具有参考意义。

     

    Abstract: Deep coalbed methane (CBM) production behavior differs markedly from that of shallow CBM. Coal seam burial depth, as a core factor influencing reservoir physical properties, fluid occurrence, and seepage patterns, plays a particularly important role in regulating gas and water production. In-depth analysis of the factors influencing the differences in production characteristics between deep and shallow CBM, as well as the depth effect on production dynamics, is of great significance for productivity prediction and production in subsequent CBM development stages. To clarify the depth effect on the dynamics of deep and shallow CBM in the Daning-Jixian Block, research was conducted based on geological and engineering factors such as coal seam burial depth, gas content, reservoir pressure, vitrinite reflectance, structural characteristics, hydrochemical characteristics, and fracturing stimulation intensity. The results show that compared to shallow coal reservoirs, deep coal reservoirs in the Daning—Jixian Block are characterized by “high gas content, high adsorption saturation, high reservoir pressure, and low permeability,” and there are significant differences in the gas and water production dynamics between deep and shallow CBM wells. Using a support vector machine (SVM) regression model and a data augmentation strategy, a CBM productivity prediction model integrating multiple geological and engineering factors was constructed, achieving high-precision prediction of stable gas production rates in both shallow and deep CBM wells. The model achieved an R2 of 0.91 on the test set for shallow samples and 0.90 for deep samples, demonstrating good fitting capability and generalization performance. Based on comprehensive analysis of multiple geological and engineering parameters and machine learning SHAP attribution analysis, the main controlling factors of production characteristics in deep and shallow CBM wells and their contribution sequences were identified. Shallow CBM productivity is mainly dominated by engineering stimulation intensity, with the contribution of main controlling factors in the following order: total fracturing scale, salinity, fracture intensity coefficient, gas content, burial depth, reservoir pressure, and vitrinite reflectance. This indicates that in shallow areas, artificial fracturing stimulation is the core means to break through reservoir physical limitations and enhance productivity. In contrast, deep CBM productivity is dominated by reservoir geology and fluid properties, with the contribution of main controlling factors in the following order: salinity, total fracturing scale, gas content, burial depth, reservoir pressure, vitrinite reflectance and fracture intensity coefficient. This highlights the key controlling role of reservoir fluid properties and deep enclosed hydrogeological conditions in maintaining high and stable production. The above research results reveal the depth effect and main controlling factors of CBM production dynamics in the Daning—Jixian Block, providing reference significance for CBM production in the study area and adjacent regions.

     

/

返回文章
返回