油气藏评价与开发 ›› 2018, Vol. 8 ›› Issue (3): 12-17.

• 油气藏评价 • 上一篇    下一篇

基于遗传算法建立的凝析气藏露点压力改进模型

孙雷1,周芳芳1,夏静2,潘毅1,罗强3,冷捷1   

  1. 1. 西南石油大学油气藏地质及开发工程国家重点实验室,四川 成都 610500
    2. 中国石油勘探开发研究院,北京 100083
    3. 中国石油塔里木油田油气工程研究院,新疆 库尔勒 841000
  • 收稿日期:2017-06-01 发布日期:2018-12-07 出版日期:2018-06-26
  • 作者简介:孙雷(1954—),男,教授,油气田开发工程。
  • 基金资助:
    中国石油天然气股份有限公司十三五重大科技项目“复杂凝析气藏提高开发效果关键技术”(2016B-1504)

An improved dew point pressure model for condensate gas based on genetic algorithm

Sun Lei1,Zhou Fangfang1,Xia Jing2,Pan Yi1,Luo Qiang3,Leng Jie1   

  1. 1. State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Southwest Petroleum University, Chengdu, Sichuan 610500, China
    2. Research Institute of Petroleum Exploration and Development, PetroChina, Beijing 100083, China
    3. PetroChina Tarim Oilfield Oil and Gas Engineering Research Institute, Korla, Xinjiang 841000, China
  • Received:2017-06-01 Online:2018-12-07 Published:2018-06-26

摘要:

利用前人实验测试结果,总结国内不同地区凝析气藏的露点压力数据,运用最优化算法遗传算法进行计算得到预测凝析气藏露点压力的模型,该模型考虑的影响因素比其它模型多,其中包括温度、流体组成、重组分分子量及相对密度、气油比,且其露点压力计算值与前人实验测试值接近,相对误差绝对值的平均值为3.05 %。为进一步验证该模型的预测能力,另外将TLM油田10口井样品的实验测试值与该露点压力改进模型、N-K、Elsharkawy、孙志道和王海应模型计算结果进行对比。其结果表明,本模型预测精度相对较高,其相对误差绝对值的平均值为1.79 %,其它模型的平均相对误差分别为49.53 %、27.05 %、3.30 %、7.05 %。

关键词: 露点压力, 凝析气藏, 遗传算法, 优化算法, 预测模型

Abstract:

By the experimental data of the forefathers and the dew point pressure of the gas condensate reservoirs derived from different parts of the domestic, we make an empirical correlation expression for the calculation of the dew point pressure based on the genetic algorithm(GA). This expression considers more factors than that of other existing models including the temperature, fluid compositions, molecular mass and density ratio of heavy constituent, gas-oil ratio. The calculation results are more closed to the experimental data than the previous test values. The average of the relative error's absolute value is 3.05 %. In order to further verify the new empirical expression's prediction ability, the experimental data of other 10 groups of dew point pressure are compared with the results calculated by four models that is the N-K, Elsharkawy, Sun Zhidao's model and Wang Haiying's model. The results are relatively high in accuracy with the absolute value of 1.79 %. The absolute values of other four models are 49.53 %, 27.05 %, 3.30 % and 7.05 %.

Key words: dew point pressure, gas condensate reservoir, genetic algorithm(GA), optimization algorithm, prediction model

中图分类号: 

  • TE31