基于增强型生成对抗网络地质建模框架的多模式非平稳储层随机模拟
|
宋随宏(1990-),男,陕西延安人,博士,斯坦福大学研究科学家(Research Scientist),主要从事人工智能、储层地质建模、渗流模拟和人工智能大模型研发与应用方面的工作。地址:Stanford University, 367 Panama St, Stanford, CA 94305, USA。E-mail:suihong@stanford.edu |
Copy editor: 魏玮
收稿日期: 2025-08-24
修回日期: 2026-01-19
网络出版日期: 2026-01-22
Geomodelling of multi-scenario non-stationary reservoirs with enhanced GANSim
Received date: 2025-08-24
Revised date: 2026-01-19
Online published: 2026-01-22
为了将GANSim(基于生成对抗网络的直接条件化地质建模框架)用于多地质模式、非平稳性储层地质建模,并克服该方法易于忽略占据单网格的井点条件数据进而导致井周沉积相模拟结果局部断连的局限,提出“增强型GANSim”建模框架,以多模式非平稳浊积扇储层地质建模为例,验证增强型GANSim的有效性。针对可能存在多种地质模式的储层,提出两种GANSim地质建模流程:①训练一个覆盖所有可能地质模式的综合GANSim模型;②先进行地质模式证伪,然后针对未被证伪的地质模式训练GANSim模型。在此基础上,进行局部判别器的结构设计,以提升井周沉积相的连续性。浊积扇储层建模结果表明,两种GANSim建模流程均能产生与期望地质规律和条件数据都吻合的非平稳性沉积相地质模型实现,同时井周沉积相不连续的问题也得以解决。与多点地质统计学方法(SNESIM)相比,GANSim表现出卓越的储层规律复现能力和建模效率,尽管GANSim的训练耗时较长,但是一旦训练完成,即可用于任何具有相似地质结构、任意规模储层的地质建模,建模速度约为SNESIM的1 000倍。
宋随宏 , MUKERJI Tapan , SCHEIDT Celine , ALQASSAB Hisham M. , FENG Man . 基于增强型生成对抗网络地质建模框架的多模式非平稳储层随机模拟[J]. 石油勘探与开发, 2026 , 53(1) : 177 -189 . DOI: 10.11698/PED.202500125SD
GANSim is a generative adversarial networks (GANs)-based direct conditional geomodelling framework. To extend GANSim to multi-scenario, non-stationary reservoir geomodelling, and to address its tendency to overlook single-grid well conditioning data that can cause local facies disconnections around wells, an enhanced GANSim framework is proposed. The effectiveness of the enhanced GANSim is validated using a 3D multi-scenario, non-stationary turbidite fan reservoir as an example. For reservoirs that may involve multiple geological scenarios, two GANSim geomodelling workflows are proposed: (1) training a comprehensive GANSim model that covers all possible geological scenarios; and (2) first performing geological scenario falsification and then training GANSim models only for the unfalsified scenarios. On this basis, a local discriminator architecture is designed to improve facies continuity around wells. The modelling results show that both workflows can generate non-stationary facies models that conform to expected geological patterns and honor conditioning data, and the facies discontinuity issue around wells is effectively resolved. Compared with multipoint geostatistical methods(SNESIM), GANSim exhibits superior capability in reproducing reservoir geological patterns and modelling efficiency. Although GANSim requires a longer training time, once training is completed, it can be applied to geomodelling reservoirs of arbitrary scale with similar geological structures, achieving modelling speeds approximately 1 000 times faster than SNESIM.
| [1] |
|
| [2] |
|
| [3] |
|
| [4] |
|
| [5] |
|
| [6] |
|
| [7] |
|
| [8] |
|
| [9] |
|
| [10] |
|
| [11] |
|
| [12] |
|
| [13] |
|
| [14] |
|
| [15] |
|
| [16] |
|
| [17] |
|
| [18] |
|
| [19] |
|
| [20] |
|
| [21] |
|
| [22] |
|
| [23] |
|
| [24] |
|
| [25] |
|
| [26] |
|
| [27] |
|
| [28] |
|
| [29] |
|
| [30] |
|
| [31] |
|
| [32] |
|
| [33] |
|
/
| 〈 |
|
〉 |