SBIR Phase I: Parallel Processing of Time-Lapse Seismic Data via the Internet

Fourth Wave Imaging Corporation的小型企业创新研究（SBIR）第一阶段项目涉及在并行计算机上处理和分析时移地震数据，使用互联网控制处理流程并可视化结果。 近年来，时移地震项目活动呈指数增长。 时移地震分析通过对储层中随时间的流体运动进行成像来促进油气储层的管理。 这些结果用于指导油藏管理决策-例如在哪里放置新井或在哪里注入水，天然气或蒸汽以刺激烃类运动-并帮助最大限度地延长新油田和现有油田的寿命，同时最大限度地降低采收成本。 处理时移地震数据所需的计算机算法很复杂，需要先进的计算硬件-通常是多处理器Unix工作站或个人计算机集群-它们可以并行执行指令。 并行硬件几乎没有标准化。 客户通常根本没有并行计算机，或者计算机的体系结构与软件供应商的体系结构根本不同--这阻碍了该软件的营销和部署。 拟议的创新将允许客户在集中的PC集群上处理他们的数据，使用互联网来控制处理并可视化结果。 它还将改善延时地震工作流程各组成部分之间的联系，从而加深对这项技术的了解，使其在商业上得到更广泛的接受，拟议研究的潜在应用包括石油工业绘制绕过的石油图，监测昂贵的注入流体，以及成像流动分区和断层和裂缝的水力特性。 非石油应用包括监测地下水储量，污染物羽流的地下监测和环境清理项目。 为该项目开发的基于互联网的并行软件系统可以应用于其他计算密集型领域，如医学和卫星成像，天气预报和金融。

This Small Business Innovation Research (SBIR) Phase I project from Fourth Wave Imaging Corporation concerns the processing and analysis of time-lapse seismic data on parallel computers, using the internet to control the processing flow and visualize the results. In recent years, there has been exponential growth in time-lapse seismic project activity. Time-lapse seismic analysis facilitates the management of oil and gas reservoirs by imaging fluid movement in the reservoir over time. The results are used to guide reservoir management decisions--such as where to place a new well or where to inject water, gas, or steam to stimulate hydrocarbon movement--and help maximize the life of both new and existing fields while minimizing recovery costs. The computer algorithms needed to process time-lapse seismic data are complex and require advanced computational hardware--typically multiprocessor Unix workstations or clusters of personal computers--that can execute instructions in parallel. There is little standardization in parallel hardware. Customers typically have no parallel machines at all or machines whose architecture is fundamentally different from that of the software vendor--hindering the marketing and deployment of this software. The proposed innovation will allow customers to process their data on a centralized PC cluster, using the internet to control the processing and visualize the results. It will also improve the links between the components of the time-lapse seismic workflow, leading to greater understanding and more widespread commercial acceptance of the technology.Potential applications of the proposed research include petroleum industry mapping of bypassed oil, monitoring of costly injected fluids, and imaging flow compartmentalization and the hydraulic properties of faults and fractures. Non-petroleum applicatons include monitoring groundwater reserves, subsurface monitoring of contaminant plumes and environmental clean-up projects. The internet-based parallel software system developed for this project could be applied to other compute-intensive fields suchas medical and satellite imaging, weather forecasting, and finance.