I-Corps: Process Monitoring and Diagnosis Software for Decarbonizing Industrial Distillation

这个I-Corps项目的更广泛的影响/商业潜力是提高蒸馏设施的能源效率和碳足迹，以促进化学和炼油行业的脱碳，这两个能源使用量最大的加工行业消耗了美国制造业几乎一半的一次能源，排放了一半的温室气体。通过降低蒸馏系统在其最佳操作区域的操作风险，该软件工具有望减少能源需求和温室气体排放。受益于该软件工具的一些众所周知的工业过程包括原油分馏、天然气分离、二氧化碳捕集的胺再生、空气分离、液体生产等。此外，通过降低运营成本、提高生产吞吐量和降低安全风险，该软件工具增强了美国化工厂和炼油厂的经济竞争力，并改善了健康、安全、以及附近社区的环境健康。该软件的商业化还将有助于培训化工和炼油行业的下一代劳动力，包括工厂操作员、工艺工程师和工厂经理，以使用数字化工具进行工厂数据分析。该I-Corps项目的基础是开发用于工业蒸馏系统实时过程监控和故障诊断的数据分析技术。通过从历史运行数据中进行策略性学习，该软件工具能够在线监测传感器连续产生的多个非参数、异构过程数据流，并对各种可能的故障情况进行准确的故障分类。这是通过在线过程监测和故障诊断的整体集成在一个新的层次结构。与试图同时完成这两项任务的现有通用过程监控工具相比，这种分层架构在故障检测速度、故障分类精度和鲁棒性方面提供了上级故障检测和诊断性能，以处理多个工厂操作场景。该智能蒸馏监控软件还可以与现有的工厂控制和监控系统无缝集成，这有利于改造目的。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact/commercial potential of this I-Corps project is to improve the energy efficiency and carbon footprint of distillation facilities to facilitate decarbonization in chemical and refining industries, the top two energy use processing industries consuming almost half of the U.S. manufacturing sector’s primary energy and emitting half of its greenhouse gases. By de-risking distillation system operation at its optimal operating region, this software tool is expected to reduce energy requirement and greenhouse gas emissions. Some of the well-known industrial processes that benefit from this software tool include crude oil fractionation, natural gas separation, amine regeneration from carbon dioxide capture, air separation, liquors production, etc. Furthermore, by reducing operating costs, improving production throughput, and reducing safety risks, this software tool enhances the economic competitiveness of U.S. chemical plants and refineries, and improves health, safety, and environmental wellness of nearby communities. The commercialization of this software will also help train the next-generation workforce in chemical and refining industries, including plant operators, process engineers, and plant managers, to use digital tools for plant data analytics.This I-Corps project is based on the development of a data analytics technology for real-time process monitoring and fault diagnosis of industrial distillation systems. By strategically learning from historical operating data, this software tool enables online monitoring of multiple nonparametric, heterogeneous process data streams continuously produced from sensors, as well as accurate fault classification from various possible fault scenarios. This is accomplished by holistic integration of online process monitoring and fault diagnosis in a new hierarchical architecture. Compared to existing general purpose process monitoring tools that attempt to accomplish both tasks altogether, this hierarchical architecture offers superior fault detection and diagnosis performance in terms of fault detection speed, fault classification accuracy, and robustness to handle multiple plant operation scenarios. This smart distillation monitoring software can also be seamlessly integrated with existing plant control and supervisory systems already in place, which is favorable for retrofitting purposes.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.