Sorbent Screening and Optimization of an Oxygen Purification Process****

基准国际公司是一家总部位于埃德蒙顿的公司，以氧气解决方案公司的品牌名称生产氧气浓缩器。这些装置利用一种称为吸附的物理分离过程从大气中浓缩氧气。这些装置通常安装在没有瓶装氧气或价格昂贵的偏远地区，如水产养殖、废水处理和采矿设施。为了在这个竞争激烈的行业中获得市场份额，该公司正在寻求提高其氧气浓缩器的性能，最终目标是减少设备尺寸和能耗。这转化为氧气成本和相关的二氧化碳足迹的降低（通过减少能源消耗）。为了实现这些目标，基准公司寻求与阿尔伯塔大学（UofA）的研究人员合作，在吸附过程设计方面拥有专业知识。UofA集团将应用复杂的数学工具来模拟一个小型的氧气浓缩器，该浓缩器将由Benchmark专门建造，以支持该项目。使用这些数学模型，我们将测试三种声称可以改善工艺性能的商业材料。该项目的独特之处在于开发和部署基于机器学习的优化工具，这些工具能够探索作业空间，以确定降低总成本的作业条件。这将节省数千小时的工程人工和测试不同材料的相关成本。建模工作的结果将首先在小规模系统上进行验证，然后在商业仪器上进行放大。使用机器学习工具来优化此类系统将是“首创”，并将为基准公司提供竞争优势，并使加拿大在节能流程、减少二氧化碳足迹和人工智能等关键领域受益。该项目还通过阿尔伯塔省创新校园阿尔伯塔小企业参与计划寻求援助。******

Benchmark International Inc. is an Edmonton based company involved in the manufacture of oxygen concentrators under the brand name of Oxygen solutions Inc. These units concentrate oxygen from atmospheric air using a physical separation process called adsorption. The units are typically installed in remote locations where bottled oxygen is not available or expensive, such as aquaculture, waste water treatment and mining facilities. In order to gain market share in this competitive business, the company is seeking to improve the performance of their oxygen concentrators with the ultimate goal of reducing unit size and energy consumption. This translates into reduction of both cost of oxygen and the associated CO2 footprint (through reduction of energy consumption). To achieve these goals, Benchmark seeks to collaborate with researchers at the University of Alberta (UofA) with expertise in adsorption process design. The UofA group will apply sophisticated mathematical tools to model a small-scale oxygen concentrator to be constructed by Benchmark exclusively to support this project. Using these mathematical models, we will test three commercial materials that are claimed to improve process performance. The unique aspect of this project will be the development and deployment of machine-learning-based optimization tools that are capable of exploring the operating space to identify operating conditions that reduce the overall costs. This will save 1000s of hours of engineering labor and associated costs in testing the different materials. The results from the modeling efforts will be validated first on the small-scale system and then scaled-up for a commercial instrument. The use of machine-learning tools for optimizing such systems will be "first-of-a-kind" and will provide Benchmark with a competitive edge and benefit Canada in key areas such as energy-efficient processes, reduction of CO2 footprint and artifical intelligence. The project also seeks assistance through the Alberta Innovates Campus Alberta Small Business Engagement Program.******