Modelling, optimisation and techno-economic analysis of a process to produce oxygen-enriched air by water desorption. Application and design for a hospital in sub-Saharan Africa

摘要

Oxygen, along with oxygen enriched air, is used in multiple applications, namely chemicalindustry and healthcare industry. Currently there are three main ways to get oxygen: cryogenicdistillation, water electrolysis and membrane separation, being the first the most widely usedon industrial scale. It allows the production of highly concentrated oxygen but it has one maindownside, which is the high energy spending. The other processes present a lower costefficiency.A new process to produce enriched air was analysed. It is based on the ability of water to absorband desorb oxygen at different thermodynamic equilibria, varying temperature and pressure.Firstly, based on empirical tests carried out in laboratory it was attempted to create amathematical model. The first attempt was modelled with neural networks, but the smallavailable empirical set of information did not allow to get a trustworthy model. The secondattempt, adjusted by square errors minimising, outperforms the simulations done withcomputational software based on theoretical equations.Secondly, the process was simulated by computer solving the Rachford-Rice equation with anExcel worksheet and with the computer software SimSci PRO/II.Thirdly, the process model was applied to three different scenarios: chemical industry, shippropulsion and healthcare industry. In the chemical and the oil industry oxidation processes arecommon practice, ships use large quantities of air to oxidise the fuel in the combustion processthat propels them and oxygen is important in healthcare in almost every lung related issue.Regarding the two first scenarios, the outcomes resulted in flows of water so massive thatrendered the whole process unfeasible. The application to healthcare was focused on contextswhere there is no possibility to obtain bottled oxygen or electricity supply is unreliable, i.e. sub-Saharan Africa. The process was designed to the Sounon-Sero hospital in Nikki (Benin), as a proxyfor any hospital in a similar context.In fact, based on the information provided by that hospital, after traumatisms and malaria, it islung-related infections that account for the highest number of patients attended at Sounon-Sero. A special emphasis has been put in designing a simple and reliable system with a demandsizing based on supply side, given the maximum capabilities of the equipment that can be foundin Nikki. This opens new research opportunities regarding a more precise demand sizing orbudget, which is to be done within an on-field venture.