Robust optimization of the energy concept of an industrial process w.r.t. uncertain energy costs and environmental conditions.

In order to limit the global warming to 1.5 °C as decided by the Paris Agreement, the greenhouse gas emissions have to be dramatically reduced within the next couple of years. In order to realize this in industrial sectors with low to medium temperature requirements, such as food or pulp and paper industry, fossil fuels must be replaced by renewable energy sources to generate electricity, steam and process heat. To realize this in a most economical way, a deterministic multi-objective optimization and a robust scalar optimization of the renewable energy concept of an industrial process are carried out, whereby the robust optimization takes uncertainties in the assumptions of the price for purchasing natural gas and the solar radiation into account. To set up the automated optimization processes, suitable mathematical descriptions of generation units (e.g. photovoltaic systems, wind turbines, solar thermal systems), conversion units (e.g. heat pumps, boilers) and storages (e.g. thermal, electrical) are described first. The comparison of both optimization approaches shows that deterministic optimization is able to find very good solutions, but that the spread of the objectives is significantly larger than with robust optimization when there are uncertainties in the assumptions. The robust optimization thus expands the portfolio for selecting suitable energy concepts and enables future scenarios and developments to be considered, which is particularly necessary in dynamic and heavily changing environments. Furthermore, considering the typically long operating times of industrial plants, economically well-founded decisions can be made, which can have a positive effect on the current restraint to make necessary investments..