On the design of renewable-based utility plants using time series clustering

This work addresses the design of utility plants incorporating technologies to process biomass, manure, solar radiation, and wind to generate steam and electricity. To capture the hourly and interannual variability of solar and wind resources, an optimization framework is proposed. The framework involves clustering methods to capture the variability of solar and wind resources and a multiperiod design optimization problem minimizing operating and investment costs. Results show that the number of representative days impacts both the minimum cost and the plant topology. To study this impact, a methodology using multiple samples of representative days and two approaches to select design solutions are proposed and discussed through extensive computational results. Overall, two plant topologies were identified, one integrating syngas and the other a biomass boiler. The biomass-based plant showed 1% lower investment and 10% lower operating costs but requires additional make-up power from an external grid.