ROMEO – Recovery-robust optimization of energy production

n the ROMEO project the main goal is to develop and evaluate a novel method, including new optimization models and algorithms, for energy production planning aimed at overcoming challenges with uncertainties in input data.

Project name: ROMEO – Recovery-robust optimization of energy production

Research specialisation: Future Energy Center

Primary funding organisation(s): Future Energy Center

Secondary funding organisation(s): Mälarenergi AB, Vattenfall AB, Frontway AB

Co-operating/Co-production partners: SICS Swedish ICT Västerås AB

Project leader:  Markus Bohlin

Project period: 20150201-20180131

Funding: 8,5 MSEK of which 3,99 MSEK from KK-stiftelsen

 

Description of the project

Energy production planning is the process for scheduling when one or several interconnected production units, such as boilers and steam and water turbines, should be active (the Unit Commitment Problem), as well as deciding the level of production for the active units (the Economic Dispatch Problem). However, an optimal solution to the energy production problem may easily become invalidated given that input data, such as heat demand or electricity price, differ from the predictions.

In the ROMEO project the main goal is to develop and evaluate a novel method, including new optimization models and algorithms, for energy production planning aimed at overcoming challenges with uncertainties in input data. The approach is based on recovery robustness for the Unit Commitment Problem, in where recovery actions from the load distribution are taken into account already at the planning stage, and decomposition strategies for optimally solving large-scale production planning problems. Furthermore, it will aim at solving large problem instances in reasonably short time, and allow a high level of robustness against prediction errors with relatively small safety margins.

In the ROMEO project, Mälardalen University will collaborate with the companies Mälarenergi AB, Vattenfall AB and Frontway AB, and the developed approach will be used for several experimental studies of the two cases provided by Mälarenergi AB and Vattenfall AB, respectively. From Mälardalen University, Markus Bohlin, Maria Nelénius, Daniel Andrén and Konstantinos Kyprianidis will participate in the project.