CHP process evaluation, optimization and troubleshooting in OpenModelica

Abstract: As an important part in creating a sustainable energy sector the most prevalent sources of energy production must be in the forefront of being both reliable and effective. In Sweden district heating is utilized in more than half of all households and must therefore be subject to the same scrutiny and analyses as other energy producing technologies. A big challenge here is to optimize a power plant with regards to its specific layout through changing fuel types, amount recirculated flue gas, air flow etc. while also being able to analyse the measured data to draw conclusions on component performances in steam production or flow handling with the goal of predicting unplanned downtimes. To address this in a way that provides a user with a simple interface, a flexibility in handling inputs while producing close to ideal results for conventional boiler components this thesis encompasses the development of a CHP modelling package in OpenModelica, an analysis of the program, the method of model design and of the results produced. As OpenModelica is an open source program and free for anyone to use while it could, if a model is designed reliably, provide the necessary groundwork for closer analysis of boiler components and close operation optimization. To make a viable analysis and comparison a case study in the form of Boiler 1 at Hedemora Energi has been used. The designing procedure of the model involves producing correlations of heat transfer, pressure drops and mass flow rates of water. These are then implemented in components to produce output temperatures, pressures and loads close to what should be in a clean and recently repaired boiler. Many tries in using different numerical and thermodynamic solution methods in different components has been conducted leading to a model that can take any solid fuel compositions, mass flows, water contents, flue gas recirculation fractions and other input values into account while producing results with error margins below 2.5% in validation. Running the model during periods preceding unplanned downtimes also indicated divergences in steam output temperatures and flows which in analysis may be useful in preventing the different downtimes while diagnosing the boiler for future maintenance. As the modelling library allows one to dimension components freely the library could fit any boiler using components of the same sorts thus showing flexibility.