Validation of Models for Analysis of the Flexibility of the Swedish Power System

Abstract: The Swedish parliament has passed a planning framework to increase wind power production and have the annual production of 30 TWh wind power in 2020. The expansion of a continuously varying generation would result in an increased need for the capability of power system to keep the balance between generation and consumption. Therefore, it is important to study the flexibility of Swedish power system. Two models of Swedish power system are studied in this thesis work. The first model is a model of Swedish hydro power system which has been developed at KTH. The KTH model is formulated as a large linear optimization problem simulated in GAMS platform. It has a detailed representation of large hydro power plants but presents a simple model of electricity market and trading to other areas. The other model is Apollo which is developed by Sweco Company. Apollo is also formulated as an optimization problem and is a market model which uses a simplified model of hydro power system. The objective of this thesis work is to exchange data between the two models in order to compare, validate and if possible improve the models. To exchange data, the inputs and some outputs of Apollo are used as the inputs of KTH model and finally the outputs of KTH model is compared with the corresponding outputs of Apollo. There are some differences between the two models that must be removed in order to exchange data. All of differences except one of them are removed by data adjustment. The different methods that are used to remove those differences are discussed in the report. Due to the remaining difference and different efficiencies in the two models, scenarios cannot be directly transformed from Apollo to the KTH model. Therefore, three methods are introduced as compensation for the remaining differences. After applying those methods the same results can be obtained in the two models. As a result of the work on the data exchange some improvements are implemented in the KTH model and some improvements are identified and proposed for future work. The improvements are toward removing all the differences between the two models and make the models more similar to the real Swedish hydro power system. It is also concluded from the results that the Apollo hydro power schedules are feasible according to KTH model of hydro power system. This shows that Apollo does not overestimate the flexibility of Swedish hydro power system in the tested scenarios.