Representing variable renewable energies in long-term energy systems modelling. Development of a modelling tool for providing initial consultation to German federal states
Summary
With increasing deployment of variable renewable energies (VRE), as well as ongoing structural changes in energy demand, decision-makers in German federal states must make difficult decisions with regard to long-term power supply investments and policy design. Model-based scenario analyses are a common means to derive long-term strategies in this matter. However, considering their typically high computational complexity and data needs, established modelling tools can be considered inappropriate for providing initial consultation, i.e. a first approximation of developments necessary for attaining certain long-term outcomes. For this reason, the present study investigates to what extent a model with reduced levels of detail can yield plausible results regarding the dynamics of VRE-based power supply and energy demand.
The centrepiece of the study is the development of the novel modelling tool ‘FederalPlan’. As for energy demand, it features a bottom-up accounting framework for projecting energy needs in four major demand sectors. Exogenous parameters include fuel and technology substitution towards electrification, improvements in energy efficiency, as well as socio-economic effects. These projections are used for endogenous projections of future electrical load curves and demand-side-management (DSM) potentials. Considering power supply, the model optimises investments and dispatch of generators and flexibility options, in order to minimise power system costs from a socio-economic perspective. Technologies are selected from a diverse portfolio of thermal power plants, cogeneration plants, storage facilities, DSM interventions, as well as power-to-gas and power-to-heat converters. One model run results in a cost-effective power supply configuration at given parameters and policy constraints for a single target year.
Robustness of the model outputs is evaluated by performing a comparative assessment with an established model-based scenario study for the state of Baden-Württemberg. Under equal boundary conditions, the two models match with regard to the deployment of power-to-heat and CHP plants in the target year 2050. Deviations include the extent and composition of reserve capacity needed to cover system load. Accordingly, using either of the two approaches, policy recommendations may differ in providing initial consultation to decision-makers. In addition to the model comparison, sensitivity analyses are carried out for the FederalPlan tool. Among other parameters, the hourly shape of the endogenously modelled load curve is found to have an influence on the deployment and dispatch of power plants and flexibility options. Overall, the modelling tool is considered to yield reasonable results. Given its low computational complexity and data needs, it may proof useful for providing initial consultation to federal state decision-makers with regard to long-term planning of power supply and associated policy design.