Felix Müsgens

Quantifying Market Power in the German Wholesale Electricity Market Using a Dynamic Multi-Regional Dispatch Model

This paper quantifies the degree of market power in the German wholesale electricity market. A dispatch model simulates competitive marginal costs. In addition to common input factors like plant capacities, fuel prices and load structures, the model also incorporates international power exchange and dynamic effects like start-up costs and hydro storage plant dispatch. The simulated prices are subsequently used as a benchmark for observed electricity prices. The analysis reveals significant market power in the German electricity market, mainly exhibited during peak periods. Producer surplus is also increased significantly due to strategic behavior.

Modelling Dynamic Constraints in Electricity Markets and the Costs of Uncertain Wind Output

Building on models that represent inter-temporal constraints in the optimal production decisions for electricity generation, the paper analysis the resulting costs and their impact on prices during the day. We linearise the unit commitment problem to facilitate the interpretation of shadow prices. Analytic research gives insights for a system with one technology and numeric implementation provides results for the German power system. The model is expanded to a stochastic optimisation with recourse. The model is used to calculate the cost of wind uncertainty and the value of updating wind forecasts.

Modelling start-up costs of multiple technologies in electricity markets

This paper deals with the dispatch problem in providing electric power with minimal costs using different technologies. Initially, we describe this problem in terms of a linear program. This enables us to take generally neglected start-up costs into account. The main result is the explicit solution of a simplified linear program which provides us with a better understanding of the ‘start-up cost’ effects. Furthermore, we show that dominated technologies should be used in the case of limited availability of efficient technologies.

Increasing flexibility of combined heat and power plants with Power-to-Heat

Increasing shares of renewable energy sources in the German electricity market increase the need for flexibility in order to reduce curtailment of low variable cost RES generation. Power-to-Heat (PtH) is an option to provide additional flexibility. This paper evaluates the flexibility gain a PtH installation provides using the example of adding it to a district heating facility. Therefore we compare a setting consisting of a CHP unit and a conventional gas-fired boiler unit with an extended setting including the PtH unit. We present all relevant changes between the two unit endowments. The results show that the regulatory framework has a crucial influence on usage and profitability of a PtH unit.

The effect of variable renewable energy sources on the volatility of wholesale electricity prices — A stylized full cost approach

We analyze the effect of an increasing share of variable renewable energy sources (RES) on the electricity price, particularly on the price variance. To that end, we develop and apply an electricity spot market dispatch and investment model which is formulated as a linear optimization problem (LP) and minimizes the total system costs. We use a full cost approach, i.e. investment costs of the installed capacity are taken into account. In order to focus on theoretical effects, we analyze a stylized electricity system with three generation technologies. In a system with low shares of wind generation, the electricity price variance decreases with additional wind. However, we observe the opposite at higher shares of wind mainly driven by wind curtailment and start-up effects in the thermal system.

Political shocks and efficient investment in electricity markets

This paper presents a fundamental electricity market model capturing many essential features of investments in electricity markets - fuel price developments, investment and generation costs, demand and dynamic effects such as start-up costs and (pump) storage dispatch. The partial equilibrium model minimizes the total costs of the electricity system ensuring demand coverage. The model optimizes both long-run investment as well as short-run dispatch decisions. From the optimal solution, the optimal future generation technology mix and dispatch can be derived. Furthermore, using the concept of shadow prices in mathematical programming, we calculate electricity price predictions. The model is applied to analyze the effect of recent political shocks in German nuclear energy policy on the electricity market and the power plant portfolio. The effect is quantified using the example of a specific combined cycle gas turbine project. We find a significant impact on the future net revenues.

Electricity market equilibria and intermittent renewables — A stochastic approach

The electricity system moves step-by-step towards a system built on the use of intermittent renewable energy sources (RES). The implications raise a variety of questions. We provide deeper insights into the impact of uncertain wind power generation on wholesale electricity prices. In particular, we analyse long-term equilibrium prices and their volatility. We develop and apply a stochastic electricity spot market dispatch and investment model with recourse. We perform our calculations on a “green field” and apply a full cost approach. Thus, we investigate market equilibria in the electricity market. We show that the variance of electricity prices remains stable with the introduction of intermittent wind power production, but increases with the appearance of wind curtailment. As we come closer to a real world application, the price variance increases with increasing uncertainty in the market.

How to improve standard load profiles: Updating, regionalization and smart meter data

Currently, most distribution system operators in Germany estimate non-real-time metered consumption profiles based on “Standard Load Profiles” developed in the late 1990s by the German Association of Energy and Water Industries. However, as both consumption behavior and consumer structure change over time, their predictive power may have deteriorated. In addition, they do not account for regional differences within Germany. Therefore, we compared their forecasting accuracy with two newly developed alternative standard load profiles, differentiating between households and commercial enterprises. We calculated the new profiles based on regional, more up-to-date aggregated consumption data and a limited set of smart meter data. Furthermore, we varied the number of seasons and day types included in the profiles. A comparison of our new load profiles with the existing Standard Load Profiles revealed significant improvements in forecasting accuracy. Improvements are mainly resulting from improved input data (regional and more recent data set), but the utilization of smart meter data as well as variations in day types and seasons also reduced forecast errors.

Dynamic forecast combinations of improved individual forecasts for the prediction of wind energy

We study the prediction performance of different improved individual wind energy forecasts in various static and dynamic combination processes. To this end, we develop a combined error minimization model (CEMM) based on nonlinear functions. This approach reflects the nonlinear nature of weather and especially of wind energy prediction problems. Based on the model, we construct significantly improved individual forecasts. The corresponding time dependent model coefficients are determined by dynamic OLS (ordinary least squares) regression and Kalman filter methods. The former method shows a slightly better performance than the Kalman filter based approaches. Further improvements can be achieved by a combination of these improved wind energy forecasts. In this case, the combination coefficients are calculated from a static and two dynamic OLS regressions. The resulting forecasts are characterized by a further increased prediction accuracy compared to the combination of the uncorrected forecast data and can outperform a given benchmark.