Johannes Schwippe

Pan-European market and network simulation model

The liberalization of the energy market enables a pan-European energy trade. Together with an increasing installed capacity of renewable energy sources this provokes new challenges for the European transmission system. New technologies and legal conditions can accomplish them. A combined simulation of the European energy markets and the energy system, regarding their interdependencies, is necessary to evaluate possible situations. In this paper a combined market and network simulation tool for the pan European Transmission network is presented. The applicability of this combined model is verified with published load flow results and market situations. The main scope of application is the network extension planning of the western European transmission network. The aggregated network model provides a sufficient macroscopic perspective with typical time series, because detailed information of single lines is not required.

Extension of a probabilistic load flow calculation based on an enhanced convolution technique

Traditional algorithms used in grid operation and planning only evaluate one deterministic state. Uncertainties introduced by the increasing utilization of renewable energy sources have to be dealt with when determining the operational state of a grid. From this perspective the probability of certain operational states and of possible bottlenecks is important information to support the grid operator or planner in their daily work. From this special need the field of application for Probabilistic Load Flow methods evolved. Uncertain influences like power plant outages, deviations from the forecasted injected wind power and load have to be considered by their corresponding probability. With the help of probability density functions an integrated consideration of the partly stochastic behaviour of power plants und loads is possible. In this context an extension to a convolution based probabilistic load flow is present in this paper. The extension reduces the already limited inaccuracy introduced by network model simplifications. Aspects like accuracy improvement and computation time in comparison to existing method are covered in detail.

Probabilistic Load Flow Calculation based on an enhanced convolution technique

The increasing uncertainties grid operators have to face in their every-day work lead to the necessity for fast and accurate information about the probability of certain operational states of the operated network. Algorithms traditionally used in grid operation and planning are only able to evaluate discrete operational states of the particular power grid. A consideration and integration of probabilistic data can only be done by the analysis of selected discrete states followed by an interpolation. The algorithm presented in this paper falls into the category of Probabilistic Load Flow Calculation with a new approach for the modeling of the underlying network, as well as measures to reduce the inaccuracy introduced by linearization. Aspects like accuracy and calculation time in comparison to existing approaches are covered in detail.

Network extension planning considering the uncertainty of feed-in from renewable energies

The integration of fluctuating renewable energy sources is a challenge for grids on all voltage levels. The feed-in of these energy sources is difficult to forecast and lies in certain limits of variation. Renewable energy plants are built at locations with a high potential e.g. high wind speeds that allow high utilization factors. However, grid constraints caused by the construction of energy plants are not considered for the choice of the location. The increasing amount of installed renewable energy plants causes the need for network reinforcements. For network planning, the grid operators have to anticipate the highest possible electricity feed-in. But this scenario only happens very rarely when several types of fluctuating generation, such as wind and photovoltaic, are installed. This paper presents a new approach for determining scenarios for an efficient network planning process which depends on the allowed feed-in from renewable energies. The reduction of the injected power and energy of renewable energies in relation to a reduction of network extension costs is discussed.

Transmission system congestion analysis based on a European electricity market and network simulation framework

The coupling of electricity markets and the increasing feed-in of renewable energy sources (RES) cause a rising number of congestions in the European transmission network. In order to ensure a secure network operation suitable measures for resolving these congestions need to be found. In this context, a methodology for analyzing congestions in transmission systems is presented in this paper. The methodology is based on a European electricity market and network simulation framework including internal congestion management by redispatch and control of high-voltage direct current (HVDC) links. In particular, the methodology enables to determine the frequency of occurrence, the causes as well as a ranking of existing congestions regarding their constraining effect on social welfare. The simulation results of a Europe 2030 case study underline that the introduced methodology offers a detailed insight into the system and can help to identify suitable network extension measures for avoiding overloads.

Analyzing the impact of forecasting errors on redispatch and control reserve activation in congested transmission networks

Uncertain forecasts regarding the feed-in by Renewable Energy Sources (RES) can cause overloads in the transmission system. Thus, forecasting errors influence both the decision on redispatch for congestion management and on permissible control reserve activation. In this paper, a probabilistic approach is applied for investigating these two effects. After evaluating probabilistic models of forecasting errors for wind power, probabilistic load flow methods are applied for mapping the domain of credible deviations from forecasts to the domain of permissible control reserve distributions. In a case study of the German transmission grid, the impact of forecasting errors on necessary preventive redispatch and RES curtailment as well as on the permissible space of control reserve activation is analyzed. As part of this, the benefits of dividing the decision on preventive redispatch into a day-ahead and an intraday decision is evaluated.

Benefits of coordinated control reserve activation and grid management — A probabilistic load flow analysis

Joint control reserve procurement and activation by several Transmission System Operators can have significant technical and economic benefits. In congested transmission networks, however, the activation of control reserve can be constrained by impending overloads. In this paper the benefits of an optimization to coordinate control reserve activation and grid management considering uncertainties caused by forecast errors is outlined. The effect of these uncertainties are measured by using probabilistic load flow methods based on convolution technique. A joint optimization is presented to minimize the costs of supplying control reserve and conducting redispatch, taking into account HVDC lines, variable costs of power plants and the impact of control reserve activation on lines instead of optimizing it separately as it is the current proceeding by European Transmission System Operators. In addition, a method to determine the solution space for permissible control reserve distributions is presented. In a case study of the German transmission grid for the year 2024 the costs of the joint optimization are contrasted to the costs using the normal merit order and additional redispatch as it is the current market scheme. With the simulation for one year, the solution spaces are calculated to show, which power plants can supply control reserve in every hour without causing redispatch measures and whether the existing power plants in 2024 are sufficient to provide control reserve activation without impending overloads.

Load Flow and Dynamic Model

Based on the general requirements, particular consideration has to be given to different scenarios of joint operation of the ENTSO-E- and IPS/UPS-systems. To provide a basis for the following investigations in this book, this chapter focuses on the development of an aggregated network model of these power systems which will be used as general test case in the following chapters.

Scenario-based analysis of investment needs for the German transmission grid

The increase of renewable energy sources in Europe is connected with a new challenge for the transmission system. In Germany, the decrease of generation capacity in central and southern regions and new conventional and renewable power plants located in the North necessitate investments in the transmission grid. In this paper, the need of grid expansions is identified using a genetic optimization model. The algorithm finds the cost-minimal investments projects for a given network topology. The model is applied on a reduced network model of the German transmission system. A wholesale market model generates the input data that consists of local distribution of generation and load. The model is applied on different surveys assigned by the federal ministry and finally, the quality of the results is discussed.

Enhancement of load flow computation methods to improve the convergence with automated measures based on stability analyses

Future scenarios of generation and load are predicted with market models which consider possible prices and demographic developments. However, nodal reactive powers which are critical in terms of the convergence of load flow calculations can only be estimated. The reactive power demand of heavily stressed power systems differs from regular burdened scenarios, whereby the estimation of the nodal reactive powers is complicated and minor differences affect the convergence characteristics of load flow calculations significantly. The algorithm that is presented in this paper modifies the reactive nodal powers during the load flow calculation if it does not converge. Regions of instability and measures for convergence achievement are determined with a modal stability analysis.