Einfluss von Unsicherheiten auf die Anfahrtsentscheidung der Netzreserve

• Impact of uncertainties on the start-up decision of the grid reserve

Siemonsmeier, Marius; Moser, Albert (Thesis advisor); Schürhuber, Robert (Thesis advisor)

1. Auflage. - Aachen : printproduction M. Wolff GmbH (2022)
Book, Dissertation / PhD Thesis

In: Aachener Beiträge zur Energieversorgung 218
Page(s)/Article-Nr.: x, 154 Seiten : Illustrationen, Diagramme

Dissertation, RWTH Aachen University, 2022

Abstract

As a result of various changes in the electricity supply system, the transport volume in the German transmission grid is increasing. As the necessary grid expansion projects are delayed, the need for congestion management is growing. This includes market and grid related measures as well as additional reserves, whose operations are determined within the short-term operational planning. In this context, the start-up decision of the grid reserve poses a major challenge because it has long lead times and may only be dispatched in lower priority in the legal order of measures. In operational practice, the grid reserve's start-up decisions are derived by means of forecast-based grid operation simulations. To manage the high uncertainties, safety margins based on empirical values are used. On the one hand, this approach to taking uncertainties into account can lead to start-ups that would not have been necessary ex post and thus do not comply with the legal order of measures. On the other hand, insufficient safety margins can underestimate the uncertainties and endanger the grid security. Against this background, the impact of uncertainties on the grid reserve's start-up decision is investigated in this dissertation and different approaches for the differentiated depiction of uncertainties in grid operation simulations are compared. For this purpose, a two-stage methodological approach is developed. The first stage models the uncertainties on the basis of scenarios. Here, a Monte Carlo simulation using copulas is applied to generate consistent scenarios of the feed-in of wind mills and photovoltaic units, the consumer load and the weather-dependent current limits of transmission lines using historical weather and load forecast errors. On that basis, a power generation dispatch simulation complements the respective power plant dispatch and electricity exchanges. These scenarios enter the second process stage, which can be executed as a probabilistic or stochastic grid operation simulation. In their core, both simulations are based on a two-stage optimization problem that minimizes the grid reserve starts and the dispatch of the measures in terms of quantity and costs to relieve all congestions. In the probabilistic simulation, the optimization is performed separately for each scenario, whereas in the stochastic simulation, the approach optimizes expected values over all scenarios. The exemplary investigations based on a future scenario in 2025 show that the range of uncertainties in the grid reserve's approach decision are considerable and extreme scenarios in particular show high deviations. Thus, the manifold uncertainties lead to grid congestion at different levels, at different times or in different geographical locations. As a consequence, there are also significant differences between the start-up decisions of the grid reserve power plants in the number and selection of plants. Overall, it appears that the derivation of the start-up decision based on the probabilistic simulation is not adequate to the trade-off between the order of measures and grid security. In contrast, the stochastic simulation considering jointly the uncertainties in the optimization problem is superior because, on the one hand, the start-ups are more often in accordance with the legal order and, on the other hand, possible combinations of required grid reserve power plants are identified, contributing to grid security.