Ziel der Dissertation ist, die Bereitstellung positiver Minutenreserve durch dezentrale Klein-KWK-Anlagen unter technischen, wirtschaftlichen und ökologischen Gesichtspunkten zu bewerten. Die Grundlage einer solchen Bewertung bilden detaillierte Kenntnisse des Abrufverhaltens positiver Minutenreserve. Deshalb wird die Inanspruchnahme positiver Minutenreserve untersucht und als stochastischer Prozess modelliert. Auf dieser Basis erfolgen Betrachtungen zur Dimensionierung der Wärmespeicher, zum Netzbetrieb mit hoher Dichte dezentraler Klein-KWK-Anlagen und zur Wirtschaftlichkeit eines solchen Konzeptes. Die abschließende ganzheitliche Betrachtung liefert Hinweise zur optimalen Betriebsweise der Klein-KWK-Anlagen. / From the technical point of view, virtual power plants consisting of small decentralized co-generation plants are able to provide positive tertiary reserve power for the European electricity transmission grid. For serious analyses, detailed knowledge about the load-characteristic of called reserve power is essentially. In order to examine grid operation, heat storage capacity and optimized power plant operation, the switch-on times of co-generation plants and the co-generated heat during reserve power provision have to be estimated. Aiming this, the called positive tertiary reserve power in Germany is analyzed and a mathematical model of the call-characteristic is synthesized. Furthermore, the results of examining grid operation, optimizing heat storage capacity and power plant operation are given. Calls of positive tertiary reserve power usually occur suddenly, non-scheduled and jumpy. Sometimes, there are single calls. Usually, calls occur clustered, i.e. one call is directly followed by further calls. Positive reserve power is much higher frequented under peak-load conditions than under base-load conditions. The characteristic of calling positive tertiary reserve power deeply depends on the control area. From the mathematical point of view, a Poisson-process fits non-scheduled and jumpy occurring events. Each jump marks a call date of positive tertiary reserve power. The values of the called power fit a logarithmical normal distribution. The lengths of the call-clusters satisfactorily fit a geometrical distribution. The expected value of called reserve energy is modeled dependent from the time of the day. The model is essential for simulating all combinations of switch-on times of co-generation plants and of co-generated heat volumes that might occur during providing reserve power. Aiming to optimize the installed heat storage capacity, the quote of heat use has to be examined. From both technical and ecological point of view, installing large heat storages is desirable in order to use all co-generated heat. From the economical point of view, installing any heat storage is not sensible. The solution of this trade-off is installing a heat storage that guarantees less or equal CO2-emissions than a conventional power plant fired with natural gas. The results of this thesis lead to 1 kWh heat storage capacity per 1 kW installed electrical power as rule of thumb. Concerning grid operation in steady state, a much higher density of co-generation plants than expected is technically installable. A general rule for extending the installable decentralized power cannot be deducted. Examining economics, decentalized co-genertation plants are desired to provide balancing power during their stand-by times. Building a virtual power plant only in order to provide reserve power is not economically sensible. From the power plant owners' view, providing positive tertiary reserve power by small decentralized co-generation plants is generally sustainable.

Ziel der Dissertation ist, die Bereitstellung positiver Minutenreserve durch dezentrale Klein-KWK-Anlagen unter technischen, wirtschaftlichen und ökologischen Gesichtspunkten zu bewerten. Die Grundlage einer solchen Bewertung bilden detaillierte Kenntnisse des Abrufverhaltens positiver Minutenreserve. Deshalb wird die Inanspruchnahme positiver Minutenreserve untersucht und als stochastischer Prozess modelliert. Auf dieser Basis erfolgen Betrachtungen zur Dimensionierung der Wärmespeicher, zum Netzbetrieb mit hoher Dichte dezentraler Klein-KWK-Anlagen und zur Wirtschaftlichkeit eines solchen Konzeptes. Die abschließende ganzheitliche Betrachtung liefert Hinweise zur optimalen Betriebsweise der Klein-KWK-Anlagen. / From the technical point of view, virtual power plants consisting of small decentralized co-generation plants are able to provide positive tertiary reserve power for the European electricity transmission grid. For serious analyses, detailed knowledge about the load-characteristic of called reserve power is essentially. In order to examine grid operation, heat storage capacity and optimized power plant operation, the switch-on times of co-generation plants and the co-generated heat during reserve power provision have to be estimated. Aiming this, the called positive tertiary reserve power in Germany is analyzed and a mathematical model of the call-characteristic is synthesized. Furthermore, the results of examining grid operation, optimizing heat storage capacity and power plant operation are given. Calls of positive tertiary reserve power usually occur suddenly, non-scheduled and jumpy. Sometimes, there are single calls. Usually, calls occur clustered, i.e. one call is directly followed by further calls. Positive reserve power is much higher frequented under peak-load conditions than under base-load conditions. The characteristic of calling positive tertiary reserve power deeply depends on the control area. From the mathematical point of view, a Poisson-process fits non-scheduled and jumpy occurring events. Each jump marks a call date of positive tertiary reserve power. The values of the called power fit a logarithmical normal distribution. The lengths of the call-clusters satisfactorily fit a geometrical distribution. The expected value of called reserve energy is modeled dependent from the time of the day. The model is essential for simulating all combinations of switch-on times of co-generation plants and of co-generated heat volumes that might occur during providing reserve power. Aiming to optimize the installed heat storage capacity, the quote of heat use has to be examined. From both technical and ecological point of view, installing large heat storages is desirable in order to use all co-generated heat. From the economical point of view, installing any heat storage is not sensible. The solution of this trade-off is installing a heat storage that guarantees less or equal CO2-emissions than a conventional power plant fired with natural gas. The results of this thesis lead to 1 kWh heat storage capacity per 1 kW installed electrical power as rule of thumb. Concerning grid operation in steady state, a much higher density of co-generation plants than expected is technically installable. A general rule for extending the installable decentralized power cannot be deducted. Examining economics, decentalized co-genertation plants are desired to provide balancing power during their stand-by times. Building a virtual power plant only in order to provide reserve power is not economically sensible. From the power plant owners' view, providing positive tertiary reserve power by small decentralized co-generation plants is generally sustainable.

Electricity grids require the ancillary services frequency control, grid operation, re-establishment of supply and voltage stability for a proper operation. Historically, conventional power plants in the transmission grid were the main source providing these services. An increasing share of decentralized renewable energy in the electricity mix causes decreasing dispatch times for conventional power plants and may consequently lead to a partial replacement of these technologies. Decentralized energy sources are technically capable of providing ancillary services. This work focuses on the provision of reactive power for voltage stability from decentralized sources. The aim is to answer the question of how voltage stability and reactive power management can be achieved in a future electricity system with increasing shares of decentralized renewable energy sources in an economical and efficient way. A methodology that takes reactive power and voltage stability in an electricity system into account is developed. It allows for the evaluation of the economic benefits of different reactive power supply options. A non-linear and a linearized techno-economic grid model are formulated for this purpose. The analysis reveals an increasing importance of reactive power from the distribution grid in future development scenarios, in particular if delays in grid extension are taken into account. The bottom-up assessment indicates a savings potential of up to 40 mio. EUR per year if reactive power sources in the distribution grid provide reactive power in a controlled manner. Although these savings constitute only a small portion of the total cost of the electricity system, reactive power from decentralized energy sources contributes to the change towards a system based on renewable energy sources. A comparison of different reactive power remuneration mechanisms shows that a variety of approaches exist that could replace the inflexible mechanisms of obligatory provision and penalized consumption of reactive power that are mostly in place nowadays.

Electricity grids require the ancillary services frequency control, grid operation, re-establishment of supply and voltage stability for a proper operation. Historically, conventional power plants in the transmission grid were the main source providing these services. An increasing share of decentralized renewable energy in the electricity mix causes decreasing dispatch times for conventional power plants and may consequently lead to a partial replacement of these technologies. Decentralized energy sources are technically capable of providing ancillary services. This work focuses on the provision of reactive power for voltage stability from decentralized sources. The aim is to answer the question of how voltage stability and reactive power management can be achieved in a future electricity system with increasing shares of decentralized renewable energy sources in an economical and efficient way. A methodology that takes reactive power and voltage stability in an electricity system into account is developed. It allows for the evaluation of the economic benefits of different reactive power supply options. A non-linear and a linearized techno-economic grid model are formulated for this purpose. The analysis reveals an increasing importance of reactive power from the distribution grid in future development scenarios, in particular if delays in grid extension are taken into account. The bottom-up assessment indicates a savings potential of up to 40 mio. EUR per year if reactive power sources in the distribution grid provide reactive power in a controlled manner. Although these savings constitute only a small portion of the total cost of the electricity system, reactive power from decentralized energy sources contributes to the change towards a system based on renewable energy sources. A comparison of different reactive power remuneration mechanisms shows that a variety of approaches exist that could replace the inflexible mechanisms of obligatory provision and penalized consumption of reactive power that are mostly in place nowadays.

Die Dissertation liefert einen Beitrag zur Modellierung und optimalen Ansteuerung von vernetzten hybriden Energiesystemen. Die Arbeit beschreibt die Entwicklung einer modellprädiktiven Regelung (MPC) für konkrete Energiesysteme. Dafür wird eine Betrachtung zu berücksichtigender wirtschaftlicher und technischer Rahmenbedingungen vorgenommen, die zur Formulierung notwendiger Nebenbedingungen für die MPC genutzt wird. Für den Umgang mit dem ansteigenden Rechenbedarf der MPC bei steigender Systemzahl wird ein alternativer Ansatz auf Basis eines auktionsbasierten Algorithmus vorgestellt. Die Modellierung der Energiesysteme wird ausgehend von einer bestehenden Laboranlage vorgenommen. Die Erprobung der vorgestellten Ansätze erfolgt in einer Simulationsumgebung, die die Untersuchung verschiedener Szenarien erlaubt. Im Rahmen der Simulationsszenarien mit unterschiedlicher Systemzahl und Zusammensetzung der Energie-systeme wird eine Sensibilitätsanalyse der vorgestellten MPC vorgenommen. Die Interpretation der Ergebnisse erfolgt auf Basis numerischer und empirischer Bewertungskriterien.