With an increasing share of photovoltaic and wind in the German generation mix, the power supply reduces its CO2 emission and becomes more weather dependent. In order to guarantee the security of supply, the volatile power output by the renewables asks for compensation by other sources. These backup technologies have to cope with a variety of scarcity cases, ranging from single peaks of uncovered load to longer scarcity periods. During winter times, these scarcity incidents show their most extreme form and overlap in some weeks. The German news calls this phenomenon “kalte Dunkelflaute” or only “Dunkelflaute” (analogously translated “energy drought during darkness and cold”). The need for backup technologies goes not along with a certain cost recovery. As the last dispatched entity in the market, they need to cover their fixed costs by scarcity prices. Depending on various factors, such as the weather conditions and the availability of short-term flexibility, it is difficult to predict the level of scarcity prices. Uncertain investment conditions lead to investment restraints and harm the security of supply. At the same time, frequent scarcity prices would contradict the principle of a secure energy supply with a limited financial burden for the consumers. The regulator is in a dilemma. Rewarding the contribution to the security of supply is an option to stimulate investments in backup technologies. The so-called capacity mechanisms are an intervention into the energy-only-market (EoM) which is known to organize the energy supply in the most efficient way. The principle decision of an intervention and its design requires a sounds judgment and the consideration of many factors. By presenting the challenges of future scarcity moments, the cost recovery conditions for backup technologies and examples for cost recovery measures, the thesis aims to contribute to the on-going discussion by answering the research question “How to maintain the security of supply under extreme weather conditions in a renewable dominated electricity system in the most cost-efficient way?”. It is addressed by a simulation with the agent-based model AMIRIS and accompanying analyses based on literature reviews and calculations. In the beginning, a cost comparison of backup technologies (chapter 3) and a multi-criteria decision analysis for capacity mechanisms (chapter 4) are examined. The most cost-efficient forms of each analysis are processed subsequently. After an introduction into the model (chapter 5) and the simulation process (chapter 6), the variations of scarcity (chapter 7) and cost recovery (chapter 8) according to two contrary weather years and the availability of battery storages are presented. Three scarcity indicators help to characterize the scarcity in the first step and indicate the dimensions of the backup capacity in the second step. A sensitivity analysis to test the robustness of the simulation results is examined in chapter 9. Lastly, two measures to bridge the lacking cost recovery for the backup technologies are presented. The resulting costs of the energy supply for the consumers are compared for the two measures and the system with scarcity prices without any intervention (chapter 10). The thesis describes the challenge of increasingly volatile market conditions and recommends measures to mitigate their risks. The main findings are presented by the four guiding hypotheses in the following. 1. The level of scarcity varies substantially with the weather conditions For every weather condition, single moments occur in which a high residual load and limited flexibility lead to an extreme peak of the uncovered load. The simulation shows the same level for the maximum hourly peak for every scenario. In contrast to that, the aggregated uncovered load per year varies significantly for the extreme weather year 2010 with a low renewables output and the mild weather year 2007 with a high renewable output. The scenario with the weather year 2010 asks for 25 percent more additional energy to cover the demand than the one with the mild weather year 2007. The severity of scarcity becomes explicit when one considers the uncovered load of the longest scarcity period of each simulated year. In the scenarios with the weather year 2010 and 15 GW installed battery storage, it lasts for almost three days and contains 0.25 percent of the yearly uncovered load. This period by mid-February includes the maximum peak of uncovered load and is surrounded by other long scarcity periods. In this sense, it is a stress test for the electricity system. The situation changes for the mild weather year 2007. The longest duration of scarcity asks for less than a half of the energy of 2010. The difference of additionally requested energy and capacity makes it explicit how difficult it is to design a well-tailored backup mix for these scarcity moments. 2. The short-term flexibility providers lower the need for backup capacity but cannot substitute it The emergence of short-term flexibilities, such as battery storage, is a mixed blessing for the security of supply. On the one hand, it balances the energy supply in moments with high and low prices and closes gaps between demand and supply. Looking at the scenario with the extreme weather year, the implementation of 15 GW battery storage divides the uncovered load per year in a half. On the other hand, it reduces the income basis for the backup technologies and is, therefore, a further factor of uncertainty for investments. At the same time, battery storage can hardly address situations of extreme scarcity alone. Only 10 percent of the missing energy can be covered by the battery storage during the longest scarcity period. The maximum uncovered load cannot be reduced at all by the battery storage. The correlation of extreme scarcity peaks and long scarcity periods gives the battery storage hardly any opportunity to charge during the extreme scarcity periods. For instance, more than one-third of the extreme peaks (in this case defined as 20 GW and more) is surrounded by a period of scarcity of 5 hours and longer. The knowledge about future prices impacts the ability of the battery storage to address scarcity hours. If the so-called foresight is extended from one day to one week, the uncovered load per year can be reduced by 20 percent. However, the positive effect of the longer foresight is bounded by the technical limitations of the battery storage. In the presented scenarios, no additional mitigation of the scarcity can be achieved in times of the Dunkelflaute by a foresight of one week. Also, a longer foresight than one week cannot reduce the uncovered load per year any further. Conclusions can be drawn about the suitability of the price sequences to support the bidding of the battery storage. It is noticeable that the longer foresight increases the number of used high prices but does not have the same effect on the low prices. Upcoming high prices seem to be missing to create an additional value by charging more energy during the low prices. This indicates a limited fit of the prices sequences for charging and discharging of the battery storage. All in all, the battery storages are restricted by their technical limitations and the sequence of market prices. Despite their positive effect on scarcity, they cannot substitute long-term backup technologies. 3. The level of cost recovery for backup technologies on the EoM is not sufficient and depends strongly on the weather conditions and the availability of battery storage Due to its ability to react fast to scarcity signals and its relatively low fixed costs, gas turbines supplied with fossil gas are selected as backup technology. The constant level of the scarcity peak for all scenarios gives a clear indication for the required backup capacity. At the same time, the changing request of the backup energy for the different weather conditions and availability of short-term flexibility impacts their cost recovery. In the simulated case, the backup technologies do not leverage their dominant market position in times of scarcity and only bid their marginal cost. This leads to a negative profit margin for every scenario negative. With the implementation of 15 GW battery storage, it becomes four times lower. The optimized bidding with the longer foresight decreases the margin additionally by 25 percent. The lower request of the mild weather year leads to a decrease of 83.5 percent. Assuming that a further expansion of renewables would lead to a larger price spread and a more constant level of scarcity, the implementation of a power-to-methane to create synthetic gas from excess energy is an interesting option. Due to limitations of the model, power-to-methane was not simulated as backup technology, but the insights from the simulation with the gas turbine are transferred in a simplified calculation. It results in a higher (but still negative) profit margin for power-to-methane. At the same time, it needs to be considered that the smaller fixed costs base of the fossil gas application can adapt better to changing requests of backup energy. An implantation of a gas turbine supplied with fossil gas and later with syntactic methane would address both incidents. Solely the availability of suitable backup technologies will not lead to investments without a positive business case. Therefore, some measures to handle the lacking cost recovery are presented in the next paragraph. 4. A well-designed regulatory intervention which rewards the contribution to the security of supply can reduce the costs for the consumer and improves the supply ratio The skepticism about the ability of the EoM to recover the costs of capital intense backup technologies is reinforced by the analysis. At the same time, no final assessment to which extent the invisible hand of the market can trigger investments in backup capacity and at which point a regulatory intervention is needed can be given in the course of the thesis. It only aims to pinpoint measures to improve the cost recovery and indicate their financial burden on the consumer. If a capacity mechanism does not only trigger investments in backup capacity but incites a system friendly behavior by all market participants, the security of supply can be maintained in the most cost-efficient way. In this sense, the MCDA highlights the concept of capacity subscriptions. By its self-rationing approach, it gives the consumers the option to decide whether they prefer to invest into backup capacity or contribute to the security of supply by reducing their consumption at peak times. The simplified calculation for the industrial consumers demonstrated its efficiency enhancement. As an alternative for improving the cost recovery, backup technologies can exhaust their dominant market position in scarcity times by including markups in their bids. As other market participants benefit from the increased prices as well, the measure with markups leads to higher costs of the energy supply for the consumers than the one with capacity subscriptions. Comparing the costs of energy supply in case of the EoM with and without a capacity subscription, the acceptance of scarcity prices leads to significantly higher costs for the consumer every year. Assuming that the scarcity prices would trigger investments in backup capacity, this would be at least the case during the lead time between the investment decision and commissioning of the backup technology. For the simplified calculation, the costs of energy supply are 12.9 times higher in the extreme case of scarcity and 5.2 times higher for a sensitivity of only 50 percent of the missing secured capacity. Overall, the calculations show the increases in efficiency resulting from capacity subscriptions and the reduction of the financial burden for the consumers. The design of a well-tailored regulatory intervention is a complex matter that carries the risk of maladjustments and false incentives. The design of capacity subscriptions, the extent of market power during scarcity time and the composition of a well-balanced flexibility mix which considers short-term flexibility potentials and long-term backup technologies is subject for further research. In conclusion, the thesis shows that suitable backup technologies are available and needed in an electricity system with a high share of renewables. At the same time, their cost recovery by the EoM is insufficient and is deteriorated in case of a mild weather year, a high level of battery storage and their qualification of forecasting the future dispatch accurately. The concept of capacity subscriptions is recommended as a measure to improve the cost recovery, decrease the risk and enhance efficiency.

Prosumers with photovoltaic systems can reduce their electricity expenses by increasing their consumption of self-generated electricity. This makes them more resilient to price shocks, like the 2022 European energy crisis. We evaluate how prosumers adapt their consumption behavior in response to such political uncertainty and increasing electricity prices. The collected survey and smart meter data allow us to evaluate the perceived self-reported and measured impact on self-consumption.Saving intentions due to the energy crisis are more clearly displayed by the survey than by the measured self-consumption. While solar radiation predominantly explains self-consumption changes, Google searches on electricity-related topics have limited explanatory power. However, considering time lags and the interaction with solar radiation leads to more nuanced insights on the effect of Google searches. Depending on the level of solar radiation, the effect of Google searches ranges from decreasing the daily self-consumption by 26.45 Wh to increasing it by 69.45 Wh. @en