Assessing the flexibility of coal-fired power plants for the integration of renewable energy in Germany
More than 80% of a total of nearly 24.2 GW of installed coal-fired power plant in Germany (end-2018) were built before wind and solar power made any meaningful inroads to the Germany power market. Their initial design was thus optimized for baseload and mid-load provision rather than cycling operations, now generally raising questions about the ability of coal-fired power plants to provide the necessary flexibility. The shares of wind and solar power can be expected to continue growing. Not only are the decreasing costs for these technologies gradually tilting the economic calculation in their favour, but the German government (alongside other European nations) has also set itself ambitious targets to further expand deployment of renewable energies.
This study has two primary objectives: first, assess how the need for power system flexibility grows in Germany as wind and solar power are further expanded. Second, assess whether and how the existing coal-fired power plant fleet in Germany can accommodate and integrate growing shares of variable renewable energies, without jeopardizing reliability of electricity supply. Specifically, we are studying whether renewable energy shares of 50%, 60% or 70% (as compared to the 38% reached in 2018) would alter the way coal plants in Germany are operated and whether their technical characteristics are compatible with a further increase in wind and solar power.
The main technical characteristics that determine the flexibility of a thermal power plant are start-up duration and cost, minimum load level, ramping speed, minimum operation time and minimum down time. We collect data on all of these parameters through literature review and expert consultations, discuss and feed those into our in-house power system model DEEM. DEEM is a mixed-integer, linear optimization model of the European power system that dispatches power plants on an hourly basis, explicitly taking into account the above-mentioned operational constraints.
Taking stock: model runs, carried out for the years 2015 and 2018 demonstrate that in these two years coal-fired power plants have been instrumental for integrating fluctuating output from variable sources into the grid. The years were chosen to obtain a maximum difference in renewable energy share (32% in 2015 vs. 38% in 2018) while still having a complete set of power market data available for model validation (data prior to 2015 is incomplete, notably regarding hourly infeed of wind and solar power).
Looking ahead: model runs, carried out for renewable energy shares of 50%, 60% and 70% suggest that the existing dispatchable fleet in Germany, including the installed coal plants, pose, from the point of view of flexibility, no barrier to further expansion of variable renewables. The flexibility metrics developed in this study indicate that coal plants ramp more often and cycle more intensely as the share of renewables increases. (Technically speaking, the plants change their operational status – offline, online, minimum load, full load etc. – more often). Hence, we find that coal-fired plants contribute to the provision of “short-term” flexibility by adjusting their energy contribution and by being dispatched more flexibly and at moments when they are most valuable for the system.
Moreover, by analyzing the contribution of coal plants during “dark cold doldrums”, (i.e. cold spells coinciding with meteorological conditions that result in limited output from wind and solar plants) we find that their role for providing “mid-term” flexibility, becomes more important with increasing shares of renewables. By analyzing such periods of tightness with durations of one to three days, we find that during such events, coal-fired power plants generate twice as much power than on an average day if the renewable energy share is 50%, and three-and-a-half times more power if the renewable energy share is 70%. Another finding is that the weather conditions leading to “dark cold doldrums” also affect Germany’s neighbouring countries to a degree: although Germany becomes a net-importer during such events, the scope for balancing via higher imports is constrained by the availability of dispatchable plants in other countries and congestion of interconnectors.
Coal-fired power plants contribute, as other dispatchable plants, to system security via flexible adjustments of their output and through the provision of firm capacity. Nevertheless, output from coal-fired power plant drops as the share of renewables increases but remains, with 45 TWh in a 70%-renewables scenario (compared to 72 TWh today), significant. The average load factor of the fleet drops to just over 30% in the 50% renewables scenario (in comparison, the average load factor stood at 35% in 2018) and further to around 20% and 15% in the 60% and 70% renewables scenarios respectively.
However, the coal fleet is not homogenous; neither in its age structure nor in its technical characteristics. The various plant types thus react differently to increasing shares of renewables. Unsurprisingly, modern and thus more flexible plants, adapt more easily to the changing market conditions. The latest designs achieve load factors far above the fleet average. In contrast, some of the oldest plants hardly run at all, being dispatched only in the tightest hours of the year. In our modelling framework such plants contribute to the system stability and adequacy but whether they could be profitably operated based on energy-only market revenues is questionable. This problem is not unique to old coal plants but also affects gas-fired power plants. It forms the heart of a debate around what market designs can safeguard the integration of variable renewables without jeopardizing the economic viability of the dispatchable fleet.
Nearly three quarters of the installed coal fleet in Germany produce heat and electricity at the same time, which is both a challenge and an opportunity for flexibility provision. Most of the Combined Heat and Power plants are so-called extraction turbines which can switch between heat and electricity flexibly and seamlessly (i.e. minimizing losses during times of high renewables infeed) as long as electricity output is not constrained by heat demand during cold weather periods. In either case, thermal storage retrofits can improve the operational flexibility of coal plants.