|Design for a cogeneration plant in Belgium Credit: GE|
In a future where our energy increasingly comes from renewables, flexible production is one of the key assets for facing these challenges, but flexibility solutions all come with a cost both energetic and financial, writes Filip van den Borre
Over the past years, we have seen a large increase in electricity production based on renewable energy sources (RES) like solar photovoltaics (PV) and wind – and policy targets at the European and national levels aim for a further increase. For example, Europe set a target of 20% renewable energy production by 2020 and 27% by 2030. The final goal of these targets is to reduce, as much as possible, primary energy use from fossil sources.
Nobody questions the useful contributions of these renewable sources to lowering the emissions of CO2 and polluting substances (NOx, SO2, dust, etc). There are, however, some drawbacks that need tackling as well. A system with a large share of intermittent production demands a large amount of flexible downward and upward capacity. Besides demand side management and storage solutions, both still strongly in development, flexible production is one of the key assets for facing these challenges. It is important to note, however, that these flexibility solutions all come with a cost both energetic and financial. E.g., the electrical efficiency of a flexible operated CCGT is 5% to 25% lower depending on the operation mode. Furthermore, the investment cost for a flexible combined heat and power (CHP) system can be 15% to 20% higher with respect to a baseload CHP installation.
The future role of baseload CHP
One of the main questions for the CHP sector is about the role of baseload CHP. In the foreseeable future, we will certainly have moments where – due to a large increase in renewable capacities – renewable electricity production outgrows overall demand. At such moments, using fossil energy sources to fuel electricity production in a CHP installation equates to destruction of exergy. So what about baseload CHP?
As a thought experiment, we consider an industrial site with a baseload CHP which runs continuously, but cannot be modulated or stopped. Theoretically, this installation could deliver 8760 hours of primary energy savings per year. In a future scenario with larger shares of wind and sun, there will be times when there is enough renewable electricity to cover the demand. Continuing operation means primary energy will be destroyed instead of saved. However, as long as the total number of hours where demand is fully covered is limited, the net annual primary energy savings of this installation will be positive: the losses during these hours are smaller than the savings realized during the remainder of the year. Therefore, if one should choose between a pure baseload CHP and no CHP at all, and looking at overall primary energy savings alone, one should opt for the baseload CHP when the number of hours of overproduction is limited.
This is shown by the blue curve in Figure 1: the net primary energy savings will decrease as the number of hours in which renewable electricity is sufficient increases. In this example, the net annual energy savings will be become negative when the number of hours of overproduction increase above about 3800 hours.
This also means that, as the number of hours with sufficient renewable electricity production increases, the fixed (investment) costs per unit of primary energy saving increases as well: the fixed cost of the system remains the same, and if they realize fewer savings, the relative cost per unit of savings will be higher, as shown in Figure 1 by the red curve. So economics put a more stringent restriction on the maximum number of hours of ‘overproduction’.
Of course, this scenario is all too simplistic. First of all, from a system point of view, it is not possible for both intermittent sources and CHP to continue producing at times of overproduction. Secondly, it is of course possible to design a CHP plant in such a way that it can be shut down or modulated. But this flexible installation will have an additional cost with respect to a baseload CHP.
|Figure 1. The net primary energy savings in function of the number of hours with sufficient electricity is produced, and the relative investment costs per unit of primary energy savings|
|Figure 2. Potential for baseload CHP in Belgium for different scenarios|
If, on the one hand, we agree that CHP plants are useful for the system due to their primary energy savings and the system services they offer, and on the other hand want to keep the total amount of renewable electricity the same, we can pick between two options:
• We pay the additional cost for the flexibility of the CHP plants, and stop them at times when there is plenty of renewable electricity production; or
• We curtail renewable production at times of overproduction and invest in additional renewable production capacity to compensate for the loss of production.
Both scenarios realize primary energy savings by means of CHP and produce the same amount of renewable energy. However, both scenarios involve some costs. In the first scenario, there is the cost of making the cogeneration plant flexible. In the second scenario, there is the cost of installing additional intermittent production capacity. If both costs are compared with each other, the cost of the second scenario is lower than that of the first scenario when the number of hours of overproduction on an annual basis is limited. With increasing capacities of renewables, however, this option will become less attractive.
We calculated the theoretical capacity available for baseload CHP in Belgium based on different renewable energy scenarios from a study for the Flemish Energy Agency. The results are shown in Figure 2. The blue bars indicate the potential for pure baseload CHP that can run during 8760 hours without causing a single hour of overproduction. While there is still potential for a pure baseload in the current installed capacity and the reference scenario, this is not the case for the high RES scenarios.
If we assume baseload CHP is meaningful as long as it delivers a net primary energy saving (purple bars) there’s plenty room for baseload CHP in all scenarios. If we compare the cost of investing in flexible CHP plants with the investment in additional renewable (while keeping the primary energy savings and renewable energy saving at the same level) (red bars) then baseload CHP is interesting in all scenarios except the 2030 high scenario.
An important lesson is that the potential for baseload CHP based on an energetic or cost optimization is larger than the amount of pure baseload CHP that can be fitted within the residual demand curve. This implies that, although the potential for baseload CHP decreases with increasing intermittent energy production, it can be advisable to reduce the highest production peaks of intermittent RES installations during a limited number of hours a year.
|Figure 3. Theoretical, real and capped power production of a 2 MW wind turbine|
|Figure 4. Full load hours of overproduction and required flexible capacity in function of the growth of RES|
This should not be frowned upon, however, since the current sizing of wind turbines and PV systems is far from optimal. Better sizing of these plants would automatically decrease the demand for flexibility (and the related costs) whilst at the same time increasing the full load hours of the flexibility providers when flexibility is needed.
Optimizing the production profile of RES
We explain our reasoning below with a wind project, where the power of the generator is optimized for blade diameter, but a similar argument can be built for PV installations, where the inverter output is optimized for the peak power of the panels.
The electricity production of a wind turbine is – for a given wind speed – determined by the diameter of the blades and by the capacity of the generator. The diameter of the blades determines how much energy, at any given moment, can be captured theoretically by the wind turbine. This equals to the kinetic energy of the wind that passes through a circle formed by the blade tips of the turbine and increase with square of diameter of the blades.
The blades’ diameter determines the amount of energy that can be captured, but the power of the generator in the turbine determines how much of this energy is converted into electricity. Since the capacity of the generator is always smaller than the maximum theoretical capacity of the rotor, this determines the maximum capacity of the overall turbine.
A monotone of a typical production profile of a standard 2 MW wind turbine is shown in Figure 3. The theoretical power output of the wind turbine is shown in purple. It shows that peak velocities occur only during a limited number of hours per year. The real power output of a 2 MW wind turbine is shown in red. It shows the result of the trade-off between the additional electricity production with a larger generator on the one hand, and the additional cost of this larger generator on the other hand. In practice, this assessment is made by the equipment manufacturer offering a limited set of possible combinations.
This optimization should take into account several factors: first of all, there is the cost of the generator itself: a smaller generator entails lower production costs due to reduced material consumption (e.g., amount of copper). In addition, many other components will be less costly: thinner cabling, smaller transformers, … And also the connection cost is obviously lower, which means the same network infrastructure can connect more turbines. Optimizing from a system point of view would lead to much smaller generators.
The 2 MW wind turbine, which we used as example, achieves approximately 2000 full-load hours per year, which is seen as a standard for new installations in Flanders (maybe increasing towards 2300 hours for new installations). Yet the question is whether this is the optimal installation: if the generator power is limited to 1500 kW (a reduction of 25%), only 6% less renewable electricity is produced. A limitation to 1200 kW (a 40% reduction) means a loss of renewable power production of 13%.
For the same grid connection, two turbines of 1 MW could be installed instead of one turbine of 2 MW. Together, those two turbines produce 60% more renewable power. This equates to 3200 full load hours.
Another point to take into account is the loss of revenue from the electricity using a smaller generator. Since electricity is only lost during times of high wind power, and considering the price of electricity is relative low with a large share of renewable electricity, the market value of the lost electricity is relatively limited. In terms of income (whether in terms of economic value or of societal value), the loss is even smaller than the percentages mentioned above.
Taking into account the above factors, the optimum number of full load hours for on-shore wind projects in Flanders would probably amount to 3500 hours.
To reach a high share of renewable electricity, we need a sufficiently large installed capacity of renewable sources. In the high RES scenario used before, the total installed capacity is significantly higher than the peak demand. This means production frequently exceeds demand, and measures should be taken to keep the grid in balance: modulating CHP, shifting demand, storage, etc. All these technologies require additional investments. For their profitability, these depend heavily on the total number of hours per year they can be utilized effectively. Batteries may be used to store electricity when production exceeds demand, but if this only happens a few times a year, they will never pay for themselves.
To analyze this, we apply the principle of full load hours to overproduction: the total amount of overproduction over the year divided by the maximum overcapacity. Reaching higher full load hours means a greater utilization rate of the flexibility providers. This is strongly influenced by the production profile for the intermittent sources. Using installations with a flatter production profile has a positive influence on the full load hours of flexibility solutions.
With a RES growth of 0.1 (it is installing 10% of the additional capacity required to reach the high RES scenario), overproduction occurs during a limited number of hours per year with the current design, while there is still no overproduction when reducing the maximum power with 50%. Again, this is for the same total production of renewable electricity. Furthermore, the flexible capacity that is needed also decreases, so there’s less need for investment in batteries, DSM, power to gas, etc. or less renewable electricity to be curtailed.
If we consider renewable generation at higher RES shares, we see that this trend will continue: the required capacity for flexible solutions remains lower, but the number of full load hours of this capacity is greater. Flexibility solutions are therefore required later, and to a lesser extent, but when they are installed they will be more cost-effective due to a greater utilization.
It goes without saying that the lower cost for flexibility solutions need to be compared with the higher cost to place additional wind and PV installations. On the other hand, the loss of intermittent electricity production by reducing the maximum capacity of the current design is very limited. By halving the power, less than 20% of energy production is lost, so for the same installed capacity 60% more renewable electricity is being produced. And as mentioned, the renewable electricity that is given up has a lower market value than that of the remaining production because of the simultaneity of wind and PV production. In other words: a flatter production profile reduces overall installation costs and increases the average value of the produced electricity. In this respect, current production facilities can still be further optimized…
Filip van den Borre is Cogeneration Advisor at COGEN Flanders
This article is available on-line. Please visit www.decentralized-energy.com