1 Intelligent Well Technology: Status and Opportunities for Developing Marginal Reserves SPE
EFFECT OF WIND ENERGY ON CAPACITY PAYMENT. THE CASE OF SPAIN.
Julio Usaola, Universidad Carlos III de Madrid, +34916249404,
J. Rivier, Iberdrola Renovables, Iberdrola Renovables, 34 91 577 65 65 ,
M.Á. Moreno, Universidad Carlos III de Madrid, +3491249991,
M. Bueno, Universidad Carlos III de Madrid, +34916248853,
Gonzalo Sáenz de Miera, Iberdrola +34 915 776 500,
Overview
The effect on electricity costs that a great penetration of wind energy would produce on a system has been the object of many studies. Apart from the operational costs (operational reserves), the main component of these effect (apart from the needed grid reinforcements) is related to the impact of wind energy on capacity costs, or how to pay generators that must exist because of the low capacity credit of wind energy, but that has less capacity factor due to the production of wind energy.
In this paper we argue that in a simplified market with a system of marginal prices, uder equilibrium conditions, the capacity costs of a system with wind energy are equal to those of a system with the same demand level and installed power. The reason of this conclusion is that capacity costs are the fixed costs of the most expensive units in the system (expressed as a price per MWh), and, as they must be received by all the generation to guarantee the cost recovery, if the installed power of the system is not changed, the capacity costs do not change, either, for the consumer. The electricity prices are also the same for both scenarios.
This conclusion is different from most of other studies because in them, the need of changing the generation mix was not considered. Under this assumption (same generation mix in both situations) there would be an increase in the capacity costs to be paid by the consumers.
The paper develops the arguments that lead to the aforementioned conclusion, and gives information about the Spanish electricity market, where a great amount of wind energy has been already installed, and more is still planned. Comparisons with other studies will be also included and commented.
Methods
The approach followed through this paper is based on the classical analysis of the optimal generation mix using the load (or thermal demand) duration curve, as presented in [1] , for instance, which has been applied to systems with wind energy in [4] . Our approach is made assuming very simplified conditions, that, however, allow us to draw interesting conclusions. These assumptions are that the generators receive the marginal price, that there are only two kinds of generation (apart from wind), both thermal: baseload units and peak units, and that the market is in equilibrium. This equilibrium is defined as a situation where the installed capacity, as well as the load factor of each plant, is optimal, and where there is a full cost recovery by all that generation. Availability of power plants has not be considered, but this does not change the conclusion significatively. Demand is considered inflexible, although this last point is discussed along the paper.
Costs of each plant are modelled by their screening curves, that have the following expression
ACC = FC + VC(€/MWh)
Where ACC is the Average Capacity Costs, FC are the fixed costs, is the capacity factor of the plant (0 < < 1) and VC are the variable costs of the plant. It can be seen that these costs are expresed in price per unit of energy, following the methodology of [1] . Then, ACC shows the average ost of using the plant’s capacity.
In this classical analysis, the different capacity costs of the units in the system, together with the load duration curve, give the optimal installed power, the load, or capacity factor of each technology, and the level of the prices of electricity throughout the year. This prices are set by the variable costs of the most expensive unit present in the market to cover the demand. Then, the fixed costs of the units are recovered during the times where prices are set by a more expensive unit. This means that there must be price spikes (or demand response prices) that will compensate the fixed costs of the most units with higher variable costs (peak units).
Anotherpossibility is to pay all generators (except the wind generators) an amount equal to the extra revenue that they would have during the price spikes in the equilibrium. To guarantee the cost recovery, this extra payment should be equal to the fixed costs of the peak units, FCP (€/MWh). This amount must be paid to all the generators (except the wind generators), since all of them would receive the revenues during price spikes.
The study consists in comparing the capacity costs in two situations: with and without wind energy, with the same kinds of conventional generation and the same demand level (or load duration curve).
Results
As explained before, in an equilibrium situation where no wind power has been installed, the cost recovery may be guaranteed by paying the fixed costs of the peak units, FCP(€/MWh), to all the units present in the system. If wind energy is added to this system, then another equilibrium point should be reached, where:
- The total conventional power installed does not change. This means to assume, in a conservative way, that wind does not have capacity credit. See [2] for a discussion on this subject.
- That the installed power of peak units increases, whereas the installed power of baseload units decreases. In this paper this conclusion will be explained, but a detailed explanation may be found in [4]
- That the electricity prices, in the equilibrium, do not change in comparison with the situation without wind energy. That is, the wind energy does not affect, in the long run, the price levels.
The capacity payment, when wind energy is included, must be also equal to the fixed costs of wind power, FCP(€/MWh), and it must be given to all the installed power in the system. Since the costs remain the same, as well as the total installed power, it can be deduced that no additional capacity costs should be paid to the consumers, even under large amounts of wind penetration, and considering null the capacity credit of wind power.
This conclusionis different from those, for instance, found in [2] or [3] . In this last work, an estimate of the capacity cost for the market in Great Britain with a 20% of wind penetration yields costs of 4.81 GBP/MWh of additional costs of capacity. If we compare this to the marginal cost of the representative average CCGT used in the study, 13 GBP/MWh, we can conclude that the increase is very high. This could be because there are no changes in the conventional generation mix between the ‘no wind’ and ‘wind’ scenarios in these studies.
In order to briefly describe the Spanish situation, we can recall that the Spanish penetration of wind energy for the year 2008 were the 11% of the energy produced. Capacity costs in Spain were the 7.62% of the final price of electricity, and a 9.2% of the daily wholesale market price in 2007. Due to regulatory changes, capacity payments were only the 1.14% of the average wholesale market price. Installed wind power in the Spanish peninsular system 2008 was 15576 MW. A longer description of the Spanish peninsular electrical system will be given in the paper.
Conclusions
In a simplifyied electricity market, the equilibrium is reached when every generator cover their costs. When electricity prices are set in a marginalistic market, the units with highest variable costs (peak units) should recover their costs in occasional price spikes, or if they are paid an additional amount, called here capacity payment.
The effect of wind power on an electricity market in equilibrium is to change the optimal generation mix to a situation with more peak power and less baseload power. Under the simplifying conditions of the study, and if no capacity credit is given to the wind energy, there would be no effects on the wholesale prices. It can also be drawn that the capacity costs would not suffer any change.
This conclusion is different from those of other studies, but this is due to the fact that most of them did not take into account the need of having a different generation mix when wind energy is present in the system.
References
[1] Stoft S. Power System Economics Wiley Interscience. New York, 2002.
[2] Gross, R., Heptonstall, P., Anderson, D., Green, T., Leach, M., Skea, J. The Costs and Impacts of Intermittency. UK Energy Research Centre, London, 2006.
[3] Skea J., Anderson D., Green T., Gross R., Heptonstall P. Leach M. “Intermittent renewable generation and the cost of maintaining power system reliability.” IET Ger. Transm. Distrib. Vol 2 (1) (2008). pp. 82-89.
[4] Sáenz de Miera G., del Río González P., Vizcaíno I. “Analysing the impact of renewable electricity support schemes on power prices: The case of wind electricity in Spain” Energy Policy. Vol. 36 (2008) pp. 3345-3359.