Long-Term Security of Supply and Strategic Policies in Liberalized Electricity Systems

LONG-TERM SECURITY OF SUPPLY AND STRATEGIC POLICIES IN LIBERALIZED ELECTRICITY SYSTEMS: CAPACITY PAYMENTS, CAPACITY MARKETS AND RENEWABLES PROMOTION MECHANISMS.

Álvaro López-Peña, Instituto de Investigación Tecnológica,

Phone +34915422800, E-mail:

Efraim Centeno, Instituto de Investigación Tecnológica

Julián Barquín, Instituto de Investigación Tecnológica

Overview

“Energy security of supply” is not a clearly defined concept, and the measures and policies that would have to be put in place to strengthen energy security are a source of confronted interests among the agents involved (governments, policy makers, companies and consumers, for instance). Anyhow, some consensus on the best ways to enhance security of supply is being reached, primary energy sources’ diversification being one of them. Focusing on liberalised electric systems, this is mainly related to the generation business. Staying within liberalised electricity systems, security of supply is as well conditioned by the presence of enough installed generation capacity.

From another perspective the electricity generation activity is as well affected by the growing environmental concerns, making governments wanting to limit fossil fuel-based technologies by promoting renewables or supporting nuclear generation.

However, because of liberalisation, governments and regulators concerned about these issues cannot oblige private companies to consider them when deciding new investments. Instead of that, the regulator has to influence the private companies’ decisions regarding generation expansion. He may want to influence these decisions from a double perspective. Firstly to address the adequacy problem: having enough installed capacity to cover estimated demand with a fair reliability level, i.e. “having enough megawatts”. Secondly, to make the generation portfolio evolve in the desired path (foreseenby indicative planning): constraining gas dependency, promoting local fuels or renewables, or supporting nuclear generation for example; i.e. “having the good megawatts”.

The aim of this study is to assess the efficacy and efficiency of some of the suitablemechanismsfor influencing companies’ investment decisions under liberalised frameworks, considering the above-mentioned double perspective. Due to this double perspective, regulators may need to combine policies, what may lead to undesired effects, which are identified as well.

The main policy instruments that are studied are the following. From the adequacy perspective, Capacity Payments and Capacity Markets. This choice allows us to see how introducing a quantity signal in the adequacy mechanisms influences reserve margins. From the indicative planning perspective, it is supposed that the aim of the regulator is promoting renewables, what is justified in the current trends worldwide. Among the renewables promotion schemes, for systems with a low renewables penetration, feed-in-tariffs seem to be the mosteffective mechanism. For systems with a large penetration, green certificates are the best solution, because of the efficient innovation, investment and operation signals they provide. This study focuses on a system similar to the Spanish one, where a considerable renewables penetration has already been achieved. Hence, an intermediate policyhas beenstudied: feed-in-premiums, which consist on a fixed per-MWh payment perceived by renewable generators in addition to energy market prices.

Methods

For carrying out this study, the electricity sector evolution in the long run has been be simulated by using a multidisciplinary System Dynamics-based model, described in(Sánchez et al., 2008). This model represents investment decisions in a disaggregated way, i.e. it considers explicitly the companies differentiation, as well as different investment technologies, comprising both thermal and renewables (no hydro investments are considered). For doing so, endogenous profitability calculation of all agents for each technology is modelled, using endogenously calculated discount rates. This is done following innovative Credit Risk Theory concepts.

In addition, for better differentiating among companies, the short term strategies’ differences are as well considered. A Game Theory approach has been developed for modelling companies’ behaviours in the spot and forward electricity markets. The first one is modelled though a conjectural variations-based equilibrium. In these kinds of models, the conjectural variations are normally exogenous parameters obtained from observed market results. This is valid in the short and medium terms, where the market structure can be considered quite constant. But in the long term, this hypothesis is not valid. Thus, a Supply Function Equilibria approach is used in our model for endogenously calculating these conjectural variations while the simulation advances and the market structure evolves. Finally, companies’ strategies in the forward market are represented through a Cournot-like equilibrium, considering as well its influence in the spot market’s strategies.

Both the capacity payments and the capacity markets are modelled as well. For the latest, the ideas in (Hobbs et al., 2005)have been used. Concerning the feed-in-premiums, three scenarios have been studied: null, low and high premium.

Results

Six cases have been executed: three with capacity payments (CP) and three with capacity markets (CM). Each of them with a different renewable premium: null (P0), 7.5 €/MWh (P7.5) and 15 €/MWh (P15). Each simulation is done for 25 years, discarding the firstfive (greatly influenced by initial conditions). A 3% constant demand growth is considered. Non-supplied energy (NSE) price is 180 €/MWh, the same value as the considered price cap. A per-technology emission rate and a carbon price of 25 €/tCO2 have been supposed.Main results are shown in the table.

First, the results with null premium will be described (CP_P0 and CM_P0 cases). They show that capacity markets are more effective in stabilising reserve margins (mean of reserve margins of 1.117 instead of 1.081, for equivalent payments), dampening boom and bust cycles and levelling electricity prices. More peaking units enter the system, as can be seen in the table: the share of Open Cycle Gas Turbines (OCGT) is much higher in the CM than in the CP case, compared to the Combined Cycle Gas Turbines (CCGT). Other technologies’shares remain similar in both cases: no new investments in wind capacity are observed (share equals 12% in both cases). Hence, electricity prices both in peak and off-peak hours are lower in the CM_P0 case, what makes total market costs for consumers (actualised with a 6% discount rate) to be lower. Capacity mechanisms’ costs are greater in the CM_P0 case, but total costs for consumers, remain lower.However, the total CO2emissions (over the complete study horizon) are greater in CM_P0, due to the greater emissions rate of OCGTs. If the regulator wants to address this problem, renewables promotion is needed.

Introducing renewables support does not make wind capacity profitable in any of the CP cases, thus making wind shares to remain equal to the CP_P0 case in both CP_P7.5 and CP_P15 scenarios (12%). Thus, the investments are the same that in CP_P0 case, as well as reserve margins, prices and emissions. A premium bigger than 15 €/MWh may change this. Costs grow within the CP cases because of the bigger premiums paid to the existing wind capacity.

Feed-in-premiums seem to be more effective under capacity markets: the share of wind capacity in the final year increases with the premium (28% and 40%), substituting mainly CCGTs in the CM_P7.5 case and OCGTs in the CM_P15 case. This causes total emissions to decrease. However, introducing renewables premiums seems to cause boom and bust cycles to appear again, what is reflected in the decreasing mean of reserve margins. This is becausepremiums are a fully price-based signal, which may ruin the stabilising effect of capacity markets (quantity-based) over reserve margins. This implies peak prices that grow with the value of premiums, as well as renewables support and NSE costs. Off-peak prices first decrease (because wind production is subtracted from demand), and then grow again when reserve margins are low enough. This same tendency is present on market costs. Capacity mechanisms’ costs decrease with the total installed capacity (proportional to the reserve margins).

Conclusions

For achieving enough installed capacity, it seems that capacity markets are more effective (higher and more stable reserve margins observed) and efficient (from the total costs for consumers perspective) than capacity payments. This is because a wider range of capacity prices can be attained, causing peaking units to enter the system if needed. Moreover, thisneed is expressed by capacity prices growing when reserves margin fall, creating a stabilising effect. Concerning renewables support, as these technologies have relatively high fixed costs, more investment is observed under capacity markets for the same premium value.From our results, it seems that capacity markets are more desirable for obtaining enough and good megawatts, because more investment is obtained in peaking units, causing reserve margins and prices to be stabilised in acceptable levels. A slight renewables promotion suffices to transfer some of this investment to wind turbines.

Given that feed-in-premiums are a fully price-based mechanism, if set too high, they may ruin the stabilising effect that capacity markets (quantity-based) create upon reserve margins. This is the main contribution of our study, and it suggests the need for careful combination of policies aiming at addressing the adequacy problem with policies for implementing indicative planning.

References

Hobbs, B. F., J. Inon, M.-C. Hu and S. Stoft (2005). Capacity Markets: Review and a Dynamic Assessment of Demand-Curve Approaches. IEEE Power Engineering Society General Meeting. San Francisco, California USA.

Sánchez, J. J., J. Barquín, E. Centeno and A. López-Peña (2008). A Multidisciplinary Approach to Model Long-Term Investments in Electricity Generation: Combining System Dynamics, Credit Risk Theory and Game Theory. IEEE General Meeting. Pittsburgh (Pennsylvania), USA.