SECUNDARIA DE INSTITUTOS CON JORNADA COMPACTADA
21. LA POLÍTICA DE PROVISIÓN DE PROFESIONALES EN LOS CENTROS
Another approach to investigating storage value is pre-defining a function of storage and subsequently evaluating its benefits. In those cases, arbitrage revenues are used as a measure of storage profitability; however, profit maximisation is not its primary function. A common example of this is storage paired with wind generation. A large number of studies have looked at arbitrage value storage can achieve when paired with wind generation.
Within a whole system’s model (EnergyPLAN as the modelling tool), Lund & Salgi (2009) investigate the value of CAES under high wind penetration in Denmark. Primarily, the objective is not to maximise arbitrage revenues but rather to minimise wind curtailment and fossil fuel consumption. Additionally, the CAES system seeks to achieve load levelling and help integrate wind energy under one of the approaches undertaken. The arbitrage revenues derived fell far below the cost of the CAES system; however, when operated as a revenue maximising function and with the addition of reserve revenues, the storage system was found to be profitable. These findings are particularly important to the
approach taken in this thesis; pre-defining an application of storage is actually sub-optimal for revenue maximisation and therefore in order to fully investigate the economic feasibility of storage, revenue maximisation has to be the primary objective of the system.
Nyamdash et al., (2010) investigate the use of storage with wind power in the Irish system; storage charges from wind power and discharges into the Single Electricity Market. Three scenarios of low, medium and high wind penetration are explored, however, none of the technologies explored were feasible based on the arbitrage revenues derived. Also using Ireland as a case study, Foley & Díaz Lobera (2013) showed a potentially negative impact of storage; in principle under a gross pool market, storage systems can charge on cheap electricity and discharge at peak time, hence reducing the high Market Clearing Prices (MCP), on average. However, they found that CAES actually increases the average MCP; this occurs due to efficiency losses of the storage system. As a result, revenues for most generators and including CAES were shown to increase.
Using a similar approach, Grünewald et al., (2011) looked at arbitrage revenues storage can generate in GB, under a high penetration scenario using merit order of generation dispatch to calculate prices. The role of storage was to reduce wind curtailment by balancing the energy system. Arbitrage benefits were also evaluated to show that at levels of renewable energy penetration greater than 30 GW, storage technologies become economically viable. On a smaller scale, wind sited storage is explored by Hessami & Bowly (2011) who couple storage and wind farms in Portland, Australia. Arbitrage revenues can be generated from wind power and the authors show that annually, 8%-15% of return on capital investment can be expected from such a configuration.
Denholm & Sioshansi (2009) compare the case where CAES operates as a load sited system to the case where the system is sited with a wind farm, and hence save on transmission investment and reduce wind curtailment. To the wind farm owner, the benefits of choosing storage over transmission capacity upgrade become more apparent as transmission costs increase. Taking into account the intermittent nature of wind power, the authors argue, there is a danger of oversizing and thus underutilized transmission capacity. The study looks at wind and storage in two configurations: firstly, wind and storage being decoupled and operated independently as 2 separate entities and secondly, wind and storage being co-sited but operated in such a way as to maximise benefits. Under the latter configuration, transmission investment is reduced, however at the expense of storage operation flexibility. In other words, greater arbitrage opportunities are available to CAES operating as a load sited system with no restriction rather than wind sited operation constrained by transmission capacity and wind farm output. The fact that coupling storage with wind farms reduces value was also shown by Lund & Salgi (2009); They evaluated the benefits of CAES under very high wind penetration (59% and more); these benefits consist of the variable and fuel cost savings with CAES over excess electricity and higher operational costs without CAES. The results show, despite different fuel costs assumptions,
that the values are far below the annual investment cost. Even in the most optimistic case, the value is less than a third of investment costs. However, when they use CAES as a business oriented case, co- optimising revenues from spot market arbitrage and reserves, they show that the system is profitable. Fertig & Apt (2011) use a similar approach to investigate CAES co-located with wind farms. Four configurations are explored: In the first configuration, the system operates on hourly energy prices and determines when the wind farm is injecting power into the grid, to the CAES system or spilling energy, based on price thresholds. The second configuration looks at a capped maximum price with a capacity payment. The third configuration the Wind-CAES system operates at a fixed contract price equivalent to the annual average price to simulate a baseload generator operating at 80% capacity factor. Finally, the Wind-CAES system’s operation in the regulation market is investigated separately. These configurations respectively generated $900 million, £300 million, £110 million and £100 million. Comparatively, the standalone wind farm generated $245 million profit.
Besides arbitrage in the wholesale market and the provision of reserves, storage value can be calculated in terms of penalty avoidance for not meeting scheduled power for a wind farm. Turker et al., (2013) consider the direct balancing cost of forecast errors from wind power and the extent to which storage can mitigate these costs. They cite the model of the Spanish market whereby penalties are imposed at 10% of the current market price. The chosen storage technology is VRB which the authors find to be economically unfeasible; however, if the penalty multiplier was increased by 1.5 times the market price, the storage system would have a payback of 10 years, beyond which the investment would have recovered its total costs and generated profits.
A more comprehensive study on the potential of storage value, under different future energy pathways, was undertaken by Strbac et al., (2012). Using a whole system approach in GB, whereby the objective function is to minimise costs of generating electricity, the authors calculate the transmission and distribution investment deferral, generation investment foregone and operational savings arising from generators operating within their optimal range. This system approach is particularly useful to investigate the potential value of storage which is not realisable under current market conditions, as shown earlier in figure 2.1. The study, by choosing optimal capacities for storage, interconnector capacity and demand side response shows that there is a distinct need for energy storage within the future GB energy requirements, even in the presence of alternatives.
In order to understand the market value of storage under a high wind penetration scenario in GB, an understanding of the impact of wind generation on the GB markets is required. Green & Vasilakos, (2010) explored the impact of a high wind penetration scenario on the wholesale market in Great Britain. Using weather data, a 30 GW wind penetration scenario is simulated using the supply cost function for each type of generation to effectively create a merit order curve. The authors show that
under a high wind penetration scenario, the price volatility increases. They also investigate the competitive behaviour of generators under such a scenario; initially they used 6 symmetric firms to represent the supply function and later this is reduced to 2 symmetric firms. Under the latter, prices were higher and therefore, while the cost of wind generation is generally lower than mid-merit and peaking plants, prices may rise as competition falls. More recently, Cleary et al., (2016) have used PLEXOS modelling software to simulate high wind penetration scenarios for the UK. Contrasting a 14 GW and 25 GW wind penetration levels, the authors show that the System Marginal Price is lower as wind increases. This is further shown as wind energy imports from Ireland is allowed, depressing prices even more.
As shown from these previous studies, there is an overwhelming focus on energy storage value when confined to the support of wind generation. They differ from those mentioned in the previous sections as they have a primary function different from revenue generation and subsequently, their profitability is evaluated with respect to this function. Such a situation was shown to be sub-optimal for revenue maximisation, however, there is a clear application for storage. In other words, the approach to storage operation can be viewed as either one that is completely defined by market revenues or one that is defined by a specific application.