EU Energy Policy Blog

The pump-storage technology allows the transformation of low-altitude water into high-altitude water using off-peak electricity, and then the production of electricity at peak periods releasing water through turbines like in any hydroelectric plant. Because of large energy losses in the transformation of electricity into water and then of water into electricity (the cycle efficiency is of the order of 80%), this process is not generically good at saving energy but it can be profitable on economic grounds, both by decreasing production costs and by increasing consumers’ surplus.

Pump storage is also an efficient contribution to energy saving when the installed non-hydro technologies cannot be adapted to instantaneous demand. This is the case of coal plants and nuclear plants for which stop-and-go strategies would be non profitable so that a share of their off-peak output is available for pumping water up at low cost. Similarly, solar and wind energies are not necessarily available when demand is high. Therefore, if controllable plants cannot be slowed down sufficiently when the wind is blowing and/or the sun is shining, the extra production, that is the difference between the sum of thermal output and intermittent output on one hand and electricity consumption on the other hand, can be used to pump water into upper reservoirs.

All increases in hydro capacity are highly valuable because hydroelectricity can be produced almost instantaneously. This flexibility makes it a pivotal contributor to system reliability (i.e., reactive energy, voltage control, operating reserves, back-up supply). The need to increase the availability of a flexible resource for peak periods explains the interest of governments for this technology and the momentum to develop it. As the European Commission writes in its 2007 Strategic Energy Technology Plan, “Equally interesting for Europe is the need for new renewable power capacities as embedded in the European targets for renewable energy by 2010 and 2020 and the correlated needs for back-up/firming capacities to ensure grid stability due to increasing penetration of stochastic power. Hydro technologies can make a significant contribution to this topic as a storage technology. A renewed and growing interest for pumped storage schemes has been accounted for in the last 5 years in Europe.” At present, about 30 to 35 GW of pumped hydro storage capacity is installed across the EU-27, to be added to 106 GW of hydro-power installed capacity.

Given the energy losses during the transformation cycle, when and for how much to install and dispatch pump storage mainly depend on three ingredients:

i) a low opportunity cost of off-peak energy. In particular, when demand is smaller than production and the latter cannot be economically or technically reduced, energy is available “for free”. This is the case with nuclear plants selling at night to hydro-producers. For example, according to the Swiss Federal Office of Energy, on September 28 2003 (date of the Italy blackout) at 3 a.m. “the import volume in Italy was 6,651 MW, while the domestic load was 27,702 MW (3,638 MW of which was intended for pump-fed power plants).” This means that in the middle of the night more than half of the imports were dedicated to water pumping using cheap electricity mainly from France. More commonly, pumping will cost the marginal value of the fuel used in the thermal plants (including the price of CO2 permits).

ii) a high opportunity value of peak energy. In most countries, electricity prices are capped so that the system equilibrium cannot be reached by rocketing prices when some energy shortage occurs. Nevertheless the caps are very high, theoretically reflecting the value of loss load. For example, on October 19 2009, day-ahead electricity has been priced 3000€/MWh on Powernext between 8:00 a.m. and noon. Spikes are uncommon, but because the public opinion, and therefore politicians are fiercely against blackouts, the dispatcher will for sure call hydroelectricity in the merit order for adjustment making the pumped water highly valuable.

iii) the investment costs. When the two former conditions are met, the operators of pumping stations can expect a high potential cash-flow. The remaining question is to know whether the present value of the cash-flow is larger than the installation cost. Clearly technical and economic performances are dependent on the site specifications and utility operating strategies. “Average load factors of large scale hydropower plants range from 2 200 to 6 200 full-load hours per year in Europe, with an average at about 3 000 to 3 500 hrs. Capital investment costs for building large hydropower facilities (> 250 MW) are of the order of 800 to 3 700 €/kW. Capital cost for hydro-pumped storage is of the same order of magnitude. (…) Average capital costs for small hydropower plants are of the order of 1 200 to 3 500 €/kW.” [European Commission, ibid.].

Given that a pump storage station can expect an average operating margin of 300€/kW a year (below no-pump hydro but above thermal electricity), the main limitation for developing this technology is the scarcity of sites. However, water storage can be promoted at small-scale. Spread small plants would be most welcomed by the operators of distribution networks who face the challenge of an increasing volume of embedded capacity of intermittent energy sources (e.g., wind mills) encouraged by feed-in tariffs.

Claude Crampes, Toulouse School of Economics (GREMAQ and IDEI) and Michel Moreaux, Toulouse School of Economics and IUF (IDEI and LERNA)

P.S. The authors develop a theoretical economic model of pump storage in “Pumped storage and cost saving” published in Energy Economics (2009), doi: 10.1016/j.eneco.2009.10.004, available on Science Direct.

This entry was posted on Friday, October 30th, 2009 at 1:37 pm and is filed under Electricity. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.