Why CSP’s thermal storage is safe from competition… for now

Booming interest in energy storage raises fears over whether CSP’s base-load capabilities could be undercut by PV with batteries. But there are reasons to rest easy.

Gemasolar in Spain is the first CSP project to provide electricity for 24 hours

By Jason Deign

Is one of CSP’s biggest selling points at risk from an emerging industry? That is the question looming large amid growing interest in technologies such as traditional and flow batteries, flywheels and compressed air energy storage.

The market for these and other energy storage technologies is booming. According to a recent report, 91 new grid-scale energy storage projects, totalling almost 363MW, were launched or announced in 2013 and 2014.

Meanwhile, Citigroup believes PV and storage combinations will achieve ‘socket parity’ by 2020 in some places, based on a halving of battery costs, to USD$230 per kilowatt-hour, within five to seven years.

Such trends naturally lead to concerns over whether CSP with thermal energy storage (TES) will still offer value to grid operators a few years from now. Currently, TES is a major selling point for CSP, justifying the higher cost of the technology compared to PV.

This is evident from the way storage has taken off as an essential adjunct to modern CSP plants.

Whereas storage was still more or less experimental in CSP half a decade ago, today a total of 82 projects, with a cumulative 524.5 hours of TES, are announced or in development, planning, commissioning, construction or operation, according to data from CSP Today’s Global Tracker.

Molten salt dominates the CSP TES market, being used in 60% of projects and accounting for a total of 414 potential hours of storage. In a further 21% of projects the storage medium has not been specified yet.

Abengoa's Solana project in the USA has six hours of molten salt storage stored in twelve insulated tanks

 

Alternative energy sources

However, the only other TES medium seeing significant use at present is steam, which has been specified in 15% of projects totalling around 12 hours of storage. Elsewhere, three projects use ceramics and one uses pebble beds.

Could all this momentum be threatened by the advent of other types of energy storage, though? The answer depends on which alternative energy source you are looking at.

According to Bloomberg New Energy Finance (BNEF), using batteries with wind generators is already better than TES with solar thermal electric generation (STEG) in terms of levelised cost of energy (LCOE).

The analyst firm has calculated the cost of shifting half the output of wind and PV by four hours, using lithium-ion (Li-ion) batteries at current US prices. Wind comes out with a LCOE of $173 per megawatt-hour (MWh).

“Wind with battery storage could therefore be considered competitive with STEG and storage without subsidies, which currently lies at $263/MWh,” says Samira Rüdig-Sotomayor, solar insight analyst at Bloomberg New Energy Finance.

Furthermore, battery storage is one of the most expensive forms of energy storage that can be paired with wind power.

In Spain, for instance, although there is one of the most important CSP markets in the world the real renewable energy success story, from a cost point of view, has been a high penetration of wind power coupled with cheap pumped hydro storage.

PV and batteries

Given the high direct normal irradiance (DNI) zones needed for CSP, however, most project developers might be more concerned about a potential threat from PV with batteries. Here the competitive situation is less clear-cut.

Bloomberg calculates the LCOE for PV-and-battery combinations is $311/MWh, so “PV and batteries would require a drop in cap-ex to become competitive with STEG and storage,” says Rüdig-Sotomayor.

“As batteries will drop significantly in price over the next decade, we expect this race to become closer.”

But Dr Luis Crespo, president of the European Solar Thermal Electricity Association, believes there are three reasons why CSP and TES could remain in the lead for some time, at least for grid-scale generation.

The first is that PV and battery combinations do not just incur higher capital costs because of the price of the batteries, but also because an oversized solar field is required to store energy. You need one and a half times more solar panels per project, Crespo says.

A second reason is efficiency. While TES is highly efficient, batteries lose about a third of the energy involved in each charge and discharge cycle. This also means massive battery banks are needed for large-scale storage, while TES can scale more easily.

The final point is lifespan: while it is likely batteries may need to be swapped out several times during the lifetime of a solar plant, “molten salt could last decades,” says Crespo.

Not a clear competitor?

As a result, he concludes, the threat “is not obvious in the medium term. And it is not even clear if this will be a competitor.”

Others agree. Right now, says Daniel Schwab, business development consultant at BrightSource Energy and director at Kayema Energy Solutions in South Africa: “TES is more than 90% cheaper than Lio-ion batteries, is bankable and is still getting cheaper.”

In fact, Schwab sees CSP with TES as being complementary to other renewable energy sources, with its storage capabilities helping to aid the integration of wind and PV.

“The importance for CSP to have storage is critical since it both helps reduce the LCOE and offers grid operators significant value to help integrate PV and Wind into the grid,” he says. “CSP with molten salt is the best battery storage solution for PV and wind.”

Another expert who believes CSP could complement other energy sources is Stephen Tordoff, a former Falck Renewables manager who now runs a consultancy called Energy Canvass.

Given the choice between developing CSP or PV in a high-DNI location, he says: “I would aim to do both. I would aim to probably build the CSP first then I would look at slowly ramping up generation on a PV plant. That way you have much greater confidence in your resource.”

Thus it appears that for the time being the CSP industry need not worry unduly about PV crashing its energy storage party with a battery pack, provided the DNI is good enough for solar thermal to be competitive in the first place.