By Jason Deign 

The essential attraction of CSP with storage is widely understood. The storage allows power to be dispatched whenever needed. So CSP can potentially cover base-load requirements, in the same way as coal or nuclear, or provide load-following capacity, similar to hydro.

Such capability is pretty rare among non-carbon, non-nuclear energy sources. That makes CSP potentially valuable to utilities, particularly in desert areas where alternatives such as hydro might not be available.

Hence why off-takers have been willing to pay a premium for CSP power-purchase agreements, even when they could get cheaper energy from intermittent renewable generation sources such as wind or PV.

But is this the best way to get a return on the costly investment required for CSP? After all, base-load is usually the cheapest form of energy, so unless feed-in tariffs are available it would take a long time to pay off a CSP plant with this kind of supply.

Load-following power has higher value, and it is significant that most CSP plants with storage that have been built so far appear to have been designed with this in mind. But there is another area that CSP could compete in: peak demand.

Most markets meet this demand using gas peaker plants. They are much cheaper and easier to build than CSP plants, and “there’s no way on a first-cost basis you can compete with that,” says Dr Thomas Mancini, principal at TRMancini Solar Consulting.

But gas peakers have the drawback of relying on a fossil fuel that makes daily operating costs relatively high and subject to wide variations.

Mancini says current operating costs for simple-cycle gas plants are in the region of USD$0.25 to $0.35 per kilowatt-hour (kWh), whereas operating costs for the current generation of US CSP with storage is around $0.15 per kWh.

Peak demand

And while gas will undoubtedly remain the favoured option for handling peak demand in most markets for the foreseeable future, there are conceivably some places where CSP might have a window of opportunity to compete. Take California, for example.

While the wide availability of shale gas makes traditional peakers a clear favourite for covering peak demand across the rest of the US, California’s high decarbonisation targets, along with growing incentives for energy storage, might give CSP an edge.

In addition, California has to juggle major daytime energy inputs from PV. “What California is struggling with right now is the duck curve,” says Ken Ditzel, a managing director at FTI Consulting. “PV is generating more than enough power supply during the day.

“But the issue is at night and in the mornings, when demand is still high. They need this ramping capability to serve demand.”

This morning and evening demand means the California energy system is notable for regular pricing peaks that send energy prices soaring way above the lower limit of the levelised cost of electricity (LCOE) for CSP.

According to the CSP Today 2014 Markets report, CSP LCOEs now fall within the range of $90 to $170 per megawatt-hour:

Source: CSP Today Markets Report, 2014

While the average price of Californian electricity is around $40 per MWh, which is far below the profitability threshold for CSP, analysis of online pricing data from the California Independent System Operator shows almost daily peaks above $150.

These pricing peaks typically last less than half an hour but rise to between $200 and $300 per MWh. And the upticks in demand are interspersed with regular mega-peaks where electricity prices soar into the $1,000 per MWh range.

In the last quarter of 2014, for example, energy prices jumped well above $500 per MWh on at least two dozen occasions. The highest peak last year saw the price of energy topping $4,000 per MWh. That is a lot of cash.

And to take advantage of it, CSP plants do not need a great deal of storage, since the peak pricing intervals rarely add up to more than an hour in a 24-hour cycle.

Good forecasting

However, because it can take a steam-cycle CSP turbine up to half an hour to start up, a plant operator would either need to have either extremely good forecasting data or to keep the plant idling during the period when peaks usually occur, for example between 7pm and 9pm.

Other CSP markets do not boast quite as dramatic pricing highs, but nevertheless may offer opportunities for peak supply. In Chile, for example, a lack of native fossil fuel supplies means power prices are perennially high. The place is also awash with salt for CSP storage.

And while the requirement from mining operations in the north is more for round-the-clock, base-load-style generation, Chile’s stable, growing economy could lead to the kinds of daily demand peaks seen in richer nations. The problem is that right now it is difficult to find out.

“As far as I know, the Chile power prices are not public,” says KTH Royal Institute of Technology solar thermal power R&D engineer Rafael Guédez, who is developing a CSP modelling tool called DYESOPT with European funding from KIC InnoEnergy.

“The average spot prices are high, but I don’t know about the peaks,” he says. “The spot market is also new, so it’s difficult to know how it will behave. But I think it would be interesting for CSP.”

Over in South Africa, another major CSP market, policymakers have already pre-empted the need to cater for peak demand by creating a premium evening rate within the latest Renewable Energy Independent Power Producer Procurement Programme.

Guédez says that if you model a CSP plant to take full advantage of the rate then what you get is essentially a peaker.

Ilaria Besozzi, business development manager at the Swiss CSP developer Airlight Energy, says: “In the last allocation, what they tried to do with their feed-in tariff was factor in the peak to give more remuneration to the storage component. That is where I see CSP could find its value.

“I think in some countries this could already be competitive with fossil generation.”