Dispatch flexibility can triple CSP revenues
DNV GL is exploring how a concentrating solar plant with thermal storage (CSP-TES) can dramatically increase its revenues through ancillary services provision. The key is to find the optimal design, by varying turbine and storage configurations, depending on which markets will be accessed. With the opening of the Ivanpah solar power plant last week, the conversation around profitability of CSP plants, and the true costs and benefits of renewable energy, has been re-kindled (with green energy, presumably). Even as the dedication ceremony for the grand project concludes, the future for large-scale solar thermal in California is uncertain. The main culprit seems to be competition from photovoltaic (PV), which has seen a sharp decline in capital costs over the past several years. In a nutshell, if you are looking for green energy from the sun, PV is cheaper than solar thermal. But is this a fair comparison? The DNV GL study shows that revenue for the plant can more than triple if the plant is configured correctly for the ancillary markets. In addition, the grid system as a whole will benefit if CSP-TES plants focus on delivering flexibility instead of just green energy.
The increasing levels of renewable energy in California stems from an ambitious goal of reaching 33% renewable energy by 2020, and perhaps beyond in 2030 to 2050. As levels increase towards this goal, the California Independent System Operator (CAISO) is tasked with balancing a grid with more and more renewable resources on-line. This is not a trivial task, as most renewable resources are following the whim of the sun or the wind, not an hourly day-ahead schedule or an automatic dispatch signal designed to keep frequency on the grid in check. In other words, the energy market does not only buy and sell energy, it also needs to procure system flexibility, such as regulation and ramping capacity, and back-up energy, such as spinning reserves, in order to keep generation and demand in balance every second, every day. As levels of intermittent renewable energy increase, system flexibility is harder to come by and, hence, will become more expensive.
Enter concentrating solar thermal power. An important difference between this technology and PV or wind is the way electricity is produced: heat from the sun is turning water into steam, in turn driving a conventional steam turbine. The use of thermal energy and rotating machinery makes operations very similar to conventional power plants, with the difference that steam is heated with energy from the sun instead of fossil fuels. Thermal energy also has the advantage that it can be stored for longer periods of time at reasonable cost in tanks of molten salt, for instance, instead of batteries. Can these CSP-TES plants take on the role of a conventional generator, following a day-ahead schedule, providing ancillary services and reserves? What are the production cost and grid control implications for the California grid? And what are the revenue and design implications for the plant operator?
To understand these effects and choose design and operations to maximize revenue and system benefits, a holistic modeling approach is necessary, ranging from grid and plant dynamics on an intra-hour time scale to production cost modeling for future generation scenarios. Until recently, this analysis has been lacking. In an on-going study for the California Energy Commission, DNV GL is quantifying these benefits. And there may be good news for Ivanpah and the future of CSP-TES in California:
- Revenue from the day-ahead energy market for a CSP plant can be increased with 11% when 6h of storage is added
- If instead the turbine capacity is oversized relative to the solar field, revenue can increase up to 27% with only 2h of storage.
This surprising design choice, oversizing the turbine relative to the solar field, is due to the short duration of high energy prices in the late evening in California. Should the CSP-TES plant operator participate in ancillary markets, such as regulation and spinning reserves, this effect becomes even more pronounced. While some storage is necessary for participating in regulation and spinning reserves, 2-4 hours is generally enough. Instead, it is the extra turbine capacity that brings in the big bucks:
- Annual revenue can increase up to 4 times for a CSP-TES plant with minimal storage, but with a turbine oversized 4 times compared with the solar field, when the plant participates in day-ahead energy, regulation and spinning reserves markets.
Again, this is linked to the increasing cost of system flexibility. In other words, it is a better strategy to provide flexibility in the energy market, than to provide energy. Flexibility for the CSP-TES plant translates to a large turbine and modest amounts of storage. Given that the solar field, those thousands of mirrors and the solar tower collecting solar energy, is the most costly component of the CSP-TES configuration, adding 4 hours of storage and doubling the size of your steam turbine, is likely a prudent investment.
In addition, the California system as a whole will benefit from this dispatchable, flexible, and renewable generation fleet. The 33% scenario analyzed in the DNV GL study shows:
- Lower energy production costs by replacing expensive peaking units (up to 3%)
- Reduced regulation capacity needs from less variability on the system (up to 50%)
- Additional ramping capacity for the challenging morning and evening load ramps
- Additional inertia on-line from rotating machinery will promote system stability
The significance of these benefits will grow as renewables increase beyond 33% and the need for a flexible and robust grid system is put to the test.
So perhaps there is a future for large-scale solar thermal power in California after all, if we recognize that these plants should not solely be in the business of selling green energy, but in the business of selling green ancillaries.
 A Huge Solar Plant Opens, Facing Doubts About Its Future, New York Times