(The author is a Reuters columnist. The opinions expressed are his/her own)
By Gerard Wynn
LONDON, Oct 2 (Reuters) - While a U.S. Department of Energy study has concluded that fossil fuel power plants can cope with the physical stresses of balancing more renewable power, it leaves unanswered the bigger question of the financial losses utilities face.
The study focused on the costs of cycling, a term which refers to the wear and tear from shutting down and restarting power plants, as well as ramping up and down and operating at part-load.
As the grid penetration of renewable power rises, fossil fuel plants will have to cycle more frequently to balance the variability and forecasting inaccuracies associated with wind and solar power.
The latest study found that cycling costs at higher levels of renewable power penetration worked out at a small, but non-negligible 2-5 percent of operating costs for fossil fuel power plants.
But it took a grid system view, ignoring the potentially much bigger, market losses for fossil fuel generators as a result of competing with zero marginal cost renewables and from operating at partial rather than full capacity.
These costs add to the cycling expenses revealed by the U.S. study, threatening the viability of fossil fuel power plants, as shown by evidence in Germany.
These bigger costs were beyond the scope of the U.S. study, which tested scenarios including wind and solar power accounting for up to a third of electricity consumption in the U.S. portion of the Western Interconnection grid (which also includes parts of western Canada and Mexico) in 2020.
It was published last week by the National Renewable Energy Laboratory (NREL), the Department of Energy's primary national think-tank for renewable energy and energy efficiency research and development.
NREL considered it was important to demonstrate that renewable power did not impose substantial additional operating costs on fossil fuel generation, to allay concerns about the rollout of wind and solar power.
Cycling costs are also relevant to grid operators who dispatch power on the basis of least operating cost.
"Cycling of thermal plants causes temperature swings that can lead to creep, fatigue, and fatigue/creep interaction in components and equipment," the report found.
"This in turn can lead to increased operations and maintenance costs, more frequent repairs, reduced component life, and more frequent forced outages. Furthermore, some plants were designed for baseload operation and have limited ramping capability."
The NREL study calculated cycling costs by comparing annual operating and maintenance costs with operational data at 170 coal and gas plants in North America, surveying seven categories of gas and coal plant. (See Chart 1)
Cycling costs comprise start costs (from cold, warm or hot, depending on the number of hours previously off-line) and ramping costs, from increasing output at a power plant already on line.
Start costs included: start fuel (to heat systems to operating temperatures); maintenance costs from start-related thermal and pressure stress; and extra power, chemical, and water costs.
"The study found that most coal units cost more to start than gas units; however, the distributions overlap for most unit types, and some coal units cost less to cycle than some gas units. Small coal units have the highest cost per MW (megawatt) for starts," NREL said. (Chart 1)
Costs were much smaller per ramp, because of smaller heat and pressure stresses compared with starting a power plant, but ramping events were more frequent.
Chart 1: (page 56) goo.gl/mNxaRW
Chart 2: (page 24) goo.gl/mNxaRW
The study confirmed that the additional operating costs of cycling was negligible compared with the fuel saving from switching to renewables, meaning that wind and solar had a much smaller operating cost than fossil fuels.
This seemed a slightly odd conclusion, given that it could never have been seriously in doubt.
More interestingly, it calculated that cycling increased the average cost of fossil fuel power generation by $0.47-$1.28 per megawatt hour, or 2-5 percent of total operating costs, at the highest levels of renewable power penetration (a third of consumption). (Chart 2)
Starts accounted for a bigger portion of cycling costs than ramps.
Solar power caused greater supply grid variability than wind, because of sunsets which coincided with early evening peak demand, but clearly sunsets can be anticipated and so do not pose a serious grid planning problem.
Wind power posed more extreme uncertainty, because of higher day-ahead forecast errors compared with solar power, leading to a higher reserve power plant requirement.
Forecasts were much more accurate four hours ahead, however, showing the value of aligning power plant commitment to shorter timescales.
There are bigger market costs regarding the integration of renewable power, however.
In Germany, gas-fired power generation is now loss-making, triggering capacity mothballing and closures.
That is partly because utilities tied themselves into expensive long-term gas supply contracts with Russian and Norwegian exporters whose prices are tied to crude oil. It is also from competing with more renewable power.
First, renewables have caused lower wholesale power prices because they have near-zero marginal cost; and second, they are causing fossil fuel power plants to operate at lower capacity than formerly.
The higher operating cost of cycling, revealed by the NREL report, are another cost.
The impact on viability is more urgent in Europe where some countries have a much higher renewable power penetration than the United States.
The chief executives of eight European utilities in May said they faced a "perfect storm which is endangering security of supply", as a result of declining market incentives.
They called for: "a European coordinated approach... in which all assets contributing to the security of supply of European customers are fairly remunerated."
NREL acknowledged the impact on margins.
"This raises questions about who should pay for the cycling costs of incumbent plants or what happens in the marketplace to address the viability of a plant that might be needed for reliability but might no longer be profitable. These questions are not addressed in this technical report," it said.
The example of Europe suggests that the answer cannot come soon enough. (Editing by James Jukwey)