Fact check: WSJ goes astray on California's integration of wind

The Wall Street Journal carried an article yesterday by Rebecca Smith on the growth of renewable energy in California and the effect on its electric utility system. The article makes several incorrect and unsupported assertions, addressed below:

Fact #1: It is unlikely that the state will experience a shortage of flexible power by 2017.

Last year, the California Public Utilities Commission required the grid operator to examine the need for flexible capacity in four scenarios in which renewable energy provides 33% of California’s electricity, and in all four cases it found zero additional need for flexible capacity to accommodate 33% renewables. This finding was documented in a 2011 regulatory filing that was endorsed by the California grid operator and all major utilities in California.[1]

Fact #2: Nuclear and fossil fuel power plants can go offline much more rapidly than renewable energy.

In California, nuclear and fossil-fired power plants can take 2,800 MW of generation offline in 1/60th of a second, which is a far greater challenge for grid operators and a far greater cost for the power system. A 1,000-MW change in wind output can be forecast using weather modeling; the failure of a fossil or nuclear plant cannot. As explained in the section at the end of this post, data from the California system operator indicates that a 33% renewable future actually has a lower need for fast-acting reserves than an all-gas future, and a slightly larger need for far cheaper slower-acting reserves.

Fact #3: Integrating renewable energy will not negatively affect reliability in California.

As explained in detail below, to reach 33% renewable energy, the California grid operator will use slightly more of the flexible reserves that it has always used to accommodate fluctuations in electricity demand as well as sudden failures at large fossil and nuclear power plants. The quantity and use of these flexibility reserves will be slightly increased so that electricity reliability is maintained at current levels, just as grid operators have increased their reserve levels in the past to accommodate new large fossil and nuclear power plants as well as increased variability in electricity demand.

Fact #4: Renewable energy output does not change significantly on a second-to-second time-frame, while the output of nuclear and fossil plants does.

Wind and solar output do not change significantly on a second-to-second time-frame, but rather more on the hour-to-hour timescale. However, nuclear and fossil plants do.

Dozens of wind integration studies, including many conducted in California, have confirmed that adding wind and solar to the grid only results in modest increases in total system variability. This is partly because changes in wind and solar output occur slowly, as it takes hours for a weather system to move the hundreds of miles necessary to affect a large share of a region’s renewable energy capacity. In addition, changes in wind or solar output are often canceled out by opposite changes at other wind or solar plants, or by opposite changes in other sources of electricity supply or demand. Moreover, changes in wind and solar output can be forecast by using advanced weather models, allowing grid operators to plan ahead and even more readily accommodate their output. Any incremental variability that is not accommodated through these factors can then be addressed by using slightly more of the same type of reserves that are already used to accommodate the variability and uncertainty that has always been on the power system.

In contrast, failures at large fossil and nuclear power plants occur instantaneously and without warning, requiring system operators to keep large quantities of expensive, fast-acting reserves on hand 24/7/365 to be ready in case an outage occurs. For example, in February 2011, dozens of fossil-fired power plants failed in a cold snap in Texas, causing rolling blackouts, while wind plants continued to produce as expected. In another example in 2012, a nuclear power plant in California had to be taken offline after jellyfish clogged its water intake equipment.

Overview of recent CAISO integration analysis

Analysis conducted for CAISO in 2012 further documents that the reserve needs in a 33% renewable energy future are only slightly higher than in the status quo, and that large fossil and nuclear power plants impose a reserve requirement that is a dozen times larger.

The CAISO analysis compared the reserve needs and challenges associated with reliably accommodating two different generation mixes for the state. In one scenario, all of the state’s electricity needs were met from natural gas, while the other scenario 33% of California’s electricity was obtained from renewables plus shutting down many of the state’s conventional power plants with once-through cooling systems.

Analysis of data in the report indicates that moving to a 33% renewable system from an all-gas system causes a 3-4% increase in total system reserve costs.[2] In 2010, all of these reserve costs accounted for less than 1% of the total cost of energy in the California electricity market. Therefore, on an average household’s monthly electric bill of $100, the increase in reserve costs in the renewables case would work out to an increase of around 3-4 cents!

Interestingly, the need for fast-acting reserves to accommodate sudden failures of large fossil and nuclear power plants declines by 103 MW in the 33% renewable case relative to the all-gas case, while the need for fast-acting regulation reserves for all other fluctuations (including electricity demand, renewables, and deviations in other sources of supply) increases by only 53 MW. Thus, the total need for expensive fast-acting reserves declines by 50 MW in the renewable scenario relative to the all-gas scenario. The need for slower-acting and far cheaper load following reserves increases by 351 MW in the 33% renewable case.

Moreover, it is important to point out that the cost of the contingency reserves needed to accommodate large fossil and nuclear power plants account for around 45% of the total reserve costs for the state, while in the renewables case the incremental reserve need for wind and solar amounted to 3.7% of total reserve costs. Thus, the total reserve costs for large fossil and nuclear plants are a dozen times larger.

Most importantly, adding renewable energy to the power system drives down the energy costs that make up 99% of the electricity market costs in California, yielding consumers significant savings on net. Wind and solar energy have no fuel costs and adding them to the grid displaces the most expensive power plant that is currently operating. A study from 2012 found that adding wind energy to the grid in the Midwest would save consumers between $5 and $15 per month.

In summary:

CAISO is already implementing a number of market reforms to make its system work more efficiently and better accommodate large amounts of renewable energy. They are also implementing faster transmission scheduling to allow more efficient movement of power within California and between California and its neighbors. By making their market mechanisms work faster and more efficiently, CAISO will help reduce the cost of accommodating all sources of variability on the power system. Finally, CAISO has also been a leader in using wind energy forecasting.

Wind and solar energy have no fuel costs, and adding them to the grid displaces the most expensive power plant that is currently operating. Adding renewable energy to the power system drives down the energy costs, yielding consumers’ significant net savings.

[1] The rulemaking found “As requested by the Commission, the CAISO developed a methodology for assessing renewable integration resource needs (the “CAISO methodology”), and applied this methodology with the assistance of the IOUs to assess the need for flexible capacity for the four CPUC-Required Scenarios and one other CPUC scenario analyzed by the CAISO. The results show no need to add capacity for renewable integration purposes above the capacity available in the four scenarios for the planning period addressed in this LTPP cycle (2012-2020). The additional scenario studied by the CAISO did show need.” http://tinyurl.com/b64ra4t, referenced on page 2-4.

The additional scenario that was not required by the CPUC looked at a large increase in electricity demand in the state, and found that with higher electricity demand there would be an increase in the need for flexible reserves. However, that increase in reserves should not be attributed to the increase in renewables energy use, as the four renewable-only cases did not find any increase in need for flexible reserves. Even under the high load growth assumption, almost half of the needed capacity had to be built anyway to meet local voltage support needs due to conventional power plant retirements.

[2] The table on page 12 here presents the results for the 50 hours when the power system needed flexibility the most. http://www.caiso.com/Documents/Presentation_E3_CAISO_Step2NeedAnalysis_Feb10_2012.pdf

Contingency reserves are fast-acting, expensive reserves needed to reliably accommodate the unexpected failures of large fossil and nuclear power plants. Regulation reserves are fast-acting, expensive reserves used to accommodate variability in all sources of supply and demand on the power system. Load following reserves are slower, lower-cost reserves used to accommodate slower variability in all sources of supply and demand on the power system. Contingency reserves are typically provided half by spinning and half by non-spinning reserves. Load following reserves were assumed to be provided by reserves whose cost is comparable to non-spinning reserves.

In California in 2010, regulation reserves (for load, renewables, and other variations) cost $10.6/MWh, spinning reserves (which provide half of the reserves for fossil and nuclear plant failures) cost $4.1/MWh, and slower-acting non-spin reserves (which provide the other half of power plant failure reserves, and we also use this cost as a proxy for load-following reserve costs) cost $0.6/MWh on average.  Including the costs for these reserves allows for a direct comparison between the reserve cost for the all-gas case versus the renewables case, which shows a 3.7% increase in reserve costs for the 33% renewables case:

All Gas = $14,317 per hour
Renewables = $14,850 per hour

Related articles: