Search for wind-related grid problems finds a bigger concern

A national laboratory report targeted at wind power integration has found unexpected evidence of the electricity grid's vulnerability to potential blackouts due to the current operations of conventional -- not renewable -- generation.

The study by Lawrence Berkeley National Laboratory, issued last month, reports that frequency levels are dropping further than realized on the nation's two largest grid systems, particularly at the start of the day when electricity demand ramps up and when it ramps down in the evening.

The speed and timing of spinning power generation are closely regulated so that power can be shared across the nation's three synchronized grids -- east and west of the Rocky Mountains, and in most of Texas, which has its own grid. Keeping the lights on requires that power generators respond automatically within seconds to halt a swing in frequency and restore levels to the 60-cycle-per-second standard.

But the requirement for instantaneous response is being eroded by a combination of factors, including the ways nuclear plants and merchant gas generators are typically operated to maximize operators' revenues and the lack of a settled way of paying generators to provide instantaneous frequency support, the study says.

The emerging threat is a case study in the challenges growing out of the patchwork "deregulation" of the power sector and the lack of comprehensive national policies for managing the grid. All this is emanating from concerns that once pointed to the disruptive shift toward renewable power and smart grid technologies aimed at energy conservation, experts say.


Wind power is still too small a factor to be playing a part in the decline in frequency response, the study says. But as wind and solar power grow, they could add to the problem -- or help prevent dangerous frequency swings if their operations are managed to do so, officials said.

The study results suggest that if the grid were hit with a sudden shutdown of a large generating plant just at a moment when frequency was already declining, "a much bigger frequency excursion would have been recorded, potentially one that would have initiated under-frequency load shedding." In that case, "these customers will experience a blackout," which could spread to other areas, the report says.

Vulnerability begins at dawn

"Our feeling is we are not at that point, nor do we want to allow that point to occur," said an official of the PJM Interconnection, the grid manager in the mid-Atlantic and parts of the Great Lakes region.

"It would be an usual event, but in the case of that event, we would have seen a problem," said a Federal Energy Regulatory Commission official, speaking not for attribution.

The frequency issue, which was initially highlighted by FERC in 2007, has been under study by industry groups working through the North American Electric Reliability Corp. (NERC), which is delegated by FERC to oversee grid reliability. Industry agreement on new standards had moved slowly until a new deadline was set by FERC last year. NERC is now preparing to issue proposals to FERC by May 2012.

FERC commissioned the Lawrence Berkeley study to see whether measurements of frequency variations could be a useful tool in determining how much wind power could safely be brought onto the grid. The study concludes that this approach could serve that purpose.

But in the course of its detailed investigation, the laboratory team headed by Joseph Eto discovered that the problem of "frequency deviation," well-known within the industry, was direr than regulators and industry officials had realized.

The grid becomes vulnerable at the start of each day, when cities awaken and electricity demand rises rapidly. Grid managers schedule generation to come online according to historical patterns and weather forecasts, but there often are surprises.

If there is suddenly too much demand for the amount of generation online, control systems send an automatic signal to speed-regulating governors at power plants to get them to ramp up power within seconds. This "primary" frequency response is designed to halt a drop in frequency until other automatic but slower-responding "secondary" controls come into play. Finally, control room personnel acting within the hour can issue orders to bring other generators online to restore frequency to the proper level.

Big coal-fired generating plans equipped with governors, whose spinning generators have sizable inertial power, were ideal for answering the need for primary response.

'Head room' to prevent blackouts is shrinking

Robert Cummings, NERC's director of system analysis and reliability initiatives, said in an interview that this capability has grown weaker.

"The nuclear plants generally don't operate with the governors on," he said in an interview. Gas-fired generators, particularly the more complex combined cycle units, are a "mixed bag," he added. "Some want to sell megawatts. They are going to top out the units at the maximum capability whenever they can. Their ability to respond is usually limited by that."

"We are coming away from a traditional large generation plant that has quite a lot of command and control resources," said the PJM official. "Combined cycle and wind are operated more at their peak efficiency, where the coal plant has the ability to back off and provide that reserve quickly." That reserve capability is called "head room," the Lawrence Berkeley report says, and head room is shrinking because of changes in the kinds of generation on the grid.

Additionally, many plants operate to produce a set output, and even if they do respond initially to a signal to provide frequency support, they will quickly go back to the original set point, and that can reduce support when power demand is ramping up.

Generators may be paid to provide slower-responding frequency support, but are not currently compensated for the instantaneous response needed to halt a sudden deviation in frequency, the FERC official said.

"Taking their side," the PJM official said, "unless they are paid for that head room position, they see a loss of opportunity to gain a return from existing market prices."

The wind industry had awaited the Lawrence Berkeley report with misgiving, fearing that it might provide ammunition to renewable energy critics. But the report concludes otherwise, at least for the present. "This pattern of frequency deviations is not a function of wind operations," it says.

"In the Eastern Interconnection because there are such small amounts of wind, it [declining frequency response] can't be blamed on wind. It's not wind's fault," the FERC official said. The problem would be expansion of wind power coupled with a failure to deal with the frequency response challenge, the official said.

"One of the fundamental findings of the report was that as the generation mix changes, as more renewables are brought online, or mandated to come online, they have the potential to displace generation that provides frequency response," the FERC official said. "If that occurs, then this problem may become exacerbated, so your reliability problem materializes. ... It's the combination of the two that can create a problem."

Wind power might be a solution

Cummings said that on the other hand, wind power can be a vital ally in maintaining frequency levels. "Wind has a quick capability of moving down," he said. And it can also respond quickly to add energy if there is head room. "You could make that very, very influential in the one- to two-second period," he said.

Other technologies such as flywheel units and battery storage can also be put into the mix, he added. Sophisticated demand management programs can be used to stabilize frequencies quickly, he said. Grid operators are adding new monitoring equipment that greatly adds to their ability to track changes in frequencies and voltages in fractions of seconds.

The challenge, officials said, is to get agreement within the industry, and among regulators, on what is needed and how to pay for it.

The problem extends across engineering, operating and market issues, said Peter Fox-Penner, chairman emeritus of the Brattle Group and author of "Smart Power," about the future of U.S. utilities.

"I do believe the engineering problems are solvable, if only by adopting constraints on operating practices. What is very, very difficult is determining the incidence of costs and getting people to pay those costs. So it's fundamentally an institutional and regulatory challenge," he said.

"Without a clear obligation [to provide support] in this day and age, and without some measured performance standard -- which to date is not a requirement -- it's difficult to make it an objective of the utility," the FERC official said.

Cummings said the NERC standards development committee expects to meet its 2012 deadline for reporting proposals to FERC. It is also taking interim steps. "We just completed a survey of the generators' owners and operators for their governor settings. Whenever you shine a light like this on a problem, an awful lot of people pay attention, and you end up with an improved performance," he said.

Creating new grid reliability standards follows a process under the Federal Power Act that requires proposals from the industry to FERC -- the commission cannot order them on its own. "This one has taken longer than some, probably longer than most," the FERC official said. "It's been three years. We need some action on this."

Like what you see?

We thought you might.

Start a free trial now.

Get access to our comprehensive, daily coverage of energy and environmental politics and policy.



Latest Selected Headlines

More headlinesMore headlines

More headlinesMore headlines

More headlinesMore headlines

More headlinesMore headlines