How climate change will stress the grid and what ISOs are doing about it
Global warming is expected to increase peak demand and extreme weather events, but new grid planning practices can help ensure power reliability
President Trump’s arrival in Washington was supposed to signal a reduced emphasis on environmental regulation. But as conservatives gear up to dismantle the Obama administration’s climate initiatives, the town has been abuzz with talk of the costs of climate change.
Last week, House Republicans held a hearing on the social cost of carbon — an administrative tool used to estimate the costs of global warming when making federal regulations — a discussion that reportedly spiraled into a debate on basic climate science.
Days later, the Heritage Foundation — a conservative think tank influential in the Trump White House — released a recommendation letter for the Department of Energy, advising that the agency stop using the social cost calculation because “minor changes” to the methodology could result in “negative values which indicate CO2 emissions are a net benefit to society.”
But while climate change skeptics wring their hands over the theoretical costs of global warming, the nation’s electricity system operators are already feeling it — and planning for its future effects. Average temperatures are rising across the U.S., pushing grid operators to examine whether they have adequate capacity to meet higher power demand and sharper spikes in peak load.
Projected temperature increases will raise average electricity demand 2.8% across the U.S. by the end of the century, according to a new National Academy of Sciences study. The impact is expected to be greatest in the summer months, when cooling load is already high; the study projects a 3.5% increase in average peak demand over the same timeframe.
The demand increases are compared against a business-as-usual scenario. In more extreme cases, peak demand could rise 7.2% to 18%, making the cost of needed generation even higher.
Meeting that demand could require $120 billion to $180 billion in new natural gas peaker plants. But smart planning could save a lot of that cost, according to experts.
“This is a thought exercise,” University of California Professor and study co-author Maximilian Auffhammer told Utility Dive. “We observed the system peak demand increase when it was hot, applied that to climate data, and calculated what several load balancing authorities’ peak demand would rise to.”
The study asks what pressure that increased peak demand will put on the U.S. electric power system, Auffhammer said. “It is not what is actually going to happen, but what the system is working against.”
Outside the thought exercise, the real-world question is how well utility and balancing authority system planners can respond. James Hoecker, former chairman of the Federal Energy Regulatory Commission (FERC), said that it is often difficult to get policymakers to respond to such long-term concerns when “it is stretch for them to deal with 5-year and 20-year planning.”
Even so, Hoecker and other experts stressed that the warnings in the National Academies paper are real, and that stakeholders should plan today to prevent the need for costly system investments later.
“The best way to deal with climate change is to prevent it,” he said. “There will likely be enormous stresses on the electric system. A resilient, flexible, and fully integrated grid is a good bet long term.”
Climate change fuels rising demand, especially at peaks
“Our results show that much greater generation or storage capacity will be needed to meet the demands that will come with the warmer climate that scientists predict,” Professor Auffhammer said.
The National Academies paper estimated end-of-century changes in average load, daily peak load, the most extreme changes in load, and the number of days of the most extreme load changes.
“For nearly all regions, increases in the mean of the temperature distribution will increase average and peak loads at higher temperatures,” it concluded.
In the most extreme cases, increases in peak load averaged between 153% and 389%. This means “levels of demand that are currently considered unusually high will become much more common, even absent changes in population or income,” the paper reports.
In the single most extreme case, the number of days of increased electricity demand rose over 1,500%. In this case, the Electric Reliability Council of Texas, which serves over 80% of the state, would have 65 days a year with “a peak load of or in excess of the currently four highest load days.”
Impacts on peak loads will by region, the researchers found. The biggest impacts would be in the U.S. South, while the Northwest would likely see a load decrease. Key factors in regional variations, researchers found, were whether heating is primarily done with electricity or natural gas, how much air conditioning is used, and how much of load is industrial process heating and cooling.
“These results are indicative of a need for regionally distinct strategies,” the paper reports. There will be substantially more “peakiness” of electricity demand in the U.S. South and less in the Northwest. “These regional changes imply shifts in the need for new transmission and generation (or storage) capacity in particular.”
The researchers found meeting a nationwide increase of 7.2% in peak load would require a $70 billion investment in new generation capacity. The most extreme case of an 18% peak load spike would require a $180 billion investment in new generation.
Many economic and societal factors outside the study’s purview make impacts of adaptation “ambiguous,” the paper reports. Higher temperatures could cause more use of air conditioners and bigger peak demand spikes, but that might lead to more efficient air conditioning technologies.
Economic growth would likely increase peak demand spikes “because heating and air conditioning use tends to increase with income,” the paper reports. On the other hand, advances in battery energy storage or electric vehicles for storage that might come with economic growth and technology advances would likely smooth peak demand.
Renewables’ impacts on peak period spikes are “ambiguous” because their generation profiles do not necessarily match peak periods, the paper reports. But their use with energy storage technologies and/or time-varying rates could shift load and smooth peaks. Climate change could, therefore, increase demand “for storage technology, demand response programs, and alternative pricing schemes.”
Grid operators plan for hotter climate
One of the drawbacks of load forecast studies like the National Academies report is that they
“don’t have tools to forecast out-of-historical-pattern ‘shocks’ that could dramatically affect system planning,” said Judy Chang, a principal at the Brattle Group, a consultancy. In the real world, system planners have to prepare for more than just forecasted population growth, GDP increases and other predictable factors.
But the nation’s grid operators say they are on the case. Five transmission system operators reached by Utility Dive their planning is done in a way that can include preparation for Chang’s “shock” events.
Electric Reliability Council of Texas (ERCOT) 10-year and 15-year scenario planning processes were introduced by Brattle and Chang. They develop cases based on a range of economic, technological and regulatory trends that could impact demand and generation.
“We update this Long-Term System Assessment every two years,” ERCOT Communications Manager Robbie Searcy said. “For more near-term planning and operations, our meteorologist assesses weather trends and their potential impacts on consumer demand and system performance.”
The California Independent System Operator planning is done through a stakeholder effort, but “we can't guess at this point what inputs, if any, regarding climate change impacts might influence our study processes,” Senior Public Information Officer Steven Greenlee said.
Planning in the Southwest Power Pool (SPP) includes predicting, modeling, and analyzing future needs and “more frequent and higher peak demands will certainly have an impact on transmission planning,” Vice President of Engineering Lanny Nickell said. They “will likely increase the need for transmission upgrades that address needs related to both reliability and economics.”
SPP’s planning for reliability would necessitate adding capacity “if peak demand rises,” Nickell added. And because peak demand spikes would likely increase system congestion and costs to SPP’s customers, they would also drive new planning.
PJM Interconnection planners recently shortened their historical weather data assessment period from 40 years to 20 years when they observed a significant increase in extreme hot days in the more recent two decades, Spokesperson Paula DuPont said.
Due to this shortening, which planners expect to continue, PJM’s “50/50” summer load forecast was increased 1%, though net load decreased 4% because of other system factors, she added.
Midcontinental Independent System Operator (MISO) planners and system operators are constantly reacting to the changing energy landscape and factors like environment, regulations, economics, aging facilities, and new technology, J.T. Smith, director of policy studies J.T. Smith told Utility Dive.
MISO’s 20-year planning, which always remains in compliance with national reliability standards, considers those kinds of factors’ impacts on operations, resource adequacy, and transmission, he added. It must also enable MISO’s competitive electricity markets to benefit all customers.
Changes in climate that affect future forecasts for demand and energy “will be captured through the forecasting of system conditions and incorporated into planning analyses and decisions,” Smith said.
Hoecker, currently counsel for transmission advocacy group WIRES, is urging planners to think about grid modernization as a tool to both mitigate and adapt to climate change. There could be a big opportunity if the Trump administration follows through with its plan to designate $100 million for transmission as part of its infrastructure renewal spending.
“We know that we can look ahead two or three decades and do effective planning for the transformation, that will take place between now and the mid-2030s,” he said. “An optimized, cost-effective grid would be ready for the changing generation mix, the new technologies, the changes in demand, and the need to accommodate new markets.”
The demand spikes described in the research will not occur everywhere at the same time, he added. “A robust grid offers a way to deal with them by moving generation to where it is needed.”
Is co-optimizing a better answer?
As policymakers prepare the grid for a warmer world, the concept of co-optimized transmission planning is particularly relevant, Chang said. The concept was recently described in a 2015 report for the Eastern Interconnection States’ Planning Council (EISPC) and the National Association of Regulatory Utility Commissioners (NARUC).
Typically, transmission is planned to meet forecasted demand, Chang said. That disadvantages generation resources which require new transmission, like utility-scale renewables energy built in remote, resource-rich regions.
Considering both generation and transmission in system planning could lower the cost of electricity delivery by decreasing investments in both, according to the 2015 co-optimization report. It also could reduce operating costs and maks decision-making on retirements and upgrades of generation more efficient.
“If you co-optimize planning for both transmission and generation simultaneously, you can save a lot of cost on generation for a bit more cost for transmission,” Chang said. “And that transmission can bring a lot of benefits to the grid along with lower cost generation.”
Co-optimization clarifies when and where the best uses are of variable renewables, demand response, distributed generation, and energy storage, the paper adds. It also guides integration of those technologies into the system’s resource mix and allows for their most efficient use in complying with environmental regulations and resource adequacy requirements.
The co-optimization paper describes a limited and a full version of the concept. Modeling on a national scale showed “full co-optimization can save up to 10% or more of total generation and transmission costs compared to generation-only planning and 5% or more compared to transmission-only planning,” the paper reports.
Co-optimization works for vertically integrated utilities because it identifies the most efficient generation-transmission investments. It also works in deregulated markets because it allows more insight into “how generation dispatch and investment will respond to changes in transmission capacity, access, and congestion,” the paper notes.
In the modeling, full co-optimization required an incremental transmission investment of roughly $60 billion but saved $150 billion from the cost of generation-then-transmission planning, the paper adds. “That is, there was a 2.5 benefit/cost ratio for the incremental transmission investment.”
Reduced capital costs for generation accounted for 40% of the savings, the paper reports. “Traditional transmission planning processes, which do not consider changes in generation siting and capital costs, miss a potentially very important benefit of transmission.”
The savings are greater than the transmission investments, according to the co-optimization paper. They are possible from co-optimization because “the most profitable locations for renewable and nonrenewable plant investment strongly depend on where grid reinforcements are made.”
This relates directly to the research on peak demand, Chang said. Though almost all transmission planning decisions are based largely on meeting system peak, “it is also about the volatility of uncertain futures.”
Transmission: the best insurance
Transmission investment is one of the best ways to guard the power system against uncertain futures, including those that may come with a changing climate, Chang said.
System-wide, there is a diversity not just in demand peaks but in load profiles as well, she noted.
“The scenario-based planning we did with ERCOT is a way to think about and prepare for the uncertainty introduced by severe weather or drought or other extreme events, whether from climate change or not,” Chang said.
Co-optimized planning is the most cost-effective way to use transmission as insurance against the uncertainties that scenario planning reveals as possible, she said. “In traditional planning, the uncertainties are planned away because forecasts based on historical year averages show things like Superstorm Sandy as so unlikely.”
Uncertainty in the system is increasing and it has to be addressed by system planners and state regulators, Chang said. “In today’s typical planning, they only see the cost and not how much transmission investments can save ratepayers. But the amount of money customers end up paying from these shocks is tremendous.”
The enormous costs are hard not to see. Studies cited by Brattle found the U.S. system would need $100 billion to $300 billion of transmission investments from 2008 through 2028 to support a capital expenditure of up to $1 trillion for generation.
But those estimates do not recognize how transmission “can help reduce overall costs in the face of a shifting generation mix, evolving regulatory requirements, and the changing demands of electricity customers,” Brattle reported. It protects customers from “higher risks of significant cost increases.”
Even before the threat of climate change introduced today’s unprecedented levels of uncertainty, Chang’s work with Brattle showed “’proactive’ or ‘anticipatory’ planning” would reduce electricity costs significantly.
Savings from proactive transmission planning and development was estimated, under a set of environmental constraints less rigorous than those the present system faces, at $30 billion to $70 billion in total generation and transmission investment costs through 2030. It also found $47 billion per year in savings for customers.
The various peak demand and load impacts described in the research are likely to become increasingly severe, Chang said. “Well-planned transmission can definitely help mitigate the costs associated with them.”