Electric customers in the Mid-Atlantic region may be missing out on $433 million in annual utility bill savings over the next three years – as much as $1.3 billion** – because their transmission grid operator seems to be undercounting the impact of energy efficiency programs that avoid the need for more power, according to an analysis released today by The Brattle Group.
The analysis, which the Sustainable FERC Project commissioned, examines the way that PJM, the entity responsible for operating the electric transmission grid that utilities use in all or parts of 13 Mid-Atlantic states and the District of Columbia (as shown in this map), incorporates energy efficiency in its analysis of future regional electricity demand.
How much energy efficiency is missing?
Using utility-reported, publicly available data, the analysis found the method PJM uses to predict future electricity demand, known as “load forecasting,” does not capture all the energy savings of current and expected future energy efficiency programs and actions that customers in the PJM region are undertaking.
In fact, the report shows PJM’s forecasting method is over-estimating the region’s electricity needs at an increasing rate into the future – in 2022 by at least 27,245 gigawatt hours, the equivalent of 3,000-megawatts of power plants running around the clock. The Brattle Group concluded the unnecessary cost to customers related to this undercounting of efficiency could total $433 million annually over the next three years and $127 million a year after that. At the same time, the true amount of “missing” energy efficiency (and therefore related costs) may be underestimated even in the Brattle analysis, for reasons discussed below.
What’s wrong with over forecasting?
Over-forecasting energy demand causes at least three problems:
- Customers pay more for electricity than necessary. The Brattle analysis predicts that if PJM’s load forecasting methods captured all of the energy efficiency savings from utilities, total customer costs would drop by $433 million annually for the next few years and, over time, continue to save customers $127 million a year because less power would need to be procured. In other words, customers are now paying for electricity they don’t need.
- Customers pay more, again, for unnecessary electric grid costs. PJM plugs its electric customers’ demand forecast data into its planning models to predict the need for new transmission lines and other grid infrastructure upgrades and additions. PJM could avoid some of these planned grid upgrades and additions by properly counting for energy efficiency. Although the Brattle analysis doesn’t estimate the potential dollar savings from deferred transmission investments, we can take a lesson from PJM’s neighbors. The electric transmission grid manager in New England, known as ISO-NE, has revised its load forecasting method to better capture the existing and planned energy efficiency in its six states. Thanks to energy efficiency investments over the last few years, in addition to energy cost savings like the PJM potential I mention above, the region avoided over $400 million in transmission upgrades, ultimately lowering customer bills by the same amount. And since PJM’s peak electricity demand in PJM is almost six times higher than ISO-NE’s, the savings in its region could be much higher.
- It harms the environment. Building new transmission and generation infrastructure can harm wildlife habitat and natural areas. While this construction often is necessary to improve grid reliability and deliver wind and utility-scale solar power to market, building it to meet demand that doesn’t exist can cause unnecessary environmental harm – and add to the financial costs, yet again. Inflated load forecasts leading to the over-procurement of energy also provide more revenue to power plants, including older, dirtier and more costly coal plants that otherwise likely would shut down for good.
Does the Brattle analysis tell the whole story?
If saving over $400 million annually doesn’t impress you, consider that Brattle’s analysis also likely underestimates the amount of missing energy efficiency and related missing cost savings.
First, as mentioned, the analysis doesn’t include savings from avoided or deferred transmission infrastructure costs or the avoidance of environmental damage. Based on the New England experience, the potential is great.
Second, the Brattle analysis uses utility-reported numbers (known as “Form 861 data”) collected and published by the Energy Information Administration, an independent federal agency. EIA’s data has several understood limitations including what is not always apples-to-apples information provided by different utilities, delayed availability (the last available reporting year is 2012) and an inability to predict trends of increasing efforts to capture efficiency. Therefore, to try to ensure credible and consistent results the economists performing the Brattle analysis made several rather conservative assumptions. For example, when Brattle economists applied the methodology used in the PJM analysis to New England’s load forecasts, they calculated significantly less energy efficiency savings than ISO-NE and its stakeholders say is actually happening in New England. Also, Form 861 reports utility energy efficiency programs, which represents only a portion of the energy efficiency that happens each year. Utilities do not usually capture in their Form 861 reports data savings from building energy code compliance, appliance standards, or in some cases third-party energy efficiency from industrial or other customers.
Third and very importantly, the analysis only looks at accounting for existing and planned energy efficiency utility programs, but doesn’t consider that many states plan to increase the amount of energy they save in coming years. For example, Ohio’s energy efficiency standard required only a .8 percent increase in efficiency savings for 2012, but includes a 2 percent increase in required efficiency for 2021. Even if there is an eventual leveling off, existing state efficiency standards and the U.S. Environmental Protection Agency’s proposed Clean Power Plan to cut harmful emissions from existing power plants, which makes energy efficiency a central tenant, should drive additional efficiency gains over the next decade. As efficiency increases, the problem I'm describing will get worse.
Can PJM solve over forecasting?
Predicting future electricity demand is harder than predicting the weather and it is impossible for PJM (or any other entity) to be exactly right. Politics, the economy, fuel prices, customer behavior, and the climate all affect actual electricity demand and forecasts of load growth. The amount of energy efficiency in PJM is only one piece of PJM’s complex load forecasting puzzle. However, other regional grid operators like those in New England, and New York and California have successfully worked with their respective stakeholders to get a better handle on the energy efficiency factor in their forecasts.
The Brattle Analysis suggests that there is much to be gained if PJM were to follow the lead of these other regional grid operators and initiate a stakeholder working group to examine more rigorously how PJM’s forecasting methods could capture more comprehensively the future amount of energy efficiency rather than leave customer bill savings and air pollution savings on the table.
Fixing the problem may take time and commitment, but the payoffs could be enormous. Customers, PJM, and the environment deserve to reap all of the money-saving and clean energy benefits of our individual and collective energy efficiency investments.
** I derived the potential $1.3 billion in savings over three years from the Brattle estimate that in the short term, customer savings could equate to $433 million annually, and in the long run, savings of $127 annually are possible. While Brattle did not explicitly estimate the duration of the larger short-run benefits, I interpreted short term as three years, and assumed full achievement of the potential savings equal to $433 million annually during that period ($433M x3 = $1.29B).