October 15, 2015 by Jennifer Chen, Attorney, The Sustainable FERC Project
Accurately forecasting electricity demand is critical to ensuring that grid operators plan for adequate generation and transmission but do not over-procure or over-build, which can lead to overcharging electricity customers. PJM, the grid operator serving 61 million electricity customers in the Mid-Atlantic and Midwest, is improving its methodology for predicting future demand by more accurately accounting for energy efficiency savings throughout its footprint. PJM is nearly done with its proposed forecast methodology, which will allow customers to enjoy a cleaner electric grid at lower cost.
However, while PJM made improvements this past April in incorporating energy efficiency to decrease the forecasted demand, PJM has since introduced additional changes that increase the forecast again. Because PJM’s forecasts have historically exceeded actual demand, and over-forecasting incurs unnecessary costs, it’s important that PJM not backtrack too much.
What is demand (or load) forecasting?
Just as the weather forecast affects our plans each day, demand forecasting is an attempt to predict future electricity needs, which informs future plans for transmission and electricity generation. PJM annually forecasts the amount of customer electricity demand – the “load” on the system – years into the future by considering weather, economic, and other factors affecting customer electricity usage. The resulting load forecast helps to ensure that grid operators plan for enough power generation to satisfy electricity demand when it is at its highest (known as peak demand) and sufficient transmission (towers, lines, and substations) to deliver that electricity from power plants.
Discovering more energy efficiency in PJM
PJM has historically over-forecasted demand, and failing to account for all potential energy efficiency savings was one of the reasons. To help PJM address this problem, the Sustainable FERC Project coalition and Synapse Energy Economics quantified some of the energy efficiency missing from PJM’s forecast by surveying state-funded energy efficiency incentive programs. Synapse forecasted the future savings and showed that properly incorporating this information can significantly reduce the over-forecasted amount.
In parallel, PJM began a process to reform its load forecast methodology last fall and introduced (in April 2015) improvements to more comprehensively capture energy efficiency based on U.S. Energy Information Administration data. PJM’s approach focused on current and future energy efficiency standards for heating, cooling, and lighting equipment, along with other factors affecting customer adoption of more efficient appliances and equipment.
In its April proposal, PJM found that its forecast should be lowered by 9,000 megawatts between 2015 and 2019, a reduction largely due to projected energy efficiency savings. (The other main drivers lowering the forecast were weather and economic factors). We think that the April proposal, which is the product of months of work and stakeholder input, resulted in a conservative and reliable estimate of load reductions due to energy efficiency, because our coalition’s experts independently arrived at similar levels of reductions from energy efficiency using a different approach and dataset.
The April proposal was a big step forward. However, PJM more recently proposed further changes to its methodology that will increase its forecast again. In addition, PJM is working on other proposals that will increase the capacity it will procure in the future. These little steps back could unwind much of the benefit of cutting down on over-forecasting in the first place, and we urge PJM to not backtrack too much before finalizing its proposal.
Accurate load forecasting can save money and reduce pollution
What’s at stake? PJM’s proposed lower load forecast benefits customers and the environment. We are familiar with the immediate utility bill savings from consuming less electricity through smarter energy use in our appliances and equipment. But that same energy efficiency can help save even more money if grid operators know about it sufficiently in advance to plan for and avoid building unnecessary power plants and transmission infrastructure.
Securing enough generation to satisfy future needs (or “capacity“) is expensive; increasing energy efficiency helps by reducing future electricity demand, which in turn decreases both the price and amount of necessary capacity. The largest chunk of savings comes from the former effect – by paying less per unit of capacity, we could save roughly $2.4 billion in just one year. (I arrived at this estimate by multiplying the savings per megawatt by the total number of megawatts PJM would have to procure. In this analysis, PJM’s proposed load forecast would have reduced capacity needs by 5,000 megawatts in one year, of which 4,700 megawatts are attributed to energy efficiency.)
Accurately reflecting reductions during peak demand can also help us save on transmission costs. According to the U.S. Department of Energy, PJM spent over $28,000 on transmission investments for each megawatt of peak demand in 2013. (For reference, PJM’s April proposal would have decreased the peak forecast by several thousand megawatts.)
There are also environmental benefits to reducing capacity and transmission needs, especially during peak demand. Not only do we avoid building unnecessary power plants, we also avoid running the least efficient (and most polluting) power plants used only during peak demand. And, of course, not building unnecessary transmission infrastructure avoids negative ecological impacts (while focusing investments on truly needed transmission).
Looking to the future
Accurately accounting for energy efficiency will be even more critical moving forward as we transition to a low-carbon economy. This is especially true because the U.S. Environmental Protection Agency’s Clean Power Plan cutting carbon pollution from the electric sector could drive significant increases in planned energy efficiency. PJM’s improved load forecast model will help to illuminate this energy efficiency for the benefit of consumers, states, and the entire utility sector.