Avoided costs are rising, impacting the cost-effectiveness of managed charging and load flexibility programs. Learn why current avoided costs may double over the next several years from WeaveGrid’s Brad Harris.

‍

If you are a wonk on energy policy, the idea of reassessing avoided costs may sound redundant. Hasn’t this been a constant conversation? Nearly every proceeding on demand-side management (DSM), distributed energy resources (DER), or rate design includes some heated debate over how to measure what costs utilities avoid by reducing or shifting load. 

But the shifts occurring in capacity needs, energy prices, and infrastructure aren’t marginal. They’re significant. The result? The methodologies we’re using to determine avoided costs, many of which determine the budgets for programs like managed charging, are undervaluing resources.

These changes have real implications because avoided costs determine: 

• which programs regulators approve, 
• how much customers are incentivized to participate, and 
• how utilities and program administrators justify their investments. 

By my estimate, avoided costs could be at least double what they were a few years ago. Let me tell you why. 

What Are Avoided Costs?

Avoided costs represent what utilities don’t have to spend due to programs and technologies, like DSM or DERs. A simple example: if a customer uses less electricity due to an energy efficiency program, the utility doesn’t need to purchase as much natural gas to generate power. That’s an avoided cost. 

These savings stack across multiple categories—generation, energy, transmission, and distribution. Each of these is changing in ways that make avoided costs more valuable than what current models assume.

Avoided Generation Capacity 

Generation capacity is the cost of building enough power plants to maintain reliable service. With enough DSM resources, a utility can avoid or defer another power plant, saving their customers the cost of that plant. For years, many utilities calculated avoided generation capacity based on the cost of building and maintaining a natural gas combustion turbine plant or “peaker unit.” This made sense. If a utility needed generation capacity, the cheapest way of providing that was to build a peaker unit. 

Dramatic increases in electricity demand, in part due to data centers and investments in manufacturing, have resulted in a rush to buy electric generators. Increased demand has resulted in skyrocketing prices for natural gas turbines. An EIA report published in January 2024 estimated the capital cost of various natural gas-powered generators from roughly $900 to $1,000 per kW. However, it appears that this cost has skyrocketed to $1,578 to $1,936 per kW based on an Entergy filing to the Louisiana Public Service Commission. Jigar Shah, the former head of the Energy Department’s Loan Program’s office, recently discussed this spike in costs on the Open Circuit podcast and posted on Bluesky that 2027-2029 delivery may increase to $2,400 per kW. Manufacturers are also reporting longer lead times for utilities to acquire gas turbines, due to high demand and limited supply chain capacity.

The recent capacity auction in PJM has further demonstrated the increasing cost of generation capacity. Capacity prices spiked almost tenfold due to the retirement of fossil-fired generation, the challenges of building new capacity, and the demands of new load growth.

Mandates to reduce carbon emissions will also likely increase generation capacity costs. 

Take the example of North Carolina, which in 2021 passed legislation to achieve a 70% reduction in carbon emissions from 2005 levels by 2030 and become carbon neutral by 2050. In response to this landmark legislation, Duke Energy and other stakeholders argued that a natural gas-fired unit, which is not consistent with the state’s carbon goals, was no longer an appropriate capacity resource to use for avoided costs when evaluating DSM programs. 

The North Carolina Utilities Commission (NCUC) ruled that a hydrogen-capable combustion turbine will be used “until an alternative dispatchable clean-energy pure capacity resources is identified in future [Carbon Plan IRPs]. While the exact impact of this change is not public, given that Duke Energy filed to double its incentive for its smart thermostat demand response program, it can be reasonably estimated that this change roughly doubled the value of avoided generation capacity. 

Avoided Energy

For years, avoided energy costs have tracked the price of natural gas, since natural gas plants are often the marginal generator. That is likely going to continue for the foreseeable future. Load growth, including EV adoption, data centers, and electrification, could shift the picture.

While natural gas prices are currently low, the increase in gas prices in 2022 demonstrates the ability for these costs - and thus avoided energy costs - to increase dramatically.

Avoided Transmission Capacity 

Avoided transmission costs are another category which has become interesting and more dynamic in recent years. As the Energy Department summarizes, “the main determinants of the need for transmission expansion identified include grid reliability and resilience, congestion relief, new generation resources interconnection, and load growth accommodation.” All of these factors have seen major changes in recent years. 

Load growth and the transition to low/no-carbon generation has ramped up significantly in recent years. Meanwhile, extreme weather events, such as Winter Storms Elliot and Uri, are emphasizing the need for transmission buildouts to serve reliability and resilience needs. The U.S. grid needs 64% more within-region transmission and 412% more interregional transmission by 2035 to keep up with decarbonization and demand growth, according to the Department of Energy.

But transmission capacity has become harder to build. Challenges related to transmission planning, cost allocation, siting and permitting, as well as ratemaking and incentives all impede the rapid buildout of transmission lines. When transmission buildout slows, existing infrastructures become constrained. Therefore, there is almost certainly a higher value of avoiding transmission capacity today than there has in the recent past. 

Avoided Distribution Capacity

Distribution is the most complex avoided cost to quantify, but it’s become too significant to ignore.

As complicated as avoided generation capacity is to estimate, an analyst still only must consider one system – one fleet of generators that supply energy compared to one aggregated customer base that demands it. On the other hand, an analysis of distribution avoided costs must consider hundreds or thousands of distribution circuits, each of which may peak at different times. Furthermore, an analysis of avoided costs for a distribution primary system will be different from a similar analysis of the distribution secondary system. Finally, differences in cost due to geography, undergrounding of lines, and the density of customers can result in vastly different costs based on location. 

Getting a clear picture of the load and cost drivers on a particular distribution circuit is important because the impact of new technologies, like EVs, can be dramatic. A level 2 charger for an EV can often triple a household’s peak demand. It only takes a few EVs charging in a neighborhood at the same time to overload distribution system assets. 

A February 2024 NREL paper found that “utilities are experiencing lead times for transformers up to 2 years (a 3x increase on pre-2022 lead times) and reporting price increases by as much as 4-9 times in the past 3 years.” The NREL report cites the causes of this price increase as: “aging infrastructure, electrification and massive growth in electricity demand, increased failure due to extreme weather events, and proactive utility replacement programs.” 

Most avoided cost models don’t account for the increased need for increased distribution assets. Instead, models rely on historical cost data and simplified calculations using a single coincident or non-coincident peak will likely understate the value of avoiding distribution capacity. That’s a mistake. While it can be challenging to quantify these impacts, transformer shortages and distribution constraints keep getting worse, programs that prevent overloads, like managed charging, will be increasingly valuable.

The combination of new technologies, complex systems, and rising costs all make the subject of distribution avoided cost a fruitful one for additional research and consideration. 

How This Comes Together for Managed Charging

Avoided costs are the foundation of managed charging. A well-designed passive managed charging program—one that uses time-of-use (TOU) rates or incentives—can shift EV load away from peak periods, avoiding generation and transmission costs. If the value of avoided generation capacity increases then it will also increase the value of passive managed charging programs.

However, passive managed charging programs still leave value on the table. TOU periods are set years in advance and don’t reflect real-time grid conditions. Some days have higher-than-expected demand. A static TOU rate or passive managed charging program cannot capture that value.The first chart in the figure below shows what EV load shapes look like under a passive managed charging. When the off-peak period begins, EVs start charging en masse, creating a steep demand spike that gradually declines.

The chart below shows an active managed charging program that aligns with dynamic cost signals, in this case from PJM (the world’s largest electricity market). Instead of a flood of EVs charging the moment off-peak rates begin, charging is more concentrated in the lowest-cost hours. This maximizes the value of avoided generation and transmission capacity. It also increases avoided energy costs since it shifts the EV load to be served by the most efficient generation units, which use less fuel per kWh of energy.

But even with active price signals, EV charging can still create unintended peaks. Notice how both passive and active approaches in the figure result in concentrated demand surges. If all EVs receive the same price signal at the same time, they will respond simultaneously, causing localized demand spikes. These peaks stress distribution infrastructure and increase costs.

Programs, such as WeaveGrid’s Distribution-Integrated Smart Charging Orchestration (“DISCO”) optimize EV charging to not only reflect precise generation capacity and energy costs, but also to protect distribution assets. Local peaks are minimized, reducing overloading assets and ultimately resulting in distribution avoided costs. Critically, DISCO optimizes across generation capacity, generation energy, and distribution energy to maximize the avoided costs and thus the benefits for all ratepayers. 

Already, DISCO is proving to be cost effective. Baltimore Gas and Electric measured the cost effectiveness of their Smart Charge Management and EV-TOU program at 1.58 under Maryland’s Jurisdiction Specific Cost Test (a score above 1 is considered cost effective). If avoided costs are revised and increased, then DISCO should correspondingly achieve even higher cost effectiveness scores.

Conclusion

The debate over avoided costs is not new, but the conditions shaping them are changing fast. These are core questions for the future of the grid. If we undervalue the costs we’re avoiding, we end up spending more in the long run on power plants, expensive transmission upgrades, and overloaded transformers.