Energy Efficiency: The Latest Utility Fudge Factor

Your insider's guide to the latest number-fudging Tomfoolery by Investor-Owned Utilities

A quick programming note:

We picked up a few projects on short deadlines last week and didn’t have time to get an article up. We apologize for any disappointment, but we feel like we've got a banger for you today.

This week, we’re here to clue you into one of the latest new tricks that vertically integrated Investor-Owned Utilities (IOUs) are using to fudge the numbers in their Integrated Resource Plans (IRPs) to make them look more reliable and lower cost than they truly are: energy efficiency.

The Setup

The ploy goes like this. As we have detailed in several previous articles, IOUs are looking to maximize their profits by prematurely retiring their existing, depreciated assets and building as much new infrastructure as possible. In this respect, the fact that wind and solar are unreliable is a major bonus because it means they have to build a lot of these resources to replace the reliable plants and keep up with accelerating demand growth.

The problem? The utilities can’t admit just how much new expensive infrastructure will be needed in their IRPs, or they will not be able to pretend that they are the “least cost” way of meeting their future electricity demand. If the plan is not “least cost,” it runs the risk of being denied by state regulators, and the utility won’t be able to rake in that sweet, sweet rate of return on its rate base.

The Punchline

The solution to the problem? Artificially reduce, at least on paper, the amount of new capacity you need to meet demand by pretending you’ll implement energy efficiency measures to reduce demand in the future.

Mix this together with the intermittent resources in your IRP and voila! You’ve successfully hidden the ball from regulators as to how your IRP threatens reliability and increases costs.

The Case Study: The Arizona Public Service IRP

To demonstrate this dynamic in action, we revisit the Arizona Public Service (APS) IRP we discussed in our last article.

As we mentioned last week, the APS “Preferred Plan” consists of building nearly 20,000 megawatts (MW) of new capacity to replace retiring capacity and meet electricity demand growth, which only increases by 5,000 MW, rising from 8,100 MW in 2023 to 13,100 MW in 2038 (See Figure 1).

Figure 1. So much new capacity to meet load growth, so little of it reliable.

The total installed capacity on the APS grid increases by more than 2.5 times from 2023 to 2038 (See Figure 2). For energy efficiency and demand response, these resources grow from 360 MW in 2023 to 1,697 MW by 2038, which means these “resources” increase their share of the system’s peak demand from just 4.4 percent in 2023 to a whopping 21 percent of peak demand in 2038.

Figure 2. The amount of dispatchable thermal capacity on the system remains about the same as in 2023, despite a 62 percent increase in electricity demand.

The rising share of energy efficiency and demand response means these resources will no longer be a “nice to have” emergency measure to give the grid operator more room to maneuver during times of high system stress; they will be an indispensable part of keeping the lights on during periods of low wind and solar output.

What Does Energy Efficiency Even Mean?

The energy efficiency estimates used by APS stem from a 79-page report prepared for the utility by Guidehouse. In the report, Guidehouse identified a series of measures APS could use to reduce system demand.

For residential customers, these programs frequently take the form of allowing the utility company to control household devices and appliances like thermostats, air conditioners, water heaters, and pool pumps in exchange for up to a $50 annual bill reduction. What a deal!

In the future, APS believes these programs could also include having the company control when electric vehicles can be charged and potentially draining the batteries of charged cars during periods of high electricity demand and low supply from intermittent sources. This transfer of power from EVs to the grid is frequently called vehicle-to-grid (V2G).

Guidehouse estimated the “achievable” one-hour peak demand reduction available to the utility in 2038 would be nearly 1,500 MW, with most of these “resources” consisting of residents allowing the utility to control their thermostats and other devices (darker green), and allowing the utility to drain the batteries in their electric cars (Figure 3).

Figure 3. A graphic from Guidehouse showing the *totally realistic* achievable demand reduction.

But what if people don’t want to swelter in their homes during heatwaves or allow the power company to siphon their electric vehicle’s “gas tank” when the sun isn’t shining? Will there be enough power to meet demand, or will Arizonans be left in a situation where there is not a square to spare?

The Blackouts

When we first evaluated the reliability of the APS IRP without including energy efficiency and demand response resources (we sometimes don’t include these “resources” in our capacity stack because they’re not really generators or storage devices), we found massive rolling blackouts on the APS system. The blackouts were so big we knew we had to be missing something, which led us on the hunt for our utility fudge factor.

At first, we looked for imports from neighboring regions, but this wasn’t a big factor. Then we realized APS acts as if energy efficiency and demand response are basically dispatchable peaker plants in their modeling.

If these resources, which constitute 21 percent of peak system demand in APS’s modeling, don’t show up as APS is planning on, the lights won’t stay on.

The graph below shows electricity demand and supply by type for a hypothetical period in the future stretching from July 21, 2038, through July 23, 2038. As you can see, wind, solar, battery storage, nuclear, natural gas power plants, and energy efficiency are able to meet electricity demand, shown in the black line, despite a wind drought impacting wind generation.

Figure 4. Hourly electricity demand in 2038 is calculated by scaling up 2022 hourly demand from EIA to meet the projected peak demand in the APS Preferred Plan. Hourly wind and solar generation is determined by applying the 2022 capacity factor of each resource to their respective installed capacities in 2038 per the IRP.

However, the reliability of the grid during this stretch is reliant upon APS meeting its target of 1,697 MW of energy efficiency, shown in dark blue in the graph. If the company fails to reach this level of energy efficiency, the system will not be able to reliably meet demand.

For example, we ran multiple sensitivity analyses and found that if only 50 percent of APS’s modeled energy efficiency comes to fruition, then the system would have multiple rolling blackouts during the same week shown above, with the largest shortfall weighing in at 1,217 MW, which would constitute 12 percent of power demand at the time of the blackout.

Figure 5. If energy efficiency is only half as prevalent as APS is estimating, the lights won’t stay on.

The situation is even worse if we assume only 25 percent (or 424.25 MW) of the energy efficiency resources assumed in the IRP come online. The size of the maximum shortfalls in this scenario is 3,701 MW occurring at 11:00 P.M. on July 21, 2038, after the sun sets and there is no more storage to back up the wind and solar fleet. This shortfall constitutes 33 percent of the total demand during this time period, meaning hundreds of thousands of families and businesses would be without power.

Figure 6. The less fudge in the numbers, the bigger the blackouts.

Is it realistic for APS to assume their energy efficiency “capacity” will grow from a mere 147 MW of energy efficiency available to meet peak demand in 2023 to 1,697 MW by 2038? We think not.

For starters, APS has been implementing energy efficiency programs since 2005 and has only managed 147 MW of peak demand reduction thus far. Furthermore, we doubt many Arizonans will turn the control of their thermostats over to APS for only $50 per year. Lastly, the Guidehouse study indicates that not all of the energy efficiency capacity assumed in the IRP is cost-effective. If it’s not cost-effective, then it’s probably not going to happen.

As a result, we strongly suspect that the energy efficiency in the APS IRP is less about reducing peak demand and more about hiding the ball from regulators about the true reliability and cost impacts of their Preferred Plan. They can submit this IRP, start spending capital, and then cover their posteriors when they file the update to their IRP in three years.

The Utah Way

In neighboring Utah, however, lawmakers have wisely decided that energy efficiency and demand response are not capacity resources; they are resources of last resort.

As a result, they passed House Bill 201 in 2025 that prohibits utilities from fudging their capacity numbers by counting energy efficiency and other demand-side management as dispatchable capacity. Instead, they are treated as voluntary ways to reduce demand on the load side, meaning anticipated demand reductions cannot be counted toward generation capacity.

This change is beneficial because it will force Utah utilities like PacifiCorp to actually model the reliability or cost implications of the resource plans they are proposing, rather than hiding the ball, building the infrastructure, billing the ratepayer, and repeating the cycle all over again in three years.

Conclusion

Energy efficiency can be a helpful tool for reducing electricity demand during periods of high stress on the system, but there is considerably more uncertainty surrounding the adoption of energy efficiency measures than in building supply-side resources.

However, utilities should not be allowed to use energy efficiency resources to hide the fact that their grids are not reliable, and that making them reliable while also indulging their Net Zero bilking of ratepayers ESG goals will cost more money than regulators may be willing to approve.

Report: Congress’ Mega-Bill Puts 600 GW of Clean Energy Capacity at Risk by Cleanview: You. Love. To. See. It.

EPA Announces Coal Power Plant Deregulation by

Nathan Hammer

. Hammertime lays out the facts about how pollution control equipment has drastically reduced emissions of traditional pollutants from coal plants.

One Coal Plant Provides Nearly As Much Power as All of Michigan’s Wind Machines by Ken Braun at

We Bicker for Food

. This one is pretty self-explanatory.

How California Regulated Itself Into an Energy Crisis by Lauren Teixeira. California lawmakers enter the “Find Out” phase after years of layering regulations on fuel producers.