Demand Side Analytics

Heat Pump Performance in Maine Homes

Heat pumps are increasingly seen as a cornerstone of building electrification. When designed and used effectively, they can dramatically reduce fossil fuel consumption, lower emissions, and improve household comfort and economics, even in cold climates like Maine. But as adoption accelerates, utilities and policymakers face a critical question: are customers using heat pumps as their primary source of heat, and how do these systems perform in real-world conditions?

The answer matters. If heat pumps sit idle during cold winters or are used as supplemental heating equipment, the expected fuel savings and emissions reductions are not realized. If, however, they reliably displace fossil fuels when it matters most, they can become a powerful and predictable resource for decarbonization and grid planning.

Efficiency Maine is one of the most aggressive promoters of cold climate air source heat pumps in the country. In fall 2023, following a series of heat pump utilization studies commissioned by Efficiency Maine and performed by Demand Side Analytics (DSA) and Ridgeline Energy Analytics (Ridgeline), Efficiency Maine redesigned its residential heat pump program strategy for individual heat pump incentives to a “whole home” design. After two winters using the new approach, Efficiency Maine again contracted DSA and Ridgeline in 2025 to evaluate the performance of its Whole-Home Heat Pump (WHHP) Rebate Program. The evaluation combined three complementary perspectives:

  • Advanced Metering Infrastructure (AMI) data from about 1,000 homes were used to measure how electricity use changes after heat pump installation. The full report is available online here.
  • End-use metering data from 78 homes were used to isolate heat pump electricity consumption from other household loads. The full report is available online here.
  • A survey of 1286 WHHP participants to understand customer behavior, comfort, confidence, and decision-making. The full report is available online here.

The goal was straightforward but critical: determine whether Efficiency Maine’s shift from its legacy program design allowing rebates for supplemental heat pumps to its new requirement that heat pumps be designed and sized to heat the whole home improved the system economics, and changed how Mainers heat their homes, and how that change shows up on the electric grid.

Evaluation Approach

The evaluation followed a structured, evidence-based framework to connect customer experience with system-level impacts.

Step 1: Measure Real-World Load Impacts
AMI data were used to compare electricity usage before and after heat pump installation, with a focus on winter months and peak demand periods. This made it possible to quantify how whole-home heat pumps affect system load under real operating conditions.

Step 2: Isolate Heat Pump Behavior
A sample of households was selected for the installation of metering equipment to investigate suspected underutilization of heat pumps. Metering data were used to separate heat pump electricity consumption from other household loads. This step was essential for understanding when heat pumps were running, how intensively they were used, their efficiency, and whether they were meaningfully contributing to winter heating demand.

Step 3: Ask Customers Directly
A WHHP participant survey provided insights that AMI data alone cannot capture. Participants were asked about their heating behavior, comfort levels, confidence in their systems, and interactions with installers.

Step 4: Compare Against a Legacy Supplemental Program
Finally, results were benchmarked against findings from Maine’s earlier legacy supplemental heat pump rebate program, which did not require heat pump systems to be the primary heating source. (The new program design requires that the heat pump system meet at least 80% of a home’s peak heating load). This comparison made it possible to isolate the impact of the program design change.

Together, these steps linked what customers said with what their meters showed, creating a more complete picture of how whole-home heat pumps perform in practice.

Key Finding #1: Heat pump electricity use increased in expected ways.

On average, homes participating in Efficiency Maine’s WHHP program use 4,904 kWh annually for heating with heat pumps, delivering approximately 52 MMBtu of heat. Compared to the earlier generations of Efficiency Maine’s heat pump rebate program, which had smaller incentive amounts and did not establish any minimum requirements for design load capacity, total kWh consumption for heating with heat pumps nearly doubles in the new WHHP program (Figure 1).

Under the new WHHP program, electricity consumed per kBTU of rated heat pump capacity rises from approximately 109 kWh/kBtu_Rated@47 to 143 kWh/kBtu_Rated@47 compared to the legacy program design. Alongside improved COPs, the higher electrical usage by the heat pump systems will improve electric system utilization and place downward pressure on electricity distribution rates. The increased winter electric consumption is also a direct proxy for reduced fossil fuel usage, which is largely fuel oil given Maine’s limited natural gas network.

For homes that were included in the metering sample, we compared our AMI-based estimate of annual heat pump heating kWh with the annualized meter-based estimate. We found strong alignment between the two estimates. The average AMI-based estimate of annual heat pump heating kWh was 96% of the average meter-based estimate.

Figure 1: Heating Season Average Household Daily kWh by Day of Year and Period

Key finding #2: Whole-home program design drove higher utilization and greater fossil fuel displacement.

Both metered data and survey responses show that WHHP program participants rely on heat pumps far more frequently than participants in the legacy supplemental program on typical winter days and during very cold conditions. Reported use of oil, propane, and other fossil fuels declined sharply after installation.

AMI and metering results indicate that heat pumps in the WHHP program provided over 70% of total heating needs, with wood heat accounting for roughly 10% and fossil fuels making up the remaining 19%.

By requiring heat pump systems to serve as a home’s primary heating source (meeting at least 80% of the home’s peak heating load), the WHHP program encouraged installations sized to meet real heating needs. As a result, systems operated closer to their full capacity instead of sitting idle, leading to significantly greater displacement of oil, propane, natural gas, wood, and kerosene. For households previously reliant on fossil fuels, this translated into substantial cost savings and emissions reductions.

Figure 2: All heating sources in 2023 Legacy Supplemental Rebate and 2025 WHHP Rebate

Key finding #3: Better system design leads to improved comfort and higher utilization.
Survey results show that WHHP program participants experienced broader heat distribution, higher satisfaction, and a greater likelihood of recommending the program. Homes where installed heat pump capacity met or exceeded the home’s design heating load were significantly more likely to rely on heat pumps as their primary heating source.

This reinforces a key lesson: comfort and performance are inseparable. Systems that keep homes warm throughout the winter are the systems customers use, and those are the systems that deliver the largest grid and emissions benefits.

Efficiency Maine’s Whole-Home Heat Pump Rebate Program demonstrates that electrification can deliver real, measurable benefits when programs are designed with performance in mind. By pairing AMI analytics, end-use metering, and survey responses, this evaluation shows not just what was installed, but how it is used, when it matters, and why it works.

Pennsylvania Transmission and Distribution (T&D) Avoided Cost Study

Demand Side Analytics (DSA) conducted a transmission and distribution (T&D) avoided cost study in Pennsylvania. The focus of the study is on quantifying the change in T&D costs associated with an increase or decrease of kW coincident with location-specific peaks. It employs methodologies that are novel for Pennsylvania but have been applied and approved in New York and California for load forecasting and distributed energy resource (DER) valuation. The full study is available online here.

A vital role of the Electric Distribution Company (EDC) is to ensure that regional electricity supply makes its way to homes and businesses safely, reliably, and cost-effectively. By projecting future demand and reinforcing the local transmission and distribution network so that sufficient capacity is available to meet local needs as they change over time, costly outages are avoided.

What are the objectives of the study?

The study was designed to meet the following objectives:

  • Analyze load patterns, excess capacity, load growth rates, and the magnitude of expected infrastructure investments at a local level
  • Model location-specific forecasts of growth with uncertainty
  • Quantify the probability of potential need for infrastructure upgrades at specific locations
  • Calculate local avoided distribution costs by year and location
  • Identify beneficial locations for demand reductions

Which methods did we use?

The deferral value approach focuses on quantifying the value of load relief on ratepayer costs (i.e. revenue requirements). It effectively compares revenue requirements with and without load relief. While infrastructure upgrades can be temporarily avoided or deferred via load relief, they cannot be avoided indefinitely because equipment eventually ages and needs to be replaced. The marginal cost of service study approach quantifies the supply cost of additional distribution or transmission capacity on the system. At the simplest level, it involves classifying infrastructure investments as growth related or not and dividing the costs of those investments by the incremental transmission or distribution capacity added. The approach uses the cost of adding additional transmission capacity to the system as a proxy for the cost avoided by reducing peak demand.

Figure 1: T&D Avoided Costs Methods Considered

What did we do?

Figure 2 describes the main steps in developing location-specific avoided distribution costs using probabilistic methods. These steps help identify the magnitude of reductions at the right location at the right time and right season to delay upgrades. The process was implemented for each feeder and substation (transformers or terminals if applicable) that had a valid growth rate and operating limit, then layered two levels to get the distribution avoided cost for each site. For system-wide values, the estimates consider the likelihood that reductions would be in locations with or without value due to random chance. We emphasize that system-wide value is a load-weighted average of areas where reductions do or do not lead to deferral of distribution investments.

Figure 2: Key Steps in Estimating Location Specific Avoided Costs

What are the results?

The avoided cost of transmission capacity estimates simply increases with inflation and escalation. The avoided cost of distribution capacity values changes at varying rates based on the outcomes of the probabilistic deferral analysis methodology.

While the final study outputs are territory-wide average values for each EDC, the granular forecasts are useful for identifying locations and timing when demand reductions or injections of distributed generation are beneficial. Figure 3 shows the total deferral value by local systems. Shades of blue indicate relatively low deferral value while orange and red tones indicate high deferral value. The values range from a lower bound of $100 to a maximum of $200 in 2026 nominal dollar. To calculate the total deferral value, we aggregated the deferral value at feeder and substation levels, then incorporated the deferral value of transmission. A key outcome of the study was to highlight the fact that the avoided T&D costs associated with peak load relief vary widely within each EDC territory.

The value of avoided T&D costs associated with an increment or decrement of peak load is a key component of benefit-cost analyses. In practice, T&D capital costs resources are concentrated in pockets that are experiencing growth but lack the capacity to accommodate additional growth. Most utilities have a mix of areas where loads are growing and areas where loads are declining, which may or may not overlap with highly loaded components. In locations with excess distribution capacity or where local peak demand is declining, the potential to avoid T&D costs is minimal. In areas where a large, growth‐related investment is imminent, the avoided T&D costs from reducing peak demand are much higher.

Figure 3: Heat Map of Total Deferral Value ($2026/kW-year)

What did we find?

    Load growth varies by location. Some pockets are experiencing load growth, and some are experiencing load decreases. We received granular growth rates for PECO, PPL, and FirstEnergy. In each EDC territory, growth trends varied by location. As a result, growth-related T&D investments are required even when overall EDC loads are flat or declining.

    The T&D avoided costs are concentrated in locations that are more heavily loaded. A key component of distribution planning is the loading factor: the weather-normalized peak demand divided by the operating limit. Not surprisingly, avoided costs are concentrated in more highly loaded locations. Conversely, locations with ample capacity to accommodate additional loads had lower avoided T&D costs.

    Individual locations are generally winter or summer peaking, not both. Most distribution locations – feeders, transformers, substations – can be classified as winter or summer peaking. Few feeders are dual peaking. The implication is that the avoidable T&D cost for a specific location is concentrated in the summer or winter, but not both.

    Resources that deliver load relief at the right location, in the right season, and at the right hours are more valuable. The same energy efficiency resource can deliver different T&D benefits at two locations based on how well it coincides with the local peak load. To illustrate, a more efficient air conditioner does not provide T&D load relief on a winter-peaking substation but does so on a summer-peaking substation. Likewise, measures with load shapes that better coincide with the need for load relief are more valuable.

    Lump loads are a key driver of distribution upgrades. Lump loads are simply new, large loads. They vary widely in size, and it can be difficult to predict in advance when and where they will show up. When they are built, they often trigger distribution and even transmission upgrades. Reducing demand via EE&C program efforts can create room for additional loads and help avoid upgrades due to smaller lump-load projects.

The Pennsylvania PUC leveraged the results of the study for its 2026 TRC Test Tentative Order. The TRC Test Order provides utilities with directions for calculating avoided costs and performing benefit-cost analysis when planning for Phase V of Act 129 programs. The avoided T&D values will play an important role in the Phase V DR Potential Study, which DSA is currently working on.