With their ability to use energy flexibly, heat pump water heaters can “charge” during off-peak times with cheaper, cleaner electricity and act as small thermal batteries. Doing so not only saves homeowners money, but it also reduces grid congestion and emissions.

In a perfect world, HPWHs would turn on each day to absorb inexpensive renewable energy – mitigating the infamous “duck curve” effect (https://tinyurl.com/2cwj2uyn) – then coast through peak demand periods without needing to turn on compressors or resistance elements. However, achieving this harmonious load curve with a fleet of HPWHs is incredibly difficult. 

To better understand the barriers to HPWH load-shifting at scale, the Advanced Water Heating Initiative compiled a comprehensive repository of HPWH load flexibility and demand response studies. Each row in the link below contains a link to a field or lab study report, as well as synthesized findings and key statistics. We hope that this resource will help find fast answers to complicated questions and avoid duplicative work. You can access the repository here: (https://tinyurl.com/44u3m9rb).

Let’s discuss some of some key challenges and learnings that arose time and again throughout these reports (including in our own 120-volt HPWH load-shifting study).

Module connectivity can be a challenge – especially with Wi-Fi 

Most load-shifting pilots we documented used EcoPort modules to enable load-shifting signals (“Advanced Load Up,” “Load Up,” “Shed,” etc.) from the CTA-2045 standard. Several study teams reported connectivity challenges, especially with Wi-Fi modules. A router replacement or password change can result in the EcoPort losing connection and failing to receive signals. Connecting the EcoPort to the water heater can be confusing, and many homeowners simply won’t bother to troubleshoot the process in the event of disconnection. 

Cellular EcoPort modules offer a more reliable connection at the fleet level, especially for long duration programs. However, it’s important to keep in mind certain walls or closets can block cellular signal. For instance, an SDG&E load-shifting study (https://tinyurl.com/mrx3xdv5) found that the cellular communications modules installed on electric resistance water heaters failed to establish a reliable signal. 

Advanced Load Up needs a more consistent rollout 

One of the signals available in the CTA-2045 lexicon is Advanced Load Up (https://tinyurl.com/3xpzkeft), which typically directs a HPWH to raise its tank temperature above the user-specified setpoint – usually to 135 degrees or 140 degrees – depending on manufacturer controls.

High temperature storage is critical for load-shifting, as it extends the time a water heater can coast through a higher price or peak period without drawing energy. Basic Load Up directs a water heater to heat to setpoint, often defaulting to 120 degrees. While Load Up is proven to be effective at partially shifting load away from peak periods, the HPWH will often activate part way through a longer (three-plus hour) shed period. 

Researchers haven’t conducted much real-world ALU testing for two reasons: Firstly, ALU requires a mixing valve, which most 240V HPWHs do not come with; Secondly, it’s only available in CTA-2045 version b. Although CTA-2045b was released in late-2020, HPWH manufacturers have not all moved over to the updated standard. In studies conducted as recently as 2024, HPWHs in the field did not have access to the ALU command. 

Recent results from PG&E’s WatterSaver program shows promising results, when HPWHs properly conform to the ALU signal. HPWHs given the ALU signal shifted roughly 500 Wh away from the afternoon peak, vs only 300 Wh for those given the basic Load Up signal. Even more importantly, the ALU group consumed far more energy in the early afternoon (during the ALU period), essentially using up abundant solar energy that might otherwise be curtailed. These findings echo earlier lab testing (https://tinyurl.com/5cxz8n3v) done in 2021 by the FSEC Research Center in Florida, in which a CTA-2045 b compatible HPWH shaved peak demand by 510 watts.  

HPWH load-shifting savings are tricky to quantify 

Load-shifting isn’t always free. For more sophisticated utility programs, a universal communications module or API is usually required. The former usually costs $150 for the hardware on top of a one-time licensing fee, and manufacturer APIs can exceed $10/month per device. The question then is whether that cost is worth it, and for whom. 

Based on our repository, HPWHs can typically shift between 200 and 700 watt-hours away from a three- to five-hour high demand window. For that load shift to save a consumer money, they must be on a time-of-use rate. TOU rates commonly have a 10-50 cents difference between on-peak and off-peak pricing. Using Southern California Edison’s TOU-D-5-8PM (https://tinyurl.com/56vszkp6) rate plan, customers can save 19 cents a day by shifting 500 Wh from on-peak to off-peak hours. That might not sound like much, but it equates to nearly $70 a year, or potentially $700 over the life of a HPWH.  

Homeowners aren’t the only ones financially benefiting from load-shifting. Wholesale electricity prices can be extremely volatile, and those costs are borne by utilities. Monthly average wholesale prices per megawatt-hour typically range from $30-$80, depending on the region.

However, an extreme weather event can cause prices to skyrocket, such as in Texas in 2023 when electricity prices briefly hit $2,500 per MWh during a heatwave.

So, while load-shifting savings can be hard to evaluate, the value proposition is clearly there. And once ALU is rolled out and connectivity improves, that value will only continue to grow.