Heat Pump Water Heater (HPWH) Interaction Factor

[jmaguire1]

There's been some work that shows in lab homes (and as best as we can tell in field installations, although it's really hard to tease out) 100% of the cooling from a heat pump water heater isn't "seen" by the thermostat, so the HVAC doesn't have to make up for/benefit from the HPWH running if the heating/cooling HVAC is running respectively. The way we generally characterize this is an "interaction factor" between the HPWH and the thermostat/HVAC.

PNNL has done some work on this with their lab home where they were able to see interaction factors < 1 for several installation locations. Interaction factors > 1 are technically possible if you point the HPWH exhaust directly at the thermostat for example, but extremely unlikely to happen in practice unless the thermostat was extremely close to where the HPWH is installed. When Ben Larson and I have talked about this our best guess is 0.75 if there's no other information available, but what you'd really want is to know something about the distance between the HPWH and the thermostat ("thermal distance" would be even better, in that there are things like closed doors that could be in the way that affect airflow beside just physical proximity). However, there's not good correlation between this distance and interaction factor, this isn't data that's usually gathered during an audit, and it is different from how we treat every other internal gain (say cooking) in that it breaks the energy balance.

Mechanically this isn't hard to do, we even have an EMS program to help this out. We'd just have to add interaction factor as an input, and then use it to multiply the sensible (and latent, although less important because SHR is so high for HPWHs) here: https://github.com/NREL/OpenStudio-HPXML/blob/5141bcd2f2a2c607b72dde834e2aaf79cc617d54/HPXMLtoOpenStudio/resources/waterheater.rb#L1290.

This EMS program is there to make interaction factors (and ducting) easier for HPWHs, so a lot of the hard work is done. We'd still have to add this to HPXML (or use extension elements) and add test cases for this.

We now have some information about whether the HPWH is installed in a confined space without proper ventilation: #2086. So this reminded me that we had this as a potential outstanding issue. ResStock might want this at some point? I've had some discussions but not sure if this is a priority right now, so mostly just getting this tracked.

[jmaguire1]

See #2086 (comment) if we want to try to account for element usage in confined spaces if we tackle this. It'd be more technically correct to cover both at the same time, although interaction factor is by far the bigger issue in terms of amount of housing stock affected. The linked issue is really only relevant if you install a HPWH in a very small closet with no airflow, which we'd generally argue no one should do.