Methodology for Household Savings Calculations

The costs of running a residential heat pump are dominated by its coefficient of performance (COP), that is, the amount of electricity required to deliver a unit of heat (no matter whether in the form of a heated space, hot shower, or a dry shirt). As described above, the COP is a function of ambient dry-bulb temperature, often reaching above 5.0 at 50°F and down to 1.5 at -20°F. Because of this temperature dependence, to calculate annual electricity bills we must look at every hour of the year individually, at every location in the U.S.

To do this, we obtain hourly temperature profiles on a 0.5 x 0.625 degree latitude-longitude grid from the NASA MERRA2 dataset, an incredible collection of meteorological measurements including temperature, humidity, windspeed, incident radiation, and many more dating back to 1980. We then correlate these hourly temperature profiles with hourly demand profiles to capture when households run their heating equipment. These hourly load profiles are available for every location in the typical meteorological year dataset (TMY3) in the U.S. from the National Renewable Energy Laboratory (NREL). These two hourly series are used together with manufacturer-published performance curves to determine how much electricity is required to run the heat pump at every hour of the year. Critically, these performance curves must include parasitic effects, like the heat exchanger defrost cycle and other system losses.

For a given location, space heating heat pump models and sizes are selected by evaluating the number of hours with any unmet heating demand. (This is analogous to how heat pumps are sized during installation.) The small remaining unmet heating capacity is assumed to be supplied by backup electric resistance heat with COP=1.0. We then add up each of these hourly contributions to calculate total electricity use. Consumption for other fuels is obtained from the U.S. Energy Information Administration's (EIA's) 2015 residential energy consumption survey. These fuel consumptions are converted to expenditures using local residential energy price data, obtained from the EIA's application programming interface (API).

For water heating heat pumps, we use the published uniform efficiency factor (UEF) rating to calculate baseline performance and add an interaction factor to account for the added load placed on the space heating system when a water heater is run inside a conditioned space.

Critically, we classify a household as "in the money" if it has at least one electrification project that saves money compared to current bills. In many cases, this electrification project is water heating. Savings are evaluated against an estimation of existing appliance stock. Savings due to improved air-conditioning efficiency of a new heat pump are not yet included, which will add to the amount of households saving money. Dehumidification and air-conditioning effects of heat pump water heaters are not included, which can also add to the savings.