The takeaway from this post is a single estimate: The modest shift from fossil heating to electric appliances from 1972 to 2009 accounted for only approximately 1.2 kbtu/sqft/year reduction in fossil EUI in 2009 — a relatively small contribution to the overall drop in fossil EUI of 77.9 kbtu/sqft/year from 1972 to 2009. Table 1 shows the basis of this estimate. As discussed below, it does not appear that there was statistically significant change in the electrification of appliances other than space heating.
Table 1: Contribution of electrification to the decline in fossil fuel EUI in Massachusetts homes, 1972-2009
|1||1972 share of homes with electric appliances for primary space heating||4%|
|2||2009 share of homes with electric appliances for primary space heating||14%|
|3||1972-2009 home share increase: (2)-(1)||10%|
|4||1972-2009 home share increase as % of 2009 share: (3)/(2)||71%|
|5||2009 heating electricity use as % of total electricity used (from RECS)||7%|
|6||Total 2009 residential electricity use (TeraBTU from EIA)||66.4|
|7||2009 electricity use attributable to appliance increased share:(4)*(5)*(6)||3.32|
|8||Total 2009 site energy use (TeraBTU from EIA)||303.9|
|9||2009 electricity use attributable to appliance increased share as % of total 2009 site energy use: (7)/(8)||1.1%|
|10||Total 2009 Massachusetts residential site EUI |
(kbtu/sqft/year from EIA and previous square footage analysis)
|11||Estimated 2009 fossil fuel appliance efficiency relative to electric efficiency||69%|
|12||Fossil fuel EUI decline attributable to electrification (kbtu/sqft/year): |
(9) * (10) / (11).
Methods — see also attached spreadsheet.
All of the major fossil energy uses in homes can be electrified — space heating, water heating, cooking, and clothes drying. There are two ways that a shift to electric appliances could occur — conversions in existing homes or greater use in new construction or. Our estimates below do not depend on whether the shift is due to conversion or new construction, but it is worth noting that during the period we are studying, 1972 to 2009, conversions from existing fossil appliances to new electric appliances were not likely a material factor. In that period electricity was never less than 110% more expensive than the most expensive fossil fuel and was typically 250% more expensive. See the “Prices” tab in the attached spreadsheet. So, on top of capital expenditures for conversion, an existing fossil-heated home would face a significant operating cost increase if they converted to resistance heat. The same basic economics would apply for conversion of any other electric home appliance. Moreover, resistance electric appliances were more environmentally more harmful than fossil appliances given the fossil-intensive New England electric grid, so few would be motivated to convert for environmental reasons. Cold climate heat pumps had not yet emerged.
We have to combine data from several different sources to estimate the contribution of electrification to the historical decline in fossil energy use intensity. The sources are detailed in the attached spreadsheet. The basic reasoning in Table 1 above is as follows:
- Estimate the share of the particular appliance type — space heating, water heating, other — that was powered by electricity in approximately 1972. Here one is using saturation (prevalence) as share — the denominator is the total number of homes.
- Estimate the share in 2009.
- Compute the increase in share from 1972 to 2009: (2) – (1)
- Compute the increase as a % of 2009 share: (3)/(2)
- Estimate 2009 electricity use for appliance as percentage of total residential electricity used (at the population level, so reflecting both the per appliance electricity use and the saturation/prevalence of the electrified appliance). Here the denominator is energy in BTUs (one kilowatt-hour = 3412 BTUs).
- Estimate total residential electricity use in 2009. (Note that steps 5 and 6 interact and may go in either order depending on the data source.)
- Multiply the prevalence share increase by the 2009 use percentage and then by total residential electrical energy use in 2009 to derive 2009 electricity use attributable to appliance increased share: (4) * (5) * (6)
- Estimate total residential site energy use.
- Compute the 2009 electricity use attributable to appliance increased share as % of total 2009 site energy use: (7)/(8).
- Estimate total residential site EUI in 2009
- Efficiencies based on 2015 installed base estimates from Guidehouse national technology study for DOE without any discounting for progress since 2009 — since relative efficiency is the issue, progress from 2009 is less of an issue since likely all categories made progress and installed base changes only slowly anyway. For gas fired furnaces, the “North” efficiency was used. Otherwise the efficiency results are national averages and may slightly understate Massachusetts appliance efficiency, but again relative efficiency is the issue across furnace types so inaccuracy should be less. As one illustration of variation, the average fossil space heating efficiency for 2019 from the Baseline study appears to be 85 AFUE, as compared to the 82 AFUE for 2015 used here. See further discussion of burner efficiency below.
- Apply the electrification percentage to the EUI to compute the fossil EUI decease attributable to electrification — (9) * (10) / (11)
Space Heating Electrification
Electricity expanded as a primary heating source in Massachusetts, from 4% of homes in 1970 to 14% in 2009, likely due to new construction (Census data, see “E-heat-prev” tab of attached spreadsheet.) Electric baseboard heat is relatively inexpensive to install. Builders who are unconcerned about future operating costs — often builders of rental housing, where the tenants will bear the costs — may make an economically rational choice to install electric heat. In 2009, only 7% of owner-occupied homes relied on resistance heat, as compared to 26% of rental homes; rental homes accounted for 71% of all homes relying on electric heat.
Since rental units skew heavily towards apartments with smaller room counts, it follows that homes heated with electric heat are likely to be smaller homes. Consistent with that inference, electric heat accounted for a disproportionately small portion of population-level energy use for space heating. The Residential Energy Consumption Survey (“RECS”) (see Table HC6.8) estimated that 12% of Massachusetts homes had primary electric heat in 2009 (roughly consistent with the Census estimate of 14% above). However, RECS further estimated that electric space heating accounted for only 2.5% of total site energy consumption for space heating and only 1.5% of total site energy consumption (see computations in “2009 RECS” tab of attached spreadsheet). Those shares sound low, but they are consistent with the same shares estimated for the entire Northeast region (3.1%, 1.7%), a larger sample slice with a much lower sampling error. Moreover, the RECS data for 2009 are consistent with Table 2-1 in MassSave’s Residential Baseline Study (“Baseline Study”) for 2019 as to the population share of electricity use attributable to space heating — 7%. Additionally, according to the EIA consumption data referred to in charts 1 and 2, electricity was 22% of total residential energy consumption; applying the 7% to that 22%, one ends up again with electric space heating being 1.5% of total residential site energy consumption in 2009 — a tidy, perhaps lucky, concordance across RECS, the Baseline survey, and EIA statewide totals.
Just a wrinkle to explore: The 7% figure for heating use as a percentage of total population electricity use in the preceding paragraph includes heat pump lines in both surveys, but it excludes the line labeled “Central AC/heat pump” in the Baseline Table 2-1. That line accounts for 6% of annual electric consumption. It includes the 33% of homes that have central AC and within that 33% to include the 2% of homes that had a central heat pump providing both AC and heating (compare Table 2-1 to the saturation data on page 7 of the Baseline Study) We know from fhe Fuel Displacement Study that true whole home heat pumps were rare before 2020; in particular, from Table 4-3 of the study we know that the one central pump installation they did meter was more heavily used for cooling than for heating. It follows that the heating component of the 6% annual consumption in the “Central AC/Heat pump” line was likely less than 2/33 of 6%, in other words a negligible fraction in our rough computation.
Water heating electrification
The same sources and methods give us insight into water heating electrification. According to RECS, in Massachusetts in 2009, water heating with electricity was more common (20% of homes) than primary space heating with electricity (12% of homes), accounting for 9% of water heating energy use but still only 1.5% of all energy use. RECS puts the water heating share of electricity use at 6.8% — a little higher than the 2019 Baseline study which puts it at 5% (the Baseline also estimates a slightly lower share of homes, 16%). If one uses the RECS estimate for the water heating share of population electricity use, and follows the same computation as above for electric space heating, one gets the same tidy concordance across sources as to the electric water-heating share of total site energy consumption.
The 1980 Census of Housing, Detailed Housing Characteristics for Massachusetts volume (not accessible online, but available at the Boston Public Library), reports home prevalence of electric water heating in 1980 in Massachusetts at 19.3%, not statistically different from the 20% RECS estimate for 2009. We also have a report authored in the mid-80s by the Massachusetts of Energy Resources that estimates (at Table 14) that water heating electricity use declined on a consumption basis from 1976 to 1980. We have no basis for believing that the 1980 electric heating share was higher than the 1972 share — as expensive as oil and gas were, electricity was three or more times more expensive in the 70s (see Prices tab of attached spreadsheet). This suggests that the most reasonable assumption is that the electric water heating share of homes has been net flat from 1972 to 2009 at around 20%.
Laundry and cooking electrification
The other major fossil appliances that could be electrified are laundry dryers and ranges/ovens. The RECS usage report for 2009 does not break out these appliances as to electric use, bulking them in a large “other” category which includes many uses which do not run alternatively on fossil fuel (e.g., lighting and office equipment). The Baseline 2019 report puts electric dryers at a slightly higher share of population electricity consumption than electric water heaters (6% vs 5%), due to their much higher presence in homes (61% vs 16%). The Baseline 2019 report estimates electric ranges or ovens are present in 51% of homes (p.7), but apparently did not meter their use — they do not appear in Table 3-28 analyzing kitchen end uses. The 2022 update of the Baseline Report does meter ranges and Table 2-1 in that report shows cooking use at 3% of overall load and we use that estimate for load share although we otherwise use the slightly different 2019 estimates for saturation/prevalence.
Electric ranges are today more common than electric space and water heating and were also more common in the 70s. In fact, the 1980 Census of Housing, Detailed Housing Characteristics for Massachusetts volume reports that 50% of homes used primarily electricity for cooking, not statistically different from 2009. The Massachusetts of Energy Resources energy resources report estimates that as for water heating, electricity consumption for cooking was falling in the late 70s. We again infer that the most reasonable assumption as to cooking was that it was net flat from 1972 to 2009.
We have not located data for historical electric laundry dryer home prevalence, but the same Energy Resources Report shows laundry electricity consumption flat or down in the late 1970s and early 1980s. The only reasonable assumption is that it was net flat from 1972 to 2009 like electric water heating and electric cooking.
Note: From the Annual Housing Survey, we have estimates of the electric water heating and electric cooking shares for the Northeast Region in 1978 that were considerably lower than those we have for Massachusetts from the Census for 1980. An earlier version of this post made inferences from those estimates as to Massachusetts, but the Decennial Census data which we now rely on is more specific. It appears that Massachusetts had a higher appliance electrification rate than the Northeast as a whole. This may go along with Massachusetts heavier reliance on oil for heating. The largest states in the Northeast (NJ, NY, PA) had much lower reliance on oil for heating correspondingly higher home heating with gas. See “1970 heating fuels tab” of spreadsheet below.
Click here for spreadsheet of computations and sources.