What do American households use electricity for?

It's all about controlling temperature and humidity.

The more you look at data on current and future energy demand, the more you realise that it’s nearly all about controlling temperature.

Industry is mostly about achieving high temperatures. Supply chains are about keeping stuff cool.

But even within households, nearly all of our energy is used for heating, cooling, or controlling humidity.

I’ve been looking at the Residential Energy Consumption Survey (RECS) that the Energy Information Administration in the United States carries out periodically. It’s an impressive effort that tries to capture what Americans use energy in their homes for.

In the chart below you can see the electricity breakdown for the average household. This is measured in kilowatt-hours (kWh) per year.

Note that this is based on the average use for households that have a given technology.¹ For example, the electric vehicle charging figure is for households that have one. If you were to average the electricity use for EVs across the US it would be really low, since most households don’t have one yet.

The two end uses I’ve left out are pools and hot tubs, because most households don’t have one (and unlike EVs, I don’t expect a major ‘hot tub’ boom over the next decade). But the chart with them included is in the footnote, if you’re interested.²

With the exclusion of EVs, it’s nearly all about heating and cooling. Either controlling air temperature, water temperature, humidity, or refrigeration.

Stuff like clothes washing, cooking, watching TV and even lighting are small in comparison. Add all of these uses up and it’d still be less than water heating on its own.

Data Source: US EIA (2020). Residential Energy Consumption Survey (RECS).

A quick way to visualise the importance of temperature control is to group all of these related uses together. In the chart below, ‘heating, cooling, humidity’ includes ACs, fans, space and water heating, dehumidifiers and refrigeration.

You can see that it completely dominates electricity use.

Data Source: US EIA (2020). Residential Energy Consumption Survey (RECS).

The figures for electric vehicles also seem surprisingly small. How credible are they? I estimate that the 2,363 kWh is equivalent to a bit less than one electric car, based on the average American driving patterns.

The Tesla Model 3 uses around 144 Wh per kilometre. The average American drives around 60 kilometres (37 miles) per day. That’s equal to around 3,100 kWh per year – a bit more than the average value in the survey. Some of the difference here might be explained by public charging: if drivers are occasionally charging at work or using public outlets, their home charging use will be lower than expected.

In the future many households will have two electric cars. In that case you’d be looking at a combined consumption of 5,000 to 6,000 kWh per household.

Note that in other countries, electricity demand for EVs will be lower. UK drivers cover just 22 miles per day – about one-third less than Americans.

---

The timing of supply and demand will be the biggest hurdle

Countries across the world will need to produce a lot more electricity in the decades ahead. This is no secret. The adoption of electric cars, heat pumps, and growing demand for air conditioning will put increasing strain on the grid.

But it’s not just the total electricity demand that’s important, but it’s timing.

I think this balancing act will be most challenging for heating. It’s the power user that’s most out-of-sync with renewable supply, and is hard to shift to another time of the day.

Electric vehicles won’t need to be charged every day – the range of most cars is much larger than the daily mileage. Drivers aren’t going to need a full top-up at the same time. And charging can be managed with smart vehicle-to-grid networks. Batteries can charge during low-demand, high-supply times, such as during the night. Smart technologies should be able to manage this across the grid on their own, without much human input.

Air conditioning demand is going to grow a lot as people across the world get richer, and the world gets hotter. The United States is quite unique in the fact that most households already have air conditioning. By contrast, AC adoption in Europe is low. And in middle-to-low income countries, most people can’t yet afford it.

This will be a challenge for developing grids. But perhaps the saving grace is that solar supplies should be abundant in some of the hottest periods of the day. Supply and demand should match quite nicely.

I’m not sure the same is true for heating. Most people heat their homes in the morning before work or school, and in the evening when they get home. These are the typical periods of ‘peak demand’. Heat pumps, then, add to the hardest part of the day. There likely won’t be much solar to top up energy supplies.

It’s therefore heating that will be the hardest peak to tackle, especially at high latitudes.

But to bring it back to what this data shows most clearly: future energy demand is mostly a challenge of temperature control. This is true in both industry and households. Solutions that help us to do this more efficiently will be crucial.

---

1. This is an interesting chart from the US EIA that shows the share of households that have a given technology. To get the total consumption of energy, you’d have to multiply the use per household by the share that has one.

You can see, for example, that almost 90% of households have “space cooling” (mostly air conditioning). Space and water heating appear low because many households still use gas as their heating source. Hopefully, this will change in the future as more switch to electric heat pumps.

Note that even when looking at total electricity consumption across the US, cooling and heating still dominate.

2. Electricity demand per household (for those that have a given technology) with pools and hot tubs included.