---
title: 'Energy Consumption in Commercial Buildings: What CBECS and Eurostat Data Actually Show'
description:
    'How much energy do commercial buildings actually use, and where does it go? CBECS, Eurostat and ENERGY STAR data on
    end use, building type and climate.'
image: /images/articles/energy-consumption-commercial-buildings.webp
image_alt: Modern city buildings with illuminated windows at night
published_date: '2026-08-19 12:00'
---

_Photo by
_[_Sebastian Jacobsen_](https://unsplash.com/@hejsfj?utm_source=frostlogic-mission-control&utm_medium=referral)_ on
_[_Unsplash_](https://unsplash.com/?utm_source=frostlogic-mission-control&utm_medium=referral)_._

Commercial buildings accounted for about 17.2% of total US energy consumption in 2023, once electrical system losses are
counted, according to the Energy Information Administration's Monthly Energy Review. That figure sits underneath every
capital plan, disclosure filing and decarbonization roadmap a building owner produces, and almost none of those plans
start from a blank page. They start from a small set of federal and EU surveys that measure, in detail, where that
energy actually goes.

Four things depend on getting that breakdown right: benchmarking a building against its real peer group instead of a
rule of thumb, building a capital or utility budget around where the money actually goes, meeting a growing list of
disclosure ordinances that ask for exactly this kind of data, and prioritizing a decarbonization plan against the end
uses that carry the most weight. None of that works from guesswork. It works from the Commercial Buildings Energy
Consumption Survey (CBECS), Eurostat's energy balances, and the ENERGY STAR Portfolio Manager benchmarks built on top of
CBECS, all cited directly below. Where a figure isn't published in one of those sources, we say so rather than filling
the gap with an estimate.

One caveat up front on vintage. CBECS runs on roughly a four-year cycle, and the 2018 survey, released in stages between
2021 and 2023, remains the most recent complete cycle at time of writing; EIA has published exploratory follow-up data
since, including a late-2024 look at office and education buildings' post-pandemic energy patterns, but no full
replacement survey. Every US figure below is 2018 CBECS unless stated otherwise, which matters for a building portfolio
whose occupancy or operating hours have shifted materially since.

## Where the energy goes: breakdown by end use

The EIA's 2018 CBECS, the most recent complete survey cycle covering the full US commercial building stock, splits
energy use into ten end-use categories: space heating, cooling, ventilation, water heating, lighting, cooking,
refrigeration, computing, office equipment and other. Three categories dominate. Space heating alone accounted for close
to a third of end-use consumption nationally, about 32% in 2018, more than any other single category. Ventilation and
lighting followed, each accounting for roughly 10% of total commercial building energy use.

Cooling's national share is smaller than either ventilation or lighting: commercial buildings consumed 589 trillion Btu
for cooling against a 6,787 trillion Btu total in 2018, close to 9% of the total, electricity supplying 98% of it. That
national average hides enormous regional variation, covered below.

Fuel source doesn't split evenly across those categories. Electricity accounted for 60% of total commercial building
energy consumption in 2018 and natural gas for 34%, but natural gas fueled nearly three-quarters of space-heating energy
and was the dominant fuel for water heating (67%) and cooking (81%). Electricity, by contrast, was the only energy
source used for ventilation, lighting, refrigeration, office equipment and computing, and it supplied 98% of cooling.

A note on what's not published: EIA's summary releases give top-line national percentages for space heating, ventilation
and lighting specifically, and cooling is calculable from the total-Btu figures as done above. Water heating,
refrigeration, cooking, computing and office equipment aren't broken out as rounded national percentages in EIA's
summary materials, only inside the full end-use tables by fuel and building type. We're not rounding those to a headline
percentage here, since EIA itself doesn't.

Energy intensity per square foot tells a different story than the aggregate shares. Space heating was the most intensive
end use at 25,000 Btu per square foot annually, an effect most pronounced in colder climates. Office equipment was the
least intensive at just 500 Btu per square foot, with computing, including servers, at 3,100 Btu per square foot, both
lower than the attention IT load gets in efficiency conversations would suggest.

Efficiency has moved in the meantime. Average energy consumption per square foot across the US commercial stock fell 12%
between the 2012 and 2018 CBECS cycles, electricity intensity down 14% and natural gas intensity down 11%, a shift the
EIA attributes partly to LED lighting adoption and building automation systems.

## Breakdown by building type

Energy consumption doesn't track building count or floorspace in any simple way. In 2018, office, mercantile and
education buildings together accounted for 43% of all US commercial building energy consumption. Building count skews
differently: warehouse and storage, office, and service buildings together made up 48% of the commercial building stock.
Floorspace differs again: warehouse and storage, office, and education buildings accounted for about half of total US
commercial floorspace. The three rankings don't line up, which is the CBECS data's clearest warning against assuming a
building type's share of one metric predicts its share of another.

Building size compounds the effect. CBECS's own averages show why: education buildings average 31,100 square feet and
health care 29,300, a figure skewed heavily by hospitals, which average 264,800 square feet against just 13,700 for
standalone outpatient health care buildings. Lodging buildings average 33,700 square feet, while food service buildings
average just 4,800, the smallest category the survey tracks. Mercantile buildings punch above their footprint: 9% of the
building count and 11% of floorspace, but 14% of major fuel consumption and 13% of energy expenditures, consistent with
retail's above-average energy intensity from lighting and refrigeration loads running most of the operating day.

Per-square-foot intensity flips the picture again. CBECS identifies food service, food sales and health care as the most
energy-intensive building types per square foot, categories running dense equipment loads, commercial kitchens,
refrigeration, medical equipment, for most of the operating day. The least intensive are vacant buildings, warehouse and
storage, and religious worship space, categories with light equipment loads and often partial conditioning of the floor
area.

ENERGY STAR's Portfolio Manager converts building-type intensity into the specific benchmark most owners actually use:
median site EUI is 52.9 kBtu per square foot per year for an office (116.4 on a source-energy basis) and 51.4 (120.0
source) for a retail store, both drawn from CBECS survey data. We've published the full EUI formula, the
site-versus-source distinction, and a worked example elsewhere; see our guide to
[energy use intensity](/articles/energy-use-intensity/) rather than have us re-derive it here. The practical point for
benchmarking against these figures: a hospital or full-service restaurant routinely posts an EUI several times an
office's, not from mismanagement but because the underlying operation is more energy-dense, so EUI comparisons only mean
something within the same building type.

## Regional and climate variation

Climate moves the needle on commercial energy use more than almost any other single variable. EIA's climate-zone
analysis of 2018 CBECS data found that buildings in hot or very hot US climate zones were more than six times as
cooling-intensive as buildings in cold or very cold zones, 14.2 thousand Btu per square foot annually against 2.3
thousand Btu per square foot. Cooling coverage differs just as sharply: 52% of buildings in hot climate zones reported
cooling their entire floorspace, against 25% in cold climate zones. A single national cooling average describes no real
building well. A Texas portfolio and a Minnesota portfolio benchmarked against the same blended national figure will get
a badly misleading read on at least one of them.

Fuel mix shifts by region too, not just cooling load. Nationally, commercial buildings drew 60% of their energy from
electricity and 34% from natural gas in 2018, but the South Census region's mix skewed further toward electricity: 69%
electricity, 26% natural gas, 4% district energy and 1% fuel oil. That's consistent with a warmer, more
cooling-dominated climate that leans less on gas-fired heating. A national fuel-mix figure obscures a real regional
split in the US, on top of the climate-zone split above.

The EU's data is structured differently but tells a related story. Eurostat tracks a services sector as the closest
equivalent to US commercial buildings, and it accounted for 13.5% of the EU's total final energy consumption in 2023,
smaller than transport (32.0%), households (26.3%) or industry (24.6%). Within services, electricity supplied 51.2% of
final energy consumption and natural gas 26.0%, with renewables and biofuels (8.4%), purchased heat (7.7%) and oil
products (6.2%) making up most of the rest, a more electrified fuel mix than the US commercial sector's roughly 60/34
electricity-to-gas split. Wholesale and retail trade was the single largest sub-sector, at 20% of EU services energy
consumption. The two datasets use different survey methodologies and sector definitions, so treat any US-EU comparison
as directional, not a precise like-for-like read. For a Nordic framing that compiles Swedish lokaler intensities with
Nordic services-sector totals from public statistics, see the
[Nordic commercial building energy benchmark](/articles/nordic-commercial-building-energy-benchmark/).

## How consumption patterns are shifting

Two forces are moving this data faster than any single survey cycle can track. The first is electrification. Cooling
already runs almost entirely on electricity, 98% per CBECS, and as heat pumps displace gas-fired boilers and furnaces in
new construction and major retrofits, more of that roughly 32% space-heating share migrates onto the same electrical
service that already carries cooling and lighting. EIA's own 2018 CBECS data shows the starting point: 31%, about 1.8
million buildings, were already all-electric, using no on-site fossil fuel for any end use electricity can serve. That
changes what "efficiency" looks like operationally: a building that electrifies heating can improve its site energy
efficiency on paper while its electricity demand, and its exposure to peak-demand utility charges, actually goes up.

The second is regulatory. Benchmarking mandates, built on the same Portfolio Manager infrastructure this data feeds,
have moved from voluntary to required filing across a growing list of US cities and states; we cover how that scoring
system works, and where it stops being a useful diagnostic, in our
[guide to commercial building energy benchmarking](/articles/energy-benchmarking-commercial-buildings/). CBECS and
Eurostat both describe the building stock in aggregate, once every several years. Neither tells an owner whether a
specific building's HVAC schedule drifted last month or whether this quarter's lighting load is trending somewhere the
next survey won't catch for years. That's the gap continuous [energy monitoring](/platform/energy-monitoring/) is built
to close: decomposing the same annual totals these surveys report into the week-to-week behavior driving them. Any
automation built on top of that monitoring data starts with write access off. Write access is granted per scope; until
then it returns a ranked queue for an operator to review.

## FAQ

**How much energy do commercial buildings use in the US?** About 17.2% of total US energy consumption in 2023, once
electrical system losses are included, per the EIA's Monthly Energy Review. The most recent full CBECS survey (2018)
measured 6.8 quadrillion Btu of commercial building energy consumption across roughly 5.9 million buildings and 96
billion square feet of floorspace.

**What uses the most energy in a commercial building?** Space heating, which accounted for close to a third of end-use
energy consumption nationally, about 32% in 2018 CBECS data, more than any other single category. Ventilation and
lighting followed, each around 10%.

**How much of a commercial building's energy goes to cooling?** About 9% of total commercial building energy consumption
nationally in 2018 (589 of 6,787 trillion Btu), almost entirely supplied by electricity (98%). That national figure
hides large climate variation: buildings in hot US climate zones are more than six times as cooling-intensive per square
foot as buildings in cold zones.

**Which commercial building types use the most energy?** By total national consumption, office, mercantile and education
buildings together account for 43%. By intensity per square foot, food service, food sales and health care run the
highest loads; warehouse and storage, vacant buildings and religious worship space run the lowest. Specific EUI
benchmarks by building type are in our [energy use intensity guide](/articles/energy-use-intensity/).

**How does US commercial energy consumption compare to the EU?** The EU's services sector, Eurostat's closest equivalent
to "commercial," used 13.5% of the EU's total final energy consumption in 2023, with electricity (51.2%) and natural gas
(26.0%) making up the bulk of the mix, a more electrified split than the US commercial sector's roughly 60/34
electricity-to-gas ratio. The two surveys use different methodologies, so read any comparison as directional.

**How do I calculate my own building's energy use intensity?** EUI is total annual energy consumption divided by gross
floor area. We cover the full formula, the site-versus-source distinction, and a worked example in our guide to
[energy use intensity](/articles/energy-use-intensity/), rather than repeat it here.

**Is this the same data used for benchmarking compliance filings?** Largely, yes. ENERGY STAR Portfolio Manager, the
tool behind most US benchmarking ordinances, calculates its 1-100 score from source EUI benchmarked against CBECS survey
data by building type. We cover how that scoring works, and where it runs out of road as a diagnostic, in our
[benchmarking guide](/articles/energy-benchmarking-commercial-buildings/).

**Does tracking commercial building energy consumption data require a CMMS?** No. Consumption data comes from utility
meters, submeters and building management systems, not from a maintenance management platform. FrostLogic Explore is a
sensor-intelligence platform built to ingest and analyze that data continuously; it can flag when a consumption pattern
points to a likely maintenance issue, but it isn't a CMMS and doesn't replace one.
