Originally published March 3, 2013. Updated July 30, 2026.
About two of every three units of energy the US uses never do any useful work. The country consumed 94.6 quadrillion Btu of primary energy in 2024, and Lawrence Livermore National Laboratory’s annual flow chart counted 62.27 quadrillion Btu of that as rejected, mostly lost as heat, for a rejected share of 65.8%. That share hasn’t moved much in years, even after a decade of cheaper LED bulbs and tighter appliance rules.
How much energy the US wastes
Total US energy use splits into two buckets on that LLNL chart: energy that does something, like running a motor or heating a room, and energy that’s rejected along the way, mostly as waste heat from combustion and power generation. In 2024, 32.34 quadrillion Btu counted as useful energy services, while the remaining 62.27 quadrillion Btu got rejected. A year earlier, in 2023, the chart put total consumption at 93.6 quadrillion Btu with 65.7% rejected, almost identical to 2024’s 65.8%.
Behind that flat total, the generation mix shifted. Natural gas supplied a record 34.3 quadrillion Btu in 2024, and renewables supplied a record 8.7 quadrillion Btu. Coal fell to a record low of 7.9 quadrillion Btu. None of that moved the rejected share.
Where the wasted energy goes, sector by sector
Two big categories eat most of the waste: turning fuel into electricity, and burning fuel directly for transportation. Converting fuel to electricity lost 19.21 of 32.77 quadrillion Btu in 2024, a 58.6% loss rate that works out to roughly 41% conversion efficiency for the average power plant. That conversion-efficiency number isn’t the same as a power plant’s capacity factor, which measures how much of a plant’s yearly potential it delivers rather than how much fuel it turns into power. See Energy Collective’s capacity factor breakdown for that number by source.
Transportation rejects more than any other sector, 22.35 of 28.29 quadrillion Btu in 2024, a 79% loss rate that LLNL attributes to an assumed 21% end-use efficiency for engines and drivetrains. Industry rejected 13.46 of 26.39 quadrillion Btu, a 51% loss rate, well below transportation but still substantial. Buildings pull the average down: homes rejected 3.93 of 11.23 quadrillion Btu, a 35% loss rate, and businesses rejected 3.32 of 9.49 quadrillion Btu, also 35%.
| Sector | Energy in (quads) | Rejected (quads) | Share rejected |
|---|---|---|---|
| Electricity generation | 32.77 | 19.21 | 58.6% |
| Transportation | 28.29 | 22.35 | 79% |
| Industry | 26.39 | 13.46 | 51% |
| Residential | 11.23 | 3.93 | 35% |
| Commercial | 9.49 | 3.32 | 35% |
Source: Lawrence Livermore National Laboratory, 2024 US Energy Flow Chart
Is US energy efficiency improving
By some measures, yes. US energy intensity, the amount of energy it takes to produce a dollar of GDP, fell from 5.03 to 4.05 thousand Btu per dollar between 2015 and 2024, a drop of about 19.5%, according to EIA. Energy use per person fell too, from 295 to 278 thousand Btu over the same years.
The rejected-energy share hasn’t followed that trend down. A 2013 piece from Energy Collective writer Sarah Battaglia put US energy efficiency at 42%, citing a 2008 study reported by the New York Times that found 71% of transportation energy and 66% of electricity-generation energy wasted. LLNL’s 2024 chart puts useful energy at 34.2%, lower than that 2013 estimate. The honest read is that rejected energy hasn’t visibly declined in over a decade.
Why bigger efficiency gains haven’t moved the total
LED bulbs and better appliances deliver real, measured savings, but they’re small against the two categories that dominate rejected energy: transportation and power generation. LEDs saved an estimated 1.1 quadrillion Btu in 2017, about $12 billion in avoided energy costs that year, according to the Department of Energy’s solid-state lighting forecast. Transportation alone rejected 22.35 quadrillion Btu in 2024. Those numbers come from different years and aren’t a direct comparison, but lighting is a small slice of a total that transportation fuel and power-plant conversion losses dominate. Even DOE’s most optimistic case, cumulative LED savings of up to 78 quadrillion Btu and about $890 billion in avoided costs across the 18 years from 2017 to 2035, is measured against a rejected-energy total that topped 62.27 quadrillion Btu in 2024 alone.
LED lighting is the easy win worth taking
At the bulb level, the payoff is dramatic and well documented. ENERGY STAR-certified LEDs use up to 90% less energy than standard incandescent bulbs and last 15 times longer. A single LED saves roughly $55 over its lifetime compared with an incandescent, and swapping a single bulb per household nationwide would save an estimated $580 million a year.
The federal floor is rising too. A new Department of Energy standard raises minimum bulb efficacy from 45 to more than 120 lumens per watt, a separate action from the incandescent phase-out already in force, and takes effect in 2028. DOE projects it will save $1.6 billion a year and $27 billion cumulatively over 30 years. Over that same span, the agency expects it to cut 70 million metric tons of CO2.
Adoption still has room to run. As of 2020, LEDs made up about 48% of installed lighting units in the US, meaning the other 52% of fixtures were still due for the swap, the most recent figure DOE has published.
Buildings already run more efficiently than transportation or industry
Residential and commercial buildings account for 27.6% of US end-use energy, or 36.9% once you count the electricity-generation losses baked into every kilowatt-hour they draw from the grid, according to EIA. Even so, homes and businesses reject a smaller share of what they use than transportation or industry does. Residential use rejected 35% in 2024, and commercial use rejected 35% too, both well below transportation’s 79% and industry’s 51%.
That 35% figure is one of LLNL’s fixed modeling assumptions for how efficiently buildings use the energy delivered to them, not a number the lab remeasures every year, so it can’t say on its own how fast the US building stock is improving. A lot of what a building wastes leaves as heat. If you want to know where that heat escapes a house, Energy Collective has a separate breakdown.
What would cut US energy waste next
The biggest single number to watch is transportation, since it rejects more energy than any other sector and runs almost entirely on combustion. Swapping combustion for electric drivetrains moves that loss out of the tailpipe and into the power plant, where the average conversion efficiency is about 41%, better than an engine’s assumed 21% efficiency but still short of full use.
On the policy side, the lightbulb standard taking effect in 2028 is a small, concrete test case, and it’s the one federal efficiency rule here with clear, sourced before-and-after numbers behind it. Broader levers like appliance and vehicle-efficiency standards, and utility rate design, matter for the same total, but they don’t have a comparable public before-and-after number to check yet. If DOE’s projected $1.6 billion a year in savings shows up once the standard is in force, that’s evidence a single federal standard can move a number that’s held close to 66% rejected for years. The next LLNL flow chart, published every year, is what will say whether it did.
