Landfill Gas to Energy Is the Cheapest Megawatt in Waste. So Why Are Operators Leaving?

About 540 landfill gas to energy projects were running in the United States as of September 2024, drawing methane off nearly 490 active and closed landfills, according to the EPA's Landfill Methane Outreach Program. Most of them do one thing: burn the gas in an engine and sell the electricity. And on a straight cost-of-generation basis, that electricity is close to the cheapest renewable megawatt anyone can build in the waste sector.

The fuel is free. Better than free, really, because federal rules make you collect and destroy the gas whether or not you ever sell a kilowatt-hour, so the collection system lands on the compliance budget, not the energy budget.

Which raises an odd question. If landfill gas is the cheapest baseload renewable in waste, why are the operators with the best sites quietly walking away from the engine?

The megawatt is cheap because the gas is a liability first

A large landfill is a methane factory that can't be switched off. Buried organic waste breaks down into a gas that runs about half methane and half carbon dioxide (the split drifts with the age of the waste and how wet it is, but 50 to 55 percent methane is typical). Methane is a strong greenhouse gas and a fire hazard, so the EPA doesn't leave capture to the operator's discretion.

Under the 2016 New Source Performance Standards at 40 CFR Part 60 Subpart XXX, a landfill above 2.5 million metric tons of design capacity that emits 34 megagrams a year or more of non-methane organic compounds has to install and run a gas collection and control system. The wells, the vacuum blowers, the flare: those go in because the rule says so, not because anyone has decided to sell energy.

So by the time you're choosing what to do with the gas, the expensive part is already built and paid for as a cost of staying legal. Flaring earns nothing. Send the same gas through an engine and a mandated liability turns into a revenue line. That's the whole case for landfill gas to energy, and it's a strong one.

The power itself behaves well, too. Landfill gas flows around the clock, so an engine running on it looks like baseload rather than the on-again-off-again output you get from solar or wind; availability north of 90 percent is normal for a well-run site, per LMOP's figures. More than 70 percent of these projects use reciprocating engines, mostly GE Jenbacher and Caterpillar units, because they're cheap per kilowatt and they tolerate a dirty, variable fuel. Electrical efficiency lands around 30 to 35 percent, which reads as poor until you remember nobody is paying for the fuel. It's about as dull and dependable a form of renewable energy from waste as the sector has.

Capture efficiency is the number that actually moves

All of that assumes the gas reaches the engine. That assumption is where most of the money is won or lost, and it's a data problem before it's an energy problem. Every landfill gas to energy project lives or dies on how much of the methane the wellfield actually pulls in.

A wellfield is dozens to hundreds of extraction wells, each drawing gas under vacuum. Pull too little and methane slips out through the cap and off your revenue. Pull too hard and you drag air into the waste mass, which can push a section aerobic and, at the extreme, start a subsurface fire. The tuning window is narrow, it moves as the landfill settles and as it rains, and it sits somewhere different at every wellhead.

This is my actual corner of the business, so let me be blunt. In 2022 I helped a Midwest site instrument its wellfield: methane, oxygen, and pressure sensors at every wellhead, feeding a model that recommended vacuum set points. Three months in, the model began calling for settings that walked the oxygen readings toward the 5 percent shutdown limit, and we nearly throttled a row of good wells over it. The trigger looked like a modeling error. It wasn't: the calibration gas for the methane sensors was two years expired, the readings had drifted, and the model was faithfully optimizing on garbage. We lost 4 weeks of usable capture before anyone thought to check the bottles.

Sensor drift always wins eventually, so on a wellfield the recalibration has to live in the maintenance schedule, not inside the model. Good waste intelligence here isn't a fancier network; it's instruments you can trust and a control loop that closes fast. The same discipline runs downstream, in the gas conditioning and upgrading technology that strips moisture and siloxanes before the gas ever reaches an engine or an upgrader.

I'll say the harder thing too. A lot of what gets sold as AI wellfield optimization is a PID control loop with a dashboard on top. The gains in the case studies almost always trace back to fixing instrumentation and closing the loop faster, not to the model architecture. Tune the inputs before you buy the network.

The money moved from the electron to the molecule

So, back to the odd question. If the megawatt is cheap and dependable, why leave it? Because the same gas is worth far more as a molecule than as an electron, and the gap is wide.

Clean landfill gas up to pipeline spec and inject it, and it counts as renewable natural gas. An RNG project sells the commodity gas, then stacks two subsidy streams on top: a federal cellulosic (D3) RIN under the Renewable Fuel Standard, and, where the gas reaches a low-carbon-fuel market like California's, an LCFS credit. The American Gas Foundation's July 2025 supply assessment put the combined credit support for landfill RNG north of $13/MMBtu, on top of the gas itself, and found landfill and wastewater gas to be the cheapest RNG feedstocks going, most of the achievable volume produced under about $20/MMBtu.

Set that beside a wholesale power price and the comparison stops being close. Electricity earns an energy price plus, in some states, a renewable energy credit. RNG earns the gas plus credits worth several times the gas. It's the lesson operators keep relearning across the sector: the kilowatt-hour is usually the byproduct, and the real product is something else. In a mass-burn plant that product is the gate fee. On a landfill there's no gate fee on the gas at all, so the credit is the whole product.

But there's a catch, and it's geographic. RNG only works if a pipeline interconnect sits close enough to reach without spending the whole subsidy on a lateral. That's the same constraint that decides whether carbon capture on a waste plant ever pencils, which comes down to what's next door more than what's in the flue. A landfill 30 miles from the nearest gas main is still, in practice, an electricity site or a flare. Across the waste-to-energy projects I've worked, geography and offtake pick the technology long before the engineering does.

Where the cheap megawatt still wins

None of this retires the engine, and the RNG pivot has real limits. Below a couple of megawatts of gas, the upgrader's capital and the pipeline lateral stop penciling, and electricity or a direct use for the raw gas (a neighboring factory boiler, say) stays the better call. Small and older landfills, wet-climate sites with erratic gas, and anything far from a pipeline are electricity or flare cases for years yet.

But the RNG economics aren't as bolted down as a sub-three-year payback makes them sound. Those D3 RIN credits float on a policy market. Through 2025 their price softened enough that waste companies openly paused RNG buildouts, waiting to see where the Renewable Fuel Standard volumes settled; Waste Dive tracked several of those hesitations in the fall. A megawatt sold on a power contract is boring. A molecule sold on a credit is exposed to Washington. Some operators will take boring.

The EPA reckons another 444 landfills could support a cost-effective project and don't have one yet. Watch where those go over the next two or three years. The ones near a pipeline will mostly chase RNG. The ones that aren't will take the cheap, dull, dependable megawatt, or they'll flare, because doing nothing with the gas was never legal and an engine is the lowest bar to clear. The cheapest megawatt in waste isn't going away. It's just stopped being the prize.

Sources & Notes

• Project counts, the 90-percent-plus availability figure, and the 444 candidate landfills come from the EPA's Landfill Methane Outreach Program project data (September 2024 update) and its benefits summary.
• The collection mandate and its triggers sit in the 2016 New Source Performance Standards for municipal solid waste landfills, 40 CFR Part 60 Subpart XXX.
• Credit support north of $13/MMBtu and the sub-$20/MMBtu production cost for landfill gas are drawn from the American Gas Foundation's Renewable Natural Gas Supply Assessment (July 2025).
• For the 2025 RIN-price wobble and the projects it stalled, see Waste Dive's reporting on RNG pricing pressure.
• The 2022 wellfield anecdote is from my own project file; the figures there are that site's instrumentation data, not a published source.