Burning Wood to Run AI (Hear Me Out)

Everyone in this industry has had to become an amateur energy planner. Power is now the first question on every site search, often ahead of land, fiber and zoning. In the last year I've seen campuses planned around gas turbines, on-site substations, battery yards and fuel cells, with small modular reactors on the long-range horizon.

Biomass rarely comes up. I think it deserves a seat at the table, and lately I've had a reason to look at it closely. A contact from a previous project is researching a biomass development in two states with good access to fuel and transportation, and a few developers have already shown interest. Watching that come together has given me some opinions.

What biomass power actually is

At its simplest, a biomass plant is a steam power plant. Instead of coal or gas, it burns organic fuel such as wood pellets, sawmill residue, forest thinnings or agricultural waste. The heat makes steam, and the steam turns a turbine.

The technology isn't experimental. Circulating fluidized bed boilers have burned wood at utility scale for decades. They're well suited to it because they tolerate fuel that varies in moisture and energy content, which is exactly how waste wood behaves. In the UK, Drax converted several large coal units to wood pellets and now runs one of the largest biomass stations anywhere.

Why it fits data center load

Data centers need power that runs around the clock, regardless of the weather. Solar and wind can't do that alone without a lot of storage. Gas can, which is why gas is back in fashion. But gas carries a carbon story that many buyers with 24/7 carbon-free goals would rather not explain to their shareholders.

Biomass sits in an unusual position. It's firm, dispatchable baseload like gas, but the carbon it releases was recently absorbed by growing plants. Pair it with carbon capture and geologic storage, known as BECCS (bioenergy with carbon capture and storage), and the plant can be net negative: it removes more carbon from the atmosphere than it emits. In the US, captured and stored carbon also qualifies for the federal Section 45Q tax credit, which changes the economics considerably.

For a buyer who wants clean, firm, always-on power, very few options have that profile.

Fuel is the whole story

What I learned quickly is that a biomass plant is a fuel logistics operation with a power plant attached.

The economics depend almost entirely on the cost of fuel delivered to the site. Pellets are dense and consistent but expensive. Waste wood is cheap but wet and bulky. Blending the two, and using the plant's own waste heat to dry the wet fuel before it's burned, is one of the smartest moves available. Drier fuel means more energy per ton, fewer tons to haul and a more efficient plant.

The volume is enormous. A large station burns millions of tons a year, which makes it a rail problem, not a truck problem. Sites with direct rail access in active timber regions start with a real advantage.

What it looks like on the site plan

This is where my architect's instinct takes over. A biomass plant's footprint is dominated by things a typical data center campus never has to think about: a rail spur and unloading station, covered fuel storage sized for weeks of inventory, screening, sizing and drying equipment, conveyors running across the site, ash handling and a stack. If carbon capture is added later, you also need a capture and compression plant plus injection wells, and that land has to be reserved from day one.

That's heavy industrial land use. Zoning needs to be in place and setbacks need to be generous. Dust, noise and the visual impact of fuel piles and conveyors belong in the massing studies, not the mitigation report. If a data center shares the campus, the relationship between the clean, quiet, secure compute buildings and the working fuel yard becomes a real master-planning problem. Prevailing winds, separated circulation and landscape buffers all become design decisions.

But the site plan is only half the job. The other half is the community. A fuel yard, a stack and a rail line are hard things to spring on neighbors, and they shouldn't find out about them from a zoning notice. Bring the community in early, show them the plans, including the parts that aren't pretty, and be straightforward about trucks, trains, noise and emissions. In my experience, being open from the start goes a long way toward getting people on board. Trust lost at the first public hearing is very hard to win back.

The honest drawbacks

I don't want this to read like a sales pitch, so here are the criticisms.

The sustainability of biomass depends entirely on where the wood comes from. Residues and mill waste are one thing; harvesting whole forests for pellets is another. Critics have raised legitimate questions about carbon accounting and forest impacts, so certification and supply chain transparency matter. My own view is that a project like this needs a full environmental impact analysis before it moves forward. That analysis should look not only at the plant but also at the forests and supply chain that feed it. Sourcing claims should be backed by a study, not just a promise.

Biomass plants still produce local air emissions, including NOx and particulates, which need to be controlled and permitted. Neighbors will ask hard questions.

Carbon capture adds substantial capital and energy cost. Geologic storage requires the right formations underground and a long federal permitting process, so not every site qualifies.

Fuel supply is a long-term commitment. A plant built around one regional wood basket is exposed if that market shifts.

Where it makes sense

Biomass won't power every data center, and it shouldn't. It makes the most sense where several conditions line up: a strong regional wood products industry, rail access, heavy industrial zoning, geology suited to carbon storage and a buyer willing to pay for firm, low-carbon or net-negative power.

In those places, I think biomass deserves more attention than it gets. The industry is already spending fortunes to secure power, so it makes sense to look seriously at the one firm source that can also draw carbon down.

Final Thought

In the end, biomass isn't the answer for every site, and neither is gas, solar, nuclear or anything else. The country needs a balanced national approach to power generation, one that considers every source and fits each to the location and the resources it has. In the right place, biomass has earned its spot in that mix.

About the Author

Frederick is an award-winning architect and planner with over 35 years of archite

ctural design experience and nearly 20 years in the data center sector, bringing an international portfolio that spans large-scale facilities and diverse project types. His background across global markets gives him a broad, practical lens that shapes the insights he shares.

Throughout this series, he closely follows the evolution of the data center market, drawing on industry events, newsletters and ongoing dialogue with fellow experts to distill what's worth paying attention to and why it matters.