Wood pellet capacity by country

What the review found

The pellet register was built in October 2026 from three kinds of source: public registers (the ENplus register of certified producers, the US Energy Information Administration's monthly densified biomass fuel report), industry plant tables (Bioenergy International's World of Pellets map, Holzkurier's table of plants in the German-speaking countries, Bioenergitidningen's Swedish pellet map, Canadian Biomass's pellet mill map) and the operators' own sites and reports. Plants above 50 000 tonnes per year were sought actively; smaller plants were taken in when a register or table named them, so the register is complete above the threshold and representative below it.

Every row was then verified in two passes: one agent checked existence, status, capacity and raw material against the operator's own sources, and a second, independent agent tried to prove a random sample of the approved figures wrong and challenged every proposed correction before it was written. The verification corrected some two dozen capacities, moved about forty plants to the right town (most often from the company's seat to the plant), found more than a dozen duplicate rows that industry tables had listed under two owners or two spellings, and set fourteen European rows to unknown status where the company had been deregistered or the only certificate had lapsed and no other source existed. The error rates of the review itself: 0 of 15 approved figures fell in the Nordic and Baltic pass, 0 of 15 in the German-speaking and Central European pass, 1 of 14 in the rest of Europe, 1 of 10 approved and 2 of 8 corrections in the World of Pellets supplement, and 0 of 13 in North America.

Two findings shaped how the register is kept. First, the operator's own capacity is the only figure that replaces a table figure. A production volume, a delivery figure or a press rating in tonnes per hour never does, because the table figure is at least a stated capacity and the others are not. Second, a lapsed certificate is a signal, not a closure. Fifty-four plants carried an ENplus certificate that the register had since terminated or suspended; the certificate status is recorded as such, and the plant's own status is changed only when a source says the plant stopped.

The locations were verified the same way as the sawmills': the address from a register or the operator, a satellite image around it, one agent's verdict and an independent control pass on every site-level hit. The control passes downgraded about a third of the first-pass hits to town level, mostly sawmills with a pellet line where the image shows the sawmill but not the line. Where the first image showed the plant beside the point rather than on it, an agent pointed to it in the image, a script converted the pixel to a coordinate, and a second agent judged a new image around that point; no agent ever wrote a coordinate.

Capacity by country

As of 2026-10-19, WoodComp's register holds 879 published pellet plants in 34 countries. 827 are operating and 817 of those carry a capacity, together about 47 702 000 tonnes of pellets per year. 74 of the plants, all in the United States, report their capacity to the US Energy Information Administration in short tons per year, together 10 612 000 short tons; the register keeps that figure as reported and converts it to tonnes at 0.90718474 for the table, so the countries can be compared. 10 operating plants carry no figure, because no source we have read publishes one.

How to read these numbers

• Short tons are converted for display, never in the register. Plants in the United States report to the Energy Information Administration in short tons per year, and the register keeps that figure as the source states it; each facility page shows it. In the table and in every list on the site the figure is converted to tonnes at 0.90718474 tonnes per short ton and rounded to the nearest thousand, so that a plant in Georgia stands next to one in Latvia on the same scale. The column "Of which reported in short t/y" gives the reported sum, so the conversion can be redone with another factor.
• Capacity is pellets out, not wood in. A pellet plant's capacity is tonnes of pellets per year. Its wood consumption is recorded separately, as raw material with its origin, and is never summed with a sawmill's or a pulp mill's consumption, because much of it is the same wood a second time (see the double counting rule on the methodology page).
• The table follows the source. Where the operator publishes a capacity, that figure stands; where only an industry table does, the table figure stands and the facility page says so. Where neither does, the field is empty.
• Only operating plants are counted. Idled, closed, unknown and projected plants are in the register with their history and sources, but not in these totals.
• A shared site is one place, two facilities. A pellet line on a sawmill's plot is its own row with its own capacity, status and products, and shares the sawmill's site key. The last column counts such rows; group the dataset on `site_id` to count places instead of facilities.

Where the plants stand

Of the 879 published pellet plants, 106 (12.1 percent) share their industrial site with another facility in the register: 105 stand on a sawmill's site and press that sawmill's residues, 1 on a pulp mill's. The rest are free-standing plants that buy sawdust, chips or roundwood. 220 plants carry a raw material record with its origin (on-site residues, purchased or mixed). 204 plants (23.2 percent) are located to the plant itself after satellite verification and appear as pins on the map, 210 are placed at their town, and 465 have no verified location yet. 254 plants hold an active ENplus certificate in the public register.

Where the gaps were

The register's weakest region is the one with the most plants: southern and eastern Europe. Bosnia, Serbia, Ukraine, Portugal, Spain and Poland together hold a few hundred small plants that appear in industry tables but have no website and stand in company registers that cannot be read automatically. For about a third of the European rows the table is still the only source for existence, and those rows carry a flag saying so. The United States is the opposite: the EIA report gives every plant's capacity and status but withholds its operator, town and address, so the American rows were populated from the report and then located one by one through the EPA's facility register and the operators' own sites.

The geolocation stopped where the registers did. A little over half of the published plants carry a verified location; the rest are plants whose address no register or company page we can read discloses. Those rows are listed with their town only and marked unverified, and the register says so rather than placing a pin at the town centre and calling it a plant.