The Claims On The Char

Hydrothermal carbonisation has been patented since 1913, when Bergius filed the original process and took the Nobel Prize in Chemistry for the work in 1931. More than a century later the granted record is still short enough to read in an afternoon. The named holders fit on one hand: Ingelia in Spain, C-Green in Sweden, TerraNova Energy in Germany, Suncoal under UPM, and HTCycle, the former AVA-CO2 business. That is the first fact worth having. A technology running commercially with five significant portfolios is a technology where the ownership question has an answer, and the answer is written in the claims.

Ingelia holds ES2768236T3, a continuous hydrothermal carbonisation reactor and process, granted and active in the European Union with a family of national phase entries behind it. Read the claim and the commercial position becomes obvious. The patent covers continuous operation rather than batch. Continuous is the configuration the industry actually runs, because a batch vessel cannot hold a steady thermal profile and cannot be fed without stopping. So the most important claim in the field is not on a chemical reaction. It is on the shape of the machine, and the shape of the machine is the part most operators have never thought to check.

Two more grants make the pattern clearer. C-Green holds US11724952, hydrothermal carbonisation combined with wet oxidation, granted in the United States in 2023 with coverage in Sweden, the European Union and the United States. The load-bearing part is the second half: wet oxidation removes the phytotoxicity that otherwise blocks hydrochar from soil. TerraNova Energy holds US10343917 for carbonisation coupled with thermal hydrolysis to raise biomethane yield, and US10287505 for the integration of the process with municipal wastewater treatment. Both are granted, both run in the United States and Europe, and the second carries SUEZ, a water utility, as a co-assignee. The published gain from the hydrolysis pairing is 50 per cent more biomethane. Nobody is patenting the base reaction. They are patenting what happens when it is bolted to something else.

Suncoal went the other way. US12545789 covers carbonised lignin materials sold as CarboREN and US12584021 covers lignin-based rubber fillers. The portfolio runs to 47 or more active patents, all granted, and UPM acquired the business in 2023. Read that as a signal about where the margin sits: the same chemistry, sold into materials rather than energy. HTCycle took the co-product route instead, with a portfolio built around phosphorus recovery from the process water and the conversion of hydrochar into activated carbon. Two different answers to the same question, which is what a char is worth once it leaves the reactor.

A field with five portfolios has an answer to the ownership question.

Sort the record and the shape of the field appears. Reactor design takes about 40 per cent of the patents, process integration about 25, product valorisation about 20, process water treatment about 10 and novel feedstocks the last 5. Geographically Europe holds 55 per cent, the United States 25, China 12, Japan and Korea 5 and the rest 3. Filing was steady from 2010 to 2020 and then accelerated sharply from 2020 to 2025. The applicant mix is the part most readers skim past: universities file about 60 per cent of it, with Chinese universities dominant, corporates about 35 and individuals about 5. A field where the universities lead is a field still in its research phase, and the research phase is where a filing can still define a position rather than defend one.

The white spaces are as legible as the claims. In-situ catalysts are barely covered, so metal-doped and acid-base catalytic carbonisation is close to empty ground. Magnetic hydrochar composites, the ones built for pulling char back out of water, are nearly unpatented. Integration of carbonisation with carbon dioxide capture from flue gas carries zero filings. Machine-learning control of the reactor carries none. Reprocessing gasification char through carbonisation carries none either. That list is the useful half of a patent search, because it is the half that is not being enforced by anyone. The probability that a white space stays white is not zero, and most people price it at zero. The claims tell you what is owned. The absences tell you what is still available, and only one of those two lists arrives pre-sorted.