The Record Of What Has Been Tested

Biological methanation has a verified record, and it is 68 documents long. Not 68,000. Sixty-eight publications, counted by hand and split into ten groups by what they measure: twelve reviews, thirteen trickle bed studies, eight stirred tank studies, five membrane and biofilm papers, three plug flow and novel reactor papers, six in-situ injection papers, five kinetics and additive papers, four syngas papers, five techno-economic papers and eight supporting methodological studies. The first batch was 45. The second batch added 23 on 2026-07-25. Every reactor family in the sector, every full-scale plant on the record and every cost projection in circulation traces back to something on that list. When an operator says the technology works, he is pointing at one of 68 entries.

Citation counts are the closest thing this record has to a consensus, and they are lopsided. Angelidaki 2018 on biogas upgrading sits at 1,382 citations in Biotechnology Advances. Lecker 2017 on biological hydrogen methanation sits at 214. Rittmann 2015 on the prerequisites for bioprocess development sits at 143. Thema 2019 on a standardisation framework for methanation rates sits at 102, and the same author surveyed 153 power-to-gas projects in a second 2019 paper at 114 citations. Jensen 2021 on gas-liquid mass transfer sits at 104 and is the reference every other paper cites for the transfer coefficient. Twelve reviews out of 68 means one entry in six is not an experiment at all. It is a summary of experiments. That ratio is the most honest number in the file.

Mass transfer is the arithmetic that separates one reactor from another. Jensen 2021 collected it into a single comparative table, which is why it became the citation. Alfaro 2018 holds the highest membrane figure on record at 268 per hour with a flat membrane module. Diaz 2018 measured 65 to 120 per hour in a stirred tank at 300 to 500 revolutions per minute. Orgill 2013 measured 50 to 80 per hour in a bubble column with a fine sparger. Karakashev 2020 put the biological ceiling underneath all of it: a specific hydrogen consumption rate of 0.2 to 0.36 mole per gram per hour. Every design in the record is an attempt to hand the archaea their hydrogen faster than the gas will dissolve into liquid. That is the entire engineering problem, and it is written down.

Struebing 2017 holds the record methanation rate in a trickle bed reactor at 15.4 normal cubic metres per cubic metre per day, above 98 per cent methane, on a gas load of 6 cubic metres per square metre per hour. Every later paper measures itself against that line. Seifert 2014 is the benchmark on the other side: 0.51 in a stirred tank at 65 degrees and one bar, at 85 per cent methane. Thema 2021 ran a pure culture at 13.5 bar and got 0.35, which is the answer to a question people keep asking. Pressure and purity do not beat mixed culture and surface area. Keramati 2024 put pre-colonised biochips into a trickle bed and tripled the rate to 15.7 at a 0.3 hour gas contact time. Savvas 2024 ran a tubular baffled reactor above 90 per cent at 270 litres per litre per day. Nine years passed between Struebing and the first serious attempt to pass him.

The whole verified record of the technology is 68 documents.

A rate is a point in time. Continuous operation is the record. Burkhardt 2019 published 1,200 days of uninterrupted trickle bed operation at 55 degrees and 5 bar, at 97.7 per cent methane. That is three years and three months on one biofilm. Rachbauer 2016 ran 240 days. Fenske 2023 ran a 0.8 cubic metre pilot at a wastewater treatment plant for 450 days, including shutdowns, and reported hydrogen sulfide removed biologically as a side effect. Jiang 2022 ran 151 days with no external additives and watched a loading shock push the community into homoacetogenesis. Savvas 2017 held 40 volumes per volume per day in a plug flow design on the lowest parasitic energy in the file. Four studies past the four-month mark. That is the whole long-duration record.

Two papers describe a way around the transfer limit, and both come from the same group. Tucci 2023 added a quinone-based redox mediator and measured a 110 to 181 per cent increase in methane rate, because the electron shuttle bypasses the gas-liquid interface entirely. Tucci 2025 added magnetite nanoparticles and measured a 13-fold increase, with the bacterial population shifting from acetogens to hydrogen oxidisers. Mauerhofer 2021 took the opposite route: a hyperthermophilic archaeon resistant to 50 bar, at 80 to 85 degrees, with a published potential for two to four times the rate. The source notes something else about that paper. No patents have been filed on the concept. A 13-fold improvement is a manufacturing question. A two-to-four-fold improvement at 50 bar with no filings is open ground, and nobody has planted a flag in it.

Sixty-eight papers do not contain a winner. Three reactor families are still running against each other, one at 0.5, one at 15.4, one at 0.8 cubic metres over 450 days, and nothing in the literature says which one is right at industrial size. The record holds another hole, and the source names it: the entries without a verified digital object identifier are cross-referenced from an earlier 45-paper database and from citation analysis, and the source instructs that they be confirmed before use in a competitive filing. Anything built on them is built on a reading list rather than on a paper. The commercial side is smaller than the academic side. Europe counted 35 operational e-methane plants in 2023 producing 449 gigawatt hours a year, against 1,678 biomethane plants in June 2025 and a published European target of 35 billion cubic metres. The biology is proven at 0.8 cubic metres. The market is priced in billions. The probability that this record is complete is not zero. The probability that it is a forecast is.