Any technology thesis which divides a proven service from a contingent yield owes its audience a straight account of what would convert one into the other. For the GH2E platform, that division is the Track A / Track B distinction used throughout this series: treatment-as-a-service built on electrocoagulation (EC) is considered proven, while hydrogen-yield-at-scale is conditional, and remains contingent on its characterisation work — the total and dissolved organic carbon from the EC sludge cake and complete waste chemistry — that has not been fully explored across all treatment ecosystems. A previous article made the gap wider by showing how paunch manure offers a different feedstock chemistry from dairy manure, so the dairy-derived analysis simply does not carry across. This article asks the next question: what would a campaign to close that gap actually contain, and how should the data it produces be recorded if the platform intends, as the wider architecture does, to tokenise its outputs?
The short answer, developed with the same hedging this series has applied elsewhere: closing Track B takes a staged, independently witnessed campaign rather than a single laboratory report, with pass/fail thresholds fixed before the data arrives. Tokenisation belongs at the end of that sequence, not the beginning. A token can preserve, move, and make auditable a verified measurement; however, it cannot make an unverified measurement true, and because ledger records are hard to erase, it raises the cost of getting the underlying characterisation all wrong.
What Track B is actually waiting on
The unresolved item is narrower than “does the technology work.” Track B needs three separate questions answered, in order. First, how much organic carbon does the solid stream carry, and in what form? Second, when those streams are gasified, does the hydrogen yield reach a level that the commercial case assumes? Third, can that result be repeated, metered, and recorded to a standard that an investor, an insurer, or an independent verifier will accept without relying on the operator’s word? Each question depends on the one before it, and the first hasn’t been answered for either of the two solids that matter here: the paunch itself, and the electrocoagulation sludge cake that is produced by the water-treatment half of the platform.
The reason the elemental accounting method matters more than the headline yield is simple chemistry. Typical solid biomass and manure feedstocks carry on the order of 6–7% hydrogen by dry mass, with carbon in the high-40% range, as reported from a published analysis of manure, sludge, and straw.1 Hydrogen yields that exceed what the feedstock’s own hydrogen could supply must come from somewhere else: from carbon reacting with steam, from the water in or added to the process, or from any separate water-splitting stage operating operating in parallel. None of those sources are illegitimate, so they should not be ignored. But they are governed by a different set of inputs — the feedstock’s carbon content in the first case, energy and water in others — and a characterisation campaign that blends them into one yield figure may not survive a diligence reader’s first question about where the hydrogen came from.
Some evaluators give the chemistry less weight than it deserves, not because they doubt it, but because the real test is whether they have a credible level of expertise in a field where innovation often arrives in unfamiliar form. Advances in the energy space are frequently obscure at first. They may not map to an existing template, even as they fit a familiar technological bias about what a “proven” pathway looks like.
This isn’t a criticism. It reflects an investor’s comfort level and a reluctance to stray far from a risk framework that is built to protect capital, or declining a technology which resides outside that framework is a sound decision. But a decision made because the innovation doesn’t fit the template is different from one made because it fails on its merits, and the two should not be confused. The strongest evidence of seriousness is not a favorable headline yield but a team that can show its elemental accounting, define what it has and hasn’t established, and say precisely where each unresolved question will be answered. Advances that meet that standard deserve to be evaluated on their own terms, even when they arrive without a template.
Why the dairy numbers cannot be borrowed
Article Four set out the reasons paunch is its own case, and they bear directly on test design. Paunch leaves the slaughter floor as a wet, dilute stream at only 2–5% solids and is typically dewatered to roughly 23–24% dry matter before land-spreading or composting.2 That single fact imposes a discipline on every number in a commercial case: the figure quoted “per tonne of paunch” can differ by a factor of four or more depending on whether the tonne is wet, dewatered, or dry. Any yield, revenue, and throughput claim built on paunch needs its moisture basis stated alongside it, and a test campaign should report on a dry, ash-free basis with the as-received moisture recorded separately.
Paunch is also lignocellulose-rich and only partially digested, and published work on its composition documents substantial variability from one source to another.3 That variability is the point. If the dairy characterisation produced one number with a tight range, the paunch campaign cannot assume the same tightness; it has to measure the range itself, across the plants, seasons, and feeding regimes that a commercial deploy-ment would likely encounter.
The second solid: the electrocoagulation cake
The sludge cake deserves its own treatment, because it is the less studied of the two solids and the one the Track B contingency specifically identifies. Electrocoagulation works by passing current through metal electrodes, typically aluminum or iron, which dissolve and form metal hydroxides that bind suspended solids, fats, and phosphorus; the cattle slaughterhouse literature describes both the field technique and the sludge it produces.45 The consequence for classification is that the cake is not concentrated organic matter. It also carries the electrode metal, which adds to the ash fraction and dilutes the organic carbon per kilogram of cake. How much it dilutes, and how these proportions shift with influent strength, current density, and electrode material, are all empirical questions that a single batch cannot answer.
One further caution belongs here. Vendor literature in this sector often describes the electrocoagulation sludge as typically non-hazardous and cheaper to dispose of than chemical-coagulation sludge once dewatered.6 That may well prove true for a given installation, but it is the supplier’s characterisation, and a diligence-grade file should replace it with an independent analytical study on the actual cake itself, including free metals, before the claim appears in an investor-facing document.
The measurement panel
A credible panel is not exotic; it is a set of standard methods applied consistently, by an accredited laboratory, with the basis of every result stated. For the solid feeds, the established approach combines proximate analysis (moisture, ash, volatile matter, and fixed carbon), calorific value, and ultimate analysis, as in published characterisations of lignocellulosic biomass carried by ASTM methods.7 Carbon, hydrogen, and nitrogen are commonly determined instrumentally under the European solid biofuels standard EN 15104, with oxygen, then calculated by difference from the other results and the ash content, a convention which makes the ash determination part of the elemental result, rather than a separate detail.8 Ash behaviour matters independently: animal manures can carry high ash contents, which complicate thermochemical processing, and ash fusion temperatures are a routine part of assessing how any given solid will behave at temperature.9
Organic carbon needs methods suited to each phase. For the solid cake, EN 15936 has specified two methods for total organic carbon in sludge, treated biowaste, and waste by dry combustion,10 with a notable caveat that elemental carbon, carbides and some other carbon-bearing compounds are counted as organic carbon by method, so that an interpretation can be problematic where the sample contains relevant levels and they should be determined separately.11 That caveat is directly relevant to a thermo-chemical platform, where char and carbon-black formation are part of the story. For aqueous fractions, TOC is determined under the EPA 415 series and SW-846 Method 9060,12 and European guidance in EN 1484 covers both total and dissolved organic carbon in water.13 Dissolved organic carbon is, in practice, the fraction that passes a filter, which means the filtration step and its pore size are part of the result and need to be recorded.
Beyond carbon, the panel should also capture what might damage equipment or the output stream: sulfur, chlorine, nitrogen, alkali metals, and heavy metals in each feed, plus present aluminum or iron in the cake. These are not characterisation extras. They determine whether the syngas needs cleanup, which the platform’s design assumes it can avoid, and whether the solid residues have a disposal route.
Sampling is where campaigns quietly fail
Most failed characterisation programs do not fail inside the laboratory; they fail at the sample. A single grab sample from a single plant on a single day is an anecdote, and the variability documented for paunch makes it a particularly risky one.14 A defensible design samples across multiple plants, multiple seasons, and multiple lots within each plant; takes composite samples alongside discrete ones so that within-lot variation is visible; retains split samples for independent re-analysis; and runs field duplicates and blanks so that the laboratory errors can be separated from real variation. The EC cake needs the same treatment against its own drivers — influent strength, current density, electrode material, and operating hours — because a cake’s composition is a function of how it was made as much as of what was treated.
The reporting standard follows from design. Results should be presented as ranges with their sample counts, not as point values, and any figure that the commercial case relies on should be the conservative end of the measured range rather than the mean. This is the same convention that this series applies to every documented estimate: the number a diligence reader should carry away is the one that the data can defend on a bad day.
From characterisation to a closed balance
Once the feeds are characterised, the campaign moves to thermochemical testing in stages. Small-scale tests bound the achievable yield for each feed and for the blend of paunch and cake that the platform would actually see. A pilot-scale run on the actual platform, at a host plant, long enough to reach steady operation, then tests whether the bounded yield is achieved under real conditions. The governing test at every stage is an elemental mass balance: carbon, hydrogen, and oxygen entering through feed, steam, and water should be accounted for in gas, condensate, char, ash, and carbon-black outputs, within a closure band fixed in advance by an independent engineer.
The published literature gives that balance its useful external reference. Experimental hydrogen yields via biomass steam and oxy-steam gasification in downdraft reactors have been reported at up to roughly 45 grams of hydrogen per kilogram of biomass in one widely cited study,15 and up to about 104 grams per kilogram at high steam-to-biomass ratios in another,16 while modeling studies from steam gasification have projected figures in the range of 80–130 grams per kilogram.17 These are all different feedstocks, reactors, and conditions, and none validates one specific platform.
But they have defined an envelope. A reported yield that sits inside it needs ordinary verification; a reported yield that sits above it needs an explicit, metered explanation of where the additional hydrogen originated, which is exactly the separation between feedstock-derived and water-derived hydrogen described above. It follows that any water-splitting stage running in parallel with the gasifier must be metered on its own electrical and water inputs, so that its contribution is visible and is not credited to the feedstock.
There is also serious precedent for thermochemical waste-to-hydrogen at pilot scale: a UK instance on refuse-derived fuels reported bio-hydrogen yields of 33.6 mol/kg with hydrogen concentrations up to 82% in the product gas.18 As Article Four noted, that’s a different feedstock and not validation of GH2E’s yields. A more sobering precedent runs the other way. Plasco Energy, which raised $300 million to commercialise plasma gasification of municipal solid waste in Ottawa, filed for creditor protection in 2015 without having secured financing for its commercial plant, in the same coverage in which the city report noted that only four firms had documented the operation of a commercial-sized waste-processing facility.19 The lesson isn’t that plasma gasification cannot work, and the feedstock and business model were different. It is that buyers, lenders, and municipalities once asked for operating documentation, and the distance between the pilot result and a documented commercial operation is where capital has been lost.
Tokenisation: what the characterisation has to supply
The tokenisation question cannot be answered as one question, because “tokenising the platform’s output” covers at least three distinct things with different risk profiles, and the characterisation campaign feeds each differently. None of what follows here is legal advice, and the third layer in particular requires specialist counsel.
The first layer is the record. Here, the token or ledger entry is a tamper-evident, time- stamped anchor for data: the feedstock bundle, the sample identifiers drawn from it, the accredited laboratory’s certificate, the run log, and the meter readings. This is the lowest-risk and most immediately useful form, and it dictates how the characterisation data should be generated from the first sample. Every result needs a unique sample identifier, a stated basis (as-received, dry, or dry ash-free), a unique method code, the laboratory’s accreditation reference and units that never change between records. Any corrections should be added as superseding entries that reference the original rather than as overwrites, so that the history is visible. Data designed this way in Stage One could be anchored later at almost no cost; data designed loosely cannot be repaired afterwards.
The second layer is an environmental attribute: a hydrogen certificate, a carbon credit, or the water-reuse volume, represented digitally. Infrastructure for this exists in early form. JPMorgan’s Kinexys unit announced in 2025 that it was developing and testing an application to tokenise carbon credits at the registry layer, in partnership with S&P Global Commodity Insights, EcoRegistry, and the International Carbon Registry.2021 The same reporting records JPMorgan’s own concern that earlier tokenisation efforts by other players raised market-integrity problems, including the double counting and trading in credits which had already been retired.22 That concern bears directly on a platform that describes its outputs as stackable assets. If a single conversion batch is to generate hydrogen, carbon, and water attributes, the characterisation and metering data must also support allocating each attribute once and only once, and must make that allocation reconstructable years later. It is also worth stating plainly that these are exploratory industry pilots: nothing in them implies that any bank, registry, or security exchange has reviewed or endorsed GH2E, and the platform’s batch-record schemas were prototyped independently of those institutions’ infrastructure.
The regulatory side of the attribute layer is a data-quality problem. From California’s LCFS, the demand driver this series has already examined, requires the Tier 1 pathway applicant to submit at least three months of fuel production data and corresponding feedstock procurement records, which are provided to a CARB-accredited verifier,23 and the regulation’s requirements for pathways relying on waste or residue feedstocks include feedstock transfer documentation and access for CARB staff and verifiers to audit suppliers.24 Whether a plasma-gasification hydrogen pathway on animal-waste feedstock could be certified, and under which tier, is untested here and is a separate piece of work. The point for this article is narrower: the attribute is only as marketable as the measurement behind it, or measurement of the kind (described above) is what a verifier will ask to see.
The third layer is the financial claim: tokens representing ownership, revenue rights, or other interests in the platform or its projects. That layer sits inside those securities and commodities regulation in most jurisdictions. The correct sequencing for a diligence- minded issuer is simple: nothing at this layer should be offered before Stage Three data exists, because a financial instrument referencing an uncharacterized yield is a disclosure problem, well before it becomes a technology one.
Across all three layers sits where the blockchain literature calls the oracle problem: the ledger cannot natively check the real-world data written to it, so an on-chain record is only as reliable as the process that supplied it, and bad inputs produce bad outputs.25 Practitioners working on tokenized real-world assets have argued that independent auditors have a role in validating the off-chain components, including analog data that was never natively digital.26 For a characterisation campaign, that translates into concrete requirements: results signed by the accredited laboratory rather than simply transcribed by the operator, sensor data signed from the point of capture, a named independent witness for pilot runs, and a correction mechanism. Immutability slices both ways. It protects an honest record, but it can also preserve a wrong one.
Pass/fail, sequence, and who pays
The sequence that follows from all of the above has four stages, each gating the next. Stage One is feed characterisation across plants and seasons, the cheapest stage, and one that determines whether the later stages are worth running. Stage Two is bench-scale thermochemical testing to a bounded yield. Stage Three is the metered pilot at a host facility with an independent witness. Stage Four, recording or attribute issuance, begins only on verified Stage Three data. Funding can follow the same guardrails, with each tranche released on the prior stage’s result, which gives investors a defined exit if these characterisations disappoint and gives the platform a credible reason to publish the result either way.
Thresholds must be written down before the first sample is taken, by an independent engineer and not by the system. For illustration only, and not drawn from a standard, a defensible set might require elemental balances to close within a pre-agreed band, yield results to be reproducible across independent runs and across multiple feed lots, and the conservative end of the measured range, not the mean, to carry a commercial case. Setting and publishing the criteria in advance is the single most credible thing a platform can do, because it makes the eventual result mean something, regardless of its direction, context or ramification.
Where this leaves Track B
The campaign described here is, by the standards of the capital the commercial case contemplates, small, and it is the item on the critical path whose absence creates the largest uncertainty. That asymmetry is the argument for doing it first. It is also worth being exact about what completing it would and would not establish. It would convert Track B from contingent claims into a bounded, evidenced range for the specific feed- stock(s) tested. It would not demonstrate commercial-scale reliability, resolve the key regulatory questions around certifying a pathway, or establish a full revenue model; those remain separate gaps, several of which Article Four has flagged.
In the vocabulary this series uses for natural hydrogen exploration, the chemistry of a waste-to-hydrogen conversion technology is well supported in published literature: and the risk presence is low. Deliverability risk, meaning engineering and economic proof for these specific feedstocks on this specific platform, is what a characterisation campaign exists to retire, and until it has been commissioned, the right posture for a diligence reader is to treat Track B as a leading indicator with a defined test, and for tokenisation to be treated as infrastructure waiting for data worth recording.
The hydrogen transition waits for no one.
Sources and references
PMC (Energy Conversion and Management, 2016), “Inputs of biomass feedstocks for model validation,” Table 1 (ultimate analyses of manure, sludge, straw, and other feedstocks): https://pmc.ncbi.nlm.nih.gov/articles/PMC4892428/table/t0005
Frontiers in Sustainable Food Systems, “Current Progress in Optimising Sustainable Energy Recovery From Cattle Paunch Contents, a Slaughterhouse Waste Product”: https://www.frontiersin.org/journals/sustainable-food-systems/articles/10.3389/fsufs.2022.722424/full
PMC, “9 Reduce Reuse ReRumen: Variability in nutritive value of paunch manure”: https://pmc.ncbi.nlm.nih.gov/articles/PMC6663834/
Hellal et al., “Techno-economic evaluation of electrocoagulation for cattle slaughterhouse wastewater treatment using aluminum electrodes,” Sustainable Environment Research (2023): https://sustainenvironres.biomedcentral.com/counter/pdf/10.1186/s42834-023-00163-0.pdf
Indonesian Journal of Chemistry (Universitas Gadjah Mada), “Electrocoagulation for Cattle Slaughterhouse Wastewater Management: A Review”: https://journal.ugm.ac.id/ijc/article/download/102193/41481
Genesis Water Technologies, “Meat Processing Companies: Benefits Of Electrocoagulation,” Water Online guest column (October 30, 2020) (vendor source): https://www.wateronline.com/doc/meat-processing-companies-benefits-of-electrocoagulation-0001
Fajobi et al., “Investigation of physicochemical characteristics of selected lignocellulose biomass,” Scientific Reports (February 2022): https://doaj.org/article/460f8cfd89914048843716c0f41bd9d7
Celignis, “Analysis Package P33: Ultimate (Elemental) Analysis” (EN 15104:2011 method description): https://www.celignis.com/archive/package.php?value=19
Celignis, “Analysis of Animal Manures” (ash content and ash fusion analyses): https://www.celignis.com/archive/feedstock.php?value=11
CEN, EN 15936:2022, “Sludge, treated biowaste, soil and waste — Determination of total organic carbon (TOC) by dry combustion” (scope): https://genorma.com/en/standards/en-15936-2022
State Quality Inspectorate (Albania) standards listing for EN 15936:2012 (scope and note on elemental carbon and other carbon-bearing components counted as organic carbon): https://dps.gov.al/en/project/show/dps:proj:49329
U.S. EPA, “Total Organic Carbon in Water: EPA Methods 415.1, 415.2 and SW-846 Method 9060” (data quality indicators summary): https://19january2017snapshot.epa.gov/sites/production/files/2015-06/documents/415_1dqi.pdf
Analytik Jena, “DIN EN 15936” standards overview (references DIN EN 1484, guidelines for determination of TOC and DOC in water): https://www.analytik-jena.com/knowledge/standards-norms-productfinder/standards/din-en-15936/
PMC, “9 Reduce Reuse ReRumen: Variability in nutritive value of paunch manure”: https://pmc.ncbi.nlm.nih.gov/articles/PMC6663834/
Renewable Energy, “Hydrogen-rich gas production from biomass air and oxygen/steam gasification in a downdraft gasifier” (2007), vol. 32, no. 13: https://ideas.repec.org/a/eee/renene/v32y2007i13p2173-2185.html
Indian Institute of Science, “Oxy–steam gasification of biomass for hydrogen rich syngas production using downdraft reactor configuration,” International Journal of Energy Research: https://cgpl.iisc.ac.in/dasappa/img/pdf/journals/Oxy%E2%80%93steam%20gasification%20of%20biomass%20for%20hydrogen%20rich%20syngas%20production%20using%20downdraft%20reactor%20configuration,%20International%20Journal%20Energy%20Research.pdf
Naterer, Dincer, and Abuadala, “Exergy Analysis of Hydrogen Production from Biomass Gasification,” International Journal of Hydrogen Energy 35(10), 2010: https://doi.org/10.1016/j.ijhydene.2009.08.025
Materazzi, Taylor & Cairns-Terry, “Production of biohydrogen from gasification of waste fuels: Pilot plant results and deployment prospects,” Waste Management (2019): https://pubmed.ncbi.nlm.nih.gov/31279400
Ottawa Citizen, “City cuts ties with Plasco as company can’t pay bills” (February 11, 2015): https://www.pressreader.com/canada/ottawa-citizen/20150211/281487864777883
ESG Today, “JPMorgan Launches Blockchain Tokenization for Carbon Credits”: https://esgtoday.com/jpmorgan-launches-blockchain-tokenization-for-carbon-credits/
The Asset, “Tokenization seen as key to voluntary carbon market growth”: https://www.theasset.com/article-esg/54394/tokenization-seen-as-key-to-voluntary-carbon-market-growth
ESG Post, “JPMorgan pilots blockchain platform to tokenise carbon credits”: https://esgpost.com/jpmorgan-pilots-blockchain-platform-to-tokenise-carbon-credits/
California Air Resources Board, “Apply for an LCFS Fuel Pathway”: https://ww2.arb.ca.gov/resources/documents/apply-lcfs-fuel-pathway
Cal. Code Regs. tit. 17, § 95488.8, “Fuel Pathway Application Requirements Applying to All Classifications”: https://www.law.cornell.edu/regulations/california/17-CCR-95488.8
Chainlink, “What Is a Blockchain Oracle?” (definition of the oracle problem): https://chain.link/education/blockchain-oracles
LedgerLens, “Bridging the Gap: The Critical Role of Auditors in Tokenizing Real-World Assets”: https://ledgerlens.io/the-role-of-auditors-in-tokenizing-rwa


