Certification is a Day 1 design decision: what GHIS 4.0 left open
Ekansh Sharma · Founder, HyGOAT
Reviewed
ghis · rfnbo · ghci · certification · dmrv
The question from the floor
GH2 India hosted GHIS 4.0 in Mumbai on 1 September 2026. The closing session, Panel 6, covered scaling adoption from pilots to markets. Nishaanth Balashanmugam, CEO of GH2 India, moderated it. Sanjay Nagrare of Ocior Energy, Anuj Sharma of Waaree, Harish Jayaram of Hygenco, Sachin Chugh of Arup and Roochi Loona of DSK Legal spoke on it.
I asked one question from the floor.
How are we supposed to sell into the EU, Japan or Korea when RFNBO, CHPS and JCM compliance cannot be retrofitted after commissioning? What would it take to make certification a Day 1 design decision?
That is the question as recorded at the event. Its warning is sound, but “cannot be retrofitted” is too absolute: some contracts, monitoring and operating controls can change prospectively after the commercial operations date (COD), while missing historical proof may not be recoverable. A physical sale can remain possible subject to buyer and import rules; a clean claim, tender, incentive or project credit needs its own acceptance.
The panel and room agreed it deserved an answer. This post is my attempt.
The gap belongs to the whole industry: my field notes record no dedicated GHIS 4.0 session on certification, standards or MRV. That absence shows where attention sits in 2026.
Why the question stayed open
Offtake and the final investment decision (FID) settle whether a plant gets built. Market rules settle which claims, tenders or incentives can accept its molecules. Meters, contracts and data records settle what can be proved. Those choices are cheaper before FEED freeze.
My notes record the order: get the offtake, get the plant running, align to standards later. Rational for a first project; expensive on the second.
The rest of the agenda showed why the pressure sits there. Panel 2 covered the electrolyzer value chain: current density, stack life, imported catalyst coatings, nickel and membranes, and the open localisation question. Green ammonia tenders still await execution, with USD-INR volatility named as one cause. An Alfa Laval spotlight showed ultra-pure water from waste-heat recovery as a lower-LCOH route. One developer could not commit to LCOH because central, state and local subsidies remain unclear.
Every one is a real cost problem. Adding an evidence system after COD may not restore records never captured.
Seven measures for the operational evidence layer
Seven measures, captured at project start
The author recommends all seven evidence domains for a design basis. They are parallel measures, not a sequence.
Seven parallel evidence domains: boundary metering, power contracts, data custody, allocation, chain of custody, verification workflow, and a named owner.
Boundary metering
Power contracts written for correlation
Data custody
Batch-to-market allocation
Mass balance and chain of custody
Verification workflow
A named owner
Seven author-recommended operational evidence measures for a design basis. Delay can widen an evidence gap.
ScopeAuthor recommendations for a project evidence design basis. They are not a certificate or project assessment.
My answer: seven measures for the design basis. Some can change later; delay grows the evidence gap.
1. Boundary metering. Put meters on every stream that crosses the certification boundary: electricity per source, water, feedstock, product and by-products. Size the design for the temporal and geographic evidence required by each target market. The RFNBO rules contain route-specific correlation requirements.
2. Power contracts written for correlation. The power purchase agreement has to deliver audit evidence, not only electrons. For RFNBO, ask for time-stamped generation data, bidding-zone data, grid-mix access and the renewable asset's commissioning date. A bankable contract still needs auditable evidence.
3. Data custody. Map historian tags to each scheme's quantities. Fix time synchronisation, calibration records, append-only storage and retention. This is the monitoring plan a verifier asks to see. Writing it from a running plant costs more than writing it at FEED.
4. Batch-to-market allocation. Design batch and allocation records around each target scheme. A lot can support a claim only after that scheme accepts its inputs, boundaries and evidence; tags or low emissions alone do not establish acceptance. Raw history may not recreate missing proof.
5. Mass balance and chain of custody. Where a scheme requires it, batch identity has to survive storage, synthesis into ammonia or methanol, and shipping. Put scheme-specific tracking in the plant data model, not in a spreadsheet at the port.
6. Verification workflow. Decide verifier access, sampling rules and a pre-audit self-assessment before COD. Confirm the current GHCI cycle and ACV requirements with the responsible reviewer. If a project uses ISCC EU RFNBO pre-certification, treat it as a document-based readiness assessment. ISCC says it is not a certificate, certification decision or basis for a sustainability claim. It is not a guarantee of later operational certification.
7. A named owner. One person in operations owns the evidence pack. Not the EPC contractor, who leaves after handover. Not the auditor, who arrives once a year. Evidence without an owner decays at the first shift change.
What this means for a producer at FEED today
Separate market review paths
Three parallel market-review lanes for EU RFNBO, Korea CHPS, and India–Japan JCM.
RFNBO, CHPS, and JCM support different decisions and require separate checks.
ScopeAuthor comparison of distinct market paths. It does not establish approval or market access.
Start with the intended market and its methodology. The Green Hydrogen Certification Scheme of India (GHCI) and the EU's Renewable Fuels of Non-Biological Origin (RFNBO) rules use different eligibility and evidence requirements. A single emissions number does not establish acceptance across markets.
These routes answer different questions:
Korea's Clean Hydrogen Portfolio Standard (CHPS) uses Korean certification for power bids; India–Japan Joint Crediting Mechanism (JCM) is bilateral project crediting, not a hydrogen product certificate or general Japanese import prerequisite.
The strongest case for an RFNBO-designed plant is practical reuse: detailed meters, time-stamped renewable-power records, lifecycle inputs and traceability may already cover much of a second review. The operator still has to map its boundary, electricity accounting and evidence to that scheme and obtain its approval. The FEED question is whether each target claim can be proved per hour, batch and market, and what later changes can cover.
What this means for your project
If your project is at FEED, write the target market and evidence owner into the design basis. Map each electricity, water, feedstock and product record to the scheme that will review it, then test the declared inputs before the procurement package is fixed. A short project brief can make that first decision concrete; discuss your project with the founder-owned HyGOAT team.
Where HyGOAT sits, and where it does not
SCREEN checks declared project data against RFNBO and GHCI criteria before FID. It provides an indicative readiness result and names evidence gaps. SCREEN is not Korean CHPS or India–Japan JCM approval.
HyGOAT does not issue certificates. Formal verification and certificate issuance follow the applicable scheme and its authorised organisations. HyGOAT supports readiness preparation; a SCREEN result does not establish regulatory or buyer acceptance.
Thanks
Thanks to GH2 India for hosting GHIS 4.0, and to Nishaanth Balashanmugam for running Panel 6. Thanks to Sanjay Nagrare, Anuj Sharma, Harish Jayaram, Sachin Chugh and Roochi Loona. They treated a hard question from the floor as a fair one. A certification and MRV session would sit well on the GHIS 5 agenda.
Start with a screen
If your project is at FEED, start by identifying the intended market and the evidence owner. Choose GHCI Gap Analysis or RFNBO Export Screen. For help defining a bounded readiness engagement, discuss your project.
Sources and scope
The event observations are the author's field notes. Primary references are the MNRE GHCI scheme publication, EU RFNBO regulation, Korean clean-hydrogen certification notice, Korean power-bidding release, India–Japan JCM Rules of Implementation and ISCC's RFNBO pre-certification description. Apply the current rule version, transition provisions and project-specific methodology with the relevant reviewer.
Photographs from GHIS 4.0
