From turnaround RFI to aviation-system RFI
Rev 1 was strong on stand operations, geometry, power and bowser logistics — but it stopped at “can an LH₂ aircraft turn at this stand?” The lesson carried over from the CASCADE/HyFlux programme is that this is the wrong stopping point. A defensible sustainability case must also answer four further questions, and rev 2 adds the demand, provenance, lifecycle-boundary, infrastructure and residual-fuel layers needed to do it.
Two new fields on every line
Data owner — BRS Operations / Fire / Engineering / Sustainability / Commercial / ground handler / easyJet / DNO / fuel supplier. And where BRS does not hold it, we ask BRS to nominate the organisation or contact who does rather than return a blank. Several of the highest-value fields — fleet plans, hydrogen specification, connection capacity, fuel-farm operation — sit outside the airport team receiving this request.
Critical-path minimum — items 1–7 (was 1–3)
1Turnaround milestone data, current narrowbody operations
3–6 months of A-CDM milestones (AIBT, TOBT, AOBT, ground-handler activity timestamps where held) for A320neo/737-class turns, plus doors in use by stand type. Aggregated or anonymised is fine. Calibrates phase durations against BRS's real 25–35 min baseline instead of a generic industry one.
2Candidate trial stand — identity and geometry
Nominated stand (FlyZero guidance: off-pier pre-2030), dimensions, adjacent-stand spacing, jetbridge/stairs configuration. Settles whether the 20 m connect/disconnect zone fits without sterilising the neighbouring stand — which decides whether simultaneous or serial turns are BRS's planning case.
3Policy position on parallel activities during fuelling
Written steer — provisional is fine — from airside ops and the fire service on which cabin activities are permitted inside an 8 m flow zone with ATEX Cat-3 GSE. The single highest-leverage input in the model.
| Case | Fill | Stand time | Critical path |
|---|---|---|---|
| Serial, 0.8 kg/s (conservative) | 72.9 min | 89.0 min | disembark › clean › fill › board |
| Single 6" hose @ 5 m/s | 11.7 min | 27.8 min | disembark › clean › fill › board |
| Dual 6" + simultaneous | 5.8 min | 16.1 min | disembark › clean › board (fill leaves the path) |
4One-year flight and activity extract new
12 months of departures: date/time, route, aircraft family/type where available, scheduled and actual movements, stand allocation, turnaround duration, passenger/freighter flag, and seats or an agreed capacity proxy. This is the BRS equivalent of the CASCADE shrinking-activity pool — when an LH₂ flight replaces an A320-family Jet-A flight, the model removes that exact activity from the conventional fuel pool rather than adding hydrogen on top.
5Current airport energy and fuel baseline new
Electricity: half-hourly/hourly consumption if available, contracted product, annual grid import, on-site and backup generation, peak demand, substation constraints, location-based and market-based Scope 2 factors. Fuels: annual and peak-day Jet-A throughput, storage capacity, tanker deliveries, hydrant/bowser split, any SAF volumes. Without this we can size an LH₂ installation but cannot state what percentage of BRS aviation energy or emissions it actually changes.
6Hydrogen accounting and procurement basis new
Intended procurement standard or policy: renewable electrolysis, RFNBO compliance, grid-connected electrolysis, imported hydrogen, certificate/PPA treatment, additionality, temporal matching, geographic matching, fallback supply. And whether BRS wants WTW, TTW, or both reported. CASCADE testing showed the hydrogen-aircraft result moves dramatically with production CI and emissions scope — HyFlux will compute both bases, but BRS must say which it would defend. This decides whether the demonstrator is climate-beneficial rather than merely technically feasible.
7Trial metering and loss-fate capability new
What BRS could actually measure during a trial: delivered mass, storage inventory, boil-off generation, recovered vapour, vented mass, purge/chill-down mass, transfer loss, electricity to recovery/reliquefaction, aircraft uplift. Plus what atmospheric fate BRS and the fire service expect for transfer, purge and emergency-release streams. "Loss" must not automatically mean "emitted" — H₂ is an indirect GHG and the fate split is one of the most consequential unresolved areas in the HyFlux LCA work. This is what turns literature values into measured evidence.
Supporting requests — items 8–17
8Airside electrical capacity
Available kW at the candidate stand and NW-field compound, substation headroom, upgrade lead times. Determines whether a 5 MW-class liquefier is a substation project or a grid-connection project.
9Airport-wide spatial and safety envelope expanded
Protected development areas, safeguarding surfaces, underground services, drainage, fire-water provision, hazardous-area classification, emergency access, prevailing wind, public-road proximity, terminal/stand future plans, planning constraints, land ownership and lease boundaries — plus sensitive receptors, ecology and biodiversity constraints, watercourses, noise-sensitive areas and flood risk.
10Liquefaction concept — beyond "5 MW-class" new
Does the concept assume LH₂ delivered to site, or gaseous H₂ liquefied at BRS? Any existing position on on-site production. For on-site: electrical connection, operating hours, redundancy philosophy, reject heat, water and cooling demand, noise, maintenance access, availability expectation. This lets us compare conventional liquefaction against the BRSupercool route on the same functional unit rather than treating the latter as a lower-energy slider.
11Fuel-quality and operational assurance new
Receiving specification, sampling, custody transfer, contamination control, pressure and temperature requirements, tank conditioning, boil-off pressure management, refuelling rate, hose and connector assumptions, grounding/bonding, isolation time, emergency disconnect, post-transfer purge. Otherwise the turnaround result is driven by generic FlyZero assumptions rather than the chain BRS would approve.
12Airline and aircraft adoption data new
Provide or nominate owner. Anchor-carrier fleet plans relevant to BRS, aircraft family mix, route-range distribution, intended zero-emission demonstrator types, expected EIS windows, likely daily utilisation, overnighting pattern, and which services would realistically be first adopters. Prevents the trial case from silently becoming an all-flights hydrogen scenario.
13Residual-system interaction: Jet-A and SAF forecast new
Enough throughput and forecast data to model a mixed fleet. Hydrogen aircraft remove demand from the liquid-fuel pool before SAF acts, so each scenario must show LH₂ demand, residual Jet-A demand, SAF demand, electrical demand and infrastructure simultaneously — otherwise the project double-counts future abatement or over-sizes multiple fuel systems independently.
14Cost and commercial boundary new
Electricity tariffs and demand charges, connection costs, land value and lease assumptions, fuel-farm operating costs, bowser operating cost, staff cost categories, maintenance expectations, infrastructure hurdle rates, asset-life expectations. Agreed envelopes are sufficient where actuals are commercially sensitive — but a sustainability case that ignores recurring energy and abatement spend is incomplete.
15Early hydrogen movements per day
For buffer sizing and bowser count. FlyZero puts a small airport at 1–2 bowsers through 2035; we currently assume 2 flights/day against a 2-day buffer.
16GSE fleet and per-turn diesel burn
For the Scope 1 co-benefit, currently an assumed 35 L/turn — plus an introduction to the easyJet ACT trial data (1,721 kg CO₂e over six months), the best BRS-specific datum in existence.
17Airside access rules for a repeating bowser crossing
Gate, escort, or fixed transfer line at the fence boundary.
From the system-flow review — three boundary questions we were not asking
18Custody transfer point and metering boundary new
Our chain runs Avonmouth → aircraft as one continuous calculation with no ownership boundary in it. The system flow is right that there must be one. Where is custody transferred, and what is metered on each side? That boundary is where the mass balance closes — and it is the single point at which energy, CO₂e and cost should be charged exactly once. (Reviewing the operational twin, charging at both delivery and uplift overstated a day by roughly 57%.)
19Conditioning state required at the coupling new
We model fill as a mass flow (0.8–10 kg/s) with no required inlet state, so our fill rates are asserted rather than derived. What pressure and subcooling does the aircraft require at the coupling, what tank conditioning precedes it, and who owns that duty — supplier, airport or bowser?
20Which architecture phase is BRS planning for? new
Demonstrate (2026–30, road-delivered + bowser) · Prove (2031–35, on-site liquefaction) · Scale (2035+, backbone supply + hydrant). These are different topologies, not different parameter values. Our model currently has no time axis at all — every scenario is one static architecture. Knowing the target phase tells us which architecture is worth building.
From the CHOSAN ZEFD announcement — four questions that did not exist last week
21Project Acorn data — the trial you already ran new
Bristol Airport hosted Project Acorn in 2024, the UK's first airside hydrogen refuelling trial. That is measured hydrogen data, on this apron, already collected. Whatever exists — timings, mass, losses, procedures, lessons — is more valuable to us than anything we can assume, and it should have been the first thing we asked for.
22CHOSAN interface — who now owns the on-airport chain? new
With Unitrove delivering LH₂ refuelling and supply at BRS for the CHOSAN ZEFD flights, several of our assumptions may already be superseded by real infrastructure: compound siting, transfer mode, conditioning state and custody boundary. Can BRS introduce us, and should items 2, 9, 11, 18 and 19 be answered jointly with Unitrove rather than by BRS alone?
23Reference aircraft — regional first, or single-aisle? new
We model a 150-pax single-aisle at 3,500 kg per sector. CHOSAN flies a light aircraft on regional routes, and NAPKIN concludes regional-first. Which should BRSupercool treat as the primary case for BRS — and should we carry both? Note a dual-fuel combustion engine (Weslake) is a third architecture we don't currently model, and it produces NOₓ where a fuel cell does not.
24Leakage, NOₓ and contrails — coordinate with Badgerworks new
Badgerworks is running an environmental assessment for CHOSAN covering NOₓ, contrails and hydrogen leakage estimates — which is exactly the open question we flagged as our largest gap (loss fate, H₂ as an indirect greenhouse gas). A shared basis would strengthen both studies and give PXLICE an external reference rather than a HyFlux-internal assumption.
Following the ZEFD announcement
25Connect us to the three ZEFD projects supersedes
CHOSAN/Unitrove, ZeroAvia HyPRIME and Equilibrion are now answering much of what items 2, 9, 10, 11, 18 and 19 were asking. Rather than have BRSupercool produce a fourth independent set of infrastructure assumptions, the most valuable thing BRS can give us is an introduction to each — and a view on how the four projects should share a common assumption base. Specifically: Equilibrion's storage and supply findings, HyPRIME's real refueller cycle time and capacity, and Badgerworks' leakage/NOₓ/contrail basis.
Sustainability annex — identify datasets or owners (need not block G1)
Water and thermal rejection deserve explicit attention
If an advanced liquefaction route is compared against a conventional one, the model must capture not only kWh/kg H₂ but cooling duty, heat rejection, water demand, magnet/regenerator material requirements, availability, maintenance and plant footprint — otherwise an electricity advantage simply moves burden into another system.
Model note, stated plainly: brsupercool v0 implements a TPMS recuperator liquefier (6.03 kWh/kg, ε-based, REFPROP-verified) and a conventional 11 kWh/kg baseline. A magnetocaloric route and these non-energy burden vectors are a v0.5 extension, not a present capability.
Model mapping — six layers, one propagation chain
Rev 1 mapped requests to individual knobs (clean_parallel_to_fuel,
compound_to_stand_m, gse_power_kw). Rev 2 maps every dataset to one of six layers:
A single BRS flight record should ultimately propagate all the way through:
flight record → route eligibility → aircraft architecture → fuel demand → stand turnaround → airport LH2 throughput → electricity / liquefaction → WTW emissions → residual Jet-A / SAF pool → infrastructure → cost
That is the point at which BRSupercool becomes a full aviation-system demonstrator rather than an LH₂ ground-handling study.
Current coverage, honestly: layers 2–4 are built and passing tests (coil envelope, turnaround DES, five-stage cold chain with transfer). Layer 5 is partial — per-departure WTW Cat 3 + Cat 11 against a Jet A1 baseline exists, but the fleet-level residual-pool logic does not: today the model displaces one flight, not a shrinking conventional pool. Layers 1 and 6 are not yet built. Items 4, 5, 13 and 14 are what make them buildable.
The ask
Items 1–7 by 7 August, a named contact per function, and nominated external owners where BRS does not hold the data. Everything lands in versioned YAML in the repository BRS keeps, so each answer visibly moves the numbers BRS will defend in its own Scope 3 inventory.
Priority order if capacity is limited: real flight/fleet activity (4) · airport energy and fuel baseline (5) · hydrogen procurement and accounting basis (6) · loss and metering capability (7). Those four connect the physical trial work to the sustainability and aviation-system model developed through the CASCADE/HyFlux programme.