A new GCMD/BCG fleet-economics model makes a clean argument: engine orders placed this decade lock in most of the propulsion capacity the 2050 fleet will have, but capacity to burn a new fuel and actually burning it are different things — that gap is closed by carbon pricing and, more fundamentally, by the cost of the hydrogen every e-fuel is built from. The report never mentions hydrogen-fuelled ships. It doesn’t need to — its entire adoption forecast collapses to a single number: the Levelised Cost of Hydrogen. We happen to have just spent a week pricing that number from the supply side.
⚡ TL;DR
- What: GCMD and BCG's joint fuel-pathway model finds that more than half of the 2050 fleet will come from vessels ordered before 2035 — engine choice today locks in most of what's technically possible for decades.
- The real constraint isn't engines, it's economics: at today's carbon price (~$380/tCO2e), methanol dual-fuel engines are ~10% of fleet capacity but only 2% of actual fuel consumption. Reaching 61% new-fuel consumption by 2050 requires carbon pricing near $700/tCO2e.
- The pivotal variable: at $2/kg LCOH, methanol+ammonia could reach 36% of global fleet energy by 2050; at $3/kg, only 4%. A $1/kg swing in hydrogen production cost is the difference between transformation and irrelevance.
- Our cross-check: the Fraunhofer RE-Chain study we covered last week puts real 2030 LCOH at €3.50-4.00/kg (~$3.8-4.3), above BCG's pessimistic case — and the 2050 LCOH of €2.00-2.50/kg (~$2.2-2.7) only reaches BCG's favourable band in the second half of the 2040s, not before.
- Watch for: whether BCG's $2/kg scenario is achievable on any published timeline, or whether it's a target rather than a forecast.
What the Report Actually Argues
GCMD (Global Centre for Maritime Decarbonisation) and BCG built a joint model that stress-tests shipping’s fuel transition against fleet renewal rates, carbon pricing, and fuel economics. The headline finding is about timing: at roughly 4% annual fleet renewal, more than half of the vessels sailing in 2050 will have been ordered before 2035. Whatever engine technology gets specified on newbuild contracts over the next decade is, in large part, what the 2050 fleet has to work with — there’s no fast do-over.
But the report’s more interesting finding is about the gap between capacity and consumption. A dual-fuel engine can run on methanol or ammonia, but whether it actually does depends entirely on whether that fuel is cheaper than the compliance cost of staying on conventional fuel. In the base scenario, with the IMO’s Tier-2 penalty at $380/tCO2e, methanol dual-fuel engines already represent about 10% of installed fleet capacity — but only 2% of actual fuel consumption. Owners are buying optionality, not commitment. Getting new fuels to 61% of fleet energy by 2050 requires carbon pricing to climb toward $700/tCO2e, a level GCMD’s own CEO, Professor Lynn Loo, calls “particularly challenging in today’s geopolitical environment.”
The Number That Actually Decides This
Strip away the carbon-price scenarios and one variable does more work than any other in the model: the Levelised Cost of Hydrogen. Both e-methanol and e-ammonia are hydrogen derivatives — hydrogen produced from renewable electricity, then combined with captured CO2 (for methanol) or nitrogen (for ammonia). Their production cost is a direct pass-through of whatever the underlying hydrogen costs.
BCG’s sensitivity analysis is stark:
| LCOH | Methanol + ammonia share of global fleet energy demand |
|---|---|
| $2/kg | ~36% |
| $3/kg | ~4% |
A one-dollar swing in hydrogen production cost is the difference between e-fuels becoming a meaningful share of the fleet’s energy mix and staying a rounding error. BCG’s Anand Veeraraghavan frames this directly: “The maritime fuel transition is shaped as much by policy and cost uncertainty as by technology readiness.” On this specific variable, “cost uncertainty” understates it — this is the single lever the entire adoption curve pivots on.
Cross-Checking Against Real Cost Data
This is where our own recent work is directly useful. Last week we read the full Fraunhofer UMSICHT RE-Chain study — a bottom-up engineering cost model for hydrogen imports from Spain and Oman into Europe, commissioned by the ports of Amsterdam and Duisburg. It publishes its own Levelised Cost of Hydrogen projections, independent of BCG’s model and built from a completely different methodology (production-cost engineering rather than fleet-adoption economics):
| Year | RE-Chain LCOH, Oman (Trend) | RE-Chain LCOH, Spain (Trend) | ~USD equivalent (Oman) | BCG’s threshold this sits nearest |
|---|---|---|---|---|
| 2030 | €3.50/kg | €4.00/kg | ~$3.8/kg | Above the $3/kg (4% share) case |
| 2035 | €3.00/kg | €3.50/kg | ~$3.2/kg | Above the $3/kg case |
| 2040 | €2.50/kg | €3.00/kg | ~$2.7/kg | Between $2-3/kg |
| 2045 | €2.25/kg | €2.75/kg | ~$2.4/kg | Between $2-3/kg, closing |
| 2050 | €2.00/kg | €2.50/kg | ~$2.2/kg | Approaching the $2/kg (36% share) case |
(EUR-to-USD conversion here is an approximate ~1.08 rate for orientation, not a precise real-time figure — treat the USD column as indicative, not exact.)
Two independently-built cost models — one a top-down fleet-adoption framework, one a bottom-up engineering supply-chain cost estimate — land in the same neighbourhood, which is itself a useful piece of triangulation. But read together, they produce a more sobering timeline than BCG’s headline numbers might suggest on their own: RE-Chain’s own 2030 LCOH sits above BCG’s pessimistic $3/kg case, not between the two scenarios. By RE-Chain’s numbers, hydrogen doesn’t get cheap enough to approach BCG’s optimistic 36%-adoption band until somewhere in the second half of the 2040s — a full decade-plus later than the “engine choices made today” framing might imply.
This doesn’t mean BCG’s $2/kg scenario is wrong — it’s a sensitivity case, not a forecast, deliberately chosen to bound the range. It means that on the best independent engineering cost estimate we’ve reviewed, the favourable end of BCG’s own range is a 2045-2050 outcome, not a 2030s one. The dual-fuel engines being ordered now will spend a long first decade running mostly on conventional fuel, exactly as the report’s own 2% consumption figure for today’s methanol-dual-fuel fleet already shows.
Why This Matters for Hydrogen Specifically
Every reader of this site who’s watching e-ammonia or e-methanol adoption as a proxy for green hydrogen demand should take the LCOH sensitivity seriously as the actual bottleneck — not engine availability, not classification society rules, not even carbon pricing on its own. Cheap hydrogen production is the precondition for all of it. That’s consistent with what we found applying the same RE-Chain data to LH2 as a direct bunker fuel: landed LH2 fuel cost came out around €7-8/kg for 2030, which includes liquefaction, storage, and shipping on top of the LCOH figures in this comparison. Methanol and ammonia avoid the liquefaction and cryogenic-storage cost stack entirely — which is exactly why BCG’s model has them clearing a much lower total cost bar than direct hydrogen fuel does, even though both are gated by the same underlying LCOH number.
Challenges and Open Questions
- The EUR/USD conversion here is approximate. A precise comparison would need the actual exchange rate assumptions BCG used in its own model, which the press release doesn’t disclose.
- Geography isn’t identical. RE-Chain prices Spain and Oman specifically; BCG’s LCOH sensitivity is presented as a generic global variable without naming source regions. Different renewable-electricity-cost geographies could shift either side of this comparison.
- BCG’s model and RE-Chain’s model measure different things at the margin — fleet-wide adoption economics versus a specific two-corridor import cost — so this cross-check is directional triangulation, not a rigorous unification of the two studies.
- Biogenic CO2 and bio-methanol costs, which BCG flags as a separate major uncertainty for methanol specifically, aren’t something we have independent cost data to check against yet.
- Carbon price feasibility remains the other half of the equation. Even if LCOH reaches $2/kg on schedule, BCG’s own model says meaningful adoption still requires carbon pricing near $700/tCO2e — a policy outcome, not an engineering one, and arguably the harder of the two variables to forecast.