Why Simple Fuels Win at Sea: Assessing LNG SOFCs, Hydrogen, Sails, and CCS Against Practical Needs

Support CleanTechnica’s work through a Substack subscription or on Stripe.

I was recently asked by someone in the maritime industry whether Bloom Energy’s solid oxide fuel cells could play a real role in maritime propulsion. The question was prompted by the announcement from Ponant, GTT, and Bloom describing a concept cruise vessel for 200 passengers that would combine hard wing sails, LNG fed Bloom solid oxide fuel cells, hydrogen fuel cells and onboard carbon capture. It is an absurdly ambitious and frankly ill considered configuration. The idea of combining multiple novel, dead end technologies for a single vessel prompts us to look at what the maritime sector actually needs, what shipyards can build, what operators can maintain and what ports can support. Curiosity is a good starting point, but curiosity needs to be followed with grounded engineering and clear evaluation of risk.

When I was first asked about putting Bloom fuel cells running on LNG on a ship, my immediate response was: “FSM on a stick, can no one involved with that initiative do math?” The problems are so obvious that it’s remarkable to me that anyone is considering this, but then to find that they were also going to put in hydrogen, sails and CCS gobsmacked me.

Maritime power is a difficult problem because ships are floating industrial sites that run continuously with very high reliability requirements. They carry people, goods and fuel across thousands of kilometers in conditions that can shift from calm to dangerous in minutes. Every part of the drivetrain has to tolerate constant vibration, corrosion, thermal cycling and mechanical wear. Power density matters because machinery spaces need to fit into hull volumes that also serve cargo, passengers and crew. Reliability matters because unplanned downtime at sea is unacceptable. Fuel logistics matter because ports have to supply energy at the right scale without excessive cost or complexity.

All of that has shaped the existing technology landscape. Dual fuel LNG engines—problematic as they are due to upstream and at sea methane slippage—methanol engines, hybrid electric systems and energy storage are climbing adoption curves and displacing straightforward diesel drive trains. Electric solutions are spreading across ferries, offshore support vessels, tugs, inland cargo and other classes of shipping, starting small but growing larger rapidly. Wind assistance is being tested on a small subset of ships but faces significant limits because the best use cases require low operational flexibility, large deck areas and predictable wind regimes. These constraints define the reality in which any new option must compete.

Bloom’s SOFCs are described by them as high efficiency—a deeply misleading claim—electrical generators that run on natural gas or biogas—almost entirely natural gas, with emissions higher than California’s grid emissions when I first encountered them 15 years ago—and produce a relatively concentrated CO2 exhaust. They operate internally at around 800° C which allows methane to reform into hydrogen and carbon monoxide inside the stack. That gives electrical conversion efficiencies close to 60% lower heating value with low nitrogen oxide emissions, but only for the hydrogen, not all of the energy in the natural gas. More on that later. The cells are packaged in modules that weigh around 15 tons and produce roughly 325 kW. The modules are bulky for their output and run very hot which means insulation and thermal management dominate the physical design. About half of the energy in the fuel leaves as high grade heat, and more leaves as CO2. That heat is useful in some stationary applications but creates significant engineering requirements in confined machinery spaces.

The modules also degrade rapidly. Field data indicates that median replacement cycles are around five years depending on duty and cycling. As they have mostly been used as backup generators historically, a median failure rate of five years is pretty stunning. In stationary environments, the modules can be swapped with cranes. On ships, these swaps would require large soft patches in the decks and heavy lifts through the interior of the hull, which isn’t how ships are designed and built today. The technology works in stationary applications with abundant space and controlled environments but changing context to a ship introduces new questions about access, maintenance and overall system integrity.

Integrating Bloom modules into a ship exposes an engineering challenge that looks like a fundamental mismatch between the SOFC architecture and the internal geometry of most commercial hulls. Ship engine rooms are designed around reciprocating engines or gas turbines which have well understood footprints, ventilation needs and maintenance pathways. The hot boxes of a Bloom stack cannot be crowded together like engine blocks because they need air gaps, insulation and controlled ventilation paths. Ducting 800 C exhaust through confined machinery spaces while keeping structural steel within allowable temperature limits adds complexity. Keeping crew safe around hot equipment also adds constraint.

Each module must be placed so that it can be removed on a predictable path to a deck opening that can support a 15 ton load. The internal supports and overhead clearances needed for those moves reduce the available design freedom for other systems. Power density becomes a serious obstacle. A ship with a 10 MW requirement might need thirty or more Bloom modules arranged in banks, with substantial spacing and significant cooling and ventilation. The result is a machinery volume several times larger than a traditional dual fuel engine installation. This competes with revenue generating space and weights the vessel in ways that may affect stability.

And it has to be centralized and inside the ship. These are large and heavy fuel cells, so putting them on deck where thermal management might be easier would make the ship top heavy, which is not remotely a good thing. They have to be in the same space because distributing them throughout the ship and just delivering electricity to the engine room would multiply the complexity and cost of delivering the LNG and CO2 and of the massive thermal management that’s required for 800° Celsius fuel cells.

Then there’s the energy density and their fuel cell efficiency claims, which don’t stand up to the slightest scrutinty. If you treat LNG as pure methane, about 55% of its chemical energy comes from the four hydrogen atoms and about 45% from the carbon, so if you strip out the carbon up front for capture and only feed the hydrogen energy into a Bloom style SOFC you are already throwing away almost half of the fuel’s potential before you start. Take 100 units of LNG energy. Roughly 55 units are tied up in hydrogen and 45 units in carbon. A solid oxide fuel cell running at a generous 60% electrical efficiency on that hydrogen fraction will turn about 0.60 × 55 ≈ 33 units of the original 100 into electricity. A good two stroke marine engine on VLSFO can reach about 50% efficiency on the full fuel energy, so the same 100 units of VLSFO energy would deliver roughly 50 units of shaft power.

On a volumetric basis, VLSFO sits around 41 MJ/L while LNG is closer to 23 MJ/L, so to match the chemical energy of 1 L of VLSFO you need about 1.8 L of LNG, and after discarding the carbon energy and running the hydrogen through the SOFC you end up with roughly two thirds of the electrical output that a VLSFO engine gives per liter of original tank volume. That means you actually need about 2.7 L of LNG to match the energy in the VLSFO. Then there are the LNG storage and handling components, including insulated tanks, thermal management, cryogenic liquid movement and regasification equipment. That’s 1.3-1.5 times more volume than the straightforward tanks and pumps required for VLSFO and other fuels which are liquid at room temperature.

Biomethanol has even lower energy density than LNG, around 16 MJ/L or roughly 40% of VLSFO, so you clearly need bigger tanks for the same range, but in a standard maritime reciprocating engine you use both the hydrogen and the carbon energy without throwing 45% away and you do it with simple liquid fuel systems, conventional engines and well understood maintenance, which makes the biomethanol pathway much more attractive than an LNG to hydrogen to SOFC route that is both bulkier and far more complex.

Layering hydrogen fuel cells on top of an LNG fuel system creates another set of issues. Hydrogen has very low volumetric energy density and requires either large high pressure tanks or cryogenic liquid storage. Both options impose heavy penalties on naval architecture. Marine hydrogen systems must meet strict safety rules including double wall piping, specialized sensors and large exclusion zones. Combining hydrogen storage with LNG storage increases the number of gas systems onboard and the interactions between these systems must be engineered with redundancy and isolation.

A hydrogen fuel cell is a separate prime mover with its own control and safety environment. Sharing loads between the SOFCs and the hydrogen cells adds electrical integration complexity and operational modes that crews must understand and manage. Training standards for hydrogen at sea are still emerging, not that there are going to be a lot of hydrogen vessels as it’s a dead end economically and systemically. Integrating two cryogenic or high pressure gases into one vessel stretches what operators and ports can support.

Onboard carbon capture compounds the problem. A Bloom exhaust is concentrated relative to a combustion exhaust which helps, but the steps needed to go from a hot anode exhaust to a high purity CO2 stream still require substantial cooling, separation equipment and compression. Compressing CO2 to storage pressures consumes several percent of the ship’s electrical output, compounding the terrible efficiency. The equipment is industrial in scale and includes heat exchangers, knock out drums, blowers and compressors.

Liquid CO2 storage tanks take up significant hull volume and add weight that varies over the voyage as CO2 accumulates. Offloading liquid CO2 at port requires specialized equipment and trained personnel. Only a handful of ports worldwide have CO2 handling infrastructure in plan, and only a few places have any operational liquid CO2 facilities, all related to the Norwegian Northern Lights carbon capture and sequestration scheme I assessed recently. Each of these steps adds risk of leaks, failures and operational delays. Ships trading between ports without CO2 infrastructure would carry ballast CO2 until they reach a suitable terminal which impacts payload. A single malfunction in the capture plant could force the ship to vent CO2 or derate its power output which undermines reliability.

When all of these systems are combined, the risk multiplies. Hard wing sails require structures that affect ship stability, wind loading, port operations and maintenance. The Bloom SOFCs impose space, heat and maintenance requirements that are foreign to shipyards. Hydrogen fuel cells and storage introduce new safety envelopes and operational rules. Carbon capture and CO2 storage impose industrial plant complexity on a vessel that needs to operate safely at sea.

Each component is novel at maritime scale and each one demands specialized knowledge. The combined system becomes a collection of first of class technologies that interact in ways that are not remotely validated. The build schedule would be vulnerable to delays because shipyards would have to create new fabrication processes and handle new equipment. Classification societies must approve designs that have no precedent. Insurers must model risks that are not well understood. Those last two may scuttle this ship before it ever leaves the drafting table.

Operational reliability would be difficult to guarantee because a fault in any subsystem affects the whole vessel. Crew training expands beyond conventional marine engineering into cryogenics, high temperature electrochemistry, gas handling and industrial capture plant operation. It’s likely that the engineering team of such a ship would triple in size just to keep the bits in trim during journeys, likely with significant continuous land side monitoring and oversight, and land side specialist assistance on a regular basis as problems inevitably emerge. It is difficult to see how such a stack of novel technologies can deliver consistent commercial service.

In contrast, a hybrid biomethanol and battery drivetrain fits into existing maritime practice much more easily. Methanol engines from MAN and Wärtsilä are commercially available and are already being installed on container ships, tankers and ferries. Methanol is a liquid at ambient temperatures and can be stored in simple tanks that fit into hull spaces that would otherwise carry fuel oil. The fuel systems do not require cryogenics. Safety systems for methanol are extensions of existing rules for liquid alternative fuels. It’s fairly nasty stuff, but on the same scale of nasty as VLSFO, so as long as crew aren’t drinking it or bathing in it, the risks are minor and easily managed. Biomethanol can decarbonize the fuel supply without changing the ship as the carbon comes from the atmosphere.

Batteries integrate smoothly with electric drives and provide peak shaving, hotel load support and zero emission operation in port. They can fit into modular rooms distributed throughout the ship for appropriate ballasting with high voltage protection and active cooling. For container ships and roros, they  can be containerized and dropped into existing container slots at the bottom or parked evenly along the inside of the lowest roro deck for good ballasting and easy connection to power delivery mains, similar to plugging in refrigerated containers. Maintenance follows known patterns for power electronics and battery management systems. Ports can deliver methanol using tank trucks or bunkering barges without large new infrastructure commitments, although the tanks and pumps aren’t interchangeable with VLSFO. Charging batteries at port involves electrical connections that can be built out gradually, especially for containerized batteries. The machinery space is compact, the technology is stable and the crew skill requirements align with current marine engineering training pipelines.

The contrast between the two approaches becomes clear when looking at what the maritime sector needs over a vessel’s lifespan. A ship must be buildable in a commercial shipyard without long delays. It must be class approved without extraordinary exemptions. It must be insurable at reasonable rates. It must operate reliably in a wide range of ports and conditions. It must be maintainable by crews trained within standard maritime programs.

Hybrid biomethanol and battery ships meet these requirements. The multi technology stack proposed for the Ponant concept does not. The Bloom SOFCs create physical and thermal challenges that shipyards are not set up to handle. Hydrogen introduces storage and safety burdens that are not suitable for long range applications. Carbon capture adds large equipment needs and creates new logistics that ports cannot yet support. Hard wing sails offer limited benefits because wind assistance only works well for specific vessel classes with predictable routes and generous deck areas. Even there, the gains are modest and often do not justify the structural and operational compromises. Wind assistance faces constraints related to diminishing returns, variable wind fields and interference with cargo operations as I have discussed elsewhere. These limitations make sails an interesting supplement in a few narrow niches rather than a core propulsion option.

Regulatory pressure also shapes the fuel choices available to cruise and expedition vessels. Emission Control Areas in the Baltic, North Sea, North America and the US Caribbean require very low sulfur fuels at 0.10% and impose strict NOx limits for newer ships. China’s domestic control zones and the emerging Mediterranean ECA follow the same pattern. Operators meet these rules by switching to VLSFO, marine gas oil, LNG or methanol, or by adding exhaust treatment. Heavy fuel oil is not an option inside these zones without extensive abatement. For a ship trading regularly in ECAs, simple liquid fuels like methanol or compliant distillates meet the regulatory bar with minimal operational change, while LNG requires more space and complexity and multitechnology solutions like SOFCs, hydrogen fuel cells and onboard capture introduce risks that are not rewarded by the regulatory structure.

I was deeply unimpressed with Bloom’s fuel cells 15 years ago. They were showing up as backup power generators in California, but they just weren’t delivering any real value that I could see. Their emissions were higher than California’s grid emissions, they were throwing away a lot of the energy in the natural gas they were consuming, they were introducing additional safety concerns due to their heat, their stacks didn’t last—and still don’t—and their claims of being an environmentally friendly solution were clearly nonsense. They rode they hydrogen hype recently, but 99.9% of their business is still turning natural gas into energy. Now they are riding the AI data center hype cycle to the benefit of their stock price, with several big deals inked for their rapidly degrading stacks running 24/7. That doesn’t bode well for their clients, but presumably they’ve done their due diligence and priced frequent replacement in against their belief in a world where every man, woman and child is paying thousands monthly for AI pictures of kittens. None of the glaringly obvious limitations of SOFCs or Bloom’s misleading claims of environmental friendliness and efficiency have changed. When they actually try to extend into the maritime industry, where they are so clearly unfit for purpose, I’m even less impressed.

Decarbonization in shipping will be shaped by the solutions that are buildable, maintainable and scalable. The sector has little tolerance for systems that depend on coordinated breakthroughs across many domains at once. Bloom’s SOFCs do not fit the maritime environment in terms of size, heat, access needs or lifecycle logistics. Hydrogen fuel cells remain constrained by storage density, safety rules and the unbending realities of their bad economics. Onboard CCS adds mass, volume and operational risk to vessels that already operate under tight constraints. These pathways do not align with the needs of operators or with port capabilities. Methanol engines and batteries do.

They fit the physical and operational realities of ships. They integrate with ports that are already expanding methanol supply chains and grid connections. They support incremental improvement instead of wholesale reinvention. They offer a path that can scale across vessel classes and regions without forcing operators into untested configurations. Shipping tends to adopt technologies that lower operational risk and simplify compliance. The solutions that succeed will be those that deliver reliability and practicality first and emissions reductions as a consequence.