Over the past decade, solid oxide electrolysis (SOEC) development has focused on low-carbon hydrogen production and power-to-fuels applications.
Rising data center power demand and grid capacity constraints are renewing commercial interest in solid oxide fuel cells (SOFCs) as a distributed generation option. At commercial scale, feedstock selection and reforming architecture materially affect system efficiency, equipment requirements, and operating reliability.
Reforming pathways across diverse SOFC feedstocks
Many large stationary SOFC systems use natural gas as a methane feedstock. Depending on the system, methane may be converted through steam methane reforming (SMR), catalytic partial oxidation (CPOx), or a combination of reforming reactions.
In SMR, methane reacts with steam to produce CO and H₂:
CH₄ + H₂O ⇌ CO + 3H₂
SOFC systems can also be configured around other feedstocks. Some rural and agricultural sites have access to biogas containing methane and CO₂. The dry reforming of methane (DRM) reaction uses CH₄ and CO₂ to produce syngas through an endothermic pathway. Other mobile or remote applications may use liquid fuels such as methanol or jet fuel. The appropriate reforming conditions and catalyst formulation depend on the fuel. For jet fuel and other heavier hydrocarbon mixtures, coking resistance is an important catalyst requirement.
SMR, DRM, and CPOx differ in reaction chemistry and heat effects. Catalyst formulation and operating conditions must be matched to the feedstock, reforming pathway, and system requirements.
External, internal, and hybrid reforming architectures
Reforming location affects mechanical complexity, thermal efficiency, and system cost. Depending on the pathway and system design, reforming can occur in an external pre-reformer, in situ within the stack using an internal reforming catalyst, or in a hybrid configuration that combines external pre-reforming with internal conversion.
External reforming can use mature reactor designs that operate semi-independently of the SOFC subsystem, but it may add balance-of-plant equipment, pressure drop, heat loss, and cost depending on the system design.
Moving some or all reforming into the stack can reduce the size or footprint of external equipment and may lower CapEx and OpEx. When an SOFC operates in an exothermic regime, endothermic internal SMR or DRM can absorb part of the heat generated in the stack and reduce external heating demand. CPOx is exothermic and has a different thermal integration role.
This integration can increase design complexity. Local cooling and thermal gradients affect heat and mass transfer within the stack, so the thermal balance still depends on stack design and operating conditions.
Operating conditions that determine catalyst performance
Predicting catalyst performance beyond bench scale requires accounting for operating temperature, gas flow, feed composition, and system geometry. Together, these factors determine how much reforming can occur within the stack and how much pre-reforming is required upstream.
Operating temperature constraints
There is a commercial incentive to operate at lower temperatures because doing so can reduce operating costs. However, SMR conversion depends on both thermodynamic equilibrium and reaction kinetics, each strongly influenced by temperature. At sufficiently low temperatures, reduced equilibrium conversion and reaction rates may require additional external pre-reforming for the selected catalyst, feed, and system.
Flow rate effects on residence time and catalyst loading
For a given flow-path volume, increasing gas flow reduces residence time. Higher space velocity may therefore require changes to catalyst loading or system geometry. These parameters must be evaluated together.
Feed composition and steam-to-carbon ratios
Pipeline natural gas differs from controlled laboratory feeds; its composition and trace contaminants can vary within applicable specifications. Sulfur-bearing species may require upstream removal with a sulfur guard bed, depending on catalyst and stack tolerance. Separately, the steam-to-carbon ratio and operating conditions influence carbon-formation risk during SMR. The required ratio is system-specific and should be evaluated alongside feed composition and operating conditions.
Modeling and testing for commercial SOFC scale-up
Process modeling narrows the design space for a candidate feedstock and reforming architecture before hardware is built. It can estimate equilibrium conversion and heat duty, then evaluate whether the proposed residence time is practical at the target flow rate. Reactor testing under representative conditions shows whether actual catalyst performance matches those predictions.
One approved commercial-scale project is AEP Ohio’s planned 72.9 MW Bloom Energy SOFC installation for an Amazon Web Services data center in Hilliard, Ohio (source). The project will use dedicated natural gas infrastructure. At this scale, discrepancies between modeled and measured reactor behavior can force redesign after catalyst loading and equipment dimensions have been set.
