Düren’s Hydrogen Bet: The Math Behind a Looming Liability

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

Doing the math on Aberdeen’s abandonment of hydrogen buses led to a question from someone living in Düren, Germany about their hydrogen program. On the surface the situations look different. Aberdeen was a city trying to build a hydrogen ecosystem largely on its own. Düren, a district of about 270,000 sits inside North Rhine-Westphalia, near Cologne, Düsseldorf and the wider Rhine-Ruhr industrial cluster. It has federal funding, a regional hydrogen narrative and a new 10 MW electrolysis plant under construction at Brainergy Park in Jülich. The question is whether that embedding changes the math.

Aberdeen provides a completed case study. Twenty five hydrogen double deck buses were purchased at roughly £500,000 each. The Kittybrewster refueling station cost on the order of £1m to build and was recording about £325,000 per year in operating costs, roughly 30% of capex annually. That aligns with what I previously calculated for California’s high pressure stations, where compression, storage and dispensing hardware with seal integrity loss drives 10% to 30% of capex per year in O&M. Using realistic electricity pricing and full system electrolysis assumptions, hydrogen in Aberdeen was landing in the £20 to £25 per kg range. At 6 to 7 kg per 100 km, energy cost alone was £1.3 to £1.7 per km. Diesel was about £0.7 per km in fuel and battery electric about £0.14 per km in electricity at Scottish business tariffs. The fleet sat parked for more than a year before being retired. That is what low utilization and high fixed cost look like in practice.

Düren’s hydrogen story began with five Caetano fuel cell buses entering service in 2022. Local reporting later noted those first five were being withdrawn because of technical defects, a not unusual occurrence. The district expanded the fleet to about 20 Solaris hydrogen buses. Hydrogen is dispensed at a Shell forecourt site in Düren that is built and operated by H2 Mobility Deutschland, the same network operator that has closed 22 passenger car hydrogen stations across Germany due to low demand. The Düren station is a mixed 350 bar and 700 bar site intended to serve buses, trucks and passenger vehicles. Passenger fuel cell vehicle adoption in Germany remains small, so the bus fleet is the only offtaker of scale.

Parallel to the forecourt station, the district formed HyDN GmbH with Messer to build a 10 MW green hydrogen production plant at Brainergy Park in Jülich, about 27 km from the transit bus garage. Public documents list total investment of about €35m with €14.7m in federal funding. Early planning documents indicated production could begin in 2024 at the earliest. Later local reporting has shifted expected startup into early 2026. Participation reports from the district acknowledge cost pressure from supply chains and state that EPC contracts expose the project company to actual construction cost plus a 10% surcharge. No formal overrun number has been published, but the timeline has slipped.

The Jülich plant was not conceived to serve 20 buses. Early framing pointed to hydrogen trains, heavy goods vehicles, municipal fleets, passenger cars, regional hydrogen rail and broader industrial uptake as anchor customers that would absorb hundreds of tons per year. The hydrogen train program in the region has been shelved, passenger fuel cell vehicle adoption in Germany remains marginal, and there is no public evidence of a large industrial offtaker in the immediate vicinity committing to the plant’s output.

Rurtalbus itself already operates battery electric buses, and the large regional operators in Cologne, Düsseldorf and the Ruhr are scaling battery electric fleets at far greater volume than hydrogen. Given that procurement pattern, and given the operating cost gap between battery electric and hydrogen at realistic utilization levels, it is difficult to construct a credible pathway from roughly 20 hydrogen buses today to the 150 or more that would be required to materially lift plant utilization. Without new classes of customers, the likely steady state is a small hydrogen fleet served by a large production asset built for a market that has not emerged.

The 10 MW plant is designed to produce up to 180 kg of hydrogen per hour. At continuous operation that is 4,320 kg per day and roughly 1,577 tons per year. HyDN’s own documents describe a long run target of about 1,000 tons per year by 2028. Now look at bus demand. A hydrogen bus consuming 6 kg per 100 km and traveling 60,000 km per year uses 3.6 tons per year. Twenty buses consume about 72 tons per year. Against 1,000 tons of capacity that is 7.2% utilization. Against the engineering maximum it is about 4.6%. A 10 MW electrolysis plant serving only 20 buses is operating at single digit percentage utilization.

When utilization is that low, capex dominates. Annualizing €35m over 10 years at a 7% discount rate produces an annual capital charge of about €4.97m. Divide that by 72,000 kg per year and capex alone is €69 per kg. Electricity at 65 kWh per kg and €0.15 per kWh adds €9.75 per kg, a bit of a rounding error. That still ignores maintenance. Reference class forecasting from combined electrolysis and refueling projects suggests that 25% to 45% of total capex is tied to compression, storage and dispensing. Applying 10% to 30% per year O&M to that portion and 4% to the electrolysis balance yields total hydrogen cost in the €105 to €140 per kg range when only 20 buses are offtakers. At 6 kg per 100 km that is €6.3 to €8.4 per km in fuel cost. Diesel at €1.73 per litre and 0.5 L per km is about €0.87 per km. Battery electric at 1.5 kWh per km and €0.17 to €0.18 per kWh is about €0.25 to €0.27 per km. Even before adding vehicle capex, the gap is extreme.

Emissions tell a more nuanced story. Diesel buses at 0.5 L per km and 2.65 kgCO2 per litre emit about 79.5 tons CO2 per year at 60,000 km. Battery electric buses at 1.5 kWh per km on Germany’s 2023 average grid intensity of 380 gCO2 per kWh emit about 34.2 tons per year. Gray hydrogen produced from natural gas at about 11 kgCO2 per kg H2 yields about 41.7 tons from hydrogen production per bus year. With 5% hydrogen leakage and a GWP100 of 11.6, leakage adds about 2.2 tons, bringing the total to about 43.9 tons per year. Using a GWP20 of 37.3 raises leakage impact to about 7 tons and total to about 48.8 tons per year. If Jülich electrolysis runs on the average German grid instead of certified renewable power, the emissions are higher. Fifty five kWh per kg at 380 gCO2 per kWh, just the electrolysis, yields about 20.9 kgCO2 per kg H2. At 3,600 kg per year that is roughly 75 tons from electricity alone. Including 5% leakage raises that to 81 to 86 tons per year depending on GWP horizon. In that case hydrogen is roughly on par with diesel in operational emissions. At 65 kWh to accommodate balance of plant, compression and recompression, it’s over 100 tons CO2e for electrolysis, well above diesel.

Note that Germany is working to decarbonize its grid and has seen significant reductions over the past 30 years, so these emissions will diminish, but battery electric will always be lower than green hydrogen because of the much greater efficiency of the drivetrain.

Regionally, large transit operators such as KVB in Cologne and Rheinbahn in Düsseldorf are scaling battery electric buses in significant numbers. Nationally, the Federal Court of Auditors has called for a reality check on Germany’s hydrogen strategy. The Bundesrat has urged Brussels to double green fuel quotas to try to support hydrogen markets. H2 Mobility is consolidating its retail network. Düren is not alone in believing in a hydrogen economy, but it is a district with 270,000 residents committing €35m to a production asset that requires an order of magnitude more demand than its current fleet provides.

This is where sunk cost dynamics emerge. The buses are in service. The electrolysis plant is under construction. Federal funding has been secured. Political capital has been invested in a hydrogen identity. Scaling back becomes harder as more capital is committed. The economics, however, do not change because of commitment. A 10 MW plant serving 20 buses operates at 5% to 7% utilization. At that level the cost per kg remains structurally high. The plant requires either hundreds of buses, industrial hydrogen customers, or a long term subsidy framework to avoid becoming a stranded asset.

There is no reason for H2 Mobility to indefinitely keep the hydrogen refueling station open knowing that it’s going to be made obsolete and lose its only scaled customer. It’s undoubtedly losing money on it. With the Julich plant delayed and requiring shake in when it finally goes live, there’s a strong potential for Düren to have no available hydrogen refueling and to have to park their hydrogen buses for months, just as happened in Aberdeen.

The lesson from Aberdeen was that pilots do not scale automatically into viable ecosystems. The lesson from California refueling stations was that high pressure infrastructure has maintenance burdens far above modeling assumptions. Düren’s hydrogen program has better regional backing and larger capital scale, but the core arithmetic remains the same. If demand does not materialize at scale, the infrastructure becomes an economic liability. Doing the math does not predict the future with certainty. It does make clear what must happen for the program to justify itself. If that demand does not arrive, the district will carry the cost for a long time.