Canada’s Hydrogen Bus Trials Canceled Due To High Costs & Emissions

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

---

For several months I’ve been digging through what’s going on with a Canadian transit ‘think’ tank, CUTRIC, and why it’s pushing hydrogen buses into city plans across the country. It keeps claiming hydrogen buses have been a success and are cheap, but cities keep canceling their plans for them as they are too expensive.

Without nuance, CUTRIC tells cities that there are multiple successful uses of hydrogen buses in Canada, citing Whistler BC, Edmonton / Strathcona County and Winnipeg. Here’s the summary of the cities’ experiences:

Whistler

• 20 $3 million buses
• $55 per kg hydrogen
• 68% more expensive maintenance
• only 50% emissions reductions
• significant range loss in winter
• traveled a quarter as far before breaking down by the side of the road
• abandoned after four years
• bought diesel instead

Edmonton/Strathcona

• 2 expensive buses not servicing regular routes
• gray hydrogen with 3-5x emissions of diesel
• 80 km round trip refueling every 2-3 days
• no costs of hydrogen available
• cancelled order for more hydrogen buses
• cancelled plans for hydrogen refueling station
• bought diesel buses instead

Winnipeg

• No hydrogen buses
• electrolyzer canceled because too expensive
• alternative plan for hydrogen would have seen 3.2x diesel emissions
• canceled hydrogen refueling depot
• canceled hydrogen bus order
• bought diesel buses instead

CUTRIC cites each of those and claims they are successes and that Edmonton and Winnipeg are expanding their hydrogen bus fleets when they’ve canceled their orders. Let’s look at each in more detail.

Whistler Olympics Hydrogen Bus Trial

As is typical for Olympic Games, very expensive choices that aren’t particularly wise were made. The hydrogen buses were one of them. To be clear, it wasn’t obvious in 2010 that hydrogen buses were a dead end and that they would be displaced rapidly by battery electric buses. This was one of the useful trials of the approach at the time. Every trial after 2020 doesn’t pass this sniff test.

The Whistler hydrogen bus trial ended due to a combination of high operating costs, maintenance challenges, and infrastructure limitations, according to a report by the National Renewable Energy Laboratory (NREL). The BC Transit Fuel Cell Bus Project Evaluation, which analyzed three years of service data from 2011 to 2014, found that the hydrogen-powered buses were significantly more expensive to operate than conventional diesel alternatives. Maintenance proved to be a major hurdle, as the complexity of fuel cell technology required specialized servicing, adding to the costs. The hydrogen fueling infrastructure posed logistical and financial challenges, making it difficult to sustain the project.

The fleet’s average availability was 64%, and issues with the air supply system led to several buses being permanently removed from service toward the end of the trial. BC Transit opted not to repair them due to high costs and long wait times for parts. Diesel buses in transit fleets typically achieve an availability rate of 85% to 90%, with well-maintained fleets often exceeding 90%. Most transit agencies aim for at least 85% to ensure reliable service, while older buses or those requiring frequent maintenance may dip closer to 80%.

The air supply system, consisting of the air compressor, motor, and controller assembly, was one of the most problematic components of Whistler’s hydrogen buses. Initially, water infiltration caused failures in motor controllers, requiring retrofits in the first year. Later, compressors began failing between 1,200 and 1,500 hours due to low oil levels, exacerbated by a lack of maintenance guidance and monitoring tools. This was addressed by adding a sight glass and oil port. However, motors subsequently failed at around 3,000 to 4,000 hours, creating a cycle where failures in one component stressed others, leading to further breakdowns.

The liquid hydrogen for the Whistler bus trial was transported from Air Liquide’s facility in Bécancour, Quebec by truck. The plant produced hydrogen using renewable methods, including water electrolysis and chlor-alkali waste recovery, powered by Quebec’s grid, which is 98% hydroelectric or other renewables. As key points, hydrogen is a waste byproduct of the chlor-alkali process and hence is free from Air Liquide’s perspective and the volumes aren’t very small, likely in the range of 12 tons a day for the plant, but are still vastly smaller than the input requirements for ammonia plants or refineries per day. Finding things to do with inconveniently scaled but hypothetically valuable chlor-alkali waste hydrogen is an ongoing challenge, as is the case with the Prince George Teralta case.

The diesel trucks had one way trips of around 5,000 kilometers, plus or minus depending on which route they took. Among other things, liquid hydrogen trucks are not allowed in tunnels, so must plan routes to avoid them. A Class 8 diesel truck towing a liquid hydrogen trailer over 5,000 kilometers would consume 2,100 to 2,600 liters of diesel over the five days of the journey, depending on road conditions and fuel efficiency. With diesel emitting 2.68 kg of CO₂ per liter burned, this journey would produce between 5.63 and 6.97 metric tons of CO₂ emissions. Then the truck turned around and drove back to Quebec, creating a total of ten days, 10,000 kilometers and ~12 tons of CO2 emissions. This isn’t unusual, by the way, just an extreme example. Hydrogen supply is a consistent challenge, electrolyzers are expensive and hydrogen is trucked long distances as a result. Norway’s MF Hydra ferry receives its hydrogen from 1,300 kilometers away in Germany. Mississauga is planning to truck hydrogen from 60 kilometers away in Markham.

The liquid hydrogen trucks used in 2010 for deliveries to Whistler were not equipped with boil-off recovery units. During transit, some boil-off occurred due to the inherent nature of liquid hydrogen storage. For long haul trucking, boil off is often cited as around 1% per day, so 200 kg of hydrogen was vented for each trip, 5% of the load.

Peer reviewed literature now shows leakage rates of 2% to 10% before remediation and 2% to 4% at small hydrogen manufacturing facilities, and US DOE data shows 2% loss of liquid hydrogen due to boil off at points of pumping into and out of trucks in the best case scenario, and up to 15% losses for small volumes. South Korean testing found 15% of buses and cars leaked as well. Likely total leakage of hydrogen was 10% or more, but let’s stick with 10%. That brings the total hydrogen emitted into the 400 kilogram range.

At the time atmospheric scientists understood that hydrogen interfered with methane’s degradation, but it hadn’t been quantified. Now we know that hydrogen has a global warming potential of 12-37 times that of carbon dioxide over 100 and 20 years respectively. That means that boil off venting added another 5 to 15 tons of greenhouse gases to the the diesel emissions. Twenty tons of CO2e for 36 trips a year for four years turns into 2,900 tons of CO2e, most likely the low end of the range. That wasn’t accounted for in the claimed 5,835 tons of CO2e avoided by not operating diesel buses. 50% emissions reduction doesn’t sound nearly as useful.

Whistler’s transit agency has an out on part of this, although not all of it. The global warming potential of hydrogen was little known outside of the atmospheric science community at the time and was unquantified, so hydrogen vented to the atmosphere being benign was what almost everybody thought was accurate. The 10,000 km round trip bus trip, on the other hand, makes no sense at all.

As a note, these emissions rates are fairly common for my workups of hydrogen transit emissions around the world. There’s always some combination of gray hydrogen or high-carbon electricity that makes the hydrogen high emissions and leaking hydrogen along supply chains or at tiny local electrolysers that adds up to very low actual greenhouse gas emissions for hydrogen vehicles. In California, for example, one hydrogen refueling station was leaking up to 35% of the hydrogen and after a couple of years of remediation they managed to get it down to 10% leakage, and that was only one touch point.

A full liquid hydrogen tanker generally carries 3.5 to 4.5 metric tons of hydrogen, depending on the trailer size and design. Whistler’s 20 buses averaged 4,200 kilometers per month, or about 140 kilometers per day. Whistler isn’t a big place and skiers and mountain bikers take more time to embark and disembark so these are below the averages for buses in larger cities. The FCEBs had an average fuel consumption of 15.67 kilograms of hydrogen per 100 kilometers. The 140 kilometers per bus per day for 20 buses would have required 440 kilograms of hydrogen. Every ten days they would have required 4.4 tons, so likely more than one truck was crossing Canada back and forth for the four years of the trial quite regularly.

The total maintenance cost for Whistler’s hydrogen fuel cell buses was $1.10 per kilometer ($1.70 per mile) over the three-year trial period, including parts and labor. In comparison, BC Transit reported that diesel buses in similar service had an average maintenance cost of $0.65 per kilometer ($1.01 per mile). This means the fuel cell buses’ maintenance costs were 68% higher than those of conventional diesel buses. This is even higher than California’s much more recent hydrogen bus fleet experiences, where maintenance was 50% more expensive.

The propulsion-related maintenance costs for a typical diesel transit bus range from $0.25 to $0.35 per kilometer ($0.40 to $0.56 per mile), making up roughly 35% to 50% of total maintenance expenses. In contrast, the hydrogen fuel cell buses in the Whistler trial had propulsion maintenance costs of $0.62 per kilometer ($1.00 per mile)—nearly double the upper range for diesel buses.

Diesel transit buses typically travel 9,650 to 12,875 km (6,000 to 8,000 miles) between road calls, with well-maintained fleets exceeding 16,000 km (10,000 miles) before mechanical failures. In contrast, the hydrogen fuel cell buses in the Whistler trial averaged just 2,393 km (1,487 miles) between road calls, with propulsion-related failures occurring every 3,082 km (1,915 miles). They only traveled a quarter as far before breaking down by the side of the road.

Just as with Quebec’s fleet of hydrogen fuel cell cars that they abandoned after the four year trial was over, Whistler’s buses saw much lower fuel efficiency in the winter time, with 20% to 30% range losses. Hydrogen vehicle advocates assert that hydrogen fuel vehicles don’t lose range in the winter as a purported advantage over electric buses, but in the real world, they are comparable in that regard. As a note, fuel efficiency for diesel buses also drops from 10% to 30% or higher in cold temperatures. In other words, all vehicle drive trains are less efficient when it’s cold.

The buses cost $2.1 million each in 2010 dollars, or $3 million in 2025 dollars. The dispensed cost of hydrogen was $10.55 per kg, about $15 today. That excluded the cost of hydrogen storage tanks and refueling systems. The hydrogen storage and refueling system for the Whistler hydrogen bus trial cost $11 million CAD or $15 million 2025, funded as part of a $45 million CAD investment from the Canadian and British Columbia governments. Air Liquide supplied the liquid hydrogen, which was transported from Bécancour, Quebec, but did not directly fund the refueling infrastructure.

The total consumed fuel was 428,941.7 kg. That means that the $11 million amortized over dispensed hydrogen added another $27 per kilogram to the cost of the fuel, bringing it to $37 per kilogram, about $51 in 2025. Remember that the hydrogen itself was waste from the Quebec chlor-alkali plant and hence free. Even in Quebec with its low industrial electricity rates and low-carbon electricity, the best numbers I could come up with were $3.55 per kilogram to make it at an industrial scale facility with very high utilization. That’s close to $55 per kilogram for real hydrogen situations.

That means that one Whistler hydrogen bus would cost about $1,200 per day to fuel for its 140 kilometers of travel today. By contrast, a diesel bus would have a cost of fuel of about $140 per day. Meanwhile, at BC’s low industrial rates for electricity and the high efficiencies of battery electric drive trains, 140 kilometers would have cost around $14 per day, or 86 times less. That’s a staggering difference. No wonder Whistler abandoned it and no BC city has considered hydrogen buses since.

This is merely one of the more extreme examples of how much more it costs to run hydrogen buses than battery electric ones, and one of the key reasons why all of CUTRIC’s modeling is obvious nonsense. Mississauga’s trial isn’t as ludicrous because Enbridge is trying to get them hooked on hydrogen and has pegged the hydrogen as a service model per kilogram at around the same cost as diesel per kilometer traveled, $27 per kilogram, half of the real cost that Whistler experienced. But if Mississauga just put electricity into buses at night it would cost 24 times less per kilometer traveled. Norway’s MF Hydra is 10 times cheaper for electricity than the hydrogen it’s getting.

So Whistler had incredibly expensive $3 million buses, incredibly expensive $55 per kg hydrogen, incredibly expensive 68% pricier maintenance and only 50% emissions reductions. This is a massive economic failure with low climate benefits.

What does CUTRIC have to say about this clear lesson?

“The lack of an effective integrator in the project led to a national narrative of technology failure that conflated financial, business and project execution failures with overall FCEB technological failures. From 2010 to 2019, this narrative had become so profoundly ensconced in Canadian transit culture— despite the FCEBs actually performing well during the Olympics, and despite several successful FCEB procurements and deployments worldwide in the years following—that it created a legacy of inertia against FCEB procurement in Canadian transit.”

Yes, this massively costly, low benefits exercise that’s been repeated by at least 20 other transit organizations globally with similar results was unfortunate because it slowed the adoption of more hydrogen buses in Canada. You would think that CUTRIC had a fuel cell provider, a sole source procurement hydrogen bus provider and a hydrogen provider on its Board and in its high-paying membership, and you would be correct.

Edmonton Hydrogen Buses

Edmonton first attempted to operate electric buses, but ran into a snag by buying buses from ill-fated electric bus startup Proterra in 2020. That firm, founded in 2004, sold approximately 1,300 electric buses to more than 130 transit agencies across the United States and Canada, and Edmonton was unlucky enough to have purchased 21 of them.

Proterra’s bankruptcy in August 2023 stemmed from high production costs, supply chain disruptions, inflationary pressures, and intense competition from international manufacturers. Delays in obtaining essential components led to production setbacks, while rising material and labor costs strained its finances. Competing against larger, more efficient manufacturers, particularly from China, further eroded its market position, ultimately forcing the company into Chapter 11 bankruptcy. Basically like any other California startup with a 50% chance of failure, but also attempting to compete in bus manufacturing with China which has built approaching 100% of all electric buses ever built across its industry giants like Yutong and BYD, and many smaller players. Edmonton took the wrong lesson from this, unfortunately.

Edmonton’s attempt to move into hydrogen-powered public transit began with the Alberta Zero Emission Hydrogen Transit (AZEHT) initiative, a collaborative effort with Strathcona County. This pilot project, supported by Emissions Reduction Alberta with a $4.6 million grant, introduced two hydrogen fuel cell electric buses—one for each municipality—to assess the viability of hydrogen technology in Alberta’s climate.

The buses commenced service October, 2023. The hydrogen fuel is sourced from Alberta’s Industrial Heartland, and they are coy about it’s cost and emissions. In 2023, Alberta produced about 2.5 million tonnes of hydrogen, with 500,000 tonnes classified as blue hydrogen from the Shell Quest facility. Alberta is making that much hydrogen because desulphurizing and cracking Alberta’s crude oil requires a lot of it. That’s the biggest real market for hydrogen today. At present the cost and CO2 emissions of the hydrogen aren’t being shared that I can find, although the City Manager has said that they it can be very expensive per kilogram.

The hydrogen is also being trucked, but there’s some fun there too from the looks of it. It’s not going to the bus garage, but to the Suncor operated hydrogen refueling station in Leduc, 40 kilometers and a 40 minute drive south of Edmonton. The hydrogen is being trucked from the Industrial Heartland, north of Edmonton. As it’s Suncor, it’s not blue hydrogen but gray hydrogen. The hydrogen trucks are driving a 180 km round trip to deliver the hydrogen, then the buses are driving an 80 kilometer, hour and twenty minute round trip to get refueled. That’s a lot of driving for gray hydrogen and a lot of operational disruption for buses whose only selling feature is no changes in operations. That’s probably part of the reason why the buses aren’t actually servicing specific routes but being tried out on a handful.

This suggests that the service kilometers per day are at the low end, with the 80 kilometer round trip to the refueling station added on every other day. The CO2e debt of Alberta’s hydrogen is understated of course. Steam reformation of natural gas produces about nine kilograms of CO2 per kilogram of hydrogen. Alberta’s natural gas system isn’t as leaky as US shale oil emissions, but it’s still about 1.5% leakage. About seven kilograms of natural gas is required per kilogram of hydrogen for steam reformation, with a third or so for heat and two-thirds as a feedstock.

That’s not all, however. Methane reformers slip unreacted methane to the atmosphere 0.5% to 2%. Total leakage is 2% to 3.5%. Let’s call it 3%.

That means that every kilogram of hydrogen comes with about 0.2 kilograms of methane leakage. Methane is a very potent greenhouse gas, with 30-90 times the potency of CO2 over 100 and 20 years respectively. That means that kilogram of hydrogen also has 6 to 18 kilograms of CO2e from the leaking methane embodied in it.

The carbon in the methane gets combined with oxygen to make carbon dioxide and unless carbon capture is bolted on, gets vented to the atmosphere. That’s another 9 kilograms of CO2e. We’re up to 15 to 27 kilograms of CO2e per kilogram of hydrogen and we haven’t even left the hydrogen manufacturing facility.

There’s another gotcha in there. Liquification requires a full third of energy in a kilogram of hydrogen, about 12 kWh. That’s at Alberta’s grid intensity, which has fallen a lot in the last few years, but is still a high 470 grams CO2e / kWh. That’s another 6 kg of CO2e added on, bringing the total before it’s left the hydrogen manufacturing facility of 21 to 33 kg of CO2e per kg of hydrogen.

Then there’s the leakiness of hydrogen, typically 1%+ at every touch point. Steam reformation, site storage and compression, liquifying, pumping into trucks, boil off, pumping into tanks in Leduc and then pumping into the buses.

Then up to 10% hydrogen leakage per kilogram of hydrogen, and hydrogen is a potent greenhouse gas too as it interferes with the decomposition of methane in the atmosphere. As noted in the Whistler section, it’s 12-37 times more potent than carbon dioxide. That’s another 1 to 4 kg of CO2e added to the mix, bringing the total to roughly 22 to 36 kilograms of CO2e per kilogram of hydrogen in the buses. As noted, the buses are likely to leak during their service lives too.

New Flyer doesn’t commit to efficiency ratings, but the same fuel cells are in the Edmonton buses as the Whistler buses so 15.67 kg per 100 km is a reasonable assumption. California’s fuel cell buses get better mileage, but as noted fuel cell buses lose 10% to 50% of range as it gets colder, and Edmonton isn’t known for its California climate.

With short test routes and the 80 km refueling trip every other day, the buses probably are seeing 200 kilometer days as a guess, assuming that they are actually working every day. That means that they are using about 30 kg of hydrogen a day. That means well to wheel emissions of 0.7 to one ton of CO2e per day.

By contrast, diesel buses running 200 kilometers a day emit about 0.2 tons of CO2e, a third to a fifth of the ‘green’ hydrogen buses. Battery electric buses running 200 km per day, even with Alberta’s high grid emissions, would be emitting only 0.14 tons of CO2e and declining every year with grid decarbonization.

CUTRIC is in the mix, of course. In July, the Government of Canada and the City of Edmonton announced they investing C$1.29 million to develop a comprehensive strategy for transitioning Edmonton’s public transit fleet to zero-emission buses. The initiative, supported by the Zero Emission Transit Fund (ZETF) and Edmonton’s contribution of C$258,499, will assess hydrogen fuel cell and battery-electric propulsion technologies. CUTRIC is getting most of the money, although given its lack of staff, it might be subcontracting to Deloitte again as it did with the badly botched Brampton study, for which Michael Raynor and I identified $1.5 billion modeling errors and incorrect assumptions that led to the blended fleet being cheapest.

Of course, fiscal reality about the cost of hydrogen buses and hydrogen itself reared its head along the way. In March 2024, the City of Edmonton announced a “pause” on its planned hydrogen fueling station due to budget constraints. Initially, the city had proposed acquiring 40 hydrogen fuel-cell buses, but financial limitations led to the purchase of 20 diesel buses instead.

Because the only supplier of hydrogen buses in Canada is New Flyer, the buses came from that company. As I’ve noted before, they love it when a Canadian transit agency picks hydrogen because they become the sole source provider and don’t have to compete. They are also on the Board of Directors of CUTRIC along with Enbridge which supplies hydrogen and, up until recently, Ballard Power, which supplies fuel cells. There is more suspect sole procurement in there, as the only firm approved to receive 80% coverage for transit studies under Canada’s ZETF is CUTRIC, so no competent firm can compete.

As I’ve noted, it’s a short sighted strategy on New Flyer’s part because it not only is going to make every customer who buys a hydrogen bus upset and hence consider BYD and other vendors more seriously, it also means that they aren’t focusing on make their battery electric buses competitive. My assessment is that New Flyer in North America loses three battery electric bus sales for every hydrogen bus they deliver, but they are seduced by the short term fiscals. Not only do their hydrogen buses cost a lot more to buy, the ZETF funds 50% of capital costs. Not the operating costs, which are, as noted, absurdly high.

So there we have Edmonton. Two expensive buses not operating on fixed routes, driving 80 km and hour and twenty minute round trips to refuel, with emissions 3-5x diesel buses and even further off of battery electric buses of course, a canceled electrolyzer, cancelled orders for new hydrogen buses and diesel buses purchased instead.

CUTRIC has been silent about Edmonton, despite being engaged closely with the effort and having Eddie Robar, the City Manager for the city on the Board. In an email to the City Councillors of Mississauga, they make the bold claim that Mississauga will be the fifth city in Canada to procure hydrogen buses, and list the two Alberta buses without any caveats.

Winnipeg Hydrogen Buses

Winnipeg has a special problem related to hydrogen buses and low-carbon buses in general, in that it’s the home town of New Flyer. There’s no political way for them to buy good battery electric buses from other vendors, as they must frame procurements so that only New Flyer can win.

In 2021, Winnipeg initiated plans to decarbonize its transit system by procuring both battery-electric and hydrogen buses from the firm. The strategy included installing an electrolyzer at the Fort Rouge Garage to produce hydrogen on-site for the fuel cell buses. However, escalating costs soon became a significant concern.

The financial challenges were substantial. The city allocated $33 million for just 16 zero-emission buses, averaging over $2 million per bus. Comparable battery-electric buses from manufacturers like BYD and Yutong are not only more affordable but also offer superior range and performance. Additionally, Winnipeg awarded $41 million to New Flyer for up to 43 diesel buses, each costing nearly $1 million each, which is notably higher than the benchmark global price for standard diesel buses as well.

Infrastructure expenses further strained the budget. The initial estimate for the new transit garage was $200 million, but this figure ballooned to $305 million. In September, an attempt was made to secure an additional $80 million to cover these overruns.

Complicating matters, the city faced challenges with hydrogen production. After receiving only one prohibitively expensive bid for the electrolyzer installation, officials considered alternative methods, such as extracting hydrogen from methanol.

Hydrogen electrolyzed at Winnipeg’s transit bus garages would have had a very low carbon debt for manufacturing, compressing and storing hydrogen, even with the low efficiencies of small electrolyzers at refueling stations, of about 0.1 tons of CO2e per ton of hydrogen. Adding the 0.3 tons of CO2e from leaking hydrogen arrives at 0.4 tons of emissions for the system, which is much better than the 3.1 tons of emissions from diesel per ton. An 87% improvement is actually an improvement, as long as they kept leaking under control. Too bad it was too expensive to build and operate.

However, producing hydrogen from methanol, typically derived from natural gas or coal, can result in greenhouse gas emissions approximately 3.2 times higher than those from diesel.

Of course, making hydrogen and using it is three to four times less efficient than just using the electricity in battery electric buses and battery electric buses don’t leak any greenhouse gases. Manitoba’s electricity grid is predominantly powered by hydroelectric generation, resulting in one of the lowest greenhouse gas emission intensities in the world. In 2023, the emission intensity was approximately 1.34 grams of CO₂ equivalent per kilowatt-hour.

Given this exceptionally low emission intensity, the GHG emissions associated with operating a battery-electric bus in Manitoba are minimal. For instance, if a BEB consumes 220 to 360 kilowatt-hours of electricity to travel 200 kilometers, the resulting CO₂ emissions would be approximately 0.29 to 0.48 kilograms. In contrast, a traditional diesel bus traveling the same distance would emit approximately 200 kg of CO₂.

The best case for electrolysis of hydrogen compared to battery electric is 16 times more CO2e emissions for the distance traveled, for diesel it’s 400 times better and for methanol to hydrogen it’s 1,200 times better.

Due to these escalating costs and environmental challenges, Winnipeg decided to halt its Transition to Zero Emission Buses program, redirecting the 2025 funding solely towards purchasing additional diesel buses. While the move to acquire more diesel buses allows for fleet expansion and potentially increased ridership, it also means continued reliance on fossil fuels.

CUTRIC again has been silent about Winnipeg’s experiences trying to procure hydrogen buses and facilities, despite having Erin Cooke of Winnipeg Transit as chair of their Hydrogen in Transit committee. In the email to Mississauga Councillors CUTRIC claims that Winnipeg has eight hydrogen buses on order and that they will be delivered in 2025, which as noted is not what Winnipeg is doing at all.