Hydrogen Forklifts Are A Rounding Error, Not A Rebuttal

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

Someone responded to a just-published article about yet another hydrogen fleet failure by pointing to hydrogen fuel cell forklifts in US distribution centers. The implication was clear. If hydrogen works in forklifts, then the broader critique of hydrogen transport must be flawed. It is a familiar move. It is also deeply flawed and biased. I don’t even consider the argument to be in good faith, it’s so weak. They are reaching for an example that sounds tangible. Invoking hydrogen forklifts as evidence that hydrogen transport is winning requires ignoring scale so completely that the argument collapses on contact with basic statistics.

The core problem is denominators. Any claim about success or failure that relies on a numerator without a denominator is meaningless. This is not a matter of opinion or ideology. It is elementary statistics. A few thousand units can sound impressive in isolation. The moment those units are placed against millions of comparable assets, the interpretation changes completely. Continuing to claim validation after that comparison is made is not a mistake. It is intellectual dishonesty. When hydrogen forklifts are used as a defense of hydrogen transport, the numerator is doing all the work, and the denominator is being deliberately ignored.

Globally, forklifts are a large and mature industrial market. Annual sales are in the range of 2.2 million units, according to the long-running World Industrial Truck Statistics dataset. Roughly 70% of those forklifts are battery electric. Most of the remainder are diesel, LPG, or CNG. Hydrogen fuel cell forklifts are measured in the low thousands per year. That puts hydrogen well below 0.1% of global forklift sales, hence the lack of the line for them even showing up on this chart even though they are in the data set that generated it. At that scale, hydrogen is not a niche contender. It is a rounding error. Claiming that a technology is succeeding based on a share that rounds to zero is not analysis. It is narrative defense.

It’s well worth noting that World Industrial Truck Statistics understates total forklift volumes. The undercount is overwhelmingly on the electric side, not hydrogen. China produces and deploys hundreds of thousands of electric forklifts each year, particularly Class 3 pallet trucks and walk-behind units that never enter export statistics and are inconsistently reported into Western datasets. China is far more electrified in material handling than North America or Europe. When those units are fully accounted for, the global denominator grows to around 3 million, and the electric share increases. Hydrogen does not benefit from this correction. Its relative share shrinks further.

Zooming in to the United States does not rescue the argument. The US market sells on the order of 900,000 to 1,000,000 forklifts per year, depending on whether one looks at shipments or orders. Battery electric forklifts dominate US indoor logistics, with 70% of those annual sales being electric, and only 50,000 to 60,000 hydrogen forklifts in total after more than a dozen years of sales. Internal combustion forklifts retain a presence in heavier and outdoor applications. Hydrogen forklifts become more visible than they are globally, but visibility is not dominance. Even in the US, hydrogen forklifts represent a small fraction of electric forklifts, and a tiny fraction of total forklifts. The denominator still matters. Shrinking the frame does not change the math.

At this point, defenders of hydrogen almost always pivot to the same examples. Amazon. Walmart. Home Depot. These are the largest US distribution systems, and they are often presented as proof that hydrogen forklifts have won at scale. This is where selective disclosure becomes central to the misunderstanding. Hydrogen forklift counts are sometimes disclosed by these firms or by their supplier, Plug Power. Total forklift fleets are not. That asymmetry creates a distorted picture. When only hydrogen numbers are visible, it is easy for motivated readers to assume hydrogen is dominant. It is not.

Walmart provides the clearest historical anchor. In 2016, Walmart disclosed that it had roughly 4,200 hydrogen forklifts, representing about 20% of its US forklift fleet. That implies a total fleet of about 21,000 forklifts at the time. Since then, Walmart has increased its hydrogen forklift count to roughly 9,500 to 10,000 units, according to trade press and supplier statements. What Walmart has not disclosed is its current total forklift fleet. Without that denominator, hydrogen growth cannot be interpreted as dominance. Even if the total fleet doubled since 2016, hydrogen would remain a minority technology inside Walmart’s operations.

Amazon discloses that it operates more than 15,000 hydrogen fuel cell forklifts in the United States, with a stated target of around 20,000. That sounds large until it is placed in context. Amazon operates hundreds of fulfillment centers, sortation centers, and delivery stations across North America. Logistics engineering benchmarks show that a large fulfillment center commonly operates 150 to 300 forklifts across multiple shifts. Smaller facilities operate fewer, but they are numerous. Using publicly available facility counts and conservative per-facility forklift ranges, Amazon’s US forklift fleet plausibly sits in the high tens of thousands. Under those conditions, hydrogen forklifts are not dominant. They are a subset, concentrated in specific facilities and specific classes of equipment.

Home Depot follows the same pattern, but with even less disclosure. It has used hydrogen forklifts in some distribution centers since the mid-2010s. No public source provides a comprehensive count of either hydrogen forklifts or total forklifts across its US operations. Estimation using facility counts and standard forklift densities again yields a fleet where hydrogen is present but clearly secondary to battery electric equipment. The absence of published totals does not imply hydrogen dominance. It simply reflects normal corporate reporting practices.

Estimating forklift fleets responsibly starts with facility counts that companies themselves disclose. Amazon operates on the order of 175 large fulfillment centers in the United States, alongside several hundred smaller sortation and delivery facilities. Warehouse design and material handling studies from firms like MHI members and major integrators consistently show that a large fulfillment center typically operates between 150 and 300 forklifts across multiple shifts, while smaller facilities operate between 20 and 80. Using the low end of those ranges yields a rough estimate of 175 large facilities times 150 forklifts, or about 26,000 forklifts, plus 250 smaller sites times 30 forklifts, or another 7,500. Even conservative assumptions therefore put Amazon’s US forklift fleet in the range of 30,000 to 40,000 units.

Walmart provides a useful cross-check because it disclosed both hydrogen forklift counts and total fleet share earlier in its adoption cycle. In 2016, Walmart reported roughly 4,200 hydrogen forklifts, representing about 20% of its US forklift fleet, implying a total of approximately 21,000 forklifts at that time. Walmart has since increased its hydrogen forklift count to around 9,500 to 10,000 units, based on supplier disclosures and trade reporting. Even if Walmart’s total forklift fleet grew only modestly since 2016, hydrogen forklifts would still account for well under half of the fleet. Any reasonable growth in total forklifts only pushes hydrogen’s share lower.

National forklift stock estimates provide an outer bound that keeps these calculations grounded. The United States is estimated to have on the order of 3 million forklifts in operation across all sectors. For Amazon, Walmart, and Home Depot combined to operate, for example, 150,000 forklifts would imply they control around 5% of all US forklifts, which already stretches plausibility. Placing their combined fleets in the tens of thousands aligns with both facility-based estimates and disclosed hydrogen counts. Across all of these methods, hydrogen forklifts remain a minority technology even inside the most hydrogen-friendly firms in the United States.

It is also important to understand why forklifts were ever considered a hydrogen niche in the first place. The original appeal had little to do with hydrogen being inherently superior. It was a response to the limits of lead-acid batteries. Long charging times, battery swapping labor, and space constraints created operational pain in multi-shift warehouses. Hydrogen offered fast refueling and consistent power output. That advantage was real for a time. It was also temporary. Lithium-ion batteries eliminated most of those constraints. Fast charging, opportunity charging, lower maintenance, and falling costs erased hydrogen’s operational edge. Hydrogen did not lose because of ideology. It was overtaken by better and of course cheaper electric technology.

Direct public support for hydrogen forklifts in the United States is real and well documented. The most explicit federal intervention came through the Department of Energy’s Fuel Cell Technologies Office under the 2009 American Recovery and Reinvestment Act. DOE reports show roughly $9.7 million in federal grant funding directed specifically at fuel cell lift truck deployments, paired with about $11.8 million in industry cost share. Those programs supported the deployment of just under 700 hydrogen fuel cell forklifts and included funding for on-site hydrogen refueling equipment, training, and data collection. The onsite refueling infrastructure made the fuel cell lifts sticky of course, making it lower friction to buy a few more hydrogen forklifts rather than just using battery electric as almost everybody does. Individual awards within that program included roughly $6.1 million for GENCO, $1.3 million for a FedEx Freight deployment covering 35 forklifts at a single facility, and around $1.1 million associated with Nuvera. These grants were explicitly intended to seed early markets and reduce first-mover risk, not to demonstrate unsubsidized commercial competitiveness.

Beyond direct grants, hydrogen forklifts and their fueling infrastructure have benefited from tax-based subsidies that are harder to total because they are claimed privately and reported only in aggregate. DOE materials and IRS guidance make clear that fuel cells used in material handling have been eligible for an investment tax credit of up to 30% of capital cost or up to $3,000 per kW, depending on the year and program structure. Hydrogen dispensing equipment has also been eligible for the Section 30C alternative fuel refueling credit, which after recent legislative updates can cover up to 30% of installed cost for qualifying business property, subject to location and labor requirements. These incentives can materially improve project economics at individual sites, but there is no public ledger showing how much has been claimed specifically for forklift fleets, which makes it impossible to produce a precise total subsidy figure from tax credits alone.

The broader hydrogen forklift ecosystem has also been supported indirectly through large-scale federal financing for Plug Power, which supplies most of the fuel cells, fueling equipment, and hydrogen to these fleets. In 2024 and 2025, the Department of Energy’s Loan Programs Office finalized a loan guarantee on the order of $1.6 billion to support Plug Power’s hydrogen production and liquefaction projects. This was not a grant and does not target forklifts directly, but it clearly lowers financing risk and capital costs for the hydrogen supply network that companies like Amazon and Walmart rely on for their forklift operations. Taken together, the record shows that hydrogen forklifts in the United States emerged and persisted in a subsidy-rich environment that combined early deployment grants, ongoing tax credits, state-level clean fuel programs in places like California, and large federal financial backing for the dominant supplier. What cannot be known without access to confidential tax filings is the full cumulative value of those subsidies. What can be said with confidence is that hydrogen forklifts did not scale in the absence of public support, and their continued use reflects that policy history as much as it does operational choice.

In recent months Plug Power’s financial position has shifted to desperate retrenchment and cuts to the bone, with direct implications for hydrogen forklift operators in the United States. After securing the large Department of Energy loan guarantee intended to finance multiple hydrogen production and liquefaction facilities, the company later signaled that it would not proceed with those projects under the loan structure. That decision reflects ongoing liquidity pressure and an effort to avoid taking on additional obligations tied to long-dated, capital-intensive infrastructure. At the same time, Plug Power has reduced operating and maintenance spending, deferred projects, and narrowed its focus to preserving cash and maintaining existing customer commitments rather than building out the vertically integrated hydrogen network it previously promised. As I noted, cuts to maintenance of industrial hydrogen generation facilities increase risk substantially, and my first hope is that no staff are harmed in yet another hydrogen explosion or fire, ending up Hydrogen Insights’ mounting list of incidents, injuries and fatalities.

For operators running hydrogen forklift fleets, this retrenchment introduces new uncertainty. Hydrogen supply arrangements that were originally marketed as stable, vertically integrated, and increasingly low cost now depend more heavily on third-party sourcing and short-term commercial contracts. Cost control measures at Plug Power also raise questions about the long-term robustness of maintenance, service, and refueling support for installed forklift fleets. Even if day-to-day operations continue, the shift in strategy underscores that hydrogen forklifts in the United States are tied to a supplier ecosystem under financial strain, not to a self-sustaining market. That risk profile matters when hydrogen forklifts are presented as evidence of a technology that has already won.

What remains today is path dependence. Hydrogen forklifts persist where infrastructure was built early and where replacement cycles are slow. Persistence is not validation. A technology can survive in a niche without ever being competitive at scale. Confusing the two is the deeper analytical error. Forklifts are simply the clearest illustration because the numbers are so stark and the market is so mature.

What the commenter on the previous article was doing is not subtle, and it has a name. It is base rate neglect, the most basic statistical error, where a numerator is presented as meaningful while the denominator is ignored. Pointing to a few tens of thousands of hydrogen forklifts in the United States while disregarding millions of forklifts sold globally each year, and the overwhelming dominance of battery electric equipment is not analysis.

Layered on top of that is the logical fallacy of cherry-picking, selecting the one geography and one niche where hydrogen is most visible while ignoring readily available global data and the much higher electrification rates in places like China. The behavior is driven by motivated reasoning, where evidence is filtered to defend a prior belief rather than test it. The statistical language they tend to use gives the appearance of literacy, but the reasoning violates the first rules anyone trained in statistics would apply. Once scale and base rates have been explained, continuing to present a rounding error as validation is no longer confusion. It is an intellectually dishonest attempt to signal expertise while ignoring the weight of the evidence. It’s a sign of deep cognitive biases that they refuse to even consider, never mind overcome.

A related rhetorical move appears when advocates point to claims such as a 40% year over year increase in hydrogen transportation sales in South Korea. A figure like that is designed to sound decisive, because large percentages trigger salience. They feel like momentum. What is almost always missing is the base rate. If hydrogen transportation sales start from a very small number, then a 40% increase still leaves the category trivial when set against the overall transport market. The percentage is doing the persuasive work precisely because the absolute numbers are unimpressive. This is base rate neglect and denominator blindness combined with a salience trick, where an eye-catching growth rate crowds out the more important question of scale. Growth rates only become meaningful when the underlying denominator is large enough to matter. Without that context, a 40% increase is not evidence of success. It is a way of making a rounding error feel significant.

I have addressed the irrelevancy of US hydrogen forklifts before, but buried it in articles and comments. It was time to devote some time and space to it so that next time someone tries to waste someone’s time by pointing to it, I can just paste in the article and move on. As I’ve noted in the past, far too large a portion of my publications can be explained by the XKCD cartoon Duty Calls.

This matters beyond forklifts. The same rhetorical pattern appears across hydrogen buses, trucks, trains, and other transport applications. A small surviving deployment is treated as proof of viability, while the denominator is ignored. Scale is optional only in narratives. In real energy systems, scale is everything. Technologies that cannot move beyond rounding errors do not fail morally. They fail mathematically. Hydrogen forklifts do not rescue hydrogen transport. They demonstrate its limits.

Advocates for hydrogen for energy, whether heading European hydrogen lobbying organizations or simply boneheaded commenters on the internet, persistently make these kind of rhetorical arguments, show this willing blindness to empirical reality and spew logical fallacies as if they are gotchas. It’s remarkable that they get away with it so often. It’s even more remarkable that they can look themselves in the mirror and pretend to intellectual rigor and statistical competence. Personally, I’d be deeply ashamed.