A Deep Dive into Nikola’s FCEV Design and Price Model

by thri54

Hey everyone. I’m a mechanical engineering student with a hobby interest in finance. I’ve spent the last few days figuring out if Nikola’s leasing model is actually possible. There’s some really wacky stuff going on in Nikola’s presentations and financial projections, and I wanted to share my findings.

This is an absolute wall of words, and I wouldn’t be offended if you didn’t want to read it all. In the first half, I try to tease out the cost per mile of an actual Fuel Cell Electric Vehicle (FCEV) given the specification Nikola lists. Next there is a portion where I look at the discrepancies between their Financial Projections and their Lease breakdown. Then a quick little peanut gallery where I look at their unrealistic assumptions and the hypocrisy of their comparisons. Finally, a more serious portion where I discuss the design, efficiency, and utility implications of Nikola’s chosen power output and battery capacity. Hope you guys enjoy!

Let’s get started:

Nikola claims that they have the industry first holistic leasing program, including maintenance, fuel, and use of the vehicle. They plan on leasing for $.95 per mile @ 30% margin. This implies an expense of $.73 $.67 per mile to Nikola.

Hydrogen costs:

According to the DoE, it currently costs $5.10/kg to produce, compress, and dispense hydrogen. Nikola claims they can do this for $2.47/kg. I highly doubt their estimate, and will elaborate on that later. Hydrogen has a specific energy 33.3 kWh/kg. A Fuel cell Electric Vehicle (FCEV) has an average thermal efficiency of 55%. A diesel semi tractor, which easily compares to Nikola’s offerings, consumes about 1.25 kWh of work per km (or 2.125 per mile) of useful work loaded.

This implies the Nikola truck will use 3.86 kWh of hydrogen per mile, at a cost of $0.59 per mile, or $.29 using their estimates. The DoE estimate could be pretty rosy as well, Hindenburg cited a practical price of $16 per kg for hydrogen in their report. Nikola’s estimate in the leasing breakdown is 7.5 miles per kg of hydrogen @ $2.47 per kg. That works out to $.33/mile. Our estimates are pretty close, excluding hydrogen costs. It looks like, in a surprising twist, they actually overestimated the energy consumption of a tractor. Or maybe not. We’ll get to that

ICCT Tractor-Trailer Fuel Consumption: theicct.org/sites/default/files/publications/EU_HDV_Testing_BriefingPaper_20180515a.pdf

Why do I doubt their hydrogen cost estimates? $2.5 per kg implies $.075 per kWh of hydrogen produced The average price for Industrial electricity in Arizona, the state they are headquartered in, was $.068/kWh, some of the cheapest in the US. Of course, there isn’t a 1:1 conversion of electricity to hydrogen: an electrolyzer uses about 50 kWh per kg of hydrogen ( specific energy of 33.3kWh/kg), making the electricity expense alone in excess of $.10 per kWh of hydrogen. Electricity must also be used to compress the hydrogen. This would take another ~4 kWh, though we’re already over budget. God forbid they use California electricity at an average cost of $.15 per kWh. The electricity expense for the Electrolyzers alone exceeds their estimates, much less depreciation expense, cost of capital, maintenance expense, salary expense, etc. Clearly a bogus number.

I suppose they can use renewable excess during off-hours for cheap, but the rapidly decreasing costs of energy storage will likely level out those low prices rather quickly. This also only works in Arizona and a select few other states; California not included. There is the issue of a startup paying to build huge electrolyzers that might have a utilization factor of ~30%, and additional high pressure storage will be needed. The abhorrent upfront capex needed to try and drive down operating costs is not viable for them.

EIA electricity prices nationwide: www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_5_3

NREL H2 cost analysis from 2011. This is just about the most recent research I can find. The abject lack of new material tells me it’s not exactly a hoppin’ field: www.nrel.gov/hydrogen/production-cost-analysis.html#fn3

Fuel Cell costs:

Part of the reason there are currently so few FCVs on the road today is the limited service life of a fuel cell. Fuel cells are precision manufactured components that degrade quickly when jostled & vibrated too violently. This is not good when combined with the rock-hard suspensions of semi tractors.

The DoE targets a useful life of 150,000 miles for a fuel cell. Currently, there is no information confirming this target has been met. A Toyota Murai comes with a 100,000 mile warranty on its FC. For the sake of argument, I will assume Nikola FCs can meet this target. The DoE targets a cost of $40/kW for fuel cell production in 2020, provided mass hits 500,000. This hasn’t happened yet, but I will again assume this to give Nikola the best chance. As an aside: Nikola’s decision to use exclusively GM FC technology in their Badger pickup indicates to me they have nothing “up their sleeve” to make the technology more viable, despite my optimistic assumptions.

I’ll assume the Nikola Two’s Fuel cell is 500 kW, less than the 750 kW claimed output. I think it likely their horsepower claim will be a peak power figure only achievable when the motors draw on the battery & FC. I cannot confirm this, because Nikola does not list the output of their motors and FC separately (along with myriad other questionable, or lack of, claims). I think this is reasonable, considering FC thermal efficiency is maximized between 20% and 30% load, and a semi will average ~90 kW of useful work required on the highway, translating to ~170 kW of FC usage. This is near the peak efficiency band of a PEMFC. This assumption also allows steady-state operation at 66% of the “rated” output. This implies an upfront cost of $20,000. A targeted useful life of 150,000 miles implies a depreciation expense of $.13 per mile.

NREL Stack Durability and Performance vs load chart: www.nrel.gov/docs/fy19osti/73011.pdf

Battery costs:

Using BNEF 2023 battery cost estimates of $100/kWh, that equates to $25,000 of battery expense. Assuming a useful life of .25M miles, more than any existing warranty currently covers, that results in a depreciation expense of $.125 per mile.

Chassis and the rest:

Lastly, I extrapolated an FCEV COGS of $175,000 per truck from their Financial projections, minus the $45,000 of equipment already listed, and a 15% scrap value I pulled out of my ass to try and help nikola here, leaves $104,000 depreciated over 700k miles, or a $.15 depreciation expense per mile.

Maintenance costs:

Nikola assumed a $.061/mile maintenance cost. Any engineer should be able to see such a claim and immediately question it. Tires alone should account for $.03 per mile. That leaves…. $.031 for brakes, air lines, HVAC, wiring, electrical equipment, motors, inverters, those battery and FC expenses I already calculated, sensors, etc. They make no additional provisions for the battery/FC in their leasing breakdown. Pure, unadulterated bullshit. The ICCT puts BEV per mile maintenance at ~$.19/mile. How they squeezed 70% of those costs out, as an unproven startup, by going for a more complex FC-BEV hybrid is beyond me.

Cost of trucking: www.thetruckersreport.com/infographics/cost-of-trucking/

Nikola Leasing cost breakdown, p19: d32st474bx6q5f.cloudfront.net/nikolamotor/uploads/investor/presentation/presentation_file/5/NikolaInvestorRoadShowPresentation042720.pdf

Leasing Conclusions:

Adding their laughably low per mile maintenance expense of .061 + .15 + .125 + .13 +.59 gives us an aggregate $1.06 per mile expense for Nikola. Using their fuel expense estimate of $.29, this equates to $.76; still more than their projected gross expenses. The first estimate is 50% over they need for their claimed 30% gross margin at $.95 per mile. Note: I used a projected battery expense, projected FC service life target, and projected FC production expense. None of these have been met. I used the average resitive forces acting on a US tractor-trailer, which appear to be lower than the number Nikola uses. I did not include warranty expenses in my estimate. Additionally, these are EXPENSES, and includes 0 profit for the suppliers of these parts. The GM-Nikola deal clearly shows there will be little vertical integration in their production, and such allowances would have to be made.

A more reasonable estimate, including a still optimistic 3% interest expense for truck capex, a .4% annual warranty expense (corresponding to their presented 3% estimated reserve). That reserve, btw, is very optimistic: Tesla used a higher reserve on the S for years, while building a simpler product with a warranty length/distance a literal order of magnitude lower than the Nikola truck. A *STILL* low maintenance expense of $.12, and a 10% margin for battery & FC production, we end up with an $.92 per mile expense, or more than Nikola can afford, even when using their untenable $2.5 per kg hydrogen estimate. This is before G&A expenses. Their leasing business model is not possible.

Lease Projections v. Income Projections: Internal Chaos or Outright Fraud?

It’s possible some of the folks at Nikola have already found those problems out, though. Nikola says they have plans to Lease their trucks. They’ve had presentation slides including the idea, and their truck descriptions on their website include a leasing plan. In their most recent presentation at the DB Global Auto Industry Conference in June, however, the Leasing cost breakdown slide was conspicuously missing. Their Financial projections slide showed 2,000 FCEV trucks being produced in 2023, and 470 million in revenue from FCEV sales. This represents $235,000 per truck, and their FCEV revenue scales exactly linearly into their 2024 projected sales; no room for residual from the 2023 trucks. They’re projecting to sell them! Revenue from maintenance and Hydrogen sales are also listed separately. Their Financial projections clearly show the upfront sale of trucks with additional Hydrogen fueling and maintenance revenue, and the leasing model slide has disappeared. It’s easy to see why. Their projected combined expenses and capex exceeds $7.5 billion through 2024, significantly more than their current $1 Billion in assets and a couple of lease payments would allow for. This would take some intense share dilution (not something I think Trevor would be on board with) or extremely expensive leverage.

It’s not like they’re going to get cheap loans secured against their proprietary trucks, requiring their proprietary stations, to run only their customers’ preset routes. A bank wouldn’t want that kind of collateral. The leasing idea is a real mess.

Nikla DB presentation, projections are 2nd to last page: d32st474bx6q5f.cloudfront.net/nikolamotor/uploads/investor/presentation/presentation_file/7/Nikola_and_VectoIQ_Conference_Presentation_DB_Global_Auto_Conference__6.10.2020_.pdf

One can claim that the lease model is still in the description of the trucks, but so are battery and fuel cell specifications for the Nikola one. The Nikola one was, ostensibly, never actually powered by hydrogen, and development has since been abandoned. It looks like their leasing Idea may have been abandoned as well.

I’ll also point out that an expected 2024 FCEV maintenance revenue of 56 million on 7000 trucks sold, assuming an average of 50,000 miles per truck sold in 2024 (the average mileage if the trucks are sold at a constant rate through the FY) and 100,000 per truck in 2023, equates to 12.4 cents per mile, more than double the $.061 projected maintenance costs in the april lease presentation. Either they plan on making a killing from maintenance, or there was some aggressive re-shuffling of numbers when maintenance went from an expense to revenue stream, or vice versa.

The same analysis of hydrogen expenses puts their per kg revenue at $4.08. Still low, but a hefty sum above their $2.47 cost average on the leasing slide.

If we use their projected FCEV maintenance revenue of $.124, $4.08 per kg H2 revenue, and $235,000 truck price depreciated over 7 years w/no interest expense, the cost of ownership, according to their income projections, is $1 per mile for a 100,000 mile year. More than they say a diesel will cost. OOPS!

That’s most of what I wanted to talk about. It’s pretty clear that Nikola cannot possibly make a profit with their lease model, and Nikola’s finance department has indirectly acknowledged this. Hydrogen tech is still many, many years away. Nikola’s move fast and break things approach (though I’m not convinced we’ve seen much moving outside of gravity assists) will end up a “move fast and bankrupt things” strategy.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I now want to take a few moments to look at some of the sillier things I found in my research:

In a laughably exuberant turn of events, Nikola projects 25% gross margins during their first year of production. That’s downright cute! They plan on, ostensibly, slapping GM FCs & batteries on IVECO platforms, and beating the margins of all publicly traded auto companies sans Ferrari within a few months of production start. It’s that easy. Get your shit together, Ford!

On Nikola’s website under their trucks, they have a comparison of FCs, BEVs, and Diesel where they seem to forget their trucks have hundreds of kWhs of battery storage onboard. They claim “Hydrogen acts as a buffer and grid balance,” while “batteries are a drain on the grid.” You heard it hear first, folks, batteries can’t be used as a grid buffer, only a grid drain. They also claim H2 is the “most abundant element on the planet,” while “Batteries [are] made of non-renewable resources; dangerous/costly to mine.” Makes me wonder why they chose to put so many on their trucks. Also hilariously hypocritical, considering Platinum -required for PEMFCs- is one of the rarest metals on earth.

Design considerations and odd choices:

I want to take a moment to talk about chassis design and the implications of Nikola’s set power parameters. And I want to start with a quote from Elon Musk during the Tesla Semi reveal:

“We designed the Tesla truck like a bullet,” Musk said. “A normal diesel truck is designed like a barn wall.” The Tesla Semi is more aerodynamic than a $2 million Bugatti Chiron sports car, he said.

A bold claim, but I believe this to be correct. Cheesy, but correct. Why? Cooling requirements. The more cooling air you need, the less favorable aerodynamics you’ll have. In the simple representation of an engine, there is a fuel input, a work output, and a heat output. A gas engine -especially one not designed with fuel efficiency in mind like the Chiron’s- will likely only output ~20-25% of its fuel input as a work output. The Chiron makes 1,500 hp, and needs to reject 4,500-6,000 hp of heat at full throttle. The goal of a Chiron isn’t to slip through the air, au contraire. It’s designed suck in as much air as possible for cooling and brute it’s way to 270 mph through raw horsepower. Most modern sedans have more favorable aerodynamics than a Bugatti.

On the other hand, about 90% of the inputs from an EV charger make it to the wheels — a major factor that makes EVs so efficient. This means Elon can put, for example, 1000 hp of motors in his truck and only worry about rejecting 100 hp of heat at full load, an easy task. That’s less cooling than a Prius needs, and the truck can be designed with virtually zero air input constraints. He can swing his “3x the acceleration” dick around, and the only tradeoff is beefier driveline bits for the extra torque and bigger motors.

C_D lists on wikipedia: Note almost every production car below Cd=.24 is either a small displacement diesel or electric. No supercars in sight. en.wikipedia.org/wiki/Automobile_drag_coefficient

The Chiron’s standard Cd is .38. A 1995 ford windstar minivan has a standard Cd of .35. Not exactly a prestigious club, haha:

www.roadandtrack.com/new-cars/car-technology/a32329/these-rare-development-photos-tell-the-full-bugatti-chiron-story/

Hydrogen is a different story. The thermal efficiency of a FCEV -above 50% load- will dip down into the 40-50% range. The average efficiency of a diesel truck is ~45%, and it will dip to ~35% at max rpm, full load. ‘Ever seen the video where Trevor says the production trucks won’t need as much cooling as their current prototypes? He’s lying through his teeth. A 1000 hp FC would need more than 1000 hp of cooling; as much if not more than a 500hp diesel. It’s also important to note that less heat is lost in exhaust from a PEMFC, and lower operating temps mean that actual cooling airflow required is significantly more than an equivalent diesel would require. 500 hp diesels are already built like a brick with a front aftercooler & radiator the size of a football field. Trevor’s dick swinging has major consequences. More cooling means more drag, more weight in heat sinks/radiators, and more power draw to move coolant. These all create positive feedback loops, e.g. more air requirements mean less favorable aerodynamics & more drag, which means more power draw at speed, which means more cooling, which means more air requirements… It’s not all fun and games like the pure BEV Tesla is making.

Daimler’s recently announced H2 semi offering only has 300 kW, and they can recycle their waste heat to warm up incoming liquid hydrogen, Nikola uses compressed hydrogen and won’t have the same luxury. Running so much horsepower is a real head scratcher; one would think their head of R&D’s only experience was pouring concrete or something.

It’s also important to note US tractor-trailers are hard capped at 80,000 lbs. This means that every pound of tractor weight is a pound taken from potential cargo. Y’know… the part that actually makes money. One has to wonder why Nikola is keeping ~5000 lbs of batteries on the tractor; their website showed they aren’t fans of the stuff. It’s enough to run a loaded tractor two full hours on battery alone, more than a regenerative braking system would require or power peaking during a hill climb. It’s dumb; 100 kWh would be more than enough when an on-demand primemover like a hydrogen cell is also onboard. It’s an expensive and opulent display to the vexation of customers, who would, in all likelihood, much prefer an extra 2 tons of cargo capacity.

Features like 1000 hp and 250 kWh of battery appeal more to retail investors than trucking companies.

Actual conclusion time: I think that I pretty conclusively showed that:

  1. Nikola’s hydrogen cost projections are bogus. There isn’t even enough money there to pay for the electrolysis and compression, much less maintenance, depreciation, or labor.
  2. Nikola’s leasing costs undercutting diesel is bogus. One can disprove that with their own financial projections, much less the real costs of FCs and H2 electrolysis.
  3. Nikola’s plan to lease the trucks is totally divorced from reality, according to their own financial projections.
  4. Nikola’s projected per mile operating & maintenance expenses are beyond indefensible.
  5. A myriad of odd, marketing focused design choices limit the trucks on-road efficiency and utility to potential customers.

If you made it this far, congrats!

 

Disclaimer: This information is only for educational purposes. Do not make any investment decisions based on the information in this article. Do you own due diligence or consult your financial professional before making any investment decision.