The Tesla Semi Costs – Part 4

Tesla Semi


Tesla Semi

Tesla Semi

The trucking industry is very “green”. Not green like environmental green. Green like greenback green. Though we think fondly of the trucking lore and traditions at the driver-level, at the executive level every decision is dollar-driven, especially with the larger transportation businesses and retailers with self-contained distribution and shipping branches. No stone un-turned for improving Return-On-Investment. Everything scheduled to the minute and the mile. Pickups. Deliveries. Maintenance. Equipment replacement. Right down to the cost of a cup of coffee.

*This is the fourth of four articles on Tesla’s semi truck. Links to other related articles below:

Part 1

Part 2

Part 3

We still see some heavily-chromed semi-trucks on the road with girlie-mud flaps and tall polished vertical exhaust stacks. But, for the big fleet operations like Walmart, UPS, XPS Logistics, and Knight it’s all about squeezing out that last cost-effective cent of fuel efficiency and preventing that unplanned- repair-based delivery delay. Cabs with full air dams and fairings and van trailers with belly skirts. Hidden horizontal exhausts. Real-time OTA equipment monitoring and optimized maintenance schedules. Low rolling-resistance tires everywhere. Diesel engines with 40% thermal efficiency.  Instead of the 6 mpg average of just a few years ago, the new rigs are now are getting 8.5 mpg. Not driven by CAFE standards but by accounting standards.

Tesla Semi

Tesla Semi Slide 1

So, when Tesla rolled out the new electric semi-concept that offers low energy costs, premium performance, AND simplified maintenance,  “starting at $180K,” the green eyeshades went-on and the accounting calculator key-taps flew. And then the pre-order reservations poured in. Not because Tesla spelled “S3XY” but because it spelled $PROFIT – if the production-version price matches their claims.

Estimating the cost of the Tesla Semi from-scratch would be a daunting task so we made a significant simplifying assumption: The Tesla Semi “glider” without the electric propulsion system will be similar in construction and cost to a current-generation lightweighted Class 8 diesel tractor “glider” without diesel engine and required accessories.

Tesla Semi

Tesla Semi Slide 2

Like seemingly everything else in the world, many new current-generation Class 8 tractors for sale are advertised online, with complete specs and prices. As well are the replacement components, including engines, transmissions, axles, tanks, etc. Truck buyers sites tell us expected dealer markup embedded in the asking prices. And with some costs, we just have to make a “scientific” WAG based on what appears reasonable. Note we used a diesel “sleeper” as our starting point but the Tesla Semi is a “day cab.” Most diesels with the high-end aero configurations matching Tesla’s features are sleeper units and provided the best asking-price information.

Tesla Semi

Tesla Semi Slide 3

Tesla Semi

Tesla Semi Slide 4

Tesla Semi

Tesla Semi Slide 5

For the electric propulsion system, the biggest single cost is the battery pack. We used a recent article on present and future pack costs and re-adjusted $200/kWh down 15% for 2019 and another 15% for cost-efficiencies built into a 900 kWh pack to arrive at our $140/kWh pack cost including Thermal management system (TMS).

For most of the other components, we estimated Tesla’s cost based on the fact they are mass-produced Model 3 components and what their proportion might be in a typical Model 3’s actual cost.

Tesla Semi

Tesla Semi Slide 6


500-mile-range Semi – with the electric propulsion system with battery @ $144,000  +  $52,000 for the “glider,” Tesla’s raw cost is $192,000 BEFORE any gross margin added. With  20% gross margin added, the minimum profitable “sell price” adds up to about $230,000.

Tesla Semi

Tesla Semi Slide 7

300-mile range Semi –  with the electric propulsion system with battery @ $86,000  +  $52,000 for the “glider,” Tesla’s raw cost is $138,000 BEFORE any gross margin added. With  20% gross margin added, the minimum profitable “sell price” adds up to about $166,000.

Tesla Semi

Tesla Semi Slide 8


The 500-mile Semi: Tesla definitely can build a semi that delivers the power and the range they have claimed. For cost, we were very generous at estimating they could build a 500-mile-range pack with current or near-future battery and cell technology @ 140/kWh, but Tesla’s $196K cost is just too high. Tesla will take a loss at the “starting at” $180K price. To hit a 20% gross margin, Tesla’s total costs would have to $150K, but the electric propulsion drive costs alone are $144K. That leaves NOTHING to cover the entire “glider” cost. They must bring battery pack costs down another $35K to under $100K or adjust their selling price.

The 300-mile Semi: Assuming the 300-mile version is essentially the same as the long-range version other than battery size, it is more price-competitive. Tesla’s $138K cost, but it only gives a 9% gross margin at the “starting at” $150K price. Tesla needs to get the price down another $13K to hit 20% margin*.

Because the 300-mile version price is closer to achieving a viable profit margin and more conducive to lower charge rates and distribution center charging, we predict this will be the first one Tesla releases for production and accepts production orders for. The 500-mile version will remain in concept-form until battery costs drop to $100/kWh at the pack level.

*Note that Tesla brought out two semi’s at the reveal and we’ve seen two driving around on videos. The black one is a “short cab” version and may be targeted for local-haul beverage distribution, which frequently has lower trailer heights. Local haul would also be more conducive for 300-mile range batteries. The short cab version would be even lighter and might have a less expensive “glider” construction, helping lift the gross-margin with less battery-cost-reduction.

Tesla Semi

Tesla Semi Slide 9

*George Bower contributed to this article.

Category: Tesla


25 responses to "The Tesla Semi Costs – Part 4"
  1. morrisg says:

    I believe the slide 6 costs for the drives and electronics are way overpriced. Having designed an industrial BMS a few years ago, I can tell you that your estimate is 10x too high. Also, remember that the drive units are coming from the Model 3 high volume production line (see the YouTube videos of automated motor winding production). I doubt that Tesla’s costs are $1k per motor. And the $9k for reduction gear and drive axles is also too high. On board 50kw charger? Why would they even have one? AC level 2 charging is way too slow for this commercial application, plus the money savings from not including one dictates eliminating it.

    I enjoyed the overall overview of costs, but think this slide is off based on my electrical engineering experience.

    1. HVACman says:

      Good questions on cost and deserves a thorough reply – from a mechanical engineer’s experience.

      The “Drive unit” is the 271 HP PM-rotor motor. PM material is very expensive. The rotor would cost more than an equivalent induction motor. Cost based on likely mfgr’s raw cost to make a 250 HP induction motor. $1K may not be enough per motor. I may have underestimated the cost.

      “Driver axles w/ reduction gears”: Again, We’re talking about an 4 axles (one per wheel), with EACH axle capable of carrying 9,000 lbs and able to transmit 6,000 ft-lbs of torque. Also with a big air brake assembly. That is one massive shaft, with big, big bearings. And includes a triple-reduction gear-set, with dual shafts and bearings, also rated for 271 HP/6,000 ft-lbs of torque. A lot of machined steel in there.

      I based this price on the going price for a conventional set of 6×4 Dana Class 8 driver axles rated at 40K lbs load, but only half-the torque required of the Tesla semi- axles. Picture that. These Tesla motors with the reduction gears would BREAK a conventional Class 8 driver axle. $9K is assuming that Tesla is making them and doesn’t have Dana’s O&P.

      The batterh/DU TMS is the thermal management system to keep the battery and drive unit motors/gears cool while in operation or charging. A lot of heat generated from over 1,000 HP of motor and gears and battery, or when charging at 1.4 MW power rate. That means an industrial cooling system. Including pumps, hoses, front radiator, refrigerant/glycol heat exchanger, and controls. If you watch the semi reveal undercover videos, you’ll see a lot of big blue coolant hoses running around to each drive unit. I think $2,000 probably covers it.

      About the only thing in this estimate that was a “WAG” here was the BMS. But this the a BMS covering the equivalent to 12 Model 3 packs. Sensors everywhere. Logic. Likely some power relays or power electronics. Could be $1K, could be $2K.

      Regarding a “conventional” On-board charger where they plug in a big EVSE to AC – I believe every owner would want the option to be able to charge up some without relying on Tesla’s proprietary DCFC system. Where they could have a 480 VAC outlet and just plug in. Hook up at their service centers. At remote distribution centers. At regular customers’ warehouses.

      We expect that as a minimum with our EV passenger vehicles. Truckers would expect that, too.

      1. Doggydogworld says:

        The UBS teardown report pegs the Bolt’s 150 kW PM motor at 1220-1500. Other stuff (approx):
        700 Inverter
        200 BMS
        335 HV cables
        300 On-board charger
        550 Misc power management (DC-DC, control, etc.)

      2. Doodlebug says:

        I’m replying to your article (Part 4) and February 2 post.

        First, thanks for sharing your numbers. On the whole, they seem to be in the right ballpark. Most of all, they show that the key issue wrt the economics remains battery costs.

        Second, I looked up Paccar (2016) and Navistar (2017) annual reports, which have some data that is relevant to your calculations. Both indicate that COGS (cost of goods sold) is around 85% of the OEMs’ sale price. Deduct 1.5-3% for warranty and 2.5% for plant depreciation; that leaves around 80% COGs to build a truck. So, if the OEM (say FL or Navistar) nets $100k for a basic Semi, on average that truck cost $80k to build (labor & parts).

        Third, your $40k estimation for ICE components (including the tandem) jives with my estimations & those of Piper Jaffrey. Your $18k estimation for the electric drive components is similar to Piper’s estimate also (theirs is $13k + the cost to manufacture a tandem).

        Fourth, I saw you put in some costs for what I call the Tesla Semi fancy bits (e.g. independent front suspension) but you didn’t include all of the ones I’m thinking of. Like, that Armor windshield will be expensive, spaceframe chassis and self-driving hardware all bring extra costs.

        Once again, thanks for all the work.

    2. scottf200 says:

      Re: AC level 2 charging is way too slow for this commercial application
      Maybe it would not be for ‘recharging’ but for warming up the cab, running the computers, etc while it is outside the shop (i.e. you have other rigs on the fast DC chargers). J1772 plug charges are pretty cheap to provide electricity / simplicity for my examples of uses.

      1. Pushmi-Pullyu says:

        Yeah, I think an L2 charger would be useful for a sleeper cab semi truck, so the driver can plug in during his sleeping shift to run the cab’s A/C and/or heater, plus the auxiliary power needed if he has a mini-refrigerator and/or other appliances installed.

        Many drivers might never use it, especially “day cab” drivers, but other drivers might use it a lot. Whether it should be standard equipment or an option, is perhaps a question worth discussing.

  2. Jim says:

    No worries, they’ll make it up in volume. ?

    1. Don’t ‘forgot’: the 1st Trucks are the ‘Founders Series’, at $200,000 a pop! Likely 1st 1,000 units, not likely less, but maybe more units!

      1. Pushmi-Pullyu says:

        Well, let’s see how many customers actually buy the more pricey “Founders’ Series” trucks. As I see it, commercial truck operators aren’t interested in fancy; they’re interested in profitable.

        However, all the chrome seen on some independent truckers’ rigs may indicate otherwise. 🙂

        1. georgeS says:

          They might buy a custom Tesla trailer though and that’s not included here. As Keith pointed out u need regen in the trailer to capture most of the energy on a long steep grade.

          Minimal batteries in the trailer plus it could autonomously move around the yard…..and at the same time add a little fancy accessory money to the pot to make up for the marginal economics of selling a long range cab.

  3. God/Bacardi says:

    Bottom line is there’s no way Tesla is taking a loss on producing them…Remember they have to build megachargers, they have to guarantee that seven cent electricity and have cover both bumper to bumper and long term drivetrain warranties…

    1. Pushmi-Pullyu says:

      Yes, but I don’t recall Tesla saying they were going to pay for the installation of the Megachargers. Their customers may well have to pay for those.

  4. Mike McDonald says:

    I heard a few months back that in a conference call a Tesla exec stated the cost for their gigafactory-produced batteries is less than $100/kWh right now, compared to $140/kWh in a few years. Now, I’m not sure if this is at the pack level, but if it is, then I believe this 29% reduction in the ESS amounts to $90k, a reduction of $36k in the cost (for the 500 mile one). The author estimates a materials cost of 192k, so 192 – 36 = 156k. 20% markup gives $187k. A little closer to the $180 cited. Combine that with better savings on the motors, reduction gears, and computer systems which they already mass produce (/buy?) and it becomes reasonable IMO.

    1. georgeS says:

      There’s definitely room your arguement that says Tesla is at 100$/kwh on cells NOW. Add 40% for the TMS and that agrees with the assumption in the article.

      So I definitely think Tesl’a numbers are doable….but it means the pack needs to be 100$/kwh not 140.

      Here’s an article where Argonne predicted 135$/kwh for pack price for Model 3.

      1. stan1 says:

        If they are at $100/kWh for cells now, then $90/kWh would be a conservative expectation for cell costs at launch.

        1. Pushmi-Pullyu says:

          Yes, we need to remember that the question here isn’t what Tesla’s current battery costs are, but what they will be in 2-3 years, when the Semi Truck goes into production. It seems pretty clear Tesla is planning on their battery costs dropping somewhat in the meantime.

          Some have suggested this indicates Tesla is going to put solid-state batteries, or some other radically improved batteries, into production by then. I don’t see it that way; I think Tesla is just counting on their vertical integration and economy of scale (that is, lower unit costs from high volume production) to bring down the cost that much.

          Just my opinion, of course.

    2. Doggydogworld says:

      It’s nice to dream about super-cheap batteries, but these whisper prices have no basis in reality. Tesla sells Powerpacks and Powerwalls at $400+/kWh and still has negative gross margins. Panasonic cells are $300+/kWh in bulk at wholesale. Tesla pays less than that, of course, but I can’t see them paying less than $200/kWh today for cells.

      Panasonic Nevada makes cells with the same purchased graphite and cathode powders as in Japan, on the same type of machines. The GF will have more of these machines, but having 75 cell lines instead of 25 doesn’t dramatically slash costs. The cost of the actual building in Nevada shifts from Panasonic to Tesla, but that’s just accounting. Battery pricing still has to recover that cost.

      I’m sure Panasonic uses the latest generation machines at the GF, and Tesla’s automated pack assembly machines will be more efficient once they stop destroying the subassemblies. But the cost curve is not steep at this point. $150/kWh for cells and $200 for packs would be terrific this year. This article’s $140/kWh at the pack level in late 2020 would be a great achievement.

      As this article shows, Tesla’s semi pricing needs much lower costs. Tesla is just playing their typical game – announce very low costs on a very aggressive schedule, but don’t actually sell at the announced prices for 1-2 years later than scheduled (if ever). Think 90k Roadster in 2006 vs. 109k in 2008. Or 57k Model S that was cancelled due to “lack of demand”. At the 2013 shareholder meeting Musk said the low cost Bluestar, aka Model 3, would ship in 2016 or 17, it’s Feb 2018 and the $35k model is nowhere in sight.

      At the Semi launch Musk said “order now and get it in 3 years”, roughly the same schedule the Bluestar was on in 2013. So the announced Semi prices won’t come into play until early 2023. At least for the 500 mile version. The 300 miler, which works fine for distribution centers and such, could make it a year earlier. Sure, they’ll make a few for internal use in 2019 and “start shipping to custoemrs” in 2020. But those will be Founders Series at higher prices with expensive service contracts that limit Tesla’s losses.

      1. Per: “Or 57k Model S that was cancelled due to “lack of demand”.” Point to note: They cancelled further ‘Orders’, but they made good on their promised price, for those that Were Ordered!

        Not only that, the 40 kWh Model S that they ‘Cancelled’, was not ever going to have Supercharger Access, plus it was not just shorter range, but Slower, than the 60 kWh version that they ultimately delivered to buyers of the 40, but with simple Software limits to usable energy, of that 40 kWh!

        These cars had the acceleration of the Model S 60 kWh pack cars; Plus the Option to Both, Upgrade to the full 60 kWh accessible, And add Supercharger Access! (Which many did, later, as they had the money, later!)

        One of my EV buddies bought one of the Model S ’40 kWh’ models, online, at the lower pitched price, on the very day they announced the ‘Cancellation’, even though another EV buddy tried to buy at the store, but they would not do the order! (I think the company informed staff, but the website was not updated sufficiently fast enough, allowing my buddy to slip through with a last minute order!)

      2. Pushmi-Pullyu says:

        “…I can’t see them paying less than $200/kWh today for cells.”

        I think you’re wrong. There was what appeared to be a credible claim for $180/kWh at the pack level for Tesla some time back, maybe a year or more ago. With cells from Gigafactory One, it’s reasonable to think their costs will be considerably lower.

        “Tesla sells Powerpacks and Powerwalls at $400+/kWh and still has negative gross margins.”

        I submit that the much higher per-kWh prices for PowerPacks and PowerWalls have a lot to do with low volume of production. Furthermore, those prices include the costs for power electronics, such as the inverter. Contrariwise, when figuring costs for BEVs, the inverter price is not figured as part of the battery pack. It’s more commonly lumped in with the cost of the motor, or else the price for the PEM (Power Electronics Module) is figured separately.

        1. Doggydogworld says:

          Low volume shouldn’t affect Powerpack/Powerwall gross margins much, unless they built out a huge section in the GF for them that sits idle most of the time. And inverters are a small percentage of the whole.

          Tesla’s Model S/X margins don’t indicate low battery prices, either. All the evidence I can find supports $200+/kWh for cells. The only evidence for lower is “some guy from Tesla said something once”.

          1. georgeS says:


            GM already said they are paying 140$/kwh for cells from LG. I would think GF should be able to get lower than that. Plus the link I mentioned above. Argonne already concluded it could be done at 135$/kwh including the pack. I don’t think it is unreasonable to think Tesla can do model 3 packs for 140$/kwh.

            My understanding is GF includes some changes to the CELL production line machines and processes to make the process more efficient. JB has talked about this but they are tight lipped about what it is. Plus we already know Tesla has automated the pack assembly costs.

            1. Doggydogworld says:

              George, Figure 9 in the Argonne piece shows $10k+ for a 200 mile pack. Base Model 3 is a bit over 50 kWh so I get $200/kWh not $135.

              I no longer believe all the numbers people throw around. Big contract pricing is flat for years (GM’s original chart showed $145 from 2017-20, then declining). The customer basically “buys the factory” with the big contract and gets the lower pricing from day one. But the price status flat until the equipment is fully paid off.

              Tesla is still on a 2014 contract, that’s why I say they pay $200 for cells. GF cells are cheaper, I’d guess $125ish including building depreciation, but will stay at that level for years.

  5. Bill Howland says:

    No offense, but I wouldn’t mind hearing some more information from Tesla as to the specs of the truck.

    But I’m mainly interested in the battery and charging since I’m not buying one anytime soon anyway.

  6. John Doe says:

    I’m confused about some of these assumptions. Can you clarify?

    1. Stated assumption, “The Tesla Semi…will be similar in construction and cost to a current-generation lightweighted Class 8 diesel…” – How could this be possible when the dominant cabs (and interiors) on the road today are tooled and sold in the hundreds of thousands or even millions over their lifetime? If you assume Tesla has best-in-world battery prices due to volumes, doesn’t it go the other way for the “truck” parts where the existing industry has an advantage? Are you assuming they’ll have the same volumes as the current dominant OEMs?

    2. “25% gross margin” for the OEM (slide 3). Where does number this come from?

    3. Aluminum cab panels (slide 4): $1.5k saved – How? Already industry standard.

    4. Wide base singles (slide 4): $1k saved – How? Already industry standard.

    5. No passenger seat + Tesla controls: $2.5k saved – How? Why not divide these unrelated items? A typical (non-suspended) “shotgun” seat doesn’t cost much. Compare Tesla’s fancy displays and controls to an industry standard truck today and it’s hard to imagine how it saves money.

    Some of the huge assumptions you made are just a WAG, like you said, but some of these seems like you threw the kitchen sink in to tilt the scales.

    1. Doodlebug says:

      Dear John Doe,
      I agree with your point on volumes – and hence cost advantage of today’s OEMs.

      But… did you see that in Tesla’s recent conference call they said they would like to be building 100k Semis/year 4 years from now? [of course, believe it when we see it, but you get the point about costs – they don’t intend to be at a disadvantage forever].

      One thing to point out is that Tesla probably will have an advantage in motor/inverter costs as well as in battery costs, compared to most truck OEMs.

      The exception to that is probably VW. Because of electric car volumes VW will be in the same position as Tesla: cheap batteries and electric drive components. Scania is moving aggressively on electric heavy trucks. Navistar has everything to gain strategically by being a fast mover in electric trucks in the US in order to undo its current market share woes (11%). They also have less to lose in terms of stranded asset investments. And they have access to shared technology with Scania and MAN.

Leave a Reply