Why the 1,000,000 mile Tesla Semi matters


In spite of all the debates the Tesla Semi announcement kicked off over diesel vs electric, the wild speculation over battery packs that may range anywhere from just a few thousand pounds to up over 20,000, and the “Elon is only doing this to boost his stock price” investors (the percentage of whom are short $TSLA is curiously high), one thing is becoming clearer and clearer by the day:

The Tesla Semi, based on real numbers not speculative ones, is cheaper than even next-generation semi-trucks by a longshot. Many hours of research so-you-don’t-have-to later, and “cheaper” really is what it all boils down to.

But just how much cheaper, and why does cheap matter?

Let’s dive in.

All data is as of November 25, 2017 unless otherwise noted. All facts and figures are in reference to US Truck Class 8 “Heavy Duty” vehicles also known as “Semi Trailers”, more commonly referred to as “Semi Trucks”, or the Tesla Semi as announced on November 16, 2017.

The primary source of semi-truck industry data is the American Transportation Research Institute, or ATRI. Specifically, their 2017 survey on operational costs faced by the trucking industry, aggregating the data of nearly 9 billion miles logged in 2016, by more than 80,000 trucks, across a broad sample set of transportation companies. You can get a copy of the same data I’ve referenced, for free, in the Sources section at the end of this article.

According to ATRI we can safely assume the following:

  • A typical truck logs over 100,000 miles per year
  • Trucks are replaced, on average, every 750,000 miles
  • Average fuel economy is 6.3 MPG
  • Average marginal cost per mile of repair & maintenance is $0.166

What about diesel fuel costs? According to the U.S. Energy Information Administration, the EIA, average diesel prices in the U.S. for November 20th, 2017, were $2.912 per gallon.

How do we fuel the Tesla Semi? Electricity, of course! Taking Tesla at their word until data from actual launch models proves them wrong (or right!), we can assume the Tesla Semi 500-mile range model will require 2 kWh, or kilowatt hours, of electricity per mile. The Tesla Semi page actually claims, “less than 2 kWh per mile”, but 2 is an easy number to work with and I don’t want to try and guess at whether they’ll achieve 1.5 or 1.9997, either of which is “less than 2”.

What will that electricity cost? Going back to the EIA, as of August 2017 it was $0.1007 per kWh (see: Transportation sector). At their announcement, Tesla listed the price as $0.07 per kWh. What gives? They’re citing the industrial sector price of electricity, which Tesla gets access to. It appears Tesla plans to pass this price straight the customer, provided you charge up on Tesla’s next-generation Megacharger network (which has yet to come online). I would love to assume Transportation sector pricing, since the Megacharger network doesn’t exist, however there’s a catch: it’s not known whether you can charge a Tesla Semi without a Megacharger. $0.07/kWh it is.

Last but not least, what do we know about maintenance and repairs when it comes to the Tesla?

Sadly the Tesla Semi web page and the announcement done give us much. What we can do is leverage Tesla’s current track record with the Model S against similar sedans, or reference any preventative maintenance checklist for semi-trucks.

These are totally different vehicles from sedans, so let’s do the latter.

Google brings me to, “The importance of a preventive maintenance program” by North Dixie Truck & Trailer, a family-owned semi mechanic shop in Ohio, who wrote up just such a checklist:

  • Engine oil and filter changes
  • Transmission fluid
  • Fuel system
  • Cooling system
  • Engine and transmission mounts
  • Drive shafts or CV joints
  • Belts and hoses
  • Tune-ups
  • Electrical system components
  • Braking system
  • Steering and suspension system
  • Tires, wheels, rims
  • Exhaust system
  • Undercarriage and frame
  • Exterior and interior lights
  • Body, glass and mirrors
  • Windshield wiper system
  • The horn
  • Seatbelts and seats
  • Checks fluid leaks and the auxiliary systems

We can scratch half the items off the list without much thought, and note the items we’re crossing out are the more costly ones. Let’s call it a flat 50% of the cost of repairs and maintenance. It’s potentially far less, but I’d rather stop at reasonable than stretch out into outlandish.


Let’s see how this all pans out so far:



Nearly $250,000 saved at the typical truck replacement time of 750,000 miles driven.


Wait. Just how impactful is this?

According to Tesla, with the Tesla Semi we’re getting a 1,000,000-mile vehicle. Based on that claim we can expect to save nearly $350,000, at current diesel prices and average maintenance costs per mile, which is just $50k less than two Tesla Semi trucks all-in at year 1. Combined. Imagine expanding your fleet from 1 to 2 when the trucker next-door is swapping out their one truck. Now imagine you’ve got 10 trucks and the shop next-door does too. They all hit 750,000 miles at the same time. Your aggregate savings is enough for twelve new trucks through their first year. Or nearly 10 more through year three.

This gives owner-operators and small truck shops a massive advantage over the status quo.

Digging around some more, I couldn’t help but wonder why trucks tend to get replaced every 750,000 miles. Driving hundreds of thousands of miles adds up, and all the preventative care and ongoing maintenance in the world ultimately can’t beat time. So there’s this thing called an “inframe”. An inframe is another term for an engine overhaul job that is done without removing the engine from the frame of the truck. In other words, the work is literally done with the engine “in the frame”.

Inframes average roughly $30,000 all-in: parts, labor, and warranty on the work. This is a sizeable capital outlay, and so it makes sense why fleet owners rotate out their trucks at 750,000 miles on average. Why spend $30k on an engine overhaul, not accounting for all the other parts, when you can replace the whole truck for $120k? Sifting through trucker forums, it turns out a lot of small shops and owner-operators go with the overhaul, while a lot of mid-sized fleets and larger just swap their trucks out. I’m not accounting for either/or in my calculations, it’s just interesting to note.

Where do we go from here?

Rummaging around through the world of semi-trucks, totally foreign to me until now, I started coming across talk of “super trucks” that were making leaps in fuel efficiency. So I did exactly what you would think I’d do: I jumped right in.


Enter the Next-Generation Super Truck

September 25, 2017. “Run On Less Heavy-Duty Truck Rally Hits 10 MPG Average“. A popular trucking blog posts an article confirming fuel efficiencies averaging 10.1 mpg. One driver posted 12.8 mpg. I then found this article, “Performance Update on my 2018 Cascadia“, where “Team Run Smart” averaged 9.97 mpg over 20,000 miles.

Next generation “super trucks” hit the market in 2017, at higher entry-prices due to the number of additional fuel-efficiency options required, netting as high as 10+ mpg. I specced a rig similar to Team Run Smart’s at $160,000. I’ve also looked through used truck sites and found similar lightly-used vehicles for as low as $150,000. Still, we’re assuming new pricing across the board, so $160,000 it is.

Based on more reading at the Daimler and Freightliner sites, I’ve assumed a higher repair & maintenance cost of +10% for these next-gen vehicles. These trucks have quite a few more moving parts, new fuel efficiency systems, and may also require new training for labor. All of these things lift the cost, but since we’re stretching it out over cost-per-mile, +10% is as high as I’d like to go. Note the end result of the adjustment is minor, accounting for just $16,600 over 1,000,000 miles.


Once again, the Tesla Semi comes out ahead. The difference isn’t quite enough for a new truck by itself, but remember that $30,000 inframe? Spend half that, plus the savings, and increase your fleet size by 1 brand new vehicle. Over 10 years that’s still eight new trucks, all in for year 1, or six (actually 6.8) all in for 3 years.

Wow was an understatement.


So why does the 1,000,000 mile Tesla Semi matter?

To answer that question, I need us to look at one more factor together. Let’s go back to the American Transportation Research Institute annual survey.

  • “The American Trucking Associations (ATA) estimates a shortage of nearly 50,000 drivers, with projections that the shortage could increase to 175,000 by 2025.”
  • “An ATRI study in 2014 identified alarming demographic trends in trucking – with 55.5 percent of its workforce 45 and older, and less than five percent of its workforce in the 20 to 24 year old age bracket.”

We forgot to take into account the workforce.

The estimated shortage of drivers means pent-up demand for as many as 50,000 more trucks on the road and actively contributing to the economy. That’s another few billion miles, given our earlier assumptions. Or it’s “just” another billion miles but then makes many of the other 9 billion miles that much safer for our existing truckers by allowing them to move to team-based driving.

Can we identify any other tangible effects of this shortage on the industry? Sure we can. Average wages have gone up for four years in a row, while benefits have increased 33% in the same time period. That’s as much as a billion dollar increase in employment costs across the industry in the last four years.

What technology is coming, right around the corner, that may reduce driver demand? How can we get away with miminal team driving, or even reduce solo driving down to just monitoring systems and doing diagnostic checks along the way?

Autonomous trucking. Several companies, from startups to goliaths, have been testing their tech for at least a couple of years now. Tesla themselves recently committed to launching its own fleet of 100 Tesla Semi trucks, arguably as a marketing ploy but also as a way to “eat their own dog food” due to their enormous transportation costs. They could put the autonomous fleet model to the test themselves, logging roughly 10,000,000 miles in a single year of operations, as that fleet of 100 trucks racks up anywhere from 250 to 500 miles a day each, or the distance between their Gigafactory in Nevada and their auto plant in California.

Hey, someone has to move all those batteries.



An Analysis of the Operational Costs of Trucking: 2017 Update, American Transportation Research Institute, October 2017, http://atri-online.org/2017/10/18/an-analysis-of-the-operational-costs-of-trucking-2017-update/ (requires form submission)

Tesla Semi, https://www.tesla.com/semi/

Gasoline and Diesel Fuel Update, U.S. Energy Information Administration,   https://www.eia.gov/petroleum/gasdiesel/

Average Price of Electricity to Ultimate Customers by End-Use Sector, U.S. Energy Information Administration  https://www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_5_6_a

Compare, Nikola Corp | Nikola One, https://nikolamotor.com/one#compare

  • This maintenance cost per mile conflicts with the maintenance costs published by ATRI, but does line up with our assumed 2:1 cost ratio ($0.12:0.06 = 2:1) for repairs & maintenance of diesel:electric.

Trucking Truth, The Complete Guide to a Career in Trucking, http://www.truckingtruth.com/guide-pages/

“how many miles does a tractor last?”, Truckers Report,   https://www.thetruckersreport.com/truckingindustryforum/threads/how-many-miles-does-a-tractor-last.144025/

“How much did your last in-frame cost?”, Truckers Report, https://www.thetruckersreport.com/truckingindustryforum/threads/how-much-did-your-last-in-frame-cost.122430/





Just because I dug into the numbers a bit, let’s talk about those potentially super-heavy battery packs:

Key Specs on Tesla’s Electric Semi-Truck Still Secret, Trucks.com, https://www.trucks.com/2017/11/20/key-telsa-truck-specs-still-secret/

Alright, actually I’m not going to be making any battery pack claims here as I haven’t researched battery pack technology and form factors enough to do so. What I can do, though, is talk about what you won’t find in the Tesla Semi:

A 2,500-3,200 lb diesel engine. A Cummins x15 weighs nearly 3,000 lbs dry (without fuel).

1,400-2,100 lbs of diesel fuel.

A 650-900 lb transmission. Newer transmissions are lighter, with a high-end Detroit DT12 weighing only 650 lbs dry (PDF).

Several feet of excess hi-tensile steel in the engine compartment, which is gone now, not to mention frame for mounting a transmission, and fuel tanks, and this… and that…

An instrument-laden console.

Massive seats that, accounting for weight, haven’t improved much in decades (upwards of 200 lbs!)

50+ lbs of hoses and belts.

We’re at roughly 6,000lbs of weight reduction and counting… bring on the batteries!


3D Printing, Selective Recycling, and Near-Infinite Reusability


Traditional manufacturing in its simplest form involves taking raw materials and transforming them through one or more processes into finished products. One such process is the taking of some large volume of a raw material and stripping it down into a single part. Hundreds or thousands of such parts are then combined and result in things you use every day be that your car, the gas pump, the baby stroller, your office door, or any one of the hundreds of millions of products available around the world. This particular process, known as Subtractive Manufacturing, results in billions of tons of waste annually, has enormous capital costs, and racks up potentially trillions in externalized costs such as the environmental effects of over-extracting through mining and other extraction processes.

In contrast to Subtractive Manufacturing we have Additive Manufacturing, more commonly known as 3D Printing, or the building up of parts from raw materials, literally adding layers of raw material one upon the other to create a part. Ask someone, “What is 3D Printing?” and most people will answer with a story of seeing some plastic toy being built up over a bit of time at some consumer store demo or on YouTube. What they aren’t seeing are the large-scale commercial printers that are capable of printing full-size car, truck, and construction equipment bodies; real estate “parts” such as modular rooms and even entire houses; and so on.

The many impacts of the outputs of 3D Printing are discussed quite regularly today, whether those are conversations on the reduction or elimination in tens of thousands (and more likely millions, over time) of manufacturing jobs, or the ability to have nearly anything custom-built – for enough money of course. Meanwhile the discussion of inputs largely revolves around the fraction of raw materials required and the near-zero absence of materials waste. What’s often left out of the latter conversation is the idea that, if one can build a part up by layering raw materials on top of one another, one can reduce a part to its raw materials with surprisingly similar results.

Modern recycling processes, known collectively as Selective Recycling, are focused on just that. Armed with the latest in sorting and “plucking” systems, materials identification algorithms, and chemical processes, commercial “recovery plants” are advancing to a state where in some cases nearly all raw material can be recaptured from those gadgets we return in exchange for the latest hit item, or from that industrial machine that has reached its end-of-life and is being replaced with a more advanced system.

While some materials are still at low recovery rates versus others, for example only roughly 60% of iron can be recovered through advanced recycling processes while as much as of 99.9% of copper can be recovered, what happens as these advances continue to encompass more of our periodic table of elements?

What happens in as we approach a near-equilibrium state where we recover nearly all of the raw materials used in our production of… anything? What will become of the materials extraction industry in that time? What will our approaches to waste and to recycling transform into? What of our concepts of the retail store, of online shopping, or industrial equipment sales when we are able to simply print whatever it is we want or need? How about the effects on the shipping industry?

What becomes of our thoughts on the value of a thing when scarcity is decimated?

Like the impact that SpaceX’s approach to rockets is having on the space industry, the impact of this combination of Additive Manufacturing and Selective Recycling will be monumentally far-reaching. If you’re interested in exploring the significance of these transformational technologies further you can follow me here, on Twitter, or on LinkedIn as I scout ahead into our abundant and amazing future.


Additional reading on recent advances in recycling technology

Sandra R. Mueller, Patrick A. Wäger, David A. Turner, Peter J. Shaw, Ian D. Williams, A framework for evaluating the accessibility of raw materials from end-of-life products and the Earth’s crust, In Waste Management, Volume 68, 2017, Pages 534-546, ISSN 0956-053X, https://doi.org/10.1016/j.wasman.2017.05.043

Tianzu Yang, Pengchun Zhu, Weifeng Liu, Lin Chen, Duchao Zhang, Recovery of tin from metal powders of waste printed circuit boards, In Waste Management, Volume 68, 2017, Pages 449-457, ISSN 0956-053X, https://doi.org/10.1016/j.wasman.2017.06.019

Zhi-Yuan Zhang, Fu-Shen Zhang, TianQi Yao, An environmentally friendly ball milling process for recovery of valuable metals from e-waste scraps, In Waste Management, Volume 68, 2017, Pages 490-497, ISSN 0956-053X, https://doi.org/10.1016/j.wasman.2017.07.029

Bowyer, J.; Bratkovich, S.; Fernholz, K.; Frank, M.; Groot, H.; Howe, J.; Pepke, E. Understanding Steel Recovery and Recycling Rates and Limitations to Recycling; Dovetail Partners Inc.: Minneapolis, MN, USA, 2015; pp. 1–12.

Self-driving cars are not the future


Imagine you’re in the New York area in 1805. Traveling to Philadelphia takes you roughly a day and a half. You must stop to refuel – yourself, your companions, the horses, and so on along the way. Compared to today, travel is slow, uncomfortable, and fraught with risk. Only one century later, railroads have spread across the country, like fire across a wheat field, and your travel time is cut to only several hours. Better yet, you don’t have to pay much attention any more, at least not once you board the train. But soon, just a couple of decades more go by and then there is that damnable automobile, and while it brings you far greater mobility it again steals your time away from more productive things.

Back in the modern day, travel by rail is still a widely-used mode of transport within and across many developed nations, but travel by automobile is the clear winner due to the greater access it provides. People drive across towns, cities, states, and countries. The average American drives roughly 13,000 miles per year as of mid-2016, according to the US Federal Highway Administration, all working out to roughly an hour a day per American. Looking at data from several European nations, a 2012 European Commission (pdf) study uncovered a daily automotive commute of about 40 minutes per day, per person. All of these cars and roads and parking lots take up enormous amounts of space. All for little more than the draining away of millions of collective hours per day and millions of acres of land.

Close, but… no, not quite close at all.

Self-driving cars will return some of this time, and some of this space, but by themselves they are not the future. And they certainly aren’t the 21st-century equivalent to the “Space Race”. The ‘Race, for all its great scientific achievements, was a race between superpowers seeking to claim technological superiority over one another. For Elon Musk of Tesla, for Jeff Bezos of Amazon, and for the many other lesser-known champions of the actual movement going on all around us, it isn’t about one country proving superiority over another, and it’s not about just automating transport. It’s about something far greater.

For now, let’s focus on the self-driving cars.

Imagine you’re in the New York area in 2027. You wake up, perhaps an hour later and better-rested than you had to 10 years ago. You have some breakfast, feed your kids and send them off to school. You take the dog out to run around the larger yard you have as you no longer have a need for a garage or a driveway.

You tap your watch, your glasses, perhaps a microcomputer in the back of your hand or forearm if that’s the way things go, or maybe you simply speak, but through whatever method a car just shows up. It takes you where you tell it to take you. You pay a small fee and get there in half the time it took you in 2017 because there is no traffic. You might even get there in a quarter of the time because there is no traffic and the car is so safe it can travel at 100 MPH or more. Perhaps you intentionally pay for the “slow trip” at half price to enjoy the scenery while you start your work day during the ride into work since you don’t have to pay attention to driving any more.

You get into the city. The car drops you where you told it to and speeds off. You go about your day. You had such a stress-free experience in the morning, which will almost surely lead to a longer natural lifespan, that you’re much more productive through the morning. As lunch time approaches, you leave the office and walk across the street with some colleagues without having to wait for walk signals because the few cars up here just flow around you. Most of the traffic that used to be in the city has been moved underground over the last few years, leaving you with yet another stress-free experience. You notice fewer car parks than there used to be, all replaced by more useful real estate: more offices, apartment buildings, restaurants. More regular ‘ol, green-spaced, people-parks to relax in, you think. You go back into the office and eventually your work day ends. Maybe you work longer than you used to since you’re home in 5 minutes instead of the old 30-minute trip only a decade ago.

The future I’m alluding to will be even more different than proposed here, but I’m intentionally keeping the framing small for now.

Back again to the modern day. Looking forward it’s not self-driving cars that are the future, though they are clearly a step in the right direction. The real future movement, and it’s going on right now, isn’t a race for dominance to see who builds the plastic-and-metal box that can get us from point A to point B (that race has been over for a while now and Tesla won, although we could argue the meaningful part of that race is mass acceptance or mass utilization of automated transport). It’s not actually defined by any single one of the many more technologies coming this century that play a role in it. Technologies that may democratize our energy production and our physical product creation. Others that may enable us to work and connect both more closely and yet across greater distances than ever before. Still others that may mean we never again leave a single person to suffer hunger or homelessness. And then there are those inventions and innovations that will let us take our next bold steps into that place that Sir Arthur C. Clarke once called, “the sea of stars”.

If you think self-driving cars are this century’s ‘Race, think again. New technologies are coming that will radically reshape our very concept of the world and our abilities to create, to connect, and to communicate. The real race is not about a box, once a rocket and now a car, that takes us from A to B and shows off some nation’s might. It is, instead, driven along by those advances that will free up our collective time and space. The real 21st-century race is about leading the next great expansion of human civilization.

I’ll be writing more on each of these coming technologies, several already realized and others still in their infancy, in the weeks and months to come.

Stay tuned.