Shorting Oil, Hedging Tesla

[Oil4AsphaultOnly]

It doesn't matter for oil demand whether the drop-off is coming from EVs or higher efficiencies, but I was questioning the assumption that neroden made that this was all coming from EVs, as in "This is proof that the drop in gasoline consumption is being driven by electric car adoptions" If that is really true then show me the data - how much of the reduction in oil demand comes from EV sales vs improved CAFE.

Your view seems overly Tesla-centric as you say we need Tesla to shift appetite. I love Tesla, which is why I bought a Model S, but I think that this isn't something just left to Tesla as I believe other companies need to be part of this equation as well and build, and actively promote, EVs that people want to drive, although that will matter less if few people own cars in the future if we go to Transportation as a service.

Although this is anecdotal, the people who are buying short-range EV's and PHEV's are definitely not performance oriented. These buyers are either ecology or economy focused. The economy focused ones (of which I am one) must drive enough where the fuel-savings justifies the purchase of a dedicated commuter vehicle. The sales of these short-range EV's continues to grow. It's not hard data, but it's a reasonable support.

Edit: I (and those on the leaf forum) put ~15k miles per year, so permanently losing these high-mileage commuters definitely brings down the demand for oil early.

 

[jhm]

It doesn't matter for oil demand whether the drop-off is coming from EVs or higher efficiencies, but I was questioning the assumption that neroden made that this was all coming from EVs, as in "This is proof that the drop in gasoline consumption is being driven by electric car adoptions" If that is really true then show me the data - how much of the reduction in oil demand comes from EV sales vs improved CAFE.

Your view seems overly Tesla-centric as you say we need Tesla to shift appetite. I love Tesla, which is why I bought a Model S, but I think that this isn't something just left to Tesla as I believe other companies need to be part of this equation as well and build, and actively promote, EVs that people want to drive, although that will matter less if few people own cars in the future if we go to Transportation as a service.

Sure, I'm not interested in defending @neroden's thesis, but EV sales are a response to CAFE, making it hard to separate out their effects.

Regarding Tesla's role, normally I am happy to include all EVs. However, in the cast of shifting consumer expectations, I believe Tesla plays a unique role. Virtually all other EVs are made by traditional automakers which are heavily vested in ICE models. They enter the EV market primarily as a means to satisfy regulatory requirements and are disinclined to do anything that may devalue ICE vehicles in the eyes of consumers. Tesla, on the other hand, has nothing to lose and everything to gain by making ICE obsolete and less desirable in the eyes of consumers. So they are uniquely positioned to undermine demand for conventional vehicles, and that is why I placed focus on Tesla's actions. Of course, other exclusive EV makers may rise to prominence as well, and that would be quite helpful too.

 

[neroden]

How do you know that it is EVs and hybrids that are causing this trend rather than Californians buying more fuel efficient conventional ICE vehicles? Is that not a valid explanation?

First of all, it doesn't matter.

But the reason I know this is that there is no such thing as a more fuel efficient conventional ICE vehicle; we've basically hit the limits of efficiency for them, and we hit it almost 10 years ago.

Look at the more fuel efficient vehicles in any class, they're *all* hybrids; it's now the only way to get more efficiency without making the car smaller and lighter ("econobox"). And we don't see a shift to econoboxes; that would show up in the data, and it isn't happening.

We can discuss how much of this is due to non-plug-in-hybrids versus plug-ins, which would be an interesting question. There are a lot of "mild hybrids" on the market now and I think many of the people driving them don't even realize that they're hybrids -- particularly the Ford models.

You can also look at it from another perspective.

Traditionally, more efficient vehicles have been subject to Jevon's Paradox. It's cheaper to drive them longer distances, so they *get driven more*, so the amount of gasoline used stays constant. (In other words, people's fuel budget stays constant and the amount driven adjusts to the efficiency of the car.)

Now, plug-ins break that linkage and mean that gasoline usage can actually drop, *while* gasoline prices go down, *while* spending the same amount on fuel, *while* people drive more. Because fuel spending switches over to electricity. I think it's by far the most likely explanation for the observed combination fo data in California.

 

[jhm]

Traditionally, more efficient vehicles have been subject to Jevon's Paradox. It's cheaper to drive them longer distances, so they *get driven more*, so the amount of gasoline used stays constant. (In other words, people's fuel budget stays constant and the amount driven adjusts to the efficiency of the car.)

Now, plug-ins break that linkage and mean that gasoline usage can actually drop, *while* gasoline prices go down, *while* spending the same amount on fuel, *while* people drive more. Because fuel spending switches over to electricity. I think it's by far the most likely explanation for the observed combination fo data in California.

Ah, this is good. This is what I was trying to describe about the uptake of Model 3 being higher among heavy commuters than light drivers or commercial fleets versus private passenger vehicles. Those who spend the most on fuel will value efficiency and switching to cheaper energy the most all else being equal.

I think this is basic reason why most analyses about peak oil demand are flawed. They assume average use cases and ignore self-selection preferential uptake among high use cases. Thus, the analysts will tend to under estimate utilization, assuming it to be average and not high.

 

[jhm]

The electric semi discussion is going on in multiple places. I want to set out some info that is particularly important for understanding the impact on diesel demand.

So here is a good source for averaged fuel efficiency, Alternative Fuels Data Center: Maps and Data - Average Fuel Economy of Major Vehicle Categories. It is important to not that this is in gasoline gallon equivalents (GGE). What we need is diesel gallon equivalent (DGE). So 0.88GDE = GGE is what we need.

Type mpgge mpdge
Class 8 Truck 5.29 6.011
Transit Bus 3.26 3.7045
School Bus 6.30 7.159

So firstly we note that duty cycle for transit buses is much more demanding than for class 8 trucks. The big difference is stop and go urban streets versus long highways. I include school buses as a proxy for transit buses that may not be put to such a demanding use as transit buses.

Now let's get to the question of electric trucks and buses. NREL has studied the performance of Proterra electric transit buses. They find the efficiency to be 2.15kWh/mi which is 17.48 mpdge. So note that this is substantially higher than the 3.70 mpdge. Electric Buses Efficient As He**, NREL Finds

But the main thing I want to estimate is the electric replacement ratio, and it is

7.96 kWh/dge = 2.15 kWh/mi × 3.7045 mi/dge

This is amazingly close to the 8 kWh/gal that I have been touting, though that ratio is in GGE units. Really this should be 9.1 kWh/dge. So the Proterra bus is proving to be even more efficient than I would have anticipated.

So what does this tell us about what to expect for electric trucks? Let's apply my ratio of 9.1 kWh/dge. This leads to

1.51 kWh/mi = 9.1 kWh/dge ÷ 6.011 mi/dge

But using Proterra's ratio

1.32 kWh/mi = 7.96 kWh/dge ÷ 6.011 mi/dge

So this gives us estimates for average efficiency of electric class 8 trucks across the full spectrum of use and it is based on all the diesel trucks in use. Therefore it is plausible that Tesla could produce a truck that is more efficient that the average semi or that these truck would be put to uses that have higher than average efficiency. My own view is that Tesla will produce a truck with better aerodynamics and less roll resistance than the average truck. So I remain hopeful that the Tesla Semi could do better than 1.3 kWh/mile. We'll have to see what they announced.

So finally for analyses of impact on oil demand I am inclined to use 9 kWh/dge or 8 kWh/gge.

 

[Oil4AsphaultOnly]

The electric semi discussion is going on in multiple places. I want to set out some info that is particularly important for understanding the impact on diesel demand.

So here is a good source for averaged fuel efficiency, Alternative Fuels Data Center: Maps and Data - Average Fuel Economy of Major Vehicle Categories. It is important to not that this is in gasoline gallon equivalents (GGE). What we need is diesel gallon equivalent (DGE). So 0.88GDE = GGE is what we need.

Type mpgge mpdge
Class 8 Truck 5.29 6.011
Transit Bus 3.26 3.7045
School Bus 6.30 7.159

So firstly we note that duty cycle for transit buses is much more demanding than for class 8 trucks. The big difference is stop and go urban streets versus long highways. I include school buses as a proxy for transit buses that may not be put to such a demanding use as transit buses.

Now let's get to the question of electric trucks and buses. NREL has studied the performance of Proterra electric transit buses. They find the efficiency to be 2.15kWh/mi which is 17.48 mpdge. So note that this is substantially higher than the 3.70 mpdge. Electric Buses Efficient As He**, NREL Finds

But the main thing I want to estimate is the electric replacement ratio, and it is

7.96 kWh/dge = 2.15 kWh/mi × 3.7045 mi/dge

This is amazingly close to the 8 kWh/gal that I have been touting, though that ratio is in GGE units. Really this should be 9.1 kWh/dge. So the Proterra bus is proving to be even more efficient than I would have anticipated.

So what does this tell us about what to expect for electric trucks? Let's apply my ratio of 9.1 kWh/dge. This leads to

1.51 kWh/mi = 9.1 kWh/dge ÷ 6.011 mi/dge

But using Proterra's ratio

1.32 kWh/mi = 7.96 kWh/dge ÷ 6.011 mi/dge

So this gives us estimates for average efficiency of electric class 8 trucks across the full spectrum of use and it is based on all the diesel trucks in use. Therefore it is plausible that Tesla could produce a truck that is more efficient that the average semi or that these truck would be put to uses that have higher than average efficiency. My own view is that Tesla will produce a truck with better aerodynamics and less roll resistance than the average truck. So I remain hopeful that the Tesla Semi could do better than 1.3 kWh/mile. We'll have to see what they announced.

So finally for analyses of impact on oil demand I am inclined to use 9 kWh/dge or 8 kWh/gge.

Jhm, love your work and analysis! But I have to quibble with the use of transit bus numbers for Tesla semi impact.

The transit buses benefit the most from electrification because of regenerative breaking in urban environments and lower speeds of travel (so less impact from aerodynamic drag). The proterra buses also have smaller batteries meant for local transit uses versus BYD's buses that have much bigger (thus heavier and more energy consuming) batteries for longer range operation.

The semi trucks will miss out on all 3 benefits, so expect their kWh/mi consumption to be much higher. Sorry, but I don't have sources on hand to run the calculations.

 

[jhm]

Jhm, love your work and analysis! But I have to quibble with the use of transit bus numbers for Tesla semi impact.

The transit buses benefit the most from electrification because of regenerative breaking in urban environments and lower speeds of travel (so less impact from aerodynamic drag). The proterra buses also have smaller batteries meant for local transit uses versus BYD's buses that have much bigger (thus heavier and more energy consuming) batteries for longer range operation.

The semi trucks will miss out on all 3 benefits, so expect their kWh/mi consumption to be much higher. Sorry, but I don't have sources on hand to run the calculations.

The comparison of 3.70 mpdge for buses versus 6.01 mpdge for trucks already accounts for nearly all the differences in use. In my own Model S, for urban street driving I get about 0.34 kWh/mi, but in long distance freeway driving about 0.28 kWh/mi. So we should expect that semis will get much better mileage than transit buses.

Check out this bad boy: Proterra unveils new Catalyst E2 series at APTA Annual
The E2 Max has 660kWh battery for 350 miles nominal range (1.89 kWh/mi, EPA rated), but in testing they were able to do a 600 run (1.1 kWh/mi). This has a gross vehicle weight of 39,050 lbs, Proterra | Catalyst 40-Foot Transit Vehicle. The curb weight for this E2 series runs from 29,848 lbs for 440 kWh E2 to 33,061 for 660 kWh E2 Max. It looks like the key difference is just the battery size, an incremental 220 kWh for 3212 lbs, 68.5 Wh/lb. I think Tesla can match that pack level density (66.7 Wh/lb for the Model S), so this all seems reasonable.

Note also that fully loaded efficiency goes for 1.75 KWh/mi to 1.89 kWh/mi, an increase of 0.14 kWh per 3212 lbs incremental weight. So I think we can back into incremental 0.087 kWh/ton-mile. For sanity check this should be less than 1.89 kWh/mi / 19.5 ton = 0.097 kWh/ton-mi. So this checks. Incremental efficiency is about 90% of average efficiency. Adding weight does not increase wind drag, so incremental efficiency should be less than average efficiency.

So let's suppose they tested this (600 mile test) at medium passenger load, 36,000lbs or 18 tons and got 1.10 kWh/mi, so that is an average of 0.061 kWh/ton-mile, so the incremental efficiency is probably more like 0.055 kWh/ton-mile. We see that highway conditions are more favorable to carrying a heavy load. So let's scale this up to 60,000lbs or 30 ton for a middle weight in the Class 8 range. I get 1.76 kWh/mi = 1.1 kWh/mi + 12 ton * 0.055 kWh/ton-mile. Likewise, at the top of the range 40 ton, I get 2.31 kWh/mi. So from 18 ton to 40 ton, I get a range of 1.1 to 2.3 kWh/mile. At 40 ton, the 660 kWh may suffice for 285 mile range. Also not that the 660 kWh battery weighs 9636lbs or 4.8 ton about 12% of gross 40 ton weight. At 30 ton moderate load, we get 375 mile range.

So if Tesla got efficiencies and ranges comparable to this on a 660 kWh battery, I think I'd be fairly happy. But I do suspect that Tesla can and will do better than that. For one thing, in trucking you do not have to provide air conditioning, lighting and wifi to 40 passengers, so there is some cabin energy uses that is just not comparable. I also hold out hope that Tesla has a new pack design with better density than the Model S. It does not have to be so flat and so exposed to the environment. So that implies less outer surface area and less heating and cooling per kWh. So this could save energy in two ways less weight and less thermal regulation. And finally, I do think the calculation of incremental efficiency above is a little off. At highway speeds drag is the big driver of power use, not weight. Tesla will do an excellent job on getting a small coefficient of drag, and incremental weight won't matter so much.

So we'll have to see what Tesla comes up with, but if they just get into the ballpark with Proterra's E2 series, I think that will work just fine. Proterra tripled their sales last year and look to double again this year. So heavy vehicles can be pretty exciting for investors.

 

[jhm]

2020s To Be A Decade of Disorder For Oil | OilPrice.com

Includes nice chart on investments in oil and gas. There has been a pretty strong fall off in last two years, but total investment is still where is was in 2007 when oil consumption peaked in OECD countries. Given that we're still in a three-plus-year glut, I'm not convinced there is a need to go to back to investment levels over $550B. Not sure I buy the Chicken Little routine here. In range of $400 billion looks adequate to me.

 

[adiggs]

2020s To Be A Decade of Disorder For Oil | OilPrice.com

Includes nice chart on investments in oil and gas. There has been a pretty strong fall off in last two years, but total investment is still where is was in 2007 when oil consumption peaked in OECD countries. Given that we're still in a three-plus-year glut, I'm not convinced there is a need to go to back to investment levels over $550B. Not sure I buy the Chicken Little routine here. In range of $400 billion looks adequate to me.

Yeah - the recurring theme I've seen in articles talking about the low investment levels is that they never put it within a larger context (total investment at 2007 levels) - only that it's down over the last 3 years, ergo we're going to run out of oil in a couple of years when demand explodes beyond any possible availability.

Oh - except for all the oil that could be produced, but is being voluntarily withheld from the market right now. That also doesn't get into the article somehow

 

[Oil4AsphaultOnly]

The comparison of 3.70 mpdge for buses versus 6.01 mpdge for trucks already accounts for nearly all the differences in use. In my own Model S, for urban street driving I get about 0.34 kWh/mi, but in long distance freeway driving about 0.28 kWh/mi. So we should expect that semis will get much better mileage than transit buses.

I'm assuming you're referring to the electric versions of semi's versus transit buses here? If so, then I disagree with this. Specifically because the aerodynamic drag is significantly more of an issue with these large trucks than you'd see from regular passenger vehicles.

Some Cd numbers to reference:
Drag Coefficient

The force needed to accelerate a vehicle is:
F = ma + Rr + Fd (where m = mass of vehicle; a = acceleration - assuming 1m/s2; Rr = tire rolling resistance; Fd = aerodynamic drag force)

At low speeds, Fd is negligible and it's all about overcoming rolling resistance. At higher speeds, it's all about overcoming aero drag.

The formula for Fd is:
Fd = 0.5 * Cd * p * V^2 * A (where Cd = coefficent of drag; p = density of air; V = velocity; A = frontal area of vehicle)

You'll get constant velocity when the force for accelerating a vehicle equals the aero drag plus its rolling resistance.

Fd will be 36x higher @ 60mph versus just 10mph (at low speeds, aero drag is negligible compared to rolling resistance):
Fd @ 10mph = 0.5 * Cd * p * 100 * A
Fd @ 60mph = 0.5 * Cd * p * 3600 * A

For your model S with a Cd of 0.24 (which used 0.34kwh/mi in urban traffic), you would consume the same amount of energy at ~70mph:

If your model S magically got a Cd of 0.72 (drive with all the doors and hood open?), it would start consuming the same amount of energy (ignoring the difference in rolling resistance) at 41mph:
0.5 * 0.24 * p * 4900 * A = 0.5 * 0.72 * p * V^2 * A => V = 40.4 mph

The point being, for lower drag coefficients and lower frontal areas, you'll get a higher equilibrium speed where you'd consume less energy at that speed than crawling along in stop'n'go urban traffic. Any faster than that, and you'd consume more energy regardless (that's why autobahn has the superchargers spaced much closer together).

So for trucks with high Cd values, their equilibrium would be less than 41mph. With highway speeds being 55mph, those electric trucks would be consuming more energy than an equivalent electric truck stuck in urban traffic.

Edit: All this doesn't mean that Tesla Semi won't have a positive impact against fuel consumption, because it will, but just not as big of an impact as transit busses would have. And definitely requiring more electricity in the process.

Edit 2: Just to complete out the thought. The reason the math above shows that electric transit busses would use less energy per mile than electric semi-trucks, thereby running counter to what NREL reports about fuel consumption for ICE transit busses and trucks, is because of the 3 factors that electric transit busses benefit from that electric long-haul trucks cannot take advantage of: regenerative braking, lower speeds, and lighter weight (for lower rolling resistance)

 

[jhm]

@Oil4AsphaultOnly, perhaps you can explain why metro buses get 3.7 mpg while trucks get 6.0 mpg. Trucks are heavy and going faster, right?

 

[adiggs]

@Oil4AsphaultOnly, perhaps you can explain why metro buses get 3.7 mpg while trucks get 6.0 mpg. Trucks are heavy and going faster, right?

It seems that would be because of the constant stop and go driving that most transit buses experience. The speed and weight are the primary determinants of energy consumption at otherwise constant driving speeds.

The stop and go driving being an area of strength for electric drive, rather than an area of weakness.

 

[Oil4AsphaultOnly]

@Oil4AsphaultOnly, perhaps you can explain why metro buses get 3.7 mpg while trucks get 6.0 mpg. Trucks are heavy and going faster, right?

To expand on what adiggs said, a transit bus is constantly accelerating and decelerating, to the point where the transit bus is accelerating almost 1/3 the entire transit time. So a bus that runs a 6 hr route will have spent 2 of those hours accelerating (which consumes the most amount of fuel). While a long-haul truck accelerates to speed and then just cruises at constant rpm on the highway the rest of the time. In four 2-hour duty cycles (8-hrs of driving), the truck would have spent only 8 minutes (2 minutes to 55mph each time) accelerating the entire time.

Actually, the MPG rating of the hybrid prius perfectly exemplifies this. It gets better mileage in the stop'n'go driving of the city than highway, because it's onboard battery and motor is able to use regenerative braking to slow down instead of throwing it away as friction heat and brake dust.

 

[renim]

buses are significantly more aerodynamic than trucks
(ie for a constant speed, its easiest to electrify a bike, then a car, then a bus, lastly a tractor trailer)
there are however, a multitude of municipal truck uses that are ripe for electrification.

 

[jhm]

View attachment 223325
buses are significantly more aerodynamic than trucks
(ie for a constant speed, its easiest to electrify a bike, then a car, then a bus, lastly a tractor trailer)
there are however, a multitude of municipal truck uses that are ripe for electrification.

So If I understand this chart correctly, the crossover speed is the speed at which aerodynamic resistance equals the mechanical resistance. That speed, I suspect, is where the energy per mile is lowest. Below it mechanical resistance gets worse, above it drag gets worse.

So semis have cross over speeds is in range of 80 to 90 km/h (48 to 54 mph). This seems to be largely driven by the gross weight, 29 to 36 tonne. The advantage of improving the coefficient of drag is that it goes further out on this curve, lifting the cross over speed. Thus, the optimal speed would be higher. Even in the face of great roll resistance, improving aerodynamics is always beneficial.

 

[jhm]

To expand on what adiggs said, a transit bus is constantly accelerating and decelerating, to the point where the transit bus is accelerating almost 1/3 the entire transit time. So a bus that runs a 6 hr route will have spent 2 of those hours accelerating (which consumes the most amount of fuel). While a long-haul truck accelerates to speed and then just cruises at constant rpm on the highway the rest of the time. In four 2-hour duty cycles (8-hrs of driving), the truck would have spent only 8 minutes (2 minutes to 55mph each time) accelerating the entire time.

Actually, the MPG rating of the hybrid prius perfectly exemplifies this. It gets better mileage in the stop'n'go driving of the city than highway, because it's onboard battery and motor is able to use regenerative braking to slow down instead of throwing it away as friction heat and brake dust.

Right, so it is the volatility of speed that drives poor fuel economy in city driving. An electric drive train has an advantage here because regenerative braking retains energy that would otherwise be lost. Additionally, an EV drivetrain may have an advantage in idling. Cabin energy may still be used when stopped, but turning over an engine for no use is clearly a waste of energy.

So other than regenerative breaking and low energy idling, I think that the difference on mpg has little to nothing to do with the drivetrain. That is a bus or truck may well have engines of comparable size and efficiency. Assuming that to be the case in the average case, then the difference in mpg depends on the design of the vehicle and how it is used. So sure trucks tend to haul more weight and buses spend more time in city traffic, but that is already captured in the difference in mpg.

Why this is important, is that, when we contemplate what happens when a bus and truck have a comparable battery electric drivetrain of same size, the engine efficiency is the same whole the differences in vehicle design and use will continue to drive differences in watt hours per mile. Aside from regen and idling issues, the truck gets about 62% more miles per unit of energy than buses (6 mpg ÷ 3.7 mpg = 1.62). This is the essential point I've been trying to make. So if electric buses are getting say 2.15 kWh/mi, I would expect a truck to be about 62% more efficient, specifically, 1.33 = 2.15/1.62.

Granted, there is uncertainty about this, but not so much as to expect trucks to get worse mileage than buses when electric but better mileage than buses when diesel. Vehicle efficiency, not engine efficiency, is the first order issue.

Moreover, even as a second order issue regen is important to both high and city driving. Specifically, regen is very valuable when going down hill on highways, otherwise energy spent gaining elevation is lost declining elevation. Additionally, tight traffic often requires applying the brakes to maintain a safe following distance. So regen is valuable quite broadly.

The flipside of regen is acceleration. Every time a truck needs to climb a grade or has room to accelerate back up to a cruising speed, it is using a lot more power than needed to sustain momentum. I do think that battery electric drive has an advantage here with really good torque and smooth linear acceleration. So I think they may excel in dense urban traffic and mountainous areas.

One other stray thought, I wonder if autopilot could be tuned to optimize energy efficiency. In the Model S, AP tries too hard to maintain the target speed. Opportunities to coast or ease up on acceleration are not exploited. Analysis of road curvature and elevation changes could lead to an algorithm that uses minimal energy preserve momentum.

 

[neroden]

The semi trucks will miss out on all 3 benefits, so expect their kWh/mi consumption to be much higher. Sorry, but I don't have sources on hand to run the calculations.

Well, I have been talking about the key differences between long-haul trucking and short-haul trucking. (Short haul trucking is at least 1/3 of the trucking business; I've never found harder numbers than that, it might be over half.)

Short-haul trucking shares a *lot* of duty cycle features with transit buses. Lots of stop and go. But more downtime to recharge...

 

[Oil4AsphaultOnly]

Right, so it is the volatility of speed that drives poor fuel economy in city driving. An electric drive train has an advantage here because regenerative braking retains energy that would otherwise be lost. Additionally, an EV drivetrain may have an advantage in idling. Cabin energy may still be used when stopped, but turning over an engine for no use is clearly a waste of energy.

So other than regenerative breaking and low energy idling, I think that the difference on mpg has little to nothing to do with the drivetrain. That is a bus or truck may well have engines of comparable size and efficiency. Assuming that to be the case in the average case, then the difference in mpg depends on the design of the vehicle and how it is used. So sure trucks tend to haul more weight and buses spend more time in city traffic, but that is already captured in the difference in mpg.

Why this is important, is that, when we contemplate what happens when a bus and truck have a comparable battery electric drivetrain of same size, the engine efficiency is the same whole the differences in vehicle design and use will continue to drive differences in watt hours per mile. Aside from regen and idling issues, the truck gets about 62% more miles per unit of energy than buses (6 mpg ÷ 3.7 mpg = 1.62). This is the essential point I've been trying to make. So if electric buses are getting say 2.15 kWh/mi, I would expect a truck to be about 62% more efficient, specifically, 1.33 = 2.15/1.62.

No. The inherit characteristic differences between an internal combustion engine and an electric motor makes this point invalid.

Yes, if the electric truck/bus was stopping and going frequently between 60mph and 50mph, it would recover as much energy from regenerative braking as an electric truck/bus going between 15mph and 5mph, BUT it would also expend far more energy to cover the same distance at 60mph versus covering the same distance at 15mph due to energy consumed to overcome aerodynamic drag.

An ICE truck/bus, without the benefit of regen, simply has to consume less fuel overcoming aerodynamic drag than it wastes from braking and re-accelerating. The loss from air friction @ 65mph (41 mph in my earlier calculations were for electrified trucks/busses) was simply much smaller than the loss from re-accelerating the vehicle after slowing down. That's why it would consume less fuel on the highway versus city driving.

In trying to find a real-world "proof" of what I'm describing, I point you to the Lexus RX CUV:
RX450h: https://g.co/kgs/ojSWSL (31 city / 28 highway)
RX350: 2017 Lexus RX 350 - Google Search (20 city / 27 highway)
GMC Yukon: 2017 GMC Yukon - Google Search (16 city / 23 highway)

The only difference between the vehicles is essentially their powertrains, the hybrid has regenerative braking, while the non-hybrid does not.

Similar to your transit bus versus truck situation, the non-hybrid gets 35% better fuel consumption on the highway versus city driving. The greater the weight, the greater the difference between rolling resistance losses versus aerodynamic drag losses. That's why the yukon gets 43% better fuel consumption on the highway versus city driving.

An electrified transit bus versus electrified truck would have the hybrid's characteristics - worse fuel consumption on the highway versus city driving. The RX is actually fairly aerodynamic too - 0.33!! If it had a semi truck's coefficient of ~0.75, the difference would be much worse. This weird-mobile by Walmart is only able to reduce drag by 20% versus the Peterbuilt model 386 (Cd of ~0.6): Walmart Debuts Futuristic Truck

So I'm guessing ~0.48? Still about 35% worse than the RX450h, so will have significantly worse highway energy consumption versus city driving.

Does this convince you?

 

[Oil4AsphaultOnly]

Well, I have been talking about the key differences between long-haul trucking and short-haul trucking. (Short haul trucking is at least 1/3 of the trucking business; I've never found harder numbers than that, it might be over half.)

Short-haul trucking shares a *lot* of duty cycle features with transit buses. Lots of stop and go. But more downtime to recharge...

I did manage to find NREL stats on truck fleets: Fleet DNA: Commercial Fleet Vehicle Operating Data | Transportation Research | NREL

And based on the average speed distribution, it looks like about 1/2 for short haul trucking (average speeds of 18mph or lower).

Interestingly, the median on the low-speed side for trucks was 3-9mph, while the median for transit buses were for 9-12mph. Also, the max speed distribution shows that the truck drivers were split between those driving for efficiency (54-60mph) versus speed (69-72mph), while the high-speed bus drivers were strictly for speed (63-69mph).

Anyway, my issue with jhm was not about the distribution of data nor the use cases between transit bus versus trucking, it was the extrapolation of lower energy consumption for the trucks versus the transit buses, thus possibly resulting in even lower oil consumption. For long haul trucking it does not. It might for short-haul trucking, but here Tesla will be competing against BYD, which already has trucks available for this market. BYD's LiFePo batteries, with their lower-cost and higher-cycle life, are pretty much ideal for short-haul and drayage duty. The higher weight of the LiFePo batteries only means higher kwh/mi costs, which @ pennies per mile is easily offset by capital costs (lower purchase price?).

I see Tesla doing better, both financially and from a mission standpoint, to target medium & long-haul trucking (pretty sure battery swaps will come into play here).

 

[jhm]

Does this convince you?

Thanks for trying, but not really convinced. The essence of my argument is how many kWh is necessary to replace a gallon of diesel. So that does not depend on drag or roll resistance, but simply the amount of either kind of energy used to overcome that resistance. Non-plug-in hybrids are not a good comparison because all of the energy is ultimately derived from fuel, so that is really about the drivetrain efficiency, not engine efficiency. In terms of relative efficiency BEVs use about 24% as much primary energy (3414 BTU per kWh discharged) as ICE (129000 BTU per gallon diesel). Some of this relative efficiency is due to regen and transmission to wheels, but most of it is simple engine efficiency. The output of an ICE is only 25% to 30% efficient before its power is applied to the transmission system which applies its own inefficiencies.

So for me, the key energy issue is kWh to gallons. So at 24%, 9.07 kWh/DGE = 24% * 129000 BTU/DGE / 3414 BTU/kWh. Even if the relative efficiency were 27%, this would get us to 10.20 kWh/DGE, which only increase the battery requirements by about 12%.

So the basic question is what relative efficiency can we expect from an electric semi.