Shorting Oil, Hedging Tesla

[jhm]

According to OICA production statistics, there are about 3.8M and 320k heavy trucks and busses produced each year. In the US heavy trucks consume about 11.4k gallons of diesel per vehicle each year, and buses 10k gal/year. This is 0.75 b/d and 0.65 b/d of diesel demand per vehicle.

For simplicity let's say the combined heavy segment is 4M with 0.75 b/d-veh. The aggregate diesel demand is about 3mb/d for a year of new vehicles. This segment is very important to developing countries like India where 3X as much diesel is consumed than gasoline. This is an important growth engine for the oil industry.

How long would it take for electric vehicles to claim a 10% market share in the heavy vehicle market? A 10% share would displace 300 kb/d of diesel demand in a single year. Note that with the span of 5 years EV buses in China went from 0.2% market share to 20%. So for the heavy segment we may see 10% penetration within 5 years, by 2022. Certainly Tesla Semi will move aggressively, but I expect Chinese EV bus makers to want in on heavy trucks as well.

My working scenario has mostly been about 25M EV cars displacing 1 mb/d demand in 2025. But this is about 444kb/d in 2023 and ignores the heavy segment. Suppose heavy is doubling annually and displaces 300kb/d in 2022 and 600kb/d in 2023. Combining displacements from both segments could see 1mb/d by 2023. And I'm still ignoring light commercial vehicles and about 5mb/d that are used for power generation which solar and batteries can displace.

I suspect that commercial vehicles will be the surprise attack that the oil industry did not see coming. Indeed, by the time EV cars are displacing 1mb/d, commercial EVs will be displacing more than 1mb/d.

I think I'm going to need to rework my working scenario.

 

[jhm]

OK, that brings us to 2019. It would be good to attempt to rebuild my demand-destruction model to work out when demand drops faster than supply permanently. The naive version of it says 2028, but I think (a) I've underestimated the persistence of unprofitable oil drilling, (b) I make the very bad assumption that the "average" car gets replaced, when in fact the highest-usage vehicles get replaced first, (c) I haven't included any demand-reducing effects other than electric cars, and there are a lot of them (industrial fuel-switching, peak car, CAFE improvements, etc.) Of these effects I think (b) is the most significant.

JHM, do you think we can come up with some sort of guesswork model for fuel displacement if we assume that that heaviest-usage vehicles get displaced first? That is, the tractor-trailer market gets wiped out ASAP when Tesla's semi comes out, the taxi market shifts before the "drive 2 hours per day" market, etc. I think this would give a much better oil displacement model.

Your model of each battery displacing a certain amount of oil over its lifetime is great, but it doesn't give us the *timing* -- the battery in my low-miles-per-year car will displace a little bit of oil each year and will keep doing so 20 years from now, but the battery in a semi will displace a huge amount of oil in year one.

The Tesla semi will be revealed in September, and apparently has orders already lined up. They'll have to build a factory, which will probably take two years.

There are apparently less than 2 million semi tractors in the US total. (I don't know about the world, but that means probably less than 20 million in the world.) The average lifespan is 5 to 6 years. This fleet could be replaced very quickly. How much oil would this displace?

Funny you should post this as I wrote my last post. Yes, we need to work through all this. I think peak demand comes by 2023. Lots of issues accelerate this. Decline in ICE demand could come years earlier.

 

[kenliles]

Electric truck.

Rubbish....
First all-electric garbage truck in California

 

[erthquake]

Rubbish....
First all-electric garbage truck in California

And right on the heels of that news, Ford wins contract to make 2,500 all-electric vans for Deutsche Post.

The question I have with all these truck and passenger car projections (which are reasonable) being made here, do you all foresee enough battery availability?

 

[neroden]

I still think we need to displace roughly 5.8 mb/d to put oil into the permanent supply glut. (Take 95 mb/day current production, multiply by 6% decline rate.)

At 0.75 b/d displacement per truck, a production of 7.7 million trucks is sufficient. Suppose Tesla builds a 500K truck/year factory and opens it in 2020, and suppose the competitors in total produce twice as many (total of 1.5 million electric trucks in 2020). That's 1.1 mb/d by 2020. Suppose we double every two years. 4 years and it's over: 2024. Of course this assumes that it's *just* trucks. Since we also have cars, adding the two together should give us something like 2023, as you say.

Of course, money will be fleeing the oil industry well before this. I think we actually might get a final, muted price spike for oil once the writing is on the wall, as capital flees the oil companies *before* the demand destruction exceeds the natural supply decline.

I just noticed one more point:

And I'm still ignoring light commercial vehicles and about 5mb/d that are used for power generation which solar and batteries can displace.

Geez, it's still that high? It's been uneconomical since the *1970s* to use oil for power generation. I expect those to be replaced very, VERY fast. They're the highest margin (least price sensitive) customers for batteries and solar, and production capacity for solar is already high enough to supply all of them. You can get about 568 kwh by burning a barrel of oil in an electric power plant, and you need about 1/3 of your power at night, so you need about 947 Gwh of batteries to replace all of this. OK, that's a lot. I wonder how fast that will be displaced? Probably as fast as the batteries can be manufatured.

First of all, even without using batteries, you can shut those darn oil power plants down during the day and only run them at night, so roughly 3.3 mb/d can be eliminated just with solar panels, which will be built as fast as possible.

Supplying just the power (not considering the battery storage requirements) would require 568 Gw of solar panels, more or less (less if these oil-powered locations are more tropical); world production capacity is roughly 100 Gw and doubling every two years, so it's conservative to say that this can be done in 5 years. Remember these oil-powered locations will pay the highest prices for solar.

The remainder of needed displacement is only 2.5 mb/d. I think 0.75 to 1.1 mb/d from truck displacement is quite realistic by 2020, with 1.5 to 2.2 mb/d by 2022. Stationary batteries displacing generators should be enough to get us the rest of the way, along with, of course, the contribution from replacing cars. And I haven't even thought about industrial conversions or space heating (oil for space heating has also been cost-ineffective since the 1970s).

2022 or 2023, then. Wow.

The current "short term" supply-driven oil glut is now projected to last at least into 2019. The permanent oil glut hits around 2023 if not earlier. That's not a very long time to have a final oil price spike. There might be one in 2020-2022. I'm currently guessing there won't be.

As for battery production capacity, we're talking 2022 here. That's five years. Seems short for the End of an Era, but it's pretty long for just building out some factory capacity. Tesla can spin up an awful lot of battery factories in five years. Perhaps there will be a cobalt supply crunch... if this happens I bet we'll see a bunch of vehicles and particularly stationary batteries using other chemistries for a couple of years until the new cobalt mines get built.

 

[jhm]

Actually, given the massive overcapacity of the oil & gas industry if it's only being used for petrochemicals & plastics production, I figure there will be very low pressure for that industry to get off of oil. We'll end up with a small number of oil and gas wells (the most profitable) supplying a small number of refineries specializing in chemical feedstocks... and since they will all have been built decades ago and mostly fully-paid-off, it'll be hard for renewable sources to compete for decades. These few remaining refineries will add units to crack the middle distillates into light gases for feedstock. Think of how much more ethylene you can get by cracking all that diesel and gasoline.

So refineries can crack this down to light hydrocarbons, but in addition to the relatively small volume needed, the light hydrocarbons compete directly with natural gas. So the price is natgas becomes the value of an increasing fraction of the crude barrel.

Historically, oil has been able to trade around 4 times the price of NG per unit of energy. And even at today's low WTI price it is 2.63 times gas. As crude gets heavily cracked, the ratio should fall closer to 1. At 1, today's price of gas would imply $17/b. So an awful lot of crude production would become unprofitable along the path to price parity with gas.

 

[jhm]

@neroden, remember that we are working with two very different criteria. I'm simply trying to find peak demand, the point after which consumption begins to decline. You're looking well beyond that peak to time when demand is falling so fast that all drilling must cease. In think we pretty much need to get to 100% EV penetration before your event happens. But zero demand growth, the peak, can happen with 25% penetration or much less.

Not sure if 5mb/d for electricity still holds. Mostly is problem in oil producing countries (Saudi Arabia, especially), islands and places that lack a modern grid. So solar alone would take a huge bite out of it.

 

[adiggs]

Notice how each time a revision is needed, it results from reasons why the previously aggressive and unrealistic model was not nearly aggressive enough, and the new revision results in peak demand being sooner than previously thought?

Every single time.

Any bets on whether system effects, or interactions between these different forces, takes another year or 2 off of 2023 to peak demand?

 

[jhm]

Here is an idea about how to account for high utilization uses going electric first.

Suppose a trucking fleet operator has 100 trucks. The daily routes can be broken out into three bucks: long, medium and short. The long routes require 1.5 b/d, medium 0.75 b/d, and short 0.5 b/d. Further if we assume 10% long, 60% medium, and 30% short, the fleet average is 0.75b/d.

Now suppose that a certain EV truck maker brings a product to market with a short range that is only suitable for short routes. This may provide a nice savings on fuel, it only address 30% of what this fleet operator needs. So even if this operator buys 30 of these, they offset just 15b/d for the fleet.

But Tesla comes out with trucks that are fully capable of long routes as well as shorter range options. So now the operator can buy 10 long haul Teslas to offset 15b/d plus another 20 medium haul Teslas to offset another 15b/d. So for the same number of EV trucks, 30b/d can be offset, twice as much as the competitor. Moreover, Tesla has products to address the fuel range of needs for this operator. So if the Teslas save the operator a substantial amount of total operating costs, then that client will be inclined to swap out the fleet quickly.

So I think this is Tesla's basic semi strategy, a no compromise full range truck. Fleets want to save the most money quickly. This coincides with giving the highest duty to Teslas first which minimizes fleet consumption of fuel.

Curiously, Tesla is avoiding the EV bus market. Perhaps this is because this is a much smaller market and there are already lots of competitors making serious advances. The strategy of many of these is to make due with a small battery, under 100kWh, by charging it multiple times per day. This is workable for transit buses with short routes. But Tesla is under no constraint to enter the market from the small battery side. Motorcoaches may be a worthy challenge for Tesla, but then we are still limited to a small market.

The heavy truck market seems wide open for Tesla, and they can make a big difference fast.

 

[kenliles]

Curiously, Tesla is avoiding the EV bus market.

Noticed that as well. I think Elon is still on the fence regarding how ride-share effects that market as well. It may effectively squeeze it out

 

[adiggs]

Noticed that as well. I think Elon is still on the fence regarding how ride-share effects that market as well. It may effectively squeeze it out

Maybe he's seeing good enough progress, with enough urgency, that he's thinking there isn't much value proposition for Tesla to provide. Sure Tesla COULD make it better, but can Tesla make it enough better to improve the rate of adoption meaningfully, when there are indicators the new build market is all EV in < 10 years (maybe < 5 years)?

Here's an article from Cleantechnica on China's progress:
China 100% Electric Bus Sales Grew To ~115,700 In 2016

I wonder where / when the Chinese bus makers will be looking outside of China for further growth? Already 20% market share in China in 2016, with big bus fleets already thinking and planning for their entire fleet to switch over to 100% electric.

 

[RobStark]

I wonder where / when the Chinese bus makers will be looking outside of China for further growth? Already 20% market share in China in 2016, with big bus fleets already thinking and planning for their entire fleet to switch over to 100% electric.

 

[jhm]

Maybe he's seeing good enough progress, with enough urgency, that he's thinking there isn't much value proposition for Tesla to provide. Sure Tesla COULD make it better, but can Tesla make it enough better to improve the rate of adoption meaningfully, when there are indicators the new build market is all EV in < 10 years (maybe < 5 years)?

Here's an article from Cleantechnica on China's progress:
China 100% Electric Bus Sales Grew To ~115,700 In 2016

I wonder where / when the Chinese bus makers will be looking outside of China for further growth? Already 20% market share in China in 2016, with big bus fleets already thinking and planning for their entire fleet to switch over to 100% electric.

They are expanding to the global bus market. But I thoroughly expect developments in other commercial vehicle segments. Trash collection trucks are a good opportunity if you've got to stick to short routes.

 

[jhm]

Thanks. It strikes me that this is fairly labor intensive manufacturing. It seems that the scale of semi trucks makes automation more attractive. Perhaps this is another reason why Tesla is choosing semis over buses.

 

[GoTslaGo]

Thanks. It strikes me that this is fairly labor intensive manufacturing. It seems that the scale of semi trucks makes automation more attractive. Perhaps this is another reason why Tesla is choosing semis over buses.

Also I suspect customer base. Semi's will be bought and used mostly by private companies who's sole motive is economics. Buses are often in the public sector, with all the procurement rules and limitations. So selling a Semi theoretically should be easier than a bus might be--obviously if the municipality is motivated to go green that would be less of an issue, but BYD is already moving into that market niche. Why fight another EV provider when you can tackle another ICE provider instead.

 

[RobStark]

Why fight another EV provider when you can tackle another ICE provider instead.

Proterra, headed by former Tesla Executive Ryan Popple, has 60% US market share for electric buses. I take it BYD has much of the balance. There might be some tiny outfit I have never heard of.

 

[dc_h]

Bloomberg article on energy transformation. Not as optimistic as some here, but decades ahead of IEA and other govt prognosticators.

Solar Power Will Kill Coal Faster Than You Think

 

[TheTalkingMule]

Just read that Qatar bonds jumped in yield to the same rate as Saudi's due 2026. Question.....why would you put any faith in the Saudi's being able to pay bonds 10 years from now? I'd rather bet on Burkina Faso.

 

[jhm]

Bloomberg article on energy transformation. Not as optimistic as some here, but decades ahead of IEA and other govt prognosticators.

Solar Power Will Kill Coal Faster Than You Think

Yes, they really under estimate batteries. Less than $240B in batteries over the next 23 years, are you kidding me? That's less than 42 GWh per year average. Tesla alone will produce more than that. Moreover, batteries are critical to extending the addressable market of wind and solar. So underestimating batteries underestimates renewables.

This chart is helpful.

But we need to think in total markets and not assume they got the segments right. So right now the total market for power generation is about $1.6T. The fossil segment at $0.25T is mostly about flexibility to balance load and outmoded ideas of baseload. Batteries are the key technology for cracking into that segment. Batteries can do the flexible load balancing while spot prices when solar and wind production is high. Batteries can also cut into nuclear and hydro. Solar+storage is already cheaper baseload than new nuclear, provides flexibility and can be deployed in a fraction of the time.

So batteries can address the full $1.6T, but specifically attack the $0.25T fossil market. Just 10% of the fossil segment would be $25B per year more than the $10B or so that BNEF is already allocating to the renewable segment.

Think about pairing 1MW solar with 4MWh batteries. Per W, this is $1 solar and $1 battery. So splitting up that fossil segment about $125B could go to solar (or wind) and $125B to batteries. So I would view this as a minimal need in a decarbonizing world. But applying this split to the total market, there is an upside of $800B that batteries could seize.

When Tesla gets this dialed, it will be huge.

 

[adiggs]

Another recent article on oilprice continuing our recent theme of the utility of hydrocarbons to the modern economy:
The Fourth Industrial Revolution Is Fueled By Oil | OilPrice.com

There are two points being made by this article, both of which I grant and agree with:
1- today's economy uses hydrocarbons as it's energy supply (renewables are too small of a fraction for us to go cold turkey next week)
2- there are important segments of the economy that can only be supplied today by hydrocarbons (the 2 examples used are plastics and high temperature industrial heat).

The unstated implication is that these are good reasons to believe that the O&G industry we know today will exist long into the future.

At least to me, these arguments sound more like the arguments I would expect to start hearing as the death throes of an industry begins to arrive.

For #1, we've seen from the coal industry how rapidly the value of the industry can decrease while the volume / units of the industry slowly decays. So yes, we'll be burning hydrocarbons for decades to come, and in large quantities. Probably 50+ years, simply due to the scale and the way that lowering cost of energy will both bring in new consumers, new consumption, and increased consumption of energy at the same time (more energy consumption will offset some of the replacement of fossil fuels by renewables, and the current scale of hydrocarbon consumption is stupendous).

For #2, this also is true. And who knows - maybe in the face of all historical evidence, these segments that can't be directly replaced by renewables will never be replaced. They amount to such a narrow sliver of the units of today's industry that if these are the "growth" or even maintenance uses / markets, then O&G is already hosed.


For an in-depth treatment of the scale of the industry, and a different point of view on how fast O&G will disappear from the world, I commend this 90 minute Youtube video:

There are plenty of points the speaker makes that I disagree with, but there are plenty more that are true and was worthwhile for me for getting another perspective on the scale of the industry. There are in particular several places where it seems to me that he's making an apple to oranges comparison. That doesn't change the central point - the worldwide energy system is almost all hydrocarbons, and the solar/wind that we're so excited about is more like 1 or 2%. History suggests it takes a new energy source decades to go from 5% of world energy to 25%, and we've still got years to go to start that clock for solar / wind.

(Then again, my counter to that is that technology changes today happen faster than they did 100 years ago -- LOTS faster).