Shorting Oil, Hedging Tesla

[jhm]

High-Cost Oil Faces Existential Risk | OilPrice.com
https://www.carbontracker.org/wp-content/uploads/2019/04/The-Decline-Rate-Delusion.pdf

This report from Carbon Tracker is a very important contribution to making sense of peak oil demand. In fact, the report does not even contemplate what happens as demand declines. Rather, the authors are focused on the amount of capex required to keep oil production flat. They make a distinction between the amount capex required to sustain a constant production level from the level needed to grow production. Reverse-engineering some Rystad scenarios, they estimate that the cost to growth supply 1%/y out through 2030 is about $781B more than sustaining the current level. This incremental $781B nets an incremental 27B barrels. So the marginal growth barrel costs about $29/b. But to sustain the supply requires a mere $15/b.

The very important observation that they make is that there are many low capex ways to sustain current production. The effective decline rate may be as little as 2 to 3 percent per year. The supply curve increases rapidly as you transition from non-growth to growth. The price of oil is determined by the marginal supply. So the implication here is that as the market approaches zero growth, the price of marginal oil can fall rapidly.

For example, in the 1% growth case the marginal barrel has a cost of about $29/b and gets priced above $55/b. What happens when the marginal barrel is more like $15/b? I'm thinking that prices below $30/b could be in the cards. This is my conjecture. The authors don't spell this out, but they probably don't have to. They also don't speculate about how close we need to get to peak oil to see these effects take place. As we've discussed already, much of this depends on whether OPEC will continue to cut production to keep the marginal barrel out where it still high. If OPEC is willing to keep conceding market share, they can perpetuate higher prices well past peak demand. The risk however becomes even greater. More capital is put at risk on high capex projects. Demand loss to EVs, renewables and efficiency is accelerated. And a huge glut can cast a very long shadow. So the severity and duration of crisis in the oil market can be made much worse by OPEC continuing to boost prices. So in a least pain scenario, where OPEC allows the market to set prices, I suspect we could see oil prices decline several years before peak demand. Additionally, these lower prices could actually push the demand peak out a year or two. So in this scenario, we really do have a demand plateau. Ultimately, EV uptake will break a demand peak. So we are still talking about demand peak in 2025 or 2026. So in this scenario I could see oil at $30/b in 2025, but the price will have been declining for several years. OTOH, if OPEC props up prices, the peak could happen sooner, and oil prices in 2025 could be substantially below $30 as the market tries to escape a glut even as demand declines.

 

[TheTalkingMule]

In what world would a low pain scenario ever play out? That would require coordination and planning starting right now.

Frackers will continue to inflate the bubble with the help of the Saudis who for some reason think they're making the shrewd move by keeping prices high. The 2024-27 demand peak will become obvious within 18 months and the bubble will burst crushing the entire overleveraged US oil sector.

The recession we've been pushing off with low rates, massive US fossil extraction and tax cuts will hit like a Tesla Semi and oil demand will crater.

Perhaps human civilization lives through the disruption.

OPEC+R will keep limiting production once peak is on the horizon? No chance in hell.

 

[neroden]

High-Cost Oil Faces Existential Risk | OilPrice.com
https://www.carbontracker.org/wp-content/uploads/2019/04/The-Decline-Rate-Delusion.pdf

This report from Carbon Tracker is a very important contribution to making sense of peak oil demand. In fact, the report does not even contemplate what happens as demand declines. Rather, the authors are focused on the amount of capex required to keep oil production flat. They make a distinction between the amount capex required to sustain a constant production level from the level needed to grow production. Reverse-engineering some Rystad scenarios, they estimate that the cost to growth supply 1%/y out through 2030 is about $781B more than sustaining the current level. This incremental $781B nets an incremental 27B barrels. So the marginal growth barrel costs about $29/b. But to sustain the supply requires a mere $15/b.

The very important observation that they make is that there are many low capex ways to sustain current production. The effective decline rate may be as little as 2 to 3 percent per year. The supply curve increases rapidly as you transition from non-growth to growth. The price of oil is determined by the marginal supply. So the implication here is that as the market approaches zero growth, the price of marginal oil can fall rapidly.

For example, in the 1% growth case the marginal barrel has a cost of about $29/b and gets priced above $55/b. What happens when the marginal barrel is more like $15/b? I'm thinking that prices below $30/b could be in the cards. This is my conjecture.

If the supply is truly glutted, as we anticipate by 2023, the clearing price drops to the production cost. So it drops to $15/bbl.

This is low enough to induce some price-based reswitching back to gasoline, unfortunately. Should only apply to cars with 33 mpg or above, based on the Model 3's efficiency.

The real question is whether oil companies can survive these prices; with no profit margin on oil extraction or refining, the diseconomies of scale should kill them, IMO -- they shouldn't be able to cover their fixed costs -- but I haven't worked out the math for that.

(FWIW, over $35/bbl and gasoline is simply uncompetitive in fuelling cost. Over $23/bbl and it requires Prius-like hybrids to be price-competitive. Below that, reswitching should start happening. I am, of course, assuming purchase price parity since we're there in the top half of the market -- in reality, the bottom end of the market may still be buying gasoline cars until the gas car companies go under because they don't make enough money from only selling low-end cars)

The authors don't spell this out, but they probably don't have to. They also don't speculate about how close we need to get to peak oil to see these effects take place. As we've discussed already, much of this depends on whether OPEC will continue to cut production to keep the marginal barrel out where it still high. If OPEC is willing to keep conceding market share, they can perpetuate higher prices well past peak demand. The risk however becomes even greater. More capital is put at risk on high capex projects. Demand loss to EVs, renewables and efficiency is accelerated. And a huge glut can cast a very long shadow. So the severity and duration of crisis in the oil market can be made much worse by OPEC continuing to boost prices. So in a least pain scenario, where OPEC allows the market to set prices, I suspect we could see oil prices decline several years before peak demand. Additionally, these lower prices could actually push the demand peak out a year or two. So in this scenario, we really do have a demand plateau. Ultimately, EV uptake will break a demand peak. So we are still talking about demand peak in 2025 or 2026. So in this scenario I could see oil at $30/b in 2025, but the price will have been declining for several years. OTOH, if OPEC props up prices, the peak could happen sooner, and oil prices in 2025 could be substantially below $30 as the market tries to escape a glut even as demand declines.

 

[jhm]

In what world would a low pain scenario ever play out? That would require coordination and planning starting right now.

Frackers will continue to inflate the bubble with the help of the Saudis who for some reason think they're making the shrewd move by keeping prices high. The 2024-27 demand peak will become obvious within 18 months and the bubble will burst crushing the entire overleveraged US oil sector.

The recession we've been pushing off with low rates, massive US fossil extraction and tax cuts will hit like a Tesla Semi and oil demand will crater.

Perhaps human civilization lives through the disruption.

OPEC+R will keep limiting production once peak is on the horizon? No chance in hell.

So you're thinking that the peak will be obvious within 18 months from now? So like by end of 2020?

Tesla's best contribution to that would be robotaxis and self-driving semis. A semi can displace about 1.3 b/d of diesel while a private car can displace about 0.03 b/d of gasoline. However, a self driving semi could push toward 4 b/d and robotaxi 0.13 b/d. So if Tesla is cranking on both of these, the value proposition is higher than anything fossil powered vehicles can offer and the fuel displacement impact is huge. For example, without self-driving, suppose EV sales grow 50% in a year. But with self-driving they grow say 100% in year. If just one third of these vehicle participate in self driving fleet, that doubles the fuel displacement. So put together, we could see fuel displace quadruple in a single year. So imaging 2020 come in at 200 kb/d of fuel displace. Mostly this is before FCD. The next year we could see displacement double each year for several years as FSM takes off. So 2021 sees 400 kb/d, 2022 800 kb/d, and 1.6 mb/d by 2023. This is a fast enough fuel displacement scenario to force oil demand to peak in 2022. So yeah, I could see the oil industry squatting bricks by end of 2020. The auto industry would be stunned too. In this scenario, ICE sales would plummet. This is really the Tony Seba scenario.

I think the problem Tesla has with the autonomy vision is that is it has such radical implications for massive, rapid disruption, that it causes brain freeze for most people. Most of us just slip into some sort of denial. I certainly have my own resistance to this vision. Personally, I am very attached to the notion of private auto ownership. I don't really want to rent my car out. So in my own thinking about the future I have largely ignored how autonomy can accelerate the transition. I've wanted to focus on how quickly oil markets can be disrupted even if autonomy does not materialize. This has been my own indulgence in conservatism. But if FSD can be advanced enough that some jurisdictions allow robotaxis by 2020, the handwriting will be on the wall. The politics will get even uglier. The hatred of Tesla will be even more intense that it currently is. I certainly don't know if FSD will be ready for commercialization in 2020, but it is good moment to be self-aware and try to understand what we are personally attached to, what about this frightens us. The market as a whole is no where close to embracing this. So it will be a long time before Tesla gets a fair-minded valuation.

 

[jhm]

If the supply is truly glutted, as we anticipate by 2023, the clearing price drops to the production cost. So it drops to $15/bbl.

This is low enough to induce some price-based reswitching back to gasoline, unfortunately. Should only apply to cars with 33 mpg or above, based on the Model 3's efficiency.

The real question is whether oil companies can survive these prices; with no profit margin on oil extraction or refining, the diseconomies of scale should kill them, IMO -- they shouldn't be able to cover their fixed costs -- but I haven't worked out the math for that.

(FWIW, over $35/bbl and gasoline is simply uncompetitive in fuelling cost. Over $23/bbl and it requires Prius-like hybrids to be price-competitive. Below that, reswitching should start happening. I am, of course, assuming purchase price parity since we're there in the top half of the market -- in reality, the bottom end of the market may still be buying gasoline cars until the gas car companies go under because they don't make enough money from only selling low-end cars)

A potentially moderating factor here is physical storage and the futures curve. Futures will go into contango to reward investors for storing oil. Spot prices are kept low so that inventory is kept from flooding the market. We saw this happen in the last glut. But what is really different here is that EVs will be scaling to critical level reducing demand growth for years to come. So the back end of futures really should not rise very high, perhaps only as high as about $30/b. Thus, front end prices will need to be about $10/b below this back end price to hold inventory in storage. This gets us to maybe $20/b spot for as many years as it takes for 2% decline rates to work off the glut.

Then it gets much worse as EV penetration reaches about 75% causing the decline rate of oil demand to fall faster than 2% decline rate. The authors did not work through this. But if oil producers can economically reduce decline rate to a mere 2%, then there is a real chance that EV displacement of demand can exceed this. If the natural decline rate is 4% or more, it is very unlikely that EVs could push demand decline that far. But consider that about 25M/y EV is needed to suppress demand growth by about 1mb/d. So 100M/y EV could push demand growth into range of -3mb/d or roughly a 3% annual decline.

So the authors did not spell this out, but I think that just reflects their communication strategy. Their aim is to get oil investors to question the wisdom of pouring capital down the oil hole. They don't need to spell out everything at once. If fact they can lose credibility with their target audience if they go too far too quickly. It is enough at this point to contemplate what near zero growth would do to oil prices even without a glut to really knock up the damage. The industry psychologically is only able to contemplate a peak as early as 2030, but even this suggest exiting the decade at around $30/b and reducing capex spending by as much as 80%. So at the least, the E&P subindustry is absolutely gutted by 2030 under these gentle scenarios. The big players know this. This is why for instance Aramco is pursuing midstream and downstream opportunities, expanding refining and petrochem capacity. They need to lock in captive demand for their crude. So the authors are communicating just enough that it can shift the discussion from crude volume projections to capex spending and its implication for the oil price.

 

[neroden]

The natural decline rate of fracked wells is very, very high; that of traditional wells is quite low. I haven't yet seen a good analysis which disaggregates them.

What happens if all the fracked production disappears from the market essentially overnight (within one year)? What supply level do we drop to? How many EVs are necessary to rebalance the market at that supply level?

 

[TheTalkingMule]

And there goes Brent, very logically diving south on oversupply. Why is it that Brent pricing seems to lag these very obvious indicators by 24hrs+. It's been nearly 24hrs since EIA weekly supply/import/export figures came out. Why is the market just now reacting starting at 8am? Why would it be flat all day yesterday?

 

[ggies07]

This is why for instance Aramco is pursuing midstream and downstream opportunities, expanding refining and petrochem capacity.

I came in here to post about another company that could take business away from this last area they are trying to hold on to. Anyone heard of Full Cycle Bioplastics? I just learned about them....I know there are already items made from bioplastics but from what I understand why it hasn't scaled and taken off is cost. Well, Full Cycle seems to have nailed that:

compostable-bioplastics-for-the-circular-economy

• Full Cycle PHA is 1:1 price-competitive with traditional, synthetic oil-based plastics
• Full Cycle technology uses organic waste as raw material, which is readily available, low cost, with no arable land, expensive cultivation, or food crops
• Full Cycle technology uses non-GMO bacteria, suitable for co-location and non-sterile, industrial operating environments
• Full Cycle PHA is naturally occurring, compostable, marine-degradable, and food-contact safe
• Full Cycle technology takes heterogeneous waste inputs and produces consistent polymer outputs
• After use, organic PHA becomes raw material that can be upcycled again into virgin PHA

 

[jhm]

The natural decline rate of fracked wells is very, very high; that of traditional wells is quite low. I haven't yet seen a good analysis which disaggregates them.

What happens if all the fracked production disappears from the market essentially overnight (within one year)? What supply level do we drop to? How many EVs are necessary to rebalance the market at that supply level?

You might want to read the report from Carbon Tracker. There is a lot of good stuff in it. It does address this. There is a distinction to be made between the earlier decline rates of individual wells and the decline rates of fields or basins. US shale is still at the growth edge and this incurs much higher costs than simply maintaining them. For example, local infrastructure or the lack of it is a major challenge, but that alleviates once the growth abates. The authors write, "US shale oil, notorious for individual sharp well decline rates, counteracts decline by avoiding the long lead-times of conventional oil, and should be seen more as an industrial production process with costs falling along steep technology learning curves." So this industrial-like process helps to hold the costs down.

On a more technical note (https://uu.diva-portal.org/smash/get/diva2:762320/FULLTEXT01.pdf), Eagle Ford has an average decline curve which is nearly a hyperbolic decline curve with lambda = 0.28 and beta = 1 where time is measured in months. So the instantaneous monthly decline rate is

q'/q = -lambda/(1+ lambda*beta*t)

So the rate of decline slows with time. So at first the replacement burden is high, but it eases up over time.

 

[neroden]

It doesn't really. The decline rate in fracked basins is masked by adding excessive numbers of extra wells and using even more front loading of extraction on the new wells; this means it costs more to maintain production levels essentially indefinitely. When it drops back to the point where it does not require ever increasing money injection to maintain production, production is super low and the wells no longer cover their fixed costs. They are then shuttered or abandoned. Permian looks better because it is 50% conventional oil.

 

[AwlBidnz]

It doesn't really. The decline rate in fracked basins is masked by adding excessive numbers of extra wells and using even more front loading of extraction on the new wells; this means it costs more to maintain production levels essentially indefinitely. When it drops back to the point where it does not require ever increasing money injection to maintain production, production is super low and the wells no longer cover their fixed costs. They are then shuttered or abandoned. Permian looks better because it is 50% conventional oil.

The only thing I would add is that “indefinitely” isn’t achievable as subsequent “child wells” are not as strong producers as the parent well. The new wells are smaller as more “children” are drilled.

 

[TheTalkingMule]

What does this Occidental deal say about the health of the company/industry? An 8% loan when things are likely at their best seems absurd.

Why not raise funds like Tesla just did?

Perhaps traditional funding for fracking is drying up?

 

[jhm]

It doesn't really. The decline rate in fracked basins is masked by adding excessive numbers of extra wells and using even more front loading of extraction on the new wells; this means it costs more to maintain production levels essentially indefinitely. When it drops back to the point where it does not require ever increasing money injection to maintain production, production is super low and the wells no longer cover their fixed costs. They are then shuttered or abandoned. Permian looks better because it is 50% conventional oil.

No one is saying that the life of an oil basin can be extended indefinitely. The point being made is merely about the relative costs of production. Growing production requires investment in new oil fields. At scale, developing new oil fields requires a lot more capex than merely continuing to squeeze the last barrels out of an existing oil field. It seems this should be fairly intuitive and accords with the idea that capital is very much front loaded.

If we take a step back and ask ourselves what price of oil is required to keep growing US tight oil at 1mb/d annual increase versus the price to keep production at current levels, it should be should be obvious that growth requires a higher price. Indeed whenever the price of oil looks to decline, shale producers are pretty quick to start pulling back on rig count, a proxy for capex spending levels. For sure, US frackers are some of the most price sensitive producers in the market.

So the authors of the Carbon Tracker piece are simply pointing out some very basic economic reality and doing a couple of calculations to size up the slope in the supply curve. The critical implication is that as demand growth erodes from about 1.2 mb/d per year down to 0.0 mb/d per year, this will have a material impact on equilibrium prices for oil. This is simply the price elasticity of supply. Prices traded will decline too. The question is how much downward pressure will stagnation of demand growth apply to oil prices. Carbon Tracker is right to point this out because oil company economists are obfuscating the oil price implications of peak oil demand. They tend to focus on how far off the peak oughtta be in their "view" rather than how low the price of oil could go as demand grinds to a halt. The industry has a vested interest in keeping the cost of capital low, which they do by emphasizing how many barrels will be needed and downplaying how little those "needed" barrels could be worth. An awful lot of the oil industry makes its money off of the capital flow into oil production rather than off actually selling oil. They are selling wells and field services to investors. The size of that market is measured in oil and gas capex, not barrels or cubic feet. So the strategy of Carbon Tracker is to wake up investors to how the O&G industry may be selling them lousy investments. Definitely the investor should care much more about the price of oil in changing demand scenarios than in the volume of oil "needed".

So I just want to keep this broader context clearly in mind. Within that context tight oil actually plays a very important role. Because the decline rate of these well is so high in the first year, they have the advantage of returning capital quickly. So they are capable of responding to shorter term demand without a whole lot of risk around long-term demand. This is really important for keeping the market balanced. So the relevance of US shale to the argument that Carbon Tracker makes is not that these wells have slow decline rates, but rather that they are the marginal producers, These are the most price sensitive producers in the market. OPEC concedes market share just to keep these guys busy setting a higher marginal price for oil. But OPEC needs toss them about 2 mb/d of demand growth (in very round numbers here) to keep the Brent around $70/b. So when global demand is growing 1 mb/d, OPEC+R needs to cut production about 1 mb/d to keep the shale producers busing "growing". But what the heck happens as global demand grows at 0mb/d? Will OPEC+R still be willing to cut their production 2 mb/d, just to keep frackers growing at 2 mb/d? Suppose OPEC+R production falls to about 40mb/d by the time oil demand peaks. They will be in a position of cutting production 5% in a year just to maintain fracker at "high growth" prices. The situation starts to look absurd when the oil with the lowest natural decline rates are cutting production at 5% so that rate so that the producers with the high natural decline rate can grow production at upwards of 20% per year. This is the sort of absurdity that can that undergird a true asset bubble for as long as Wiley E Coyote can be remain unaware that he has shot past the cliff's edge. So the question is whether OPEC+R will remain committed to cutting production well past their natural decline rate. I think that this point we are talking about destroying capital to preserve prices, and if properly accounted for this would be a form of asset impairment, "stranded assets." Simply allowing production to fall with the natural decline rate does not require asset impairment, but cutting further than that does. This becomes a much more transparent issue if Aramco IPOs on a credible stock exchange. How exactly do investors maintain their Wiley E Coyote stance in US shale while Aramco is busy impairing its production assets?

So I think the natural decline rates of OPEC+R are relevant marker. If you're in position of impairing your own balance sheet to prop up shale producers, maybe it could make more sense to buyout shale producers. You can use the natural decline rate of those shale assets to avoid unnatural impairments to your existing balance sheet. Either way you are forfeiting capital to shut down excess production. So it is just a question of which assets are less costly to forfeit. Now if you're OPEC+R, you could just balk at cutting production and allow shale producers to fall into financial distress. At that point you could pick up these distressed production assets for pennies on the dollar. Once you fill your balance sheets with enough of these, you can go back to cutting production as needed, but mostly the production you cut are these marginal shale assets. Your new portfolio has a lower natural decline rate. Could this be what oil majors are up to buying up shale assets? At any rate, there seems to be some logic in holding a diversified portfolio with a high enough natural decline rate that you are not forced into impairing assets as demand stalls. Within such a portfolio rational choices could be made about where to allow production to decline.

 

[RobStark]

Look. How cute. A Lego Chemical Plant.

 

[TheTalkingMule]

The question is how much downward pressure will stagnation of demand growth apply to oil prices.

We also seem to be underestimating the urgency of futures pricing in a declining market. After 2022, most likely scenario is an all-out race to see who can pull the most liquid fuels out of the ground at a cost-basis of $9-35/barrel and sell for $18-40. Isn't that how it works when peaks become reality?

What does that do to countries and organizations who've been deluding themselves into a peak 2050 scenario(or similarly lying to investors)? It creates a perma-panic in what will already be a perma-glut and futures pricing heads toward something well below the avg cost of extraction.

Then, not only will peak demand have hit by 2024-27, but we'll be well into the new better renewables-based electric universe. Even this idea of oil's future price being capped around $25 to "compete with renewables" is too rosy IMO. They won't be nearly comparable products(units of energy) in 2025, they'll need to under-bid to remain relevant.

OPEC concedes market share just to keep these guys busy setting a higher marginal price for oil. But OPEC needs toss them about 2 mb/d of demand growth (in very round numbers here) to keep the Brent around $70/b.

I don't think this is the right way to look at Saudi logic. The last three years Saudi strategy has been all over the place, the primary goal being disruption of US shale. Dumping supply on the market didn't work, now cutting supply is clearly not working. They have no next step(because there isn't one) and are therefore just maintaining course on OPEC+R supply cuts due to lack of a better plan.

I don't think SA thought US shale would do what it's done over the last 18 months. They're not really dictating anything now that US shale can clearly soak up all incremental share. MBS is a deer in headlights, though a very wealthy deer.

At any rate, there seems to be some logic in holding a diversified portfolio with a high enough natural decline rate that you are not forced into impairing assets as demand stalls. Within such a portfolio rational choices could be made about where to allow production to decline.

Yes but all that requires coordination, all out collusion, or at the very least heavy cooperation across global markets. Once peak demand is obvious, there are far too many entities who's very existence relies on oil revenue to believe anything but all out chaos and prison rules will dominate oil markets. We're decentralizing, the exit from oil will be super ugly and in no way orderly or logical.

 

[neroden]

Given that there are no new shale oil fields in the world and everyone has gone all in on the Permian as the last shale field standing, I remain suspicious that the shale oil bubble will pop quite spectacularly and suddenly.

So my question is whether the essentially-instant disappearance of *all* shale oil production, even with OPEC+R opening up their production rates to maximum, might lead to a supply shock which is faster than the demand shock. I.e. demand is shrinking, but supply gets hit with a 10% drop (equal to US shale oil dropping off the face of the earth)... if demand is dropping slower than that, this might lead to a price spike.

Now, I didn't initially think that shale oil would collapse THAT fast, but given that they've been propping it up for over 10 years by burning money, and the scam is to front-load production more and more and more, the slightest push on the house of cards could literally leave them all without enough cash to drill a single new well; some probably wouldn't even be able to run the pumps. It could crash *extremely* fast.

It crashes when people stop burning money on it. I mean, most of the fracking has never made any money at all in the first place. Peak demand should help take the money out, but the bubble doesn't burst until the dumb money stops being thrown at it.

 

[neroden]

What does this Occidental deal say about the health of the company/industry? An 8% loan when things are likely at their best seems absurd.

Why not raise funds like Tesla just did?

Perhaps traditional funding for fracking is drying up?

I certainly hope so. The Berkshire/Occidental deal is a classic Buffett move following the Bagehot Rule -- "lend freely at a punitive rate". I would go so far as to say that if it works (most people are saying it won't and Chevron will get Anadarko) it probably wipes out the Occidential common stockholders (but Buffett gets his money).

 

[adiggs]

An idea I've been seeing articulated more commonly the last months or year or so, and don't remember very often before that, is the idea that barrels of oil aren't simply fungible. When refineries are built, they're optimized around a particular source and grade of crude. There is some degree of substitutability available on the inputs (crude) for the refinery, but at the extreme, a refinery optimized around heavy crude (as I understand the Gulf Coast refineries are optimized for) can't just start running exclusively fracked crude as feedstock and have an economic result.

The world oil market is a big beast, and substitution can happen many ways. One mechanism is to export US light sweet crude while importing other countries heavy crude (such as Canadian tar sands crude).

Anyway, the point of the moment - I'm seeing changes not only in total volume, but also in mix that makes up that total volume. That change in mix is, to me, another indicator of peak oil demand. If we're at (or near) peak demand, then the actual consumers of crude are in the driver's seat when it comes to picking and choosing the flavor of crude that works best for how their plant is setup. The flavor's of crude that are least in demand are going to see their prices go down first (which will have a side effect of pulling new refineries, and refinery reconfigurations, into the market to take advantage of especially cheap non-desirable / oversupplied crude).

There isn't really one single crude oil market. There's a crude oil market that's made up of many smaller crude oil markets with some degree of substitutability between those markets. That many-in-one dynamic is, I think, going to become increasingly important as well to understanding this beast.

 

[neroden]

An idea I've been seeing articulated more commonly the last months or year or so, and don't remember very often before that, is the idea that barrels of oil aren't simply fungible. When refineries are built, they're optimized around a particular source and grade of crude. There is some degree of substitutability available on the inputs (crude) for the refinery, but at the extreme, a refinery optimized around heavy crude (as I understand the Gulf Coast refineries are optimized for) can't just start running exclusively fracked crude as feedstock and have an economic result.

The world oil market is a big beast, and substitution can happen many ways. One mechanism is to export US light sweet crude while importing other countries heavy crude (such as Canadian tar sands crude).

Anyway, the point of the moment - I'm seeing changes not only in total volume, but also in mix that makes up that total volume. That change in mix is, to me, another indicator of peak oil demand. If we're at (or near) peak demand, then the actual consumers of crude are in the driver's seat when it comes to picking and choosing the flavor of crude that works best for how their plant is setup. The flavor's of crude that are least in demand are going to see their prices go down first (which will have a side effect of pulling new refineries, and refinery reconfigurations, into the market to take advantage of especially cheap non-desirable / oversupplied crude).

There isn't really one single crude oil market. There's a crude oil market that's made up of many smaller crude oil markets with some degree of substitutability between those markets. That many-in-one dynamic is, I think, going to become increasingly important as well to understanding this beast.

I've already said that refineries optimized to maximize jet fuel are in the best position. (Although electric airplanes are coming sooner than most people think, big jumbo jets will still be running for at least 10 years, so you have time to recover your costs from optimizing for jet fuel.) Most refineries are optimized for gasoline or diesel, which is so risky financially that it's *already* hurting ExxonMobil.

 

[jhm]

WARTS--Wild Ass RoboTaxi Scenario

Let's imagine just how bad robotaxis and robotrucks could be for oil demand. For the moment let's suspend disbelief about the needed technology just to put an outer bound on the potential impact. I will use some ridiculously rounded numbers just to keep the math simple. Here goes:

A fleet of about 2 billion mostly ICE vehicles currently consume about 60 mb/d of gasoline and diesel. That's 30 vehicles per 1 b/d demand.

A robotaxi can provide the same service levels of 5 auto. So the 2 billion fleet is eventually replaced with 400 million vehicle robofleet.

Eventually all new vehicles go into the robofleet. For those who still want a private vehicle, they cheaply lease a new Model 3 or better for two years after which it goes into the robofleet.

Robofleet vehicles have a 1 million mile range. 62,500 mile per year while in fleet service. So every 16 years a robofleet vehicle needs to be replaced.

At steady state (ignoring growth in demand for vehicle miles), about 25 million robofleet vehicles are replaced.

But this 25 million robofleet market displaces need for 125 million ICE vehicles. Thus, while robofleet vehicles are replacing ICE vehicles, the impact is to displace about 4 mb/d of transport fuel.

Robofleet vehicle production in 2020 is between 500k and 1M. Demand is strong enough that it doubles every year. So by 2025 over 25 million robofleet vehicles are produced in a single year, suppressing fuel demand by more than 4 mb/d in that single year.

Cumulatively by 2025 the robofleet is about 60 million strong, cumulatively displacing oil demand by about 10 mb/d. Total oil demand fall from about 100 mb/d in 2019 to 90 mb/d in 2025. It continues to fall to about 70 mb/d by 2030.

Also past 2025 the market for new ICE vehicles falls to less than 1 million vehicles per year. There is no point in owning a new ICE because leases on new Model 3 or better are much cheap and have no risk of losing value. Also using robotaxis is substantially cheaper than operating a used ICE vehicles. So residual values on ICE are painfully low.

So there it is, WARTS and all. Sadly reducing oil demand by 30% by 2030 does not quite provide the 50% carbon emissions reduction needed for the 1.5C scenario, but, damn, it comes close.

This sort of transition would be so damn fast, I don't see how ICE makers can survive. Most EV makers won't survive either, only the ones with full autonomy and a million mile drivetrain. I also see massive losses for the oil industry. Particularly refining is totally thrown off kilter because the product mix changes so quickly.

There is almost no profitability for making either diesel or gasoline, so jet and petrochems become expensive as they carry the full burden for profitability.