-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Off Shore
Off Topic Member
Neighborhood Electric Vehicle. e.g.: (the one on the left)
Attachments
The end of natural gas is near
I've often thought that gas would only bridge to batteries. Apparently the major makers of gas turbines were rapidly scaling back production last year. We may be near the gas peaker peak. If production goes low enough then there wont be enough new capacity of gas generators to cover retired capacity. The challenge is that big batteries can now provide peak power at less cost, while integrating renewables and providing a deep stack of services. The real kicker is that while a 100MW gas plant takes about 6 years to develop, Tesla can deploy 100MW of battery power in about 100 days and wind or solar can be included within a year.
Here's an interesting chart from the Tesla battery in South Australia.
Tesla big battery moves from show-boating to money-making
This Tesla can go from 0 to 100MW in 140 milliseconds! Look at how delicate some of the trading is. These two days of rate arbitrage brought in about $1M. I doubt a gas peaker can be as nimble. With enough battery capacity on the scene, these peak prices could be pushed down even as the low early morning prices could be supported. The Hornsdale Power Reserve is supposed to be paired with a wind farm. I don't know if that farm was online at this time.
This rate arbitrage service is just using 30MW / 90MWh of capacity out of 100MW / 130MWh. That leaves the other 70MW /40MWh for other services in the stack.
I've often thought that gas would only bridge to batteries. Apparently the major makers of gas turbines were rapidly scaling back production last year. We may be near the gas peaker peak. If production goes low enough then there wont be enough new capacity of gas generators to cover retired capacity. The challenge is that big batteries can now provide peak power at less cost, while integrating renewables and providing a deep stack of services. The real kicker is that while a 100MW gas plant takes about 6 years to develop, Tesla can deploy 100MW of battery power in about 100 days and wind or solar can be included within a year.
Here's an interesting chart from the Tesla battery in South Australia.
Tesla big battery moves from show-boating to money-making
This Tesla can go from 0 to 100MW in 140 milliseconds! Look at how delicate some of the trading is. These two days of rate arbitrage brought in about $1M. I doubt a gas peaker can be as nimble. With enough battery capacity on the scene, these peak prices could be pushed down even as the low early morning prices could be supported. The Hornsdale Power Reserve is supposed to be paired with a wind farm. I don't know if that farm was online at this time.
This rate arbitrage service is just using 30MW / 90MWh of capacity out of 100MW / 130MWh. That leaves the other 70MW /40MWh for other services in the stack.
In china NEV stands for New Energy Vehicle. Ie normal vehicles using energy other than liquid fossil fuel. So its hydrogen and electric and plug in hybrid.
So a Tesla in China is a NEV.
What western define as Neighbourhood electric vehicle is named differently in China, thats a LSEV. Look up shadong LSEV to get an idea. Actually EV in china by default refers to LSEV. A full capacity in China is strictly called a NEV so not to confuse it with a 2-3 wheel 'EV'
So a Tesla in China is a NEV.
What western define as Neighbourhood electric vehicle is named differently in China, thats a LSEV. Look up shadong LSEV to get an idea. Actually EV in china by default refers to LSEV. A full capacity in China is strictly called a NEV so not to confuse it with a 2-3 wheel 'EV'
adiggs
Well-Known Member
The end of natural gas is near
I've often thought that gas would only bridge to batteries. Apparently the major makers of gas turbines were rapidly scaling back production last year. We may be near the gas peaker peak. If production goes low enough then there wont be enough new capacity of gas generators to cover retired capacity. The challenge is that big batteries can now provide peak power at less cost, while integrating renewables and providing a deep stack of services. The real kicker is that while a 100MW gas plant takes about 6 years to develop, Tesla can deploy 100MW of battery power in about 100 days and wind or solar can be included within a year.
Here's an interesting chart from the Tesla battery in South Australia.
Tesla big battery moves from show-boating to money-making
This Tesla can go from 0 to 100MW in 140 milliseconds! Look at how delicate some of the trading is. These two days of rate arbitrage brought in about $1M. I doubt a gas peaker can be as nimble. With enough battery capacity on the scene, these peak prices could be pushed down even as the low early morning prices could be supported. The Hornsdale Power Reserve is supposed to be paired with a wind farm. I don't know if that farm was online at this time.
This rate arbitrage service is just using 30MW / 90MWh of capacity out of 100MW / 130MWh. That leaves the other 70MW /40MWh for other services in the stack.
As I understand it, the wind farm was previously built and already in service. That's what enabled the "Tesla Big Battery" to be built so quickly - the wind farm already existed and had a grid interconnect already in place that the Hornsdale Power Reserve could be connected to (no grid wiring needed).
Thanks for the clarification. I wonder if the volatility in charging, then, is due to the gustiness of the wind farm. Presumably, the battery could be trimming off excess of wind production so that the net wind power delivered to grid is really smooth and reliable. This would explain the streakiness of charging.
The big blades have mass and inertia to smooth out the gustiness of wind, so the peakiness should just be its responsiveness to the needs of the grid.
adiggs
Well-Known Member
Another piece of the puzzle is that the Hornsdale Power Reserve, or at least that facility, is really more like 2 facilities in 1. I'll see if I can find an article that makes the specific sizes clear, but the idea is that part of the reserve is high power and low energy (~100MW and ~30 MWh) - that piece is contracted by the state government of SA and is used for grid services.
The other is low power (~30 MW) and high energy (~100 MWh), and is owned and operated by the wind farm owner (Neoen I think is the name), and is used to store excess wind production and smooth out peaks and valleys in the wind farm output to the grid.
That has to do with the stack of services it provided. They basically partition capacity (both power and energy) for different purposes. The state pays for the 100WM/30MWh portion for dealing with emergency power. The other 30MW/90MWh Neoen can use to trade power in the wholesale market to make a profit. I suspect that when emergency service requires it the 30MW portion for trading gets diverted to emergency use. So stacking is not exactly a fixed partition at all times.
So if you find a good article on the Hornsdale stack, let us know. I do think that stacking is one of the most important concepts for how batteries optimize value creation, and it is not entirely obvious how stacking works. Many arguments about whether batteries are cost effective run into the error of fixating on just one sort of service while ignoring the full stack potential. We've all seen those cost of storage calculations that make battery storage not look so attractive, but this ignores things like voltage regulation, frequency regulation, back up power, etc. which create multiple revenue streams that partly pay for the cost of storage, so that the net cost of storage may actually be quite affordable.
adiggs
Well-Known Member
Here's a descriptive article about the Hornsdale Power Reserve and its two components.
Yes, the Tesla battery is massive, but it can do much more beside
Turns out it's 70MW / ~12 MWh (10 minutes) for network stabilization / frequency control, and 30 MW / ~117 MWh (3-4 hours) for energy / time shifting of wind energy.
In practice, more recent articles are making the point that Neoen (owners and operators of the system) are sometimes using the battery to play in markets its not contracted for "just because they can". Such as reacting to a coal plant tripping (going offline unexpectedly) by beating the official reserve capacity into the grid to stabilize the grid, ahead of the official reserve getting in and taking over.
I think I saw an article talking about Neoen / Hornsdale Wind Farm starting to bid into the FCAS (frequency regulation) market - something wind / renewables don't typically do
Many articles on RenewEconomy about how the big battery is pushing boundaries and changing perceptions of what's possible and how to do things. Here's a reasonably good index into many recent articles (HPR gets an article quite frequently over the last month - LOTS of fascination on many people's parts of what's possible).
Search for "hornsdale power reserve" - 2/3 - RenewEconomy : RenewEconomy
For fun, here's an article about the response to the (coal plant) Loy Yang B trip and HPR response:
Tesla big battery outsmarts lumbering coal units after Loy Yang trips
The part I find most interesting - HPR wasn't part of the reserve being paid to provide that service, and the response was to an event that happened 1,000km away.
They also make the point that for all the data we have, HPR is reacting in milliseconds, in a market where the data is measured at a 4s level of detail. The market really doesn't understand technology with millisecond reaction time, and is still figuring out what it might mean.
SebastianR
Active Member
I think that is entirely possible. But more because of a weak USD rather than because we have an actual (world) price movement. In fact, the price of WTI is pretty much unchanged (+1.3%) over the year of 2017 from a EUR perspective.
If you compare WTI to Sugar and take Euro as base currency you get that both commodities - while not being totally "static" have a pretty high correlation in their price movements and are - more importantly - within +/- 4% of the prices a year ago. So I don't think this is oil specific.
Crude Oil (petroleum); West Texas Intermediate vs Sugar - Price Rate of Change Comparison - IndexMundi
Unless I missed something fundamental here. Problem will be that much of the oil accounting is going on in USD so I think there will be a problem at some point in time if the USD tanks too much...
EDIT: I need to think about this more. I don't think what I wrote above is entirely accurate. Yes, the forecast for the USD is not brilliant, but I don't think we have had a horrible 2017 so I don't think what I wrote is the complete story. What would be a good way to compare crude prices against a basket of currencies?
Last edited:
ValueAnalyst
Closed
SebastianR
Active Member
I'm not sure I follow? The higher the oil price, the lower the USD?
ValueAnalyst
Closed
BNEF Talk: The Future of Electric Vehicles
I highly recommend watching this entire 16 minute video.
Collins finally gives attention to commercial EVs. He notes that in 2016 16GWh of batteries went into Chinese EV buses while 20.5GWh all global EV autos. By my own estimate I figure that the buses may be displacing about 3 times as much fuel per day as the autos. So tracking EVs in GWh is very important. GWh is what links more closely with demand on minerals and impact on oil markets and emissions.
I highly recommend watching this entire 16 minute video.
Collins finally gives attention to commercial EVs. He notes that in 2016 16GWh of batteries went into Chinese EV buses while 20.5GWh all global EV autos. By my own estimate I figure that the buses may be displacing about 3 times as much fuel per day as the autos. So tracking EVs in GWh is very important. GWh is what links more closely with demand on minerals and impact on oil markets and emissions.
Ev-volumes.com has updated their forecast for world EV auto sales. 2017 came in at 1,218,000, up 57% over last year and reaching 1.28% market penetration. They forecast 1.9M, 1.95% market share for 2018.
I've also been playing around with building my own logistic growth models. Here is my extended forecast based on the 2018 forecast and history from ev-volumes.com.
Here I am assuming that total autos continue to grow 3.60%/y and that the logit growth continues at 0.410/y. The logit is simply the log ratio of EV share to ICE share, ln(EV/ICE), and it is remarkably stable for the global market.
Some observations about this forecast...
New ICE sales are to peak at 102M in 2022. Leading up to this ICE sales will find it increasingly hard to grow. This could become a crisis for automakers that are not prepared to compete in the EV market. After the peak, the decline picks up speed each year. Just 3 years later in 2025, the decline could be 5.4%. Massive layoffs for the unprepared could be unavoidable with such a rate of decline, and it only gets worse in following years.
Peak New ICE coincides with EV penetration reaching just 9.3% in 2022. By 2025 penetration is 26%, 54% by 2028, and 73% by 2030. The logistic curve is maybe to conservative about how long it takes to transition as you approach high penetration, other models transition much faster to 99%.
In any case, most of the exciting changes will happen by 2030, and ICE sales will be clearly insufficient to replace old ICE falling out of service. That is the active ICE fleet will peak sometime from 2026 to 2030. This is not modeled here, but it is simply a question when new ICE sales are insufficient to replace old ICE.
I think that peak new ICE will be an important marker, especially sense massive auto worker layoffs can be expected to happen with a few years of the peak. Yes, many auto workers will be able to find jobs building EVs, but the complexity is lower. Moreover there are many auto parts that are needed for ICE, specifically engines, transmissions, exhaust systems, fuel tanks. So many auto part makers will be faced with rapidly declining demand. It's hard to see how massive disruption to labor can be avoided.
Also as ICE makers face down this peak, it will be very difficult for the oil market to deny the imminent threat to oil demand. Traditional automakers will be desperate to build EVs just to be able to post some modest growth. So the oil markets will have to change a lot of their assumptions about automaker eagerness to make EVs. Certainly growth in demand for oil will be hard to come by as ICE sales peak. Really, the ultimate demand for oil depends on how long it takes to reach peak new ICE. The sooner it happens, the less oil will be extracted ever.
I've also been playing around with building my own logistic growth models. Here is my extended forecast based on the 2018 forecast and history from ev-volumes.com.
Code:
World
3.60% 0.410
Year EV ICE Total EV Growth ICE Growth Total Growth EV Share Logit Share Change Logit
2012 134 78,690 78,824 0.17% -6.38
2013 211 84,189 84,400 57.5% 7.0% 7.1% 0.25% -5.99 0.39
2014 325 85,201 85,526 54.0% 1.2% 1.3% 0.38% -5.57 0.42
2015 546 87,519 88,065 68.0% 2.7% 3.0% 0.62% -5.08 0.49
2016 774 89,226 90,000 41.8% 2.0% 2.2% 0.86% -4.75 0.33
2017 1,218 93,938 95,156 57.4% 5.3% 5.7% 1.28% -4.35 0.40
2018 1,900 95,536 97,436 56.0% 1.7% 3.6% 1.95% -3.92 0.43
2019 2,936 98,004 100,940 54.5% 2.6% 3.6% 2.9% -3.51 0.41
2020 4,515 100,055 104,570 53.8% 2.1% 3.6% 4.3% -3.10 0.41
2021 6,894 101,436 108,330 52.7% 1.4% 3.6% 6.4% -2.69 0.41
2022 10,422 101,804 112,226 51.2% 0.4% 3.6% 9.3% -2.28 0.41
2023 15,533 100,729 116,262 49.0% -1.1% 3.6% 13.4% -1.87 0.41
2024 22,702 97,740 120,442 46.2% -3.0% 3.6% 18.8% -1.46 0.41
2025 32,339 92,435 124,774 42.4% -5.4% 3.6% 25.9% -1.05 0.41
2026 44,609 84,651 129,261 37.9% -8.4% 3.6% 34.5% -0.64 0.41
2027 59,256 74,653 133,909 32.8% -11.8% 3.6% 44.3% -0.23 0.41
2028 75,542 63,183 138,725 27.5% -15.4% 3.6% 54.5% 0.18 0.41
2029 92,404 51,310 143,713 22.3% -18.8% 3.6% 64.3% 0.59 0.41
2030 108,780 40,101 148,881 17.7% -21.8% 3.6% 73.1% 1.00 0.41
2031 123,909 30,326 154,235 13.9% -24.4% 3.6% 80.3% 1.41 0.41
2032 137,449 22,333 159,782 10.9% -26.4% 3.6% 86.0% 1.82 0.41
2033 149,410 16,117 165,528 8.7% -27.8% 3.6% 90.3% 2.23 0.41
2034 160,020 11,460 171,480 7.1% -28.9% 3.6% 93.3% 2.64 0.41
2035 169,584 8,063 177,647 6.0% -29.6% 3.6% 95.5% 3.05 0.41
2036 178,404 5,631 184,035 5.2% -30.2% 3.6% 96.9% 3.46 0.41
2037 186,740 3,913 190,653 4.7% -30.5% 3.6% 97.9% 3.87 0.41
2038 194,799 2,710 197,509 4.3% -30.7% 3.6% 98.6% 4.27 0.41
2039 202,739 1,873 204,612 4.1% -30.9% 3.6% 99.1% 4.68 0.41
2040 210,678 1,292 211,970 3.9% -31.0% 3.6% 99.4% 5.09 0.41Here I am assuming that total autos continue to grow 3.60%/y and that the logit growth continues at 0.410/y. The logit is simply the log ratio of EV share to ICE share, ln(EV/ICE), and it is remarkably stable for the global market.
Some observations about this forecast...
New ICE sales are to peak at 102M in 2022. Leading up to this ICE sales will find it increasingly hard to grow. This could become a crisis for automakers that are not prepared to compete in the EV market. After the peak, the decline picks up speed each year. Just 3 years later in 2025, the decline could be 5.4%. Massive layoffs for the unprepared could be unavoidable with such a rate of decline, and it only gets worse in following years.
Peak New ICE coincides with EV penetration reaching just 9.3% in 2022. By 2025 penetration is 26%, 54% by 2028, and 73% by 2030. The logistic curve is maybe to conservative about how long it takes to transition as you approach high penetration, other models transition much faster to 99%.
In any case, most of the exciting changes will happen by 2030, and ICE sales will be clearly insufficient to replace old ICE falling out of service. That is the active ICE fleet will peak sometime from 2026 to 2030. This is not modeled here, but it is simply a question when new ICE sales are insufficient to replace old ICE.
I think that peak new ICE will be an important marker, especially sense massive auto worker layoffs can be expected to happen with a few years of the peak. Yes, many auto workers will be able to find jobs building EVs, but the complexity is lower. Moreover there are many auto parts that are needed for ICE, specifically engines, transmissions, exhaust systems, fuel tanks. So many auto part makers will be faced with rapidly declining demand. It's hard to see how massive disruption to labor can be avoided.
Also as ICE makers face down this peak, it will be very difficult for the oil market to deny the imminent threat to oil demand. Traditional automakers will be desperate to build EVs just to be able to post some modest growth. So the oil markets will have to change a lot of their assumptions about automaker eagerness to make EVs. Certainly growth in demand for oil will be hard to come by as ICE sales peak. Really, the ultimate demand for oil depends on how long it takes to reach peak new ICE. The sooner it happens, the less oil will be extracted ever.
So here is the forecast for USA from ev-volumes.com.
It is a bit surprising to see EVs double in 2018, but this pretty much comes down to how well the Model 3 ramp goes. Additionally, their forecast forecast total auto sales in 2018 implies a 1.7% decline over prior year. The combination of 100% growth in EVs and 1.7% decline in total cars implies that ICE sales decline of 2.9%. Short of volatility exerting a new high for US ICE sales the peak of 17,309,000 in 2015 could stand for all time. Essentially, the oil price decline in 2015 may have helped to pull ICE sales ahead after the fleet had aged quite a bit from the 2009 recession. Digesting this pull ahead likely lead to the decline in auto sales last year and could continue as headwind for 2018.
So for my extended forecast, I'll assume a 2.52%/y growth rate for total auto as this is the average growth from 2012 to forecasted 2018. This assumption has strong implications for growth in ICE. Auto sales do tend to be cyclical, but this may be more of a problem for ICE than EVs, since the later is in a very strong growth phase. Even so, the cyclical behavior of auto sales makes it hard to predict when ICE will ultimately peak.
Also I will base the change in logit on average from 2012 to forecast 2018. This is 0.309/y, which is substantially slower than the change rate for world at 0.410/y. The world rate is likely dominated by China (which I cannot find data for). Additionally, the US estimate is perhaps dampened by the decline in EV sales that were induced by oil prices, which was not so much an issue in China. So given that oil looks to be priced above $60 into the 2020s, I think the 0.309/y rate is likely too low. Even so assuming a slow rate like this would tend to understate just how quickly EVs will catch on in the US.
So here is my extended forecast...
So we see in this extended forecast that US ICE may encounter the EV-induced peak in 2022, the same peak year as the global ICE market. If change in logit is faster than I have assumed or total sales growth is slower, then this peak could arrive as early as 2020. Given the volatility of US auto sales we may never know.
Even so, EVs reach 7.7% penetration at this theoretical ICE peak in 2022, which climbs to 17.5% in 2025 and 50% by 2030. These market share forecasts are quite sensitive to the low assumption of logit change. If the global rate of 0.41/y is applied in the US, then we are looking at 11%, 30% and 77% in years 2022, 2025 and 2030, respectively. I believe this difference will be largely consumer demand driven as opposed to policy driven or supply constrained. In the US Tesla will be the dominant player. So I believe that US consumers will really want to own and drive these electrics.
So by 2025 US ICE is declining 2.3%/y and 11%/y by 2030. This does not look good for US jobs linked to production of ICE drivetrains. There is a distinct political risk that policy could turn protectionistic toward ICE and attempt to slow down EV adoption. Even so I suspect that consumers will really want EVs and balance out political shenanigans. It may be really good politics for Tesla to build a Gigafactory in or near Detroit, just saying.
It is a bit surprising to see EVs double in 2018, but this pretty much comes down to how well the Model 3 ramp goes. Additionally, their forecast forecast total auto sales in 2018 implies a 1.7% decline over prior year. The combination of 100% growth in EVs and 1.7% decline in total cars implies that ICE sales decline of 2.9%. Short of volatility exerting a new high for US ICE sales the peak of 17,309,000 in 2015 could stand for all time. Essentially, the oil price decline in 2015 may have helped to pull ICE sales ahead after the fleet had aged quite a bit from the 2009 recession. Digesting this pull ahead likely lead to the decline in auto sales last year and could continue as headwind for 2018.
So for my extended forecast, I'll assume a 2.52%/y growth rate for total auto as this is the average growth from 2012 to forecasted 2018. This assumption has strong implications for growth in ICE. Auto sales do tend to be cyclical, but this may be more of a problem for ICE than EVs, since the later is in a very strong growth phase. Even so, the cyclical behavior of auto sales makes it hard to predict when ICE will ultimately peak.
Also I will base the change in logit on average from 2012 to forecast 2018. This is 0.309/y, which is substantially slower than the change rate for world at 0.410/y. The world rate is likely dominated by China (which I cannot find data for). Additionally, the US estimate is perhaps dampened by the decline in EV sales that were induced by oil prices, which was not so much an issue in China. So given that oil looks to be priced above $60 into the 2020s, I think the 0.309/y rate is likely too low. Even so assuming a slow rate like this would tend to understate just how quickly EVs will catch on in the US.
So here is my extended forecast...
Code:
USA
2.52% 0.309
Year EV ICE Total EV Growth ICE Growth Total Growth EV Share Logit Share Change Logit
2012 55 14,419 14,474 0.4% -5.57
2013 97 15,548 15,645 76.4% 7.8% 8.1% 0.6% -5.08 0.49
2014 120 16,318 16,438 23.7% 5.0% 5.1% 0.7% -4.91 0.16
2015 115 17,309 17,424 -4.2% 6.1% 6.0% 0.7% -5.01 -0.10
2016 157 17,287 17,444 36.5% -0.1% 0.1% 0.9% -4.70 0.31
2017 200 16,894 17,094 27.4% -2.3% -2.0% 1.2% -4.44 0.27
2018 400 16,407 16,807 100.0% -2.9% -1.7% 2.4% -3.71 0.72
2019 554 16,677 17,231 38.5% 1.6% 2.5% 3.2% -3.40 0.31
2020 765 16,900 17,665 38.1% 1.3% 2.5% 4.3% -3.10 0.31
2021 1,051 17,059 18,111 37.5% 0.9% 2.5% 5.8% -2.79 0.31
2022 1,438 17,129 18,567 36.8% 0.4% 2.5% 7.7% -2.48 0.31
2023 1,954 17,082 19,036 35.9% -0.3% 2.5% 10.3% -2.17 0.31
2024 2,631 16,885 19,516 34.7% -1.2% 2.5% 13.5% -1.86 0.31
2025 3,504 16,504 20,008 33.2% -2.3% 2.5% 17.5% -1.55 0.31
2026 4,601 15,911 20,513 31.3% -3.6% 2.5% 22.4% -1.24 0.31
2027 5,943 15,087 21,030 29.2% -5.2% 2.5% 28.3% -0.93 0.31
2028 7,530 14,031 21,560 26.7% -7.0% 2.5% 34.9% -0.62 0.31
2029 9,335 12,769 22,104 24.0% -9.0% 2.5% 42.2% -0.31 0.31
2030 11,308 11,354 22,662 21.1% -11.1% 2.5% 49.9% 0.00 0.31
2031 13,375 9,858 23,233 18.3% -13.2% 2.5% 57.6% 0.31 0.31
2032 15,457 8,363 23,819 15.6% -15.2% 2.5% 64.9% 0.61 0.31
2033 17,478 6,942 24,420 13.1% -17.0% 2.5% 71.6% 0.92 0.31
2034 19,384 5,651 25,036 10.9% -18.6% 2.5% 77.4% 1.23 0.31
2035 21,143 4,525 25,667 9.1% -19.9% 2.5% 82.4% 1.54 0.31
2036 22,742 3,573 26,314 7.6% -21.0% 2.5% 86.4% 1.85 0.31
2037 24,189 2,789 26,978 6.4% -21.9% 2.5% 89.7% 2.16 0.31
2038 25,500 2,159 27,659 5.4% -22.6% 2.5% 92.2% 2.47 0.31
2039 26,697 1,659 28,356 4.7% -23.1% 2.5% 94.1% 2.78 0.31
2040 27,803 1,268 29,071 4.1% -23.6% 2.5% 95.6% 3.09 0.31So we see in this extended forecast that US ICE may encounter the EV-induced peak in 2022, the same peak year as the global ICE market. If change in logit is faster than I have assumed or total sales growth is slower, then this peak could arrive as early as 2020. Given the volatility of US auto sales we may never know.
Even so, EVs reach 7.7% penetration at this theoretical ICE peak in 2022, which climbs to 17.5% in 2025 and 50% by 2030. These market share forecasts are quite sensitive to the low assumption of logit change. If the global rate of 0.41/y is applied in the US, then we are looking at 11%, 30% and 77% in years 2022, 2025 and 2030, respectively. I believe this difference will be largely consumer demand driven as opposed to policy driven or supply constrained. In the US Tesla will be the dominant player. So I believe that US consumers will really want to own and drive these electrics.
So by 2025 US ICE is declining 2.3%/y and 11%/y by 2030. This does not look good for US jobs linked to production of ICE drivetrains. There is a distinct political risk that policy could turn protectionistic toward ICE and attempt to slow down EV adoption. Even so I suspect that consumers will really want EVs and balance out political shenanigans. It may be really good politics for Tesla to build a Gigafactory in or near Detroit, just saying.
BTW, one fun thing I can do with my extended model is run it out several more decades to see how many cars will ultimately be bought. I figure that after 2020 about 917M more ICE autos will ever be bought. This, of course depends on how fast the logit grows. So the error on this is around +/-200M. The lifetime fuel consumption of an efficient auto is about 100 barrels of gasoline. The ultimate fuel consumption for all auto built after 2020 is likely less than 100 billion barrels.
For comparison, in the 50 years following 2020, my model expect 13.7 trillion EV autos to be bought. So ICE is picking up less than 7% of this 50 year market. Of course am this is assuming that total auto sales continue to grow at 3.6%/y through 2070. I would hope that in the next 50 year the human population growth would slow down so that not so many auto would be needed.
Regardless on a long time scale, only a finite number of ICE autos will ever be built. This is part of puting an ultimate cap on oil demand.
For comparison, in the 50 years following 2020, my model expect 13.7 trillion EV autos to be bought. So ICE is picking up less than 7% of this 50 year market. Of course am this is assuming that total auto sales continue to grow at 3.6%/y through 2070. I would hope that in the next 50 year the human population growth would slow down so that not so many auto would be needed.
Regardless on a long time scale, only a finite number of ICE autos will ever be built. This is part of puting an ultimate cap on oil demand.
Similar threads
