GRA
Posts: 11176
Joined: Mon Sep 19, 2011 1:49 pm
Location: East side of San Francisco Bay

Re: Official Tesla Model 3 thread

Thu May 23, 2019 5:57 pm

lpickup wrote:
GRA wrote: I've done this calc for my own requirements.
They may not meet your own requirements. But that's not what matters. What matters is when carmakers stop focusing on just adding more battery and start focusing on reducing costs.

Even Tesla has started doing this, making the statement that they can pump out more SR+ vehicles that LR vehicles (because they are battery constrained), so that is their focus right now. I personally think that 240 miles is a bit below what I think the sweet spot is (280 miles).

I don't know exactly what the motivations were, but Nissan decided to focus on the sub-$30K niche with the 40kWh LEAF.

VW appears headed in this direction as well, but I will reserve comment until they do prove that they have battery supply and are really serious about entering the market.

And like I said, it may not be 4-5 years. Maybe it's 7-8. But it's coming. Folks that say otherwise are burying their heads in the sand.
In the meantime, as battery supply is going to be an issue while manufacturing scales up (even if raw material supply doesn't further slow things down), it makes far more sense, if reducing GHGs and local air pollution are your motivations, to put several more people in much less expensive PHEVs with smaller batteries sized to handle routine needs, instead of sticking huge, expensive batteries in a much smaller number of cars, when most of that capacity will go unused most of the time. One Niro/Kona/Soul BEV = 3.5 Volt 2s or 7.3 Prius Primes.
lpickup wrote:P.S. While I do agree with your factoring in battery longevity and temperature into required range calculations, my opinion is that is more of a local travel phenomenon. Cold temps on a trip should only really matter for the first leg of the trip (and maybe not even that if you pre-heat), so range loss is not going to be as extreme as when you are just driving around town.
No, it's the opposite. Around town you benefit from Regen, on a trip (uphill) you don't. On a road trip you want to stop as little as possible, so times between pre-heating are going to be long, and any energy for that while charging just makes your charge time longer.
lpickup wrote:Degradation could be a factor of course, but again, I don't think you have to assume any more than 20% loss. Why? Because by then a battery replacement (if needed) should be reasonably affordable for a 10 year old car, and in a Tesla you are probably only looking at 10-15% loss in that timeframe. Factoring in 50% loss in 12-15 years is far too conservative.
I'm not going to assume anything about future battery prices or the availablility of replacements. We've got plenty of LEAF owners here who did that; enough said. As noted in another post, the best capacity warranty extant AFAIK is 70% for 8 years, and we've got plenty of examples of batteries losing more than that in less time. At the moment, the oldest Model S is not quite 7 y.o., and there are Teslas that have exceeded that 30% loss. I know of no battery chemistry currently available which can provide 12 years of service at anything like the capacity % required for other than local use. which is one reason I'm a fan of battery leasing with guaranteed capacity.
lpickup wrote:I simulated my long distance trip using various vehicles as well. I found that even the Model 3 SR could handle the trip I take without any timing impact to my driving pattern (although the stops would be strictly regimented), which was a big surprise to me. That's 220 miles of range and slower charging that I can get from my LR. So even if I experience 30% degradation I am not too worried.
Unfortunately, even the Model 3 LR couldn't handle mine except when brand new, and it's far too expensive to be mass market.
Guy [I have lots of experience designing/selling off-grid AE systems, some using EVs but don't own one. Local trips are by foot, bike and/or rapid transit].

The 'best' is the enemy of 'good enough'. Copper shot, not Silver bullets.

lorenfb
Posts: 2273
Joined: Tue Dec 17, 2013 10:53 pm
Delivery Date: 22 Nov 2013
Leaf Number: 416635
Location: SoCal

Re: Official Tesla Model 3 thread

Thu May 23, 2019 6:09 pm

Oils4AsphaultOnly wrote:
lorenfb wrote:
Oils4AsphaultOnly wrote:
He's a seasoned statistician, whose analysis (the ones that I haven't linked to) have been far more accurate than anything published out there. Anyway, I've provided the reference. You just chose to mock it.
Do some reading yourself about random walk theory, and not focus on "He's a seasoned statistician".

https://corporatefinanceinstitute.com/r ... lk-theory/
You truly are in your own little world! Here's a better reference and far more relevant to my post about PEV adoption: https://en.wikipedia.org/wiki/Random_walk

What the freak does the stock market have to do with exponential growth of BEV's, which is what you asked the data for?!
Read his naive methodology:
In an effort to make a simple, but compelling forecast of plug-in EV market share, I have boiled things down to a random walk on a logit scale. May sound complex, but it actually makes very few assumptions. I examine market share from 2012 to 2018, computing the annual increase in logit market share (log(S/(1-S)). So I've got now 6 data points. I compute the mean and sample standard deviations. I make the assumption that future logit differences have the same distribution as was operative in the past. Thus, the future is modeled as a random walk with a certain mean and standard deviation. I can than predict the mean and variance (inclusive of parametric uncertainty) of the logit random walk. From this I can convert back into the market share scale, tracing out the mean path and a 90% predictive envelope.
Your "statistician" is attempting to forecast future product demand/sales based on a random walk process (historical data), similar to
forecasting future stock prices, which basically has been proven baseless. One can attempt to forecast future production volumes based
on historical data, because all the key model factors are known, e.g. worker productivities, supply chain lead times, etc. But to attempt
to forecast future demand for a product based on historical data only is naive - similar to forecasting future stock prices. One lacks key
market info such as, competitive future products/advertising, future technology breakthroughs, new market entries, etc., which are
critical for optimizing the forecasting model and are basically not know.
#1 Leaf SL MY 9/13: 74K miles, 48 Ahrs, 5.2 miles/kWh (average), Hx=70, SOH=78, L2 - 100% > 1000, temp < 95F, (DOD) > 20 Ahrs
#2 Leaf SL MY 12/18: 4.5K miles, 115 Ahrs, 5.5 miles/kWh (average), Hx=98, SOH=99, DOD > 20%, temp < 105F

Oils4AsphaultOnly
Posts: 702
Joined: Sat Oct 10, 2015 4:09 pm
Delivery Date: 20 Nov 2016
Leaf Number: 313890
Location: Arcadia, CA

Re: Official Tesla Model 3 thread

Thu May 23, 2019 8:30 pm

lorenfb wrote:
Oils4AsphaultOnly wrote:
lorenfb wrote:
Do some reading yourself about random walk theory, and not focus on "He's a seasoned statistician".

https://corporatefinanceinstitute.com/r ... lk-theory/
You truly are in your own little world! Here's a better reference and far more relevant to my post about PEV adoption: https://en.wikipedia.org/wiki/Random_walk

What the freak does the stock market have to do with exponential growth of BEV's, which is what you asked the data for?!
Read his naive methodology:
In an effort to make a simple, but compelling forecast of plug-in EV market share, I have boiled things down to a random walk on a logit scale. May sound complex, but it actually makes very few assumptions. I examine market share from 2012 to 2018, computing the annual increase in logit market share (log(S/(1-S)). So I've got now 6 data points. I compute the mean and sample standard deviations. I make the assumption that future logit differences have the same distribution as was operative in the past. Thus, the future is modeled as a random walk with a certain mean and standard deviation. I can than predict the mean and variance (inclusive of parametric uncertainty) of the logit random walk. From this I can convert back into the market share scale, tracing out the mean path and a 90% predictive envelope.
Your "statistician" is attempting to forecast future product demand/sales based on a random walk process (historical data), similar to
forecasting future stock prices, which basically has been proven baseless. One can attempt to forecast future production volumes based
on historical data, because all the key model factors are known, e.g. worker productivities, supply chain lead times, etc. But to attempt
to forecast future demand for a product based on historical data only is naive - similar to forecasting future stock prices. One lacks key
market info such as, competitive future products/advertising, future technology breakthroughs, new market entries, etc., which are
critical for optimizing the forecasting model and are basically not know.
You criticize his methodology as "naive", yet missed the first sentence of "In an effort to make a simple, but compelling forecast". You obviously didn't bother reading the rest of his case where he retested the predictability of his "naive" model using 2012-2015 data to predict the subsequent years and demonstrated how it was more accurate than BNEF's more complicated model and predictions. His entire point was that the adoption curve is sufficiently defined within this "naive" model, and that a more complicated model isn't necessary to give a good approximation of how fast EV adoption is progressing.

Pay attention to his graph! By 2025, there's a huge 25% spread of EV adoption between the low and high ends of his predictive envelope. If you factor in market info, you might end up with a more precise prediction model, but you're still going to have exponential growth that exceeds 60% market share by 2030. You're asking for the theory of relativity, when newtonian physics is enough.
:: Model 3 LR :: acquired 9 May '18
:: Leaf S30 :: build date: Sep '16 :: purchased: Nov '16
100% Zero transportation emissions (except when I walk) and loving it!

lorenfb
Posts: 2273
Joined: Tue Dec 17, 2013 10:53 pm
Delivery Date: 22 Nov 2013
Leaf Number: 416635
Location: SoCal

Re: Official Tesla Model 3 thread

Fri May 24, 2019 6:54 am

Oils4AsphaultOnly wrote:
lorenfb wrote:
Oils4AsphaultOnly wrote:
You truly are in your own little world! Here's a better reference and far more relevant to my post about PEV adoption: https://en.wikipedia.org/wiki/Random_walk

What the freak does the stock market have to do with exponential growth of BEV's, which is what you asked the data for?!
Read his naive methodology:
In an effort to make a simple, but compelling forecast of plug-in EV market share, I have boiled things down to a random walk on a logit scale. May sound complex, but it actually makes very few assumptions. I examine market share from 2012 to 2018, computing the annual increase in logit market share (log(S/(1-S)). So I've got now 6 data points. I compute the mean and sample standard deviations. I make the assumption that future logit differences have the same distribution as was operative in the past. Thus, the future is modeled as a random walk with a certain mean and standard deviation. I can than predict the mean and variance (inclusive of parametric uncertainty) of the logit random walk. From this I can convert back into the market share scale, tracing out the mean path and a 90% predictive envelope.
Your "statistician" is attempting to forecast future product demand/sales based on a random walk process (historical data), similar to
forecasting future stock prices, which basically has been proven baseless. One can attempt to forecast future production volumes based
on historical data, because all the key model factors are known, e.g. worker productivities, supply chain lead times, etc. But to attempt
to forecast future demand for a product based on historical data only is naive - similar to forecasting future stock prices. One lacks key
market info such as, competitive future products/advertising, future technology breakthroughs, new market entries, etc., which are
critical for optimizing the forecasting model and are basically not know.
You criticize his methodology as "naive", yet missed the first sentence of "In an effort to make a simple, but compelling forecast". You obviously didn't bother reading the rest of his case where he retested the predictability of his "naive" model using 2012-2015 data to predict the subsequent years and demonstrated how it was more accurate than BNEF's more complicated model and predictions. His entire point was that the adoption curve is sufficiently defined within this "naive" model, and that a more complicated model isn't necessary to give a good approximation of how fast EV adoption is progressing.

Pay attention to his graph! By 2025, there's a huge 25% spread of EV adoption between the low and high ends of his predictive envelope. If you factor in market info, you might end up with a more precise prediction model, but you're still going to have exponential growth that exceeds 60% market share by 2030. You're asking for the theory of relativity, when newtonian physics is enough.
And to sum it up and end the issue:

With just six data points, one is able to forecast a market 15 years out, right?

A totally laughable joke!

A common case where the wrong theory and background are used to solve a problem.
#1 Leaf SL MY 9/13: 74K miles, 48 Ahrs, 5.2 miles/kWh (average), Hx=70, SOH=78, L2 - 100% > 1000, temp < 95F, (DOD) > 20 Ahrs
#2 Leaf SL MY 12/18: 4.5K miles, 115 Ahrs, 5.5 miles/kWh (average), Hx=98, SOH=99, DOD > 20%, temp < 105F

Oils4AsphaultOnly
Posts: 702
Joined: Sat Oct 10, 2015 4:09 pm
Delivery Date: 20 Nov 2016
Leaf Number: 313890
Location: Arcadia, CA

Re: Official Tesla Model 3 thread

Fri May 24, 2019 11:40 am

lorenfb wrote:
Oils4AsphaultOnly wrote:
lorenfb wrote:
Read his naive methodology:


Your "statistician" is attempting to forecast future product demand/sales based on a random walk process (historical data), similar to
forecasting future stock prices, which basically has been proven baseless. One can attempt to forecast future production volumes based
on historical data, because all the key model factors are known, e.g. worker productivities, supply chain lead times, etc. But to attempt
to forecast future demand for a product based on historical data only is naive - similar to forecasting future stock prices. One lacks key
market info such as, competitive future products/advertising, future technology breakthroughs, new market entries, etc., which are
critical for optimizing the forecasting model and are basically not know.
You criticize his methodology as "naive", yet missed the first sentence of "In an effort to make a simple, but compelling forecast". You obviously didn't bother reading the rest of his case where he retested the predictability of his "naive" model using 2012-2015 data to predict the subsequent years and demonstrated how it was more accurate than BNEF's more complicated model and predictions. His entire point was that the adoption curve is sufficiently defined within this "naive" model, and that a more complicated model isn't necessary to give a good approximation of how fast EV adoption is progressing.

Pay attention to his graph! By 2025, there's a huge 25% spread of EV adoption between the low and high ends of his predictive envelope. If you factor in market info, you might end up with a more precise prediction model, but you're still going to have exponential growth that exceeds 60% market share by 2030. You're asking for the theory of relativity, when newtonian physics is enough.
And to sum it up and end the issue:

With just six data points, one is able to forecast a market 15 years out, right?

A totally laughable joke!

A common case where the wrong theory and background are used to solve a problem.
Indeed! I've led, yet the parched horse doth not drink. :(
:: Model 3 LR :: acquired 9 May '18
:: Leaf S30 :: build date: Sep '16 :: purchased: Nov '16
100% Zero transportation emissions (except when I walk) and loving it!

LeafMuranoDriver
Posts: 351
Joined: Wed May 13, 2015 8:15 am
Delivery Date: 01 May 2015
Leaf Number: 323189
Location: Los Angeles, CA

Re: Official Tesla Model 3 thread

Fri May 24, 2019 2:04 pm

This thread has really gone down hill :oops:
2015 Leaf S w/ QC
Mfg: Jan. '15 // Pur: May '15
Best: 6/27/15 - 1,416 Miles - 292 GIDs - 63.11 Ahr - 22.6 kWh - 100% SOH - 97.2% Hx - 4 QC, 29 L2
Worst: 12/12/15 - 4,616 Miles - 259 GIDs - 55.02 Ahr - 20.1 kWh - 99% SOH - 82.8% Hx - 17 QC, 96 L2[/size]

Oils4AsphaultOnly
Posts: 702
Joined: Sat Oct 10, 2015 4:09 pm
Delivery Date: 20 Nov 2016
Leaf Number: 313890
Location: Arcadia, CA

Re: Official Tesla Model 3 thread

Fri May 24, 2019 4:02 pm

LeafMuranoDriver wrote:This thread has really gone down hill :oops:
What would you like to know about the model 3?

It's constantly getting new features, so there won't be any end to this thread.
:: Model 3 LR :: acquired 9 May '18
:: Leaf S30 :: build date: Sep '16 :: purchased: Nov '16
100% Zero transportation emissions (except when I walk) and loving it!

lorenfb
Posts: 2273
Joined: Tue Dec 17, 2013 10:53 pm
Delivery Date: 22 Nov 2013
Leaf Number: 416635
Location: SoCal

Re: Official Tesla Model 3 thread

Tue Jun 04, 2019 8:12 am

InsideEVs May M3 update; https://insideevs.com/news/352626/ev-sa ... -may-2019/

Summary:

1. 2019 YTD (5 months) M3 U.S. - 46K
2. 2019 YTD Average per month - 9K
3. 2018 Q4 M3 U.S. - 62K, average per month - 20K
4. May 2019 M3 U.S. - 14K

Conclusion - Three months in 2018 outsold five months in 2019. M3 YTD U.S. demand compared to 2018 Q4 is weak.
#1 Leaf SL MY 9/13: 74K miles, 48 Ahrs, 5.2 miles/kWh (average), Hx=70, SOH=78, L2 - 100% > 1000, temp < 95F, (DOD) > 20 Ahrs
#2 Leaf SL MY 12/18: 4.5K miles, 115 Ahrs, 5.5 miles/kWh (average), Hx=98, SOH=99, DOD > 20%, temp < 105F

User avatar
jlv
Moderator
Posts: 1159
Joined: Thu Apr 24, 2014 6:08 pm
Delivery Date: 30 Apr 2014
Leaf Number: 424487
Location: Massachusetts

Re: Official Tesla Model 3 thread

Tue Jun 04, 2019 8:58 am

Really amazing negative spin in your "summary". Great job completely ignoring all the international deliveries (which these numbers do not show).

Here's the insideevs numbers for U.S. deliveries of the 3 for last year and this.

Code: Select all

Jan    Feb    Mar    Apr    May    Jun    Jul    Aug    Sep    Oct    Nov    Dec    Jan    Feb    Mar    Apr    May
1875   2485   3820   3750   6000   5902   14250  17800  22250  17750  18650  25250  6500   5750   10175  10050  13950
Looking at the whole table on insideevs, I see that LEAF U.S. sales continue to be less than 1/10 of the 3 sales. If the 3 demand is "weak", what do you call the LEAF demand?
LEAF '13 SL+Prem (mfg 12/13, leased 4/14, bought 5/17, sold 11/18) 34K mi, AHr 58, SOH 87%
Tesla S 75D (3/17)
Tesla X 100D (12/18)
85K 100% BEV miles since '14
ICE free since '18

lorenfb
Posts: 2273
Joined: Tue Dec 17, 2013 10:53 pm
Delivery Date: 22 Nov 2013
Leaf Number: 416635
Location: SoCal

Re: Official Tesla Model 3 thread

Tue Jun 04, 2019 12:16 pm

jlv wrote:
Looking at the whole table on insideevs, I see that LEAF U.S. sales continue to be less than 1/10 of the 3 sales. If the 3 demand is "weak", what do you call the LEAF demand?
So what, Nissan doesn't rely on Leaf sales for its viability does it? Please avoid non sequiturs.
#1 Leaf SL MY 9/13: 74K miles, 48 Ahrs, 5.2 miles/kWh (average), Hx=70, SOH=78, L2 - 100% > 1000, temp < 95F, (DOD) > 20 Ahrs
#2 Leaf SL MY 12/18: 4.5K miles, 115 Ahrs, 5.5 miles/kWh (average), Hx=98, SOH=99, DOD > 20%, temp < 105F

Return to “Other Electric Cars & Plug-In Hybrids”