-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Ugh. Another Model S fire - 2013-11-06
- Thread starter tommolog
- Start date
-
- Tags
- Driving Dynamics fire Model S
NigelM
Recovering Member
Yup! But we don't have the capacity to edit out snippy comments from so many posts/quotes so unfortunately some useful stuff gets thrown out with the comments; sorry about that.
dflye
S Sig Perf 414, VIN 814
Ah, but it becomes like rubber-necking while driving... you hate the folks in front of you that cause the issue, but you can't help but cast that sideways glance as you slowly roll past to see just what sort of distraction caused the issue in the first place!
gg_got_a_tesla
Model S: VIN 65513, Model 3: VIN 1913
I'll be honest: I'd prefer a metal sheet covering up those entrails (between the cross members in front of the pack) rather than the plastic that's there now.
Also, not sure if this has been discussed upthread but, is it possible that the flat underside creates more of a suction effect for road debris when compared to an ICE with similar ground clearance but, with a relatively irregular gaggle of various parts and pipes underneath?!
dflye
S Sig Perf 414, VIN 814
Similar to the supposed jet engine testing of throwing (hopefully defrosted) turkeys at running jet engines to assess damage from impact, one would hope that Tesla did at least a modicum of testing of impacts at various incident angles to the underbelly of the Model S.
Although I'd hope the Model S testing used the apparently ever-present wayward solid metal trailer hitches that slip off of their owning vehicle when someone forgets to properly set and check the locking pin!
Although I'd hope the Model S testing used the apparently ever-present wayward solid metal trailer hitches that slip off of their owning vehicle when someone forgets to properly set and check the locking pin!
Road Debris
Road debris is more common than most people seem to think. This piece from wikipedia is an eye-opener.
"Road debris is a hazard that can cause fishtailing and damage like a flat tire or even a traffic accident with injury or death. Road debris can cause loss of control crashes, rollover crashes, or penetration of the passenger compartment by the debris.
Released in early 2013, NHTSA data for 2011 showed over 800 Americans were killed that year in vehicle collisions with road debris. Mississippi, Wyoming, Arkansas, Kentucky and Louisiana were the top five states for these crash deaths to most likely occur. Also in 2011, New York and Massachusetts saw significant increases in road debris-vehicular crash deaths, unlike other big, populated states. In 2004, a AAA Foundation for Traffic Safety study revealed that vehicle-related road debris caused 25,000 accidents—and nearly 100 deaths—each year. At highway speeds, even small debris can be deadly."
http://en.wikipedia.org/wiki/Road_debris
In this perspective, Model S is a very safe car. It alerts passengers. It gives enough time to exit. Fire is (relatively) small and contained.
Road debris is more common than most people seem to think. This piece from wikipedia is an eye-opener.
"Road debris is a hazard that can cause fishtailing and damage like a flat tire or even a traffic accident with injury or death. Road debris can cause loss of control crashes, rollover crashes, or penetration of the passenger compartment by the debris.
Released in early 2013, NHTSA data for 2011 showed over 800 Americans were killed that year in vehicle collisions with road debris. Mississippi, Wyoming, Arkansas, Kentucky and Louisiana were the top five states for these crash deaths to most likely occur. Also in 2011, New York and Massachusetts saw significant increases in road debris-vehicular crash deaths, unlike other big, populated states. In 2004, a AAA Foundation for Traffic Safety study revealed that vehicle-related road debris caused 25,000 accidents—and nearly 100 deaths—each year. At highway speeds, even small debris can be deadly."
http://en.wikipedia.org/wiki/Road_debris
In this perspective, Model S is a very safe car. It alerts passengers. It gives enough time to exit. Fire is (relatively) small and contained.
Last edited:
Tesla should sue the crap out of these bozos. Faux "News" is a joke.
- - - Updated - - -
Forward to FOX Lets make sh*t up today "News"
- - - Updated - - -
Forward to FOX Lets make sh*t up today "News"
This doesn't really detract from your point, but I think the correct number is more like 122 fires per day, on average. http://www.usfa.fema.gov/downloads/pdf/statistics/v13i11.pdf (three year period of 194,000 all-vehicle fires x 68.9% of those being "standard passenger vehicles" = total number of fires over three years, then divide by 3 and again by 365 to get average fires per day). I keep seeing the 500 number but it seems wrong. I don't know if that number counts stationary vehicle fires or what .. but I think the most apples to apples comparison is a highway vehicle fire because of the greater hazard that such a fire usually presents (regardless of whether you are in an ICE or EV, I would much rather have my car flame up when I am stopped at a light versus going 65 on the highway assuming I had the choice).
Agree - I think this is the most important thing that we ought to be thinking about. 2 Model S's have had their armor plating pierced from driving over a metal object. To me, that is enough to warrant some improvement, regardless of whether or not it renders the Model S less safe statistically speaking, because piercing the armor will (always? usually?) result in a fire.
In ICE vehicles, there have been recalls were there was an identified theoretical/potential hazard even though there were no fires. The NHSTA does not always just wait around for there to be enough fires to be statistically significant.. they usually act before it gets to that point if there is a design issue. I think that 2 piercings in a month might qualify as such.
I love Tesla, but that doesn't mean we should hold them to a lesser standard. I am pretty sure Tesla is going to do something anyway.
[/QUOTE]
RoverS
Banned
gglockner
Member
As they say, "long time listener, first time caller". I'm not an owner yet but I'm trying to work a Model S into my budget.
This is a very long thread, and I tried to read as much as possible, and now I want to add my $0.02.
First, as far as I'm concerned, I have no problems with the incident in Mexico. The car had a high-speed collision with a wall. The fact that the driver had no serious injury is a testament to the safety of the car. However, this still leaves the two road-debris incidents: the first in the Seattle area and the latest in Tennessee.
The fact that many other cars have car fires seems irrelevant to me. For me, the issue is that two cars had a fire in a common road-debris situation - one that rarely leads to a car fire with any other car. To the best of my knowledge - and I welcome any evidence to the contrary - I am unaware of any road-debris situations leading to a car fire with another EV model. The Volt and Leaf have been in the marketplace long enough to have encountered road debris, and an incident would be just as big in the news.
So it seems possible that the design of the Tesla may make it vulnerable to road debris. I'm sure Tesla and/or NHTSA are investigating this, and they will make changes if anything is found, similar to what we saw with other design issues like sudden acceleration in Toyotas (about 4 years ago) or Audis (about 30 years ago), or the rollover issue with Isuzu Troopers (about 20 years ago).
And in the meantime, the fact that no one perished in these incidents is positive news.
This is a very long thread, and I tried to read as much as possible, and now I want to add my $0.02.
First, as far as I'm concerned, I have no problems with the incident in Mexico. The car had a high-speed collision with a wall. The fact that the driver had no serious injury is a testament to the safety of the car. However, this still leaves the two road-debris incidents: the first in the Seattle area and the latest in Tennessee.
The fact that many other cars have car fires seems irrelevant to me. For me, the issue is that two cars had a fire in a common road-debris situation - one that rarely leads to a car fire with any other car. To the best of my knowledge - and I welcome any evidence to the contrary - I am unaware of any road-debris situations leading to a car fire with another EV model. The Volt and Leaf have been in the marketplace long enough to have encountered road debris, and an incident would be just as big in the news.
So it seems possible that the design of the Tesla may make it vulnerable to road debris. I'm sure Tesla and/or NHTSA are investigating this, and they will make changes if anything is found, similar to what we saw with other design issues like sudden acceleration in Toyotas (about 4 years ago) or Audis (about 30 years ago), or the rollover issue with Isuzu Troopers (about 20 years ago).
And in the meantime, the fact that no one perished in these incidents is positive news.
Well it seems in the half hour it took me to type out a detailed post on my iPhone the thread got cleaned up and my post eventually went off to snipland. Apologies for that.
But I did have a good point and I want to repost without any possible snippiness. Originally I was defending the excellent post on the Poisson distribution by Mario Kadastik (someone who may very well spend his living dealing with statistics and know more than the rest of us put together).
Also to correct some math. My first degree was mathematics and although it may have been a while ago and I'm sure I'm a little rusty, I do know that units are one of the most important things in an equation.
The units are all over the place.
138,600/230,000,000
You don't introduce your units but I'm guessing this is:
'car fires / year' / 'number of vehicles'
So that is what the 0.0006 is. It is already the risk of a fire in one year to one vehicle which is what you're trying to calculate.
ie fire risk per year per vehicle
(It is NOT fire risk over the life of the car. That would be
risk per year x ave. number of years / number of cars
ie. 138,600 x 5.5/ 230,000
you don't need to go there, to multiply or divide by anything.
To divide 0.0006 by 11 (or 5.5 or any number of years) makes your new units fire risk per year SQUARED per vehicle.
Anyway If you have multiplied by the average car age 11/2 years to get the risk over the life of the car, then you have to divide by 11/2 instead of 11 later to get the risk per year. Either way the number of years cancel out. But more importantly as I said you already had what you wanted on the first equation before any additional multiplying or dividing.
Q.E.D. ;-)
Although Mario Kadastik is right about the Poisson distribution, it's not something that translates easily for the public. So you probably have to do a fire risk per year per vehicle based on just 3 fires to satisfy them. So again at this point we have that Teslas are about twice as safe as ICE vehicles.
Average fire risk per year per ICE vehicle = 0 .0006
Average fire risk per year per Tesla = 0.00024
But I did have a good point and I want to repost without any possible snippiness. Originally I was defending the excellent post on the Poisson distribution by Mario Kadastik (someone who may very well spend his living dealing with statistics and know more than the rest of us put together).
Also to correct some math. My first degree was mathematics and although it may have been a while ago and I'm sure I'm a little rusty, I do know that units are one of the most important things in an equation.
The units are all over the place.
138,600/230,000,000
You don't introduce your units but I'm guessing this is:
'car fires / year' / 'number of vehicles'
So that is what the 0.0006 is. It is already the risk of a fire in one year to one vehicle which is what you're trying to calculate.
ie fire risk per year per vehicle
(It is NOT fire risk over the life of the car. That would be
risk per year x ave. number of years / number of cars
ie. 138,600 x 5.5/ 230,000
you don't need to go there, to multiply or divide by anything.
To divide 0.0006 by 11 (or 5.5 or any number of years) makes your new units fire risk per year SQUARED per vehicle.
Anyway If you have multiplied by the average car age 11/2 years to get the risk over the life of the car, then you have to divide by 11/2 instead of 11 later to get the risk per year. Either way the number of years cancel out. But more importantly as I said you already had what you wanted on the first equation before any additional multiplying or dividing.
Q.E.D. ;-)
Although Mario Kadastik is right about the Poisson distribution, it's not something that translates easily for the public. So you probably have to do a fire risk per year per vehicle based on just 3 fires to satisfy them. So again at this point we have that Teslas are about twice as safe as ICE vehicles.
Average fire risk per year per ICE vehicle = 0 .0006
Average fire risk per year per Tesla = 0.00024
I posted these "statistics" over on the Tesla Motors forum, but it might be worthwhile to recap here since my statistics actually suggest (but don't prove) that 2 fires is not unusual relatively speaking (even disregarding the point made earlier about whether 2 incidents is enough to make a statistically valid claim either way).
Google has NHSTA data that shows crashes by year through 2010: http://www.google.com/publicdata/ex...state&ifdim=state&hl=en_US&dl=en_US&ind=false
There were 130,892,200 registered automobiles in the U.S. based on the foregoing is 2010 (this disregards trucks, etc.)
From a different report: In the three year period 2008-10, there were 194,000 highway vehicle fires! 68.9% of these were passenger vehicles (the comparable population for Tesla). That means an average of 44,555 standard passenger vehicle fires PER YEAR. See http://www.usfa.fema.gov/downloads/pdf/statistics/v13i11.pdf
Taking the 2010 data from the google link, that means 44,555 / 130,892,200 = 0.000341 fires/registered passenger vehicle in 2010.
If Tesla has 20,000 Model S on the road in the U.S. (not sure about that, but I saw it somewhere), then we have 2 / 20,000 = 0.0001 fires / registered Model S (I am ignoring the Mexico fire because the statistics for fires/vehicle are for the U.S. only and I have no idea what Mexico's rate of fire/vehicle might be). Even if Tesla only has 10,000, that is 0.0002 fires / registered vehicle. Substantially less than for all vehicles in 2010.
There are some major limitations to this analysis:
(1) I am not taking into account that the bulk of these vehicles were delivered in the last 6 months or so; I am comparing 1 year risk for ICE vehicles to a risk for Tesla vehicles that has an shorter (on average) time period. What would really be needed is a month to month analysis gong back a few years for ICE vehicles, and then compare that to 2 fires in a month for Tesla rather than going by yearly data. But I don't have the data for deliveries by month or crash/fire data by month.
(2) The fire data is only for highway fires. It seems likely that the Tesla would/will have zero non-highway fires due to the 1/4" armor which is only going to get pierced in a high speed situation (I am supposing), so that will reduce all-fire risk relative to ICE.
(3) To know whether the observed 2 fire incidents in a 30 day period is statistically significant, you would need to calculate a deviation measure, either using a standard deviation (bell curve) or the Poisson distribution (more complicated). You would need to look at something like vehicle fires per registered vehicle / month over a multiyear period; and then compute the deviation of that measure from month to month figures. WIth a calculated deviation measure, you might be able to answer the question of whether 2 fires (in the US) in one month, with a population of [20,000] vehicles, is statistically significant or not. If I can find the data to do this, I will, unless someone else beats me to it.
(4) The fire data comes from the FIRS database and is for a 3 year period that included 2010, but is averaged. The vehicle data comes from the NHSTA and is for 2010 only (someone with more time than me could probably do a 3 year average from NHSTA data ... it won't change the results much though). Comparing different data sets is always a little tricky and introduces another source of possible error.
(5) There is an underlying assumption that a Tesla vehicle is driven in the same manner and same average miles as the average vehicle in the 2010 data (about 12,000 miles per year). This may or may not be true. Do Tesla drivers drive more on the highway? More at night when it is cool (but road hazards more dangerous)? Some of the discrepancy could be related to driving habits of Tesla drivers vs others. OTOH, I would expect Tesla drivers as a whole are safer than the general population (more educated, probably a little older and more experienced, etc.)
Google has NHSTA data that shows crashes by year through 2010: http://www.google.com/publicdata/ex...state&ifdim=state&hl=en_US&dl=en_US&ind=false
There were 130,892,200 registered automobiles in the U.S. based on the foregoing is 2010 (this disregards trucks, etc.)
From a different report: In the three year period 2008-10, there were 194,000 highway vehicle fires! 68.9% of these were passenger vehicles (the comparable population for Tesla). That means an average of 44,555 standard passenger vehicle fires PER YEAR. See http://www.usfa.fema.gov/downloads/pdf/statistics/v13i11.pdf
Taking the 2010 data from the google link, that means 44,555 / 130,892,200 = 0.000341 fires/registered passenger vehicle in 2010.
If Tesla has 20,000 Model S on the road in the U.S. (not sure about that, but I saw it somewhere), then we have 2 / 20,000 = 0.0001 fires / registered Model S (I am ignoring the Mexico fire because the statistics for fires/vehicle are for the U.S. only and I have no idea what Mexico's rate of fire/vehicle might be). Even if Tesla only has 10,000, that is 0.0002 fires / registered vehicle. Substantially less than for all vehicles in 2010.
There are some major limitations to this analysis:
(1) I am not taking into account that the bulk of these vehicles were delivered in the last 6 months or so; I am comparing 1 year risk for ICE vehicles to a risk for Tesla vehicles that has an shorter (on average) time period. What would really be needed is a month to month analysis gong back a few years for ICE vehicles, and then compare that to 2 fires in a month for Tesla rather than going by yearly data. But I don't have the data for deliveries by month or crash/fire data by month.
(2) The fire data is only for highway fires. It seems likely that the Tesla would/will have zero non-highway fires due to the 1/4" armor which is only going to get pierced in a high speed situation (I am supposing), so that will reduce all-fire risk relative to ICE.
(3) To know whether the observed 2 fire incidents in a 30 day period is statistically significant, you would need to calculate a deviation measure, either using a standard deviation (bell curve) or the Poisson distribution (more complicated). You would need to look at something like vehicle fires per registered vehicle / month over a multiyear period; and then compute the deviation of that measure from month to month figures. WIth a calculated deviation measure, you might be able to answer the question of whether 2 fires (in the US) in one month, with a population of [20,000] vehicles, is statistically significant or not. If I can find the data to do this, I will, unless someone else beats me to it.
(4) The fire data comes from the FIRS database and is for a 3 year period that included 2010, but is averaged. The vehicle data comes from the NHSTA and is for 2010 only (someone with more time than me could probably do a 3 year average from NHSTA data ... it won't change the results much though). Comparing different data sets is always a little tricky and introduces another source of possible error.
(5) There is an underlying assumption that a Tesla vehicle is driven in the same manner and same average miles as the average vehicle in the 2010 data (about 12,000 miles per year). This may or may not be true. Do Tesla drivers drive more on the highway? More at night when it is cool (but road hazards more dangerous)? Some of the discrepancy could be related to driving habits of Tesla drivers vs others. OTOH, I would expect Tesla drivers as a whole are safer than the general population (more educated, probably a little older and more experienced, etc.)
Last edited:
BillHamp this is why I think a lot of folks here take your posts with a grain of salt. You handpick only the numbers you want to support your point, like in your 'statistical' analysis trying to adjust for only factors biased against your thesis. Which is fine, if you didn't state them like misleading facts to prove your point.
To rebutt this particular piece of information you provided, the $38 million loss is the GAAP number, basically where Tesla recognizes all the costs in making the car (parts, labor) but is only allowed to recognize a fraction of the cash revenue received because of the lease. We can sit here and debate the arbitrary nature of GAAP and IFRS accounting standards, but the bottom line is despite the huge amounts of money spent towards Gen3 and large infrastructure projects for future expansion, they are quite profitable. I would argue cash flow is a better indication of what's happening. They made $102 million from operations (sales less expenses) and invested $77 million in capital expenditures to lay groundwork for future growth, leaving a free cash flow of +$25 million. That would be the actual money in their bank account from company activities (note: this excludes financing activities that raised more money for them in the bank).
This would only matter if the BMW and Mercedes had gas tanks that occupied the full undercarriage of the vehicle, which is not the case. The risk of the gas tank rupturing is much lower due to the much smaller surface area that an object would need to strike. You are comparing apples and oranges here.
Highway vehicles include cars, trucks, recreational vehicles, motorcycles, and other vehicles intended for road use. “Highway vehicle fire” describes the type of vehicle. It does not mean the fire occurred on a highway based on the NFPA tterminology. Therefore, you don't need the any adjustment.
Statistics are all well and good, and will probably indicate in a weighted manner that the Model S is by a measure more safe than an ICE when the car is subject to a fire situation. However, what Tesla needs to address, through some type of simplified communication strategy, is the public perception that is forming in the general populous about Tesla cars and fires. Statistics can probably be used as part of the communication, but an explanation of the measured safety features perhaps could be part of the message strategy, i.e. that the battery pack is designed to vent downwards, etc. Similarly, one has to be cautious about the way this issue is addressed by not making it a determined focus.
Similar threads
- Question
