Discussion of statistical analysis of vehicle fires as it relates to Model S

[SCW-Greg]

Had I known you would be so rude and demeaning, I wouldn't have asked. All I wanted was an opinion on what the NHTSA would find/do. I don't want to get a car and then have to send it back in a recall if that can be avoided. I would rather wait for the fix, if one is coming, and then buy. I was hoping someone with some more knowledge of the fires would be able to offer an assessment of liklihood of a recall/fix/change in the car. Forget I asked. I had no idea people on this forum were so hostile. If this is how Tesla drivers act, I'm not sure I want to be seen in one.

I didn't see Kruggerand's response as hostile, but more as a challenge to be aware of how normal cars handle collisions and wrecks (that we, and the press, are so desensitized to). If you follow his threads you'll find him quite funny and intelligent in response.

Meanwhile, let me add, the NHSTA report showed a graph that demarks the safety factor for the Model S, that the Tesla created a significant dip in safety trends - in the likelihood of being harmed in an accident. I'll see if I can dig that graph up.

Keep in mind, batteries don't explode, not like fuel does. They start and burn slowly. So far the Model S in every situation, has told the driver, there's a problem and to pull over before the car shuts down.

This car is a 4700lb tank, simply the safest car you can be in. To me this is more important than anything, particularly for my family.

 

[lolachampcar]

I think folks on the forum need to spend a bit more time separating the "locals" from those new to the environment. I know if someone walked up to me and asked that very same question at a coffee shop, I'd be significantly softer/respectful in my response versus a more blunt data dump response to a forum member who's posts I have read and with whom I am somewhat familiar.

This place is a bit like a locker room with all the good and bad that goes along with it. It is a great place to share info and beat ideas/theories to a pulp. Regretfully, newbies can sometimes get caught in the thrashing.

 

[luvb2b]

I was hoping to get some advice from this forum about a potential Model S purchase. I am considering putting down a reserve as a Christmas gift for my husband, but then I came across the information about the fires and the NHTSA investigation. Now I'm not sure it is such a great idea. I read through most of the information on this board and a few others, but it seems to be pretty contradictory. The question is, should I put a reserve down now or wait to see what the NHTSA has to say? I mean, we wouldn't have the care for a little bit anyway, but is it enough time for this all to be wrapped up or should I wait a bit? Any advice is appreciated from those of you who understand the statistics better than I do.

hello. as the one who started this thread and a model s owner myself i hope i can offer some sensible insights.

my opinion is that the statistics to date show two things: 1. the model s is much less likely to have a fire than the average ice car on the road and 2. the model s is more likely to have a fire that results from a collision. although i have no formal work to show it, the trend should be strongly toward model s being safer than the average car in terms of injuries and fatalities of occupants.

accordingly, a letter recently made public by the nhtsa shows the specific focus of their inquiry: http://www-odi.nhtsa.dot.gov/acms/cs/jaxrs/download/doc/UCM446999/INIM-PE13037-58212.pdf

"ODI has received information on two incidents on deformations/intrusion into the propulsion battery caused by impact with roadway debris and resulting in thermal reaction and fire in 2013 Tesla Model S vehicles."

the statistics would have predicted such an inquiry, focused on fire related behavior after an impact.

the question of should you reserve now or wait... for me i would reserve now, because every day driving the s is a joy, and because the data supports it being overall one of the safest cars on the road. every day your loved one is in an s he is better protected than they are in another vehicle, and that's valuable peace of mind.

how i deal with the collision fire risk: when someone drives my car, i tell them they are in one of the safest modern vehicles, but if there is an accident just get out and slowly move away - just in case there is a fire.

regarding time and the nhtsa investigation, as of now their investigation is not done. it may find something about the battery pack that tesla must address. or it may not. i am sure if a service recall was ever required it would be done properly by tesla. a full investigation might take 6 months. if you wait that long you may have to wait another two three months for the car. that's eight to nine months of wait time.

eight to nine months of going to gas stations. eight to nine months of driving a not as fun and not as cool and not as safe gas guzzler. eight to nine months of life that cannot be rewound.

 

[brianman]

This place is a bit like a locker room with all the good and bad that goes along with it. It is a great place to share info and beat ideas/theories to a pulp. Regretfully, newbies can sometimes get caught in the thrashing.

I like this analogy. Continuing it...

The media is like the reporter that gleefully chides, "Brady says he can throw a football 600 yards." Brady then has to decide whether it's worth his time or effort to clarify, "This was in a conversation where we were talking about football on the moon."

 

[andrewt3000]

odds

luvb2b, it has been over 2 months since your stats were done. We now have a figure for Q4 2013 sales, and no new fires. Have the odds changed?

 

[kenliles]

luvb2b, it has been over 2 months since your stats were done. We now have a figure for Q4 2013 sales, and no new fires. Have the odds changed?

Post #129, pg 13 from luvb2b
"i'll say the models predict you'll see another collision-fire before the end of 2014q1."

 

[pz1975]

Will TSLA hit $280 by end of March?

Agree. I think the next fire will cause a quick dip followed by a quick rebound. I do think the first time someone dies in a Tesla (unless it is some crazy accident where there would have been no chance of survival like going off a cliff or a gasoline tanker truck exploding) will test the public's perception of its safety because the media will sensationalize it. Tesla cars are obviously being held to a much higher standard than other cars. I don't mean to sound morbid but it WILL happen someday. I hope the public will quickly see how many fewer deaths/serious injuries there are with Tesla vs. other cars.

 

[Raffy.Roma]

I do think the first time someone dies in a Tesla (unless it is some crazy accident where there would have been no chance of survival like going off a cliff or a gasoline tanker truck exploding) will test the public's perception of its safety because the media will sensationalize it. Tesla cars are obviously being held to a much higher standard than other cars. I don't mean to sound morbid but it WILL happen someday. I hope the public will quickly see how many fewer deaths/serious injuries there are with Tesla vs. other cars.

IMO this could happen only if somebody inside the cabin will not fasten the seat belt in a high speed accident. IMO this could also never happen.

 

[pz1975]

Never? If Tesla does what they plan to do (500K cars per year by 2018/2020), there will eventually be MANY fatalities (but still much less than other cars) in their cars - that is indisputably inevitable.

 

[Raffy.Roma]

Never? If Tesla does what they plan to do (500K cars per year by 2018/2020), there will eventually be MANY fatalities (but still much less than other cars) in their cars - that is indisputably inevitable.

IMO the Model S (for instance) not having an engine in the frunk like ICE cars is much less vulnerable to fatalities with respect to ICE cars. Then we should also consider that Tesla is developing the pedestrians and obstacles detection system which hopefully will be integrated with the ACC. All this means a very high level standard of safety. I guess that a fatality is very difficult to happen on a Tesla car.
Remember this post my friend pz1975 and talk to you on TMC again in 5 years. We"ll see if I am right or not.

 

[pz1975]

I hope you are right for many reasons!

 

[Mario Kadastik]

IMO the Model S (for instance) not having an engine in the frunk like ICE cars is much less vulnerable to fatalities with respect to ICE cars. Then we should also consider that Tesla is developing the pedestrians and obstacles detection system which hopefully will be integrated with the ACC. All this means a very high level standard of safety. I guess that a fatality is very difficult to happen on a Tesla car.
Remember this post my friend pz1975 and talk to you on TMC again in 5 years. We"ll see if I am right or not.

Object through windshield. Either a piece of lumber or what not goods are transported in the car in front. Instant puncture and death, nothing to do about it and not worse or better than other cars in this regard. **** will happen eventually and if I remember right about a month or two ago I did the math and found that we were at just about 1-2 deaths predicted by statistics (hence not observing one isn't significant yet). Of course as that was a while ago the number of Model S's has grown and miles driven has grown. So we might very well be in the 3+ deaths region where things start to somewhat get significant now that we've not observed anything. We'll truly start to think it interesting and can start to make estimates when we get to 5-6 expected and none observed. But if one were to happen now it would have no meaning what so ever...

 

[Raffy.Roma]

@Mario

IMO a fatality could happen on the Model S only in a high speed accident where somebody inside the cabin hadn't fastened the seat belt (and in case the assistant driving package will not already have prevented the accident).

IMO the Model S, because of her features, can overtake the statistics. Then, when the obstacles and pedestrians detection systems will be implemented and integrated with the ACC (assistant driving package), the Model S will have very high standards of safety. As I said to pz1975 talk to you in five years from now on this subject and compare the statistics with reality.

 

[pz1975]

But there won't be statistics to compare. Your argument is that there could be ZERO fatalities associated with driving a Tesla car. Thus, there will either be 0 or >0. I think the odds of >0 by 2020 is 99.99999% considering the number of cars that will be on the road by then x the number of km driven per year.

 

[Raffy.Roma]

But there won't be statistics to compare.

No Mario wrote his post on the ground of a statistic. In fact he said that by now already some fatalities should have happened on the model S (which luckily didn't happen). There are statistics to be used of course on the ground of miles driven and so on.

 

[pz1975]

I know but I meant for our discussion (0 or >0) there really aren't any stats needed since it is yes or no. Anyways, like I said I do hope you are right!

 

[AlMc]

Maybe we need mods to move some of this to new thread 'Odds of any accident or human injury'

 

[Raffy.Roma]

Maybe we need mods to move some of this to new thread 'Odds of any accident or human injury'

Agree Al. I think this is a good subject to be discussed. Also in view of the importance of the assistant driving package (that Tesla is developing) on this matter and how it will affect the safety standards of future Tesla cars.

 

[Mario Kadastik]

No Mario wrote his post on the ground of a statistic. In fact he said that by now already some fatalities should have happened on the model S (which luckily didn't happen). There are statistics to be used of course on the ground of miles driven and so on.

No, you misunderstood. I said that based on a huge statistical sample (all driven miles in US and the deaths associated to it) the expected amount of deaths in a Model S based on total driven miles is around 1-2 deaths. However the fact that none have happened is not significant because while the mean expectation is following a gaussian distribution, the observation (or lack of it) follows a poisson distribution. The probability of 0 when expecting 1-2 is quite high. So is observing 3. So right now we have no statistical handle of wether the Model S is safer or not. We all know intrinsically based on other evidence of injuries etc, but statistically we cannot claim one or the other. Now, when we get to 5+ expected deaths and we STILL observe 0. Then we have reached a region where we might start to think this is significant. However as the car seats 5-7 persons the question of what's observable granularity kind of pops up as it's not impossible for a full car to drive off a cliff and get everyone killed immediately rising the deaths to 7. So it's a bit tougher even than the whole statistics discussion of the fires was and in fact the probability from the massive statistics is far smaller for deaths so observing or not observing a death right now would have no real consequence. However the longer we go without, the more statistically significant it will become, but I think we need to wait billions of miles before we can tout to everyone that Model S is by far the safest car with regards to fatalities. Until then, let's keep the fingers crossed for avoiding flukes.

I'm talking about accidents like this where there is little the Model S could do better than its peers: Chinese Bus Driver Narrowly Escapes Death After Giant Pole Crashes Thru His Windshield! - YouTube (disclaimer, the driver survives by the skin of his teeth)

 

[Raffy.Roma]

@Mario

Ok I agree that in order to have a reliable statistic we have to wait for more miles driven. IMO main point of the discussion will be differentiating between old Tesla cars not having the driving assistant package installed experiencing fatalities (if any of course) and new Tesla cars having the driving assistant package installed and experiencing fatalities (if any again).

[To moderator: As Al asked please move some of these posts to new thread 'Odds of any accident or human injury']