AAA conflict of interest on "EV Cold Weather study"- study not credible

Discussion in 'General' started by 101101, Feb 15, 2019.

  1. 101101

    101101 Active Member

    This is so damn obvious. Tesla as it succeeds with electric autonomy will put AAA auto insurance out of business and be doing a huge public service in the process. So this crap aspect of the otherwise respectable AAA business tries to tell lies about Tesla with fibs about cold weather- or the issue is with BMW (despite coming from Germany its main EV has not been that popular in Northern climes but Teslas are very popular in icy climes) or GM EVs are having a problem and they are trying to conflate for cover these two laggards- yes presumably these issues but they are probably trying to pitch a critical advantage EVs have over ICE vehicles as a flaws (Goebbles style) probably because EVs are still running in the climate change caused cold disaster when tons of ICE cars won't start because they don't have a block heater. And while EVs are only getting better and better cold performance the same isn't true of ICE.

    You have to keep in mind Tesla has piloted in Asia turn key sales where you buy the car and the insurance and maintenance is included in the sticker price and has said if the insurance industry didn't clean up its act on Tesla where Tesla vehicles should have much lower rates (Tesla having all the safest vehicles on crashes plus autonomy accident avoidance 6.67x safer at last report and going to 10x) Tesla would self insure setting the required precedent on this stupid premium driven industry that is always making up excuses to pass along the losses on its gambling habit instead of running at cost even where it fraudulently claims to a mutual company or be a not for profit and where it presumably through bribery and other mechanisms is caught up in intentional bad investing fossil fuels to help subsidize and underwrite the endless wholly foreseeable austerity driving losses of the perpetually bankrupt (secret energy policy anyone?) with ever increasing intensity fossil fuel industry Ponzi scheme.

    AAA is probably particularly galled by AP3 impending announcements. You see the connected FUD where thee are stories accusing a government agency of lying (not the SEC this time) about accident reduction by Tesla autonomy- on the basis of one hired shill apparently intentionally misinterpreting the data?

    Both stories should be investigated and the one paying for the lies held accountable at the very least by linking to past and future stories to discredit and cut down on the paid lying. Pravdah should be well funded.
  2. Agree in concept. My PHEV by Kia has a seat heating(and cooling) option, a heated steering wheel is available, a setting for only the driver when cabin heat is desired. I can only assume the Hundi and Kia EV's are the same.
    So simply with a review of the current technology likely many of these causes of low range in cold are remediated. Cabin heat is it really used and necessary.
    Are other manufacturers behind in this, suspect not. Can we have cabin heat necessary, sure kids and all. But really is that your average drive to work? Most no.
    Curiously the video presenting when I prompted up the story, in addition to the story, a video about things EV, but things that seem to be as they were 2017. Basically selling the Gm Product as I read it.
    Nice vehicle but we are 2 years beyond 2017. things are evolving in this area dramatically.
    20 degrees is the same temp that the EPA uses in their statements on the 23% percent decline in short use ICE vehicles in cold.
    Never ever a mention of that.
  3. bwilson4web

    bwilson4web Well-Known Member Subscriber

    First read the report and then criticize. I read the report and found no technical problems.

    Bob Wilson
    DaleL likes this.
  4. No problems huh? They tested a total of five vehicles narrowed down from a initial ten. Why the five, not based on sales or use but so as not to show disproportion to any one manufacturer. They state that I can cut and paste if needed.
    Seems if I were doing a study I would see what is selling and then study that per portion. If I wanted a result that showed a thing the actual buying public may find relevant. And five whole vehicles? Where are the results of newer models? Some just came out two I know of, with 200 mile range. Five vehicles. If this was medical science I would be laughed out of the room. Such a low number and a arbitrary base of choice...did they know those five had problems and those were then the ones the chose....I would ask exactly that. Who knows. Seems strange and why then the five specifically? 100 miles sure, plenty more than five fill that.
    It is available on the AAA site. Not hard to find.
    Medical science has a peer review publication process for study qualification. Many trades of engineering and design have equal. Where is this qualification? Are they their own source of qualification on this science, AAS By what authority. Yes they used another group for conduction of the study, that group is now by participation not available for qualification. This study stinks from top to bottom with potential of bias..

    This is 2019 and it was just released, why then are two of the vehicles 2017 and were the battery functions compromised by sitting around dealer lots for two plus years uncharged and not in use? Or were they in use and then what was the use and how degraded are the batteries tested? They say there were new how that then?
    No a study, any good one, has qualification publication, and involves hundreds or even thousands, to reduce error. And primarily it is preferable they be conducted by third party independent groups, not affiliates with controls to assign validity .
    This study is so faulted with potential for error misuse intentional or nonintentional, possible mischaracterization of result, it is beyond bad, this is simply comically bad. No one with a basis in scientific study would accept these findings. They may be valid or not, truth is with such a small disproportionate arbitrary study choice, conflict of interest with the studiers and those reporting results being functionally the same, we cannot ever know. It is worthless.
    A PDF file on their site is found by clicking the public relations person statement on the study. The word they use is "here" in the article. That will prompt up the actual study.
    Please do review it as my proof of it being total nonsense is but a click away. Certainly don't take my word on it.
    Last edited: Feb 15, 2019
  5. bwilson4web

    bwilson4web Well-Known Member Subscriber

    Since I own a BMW i3, one of the five tested, I don't see a problem. The technical details in the report match my experience. Just now I checked and the EPA listed only 24 pure EVs in 2019 so they've sampled about 20% of the available models.
    This peer's opinion is based upon the report and personal experience with a BMW i3.
    If you'd read the report you'd have noticed the BMW i3 performed worse. Being a BMW i3 owner, I have no problem with their results. FYI, I also have a 2017 Prius Prime and though not included as a test article, personal experience shows it has problems in cold weather.
    I am perfectly OK if someone else wants to use a different protocol to replicate or challenge the results. Just this is the type of study I've seen similar data from:
    When I have to pay "Yankee Greenback Dollars" to run my cars, I'm pretty interested in studies about how to keep my dollars in my wallet. So between my informal studies and shared experiences in BMW i3 and Prius Prime forums, we figured out a lot of what this study has already confirmed. Of course, your mileage may vary.

    Bob Wilson
    Last edited: Feb 15, 2019
    DaleL likes this.
  6. Bob no offense this is interesting stuff and stuff you find particularly interesting for reasons stated but this is not science nor conducted in a scientific manner. If this was sponsored and published in a other grouping, some sort of applicable group who does publish scientific analysis of cars, with perhaps a speciality in EV, in this context, it would add to applicable use as anything other than interesting and perhaps fun things to know and think about.

    You finding it personally usefull, well good. Is it science, no. A sample must simply have more than five cars and they must be equal cars not ones from this year or that. And choice of cars if not all, which is best, must be on some basis other than what they stated. No new cars which came out this year are included. Impossible, that how could they....fair criticism, then hold up on the result until they can be tested. If it is wanted to appear a serious statement of fact. If not it is what it is, a interesting thing AAA printed with the polar vortex as the bite to incite viewership.
    Science not at all. They used scientific means and instrumentation, the actual conduct of the selection,who performed the testing and severe low numbers of vehicles used simply prohibit it for valid scientific conclusion.
    On would not in similar fashion feel it necessary to produce by scientific study a grouping of people in a medical context which had no control in study and insufficient number for cause to be found. It would unplublished be and be relegated to use as interesting with no import more than that.
    This is that with car as study. Five cars from different years.....who can say what they will show? Why them 2 from 17, makes no sense. If fair equal study result is wanted to be found. at the very least the same year would have been used.
    We all already know cold affects batteries. We also already know cold affects ICE mileage which is range. So what else is new?
    This is worthless.
    Last edited: Feb 15, 2019
  7. bwilson4web

    bwilson4web Well-Known Member Subscriber

    Actually I would call it an engineering study but close enough to be useful like the EPA and associated CARB studies. Since I'd already been working the problem with my BMW i3-REx and a Prius Prime, in one respect 'confirmation bias' applies. Regardless, I see it as useful insights for those without my personal experience.

    Bob Wilson
  8. Agree it is useful and interesting for you. What application to one who owns one of the newer vehicles like a kia, or hundi, zero.
    And you know being in the field, the engineering field has multiple publication source of scientific study subject to peer review, which may be endeavored if what one wants to do is provide science. You do not self publish engineering study unless you have great authority. AAA is not in any manner a authority on cars nevertheless EV technology. They may use a dynamometer in their tests that means virtually nothing if they know not how to conduct scientific study. Nothing supposes they hold this qualification. Not just anyone can do study. They must first hold qualification and then state methodology for purposes of publication. All of which must be studied, and that is just to start.
    They got five cars put them on a dyno and varied the use and temps and self published the results. So what? Anyone can do that it is not science nor valid scientific result from study. It is sort of interesting nothing beyond that.

    A article. This would serve as a good basis any good reporter may do in what back in the day was popular mechanics. A interesting fun thing to read on a very cold night in a cold winter.
    But popular mechanics would not misrepresent it as science or any sort of definitive report at all.
    A interesting fun thing to read. Thing like that have some application to readers sometimes specifically.
    This nonsense here is being used on CNBC to discredit future EV sales and sales specific to tesla. That is flat plain and simple a abomination.
    That is as equal to taking a popular mechanics article for the same result. Only in this topsy turvey world of present America can we see such things.
    Perposterous nonsense to quote or use it in this fashion. Which is exactly as some in the industry are applying it.
    AAA the media person actually qualifies it as I characterize it. We must do certain things to get best battery use, that is key to her article, a panel with five recommendations is found, for this main purpose.
    I agree with her recommendations as all here will and they do serve purpose that, being valid. The headline the 40% thing total fox news type reporting. Better left behind that nonsense.
    Study five car results from different years. how about using a already existing model, consumer reports, simply quary owners, loose battery in cold, yes no. how much? Why not that so easy to do?
    Easily found that result and thousands are replying so a good scientific base is present to draw valid conclusion...not this nonsense.
    You know those 17 cars were left overs, how could they not be. New yet 17 had to be. How much battery degredation occurs to a EV sitting in the hot sun for 2 years, southern cal, with no charge, ( I know dealers they don't charge EV's sitting on the lot) may be the better subject of their study result. What affect does that pose to battery charge in the cold.

    Why not a simple consumer reports survey.Why not that methodology employed? Makes no sense.
    Last edited: Feb 15, 2019
  9. bwilson4web

    bwilson4web Well-Known Member Subscriber

    DaleL likes this.
  10. DaleL

    DaleL Active Member

    A quick Google search finds many posts by Tesla owners concerning winter driving. All of them relate to the shorter range that cold temperatures cause. Some of the accounts dating back to 2015 are rather interesting reads. They include tales of driving at 55 mph, with no cabin heat, to extend the range enough to get to the next charging station. The AAA report appears to be based on sound science.
    bwilson4web likes this.
  11. Pushmi-Pullyu

    Pushmi-Pullyu Well-Known Member

    They can only compare cars which are actually available for the study. That's like Consumer Reports, which rates only cars it actually buys on the open market, so it can be sure the manufacturer isn't giving them a special-built car.

    AAA doesn't have unlimited funding, and it's hardly a surprise that they can't buy a lot of cars just for this one study.

    I've been skeptical of AAA's attitude towards Tesla ever since they reported that the Model S should have more expensive rates than comparable cars. But given the higher body/ collision repair costs for Tesla cars, that's probably fair.

    If you, Ron, or 101101 think AAA's study is biased, then cite what exactly you think is biased about it. Otherwise, it appears that the bias is in the mind of the reader... not in the study. 101101 sees conspiracies everywhere, so his claim of bias here, without citing any actual example of bias showing, is of zero informational/ logical value.

    bwilson4web and DaleL like this.
  12. That a thing produces a result we may find or think correct does not mean good science is produced. Good scientific study requires some certain things which are present to prevent personal bias, untoward association(two things happening at the same time due to a similar cause does not admit those two are related) and result not formed by material study but by inadvertent other factor(technical fault or break down).
    The study of five cars response to cold, some in one year and some in another, speaks of something, but it is not a finding that can be considered in any manner a proven. I ventured one guess, it is a guess as we do not know. The cars, two of them are new yet 2017's. They sat around a dealers lot for one year or two most probably uncharged. That is known to degrade batteries. The question is not that cold degrades batteries,(we know that already) but is that a solid read on percentage or is it a read on percentage to a already damaged battery?
    You simply cannot design a study with so few vehicles as source for data. If they did not want to purchase cars lease or whatever, I suggested consumer reports methodology would have worked far better and produced likely pretty good results. You simply provide a mailing to all the new EV owners with exposure to cold by geographical area. Of the several thousand in your base(provided by manufacturers records) many will respond. You quary them on this specific, do you notice loss of range in cold weather(specify temps) how much is lost. You collect the data and study, then produce findings on that. Those are then subject for review to peer publication source. The publisher reviews the data methodology used and determines if this is worthy of publication by peer review, experts in the field with qualification . This is how this works in the field of study. You cannot just self invent study and publish unless you are a recognized authority in the field. EPA for one and others in the industry, may self invent and study without any publication mechanism as they are a recognized authority with scientists and specialists which attests to qualification. AAA is not that, not in the slightest.

    This is completely preposterous to propose this as valid study. Five cars, any single one has the slightest of technical flaws and your entire base of data is 20% off. What are the chances you get five EV cars that for some reason or other have not sold without some sort of flaw that likely made that a reason that specific vehicle was not purchased?
    Think a dealer is going to give a car to AAA the best on their lot for study...of course not. How were the cars utilized. Did AAA buy them covertly, did they borrow the cars and on and on. Did AAA contact the dealers say, could you supply us some cars for study and go from there....that is most likely. Dealers manufacturers may do such things to court favor with media. AAA is media.
    Five cars is simply to small to draw upon and I can think of any number of potential compromises to data on the basis of only five cars used. Consumer reports uses survey to determine reliability for very good reasons. It is the only means that really works. Battery loss in cold is certainly the same purview. Hundreds or thousands need to provide data not five.

    In a review of scientific study the crux or burden of proof is not on others to show why a study may be biased but on the initiator to show means why bias is not present. The methods used must segregate bias from the study from those who conduct it. The methods are analyzed for this character. Controls double blind I mention two but there are various means to enable the removal of bias.
    If no means are present with the methods used to prevent bias, from presenting in the results, bias is assumed and the study is not published.
    No smoking gun is necessary. A study environment is produced in which a gun shot does not occur. Without that a gun shot is assumed with consequent smoke.
    In light of the fact car reliability concerns(certainly this fits that) all survey to make these sort of determinations(consumer reports JD Powers and others). A smoking gun, if I found one necessary, lies exactly there....why were not existing means of survey, the entire industry already, (to include the manufacturers themselves)_, uses to determine reliability of cars(they study each and every area top to bottom on the cars) not used but this highly suspect means of five cars and a dyno.
    Last edited: Feb 16, 2019
  13. That we have a finding produced which mentions a thing we may already know and suspect, in no manner means this is good scientific study. Percentages are put to this degrading with cold, and we have no way of knowing by this study if they are valid or not. Only five cars were used and they were not even from the same model years, a mix of 17 and 18. And this includes none of the newer models nor newer revamps of existing lines, three new models of which I personally know of have just been released.
    With such a variety of years how can we say we even know the past history of the cars in the study? Many practices are known to degrade batteries such as sitting uncharged for a year or two. Do we know that did not occur with these?
    A already degraded battery of course will show more percentage loss with cold. Has to, as it has less functional mass to begin with. Functional mass remaining simply is lesser by proportion, and thus any external factor will more by percent affect it.
  14. DaleL

    DaleL Active Member

    I know for a fact that when the temperature is in the mid-70s to low 90s, my Honda Clarity will get 55 to 65 miles on a charge. When the temperature is in the 40s and I am using the cabin heater, I'm lucky to get 40 miles on a charge. I would like to read "ron in new mexico" and "101101" personal experience with their EV range in cold weather. I find it rather absurd to criticize a study without evidence or experience.

    Electric vehicles are very efficient; they do not produce gobs of waste heat as do ICE vehicles. The internet has many accounts of EV range reduction due to cold weather. Thus the AAA study results should not be a surprise to anyone.
  15. I own a PHEV not a EV. My personal experience on the electric part, miles available for use upon start after charge I am loosing 8% in temps that range from 15F to 0F. this is a consistant number found.
    WE must not all be personally involved in a thing or even have qualification in the field of study being presented. Scientific study has certain things which are universal to its application that qualify one study as good and valid and another not at all. The methodology employed varies in the specific depending upon the issue regarded but the principals of sound science remain constant. A personal finding and a study which produces a similar result simply in science bares no weight.
    The field of science uses publication as the means of qualification usually. IN this specific auto industry journals engineering journals and various other sort are contacted. The study is outlined for review by the publisher. The publisher has a field of experts who look at the methodology and determine its validity. When it is verified for publication the authors then conduct the study. The results are then submitted for another qualification of results. Upon successful qualification the scientific study is then published and used by those in the field as reference considered a proven and as well used as starting point for other study in the future.

    These is one exception to this mechanism, the group providing study such as the EPA,, already has qualified experts on staff who conduct and verify study on their own. They may do so as they are expert enough to do so.
    This is not the AAA. To contend that as fact would be in the field quite comical.
  16. This is from their site, the US DOE's comment on the issue. The US DOE is of course well qualified to produce their own results of study and to self publish or internally provide validation on others study results. They then use study results to provide advice on energy use.
    Fuel economy tests show that, in short-trip city driving, a conventional gasoline car's gas mileage is about 12% lower at 20°F than it would be at 77°F. It can drop as much as 22% for very short trips (3 to 4 miles).
    The effect on hybrids is worse. Their fuel economy can drop about 31% to 34% under these conditions."

    As far as I know they have not officially made a statement on EV loss due to cold. If someone has data and/or statement by the DOE on this specific I would be glad to see it and suggest those numbers if found would likely be quite valid. These AAA numbers are completely arbitrary and not the result of proven study found.

    This listed is not at all new data I think it was 2014 or such(could check).
    Now the question becomes, do the newer batteries suffer in the same manner? KIA and Hundi I know have just produced three EV vehicles in the 200 plus range.Will they degrade as the earlier battery technology did?
    And Prius is using not a lithium but the earlier technology, nickel in their new AWD. I assume that is as AWD will be subject to more cold. Prius are all other wise using lithium. But other model lines Toyota, nickel can still be found(it is cheaper). Nickel based in very general form are considered less affected by cold but they produce less storage capacity per equal mass.

    WE would be all better off if we had new data on these, as new car purchase is a large concern. Used not as much.
    Is my experience of a phev loss of 8% at 0F with a new Kia now a new norm, or is my car a outlier? We simply do not know.
    Do not kid your self that these newer models will not be large sellers in the EV market globally to a lesser extend perhaps in the US. High mid 200's range and price point around 40K USD. And hybrid wise AWD prius is likely to be quite popular.

    We could conduct our own informal study of a sort here as well(but not to represent that as science as perhaps the AAA report is being).
    How about you who own the new KIA Hundi EV's, what are you finding on loss of range due to cold? PHEV may not apply. You may not hit 20F consistantly in calif but I bet it does get 30F consistantly most places. That would have application as loss would occur at 30 as well. How much?
    Last edited: Feb 16, 2019
  17. bwilson4web

    bwilson4web Well-Known Member Subscriber

    A favorite technique is to invent facts not in evidence and failure to site sources. For example:
    The paper states:

    All vehicles were procured and inspected to determine suitability for dynamometer testing according to the following criteria originating from SAE J1634 [5]:

    i. Check that the odometer reading is between 1,000 and 6,200 miles
    ii. Verify that the battery ampere-hour capacity is within acceptable limits before and after testing

    In addition, it was verified that no warning lights were illuminated at any point before or during testing.
    There is nothing in paper describing the state of the vehicles before procurement. Citing facts not in evidence and without a credible reference lets the poster pretend unearned credibility. You may want to revise and correct your imagined claim.

    If someone had actually studied statistics, they would not make this claim:
    There is a specific term called "confidence level", a rigorous math technique for those doing a population study. But this paper does not claim to be a population study. It is a survey of how five specific examples performed. Another misrepresentation of the paper and its scope.

    What truly suggests perfidy:
    Consumer Reports tests single examples, not a population. This false claim is called a double standard, talking out of both sides of one's mouth.

    I've grown impatient with dealing with a purely trollish post and these words are something we can agree on:
    I've got to turn up the squelch: PLOINK.

    Bob Wilson
  18. DaleL

    DaleL Active Member

    Science is a method to obtain truth (answer a question). Anyone can do it; the scientific method does not require a PhD. What it requires is a systematic approach to answer a question. It also requires an open disclosure of the materials and methods used. The AAA study was done in a systematic and open manner. If any manufacturer (or any person for that matter) of the cars tested wishes to dispute the results, they should to do their own study and publish the result.

    It is well documented that Lithium Ion batteries perform best at higher temperatures and last longer at lower temperatures. The batteries in cars, which have active temperature management (Tesla), degrade much less with time and mileage than those that don't (Nissan Leaf). However, the AAA study documents how much cold weather affects not just the battery capacity, but also the resulting total range of the vehicles tested. Cold weather driving, which does not start in a warm garage or with the cabin pre-heated, requires energy from the battery to just warm up the battery and the cabin before a single mile is driven. In the test routine that AAA used, it resulted in a typical 40% reduction in range.
    bwilson4web likes this.
  19. No sorry this is the discussion of scientific study which this is supposed to be not the study of science in general. One does not in science just go about answering questions. Study is the means to find out the proof of hypothesis. We find a thing that is observed, in this, our car is going less in cold, study it to find to what degree and how far that extends in the group of all EV cars.
    Openness does not suffice if the base used is so small any technological fault in the study group compromises result. Which is why study requires always a large base if it is to be considered to be valid and useable with findings that apply.
    One of these five were technology faulted in some manner, perhaps they were on a dealers lot for a year or two uncharged, just one, and the results are all 20% off and our 40% becomes perhaps a 32%.
    And as already discussed not one single 19 model was used despite this being published well along into 19. How applicable is this science?
  20. The onus is not on me the reviewer of a study to show or prove things did not happen but on the initiators of the study to show how means were endeavored/methodology to prevent corrupted results.

    Who did the testing and what means were employed in the testing to prove validity. Were those who did the testing without bias perhaps internal to result? Were they screened and then were means employed to prevent unconscious bias by provision of function without any relationship to result found? Were a group of like cars of the same similar sort studied as control to find if temperature was in fact not the issue, but perhaps the sequence of testing. AS per how long was the various testing sequences employed one in relation to another. Car warning lights and all are not designed nor intended to provide result in scientific study..
    Were other means of testing of the cars used to verify proper operation?
    I am just starting really, this all just elaborates why someone at AAA or some random engineer does not conduct scientific study. They have no idea what is required to produce publishable finding.

    Acceptable limits..comical that. What were the limits exactly and who determined what were acceptable limits? Their qualifications? Were they controlled or subject to review for bias or were some means employed to blind those who read the results from the study purpose itself?
    If valid results were desired why then the gross variance not only in model years but in millage as well.
    The more revealed to study is this piece(I would be hard pressed to call it a study at all) the more holes are found present in it.

    This is a pathetic excuse for scientific study. With so much variance in mileage did they even attempt to discover if like use was endeavored with each of the five cars used? A degrading of battery is a known with age. And here we have five vehicles with markedly different use patterns 1K to 6K being in study described as equal. Show me any car battery that will not suffer degrading with 5K more miles on it and I will show two equally for study purposes vehicles.
    No variance is present 5K more miles on any battery in a EV. Degree does depend upon brand and quality of manufacture.
    And say the 1K vehicle (we just do not know) is the 17 model, what says that about any results found? Where was it stored at what temp and how often was it charged?
    These are questions I have not to answer or imagine but things good study by design provide answers to before any conduction of a study even begins.
    Where are these answers and a hundred others poised that good study provides in design and methodology employed?
    Last edited: Feb 16, 2019

Share This Page