As a general rule, most catastrophic failures in a car offer some kind of warning, and most are at least somewhat avoidable if you take great care of your car. But for BMW i3 owners and owners of other BMW electric cars, there is a failure that can completely destroy their car, and not only is there sometimes no warning, but there’s absolutely nothing anyone can do to avoid it. It’s referred to by some as “The Black Death,” and it is the result of a boneheaded engineering decision.
The BMW i3 is an absolute marvel of engineering. With a carbon fiber body structure, an aluminum skateboard, a rear-mounted motorcycle engine, plastic body panels, and an interior made of all sorts of gorgeous and sustainable materials, the car still to this day feels like the future. But even an engineering masterpiece has its flaws, and in the case of the BMW i3, the three major flaws are the tires (which wear quickly), the battery longevity (which is low on early models), and the AC compressor.
Believe it or not, of those three, the AC compressor issue is by far the very worst, and it’s shared with other BMWs, too. It is a nightmare that has forced many BMW owners to say goodbye to their vehicles. Yes, failure of a simple air conditioning system component — a compressor — is enough to total a BMW i3 and other BMWs.
The Batteries Are Cooled Directly By Refrigerant, Not Indirectly By Antifreeze
To understand why the i3’s AC compressor is such a problem, we need to understand how the BMW’s battery cooling system works, and why it’s different than that of its competitors. (Note that this applies to other BMWs, too, like the i8 and 330e).
Stepping back even further, let’s establish that lithium-ion batteries like to be at about the same temperature as humans. Generally, they want to be cooler than 50 degrees centigrade (122F) and warmer than about zero (32F) — ideal temp is between 15C and 30C, per Mahle.
This isn’t easy when the outside temperature is close to 50C in, say, Arizona or California’s Death Valley. In order to create a larger deltaT (difference in temperature) between the ambient air and the battery temperature, automakers typically use what are known in the industry as “chillers.” These are generally “stacked plate-style” heat exchangers with four ports: refrigerant in and out, and coolant in and out.
The purpose of the chiller is to use the AC system to “chill” ethylene glycol-based coolant/antifreeze below ambient temperature so that the cold liquid coolant can be used to cool power electronics and, critically, the battery.
The BMW i3 and i8’s cooling systems are a little different than those of typical EVs in that the i3 and i8 each have a refrigerant loop that doesn’t cool the battery indirectly, but rather directly, like so:
Most other EVs put the battery cooling circuit in a separate liquid coolant loop that interacts with the refrigerant loop via the aforementioned chiller heat exchanger.
Here’s an example of a typical EV cooling system in which the battery is isolated from all other components except for the high voltage battery heater, the chiller, some valves, and the low-temperature radiator. (You’ll notice that the power electronics are on a separate loop that is passively cooled via a low-temp radiator out front):
My colleague Jason and I made a video in which we show the differences in the cooling circuits of the BMW i3, Tesla Model S, and Chevy Bolt:
Here’s a screengrab of the i3’s cooling system:
The setup is fairly straightforward. The i3 has a giant battery pack packaged between the axles, as is typical for any EV. At the bottom of that battery pack, on the inside, are a number of aluminum extruded channels on which the battery modules sit:
Here you can see how the batteries sit on top of these channels:
This is a close-up of those channels; notice that below the channels are electric resistance heaters to warm the refrigerant, which then warms the batteries on top, as needed.
Here is the heating/cooling setup outside of the battery pack:
Here are solely the heating elements:
And here’s just the refrigerant cooling system, with the horizontal tubes running from the front to the back of the pack. This is what the battery modules sit on top of in order to pass off their heat to the refrigerant:
These flat microchannel extrusions are named because of how small their channels are; have a look:
BMW i3 Owners Have Had To Total Their Cars Because Of AC Failure
You can probably see where this is going; with battery cooling relying on refrigerant passing through extremely tiny channels, any amount of debris running through that refrigerant loop risks clogging those cooling channels, and thus affecting how well batteries can expel their heat. Ultimately, this could ruin the car. And on many occasions, it has:
These are all screenshots from Facebook pages, i3 Forums, and Reddit. On numerous occasions folks have been charged over $10,000 to fix the results of an AC compressor failure. Here are a few quotes from the aforementioned pages:
Here’s i3Alan from the forum “mybmwi3.com“:
Here’s Daniel Heerema from the BMW i3 USA Facebook page:
“$11,71218 to fix the AC! Compressor shredded into all the lines of high voltage battery! Can’t charge most of the day in Tx. 2014 – No Warranty
Here’s SmallestWheel from Reddit:
So sad that between the stupid 12v battery bricking the whole vehicle and now this, we have to stay ahead of problems that shouldn’t exist. I like my i3 but legit thinking about selling it only due to all this insane maintenance. Paying $4-5k every 4yrs to change an AC compressor before it breaks seems absolutely loco to me.
Here’s Appropriate-Mood-69 from Reddit:
The insane cost is due to the cleaning of the system from iron shavings after the compressor disintegrates. Before this happens, you could decide to replace it earlier. This would result in a much lower bill.
Here’s rontombot, also from Reddit:
The “shrapnel” plugs up the expansion valves, both for the battery and the cabin. The repair estimate I saw for one that had the worst-case failure was well over $20k, including every aircon component and HV battery Enclosure . The battery cooling tray is part of the battery enclosure, and I believe all of the control electronics as well.
I do know some have had BMW flush the system and the compressor replaced (to the tune of $7k), only to have another failure… then told it’s not repairable.
When I get my 2015 BEV drive motor and transmission replaced (DIY, due to failed motor bearings), I’m considering replacing the aircon compressor. (with the latest. version)… Justin Case… and only IF I decide to try to keep the money pit a while longer.
I also have a good 90Ah battery that I may install at the same time if I replace the compressor… but even that’s dependant upon the outcome of having the warranty test done on my existing 64Ah battery.
ugh… too many worries… which is why I’m getting more and more serious about a used Tm3-DM.
And these were just BMW i3 issues; the BMW i8’s battery cooling method is also direct-refrigerant, and this means similar issues, as the post “AC Compressor failures on the rise” on internet-forum Bimmerpost makes clear:
I’ve received five messages in the past month from i8 owners who have indicated that they hear strange sounds from under the front bonnet, and after I refer them to the shop to have the problem diagnosed, I’m receiving feedback that the Air Conditioning (AC) compressors are failing. In some cases, the failure was so catastrophic that the AC compressor had an internal failure, scattering metal fragments inside the refrigerant lines. I posted this on the BMW i8 Facebook group and 8 people replied saying it happened to them, one of which said he had metal scattered throughout the refrigerant lines and it cost him $8,000 USD to have it repaired.
Here’s another on Bimmerpost from Patsgarageonline:
Regarding costs, someone in my comment thread mentioned they were in Puerto Rico and the AC just failed when he was driving it, shortly after he bought it. No noise or no indication of failure. He was quoted $8k at first, then a few days later said it would be $18k. That’s the cost of a new low-option VW.
BMW’s decision to use direct battery cooling instead of indirect battery cooling means the cars’ most expensive component — the battery — has refrigerant running through it. This is not an issue, and in some ways is a benefit — unless that refrigerant is somehow contaminated. And that’s exactly what happens when an i3’s AC compressor fails.
It’s a failure that BMW Blog and others have dubbed the “Black Death” (it’s actually a commonly used term in the automotive AC industry, and refers to how a failed compressor can contaminate an entire AC system, though usually with less catastrophic results than in the case of these refrigerant-cooled EVs). I first saw the term when Robin Kuthe posted this on the BMW i3 USA Facebook page:
I’m sad to be leaving this group. I absolutely *loved* my 2016 BMW i3. There is a fatal flaw on this car. It’s the compressor issue. (Search this group for more info) Once the compressor goes, it blows shrapnel throughout the system. Up to $7k to repair. My car was out of warranty by only 4K miles.My independent shop would not work on it. They sent me to the dealership. After several days of consideration, I spoke to the service manager, who told me that his mechanics call this fatal flaw “i3 Black Death” and he sees this issue a few times a month. He also said when out of warranty…he has “never had anyone decide to pay for that repair “.There should be a class action lawsuit.Just wanted to make newbies aware of this issue.
And then I saw this comment from Rontombot on the Reddit post “Looking to purchase 2016 BMW i3 – 68k miles – Anything I should be concerned about?“:
- Air conditioner compressor – many of the pre-2019 year are failing, and it has a name used by BMW techs… “The i3 Black Death“. The compressor can self-destruct, grinding the scroll compressor to powder, then spewing it throughout both systems… cabin cooling, and HV battery… which most consider irreparable. (there were 6 part numbers for the compressor from 2013 to 2019… they changed that part 6 times, but not again since 2019)
The Failure Rate Isn’t Clear, And Some Have Had Better Luck Than Others
If you’re curious what the compressor looks like inside when it fails, here’s a teardown:
To be sure, in CARB states there’s a 15 year, 150,000 mile warranty on the compressor; plus there are apparently five different versions of the i3’s AC compressor, as the company made updates through the car’s production run. It’s unclear whether the compressor failure rate is lower on later models, and it’s unclear if the failure rate on older models is even as high as 2% after 10 years — we just don’t know. (For me, personally, this doesn’t bother me, and I still consider the BMW i3 a true marvel of engineering).
It’s also worth noting that some folks have been able to repair their bad compressors for cheaper; many say they did it for around $4,000, and some DIY’d it for even less (I’ve heard of some folks installing filters to try to prevent a compressor failure from contaminating the system). Presumably, these folks didn’t have as much catastrophic damage that sent metal into the battery. Another thing to note is that some owners have been able to identify a failing compressor early based on how the compressor sounds when the AC is on.
Still, far too many folks have been blindsided by The Black Death, and left with a car whose repair costs outweight the value of their car. And that’s just sad.
Update: It turns out this can be a problem with the bMW 330e as well. From Bimmerfest:
A couple of days ago my 2016 330e with 50,099 miles started making an odd noise while charging. It sounded like the normal fan/cooling system noise that I would hear and then started to make an odd noise, like a pump that was working a bit harder. It’s relatively warm in my garage here in Vegas, so I wasn’t too concerned. However, the next day I started hearing the weird pump noise while driving and my a/c started to not work sporadically. By the next day I was still hearing the noise and the ac was barely working. The noise seemed like it was coming from somewhere under the engine – my thought was ac compressor.
[…]
When I got it to the dealer they looked at it and (4 days after I dropped it off) they called me and said the compressor for the system that cools the battery was out and they had to replace that, which would run $3862. Also, if they found metal in the cooling system from the compressor, they would have to replace the entire battery cooling system – which would cost $10,600
Hopefully I can offer some insight, as I just completed this repair on my own 2014 i3 REX and have been successfully daily driving it in triple-digit Kansas summers with no abnormal battery temperatures, level 2 charging at home.
The official BMW procedure has changed at least twice. Originally, and what the huge $10k+ quotes would be for, BMW spec’d replacing the entire AC system. Dash out, new evaporator core, new expansion valves, condenser, battery cooling extrusion, compressor, and all of the hard and soft lines betwixt.
At some point this changed. These days the BMW TIS specifies the following:
Parts to replace:
A/C CompressorSuction and pressure lines attached to the A/C CompressorCondenserBoth expansion valvesAll o-rings and seals on replaced partsProcedure:
Evacuate A/C systemDisconnect HV battery at front and rear of car using service lock-outsRemove compressorRemove condenserUse specialty fittings with a BMW A/C flush machine to flush out the evaporator core, battery cooling extrusions, and run-between lines.Replace parts as normal, re-charge by weight, perform the “compressor run-in” procedure with ISTA+ or a scan tool.
I effectively did all of this with the exception of replacing the two refrigerant lines directly on the compressor. Lacking BMWs special tool, I used denatured alcohol in a compressed-air flushing tool ($40 on ebay). The important part here is to restrict the flow of the flushing liquid out of the battery. I’d recommend shoving a sponge or some breathable cloth into the battery inlet and flushing through the outlet. This slows the flow enough to make sure the battery gets filled instead of the liquid taking the path of least resistance and missing parts of the circuit.
I actually have even more insight than many, because I used a remanufactured compressor that failed after a short time due to a manufacturing defect (it started leaking out of the case halves), which I then swapped out for an OEM compressor later on. When I removed the reman compressor I took a look inside the suction side and found absolutely no sign of any “black death” having migrated through the system and into the new compressor after a month of it being installed.
ADDITIONAL NOTE:
If you have a 2018 or earlier i3 it’s far more likely to have A/C issues. A brand new OEM compressor is $1600-ish, and incredibly straightforward to install. Far easier than a gasoline car. Go to a shop, ask them to evacuate your A/C, then lift the rear of the car on jack stands, remove a single protective panel with some e-torx bolts, and the compressor is staring you in the face. Flip the battery disconnect switches, undo three mounting bolts, two refrigerant lines, and two electrical connections and down comes the old compressor. Stick the new one in with new o-rings, go back to the shop and have vacuum pulled and the system re-charged. The car is smart enough to disable the compressor if it senses no pressure in the lines.
Thank you for the insight!
That’s very useful, thanks. Bookmarking for reference; I really don’t want my 2018 i3 to kick the bucket.
This is definitely on my short list. I need to find out if it’s been replaced already.
Not sure if there’s a cheaper way to get it evacuated, but for the refill you can actually get the gauge manifold and vacuum pump via Autozone’s loan-a-tool program. The deposit’s pretty steep, but you get it back at the end.
Variable-displacement A/C compressors cannot be properly charged via gauge set. Have to be charged by weight.
Just need a kitchen scale for that, right?
Yeah, that’ll work. Vacuum pump, manifold, and a decent scale. I always forget that’s an option, which is funny because that’s exactly how I charged mine when I repaired the A/C lol
I still would like to own an i3 despite this potential flaw. As I’d like a later (‘big’ battery) i3 w/o the REX, there’s perhaps less chance of this being an issue. Of course, those later i3s tend to be the most expensive ones, so I’m not holding my breath about getting one.
David the engineer and his “engineering marvels” that somehow all seem to have catastrophic failure modes.
I’m curious – for all the chatter of ‘just put a filter on it’ that I’ve read (I have owned a 2014 BEV for 18 trouble-free months now), I have yet to come across anyone in the Facebook Groups/on Reddit//on forums that has actually done so. To me this suggests that it’s a good dose of armchair quarterbacking, but the filter installation is, so far as I can tell, still theoretical for the i3.
The shape and routing of the lines on the i3 isn’t really conducive to easily installing a filter. FWIW, Chevrolet/GM specifies installing a filter on the suction line of their cars any time a compressor is replaced, so it’s far from snake oil.
Compressor refrigerant filter.
Holy grail?
He chose…poorly. 😉
This Grail is not exactly a bowl of cherries, but apparently is more like a bucket of pits
I don’t know people that go off, looking for holy Grail seem to bear the burden of unintended consequences more often than not.
Just curious as to whether David or anyone else has an opinion about this:
If you have an i3, do you think it’s a better strategy to avoid using the AC altogether (when possible/humane) in order to avoid the compressor failure, OR would it be better to use the AC now and again to keep it exercised?
Just wondering if anyone’s got thoughts about this. TIA! 🙂
From what I gathered it turns the ac on to cool the battery as it charges, independent of the operator, so you’d also have to be careful about outside temperatures when charging.
I think it runs the compressor for battery cooling any time you’re in hot weather, whether the AC is on or not. I don’t think it matters.
Yes, that makes sense and I should have thought of that… thanks Mantis and Rust.
When will people just accept that BMWs are toys for the rich who can afford such ridiculous repair bills? Stop torturing yourselves!
(In the meantime, looking for a hawt 5 series wagon on the cheap…)
anyways recent GM automatic transmissions have plastic check balls in them that wear out, get smaller, and then get lodged into the valve body of the transmission requiring diassembly and a transmission rebuild as early as 60,000 miles.
Every modern car has some sort of achilles heel. Even the wunderking of reliability, Toyota, completely botched the 3rd-generation Prius and they ate head gaskets like candy. Guess they’re only toys for the rich.
…guess you didn’t read the last line…
I’m guessing you waited until after your i3 was sold to post this flaw lol
AFAIK, David’s i3 is STILL for sale/available (I don’t use Facebook, so I couldn’t see the link he posted below for his Facebook Marketplace ad at my request). Also, if you’ve been keeping up w/your Autopian reading, you’d know that he recently bought ANOTHER BMW i3: what’s probably one of the most expensive ones around (in a great uncommon color FTW).
i wonder if that was a supplier/design/materials issue. For roughly the same periods Subaru AC compressors suffered a similar fate, where the compressor failure would require a complete replacement of the whole HVAC system, because the failure would result in sending metal shrapnel through every component on the cooling loop of the HVAC.
It really sounds like it. MKV Volkswagen golf/Jettas have a similar issue, the original compressor supplier for 06-08 had a fatal flaw and are known to fail (brand name escapes me) and the fix was to replace with an updated compressor from a different supplier. Metal shavings were present in the lines but flushing with acetone cleaned it out and no problems 2 years later. Though that’s much easier than on a refrigerant system that goes through the battery!
yes, it looks like the passages for the HVAC on cars are a lot larger than the passages for the batteries, soi it might be easier to clean up. Some of my HVAC friends tell me that blowing compressed nitrogen thru the system will flush it. Compressed nitrogen is also used by them to detect leaks in the system.
My N54 had 13 revisions of their fuel injectors and they still didn’t work properly.
People, stop buying BMWs unless you’re gonna lease or have deep pockets. ‘German engineering’ doesn’t mean what you think it means.
I almost bought an N54 135i, and thank the goddess every day I dodged that bullet.
N54s are really funny to me because they used goofy piezoelectric injectors that are technically superior to solenoid style but, as everyone has since found out, suffer heavily in reliability. The funny part is that the most reliable N54 car you can buy is one that somebody has modified to push 800whp because in order to do that they will have replaced the fuel injectors, the turbo, and the high pressure fuel pump, which are the three failure points of that engine. Turn it back down to, say, 400 or 450whp and it’ll run forever.
Some people are also experimenting with plugging the direct injection ports altogether and running classic port injectors in the intake manifold like an N52.
I don’t understand the “absolutely nothing anyone can do to avoid it.” Is replacing the compressor as a periodic maintenance item – like a timing belts – not an option? I know it doesn’t 100% prevent the problem and isn’t cheap, but some timing belt replacements aren’t either and it makes catastrophic failure far less likely, no?
I suppose the benefit of the timing belt is that manufacturers have done studies that put the 99% failure interval about 7 years, 100K miles (or whatever). With the compressor, data at this point suggests that failure can happen at any point, though I bet if we build a probability distribution (which nobody will ever do, though they try with internet polls) we could come up with some safe replacement interval.
Scheduled replacements as preventative maintenance are much lower than the 99% failure interval, aren’t they?
A 99% failure interval of 7 years, 100k miles wouldn’t make any sense when most belts get changed at about 90k miles. I don’t know manufacturer risk tolerance very well, but benchmarking to the 99% failure interval seems like a foolish decision. It seems you’d want things replaced while they’re still far more to the left on the statistical bell curve.
Related, I know of engines that have lasted a lot more than double the scheduled replacement interval without a new belt. Some more than triple. Of course, some simply failed… …the introduction of interference engines was a painful lesson for some old-school mechanics.
Hence the “or whatever.” You get the idea. They’ve established a failure rate at their replacement interval that spares them and the vast majority customer from failure.
Yes, you can absolutely install a later-revision A/C compressor as a preventative measure, potentially with an inline filter installed in the suction pipe. Costs about $1700 for brand new OEM parts, and takes two hours to install not counting driving to a shop to have the refrigerant evacuated and recharged.
This is good to know. I was briefly astonished that the compressor for these was somehow as expensive as the entire brand new HVAC for my 2k sq ft house. The quoted numbers in the article being for the entire AC system, rather than just the compressor, makes a little more sense.
So, David, it would seem like this is a far worse flaw that the timing belt thing that you did a while back, no?
I know you’re loving your i3, but why would any reasonable non-tool-handy person ever buy a BMW? Unless the answer is lease
I wouldn’t say far worse. Both involve the vehicle being destroyed unless you undergo a rather invasive repair that replaces a ticking time-bomb component preventatively. If anything, the AC compressor is easier to replace, though you’ll need an expert to recover the refrigerant.
Plus, ideally, if you install a filter, you won’t have to do this preventively. Mine has lasted 10 years, 145,000 miles; once I’m out of warranty I’ll probably replace it, install a filter, and stop worrying about it.
Non snark question: is it worth doing the filter first? How long until it’s out of warranty?
Definitely a way worse thing than a timing belt, since there is a known replacement interval for a timing belt that replaces it well before it is likely to fail.
Well, the good thing here is that in CARB states the compressor is warranted for 15 yrs, 150K miles, and if it does fail, you install a filter and Bob’s your uncle.
Thank you David for bringing light to this issue even when it could potentially interfere with you ever selling your two personal i3s. That is honest journalism. Bravo.
Yeesh, microfins. Minute those clog up you can say good bye to your A/C system. You can’t flush them and you can’t repair them. Leave it to the Germans to design a system for the gentle climates of Europe, then forget that they sell to different countries where the A/C is worked to the extreme and is considered essential.
This right here, my GM cars the AC is ice cold hospital grade cooling but my Polestar is meh, it feels like when I went to Portugal and the AC of my hotel room couldn’t keep up cooling under 74F for some reason.
I get the microfins are the most efficient way, but you would think someone would have thought of this issue, and maybe installed a filter from the start…or designed a system less prone to clogging (although I suspect that would be less efficient cooling)
It’s time to admit that EV technology is still so bad, prone to runaway fire and extensive destructive aftermaths, and not worth owning or driving at all.
Viva la Petrol! Viva la Diesel!
Have you forgotten about carburetors & chokes? Took some time to find engineering solutions for them, but it happened… And I guess you’ve never heard of Diesel engine runaway.
Well then…
I wish you the best of luck with your two i3s, David.
Back in 2019 when I was looking to trade my CLS 550 for an EV, I seriously considered the i3. Thank goodness I couldn’t find one at the same price point that I paid for my 2017 Bolt!
And I bought the Bolt exactly one month before the battery fire debacle broke news. I procrastinated on the battery replacement for a good 40k miles (72k total), and finally got it replaced early Feb this year. Now I have another 8 year/100k battery warranty.
You’d think if GM, of all companies, actually stood behind their engineering that BMW could do the same.
I hope David makes sure his garage is clear of rats and he adds a flea repellant to his car wash regimen.
Took me a minute to get the reference…lol.
I didn’t think it was possible to build anything more stupid than the LEAF’s battery cooling system. I stand corrected.
The concept is not stupid, and in some ways is a good idea because it reduces the complexity of the system – but they should have included a filter on the compressor outlet, which I feel in the old days all auto-ac systems had, but maybe I am wrong about that.
The best example of Germans being Germans. Build a technological masterpiece with very clever engineering and design. Convince themselves the design is flawless. Cheap out on the one crucial component to ensure that when it fails it destroys the car. Finally, don’t own up to the problem, show any remorse or issue a recall.
This is why, no matter how much I may like a German car, I will never ever buy one again!
I’ve seen generic A/C compressor black death now and then, generally on Mercedes, but obviously in that case it’s just the usual 2-4k in A/C work. But this, this is a repair I pray I never ever have to write. Unlike a lot of places, I have to build the parts list for the repair and JFC does that sound like a nightmare! And to have a half dozen supercessions on the compressor means BMW knows they done messed up good. It takes a whole lot of fuckery for us to bow out of a repair, but I suspect we would be giving a dealer referral if we saw this. This also sounds like the sort of scenario where I use my Extra Special Gentle Voice reserved for only the absolute worst possible cases.
“And to have a half dozen supercessions on the compressor means BMW knows they done messed up good.”
Parts get up-issued for a whole bundle of reasons, not just to fix problems. Adding a boss so it can be used in a new car, changing a manufacturing process, changing the supplier of a sub-component, adding a recycling label, any change at all. I’ve seen over twenty up-issues on a single part with no functional changes at all.
I once up-issued an engine part because Styling didn’t like the shade of black paint we used.
This is the correct take here.
My sweet wife had a voice like honey, and when she ran her shop it was used on every BMW owner that rolled in the driveway.
They used to take on minor BMW work, but the car owners were usually turds to deal with.
Would complain about the cost of service, repairs, and comment “I can take it to the dealer for what you charge me.”
She would say (sweetly), that might be a good idea for you to do. LOL!
When I saw the plague doctor, I was immediately compelled to read this article.
This problem was also entirely avoidable by re-designing the compressor with stainless steel components instead of aluminum and then mechanically overbuilding it so that it is unlikely to fail. Using steel would also make extracting any debris via a magnet a relatively inexpensive method to clean the channels.
Even better would be to choose a more power-dense LiIon battery less prone to generating heat. Most hobbyist-built EV conversions don’t have liquid battery cooling, and don’t need it. But that does entail design compromises(slightly less continuous power, choosing batteries with 10% less energy density, ect.).
It seems BMW designed this car to have component failures just after the warrantee is up.
Most enthusiasts dont build EV conversions that can DCFC (no cooling needed), last I checked. I admit I haven’t read up on conversions in a few years, are they doing DCFC now?? If so, awesome!
Or just use indirect cooling like others did. And a $200 Prius electric compressor instead of a $2000 unit.
Exactly!
Sometimes using a different brand of part far exceeds the money and time spent dealing with a crap design in the long run.
Great point here.
“It seems BMW designed this car to have component failures just after the warrantee is up.”
As an automotive mechanical design engineer I cannot over-stress how hard it is to design in a mechanical failure to happen at a precise time, after a precise number of cycles, or at a precise load. It’s so much harder, and costs so much more in testing, that it’s not done. Plus it’s hugely immoral and we have a responsibility as professional engineers to refuse to do it (plus, like I said, it’s really hard to do, and also no one is paying extra for a feature that benefits no one).
What we do instead is design for zero failures during the design life. We then test to validate the design for that target. We don’t know what’s going to fail after it’s out of its design life. Almost certainly nothing immediately, but who knows. Certainly not the OEMs because we only test to prove durability, not to ensure failure.
I bet the DFMEA had this failure in it, and they tested to ensure it was unlikely to happen in the vehicle’s design life.
Could we extend the design life? Yes. Would new car customers pay the extra cost so that the 8th owner has smaller bills? No.
I don’t disagree with you, but I disagree with you.
I’m sure that these BMW engineers aren’t literally breaking out the slide rule to see when something will shit itself, and/or total the car. That would be at the least immoral if not straight up actionable….see VW dieselgate and so many others.
But there is an absolute metric screw-ton of data, whether it’s merely anecdotal, or actual repair data, about sooooo many components on soooo many cars. My favorite was the Dodge Neon head gasket, where pennies were pinched to get an inferior head gasket, which lasted juuuuuuust long enough. Heads blew, cars were ruined, and so was the car’s good name.
Here and elsewhere, this has been the prevailing opinion for a very long time: German engineers and/or accountants and/or the companies themselves are coasting on how things used to be, and catering to lease customers who like shiny new things, and everyone else be damned. And you know this is happening.
Engineers, by nature, want everything they build to be awesome and the best they can be. But engineering, by nature, is always a compromise. And sometimes one of those compromises is cost (aka beancounters). Another is legal liability. And still another is when your asshole boss says you have to do it that way.
We design to pass a set durability, but the pressure to reduce cost, and mass, and assembly time never goes away. No one is coasting, but cars are more complex than ever.
All it would take to improve durability is for the same governments which keep reducing emissions targets to mandate a minimum service life of 20 years and 3000,000 miles. Cars would cost more, but they’d last longer, and it’d be better for climate change.
Since the change to the multi layer steel (MLS) head gasket during the 1998 model year, the neon’s head gasket has been rock solid. When installing the MLS gasket in an existing vehicle, one must be sure that the head and block surfaces are properly smooth and flat as specified in the service manual. Hacky repairs (ignoring poor surface quality and just tossing in a new gasket) are not the fault of the car or its designer.
Doesn’t a warranty at least marginally reflect the design life of the car? I get that it’s impossible to design something to fail at a precise time. However, If I design for zero failures during it’s design life, it would make sense to only warranty the design life. If failures can reasonably be expected outside of the (warrantied) design life, then the car has been “designed to have component failures just after the warranty is up.”
The problem I have with the phrase “designed to have component failures” is that it implies that the intention of the design engineers is component failure. This is very much the opposite of our intent. That’s why it winds me up so much.
There is a link between warranty and design life, but it’s not direct. Warranty is a marketing tool to gamble increased sales revenue against the cost of fixing stuff.
huh…I don’t think your view is unreasonable, and I see how “designed to fail” is imperfect phrasing. I’ll try to clean up my vocab going forward!
Not by design, agreed, for the reasons you listed. But cost cutting will do the rest: and when you have upscale, complex cars with ambitions that can’t be matched by their supply chain and final MSRP, that’s where corners are cut.
To stay on the i3 example:
What would they need to give up to keep the transmission oil change cap? What would have to be cut or cost added to have the otherwise perfectly foreseeable (as an engineer) compressor failure? Refrigerant filter vs what?
So for all the good, principled work an engineer will do, myriad other people’s decisions will undo that. Result is fragile, complex cars with the shelf life of a butterfly. Some models and brands (like toyota and some GM cars) don’t succumb to this because they actively work on their reputation in front of that 8th owner.
The compressor failure will have been foreseen in the DFMEA on the system. The severity score is high, so in order to approve it for production they would have had to ensure the occurrence is low. The only way to prove low occurrence is testing, which they would have done on accelerated life cycle durability. Had the AC compressor failed in testing it would have been redesigned until it didn’t fail.
You don’t fix a high occurrence score with countermeasures to give a cheaper repair, which is what adding a filter would do (you add a filter, AC compressor wears out, filter gets blocked, battery goes in to thermal de-rate, customer is stranded: it’s the same DFMEA score as doing nothing).
Great points. But still skeptical because in this case repair is not repair, it totals the car. Something must have gone wrong, because I don’t think any car manufacturer wants this reputation.
Is a reputation for generating horrific repair bills a new thing for secondhand German cars? They’ve seemed OK with it before, and it doesn’t seem to be putting off new car buyers. I’ve had five old BMWs and the last two I sold because of the disproportionate repair costs compared to the value of the car.
I suspect the percentage of i3 AC pump failures is actually very small but seems more of an issue because of how the internet works.
Or it may be that accelerated life cycle testing on this relatively new technology isn’t mature yet and the mechanism for failure only happens in real world durability. AC pumps are pretty well understood though, and usually reliable.
I’m picking this comment out to say that I so enjoy reading comments from the many engineers who visit Autopian. I don’t care if you all agree or not, it’s a gratifying learning experience.
Having said that, planned obsolescence is a fascinating topic too and a hot one around here.
Having said that, I came within spitting distance of leasing a Polestar 2 a while ago, despite the fact that my nearest service provider would have been about 600 miles away. I had gotten a little too excited for a minute there. Apparently my frontal lobes are still working, because I didn’t sign. The sales guy was intimating that he really couldn’t think of anything on the car that really was likely to go wrong. This looks like the beginning of a long learning process, eh?
isn’t this the German way?
Great discussion on this subject!
I would suggest that most design engineers are hard working, moral, and occasionally brilliant.
In my experience, the problem comes in meetings. Maybe leadership suggests that a component or system should be value engineered. When healthy discourse is discouraged, outcomes are sub-optimized.
Maybe a manager gives up on trying to explain to his boss the importance of increasing the budget on said component. Then passes down the direction to his implementation team with a bit of frustration.
Lots of unfortunate things happen along the chain of command. And remember the folks at the top are often under the most pressure to deliver financial results. After all, business exists to generate wealth for the owners. How much wealth is enough?
I own a PHEV, not BEV, but I’ve had cooling issues on my high mileage car. The repair came down to cleaning a hard to reach air filter behind the passenger rear tire that was gunked up. At the end of the day, no engineer can plan for all contingencies. Electric is still newish technology and there is a lot to learn. Early adopters likely will pay a price. They don’t call it the bleeding edge for nothin!
This is not a BMW-specific problem; I had Honda’s Black Death issue in my CR-V, where the compressor self-destructs and infects the entire AC system. However, it didn’t completely destroy my car, thank God.
From your link:
“We just got our son an ’05 CRV and the dealership told us the only issue was that the AC compressor needs replaced. They said they filled it with freon a few times and it blew nice and cold for a couple of weeks then went back to blowing hot air. Unfortunately I did not know about “Black Death” until after we got the car home….luckily we were able to negotiate a $1000 price decrease to account for the proposed cost of replacing the compressor (the dealership told us it would cost about $800).”
That sounds VERY sketchy to me. If the system blew cold for a few weeks and didn’t make any untoward noises and it’s not covered in blown out compressor oil I don’t see how that indicates a compressor failure. IMHO that sounds more likely to be a seal. The O rings and Schrader valves are the most likely culprits, then perhaps a pinhole leak in the condenser from a rock.
Next time before you take it in borrow a manifold set and vacuum pump from AutoZone, hook it up, check the pressure to make sure its empty, pull 30 inches of mercury, shut off the pump and see if the system can hold a vacuum for a half hour or so. If it can the problem is most likely the Schrader valves, if not it might be as simple as an o ring. If you suspect a leak fill with R134 and UV dye (NOT stopleak! ).
Check over the next few days with a UV flashlight ($10 on Amazon) for evidence of a leak. It might take a while to show up.
The good news is seals and valves are REALLY cheap. I had a similar issue with my ’06 Accord a few months ago where I thought the compressor had failed. I was wrong. The “compressor oil” was power steering fluid from earlier o ring failure on the PS pump (very common failure and a super cheap, super easy fix). I followed the steps I listed above and found the system held vacuum just fine with the manifold on. I replaced the Schrader valves (screw out, screw in, took 3 minutes, $10 with tool), refilled with two $12 cans of R134 from AutoZone and its still blowing ice cold today. All in I think it cost $35 and a day of mostly pumping vacuum. Even replacing everything but the evaporator behind the dash would have been under $400 in parts from Amazon. The evaporator AFAIK can be flushed out in the case of actual compressor failure with an A/C system solvent. It’s not cheap but its cheaper than a sketchy shop.
I dunno your DIY skill level or your workspace situation but if you have the resources basic DIY A/C troubleshooting and repair work is very possible. If anything you’ll be better prepared for the shop.
Many of the issues brought up with German cars have parallels with cars from other countries, but “haha German car bad!” gets people to click the smiley face.