EPA accuses Fiat Chrysler of excess diesel emissions

jack_christie

Forum Addict
Joined
Aug 1, 2006
Messages
9,620
The U.S. Environmental Protection Agency on Thursday accused Fiat Chrysler Automobiles NV of using software that allowed excess diesel emissions in 104,000 U.S. trucks and SUVs sold since 2014
http://www.reuters.com/article/us-fiatchrysler-emissions-idUSKBN14W29S


Shares down 20%
C1-9lVfXcAE5Eih.jpg
 
Soooooo ... the next company that can bend down?
 
FCA's response makes it sound like they're hoping the Trump EPA just drops it:

"FCA US intends to work with the incoming administration to present its case and resolve this matter fairly and equitably," the company said.
 
Goddammit.

VW poisoned the well for the diesel Frontier. I hope this doesn't derail the Ford diesel plans for the next generation of vehicles. I don't think it is mutually exclusive to want to breathe clean air AND be a car enthusiast.
 
This case seems a lot different than the VAG defeat systems.

According to the EPA, FCA had eight undeclared ?auxiliary emission control devices? (AECDs), namely:

Full EGR shutoff at highway speed
Reduced EGR as speeds increased
EGR shut-off for exhaust valve cleaning
DEF (exhaust fluid) dosing disabled during SCR (selective catalytic reduction) adaptation
EGR cut back due to modeled engine temperature
SCR catalyst disabled during warm-up
Alternative SCR dosing modes
Use of a load governor to delay ammonia refill of the SCR catalyst

Thus, the engine acted differently in real life than in testing; the EPA provided several examples of times when this affected emissions.

The law does allow for AECDs when needed to protect the vehicle, which is the reason for all of the cases listed. So far, nobody has determined whether FCA intended to break the law, or did so accidentally. Based on the results of the recall investigation, it seems likely that there were internal compliance errors within FCA ? that it wasn?t intentional.

It seems more likely that FCA merely failed to appropriately inform the EPA of those specific eight AECD "modes" and so the EPA testing didn't accurately account for it. That's a far cry from VAG's system that was knowingly designed to only work during test conditions.
 
Goddammit.

VW poisoned the well for the diesel Frontier. I hope this doesn't derail the Ford diesel plans for the next generation of vehicles. I don't think it is mutually exclusive to want to breathe clean air AND be a car enthusiast.



As someone with a love of cars and asthma, I completely agree. These companies know what the regulations are. If they can't meet them, don't cheat and expect to get away with it. This will end up costing us all money when the EPA changes the way they test vehicles.

I would love to have a diesel that ran on a seed oil based fuel. This sort of thing makes it harder to happen.
 
This case seems a lot different than the VAG defeat systems.



It seems more likely that FCA merely failed to appropriately inform the EPA of those specific eight AECD "modes" and so the EPA testing didn't accurately account for it. That's a far cry from VAG's system that was knowingly designed to only work during test conditions.

Pretty much this. This is a long, long, long way from VW's deliberate cheat mode when being tested and the deliberate, willful discussion/implementation of same.

However, if they did it intentionally? They need to burn just like VW.
 
This case seems a lot different than the VAG defeat systems.

It seems more likely that FCA merely failed to appropriately inform the EPA of those specific eight AECD "modes" and so the EPA testing didn't accurately account for it. That's a far cry from VAG's system that was knowingly designed to only work during test conditions.

Based on the German authorities' findings months ago, the Jeeps' and others' systems are tailored specifically to work in test conditions. For example, the cold-temperature-based cut-backs of emissions control measures mostly start just a few degrees below the standardized tests' conditions - and the hot-temperature-based cut-backs start not too far above the test conditions.

Governing the dosing of ammonia in the SCR matches what VW did in that regard as well, full blast in test conditions and hardly anything elsewhere to not bother the customer too much with refills.


Is that by accident? I doubt it, it smells of intentionally modeling the system parameters after test conditions.
 
Based on the German authorities' findings months ago, the Jeeps' and others' systems are tailored specifically to work in test conditions. For example, the cold-temperature-based cut-backs of emissions control measures mostly start just a few degrees below the standardized tests' conditions - and the hot-temperature-based cut-backs start not too far above the test conditions.

Governing the dosing of ammonia in the SCR matches what VW did in that regard as well, full blast in test conditions and hardly anything elsewhere to not bother the customer too much with refills.


Is that by accident? I doubt it, it smells of intentionally modeling the system parameters after test conditions.

It's entirely possible it was intentional, but it's still not even remotely close to the scale and depth of the VW cheats. I agree that if it's proven to be an intentional cheat, they deserve some stiff penalties. Do you have any sources on FCA's AECDs being tailored to the EPA test conditions? The VW defeat was very specific and used wheel speed parameters and other inputs to detect testing conditions, while FCA seems to merely to be based on engine temperature.
 
Last edited:
It's entirely possible it was intentional, but it's still not even remotely close to the scale and depth of the VW cheats. I agree that if it's proven to be an intentional cheat, they deserve some stiff penalties. Do you have any sources on FCA's AECDs being tailored to the EPA test conditions? The VW defeat was very specific and used wheel speed parameters and other inputs to detect testing conditions, while FCA seems to merely to be based on engine temperature.

This report from April is extremely German: https://www.bmvi.de/SharedDocs/DE/A...mission-volkswagen.pdf?__blob=publicationFile

Page 80 has a Fiat Ducato, page 92 has a Jeep Cherokee. Here's the graph for the Jeep:



The black line is our emissions limit. From left to right, these are the tests they ran (see page 14ff):

1 Dark blue: The official test as per the regulations. It passed.
2 Dark red: Right after the official test, they ran the same cycle again. It failed horribly.
3 Green: Actually, no point in continuing through the data. You run two cycles, the first cycle passes and the second fails miserably. Something's going on that made the car stop behaving in a test-compliant way right after the test was over.

The report itself is extremely coy about assigning actual blame, sadly in line with the official approach. It says FCA claims they throttle back emissions treatments to protect the emissions treatment equipment. Translated that means "we only put in emissions treatment equipment sturdy enough to pass the test, and program the software to shut it off when the test is over".
Oh, and it's not just one car. They literally write at the top of page 90 "after a first Jeep was tested and showed very high nox values, we had a second Jeep tested by a different lab company. Both passed the official test, and flunked all the others."

An attempt by FCA to explain is on page 90 as well, saying they scale back AGR/EGR below 20?C ambient temperature to protect the equipment... guess what, our tests run at 20-24?C. Coincidence that it's exactly the lower end of the test temperature range? Bad luck that they didn't fit more robust equipment?
Doesn't even begin to explain the red bar from the second back-to-back cycle failure, which was at 20-24?C...
 
Last edited:
I don't quite follow. They ran the official EPA test, which it passed, and then did the same EPA test and it no longer passed. If that's true, wouldn't that mean the FCA software was not actually detecting test conditions, as both test parameters should have been the same?

You also seem to be dismissing their claims that the AECDs are due in part to emission system longevity, do you have any data to show that's not actually the case?
 
Last edited:
I don't quite follow. They ran the official EPA test, which it passed, and then did the same EPA test and it no longer passed. If that's true, wouldn't that mean the FCA software was not actually detecting test conditions, as both test parameters should have been the same?

Not quite. The official test procedure (Euro, not EPA - but similar principles should apply) requires a very specific pre-conditioning of the car, see page 15. By running the 1180sec cycle again after the first 1180sec cycle, they tried to see if there's for example a timer in the car to turn things off after say 1200sec.

You could say they ran one 2360sec test, and it passed the first half but utterly failed the second half. Great news if you never drive for more than 20 minutes and if your garage matches pre-conditioning conditions...

You also seem to be dismissing their claims that the AECDs are due in part to emission system longevity, do you have any data to show that's not actually the case?

They say they shut down parts below 20?C - yet, the second cycle is run between 20-24?C so that can't be it.
As for longevity, how is fitting parts that won't be able to do their job for sufficiently long enough different from VW fitting insufficient nox storage catalysts that do fine for a while, but not so fine in reality? I doubt there's no sufficient equipment on the market, they just didn't fit it.
As for already starting to reduce below 20?C, other manufacturers keep up full AGR to much lower temperatures. Technically both is legal as long as they claim it's "to protect the engine"... translated, "our equipment is unable to do its job in reality".

- - - Updated - - -


For more context on those numbers, here's a Passat *with the defeat device* running the very same set of tests:



It too fails all the non-standard tests, but at less than a third of the Jeep's emissions.
 
Not quite. The official test procedure (Euro, not EPA - but similar principles should apply) requires a very specific pre-conditioning of the car, see page 15. By running the 1180sec cycle again after the first 1180sec cycle, they tried to see if there's for example a timer in the car to turn things off after say 1200sec.

You could say they ran one 2360sec test, and it passed the first half but utterly failed the second half. Great news if you never drive for more than 20 minutes and if your garage matches pre-conditioning conditions...

Page 15 was just as unintelligible to me as the rest of the document, sadly - I was hoping to pick up German around page 8 but that didn't pan out. ;) So the second test was outside the parameters of the Euro testing due to lack of pre-conditioning and it failed. The question remains if this is a specific defeat system for test conditions (illegal) or an emissions system designed to meet only the "letter of the law" of the emissions tests (legal but possibly unethical). Certainly worth investigating, but in no way the slam-dunk case the VW scandal was. VW using inputs like wheelspeed and accelerometer data to specifically detect EPA testing conditions was really the big thing that sank them.


They say they shut down parts below 20?C - yet, the second cycle is run between 20-24?C so that can't be it.
As for longevity, how is fitting parts that won't be able to do their job for sufficiently long enough different from VW fitting insufficient nox storage catalysts that do fine for a while, but not so fine in reality? I doubt there's no sufficient equipment on the market, they just didn't fit it.
As for already starting to reduce below 20?C, other manufacturers keep up full AGR to much lower temperatures. Technically both is legal as long as they claim it's "to protect the engine"... translated, "our equipment is unable to do its job in reality".

So if it's entirely legal to "protect the engine" as you have stated, what's the illegal thing they have done? Emissions testing is about adhering to a set of legal standards, if the standards are OK with that procedure it's tough to say they were in the wrong. They're not cutting emissions out of the goodness of their hearts, that's for sure.

As I mentioned before, I think they should get drawn and quartered if they've been gaming the system, but this case appears to be a lot less blatant than the VW one, specifically in the scope of EPA testing.
 
Last edited:
Page 15 was just as unintelligible to me as the rest of the document, sadly - I was hoping to pick up German around page 8 but that didn't pan out. ;) So the second test was outside the parameters of the Euro testing due to lack of pre-conditioning and it failed. The question remains if this is a specific defeat system for test conditions (illegal) or an emissions system designed to meet only the "letter of the law" of the emissions tests (legal but possibly unethical). Certainly worth investigating, but in no way the slam-dunk case the VW scandal was. VW using inputs like wheelspeed and accelerometer data to specifically detect EPA testing conditions was really the big thing that sank them.

So if it's entirely legal to "protect the engine" as you have stated, what's the illegal thing they have done? Emissions testing is about adhering to a set of legal standards, if the standards are OK with that procedure it's tough to say they were in the wrong. They're not cutting emissions out of the goodness of their hearts, that's for sure.

As I mentioned before, I think they should get drawn and quartered if they've been gaming the system, but this case appears to be a lot less blatant than the VW one, specifically in the scope of EPA testing.

Note, I didn't say "illegal", I said "intentional". In my mind, the intent is obvious by how significant and immediate the change in emissions is between the preconditioned first half of the double-length run and the second half. It's no coincidence that the change occurs about at the end of the official test cycle... I'd love to see a time series of per-minute emissions to spot the change time though.
 
The EPA isn't completely innocent in this matter IMO. If this disclosure of information was so important, why wasn't it mentioned or brought up to FCA during the 2014 homologation year?

At a certain point, the EPA needs to fall on the stake for their incompetence and focusing too much on the legislation rather than the actual scientific process to promote sound environmental policies.

narf said:
As for already starting to reduce below 20?C, other manufacturers keep up full AGR to much lower temperatures. Technically both is legal as long as they claim it's "to protect the engine"... translated, "our equipment is unable to do its job in reality".

While the public may interpret it that way, EGR systems typically do reduce engine longevity since the recirculated pollutants contaminate the engine oil in short order.
 
The EPA and California Air Resources Board told Fiat Chrysler ...

And that's where I stopped reading. Nice try. Two organizations I have zero trust or respect for based on their past track records.

The EPA isn't completely innocent in this matter IMO. If this disclosure of information was so important, why wasn't it mentioned or brought up to FCA during the 2014 homologation year?

At a certain point, the EPA needs to fall on the stake for their incompetence and focusing too much on the legislation rather than the actual scientific process to promote sound environmental policies.

While the public may interpret it that way, EGR systems typically do reduce engine longevity since the recirculated pollutants contaminate the engine oil in short order.

^ This and this.

What a non-issue this all is, because I don't trust the EPA as far as I can throw them...especially under the current administration. And don't get me started on that "California Air Resources" organization either. I will be happy to accept tests done by regular non-biased non-sponsored people that aren't employed by any government, university, or in states not on either coast. I love how so many people came in here drinking the cool aid and then try to say this was the same as what VAG pulled. I doubt there was any illegal or malicious attempt to mislead anyone. In our blame first culture how about we go back to giving the benefit of the doubt, which is what I gave VAG before the truth came out. If FCA did do this maliciously then fine them for each for 104,000 vehicles--but not enough to cripple the business and that's it. I'd feel the same way if this was GM, Ford, or any other manufacturer.
 
The EPA isn't completely innocent in this matter IMO. If this disclosure of information was so important, why wasn't it mentioned or brought up to FCA during the 2014 homologation year?

As I understand it, it was. The 8 missing AECDs were brought up in 2014, and in the meantime FCA has resolved the 1st one but failed to reach an agreement or resolution with the EPA for the other seven. The announcement yesterday was the EPA demonstrating they believe FCA had not made enough effort to fix the issues or adequately demonstrate why the AECDs are necessary. FCA, obviously, disagrees and stated they have made a good faith effort to work with the EPA and are being unfairly prosecuted. What's really gone on over those two years is anyone's guess.
 
Last edited:
I love how so many people came in here drinking the cool aid and then try to say this was the same as what VAG pulled.

If you look at the non-EPA non-CARB pollution figures, they're doing 3x worse than VAG.
 
Top