Dieselgate was a watershed moment for the automotive world. In 2015, Volkswagen was caught cheating on emissions rules by building cars that acted clean when tested, but polluted heavily the rest of the time when out in the real world. The Germans weren’t the first, though. Nope, one previous example occurred in the 1990s, and it involved General Motors.
You might have enjoyed our story about Ford’s innovative methods of meeting emissions standards back in 1975. A hunt for similar efforts at the General led me to a darker narrative around illicit “defeat devices.” This term became a hot-button topic after VW’s folly, and referred to a tool used to circumvent emissions regulations.
![Vidframe Min Top](https://images-stag.jazelc.com/uploads/theautopian-m2en/vidframe_min_top1.png)
![Vidframe Min Bottom](https://images-stag.jazelc.com/uploads/theautopian-m2en/vidframe_min_bottom1.png)
It sounds like some cool bit of hacker kit you’d use to steal millions of dollars when you’re cracking into a mainframe, but the term actually dates back much further. In fact, since the dawn of emissions regulations, the EPA has been hunting for sneaky little devices that work around those regulations. In 1993, that’s what the EPA believes it found when it was tinkering with some of the General’s premium machines.
![Gm High Technology L26](https://images-stag.jazelc.com/uploads/theautopian-m2en/GM_High_TEchnology_L26.png)
Cool It Down
It all kicked off in a routine lab test in 1993. The EPA happened to have some Cadillacs on the bench. Much to its surprise, the vehicles tested were spitting out up to 10 grams of harmful carbon monoxide a mile. This was well over the legal limit at the time, which stood at a mere 3.4 grams per mile. Further investigation revealed something curious. As explained in a contemporary EPA press release, The Cadillacs were exceeding the official limits under a very specific condition: anytime the climate control system was in use. Before long, the Justice Department felt like it had a case, and the matter hit the courts.
The matter concerned Cadillacs fitted with the 4.9-liter L26 “High Technology” V8 engine. This included a wide range of models built from 1991 to 1995, including the 1991 to 1993 Cadillac Seville, the 1991 to 1995 DeVille, the 1991 and 1992 Fleetwood, and the 1991 to 1993 Eldorado. A population of around 470,000 cars were caught up in the furor by the time the matter came to a head.
![S L1600 (3)](https://images-stag.jazelc.com/uploads/theautopian-m2en/s-l1600-34.jpg)
![Screenshot 2024 07 17 162743x](https://images-stag.jazelc.com/uploads/theautopian-m2en/Screenshot-2024-07-17-162743x.png)
GM had a great explanation, though – or some might call it an excuse. The official line was that back in 1990, Cadillac was having trouble with a number of models. Customers were finding their vehicles were stalling out when running the heating or air conditioning. Thus, GM implemented a simple fix and retuned the electronic fuel injection system on certain models, setting up the engine computer to supply more fuel while the HVAC system was on. Unfortunately, this richer fuel mixture also greatly increased carbon monoxide emissions – no surprise, given CO is a common byproduct of incomplete combustion.
It’s an odd story. Generally, you’d expect a major automaker such as GM to have thoroughly tuned and tested the engine controls prior to releasing the vehicles to customers. This would include test-driving the vehicles under a variety of real-world conditions, including with the heater or air conditioner running. You’d also expect an automaker to verify the car was emissions compliant under a wide range of conditions, including operation with the HVAC system in use. And therein lied the crux of the matter.
![Fireup2acx](https://images-stag.jazelc.com/uploads/theautopian-m2en/fireup2acx.jpg)
Back in the 1990s, standard EPA test cycles weren’t as rigorous as they are today. Cars used to be tested on regulated “city” and “highway” driving cycles, with the climate control turned off. Thus, if one of the given Cadillac models was tested under a typical regime, it would show up as compliant. Flick on the air conditioner, though, and the richer fuel mixture would kick in, and CO levels would shoot right up. This meant that Cadillac was building cars that passed typical tests, but emitted much more pollution when out on the road. It may seem somewhat like what VW did years later, though it’s worth noting that Cadillac wasn’t trying to detect when it was being tested, and it wasn’t changing its emissions solely when under the microscope.
When the EPA found out, the matter went to court. The EPA argued that GM knew exactly what it was doing, and that it had effectively built a “defeat device” to skirt around emissions regulations. It stated that the company would have known the tuning change would have increased emissions output. By 1995, the EPA stated that the trick had led to the illegal release of 100,000 tons of excess carbon monoxide pollution.
"Carbon monoxide can cause cardiopulmonary problems and can lead to headaches, impaired vision and a reduced ability to work and learn. These so-called defeat devices are not just paper violations, but result in real increases in emissions that affect real people." - Attorney General Janet Reno.
![Photos Cadillac Deville 1994 4](https://images-stag.jazelc.com/uploads/theautopian-m2en/photos_cadillac_deville_1994_4-e1721199513947.jpg)
The case didn’t exactly go GM’s way. The automaker spent $45 million to settle the charges while refusing to admit guilt. These included allegations that GM had sold vehicles that were non-compliant with the Clean Air Act, that it had sold vehicles fitted with illegal defeat devices, and that it had failed to accurately explain what it was doing to the relevant authorities. GM had also allegedly installed chips that made the tuning change in certain 1991 model year vehicles after they were sold—which the EPA alleged counted as tampering with vehicles to circumvent emissions standards.
At the time, it was the largest case ever brought by the justice department under the Clean Air Act. $11 million was paid as a fine, with $25 million allocated to recall and refit affected vehicles. A further $8.75 million was dedicated to offsetting the increased emissions from the affected vehicles.
![Scoopy](https://images-stag.jazelc.com/uploads/theautopian-m2en/scoopy-e1721414309908.jpg)
![Nto On The Test X](https://images-stag.jazelc.com/uploads/theautopian-m2en/nto-on-the-test-x.jpg)
Still, GM remained unrepentant in the wake of the judgment. “It was not in the rules, not in the regulations, it’s not in the Clean Air Act,” GM spokesperson Betsy Hemming told the New York Times in the wake of the settlement. GM’s position was that the regulations had given them no reason to test their vehicles with the climate control running. The automaker put the whole situation down to the fact it was simply playing by the rules of the EPA’s standard test cycles. “This is a matter of interpretation of current regulations regarding the complex issue of off-cycle emissions,” said Dennis Minano, GM’s then vice president of corporate affairs.
Of the $8.75 million offset fund, GM agreed to use the money to support cleaner air initiatives to help repair the damage that had been done by the heavily emitting Cadillacs. This was to potentially include things like buybacks of highly polluting vehicles like older school buses, and replacing them with newer models that ran cleaner and more efficiently. These measures were targeted to Phoenix and Maricopa County in Arizona, and Southern California, as these areas were suffering some of the worst air pollution in the US at the time.
As later approved by the Justice Department in 1996, GM funded projects in California, Arizona, Connecticut, New York, Massachusetts, and New Hampshire. The measures also saw GM providing electric vehicles for government agencies, and saw the automaker fund the purchase of electric shuttles and pick-up trucks for use at Boston’s Logan International Airport. The airport also received funds to convert ten tow tractors and a conveyor loader to electric drive.
All in all, it was believed that these projects would cut annual carbon monoxide emissions by 2,400 tons, hydrocarbon emissions by 300 tons, and nitrous oxide emissions by 175 tons. These benefits would stack year on year, but it would be some time before they made up for the 100,000 tons of excess carbon monoxide emitted by the questionable Cadillac fleet.
From the EPA’s perspective, this case was a slam dunk. It mandated certain emissions standards, and specified a test cycle to measure if cars lived up to them. The test was merely a functional tool, as far as the EPA was concerned—not the be-all and end-all of compliance. Passing the test was important, but no guarantee a car was compliant. Meanwhile, GM had taken the opposing stance—that any car that passed the standard tests was fine and legal to sell. The latter attitude is often one that works quite well in motorsports, but it’s not a winning strategy when it comes to emissions compliance.
![Blast](https://images-stag.jazelc.com/uploads/theautopian-m2en/blast-e1721414372433.jpg)
It’s not quite the same the hurdle Volkswagen would trip over decades later. GM’s attitude was that the EPA didn’t test emissions with the AC on, so emitting more in this regime was surely allowed. Volkswagen, on the other hand, set up its vehicles to specifically detect when they were under test, and artificially reduce emissions. The latter was a far more egregious example of a defeat device, whereas GM’s method fell much more into shades of grey.
Still, had VW executives examined the million-dollar disaster that had befallen GM, the company might have been less eager to press its luck. Volkswagen was caught in much the same way, too, albeit by private researchers rather than the EPA itself.
Ultimately, automakers, and the engineers within, should take this as a cautionary lesson going forward. Just because you can sneak past the test, doesn’t mean you’re doing the right thing. Still, one suspects that Volkswagen won’t be the last automaker to try and slide under this tantalizing limbo pole.
Image credits: GM, Cadillac, Volkswagen
Its the EPA’s responsibility to define the scope of the test and the required results. If the EPA’s scope included climate controls off, that’s on them.
The rich man’s version of “rolling coal”.
Seeing the photo of the engine bay makes me feel warm and happy inside.
I miss that car.
It may have had no knock sensor and run only on premium (and apparently trickery) – but that engine performed flawlessly for 210,000 miles (then was sold).
My grandmother had a 91 Eldorado with the 4.9 that this recall applied to. After the PCM was replaced, it began stalling at idle occasionally with the air conditioning on. So they seemed incapable of fixing the problem without doing it illegally.
Unrelated to the story, but I always liked that bodystyle of eldorado, just crisp, clean, simple, and beautiful.
I hadn’t always liked it, but now I do. Though I’d prefer thinner side molding.
More people need to experience the 4.9 for themselves. It may have defeated emissions standards but it is truly one of the smoothest engines to ever hit the road. Pulls like a boat motor and keeps a comfy smile on your face the whole time.
If you’re not cheating you’re not trying right?
We are set for a lot more of this due to the recent decision by the Supreme Court to throw out what was known as Chevron deference. Now instead of regulatory agency experts making rules to implement legislation, judges with no expertise will be involved. Prepare for almost unimaginable chaos. https://www.scotusblog.com/2024/06/supreme-court-strikes-down-chevron-curtailing-power-of-federal-agencies/
“Of the $8.75 million offset fund, GM agreed to use the money to support cleaner air initiatives to help repair the damage that had been done by the heavily emitting Cadillacs. This was to potentially include things like buybacks of highly polluting vehicles like older school buses, and replacing them with newer models that ran cleaner and more efficiently. These measures were targeted to Phoenix and Maricopa County in Arizona, and Southern California, as these areas were suffering some of the worst air pollution in the US at the time.”
And hey whaddya know GM just so happens to make school busses too!
https://en.m.wikipedia.org/wiki/Chevrolet/GMC_B_series
I was a mechanic for 36 years, and a long time at a GM dealer, now retired. The reason it was done is that the EPA always pushes right at the limit of what the technology at the time can achieve unnoticeably. So, in order to meet the limits, the manufacturers have to make a choice, and trust me, it’s all of ’em. The choice is whether you wanna piss off your customers or Uncle Sam. In this case it’s how to handle the extra load at idle when the AC compressor comes on. V8 or not, it’s a considerable load at idle, especially on a gasser. It’s already idling as lean as possible, but when you add more load, even a small amount, the already uneven idle is felt that much more. Can’t increase idle without adding fuel. Rough idle? Under warranty? Fix it or I’m suing!
In this case this manufacturer chose to fix those cars whose owners complained. You better believe they had an actuary run every possible scenario before they changed anything, even including the chance of winning in court due to the exact way the rule was written. By legal standards, they had a strong case. Looks like what they came up with was exactly what they did, because of a worst case scenario being better than a class action or a full blown recall. Especially since it was Cadillac, one of the highest profit margin cars in their lineup. You think the 50-some-odd mil they paid wiped all the profit they made off of all those cars they sold during that stretch of time? No, and I’m sure there’s some computer file in Detroit that shows the calculations as to why they chose how they handled it.
It’s the same today. I have a relatively new truck, and the idle sucks compared to what you would want it to be, but it’s just so they are allowed to sell it in the first place. I would rather put up with it than have it cut off at every freakin’ stop light you come too though. Won’t be long before even that won’t be enough, and the cycle begins again.
For General Motors, this is not even close to the worst thing they did for the environment. Recall that they funded studies to create the false impression that leaded gasoline was safe. GM delayed the end of leaded gas by many years. No doubt that caused much more harm to the environment and humans that we are all still dealing with today. I am down a few IQ points because of the General. Thanks GM!!!
https://www.nbcnews.com/health/health-news/lead-gasoline-blunted-iq-half-us-population-study-rcna19028
I’ve read studies that correlate the widespread use of leaded gasoline to the rise in crime in the 1960s and early 1970s, and crime’s decrease subsequent to unleaded gasoline’s introduction. Causation couldn’t be shown, but the trends matched clearly.
So is that why GM introduced the Northstar a bit early perhaps? Didn’t early Northstars have problems?
GM sux, and they certainly do outsource a lot of their R&D to their customers, but I actually am on their side here. It passed the test as written and complied to the letter of the law.
Also, didn’t they do something in the 70s where the emissions shit only worked at idle?
Eh, not really. Rules are created to work in the real world, not in a carved-out piece of existence. They cannot be interpreted in a context or a way that contradict the very sense they were created to be regulated with.
Plus, Cadillacs were designed, marketed and used with a prominent use of AC and heat devices in mind, and not in a circumspect and frugal way. GM’s position was only plausible if we accept bad faith interpretations. The law does not work that way.
By early Northstars, you mean like the first 13 years of a 19 year production run? That’s about how long it took GM to address the issues
It would be in character though, they launched the 4.9’s predecessor, the HT4100, a year early and unfinished due to the V8-6-4 problems
That is a little different than the account I heard back in the day.
What I heard was that it only did the enrichment at idle and with the AC compressor running. The first cars released to the public generated complaints of the occasional stalling but more commonly a “rough idle”.
So the first batch of cars got new chips, if the owner complained and it was a running change on the production line.
I had a 93 DTS with the 4.9. Great motor. Smooth and torquey, everything you want in a Cadillac. I miss that car. Absolute couch on wheels.
…just don’t stand behind it and breathe when the AC is on.
Ok here’s my question: haven’t manufacturers been able to make engines run an A/C compressor for decades without this happening? Maybe it’s to be expected on a 1.3L Geo Metro but a 4.9L Cadillac V8? And their solution was to just enrich the mixture so the engine got more fuel? It smacks of GM ‘band-aid’ engineering.
The way it probably played out was that the emissions folks leaned out the mixture until it passed. They ended up leaner than the engine was really meant to run, and it was prone to stalling. Add the load of a compressor, and it was more prone. Then someone had a great idea.
This is the professional level of GM’s engineering. Like that time when the windows of your rear doors woudn’t roll down. In the rest of the civilized world you’d be fired for even breathing out such ideas.
That was so they could downsize the cars without having to advertise a reduction in rear seat room, no window mechanism meant they could squeeze a little more seat width into the door cavity. GM was more worried about their customers not accepting smaller cars than they were about their customers not accepting fixed rear windows
Yes, but if you’d see any European car of the time, especially economic ones, you’d see that some serious engineering and design efforts were put in saving space without sacrificing functionality. This is why I find it unacceptable in not-cheap sedans.
Hey, it got the G-body an extra inch side to side over the Datsun 810
While being 7.6 inches wider on the outside (according to Wikipedia). So, a decent engineer could have carved enough seat width without out-cheaping a Citroen 2CV.
These are the type of issues that are created when the people who write regulations don’t consider real world conditions and base their compliance tests on ideal situations. The cars met the ideal conditions of the tests which puts the fault on the regulatory agency (but they have no faults). GM’s engineers merely saw what they had to meet and made their design fit the regulations. This is the kind of B.S. created by mindless bureaucracies that lack the vision and competence to develop regulations and monitoring methods that are in alignment.
This the kind of BS that’s developed when industry is allowed to assist in the crafting of regulations. If I had to guess, industry experts had input on the testing procedures.
You’re right. We shouldn’t bother writing any regulations. Why do anything at all about the problem when it’s never a perfect solution?
Somehow, other manufacturers managed to meet the emissions regulations for those model years without their cars stalling every time the air conditioning was on. And indeed, GM, for all of their other engines. I can only assume that there were emissions components that might have needed to be upgraded, and they didn’t want to spend the money, hence the half-assed fix instead.
They were technically correct.
The best kind of correct.
This also brings to light another issue… the efficiency a car has with the A/C turned on. Some past car A/C systems would cause a serious MPG hit.
And in 2008, they started to test the impact A/C use had:
https://www.fueleconomy.gov/feg/fe_test_schedules.shtml
I’ve heard it debated than at highway speeds, AC improves emissions v.s. having windows open due to aerodynamics. I do wonder how much it matters given the penchant for so many people to 0-MPG idle in their vehicles.
One thing I’ve always wondered about is at what speed exactly is the threshold crossed where you should close your windows and turn on the A/C.
I imagine it varies by vehicle. But I haven’t seen anything where that was studied even for one vehicle.
I thought Myth Busters did that I know they did a few other MPG tests over the years.
I’ll have to look that up.
Anecdotal, but I could’ve sworn I read it was around 40-45mph…which, coincidentally, is about the same point where you should close your windows to prevent hearing damage due to wind noise.
I live in Florida. A/C is turned on as soon as the seat belt is buckled for at least 6 months. I’ve used the heater maybe a dozen times in ten years. So the result of any such test is irrelevant to me, especially since I drive a Fit.
In the early 70s, Ford thought it was perfectly acceptable to run the AC compressor any time the engine was on. So, basically, you had de-humidified air all the time, and added hot water to the heater core (or blended heated air w/ the cold air) if you didn’t want to freeze.
I’m more on GM’s side with this one, if only because the test as-written was stupid as hell, with loopholes you could drive–well–a Cadillac through. Most people, especially in the regions were air quality was of direst concern, were not driving around with the HVAC off.
GM knew that Cadillac owners drive with the AC on so they leaned out their mixture when the AC is off to match the test. However, the EPA should have also known this and written a better test.
Is it still a cat and mouse game if only the mouse is playing?
GM played by the rules as written. Perverse outcome certainly that I will not condone. This is why we need true experts writing regulations (not judges).
Judges weren’t and aren’t writing EPA regs…
In the early 90’s, AC was not a standard feature on a lot of models. I’m also guessing the tests used were developed years earlier, when AC would have been less available.
Fuel injection itself was just becoming common. The ability for manufacturers to change the fuel curve without an observable mechanical device would have been relatively new.
I’m really not sure whose side to take on this one. I have to imagine that at any given moment, a majority of cars are driving around with the HVAC system in use, yet the EPA specifically states that cars are tested with the HVAC off. Seems kinda dumb.
I believe that every fuel injected car I’ve ever owned (and maybe some carbureted cars as well) modify, at the very least, the idle speed when the A/C compressor is engaged. Cadillac maybe just modified it a bit too much.
This is a failure of regulation and an exploitation thereof.
I guess the regulations said nothing regarding the HVAC at all. If that’s the case, GM resorted to a bad faith interpretation, that was fairly dismissed.
If the regulations said that the HVAC had to be turned off, then there was a sizable bribe somewhere.
I suspect a company as big and bureaucratic as GM probably didn’t think of it. The ‘we followed the rules’ defense came later.
Certainly in the 70s, many (if not most) manufacturers had an “idle up” solenoid that clicked in when the AC was turned on. I’m not sure if it was already on my ‘77 Accord when I installed the AC kit or it had to be added (probably already there on the carb). AC was a dealer installed kit on most Hondas including my ‘89 Civic SI (which I installed after delivery).
Follow the letter of the law, if not the spirit. I’m not condoning this behavior at all, we all need clean air and climate change is a pending an existential crisis. Yet, it appears GM met the actual testing and legal requirements for certification. Poorly written law is still law.
I gotta go with GM on this. They followed the rules. If someone did a burnout in one of their cars, should the car company take the hit for the increased particulate (rubber bits from the tires) emissions?
Agreed. They followed the law.
Most people don’t realize that full throttle acceleration is “off schedule” and the fuel mix is enriched for that bit of the operation curve. Probably don’t meet emissions levels at WOT.
#whynotboth?
It’s weird that it only seemed to affect the 4.9 and not the 4.5 as used in the Allante, since improved power had been the main reason for the enlargement. You’d have thought the 4.9 would have been more capable of powering the air conditioning instead of less. Of course, the variant of the 4.5 in the Allante was the high output 200hp one, instead of the 155hp in the sedans and coupes. The 4.9 brought power up to 200 across the board, through, again, a 1990 Deville with 155hp could run the a/c without stalling, but a 1991 Deville with 200hp couldn’t? GM was such a cluster
Those are Max HP numbers, not at idle (which is presumably when the stalling would occur). I wonder if the power delivery was different at idle. Also, maybe the AC condensor was larger for the sedans since there is more interior space to cool?
My total guess is that the larger motor was over the line for emissions, so they leaned it way out to pass, but then it was too lean to reliably idle/transition from idle without stalling when the AC was on. The easiest solution that they probably figured they could argue as being legal as the EPA test didn’t test with the AC on was to just ram more fuel in with the compressor running.
It’s really quite simple: Larger engines pollute more than smaller engines.
So if the only thing you’re doing to reduce pollutants is to lean the fuel mixture – you’re going to get performance that’s not as robust across the full range of output.