Discussions with CARB and the EPA on the future of performance parts

Interesting thread and good discussion going on about the process. Always hated the roll coalers in my area, quite annoying to be driving in traffic and some douche decides to blow smoke all over. Though Florida is lawless with how car enthusiasts have little disregard with the deleting all their emissions devices. The demographic of the rollcoalers have always been the type to ruin things for everyone in my experience

But one thing that bugs me is some of the OEM car companies are discontinuing their emissions components. I had some weird EVAP issues on my integra, when I still had the stock ECU and motor and a lot of the stuff was honda has been discontuning a tun of parts at an even quicker rate lately. Good thing the part I bought was still available from denso, but stuff like original hoses, charcoal canisters were, etc were NLA.  I wonder if there will be companies willing to remake these components again if it is a viable business option.

 All my friends think Im crazy for leaving all the emissions components and also running a highflow cat on my modded car.  My explanation to them was, for the measly 5-10 peak HP loss I may get is not worth feeling dizzy over to drive my fun car, which is what I experienced when I disconnected the EVAP stuff for diagnosing issues during the swap. 

Guess Im keeping my old CARB legal header and intake for future use just in case...

In reply to aw614 :

All manifacturers discontinue parts.  Legally they only have to meet warranty obligations, they don't have to stock anything at all.

NLA parts for 8 year old cars is common.

Yeah I know, but one can only wish to keep their old running car clean without having to take a chance buying used old parts lol.

I live in Houston.   For the annual inspection they plug  into your OBD2 plug and take a reading.   Pass or fail.  The computer reading goes into a state database instantly - you cannot renew your registration until you pass the test and the notification happens immediately.

If you don't have OBD2 and the vehicle is 1996 or later you fail..   No sniffing the tail pipe emissions to see if your "special tune" in your Magasquirt complies.  

People run two systems, one with their stand alone and then, for inspection day, switch over to OBD2, pass the test, then switch it back.  There are/were shops that catered to this market, doing the switch over in house, running the OBD2 test though the State's computer, then switching it back.

  I almost bought a turbo miata with this setup then decided it was too much trouble.  A few months later I found out that the shop that catered to this market, the one the seller  had used for 5 years,  decided to quit that service.  Maybe his arm was twisted or maybe he realized he was in danger of getting a big EPA fine.

My point of all this is the missing piece is a method of getting your vehicle emission certified outside the boundaries of your OBD2 system.   Surely this is possible without the OEM computer hook up for  OBD2? 

In reply to jharry3 :

When Megasquirt or another standalone can come out with all of the onboard monitoring, then sure.  The monitoring is the key part.

This is why it is generally better to just work within the existing computer, if at all possible.  OE computers are far, far better.

jharry3 said:

 

My point of all this is the missing piece is a method of getting your vehicle emission certified outside the boundaries of your OBD2 system.   Surely this is possible without the OEM computer hook up for  OBD2? 

The problem is that On-Board Diagnostics is part of the emissions system. If you disable that, you have disabled an emissions component. Passing a simple roller test on a hot car is easy, but OBD watches the car under all conditions and will detect failures that don't show up in a roller test. That's why you're not going to see a legal way to bypass it. 

The correct option, as noted, is to come up with an ECU that retains OBD. That hasn't been done yet. The hardware is certainly capable - heck, we had aftermarket ECUs running the charcoal canister nearly 20 years ago - but I expect the testing would be pretty heavy as you'd likely have to prove that the system can detect all potential failures. 

The aftermarket is way behind the OEs when it comes to computers and tuning and testing. As Knurled said, it's a lot better to start with a factory calibration and modify to suit. That leaves all the hard stuff - the edge cases like high altitude and bad gas and cold weather - in place. I know a few OE calibrators and you would not believe what it takes to develop a factory cal. 

This leaves a certain range of cars in a bad place. Cars that are newer than 1995 but that don't have reflashable ECUs. That's roughly a decade of production. You either have to use an auxiliary controller (as seen on our turbo kits for the NA/NB) or...well, that's about it.

aw614 said:

But one thing that bugs me is some of the OEM car companies are discontinuing their emissions components. I had some weird EVAP issues on my integra, when I still had the stock ECU and motor and a lot of the stuff was honda has been discontuning a tun of parts at an even quicker rate lately. Good thing the part I bought was still available from denso, but stuff like original hoses, charcoal canisters were, etc were NLA.  I wonder if there will be companies willing to remake these components again if it is a viable business option.

There certainly are. Those who drive mainstream cars won't have a problem. But specialty cars may have a real problem here. Replacement parts are either going to be expensive due to low volumes or they're going to be NLA. This, of course, is true for all parts of a car, but a lack of emissions components may take a car off the road permanently. You may be able to adapt a part from another car if it keeps the OBD system happy and this likely be acceptable to all involved as a one-off unless you run afoul of a California visual. Even then, you may be able to get approval through a referee - they want to see people keeping cars clean and they want to keep cars on the road.  

Keith Tanner said:

but I expect the testing would be pretty heavy as you'd likely have to prove that the system can detect all potential failures. 

You do. You have to be able to prove that the correct fault (and only the correct fault) is triggered when a part failure occurs, and that it's triggered within a certain time limit after the failure. And you have to do that for every emissions related fault. And you'll be testing for emissions at the same time by running the same emissions test cycles that are used for certification in order to verify that emissions don't exceed a predetermined amount for that specific failure.

Something like an O2 sensor can have several faults that need to be tested/confirmed, with each confirmation testing procedure taking from several hours to a couple of days. If everything is working properly, just the confirmation tests can take several weeks and require lots of part swapping between known good and bad parts that have all been approved by CARB for testing. That doesn't include development time/testing. The billing rate is about $350/hr where I work, and that's just for running the testing. It doesn't include any engineering time, etc.

Keith, that depends on the car manufacturer.  I've never had an issue flash tuning a BMW/VAG/Porsche from the early OBD2 era. 

Of course. I tend to focus on one particular manufacturer :) The transition to rewritable ECUs (and ones that can be accessed by the aftermarket) is going to vary. The Miata was always fairly unsophisticated  until the ND came along, so it was probably one of the last that couldn't be reflashed. Starting in 2019, we're locked out of the ECU again.

This was not discussed, but we may be in a golden age for emissions legal power. OEs are trying to lock down their ECUs, with Chevy claiming the C8 is uncrackable or unmodifiable (which is not the same thing). That is an ECU that will see a very concerted attack by the aftermarket, so it may not stay that way. But other platforms may not be worth the effort.

STM317 said:

Keith Tanner said:

but I expect the testing would be pretty heavy as you'd likely have to prove that the system can detect all potential failures. 

You do. You have to be able to prove that the correct fault (and only the correct fault) is triggered when a part failure occurs, and that it's triggered within a certain time limit after the failure. And you have to do that for every emissions related fault. And you'll be testing for emissions at the same time by running the same emissions test cycles that are used for certification in order to verify that emissions don't exceed a predetermined amount for that specific failure.

Something like an O2 sensor can have several faults that need to be tested/confirmed, with each confirmation testing procedure taking from several hours to a couple of days. If everything is working properly, just the confirmation tests can take several weeks and require lots of part swapping between known good and bad parts that have all been approved by CARB for testing. That doesn't include development time/testing. The billing rate is about $350/hr where I work, and that's just for running the testing. It doesn't include any engineering time, etc.

The one thing to note is that the OBD requirements changed over time- so for Keith and FM, the 96-98 Miata will be easier than the 99-2000, which will be easier than the rest of the NBs.  For the most part, the OBD requirements followed the development of the emissions that the car met.

So if anyone was going to attempt to do an aftermarket OBD, the best place to start in '96- when the requirements were less stringent.  Mind you, hardware and software developments now make it easier to do than it was for us 25 years ago.  But it's still going to be very hard.

As for emissions testing, if I were a FM, who had a chassis dyno (I think), I'd buy a Portable Emissions Measurement System- and use the combination of the PEMS and dyno to do all of your development in house as much as possible.   Saves money, and builds expertiece.  

We have a five gas analyzer and two chassis dynos :) Is that what you meant by a PEMS?

BTW, here's an example of what happens to companies found in violation by the EPA. This is Derive Systems, aka Bully Dog and SCT tuners for diesel trucks. $300k fine (chosen for ability to pay) and a bunch of products taken off the market. Bully Dog is still around, but you can't use their stuff to delete your EGR and DPF anymore.

Summary

Consent decree (for the lawyers)

I expect our friends at DIY Autotune is paying close attention to what's going on.

We are. Thank you for posting these notes, as they contain a few new developments I hadn't heard of.

This sounds like it would affect a lot more than just electronics. For example, suppose I wanted to call up Summit and order a more aggressive cam for my 1966 Dodge Dart project. All perfectly legal - the Dart predates the point where the Clean Air Act went into effect. But this is a cam that could easily be put into, say, a 1976 Dodge Aspen - which would fall within the EPA's purview. And there is no EO for this cam - or any other aftermarket slant six cam that I'm aware of. So, how would Summit be able to check that I'm using the cam for something above board?

Summit's website lists over 9,000 camshafts.

37 of them carry CARB EO approval.

In reply to alfadriver :

It would seem to me that the easy thing to do would be to use a pre-existing computer that can be hammered into the application.

R34 Skylines were being Federalized with the help of OBD-II computers from 300ZXs, for example.

MadScientistMatt said:

I expect our friends at DIY Autotune is paying close attention to what's going on.

We are. Thank you for posting these notes, as they contain a few new developments I hadn't heard of.

This sounds like it would affect a lot more than just electronics. For example, suppose I wanted to call up Summit and order a more aggressive cam for my 1966 Dodge Dart project. All perfectly legal - the Dart predates the point where the Clean Air Act went into effect. But this is a cam that could easily be put into, say, a 1976 Dodge Aspen - which would fall within the EPA's purview. And there is no EO for this cam - or any other aftermarket slant six cam that I'm aware of. So, how would Summit be able to check that I'm using the cam for something above board?

Summit's website lists over 9,000 camshafts.

37 of them carry CARB EO approval.

Given that I work exclusively in a post-Memo 1A world, I hadn't looked at that possibility. It's a sticky one. Would it be enough to verify ownership of an older car? If I were the manufacturer of that cam, I'd consider contacting the EPA and asking them. There will not be easy answers. The good news is that that's pretty far down the ROI scale, it's a lot easier to go after the guys making devices specifically aimed at removing emissions devices such as O2 sensor simulators or tuners. But eventually, that might be a problem. Cam manufacturers can either try to get ahead of the curve (har har) which might cost them sales in the interim, or they can wait until one of them gets a visit.

Drop me a line if you have any specific questions about what I learned. Hopefully DIY Autotune is a SEMA member, the SEMA Garage is quite helpful. They also published a good overview of the whole EO process including what all the tests look like that is interesting reading, I'll see if I can find a PDF version.

Knurled. said:

In reply to alfadriver :

It would seem to me that the easy thing to do would be to use a pre-existing computer that can be hammered into the application.

R34 Skylines were being Federalized with the help of OBD-II computers from 300ZXs, for example.

Different hypothetical.  But I'm not sure you can get an EO for a 300XZ in a Skyline....

The idea is more about a full emissions package for *something*- for instance a turbo, cam, injector system that makes huge power, where no OEM computer would be capable.  If something like MS is capable of doing the right year's OBD, most of what they are capable of doing is more than enough to come up with a powerful system that can meet the emissions requirements.  At least how I see it- add a specific catalyst for that, do the testing, and prove that you satify the requirements.

A hypothetical.

Or if OEM parts start failing.  

Something like that.  W

When I read Keith's first post, that's what I saw as an opportunity.

Keith Tanner said:

We have a five gas analyzer and two chassis dynos :) Is that what you meant by a PEMS?

Unless it's really portable, not really.   

I'll get the model number and drop you a line.

In reply to alfadriver :

Like people trying to make Bosch injectors work in early Mazda DISI engines for availability reasons?

Keith Tanner said:

Given that I work exclusively in a post-Memo 1A world, I hadn't looked at that possibility. It's a sticky one. Would it be enough to verify ownership of an older car? If I were the manufacturer of that cam, I'd consider contacting the EPA and asking them. There will not be easy answers. The good news is that that's pretty far down the ROI scale, it's a lot easier to go after the guys making devices specifically aimed at removing emissions devices such as O2 sensor simulators or tuners. But eventually, that might be a problem. Cam manufacturers can either try to get ahead of the curve (har har) which might cost them sales in the interim, or they can wait until one of them gets a visit.

Or consider the case of somebody who is drag racing the '76 Aspen instead. What sort of proof would Summit need to collect that it is a race car before shipping the cam? Howard Cams may not be a big target, but Summit and Jegs are more obvious targets (and I know Jegs has been hit at least once before).

Drop me a line if you have any specific questions about what I learned. Hopefully DIY Autotune is a SEMA member, the SEMA Garage is quite helpful. They also published a good overview of the whole EO process including what all the tests look like that is interesting reading, I'll see if I can find a PDF version.

Thank you, I've passed this offer along.

In reply to MadScientistMatt :

I wonder if Mazdaspeed's requirements would be sufficient.  To buy from them you need to submit race results from the previous twelve months, for the year/model that you are buying parts for.

Interesting discussion.  The responses here usually portray one side of the issue or the other; you can either be an individual-rights gearhead or an environmentalist, because those are the perspectives of the people that are shouting the loudest.  Personally I think that this is a false choice.

I own a 1992 Miata that cost around $20k when new.  I've invested another $20k and hundreds of hours in performance upgrades (turbo, ECU, etc.)  so it could perform about the same as an unmodified 2019 Miata, which costs about the same as a new 1992 Miata when adjusting for inflation.  The new one is way more reliable, and way cleaner.  This is true of just about every car out there, the performance/$ ratio is shocking compared to the old cars.  There are also hundreds of ways to improve you car's performance that don't involve messing with the emissions systems.  I had fun building that Miata.  I also view it as a completely irrational use of time and money.

We all benefit from clean air, and have to get used to the fact that the ways of the past need to change for the benefit of everyone.  You can't start with a standing assumption that all government intervention is wasteful and ineffective.  There are thousands of examples to the contrary, all you have to do is look.  I grew up in Pittsburgh in the 1970's, I'm all too familiar with the consequences of poor environmental practices.  It is possible to be a car/performance enthusiast and be a responsible citizen at the same time.

Knurled. said:

alfadriver said:With all of the experience in the OEM sphere, if that manages to pass it's way into the aftermarket (which it does), there will be innovation.  It's not as if solutions don't exist to make a very modified car fully compliant.  

 

HKS had a CARB EO'd single turbo swap setup for FD RX-7s a long time ago.

Do you remember which kit?

In reply to amg_rx7 :

I'd like to say the T51 kit?  Might have been the Greddy T78 kit.

Finding info about 15 year old turbo kits is hard nowadays.