Exhaust smell after header install
 

Hey guys so I have an issue and it’s driving me nuts. My car definitely smells like it has an exhaust leak after installing my JDL header. I installed the header and the next day I had a 13.5 hour trip. I could definitely smell exhaust fumes inside my car. A little after 4 hours into my trip I stopped at a shop and I had them take a look. They said everything looks great and I could do a smoke test for 90 bucks. I said yeah let’s do it... it passed the smoke test with no leaks anywhere. Guy at shop said it could be the anti seize I am smelling and cleaned my header of the little bit I got on it. Well I’m chugging along all on the interstate and when I wide open throttle I can definitely smell it. I froze my ass off with windows down because I definitely got a headache from the fumes. This isn’t my first header install on a car. I’ve never had this issue before. I know the header has a cat but almost every single manifold I’ve replaced was catted and I swapped it out with a cat less header without ever smelling the exhaust inside the car. The other day I went to start my car outside to let it warm up with windows up. Mind you this is outside not in a garage and when I got into my car it smelled awful.
The header is JDL ceramic coated header I bought it new, also installed grimspeed gaskets. I torqued everything to spec and I did the heat cycle and torqued it down again. Only other mods are a greddy cat back exhaust but I had it prior to header and never smelled anything. A Perrin inlet tube and K&N filter but once again never smelled anything till after the header install. I also upgraded from oft v1 to v2+ tune. I know you’ll get a little smell from taking a cat off but not enough to smell it inside your damn car when you idle it for 5 mins with windows down. Any advice would be greatly appreciated. Because I’m lost and if I can’t figure it out I’m swapping back to the original manifold and selling the headers.

is your cat clogged or burned out from running too rich?

I've dealt with this on another car.
Sometimes people like to get under the car while it's running and try to feel air coming out of possible areas that leak but that can be hard at times so try this.

1. Remove the header
2. Look at the gaskets (both the ones going to the heads and the one going to the overpipe). If they are sealing correctly you'll see black soot in a perfect circle. If it's leaking you'll see it on the gaskets spraying out of the perfect circular shape you'd hope for.

Since the header installation didn't happen too long ago not enough time may have passed before seeing any soot on the bottom of the heads or surrounding areas. The gaskets would show the leak first. So I'd remove it and examine the gaskets. Maybe buy new ones before doing this so you can put fresh ones on. It's worth noting I've NEVER had a problem with stock Subaru gaskets.

Also check the welds, there could be a bad weld.
In addition, check the overpipe to the front pipe connection.

Regarding the smoke test...where did they supply the smoke? I have another question but I'll let you answer first.

Create bubble water and spray over header to find the leak.

This will work. But I have had success getting under the car, but you gotta feel around. Helps to get someone rev while you are under there. Make sure it’s on jack stands or ramps and safe.

Also do you have the sound generator removed. Did you plug the hole to the inside of the car. Some people have gotten some smells inside the car through that hole.

Yes the sound generator is removed. I didn’t plug the hole. I’ll look into that!

I’ll definitely look into that. I had a 890 mile trip after I installed it. I’ll order new Subaru gaskets and hate my life taking it all apart again but I totally see what you are saying! I did torque the header bolts to the 22 pounds. The 02 sensor I just tightened till it felt good. I could always check the 02 to make sure it’s not leaking I just thought of this.

The smoke test they did it via tail pipe I believe. They said it pinged back with no leaks anywhere or something like that I can’t fully remember. It was a few weeks ago but I was exhausted from a 26 hour round trip in 3 days. Long story short I’m working in Maine and my car was in Ohio. I decided to stay in Maine during all this corona going on. I drove a rental to Ohio to grab my brz. While I was home I installed my header I ordered and a few other odds and ends inside. Was hoping to get my coilovers on but I was too tired and no place was open for an alignment.

I chased an exhaust leak for a while. I'd try pulling the gaskets and looking for signs of a leak. Pretty easy to see where the soot is escaping.

Since you have an OFT, you can do some datalogs and see how your fuel trims are doing. Large positive fuel trims (LTFT) at partial throttle can be indicative of a leak located near the O2 sensor.

I'd also verify that it is indeed an exhaust leak and not something else. The stock header, OP and heat shields can catch oil dripping off the rear cam cover. I remember when I replaced the header/OP with non-heat shielded versions, I could smell the oil burning on the bare exhaust parts once they heated up. The fumes would go directly to the hvac and would enter the cabin. Different smell than exhaust though, more caustic and less fuely.

I’ll look into that on my way home. I did get the ceramic coated header for heat reasons. The smell isn’t burning oil. It’s definitely exhaust for sure. I’ve had my fair share of Honda’s that burned oil in vtec and that definitely isn’t the case here.

Is it a catless header? It will smell more just because your cat efficiency is so much worse, the primary cat does most of the work after all.

There is a technique to increase the effectiveness of finding a exhaust leak using soapy water.

Plug your tailpipe or duct tape to seal it then shoot compressed air while you look for leaks. Or plug it up or restrict the exhaust while its idling.. the increase in pressure will make the leaks more obvious to find.

When I installed my JDL headers earlier this year they smelled really bad for about the first week of driving. The smell is completely gone now.

From what I've read it takes the ceramic coating several heat cycles to fully cure.

The main cat does most of the work. The precat is not necessary once the main cat heats up, but you are right about the smell. It should smell for 10 minutes (depending on temp/driving conditions) until the main cat fully heats up.

^^^ This.

Use a feather or something that is easily effected by a wind current.

If you use your hand to feel for an exhaust leak then spray alcohol on your hand. The air from an exhaust leak will be more noticeable kind of like how someone will lick their finger before putting a finger in the air to sense the direction of the wind, say for a field goal kick.

Exhaust smell after header install
 

The cat in the header IS the primary cat and does most of the work. The one in the front pipe is secondary.

The smell does not go away after 10 minutes with a catless header. I personally hated the smell and ended up going back to the stock unit cause it wasn’t worth the 10hp bump over catback and 93 tune.

I believe you are mistaken. The main cat is the one in the front pipe. The precat is in the header. Koenigsegg actually has a bypass valve that divert air around the precat once the main cat heats up.

Exhaust smell after header install
 

That’s really cool, didn’t know that, but alas this isn’t a Koenigsegg. CvK does a lot of things differently because he can afford to, and the cars aren’t held to the same regulations because of the extremely low production numbers.

Our primary cat is in the header. That’s why there’s an O2 sensor right before and right after it, and the front pipe does not have an O2 sensor.
https://www.ft86club.com/forums/showthread.php?t=71412

Maybe I missed it, but I didn't see anything in the link that shows something credible about which is the main cat.

https://en.wikipedia.org/wiki/Catalytic_converter

It is best to put the catalytic converter away from the headers because it will have a reduced negative effect, but by putting the main cat downstream there is a problem with heating it up quickly because exhaust temperatures drop quickly as they leave the combustion chamber. Manufactures will use a precat on the header to solve this problem. By putting a catalytic converter close to the headers, it heats up faster, so it is able to reduce emissions, but then why not just have a single converter? If you look at the pre-catalytic converter on the headers and compare it to main catalytic converter, typically the main cat will be denser, and the precat will be less dense, so it isn't too restrictive.

The O2 sensor locations are typically coming right off the exhaust manifold (so they can read oxygen for fuel mapping) and between the precat and main cat. Some cars have them after the main cat as well, but the precat is much more likely to fail because of the heat/proximity to the combustion chamber in terms of melting (so it is the best one to monitor), and the whole point of the o2 sensor is to detect oxygen for fuel mapping. The downstream sensors can detect catalytic health, but it is not their primary function.

If you could show me a picture of the cat on the headers and the cat on the front pipe then I would accept what you say.

This explanation was a nice read:

https://mechanics.stackexchange.com/...ixture-control

Go unbolt the exhaust on your car if you don’t believe what has been discussed hundreds of times on this forum. You can search and find dozens of threads where the primary/secondary cats are discussed.

Everywhere I've ever heard has stated the primary cat is in the header.

Too much work. I’ll probably find a picture eventually of the cats.

As in primary and secondary or 1st and 2nd? Because then the precat is the first one.

As in primary is in the header. I think @CSGmike could answer that better. But I've always read the primary is in the headers.

That's right. Or it just might be the oils and chemicals from your hands and manufacturing process. A brand new car from the dealership does the same, bu more burnt plastic smells than oils.

I'd still check for leaks anyway and retorque after a couple heat cycles.

Like I stated, if the pack density is different then we would know. If it was identical then it would depend on the materials inside. Assuming both were identical in volume or close enough in volume then the first one would naturally do the brunt of the work, but to truly know, we would have to remove one and do an emissions test then swap and run that test. If both showed an equal drop once up to operating temp then they are the same, but the first one does the brunt because it is first. If one did more by design then that one is the main one.

Emissions standards require the Primary cat to have an upstream and downstream O2 sensor. Ergo.....

Not trying to get smoke to come out of your ears, but holy crap how many people need to tell you the same thing. These cars aren't Koenigseggs, they're not 90s MR2s either. They're not EJ25 Subarus. It's been long established that on these cars, and the vast majority of NA cars made in the last decade, the main (primary/supreme/heavyweight/wearing the pants in the relationship) cat is the one in the exhaust manifold.

Lol that is funny shit...

It depends on who is telling me more than how many. There is a big difference too between making statement and providing evidence or explanations. EPA official? Work at CARB? Design cars for a manufacture or emissions equipment? I would love to see the law or a link. I wasn't aware there was a mandate.

The 3rd gen MR2 quoted was like 1999-2007. Just saying. Kinda after OBD2 adoption and not that long ago.

I mean it makes sense what you are saying, that they would just use two high density units because why not remove as much pollutants as possible, regardless of detriments to performance.

https://en.wikipedia.org/wiki/Cataly...r#Installation

I am curious if tuners use both o2 sensors for tuning like the OEMs, and I also wonder if the secondary cat is negatively effected in its ability to work by installing headers. I wonder if installing headers leads to premature failure of the other cat, and/or if the secondary cat can't do its job without the primary cat or without the ability of the oxygen sensors to properly manage the fuel delivery. I understand people often use o2 foolers, but it seems like those would provide some fixed value and not a dynamic value, which means the engine isn't getting the best feedback, which could make the engine run rich/lean.

I have a Harrop kit with E85 on a Delicious Tuning tune, so I'm curious if their tune supplements the stock tune, or if it entirely replaces it, and if so, is it not using the second sensor or is it not considering emissions at all. I feel like it much be not altering things too much because when my engine is started while cold the rpms raise and bla bla bla just like normal. Still.

Lol ok this is becoming comical. You don't want to be wrong, and that's fine. But you keep talking about how you won't believe what others have said. So I think I'll follow suit and accept what I've read countless times unless someone from CARB, Toyota, Subaru, or Tcoat says otherwise. :cheers:

The oxygen sensors are on the header. Engine does not even care if the second cat is there unless it is clogged. I have replaced my front pipe with HKS dual resonated without cat. I have not even changed the tune; no CELs, no problems, actually engine runs better, revs up easier, higher mpg with normal driving. I agree that the primary cat is on the header. I think the second cat is only there to reduce emission while both cats are warming up. My exhaust smells a little bit more (no different then our Honda Odyssey) in the first two minutes during warm up and there is no different smell after the only cat on header warms up.

I suppose the question is whether the primary and secondary cats are doing independent jobs and work as a system, or is it the case that one is doing more than the other. It sounds like you guys are saying the cat in the header is doing all the work, in which case, the secondary cat is either redundant, but capable of the same job, or redundant, but incapable of the same job.

It would be cool to do an emissions test:

-Stock
-header only
-front pipe only
-neither

Provided a situation where both cats are warm, if emissions equally dropped then both cats are identical in function and the second is redundant. If emissions dropped more with one than the other then it would be clear which one is primary. If NOx dropped more in one and less with the other, and reciprocally, if CO and HC dropped more in one and less with the other then that would suggest the two have unique rolls.

Ah yeah. I have the HKS R-Spec front pipe which has their "super catalyzer" and it does a damn good job. Of course there's still smell when it's warming up. But after it's warm there is little to no smell! It also has bung behind the cat if you wanna route the second sensor behind it.

I would like to see a study like that. It would be nice information.

Actually, I really don't mind being wrong if someone can demonstrate it. I think it is entirely plausible that I am wrong. I would readily prefer to toss any erroneous beliefs/information I have, but it is harder to do that if what knowledge I do have is conflicting with new information that is not substantiated. :cheers:

You have a catless header, right? I kept the stock header but just replaced with 2017+ improved one since I wanted to keep the JR factory tune. So you have eliminated the first cat, I eliminated the second one.

Nope I didn't say that. Second cat helps with emissions especially when the cats are cold. But the first cat is effective enough alone after warm up.

Yeah I have the ACE CS400, but the catted front pipe. It seems to make more sense doing the headers first. Had I not already bought a front pipe a while back, I'd probably just run the oem.

Using the 2017+ header, did you have to change the JR tune any? are you running the CARB or a custom tune?

That is a very nice combination, I hope you are enjoying it.

I am still running the original JR CARB tune. It works just fine with 2019 header since about one year now.

People have such trouble admitting to being wrong. It's okay to be wrong man.. it happens all the time lmao

What smells are you smelling?

Carbon monoxide has no smell, so if the catalytic converter wasn't turning CO to CO2, you wouldn't know by the smell.

I suppose if the hydrocarbons (gas) was not converted to CO2 then that would smell like gas, and if one thing is working then it is likely the other is working, but this would be an assumption.

NOx should smell like a diesel truck.

Rich mixtures have more fuel and more fuel means more sulfur and sulfur, from what I understand, can be just a natural part of the 3 way catalytic converter. It may be that all or some converters have coatings to reduce this odor.

https://www.nettinc.com/information/...e-way-catalyst

You know, typical exhaust smells when you have a catless header lol