John Pirtle wrote to the list about his 87 AT emissions test
results:

>> Reading Allowable
>> HC-ppm 71 137
>> CO-% 0.13 0.87
>> NO-ppm 948 956
>> RPM 1304 100-2500
>> CO+CO2% 14.4 6.0 min.
>>
>>
>> What would cause the NO-ppm to be so high?
>>
>> Could this be a result of a not-fully-warmed engine? The car had sat at work for about 4 hours, then I sat in traffic getting out of the parking lot and drove about 3 miles to the shop that did the test. The coolant temp gauge read normal, but maybe the catalytic converter was not fully warm? Ambient temps were about 80.
>>
>> I've read through old emails, and the Tips, but most of the discussion was on HC. Hey Dr. Bob, can you discuss Oxides of Nitrogen for us?? :)

NOx emissions are the result of high combustion chamber temps pure and simple. NOx readings go up whenever there are hot spots that exceed about 2400 degrees. Things like high compression and lean mixtures contribute to high chamber temps, as does over-advanced ignition timing.

There isn't much we can (or want...) to do about the high compression pressures in the 928 engine. Lean mixtures are usually caused by some failed components or by air (vacuum) leaks in the intake. Ignition timing on the S4 is managed by the brain trust in the passenger footwell, and isn't adjustable except by some re-mapping of some lookup tables in the controllers.

With these three mechanical things managed, some other factors contribute. Poor fuel quality contributes to detonation, a favorite cause of high chamber temps. Got a load of pretty poor gas? That will do it easily. Worn plugs allowing a little misfire? Not a problem unless the worn electrode is glowing and causing early ignition, and none of us true enthusiasts would leave old plugs in the engine. Oil contamination of the intake charge can cause an effective reduction in octane rating of the fuel. Adding a little exhaust gas back into the intake has the opposite effect by the way, effectively reducing the charge density
in the given volume in the combustion chamber. That's the EGR controller and valve that your tech is referring to as a
possible cause.

>> The mechanic said high NO usually points to a bad EGR or cat.

The cat depends on some reasonable operating temps to get a good reaction, and a bout of sitting in traffic or waiting
in line for a pass at the rollers might allow the cat to cool enough to stop the NOx conversion process. The principle component that heats the cat is HC, where the raw fuel that passes through the engine unburned finally gets a chance to do some work heating the rest of the catalyst media. With the excellent engine management and the feedback loop from the oxygen sensor keeping the HC and CO in the pipe to a minimum, the NOx reaction is probably reduced.

What to do:

Since your car actually passed and you are unlikely to have it tested again soon, I wouldn't get too excited for a while. In the meanwhile though, check to make sure the vacuum system is tight, and that may include fixing the vacuum hoses that are keeping the flappy from working correctly. The air leak may be just enough to give you NOx, and the "curve" for NOx is very steep once you get the temps up.

The upgraded performance chipset you might have installed will show a bit more NOx in those slow engine speeds and low
load ranges where the gas samples are taken. Got one of those? Take it out and replace with stock for the next test.

-- Use one of the proven octane boosters next time you test.
-- Have the tester do the rolling tests with the trans locked in first gear. Might be enough extra charge flow to get better chamber cooling. My tester guy didn't want to do that, but maybe a PKS would do the trick next time.
-- Make sure the oxy sensor is working and relatively new.
I haven't changed any on this car, but know that a new sensor pair on my Explorer paid off in improved mileage in less than eight tanks of fuel with 60k on the original pair.
Payout might be even quicker with the higher gas prices lately.

After those simple things, you may be looking at needing a new catalyst.

Another response mentioned looking at all the hoses from the air pump to the cat. Good idea, as is a look at the pump
itself to make sure it's working OK. Air is needed to complete the HC reaction, and the heat from that reaction contributes to better function of the rest of the catalyst.
The pump supplies that air.

Coincidentally, I just went through the testing with my '89 AT in California, and FAILED (!) on NOx the first time. I had never had any problem passing before with almost no readings at all on HC, CO, or NOx. I still have the zero readings on CO and HC, but the NOx was in the 1500ppm range this time. I restored the stock controller chips (were Autothority), added a bottle of 104+ to the tank, and pretty much rolled right on to the dyno for the second test from a blast up the street and back, yet still managed to reduce NOx barely enough to pass this time at 740. So I'm looking at the same problem you are facing.

My winter project list included pulling the intake for refinishing, and replacement of all the hoses under there including the vacuum and the smog pieces. Maybe I'll find something cracked or leaking in there. I'm thinking of using some stainless instrument tubing and high-temp silicone hose connectors for all those pieces that are otherwise inaccessible, including the flappy actuator hose.

Gotta find a good source for that foam that goes under the fuel rail covers too. My old stuff is turning into a grey-green crumbly mess. I may end up whittling a piece of that black hose insulation to fit there, or maybe some hi-temp neoprene cable jacketing or something. Too much to do and not enough time, that's always the problem. I may need a new 996 turbo-look cab soon so I can have a Porsche without a do-list attached... Just kidding, my list just has PM and cosmetic stuff on it so far.

Good luck!
Bob