951 High HC/NO

[vetsin]

Hey all, was wondering what step's ya'll may take to get the emissions of a 1988 951 down. It only fails at 15 and not 25 and is known to pass when very hot.

The guy I got it from did a bit to it -- including new injectors/waste-gate/fuel pump/etc. Was hoping it could be as simple as a new cat, but it seems unlikely after talking to some local shops.

emissions.png (485.09 KiB) Viewed 858 times

[whalenlg]

I had HC=182/155 and NO=1917/1823 before cat change.
id the basic things like new oil/plugs, distributor, coil, spark plug wires before changing the cat and they didn't change the results.

After changing cat:
HC=25/14 and NO=2/1

[vetsin]

Sounds worth trying, then. What did you go with, stock?

[whalenlg]

I had to find one with a Calif eo number:
Walker EO 81128

[dr bob]

From some experience --

Clean the injectors. You'd love to get a gas sample from upstream of the catalyst to prove a problem of course.

-- High CO upstream of the cat indicates a rich-combustion condition.
-- High HC upstream of the cat indicates a misfire, allowing fuel to pass through without burning.
-- High NOx upstream of the cat is a function of high combustion temps.

High HC plus High NOx tells me you have a lean misfire.



I chased a gradually increasing NOx reading behind the cats. The EZK box on my 928 controls spark timing, so no easy way to reduce ignition advance to lower those combustion temps slightly.

Solution was to clean/replace the injectors. The LH fuel injection controller reads a single common exhaust sample just upstream of the cats, and adjusts the batch-fired injector duty for all based on that sort-of average. The sensor functionally reads the CO in that sample, so reads gets the average fuel mixture and adjusts all the injectors at the same time. There was just enough flow imbalance among the cylinders to cause just a few to run slightly leaner, and those cylinders were generating the extra NOx.

I took a series of exhaust manifold/port IR-gun temp readings to see the deltas. The variations didn't seem enough to cause the extra NOx. But NOx is tied to a specific combustion temp threshold, below which there isn't much. But a 30º increase in flame-front temp causes a pretty significant spike in NOx generation. Anyway, new clean injectors brought the ~~100k NOx numbers back to what they were when I bought the car with well under 20k on it at 8 years new.

On the way, a lot of maintenance was done. New plugs, wires, caps & rotors, knock sensors, and CPS. All seemed fine, and no gains at the tailpipe as things were replaced.

The 928 comes with sample tubes from each header collector up into the engine bay. Grabbing pre-cat samples is a breeze. I worked with a neighborhood smog station to get some readings along the way, which was interesting.


But that's on a 928...?? Combustion parts are pretty similar it turns out. I just have twice as many.


Injector cleaning is inexpensive, certainly a LOT less than buying a Cali CARB-compliant catalytic converter. IMO, certainly worth the look.

[cda951]

Sorry to say, but both high CO AND NOx at 15 MPH indicates the catalytic converter is the issue. The oxygen sensor system oscillates between rich and lean of stoichiometric to reduce CO and HC emissions (lean mixture helps this) and NOx (richer mixture helps this). If both are high at a lower RPM/reduced exhaust gas temperature, I think we have the answer . . . .

[Tom]

Sorry to say, but both high CO AND NOx at 15 MPH indicates the catalytic converter is the issue. The oxygen sensor system oscillates between rich and lean of stoichiometric to reduce CO and HC emissions (lean mixture helps this) and NOx (richer mixture helps this). If both are high at a lower RPM/reduced exhaust gas temperature, I think we have the answer . . . .

CO looked 'ok' in the chart, no? Did you misread, or think the .63 is still too high even though under the limits? It looks like it had high HC and NOx on the 15 mph run. I'd say that points to a lean mixture with missing, which is worse when they are lugging the motor mercilessly at 15mph. I once logged my CA smog test and confirmed the motor is under more load at 15mph (see below). Since this car did so much worse at 15mph, I'm thinking vacuum leaks and maybe weak ignition. dr bob's angle with the injectors is aimed at the same lean condition, but if the injectors are new as you say --and correct for the car!? -- then I'd start with the odds-on 951 favorites of vacuum leaks and weak ignition. What vacuum do you see at 840rpm when the car is fully (fully) warmed up?

I once had a chance to tune a 951 real time on a live CA dyno test with no cat installed. The car had big injectors and a MAF too. We were able to get any 2 out of 3 numbers under the limits, with the third always being just barely above the limits. So while a new cat will mask a LOT of issues and lower the numbers on just about any car, I would be inclined to triple check for vacuum leaks; and change the spark plugs, cap, rotor, wires and probably 02 sensor unless you know for sure they are fresh. You can always try a new cat once the car is running in tip top shape if it still doesn't pass.


Log from actual CA dyno test on a 951... note the vacuum reading on the left (15mph) vs the right (25mph)...

IMG_0200.gif (62.87 KiB) Viewed 762 times

[vetsin]

Honestly easier to just change it all out than run to a dyno a bunch -- unless there's some way to measure this stuff "at home". Well, beyond testing a vac leak.

Honestly this does seem a little black magic as everyone has slightly different opinions.

[Belgian951]

Following. I'll soon have to get a smog test to verify I can pass inspection in another year's time. I passed it but then had an issue with the cat, which I had to remove. Now I'm unsure I'll pass smog again.

[Tom]

Honestly easier to just change it all out than run to a dyno a bunch -- unless there's some way to measure this stuff "at home". Well, beyond testing a vac leak.

Honestly this does seem a little black magic as everyone has slightly different opinions.

The way to measure it at home is a wideband O2 sensor before the cat. It will tell you if the car is running rich or lean; if the factory O2 sensor is faulty, and if you are in closed loop at 14.7:1. I struggled for years to get my '79 Mercedes to pass smog, and finally put a wideband in, and was able to see exactly what was happening and pass smog on the first try. No black magic needed that way...