NOx is produced primarily during cruise and light acceleration at the controlled stoichiometric A/F ratio. Very little is produced at idle or at WOT if the engine has full load enrichment, which most do, but WOT would not be required on the IM/240 test unless the car is severely underpowered, which would not include a 325i. Thus the idle speed setting has little or no effect on NOx. Even a small vacuum leak will be compesnated by the control system as the O2 sensor will detect a lean mixture and richen it to maintain stoich. until it reaches its limit of control authority.

I don't know the details of the 325i idle control system, but most have an adjustment that should be set to a nominal idle speed value with the idle control system disabled. On the M103 engine this is the throttle stop on the throttle valve shaft under the fuel distributior. If this initial setting is too far off the system may not have enough control authority to maintain proper idle speed. As a rule, one should not mess with this initial setting unless you have proper service documentation and equipment to perform the procedure properly.

The fact that you replaced the O2 sensor means its probably not the culprit and the change did not affect the readings that much, so the old one was probably okay. It would not be unusual to test the car twice in two days and see readings at least 10 maybe even 20 percent apart.

There's another issue that most car owners aren't aware of. Test standards are being continuously lowered, and established test standards have no precise correlation to original certification standards. It's possible that your car is functioning perfectly less some degradation of emission control components such as the catalyst. This is an issue you have to bring up with your state emission test authorities. If they are like CA, they may be uncooperative, in which case you should contact one of your elected representatives. I have and will undoubtedly do so in the future.

Questions you might ask are:

What are the average emissions for this year and model car?

How do the test standards correlate with the original certification standards? How much degradation is allowed?

My 190 is on the ragged edge of HC emissions, but is relatively low on NOx, and most M103s tend to follow this pattern. In fact, in the three ASM tests my car has had, the measured NOx and standards are all over the map, which leads me to believe that the measurements are somewhat suspect. I know that when CA first implemented the ASM test, which added NOx measurement, the standards were quite high, but they have been lowered a few times since then. But the real issue is why the measured amount has gone down so much. Leads me to suspect that the measuring technology is far from foolproof.

Duke