#DieselGate Scandal Shows Flaws In State Inspection Tests

Yeah, measuring emissions in anything other than real-world conditions will give you something other than a measure of real-world emissions. That's pretty obvious. In-lab-only tests are easily defeated.

FWIW, long before Dieselgate broke, a US-based company gathered millions of real-world emissions samples using spectroscopy. They set up operations in a van parked at the side of a long stretch of flat highway. A radar measured vehicle speed (and detected acceleration/deceleration), while other equipment took a spectral sample of actual tailpipe emissions. They reported that VW diesels had "impossibly high" emissions. They reported their data and findings to US authorities, but nothing became of it.

But measuring vehicle emissions with roadside techniques, such as Steve's study lays out, would be a nightmare if incorporated into annual state inspections. Just way too hard to implement, and even if they could, all the variables such as weather conditions, location, etc. etc. would make the results still questionable to the folks that write these articles. They would never be satisfied.

Vette71: Good thing people didn't give up on the dream of powered flight just because they believed it would be hard to pull it off, or it wouldn't be practical.

I'm not saying that lab tests should be replaced by the specific method I talked about. I'm just pointing out:

(1) Lab-only tests are easily defeated and useless for determining actual (real world use) emissions.

(2) The kind of road-side, non-invasive, mass sampling of emissions data that I described in my post is great for obtaining large pools of real world data, and could be instrumental for identifying offenders that way way beyond legal. It saw a trend with massive emissions from VW diesels years before Dieselgate broke.

Steve, I wouldn't say lab tests are useless. Governments make regulations and set standards and design lab tests that they believe if manufacturers meet will insure to a high level of confidence, not 100%, that the product performs as the regulators intended. That is if they do not cheat. The issue is VW cheated and the regulators had no process in place to do random checks to see if there was cheating. It is interesting groups in Europe have been testing diesels and are concluding other manufacturers did not cheat.

I spent my career in test and measurement in medical, chemical, gas, food, water, environmental and explosive analysis (all regulated). Things start out in the lab under controlled circumstances and methods. Moving them into the field is the dream, but there are few successes as the physics and scientific variables make the results less reliable, reduce productivity and become cost prohibitive. Random checks for compliance are usually the result.

When was the last time your briefcase got checked by a TSA sniffer machine or a dog at an airport? The issue is still there. The field solution wasn't productive and other science, as of now, cannot meet the challenge. With the risk still there, but deemed low, the field testing was dropped. Ces la Vie.

Vette71: The simple truth is as follows:

If you want to determine a vehicle's emissions under laboratory conditions, create an in-lab-only test. That's what we have today. If you want to know a vehicle's emissions while operating in real-world conditions, you have to do emissions tests on the road, or in the very least, conditions that closely resemble real-world use.

But I have to warn you: Car manufacturers will resist this, kicking and screaming, because VW isn't the only own who has optimized emissions for known in-lab tests while thumbing their nose at real-world emissions, which are not regulated not monitored. Why bother prepping for a test you're not gonna take, right?

Even simpler truth is not just car manufacturers optimize to meet the standard tests, but everything works that way. Food, beverages, medicines, medical equipment, pollution monitoring, you name it, regulators design a test, 90% in the lab, for the standard and manufacturers and service providers work to meet it. So in essence as a society we don't want to take the time and/or spend the money to be absolutely certain our regulations hold "in the real world".

Your post about being able to hyper-mile a vehicle to meet the 50+ mpg standard is a great example. It can be done as you tested. But people refuse to behave that way in everyday living as the 55 mph national speed limit proved. Society made a decision that was not real world.

The state test here in WA is simply to plug into the ODBII port and scan for trouble codes. No codes and no CEL, you're good to go. I think they actually stopped the tailpipe sniff over 10 years ago. And it's only done in the Seattle and Spokane commuting areas, so a huge number of cars are never even tested at this level.

I'd much rather be driving behind a "clean" diesel than many of the gas cars on the road. My Passat has no spell or smoke at all, while many gas cars have a terrible odor from the cat converter, or some other issue.

Indeed - the same in NJ- and that is the ONLY test done - apparently it is enough to satisfy the federal requirements

Same in Mass.