The dirty cost of one man's hypermiling?

[mechman600]

Diesel exhaust temps are lower nowadays because of this high amount of EGR (30-50%) used to significantly lower NOx. An older, non-EGR diesel will have much higher exhaust temps (at full load) than a gas engine. Gas engine exhaust temps remain fairly constant because of near stoiciometric mixtures. Diesel engine exhaust varies wildly, as you've stated above. An older (dirty) diesel can be >1200F at full load on its own.

Gas engine CAT: 3-way cat. Converts HC, CO, NOx to "nicer" things. Works whenever the engine is running and is hot enough to do so. Requires stoiciometric.

Diesel engine CAT: 2-way cat. Only oxidizes HC in order to regenerate the DPF. Cannot remove NOx because of lean diesel operation. To elevate exhaust temp at light loads, the ECM closes the VGT to create back pressure, stops all EGR flow and retards the timing to get turbo out temps above 550F. Then it either injects fuel in-cylinder during the exhaust stroke or uses an HC doser on the turbo outlet to spray raw fuel into the CAT (called a DOC, diesel oxidation catalyst). It doses enough HC to maintain 1000-1100F to allow the DPF (attached to the DOC) to regenerate. Regeneration only occurs once every 8-15 hours of operation, otherwise the CAT is basically doing NOTHING but going along for the ride.

[dcb]

It may be that a "right sized" cat would be good too, i.e. if you always shift at 3k, and say your redline is 6k, then you can theoretically use a cat that is 1/2 the size, and it should start catalyzing twice as fast, sort of.

[California98Civic]

Thanks for this discussion. I like the turn this thread took a long time ago toward figuring out balances between emissions and fuel economy that might plausibly reflect total gains in eco-friendly operation. I took a couple ideas from here: going back to exclusive engine-on P&G on the freeway to help keep the cat "lit" and only using DFCO (with the engine off) at the start of a trip (when the engine is still cold anyway) or near the end (when I am about to shut down anyway). I'm also starting to study the O2 sensors and their relationship to rich/lean fuel trim conditions. What an excellent intervention this thread has turned out to be.

[dcb]

I think it isn't really well quantified as to the effects, hardly enough information here to make an absolute decision, yet people are making decisions ?!? In lieu of the fact there are quantifiable efficiency gains, and that the cat is still 600+ degrees even with cooling dfco air going through it?!?

We know p&g can save lots of fuel

We don't have no clue how it affects total emissions, if at all.

Needs serious testing to have a serious discussion.

[SentraSE-R]

Originally Posted by dcb

I think it isn't really well quantified as to the effects, hardly enough information here to make an absolute decision, yet people are making decisions ?!? In lieu of the fact there are quantifiable efficiency gains, and that the cat is still 600+ degrees even with cooling dfco air going through it?!?

We know p&g can save lots of fuel

We don't have no clue how it affects total emissions, if at all.

Needs serious testing to have a serious discussion.

You're ignoring the elephant in the room. My engine is fine, emitting normal HC levels after normal cat-lit operation. It failed to run the cat drive cycle and set the appropriate readiness flag because P&G doesn't light the cat.

When the smog test was done the first time, after 800 miles of P&G operation, HC levels were 62, 15.5X higher than average, and just 2 below failing, and the smog test is done with the cat lit. If a lit cat is 90% efficient, then my P&G (no, every EOC P&G) that never lights the cat is dumping 155X more HCs than the average car.

100 miles of cat-lit operation brought my car's HC level down to 39, 10X the average, on the second smog test.

900 more miles of cat-lit operation brought my xB's HC levels down to average

If that's not a cause and effect situation, and you really have doubts about it, you need to propose another more likely cause and effect for us to consider and test, if necessary.

When I get home, I'll take my xB on a P&G EOC drive to Costco, or on my 10 mile test loop. I have no doubt I'll fail to light the cat at all in local low speed P&G. It takes 3 blocks of driving to raise the cat temps to 550F (light the cat). Each low-speed pulse moves me 1-2 blocks, followed by several blocks of coasting

[dcb]

You are ignoring the elephant of incredibly few tests being performed and not using A-B-A testing to begin with. You started with a conclusion, and that is never good science. You jumped to this conclusion based on two tests, one with the ECU in a not-ready state. You said you cannot explain the data, so stop and think for a minute. I'm not saying you are wrong, but it has het to be demonstrated and quantified, you still use terms like 75x while demonstrating that the cat stays fairly hot between pulses...

Is there more pollution w/p&g? As a percentage of fuel used? We don't know. Using less fuel results in less c02 generally speaking and there are a lot of wastes upstream of your tank. Your experiment has no controls.

For all we know you have a lot of oil blowby and your car need lots of extended time at temperatures to keep your cat clean on your car. I don't know what would be in the gas causing your cat a problem needing 100s of miles to correct. But the condition wasn't sufficient even with p&g to trigger the onboard self diagnostics, and it might even be the o2 sensor that is fouling and not the cat. And there were misfire codes...

[dcb]

except for HC, things generally got worse in the third test, after all the driving.

The reason I am interested in WHY you are getting these readings is because that has implications for what to do about it.

You assert it is the cat

It could be the ECU, the first run could be interpreted as a slightly rich mixture from what I can tell, and not the cat, i.e. a software problem.

It could be an o2 sensor.

It could be contamination from another system.

We simply don't know, not enough data or testing, why is this such a hard concept? You are claiming p&g makes long term changes to a cat, and really haven't demonstrated that except by repeated assertion. Make sense of the data first, I have posted links to numerous effects of catalytic converters under different conditions, the observed readings do not yet make sense to me.

[dcb]

Elephant! from typical diagnosis procedures:
AUTO REPAIR HELP - DIAGNOSE - EMISSIONS TEST FAILURE

If you have high HC AND CO, then the cat is suspect, but it is pretty low on the list of suspects. Your CO is nowhere near high in any of the tests.

High HC by itself looks like ignition mostly suspect, and then vacuum leaks.

Note, you had a HUGE spike in NOX in the third test also, like EGR went wonky or something.

You could have several things going on, emissions are not simple to diagnose, don't pretend that they are.

DIAGNOSIS
Diagnosis of high emissions can be an involved process that usually requires special tools and measuring equipment. However, armed with a test report indicating what gases are high can help you determine what systems may not be functional. If the vehicle has failed as a result of high CO and HC, the CO failure should be diagnosed first. If the "Service Engine Soon" or "Check Engine" light is on it should be diagnosed prior to any emissions diagnosis. CO Failure is the result of an excessively rich air fuel mixture. The following is a list of the most common causes:
- Defective Oxygen Sensor
- Inoperative Air Injection System (if equipped)
- Leaking or defective Fuel Injectors
- Restricted air filter (especially if high CO is present only at high RPM)
- Vacuum leaks resulting in improper MAP sensor operation
- Defective air mass or air flow sensor
- Malfunctioning fuel evaporation system or purge valve
- Defective Catalytic Converter
- Defective thermostat (cooling system), thermostat stuck open
- Oil contaminated with fuel, excessive miles between oil changes

HC Failure is the result of incomplete combustion. Any item that causes incomplete combustion can result in high HC. The most common causes are:
- Worn spark plugs
- Defective spark plug wires
- Worn distributor cap and/or rotor
- Improper ignition timing, usually over advanced
- Vacuum leaks
- Engine mechanical failure, low compression, worn valves, excessive oil consumption

NOX Failure is the result of combustion temperatures that are too high. The most common causes are:
- Inoperative Exhaust Gas Recirculation System (EGR)
- Cooling system malfunction, engine running too hot (restricted radiator, defective thermostat, etc.)
- Over advanced ignition timing
- Excessively lean air fuel mixture (defective oxygen sensor, MAP sensor or Air Mass Sensor)

I'm reminded of how many folks would replace the carb when the carb isn't even the problem.

[California98Civic]

Originally Posted by dcb

Elephant! from typical diagnosis procedures:
AUTO REPAIR HELP - DIAGNOSE - EMISSIONS TEST FAILURE

If you have high HC AND CO, then the cat is suspect, but it is pretty low on the list of suspects. Your CO is nowhere near high in any of the tests.

High HC by itself looks like ignition mostly suspect, and then vacuum leaks.

Note, you had a HUGE spike in NOX in the third test also, like EGR went wonky or something.

You could have several things going on, emissions are not simple to diagnose, don't pretend that they are.


I'm reminded of how many folks would replace the carb when the carb isn't even the problem.

I just want to say, dcb, that I agree with you that the diagnosis is not complete and that this is still just a logical, educated guess. I just think it is a good guess. But separately from that emissions-failure diagnosis-debate, there is a good eco-driving principle in heating the cat up to operating temps through light-load idling at or near the start of a drive and then begin hypermiling. If I do that, I can be much more confident that 80% load acceleration in a P&G cycle will not mean unnecessary pollution. I also appreciate better now that DFCO, especially with the engine off, will send a lot of relatively cool air through the cat, possibly interrupting its operation if the DFCO jag is long enough. So, I agree that the diagnosis has not conclusively pointed-out the cat, and I agree there could be more than one problem. However, I also have to recognize the cat is likely at least ONE of the culprits for the emissions failure. As an aside, another reason I have decided to go back to engine on coasting on the freeway is that it feels like I'm abusing my transmission by bumping it at those speeds, and I really love my car. Even engine-on, I should be able to reach this year's goal of keeping the car over 60 mpg without super radical hypermiling and stress on parts.

Sentra: when you test to see about lighting the CAT today or whenever, to see how long it takes, could you record how far you drove in miles or tenths of a mile as well as coolant temps when the cat reaches temp? My car does not have a CAT temp sensor and knowing distance and coolant would give me rough indexes of how long/far to idle to be in the right ballpark. Otherwise I'll just guess a mile or two minutes idling.

Many thanks.

[Arctic Fox]

Forgive me if this info has already been posted somewhere;

My VW Passat has a "secondary air intake". It's a small blower that pumps fresh air into the exhaust, after the turbo, before the cat, and is suppose to heat the cat quicker on cold starts.

At startup, it turns on (sounds like a hair dryer) for about 30 seconds. Sometimes it will turn on again for another 15 seconds.

Just something for you guys to think about.