The dirty cost of one man's hypermiling?

[GRU]

Its so hard to be satisfied with 40mpg when you ised to get 50 before lol

[California98Civic]

This thread convinced me to do mostly engine-on pulse and glide on the freeway, instead of engine off, in order to keep the CAT lit and control emissions better. And I just did my smog test, and here are the results and my thoughts about the results.

"Black and Green" passed on the first go. But at the 15mph 1793 rpm test hydrocarbons were 7x the ave for this model car and only 2ppm below the allowable limit. The CO and NO were close to the averages for this model, with the exception of CO at 1793 rpms, which was four times average (though still only 20% of the allowable limit). The was evidence of "unburnt fuel" the tech said.

The car has been running a WAI, 14* ignition timing advance, and non-stock specified spark plugs. And most importantly I do a lot of engine off P&G on surface roads. I have noticed that the long term fuel trim turns slightly rich under lots of engine-off P&G, which is maybe the reason for the unburnt fuel.

To get ready for the test, I reset timing to stock (12*), installed stock spark plugs, and reinstalled the stock air box apparatus. Then I drove forty miles on the freeway without engine off coasting at all. The first 20 miles I did engine-on pulse and glide, accelerating with 80% load and then coasting in neutral. Fuel trim moved from 0.00 to -1.6%. On the way back to the smog check, I held a steady 60 mph on the speedo for 20 miles. Long term fuel trim returned to 0.00% then to -0.78% before I pulled into the check station.

I don't think these results confirm or refute Sentra-SER's findings in this thread. But my gut tells me P&G, especially combined with engine off coasting using an injector kill switch plays a big role. That's because I know the ECU on my Civic responds to rapid throttling changes by slightly enriching the AFR. When I bump start and press the throttle to get to 80% load ASAP, I think the AFR gets a little rich. Then I suddenly cut the engine with the injector switch. My +0.78% to +1.6% long term fuel trim readings under engine off P&G driving would seem to confirm that P&G produces "unburnt fuel," as the tech put it. Conversely, I have observed -2% and even -3% long term trim on the freeway, doing engine on P&G, which I think gets the ECU to lean-out AFR because of the tumbling rpms each time I drop into neutral and let the engine decelerate from 2700 to 700 rpms. I imagine the ECU thinks of the rpm change as a DFCO type situation, and so it goes leaner for a second.

BTW, The tech suggested I might want to change the O2 sensor on the exhaust manifold (it has never been changed). He thinks that might return my car to better average emissions. I'm thinking a Denso (stock O2 sensor brand). And I will experiment with modifying my surface street techniques with more engine-on P&G. But considering the production and distribution of the gasoline, I still believe that on balance I am saving gas money and reducing my car's total pollution (see: 1, 2, 3).

[christofoo]

I'm sorry I didn't take the time to read this whole thread; I have two questions though, if someone can be bothered to help me catch up here:

1) A while ago I dug up 2-5 minutes as light-off time, and 25 minutes as cool-off time. (refs below). To me that infers that a engine-on duty cycle much larger than 20% with EOC cycles much less than 25 minutes will tend to light the cat or keep it lit. So I find the OP's results surprising. (Has this already been pointed out?) Yes, both of those figures are crude, real world results depend especially on load, also cat size, also the cat follows an S-shaped Wiebe function, so significant increases in emissions may occur significantly before the cat cools below light-off, but even without doing the math I expected fairly low emissions at 50% duty cycle, which is more typical (for my driving).
2) Has anyone done anything to reproduce the OP's temperature measurement? (Wondering if the OP's car is unique in this regard, or if the sensor is in a biased location.)

Sorry again for my laziness.

References:
1. 25 minutes cool-off time: see Fig 2 and Table 2 on p3:
http://www.nrel.gov/vehiclesandfuels..._emissions.pdf
2. Light-off time:
Bottom of page: NSLS Everyday Science | Catalytic Converters
Paragraph 1 on p1, Fig 3.2, 3.3, 3.4 on pp 27-29:
http://citeseerx.ist.psu.edu/viewdoc...=rep1&type=pdf

[California98Civic]

Originally Posted by christofoo

References:
1. 25 minutes cool-off time: see Fig 2 and Table 2 on p3:
http://www.nrel.gov/vehiclesandfuels..._emissions.pdf

Really interesting. Thanks. This essay on how "sufficient insulation around the converter can help the converter maintain its temperature above light off for several hours after the engine is shut off." (p.2) But of further significance to this thread is the description of current CATs as having a light off temperature of 350*C (662*F) that is typically retained for 25 minutes after the engine is shut off. That would imply that a fully grill blocked CAT, close to the exhaust manifold, allowed to warm up properly, would tend to stay lit during EOC.

james

[christofoo]

Getting off-topic now, but be careful with insulation around the cat. In that ref they used vacuum insulation with a hydride that foiled the vacuum at high temp to avoid meltdown. That approach isn't for the faint of heart, in this case even the pros messed up on outgassing and permanently spoiled their vacuum. Also of interest is their use of a phase-change-material (Aluminum) for heat retention. (Phase-change = molten, another thing we enthusiasts should be pretty careful about.) I've been wondering if a vacuum insulated cat is available OEM yet, and if it can be retrofitted on older vehicles - can't seem to find it.

[christofoo]

Back on topic, the one real data point we have here (from our OP), indicates that my research on cool-down time misled me somehow - I think it'd be nice to figure out why, or if the OP's case is isolated.

Also note, grill blocking may make a noticeable, but not gigantic difference in cat cool-down. From the cat's point-of-view, the rest of the engine is pretty cold.

[SentraSE-R]

When I made those cat temperature observations, I noted that my UG and SG cat temperature gauges behaved oddly. Indicated temperatures shot >500 degrees F within 3 blocks of driving, which seems impossible, or at least improbable.

They also indicated cat temperatures dropped on every FAS from 500, 900, even 1100 degrees F to <200 degrees F within the 15 seconds to 1-2 minutes of the engine-off glide. This is another impossibility.

I suspect the UG and SG cat temp algorithms must rely on a voltage signal that resets when the UG and SG detect the engine-off condition. Obviously, the cat cannot physically cool 1000 degrees in a few seconds.

[christofoo]

Originally Posted by SentraSE-R

When I made those cat temperature observations, I noted that my UG and SG cat temperature gauges behaved oddly. Indicated temperatures shot >500 degrees F within 3 blocks of driving, which seems impossible, or at least improbable.

They also indicated cat temperatures dropped on every FAS from 500, 900, even 1100 degrees F to <200 degrees F within the 15 seconds to 1-2 minutes of the engine-off glide. This is another impossibility.

I suspect the UG and SG cat temp algorithms must rely on a voltage signal that resets when the UG and SG detect the engine-off condition. Obviously, the cat cannot physically cool 1000 degrees in a few seconds.

The temperature rise you observe may be normal; it could take as little as 2 minutes to get >500 F, or something like that...

It does sound like your cat temp signal is erroneous after shutoff for some reason. It would be interesting to find out where your cat temp sensor is located. Or- is it possible your ECU attempts to predict cat temp, rather than measuring it, and there is no real sensor at all?

About 10-15 min after shutoff, I inserted a thermocouple into the air-gap between the cat inner and outer jackets on my 97 Civic. The reading was 200°C (392°F), and it very slowly fell to 170°C (338°F) over the course of several minutes. Since that location isn't a good thermal well, I was probably reading air temperature mixed with cat temperature.

I can say the cat's internal temp must have been higher than 200°C at least 10 minutes after shutoff, although I don't know by how much higher. I can also say that the thermal decay time is on the scale of 10-30 minutes, and certainly not <2 minutes.

So preliminarily that seems to support the 25-minute half-life I cited earlier, and casts strong doubts on the validity of the OBDII sensor (at least the particular sensor in Sentra's Scion). When I get better data I'll probably do a write-up in a separate thread, since I'm interested in smog pollution, and I'm suspicious of the conventional wisdom.

When you say FAS, the transmission is in neutral, no?

[user removed]

While 02 sensors may last considerably longer, I think they get "lazy" at 50k and degenerate from that point until they get bad enough to cause a code to be set.

regards
Mech

[metromizer]

My guess, and it's only a guess, is that your cat was not up to temperature on test #1. Cats work as a secondary combustion chamber, and need fuel that has passed through the engine to build heat and work their magic, like a mini afterburner. Taking the car out for a spirited drive, building some heat in the cat, usually fixes all that, according to the CA emissions techs I've talked to.