WVU - Diesel Car Emissions Study

WVU study problems

I have some serious reservations about the WVU study.

A minor point is that the fuel report has no measurement of cetane. VW DFP equipped engines in Australia specify 51 cetane. I understand there is no way US fuel meets this with diesel standard minimum 40 cetane. We don't get fuel with 51 cetane without adding a cetane improver or mild random luck. "Typical" value quoted by oil company is 50. Low cetane increases NOX.

I have been unable to find a VW specification sheet for US TDI diesel. Anyone got a figure.

The major problem I have is the setup of the auxiliary exhaust piping of the test rigs.

The two VWs have a right angle welded join. This is the worst possible flow interrupting join for gas flow. The BMW X5 gets two pipes bent into a nice tapered union and a vertical chimney. Not being an engineer, I don't know how much exactly this would increase exhaust gas backpressure, but it has to be something. As research, it is poorly done.





How does the engine management system cope with increased exhaust back pressure??? Would this be within the usual engine map or might it trigger more DPF regens??

Are the DPF regens reported in the study expected or unexpected??
(Seems a bit odd to get a DPF regen on a highway run.)

Nearly every small diesel blows more NOX in real world tests. So much so that ICCT recommends euro6 NOX real world NOX limits be doubled over current.
http://www.theicct.org/news/press-r...aust-emissions-modern-diesel-cars-seven-times
http://theicct.org/nox-control-technologies-euro-6-diesel-passenger-cars

I also note the VW in the real world test in the last ICCT study did fine.

I can't say how much restriction the right angle bends made on the VWs, but I can say that DPF regens are totally normal on the highway. When they happen you don't even notice except for a very slight bump in the throttle when it finishes, and the elevated exhaust temps.

-J

Can compute pressure loss with a formula like this.
http://www.engineeringtoolbox.com/minor-loss-air-ducts-fittings-d_208.html

Pressure varies with v-squared, so a lot worse with high exhaust velocities. The loss coefficient for different bends is in the table. Compare factor of 1.3 for 90deg bend with 0 for tapered reduction.

Also looks like the VW with single exhaust had the same bad pipe as double exhaust with one side blocked off. This would induce resonant chamber effects.

Maybe the WVU test equipment doesn't look great, but VW has admitted to the deliberate cheating, and the EPA has confirmed that they are, independently of the WVU study.

So even if the WVU study was completely inaccurate and their piping is ugly, it doesn't mean this whole mess goes away.

bluey said:

The major problem I have is the setup of the auxiliary exhaust piping of the test rigs.

The two VWs have a right angle welded join. This is the worst possible flow interrupting join for gas flow. The BMW X5 gets two pipes bent into a nice tapered union and a vertical chimney. Not being an engineer, I don't know how much exactly this would increase exhaust gas backpressure, but it has to be something. As research, it is poorly done.

Click to expand...

I don't think it matters as much as it would with a NA car.
http://www.dsmtuners.com/threads/exhaust-the-straight-scoop-on-backpressure.168578/
And even in NA cars, the emissions system should compensate for changes in power w/o any changes in pollutants. So a 90 degree bend could hurt your power band, but any post-cat changes should be something the emissions system can handle via changes in EGR, mixture, timing, etc...

NOx during DPF regen is very high. This is not news to anyone including EPA. It is included in the emissions calculations. The manufacturer has to do tests to show how often it regens and then the NOx emissions during regen are factored into the reported composite number. So if you can reduce regens you can reduce total NOx. If you can reduce NOx during regen it hurts you less too.

EGR pulled before the DPF cannot be used during DPF regen because it contains raw fuel which you do not want in the EGR cooler or intake. An SCR however can dose during DPF regen and significantly reduce NOx. SCRs also put significantly less soot into the DPF than an LNT system so they should regen less often.

Increasing exhaust back pressure will allow the EGR system to pull more exhaust into the intake but systems that have an intake throttle generally meet commanded EGR either way. I doubt BP would increase NOx. It might reduce it slightly although it would hurt fuel economy which would hurt NOx since the spec is in grams/mile.

bluey said:

Compare factor of 1.3 for 90deg bend with 0 for tapered reduction.

Click to expand...

Reading further http://www.engineeringtoolbox.com/minor-pressure-loss-ducts-pipes-d_624.html

The minor loss coefficient - ξ - ranges values from 0 and upwards. For ξ = 0 the minor loss is zero and for ξ = 1 the minor loss is equal to the dynamic pressure or head. The minor loss coefficient can also be greater than 1 for some components.

Click to expand...

The 2.0 liter TDI engine should be tunable without aftertreatment devices to about 1 gm per bhp-hr NOx over a reasonable torque-rpm map.

Then they can delay injection a little at the expense of about 5% fuel economy to get 50% reduction. Also can use more EGR and not turn EGR off at full power. They already switched the NOx trap to fast cycling. That will help because it can be deNOxed during operation times when the driver just happens to let up on the power. For that time, the engine will have no fuel and will generate no NOx, so you won't get NOx going through during regeneration.

Then on some vehicles where the current programming does not lead to full use of the NOx catalyst (i.e. no rich rise after the plateau at 12.6 on the rear oxygen sensor, they can actually run more rich gases to make the NOx catalyst work more. (My daughter's 2013 and my 2010 TDI are in this state. Somebody else's (probably 2009 TDI) showed rich breakthrough with the 12.6 plateaus.

Then, if necessary, they can cut out some bad areas on the torque-rpm map to meet specs. The rest can be made up with fines and NOx credits.

I am disappointed with the EPA for allowing this gamesmanship to happen -- they should have anticipataed this and made such gaming impossible with better tests.

An ideal auto pollution policy would let up on NOx, CO and HC in rural areas and push hard on CO and HC in smog-prone areas. Once they get CO and HC down, the smog chemical kinetics will THEN make it beneficial to push on NOx reduction. (Look up the weekend effect on smog in LA and other big cities. It shows that lowering NOx up to 50% from current values, in the current state of affairs, actually increases smog. Go figure.)

ronaldleemhuis said:

The 2.0 liter TDI engine should be tunable without aftertreatment devices to about 1 gm per bhp-hr NOx over a reasonable torque-rpm map.

Click to expand...

Well I would hope so since that is the standard for heavy duty vehicles. That is still a long way from 70 mg/mi.

Can anyone explain to me how this ugly duckling passed the EPA test?

??????????

TDI2000Zim said:

Can anyone explain to me how this ugly duckling passed the EPA test?

Click to expand...

Because it is not a beauty contest and they used proven emission control systems rather than willfully cheating? VW could have passed, they chose to cheat instead.

Link to an interesting documentary (unfortunately, in French) addressing diesel cars in France.

https://www.youtube.com/watch?v=5JFprj6v37Q&feature=youtu.be

The gist of it:

1) Nitrogen Oxides have severe health consequences, when levels are elevated in inner-city areas. Due to high numbers of diesels in Europe (over 50% of passenger cars), and compact cities, real health consequences from NOx & particulates do occur in significant numbers. (the inverse is likely the case in most of the U.S.)

2) It is common practice, in Europe (France anyway), for owners to have the DPF's removed if/when they clog. Illegal - but undetected in emissions tests.

3) French auto-makers specialize in diesel - even more so than VW.

4) Due to #3, there is political pressure to not crack down on diesel, even though its well understood manufacturers fudge the test results.

wrc777 said:

SCRs also put significantly less soot into the DPF than an LNT system so they should regen less often.

Click to expand...

Is this because they need less EGR so the combustion temps are higher => more efficient burn?
If so, SCR equipped vehicles should give more mileage too

kaanage said:

Is this because they need less EGR so the combustion temps are higher => more efficient burn?
If so, SCR equipped vehicles should give more mileage too

Click to expand...

Sadly no free lunch. Highly efficient combustion ends up with low exhaust temp (the exhaust cools very rapidly) and SCR only works when NOx + amonia at proper temp is in the presence of the catalyst. The temp range is very narrow and high. So the engine will do a post compression fuel injection. Not a huge amount but some which will burn late and keep the exhaust hot enough for a good catalytic reaction.

That late burning is more fuel without any useful power = slight reduction in fuel economy. The only good news is if well designed (not sure if a hack and slap retrofit counts) the effect is small (1% to 3%).

I was interested in converting NOx grams per bhp-hr to grams per mile for the Jetta TDI, as the newer high pressure direct injection engines are said to achieve 1 gm/bhp-hr before any aftertreatment. I looked up some numbers.

Typical brake specific fuel consumption for TDI engine from best of 196 gm/kw-hr to 260 gm/kw-hr at low loads and/or high RPM. I picked an average of 220.

1 bhp = 0.746 kw

diesel fuel weighs 832 gm/l or 3.8 X 832 = 3162 gm/gallon (US)

typical on road fuel economy of TDI Jetta 34 mpg average.

So 1 gm per bhp-hr becomes

1.34 gm per kw-hr.

or 1.34 gm per 220 grams fuel

or 19.3 grams per gallon of fuel

or 0.55 gm/mile.

This means that to get down to 0.06 gm per mile in actual use (not just test conditions), aftertreatment or other measures have to remove nearly 90% of the NOx.

ronaldleemhuis said:

I was interested in converting NOx grams per bhp-hr to grams per mile for the Jetta TDI, as the newer high pressure direct injection engines are said to achieve 1 gm/bhp-hr before any aftertreatment. I looked up some numbers.

Typical brake specific fuel consumption for TDI engine from best of 196 gm/kw-hr to 260 gm/kw-hr at low loads and/or high RPM. I picked an average of 220.

1 bhp = 0.746 kw

diesel fuel weighs 832 gm/l or 3.8 X 832 = 3162 gm/gallon (US)

typical on road fuel economy of TDI Jetta 34 mpg average.

So 1 gm per bhp-hr becomes

1.34 gm per kw-hr.

or 1.34 gm per 220 grams fuel

or 19.3 grams per gallon of fuel

or 0.55 gm/mile.

This means that to get down to 0.06 gm per mile in actual use (not just test conditions), aftertreatment or other measures have to remove nearly 90% of the NOx.

Click to expand...

NOx is not produced from diesel fuel. It is produced from N2 and O2 in the charge air which becomes very hot during combustion. There is no linear relationship between NOx and fuel consumption.

Even if you were calculating CO2 (which does have a linear relationship with fuel) you can't just use the direct weight. Because it isn't Fuel -> CO2. It is Fuel (H and C) + O2 -> CO2 + H20.

You are right, ChemMan, that NOx comes from oxygen and nitrogen. I am just going through the calculations to convert NOx per bhp-hr to NOx per mile based on known conversion factors. My point was to show that without aftertreatment, newer diesels would produce about 10 times the allowed amount of NOx per mile on a car like the Jetta.

Trying to figure the impact of the bad piping design handicapping the two VWs in the WVU study. Running the numbers in the engineering formula with estimates for exhaust gas temperature and velocity, I compute a pressure loss of 90 Pa. Running the formula for head loss expressed as equivalent length, I get about 1.6M of extra exhaust pipe. So adding this the the actual pipe the total WVU setup looks to add equivalent of 3m of extra exhaust pipe resistance to the engine.

Maybe it would be easier to stick a banana in the tailpipe and check performance and emissions.

In the long thread, a poster noted the presence of methane in the exhaust. Surely that is rather anomalous for unburned hydrocarbons to survive combustion and exhaust system unless there is a serious problem induced by the test rig.

When VW puts out their new ecu flash, it is going to have to be squeaky clean. It will involve an engine tune that meets specifications when NOx storage catalyst works properly. It will be possible to know whether the NOx storage catalyst works properly by monitoring the signal of the rear oxygen sensor (measuring block 38, field 4).

Perhaps VW could add an air throttle on the intake. It could include a replacement for the maf meter, a throtthe flap, motor and its own brain that operates separately from the ecu in the car. It would require only 12 volt connection, though the maf would hook up as usual. It could be programmed to produce manifold vacuum according to a formula using the maf. It would be a retrofit of the maf, along with ecu flash. Manifold vacuum would allow more clean low pressure egr instead of the sooty high pressure egr.

I wonder if there is a simple poor man's way to estimate the NOx coming out of the tailpipe. It seems that NO2 and CO2 are the main exhaust components which, when mixed with water, are acidic. Nitric acid is a strong acid, and carbonic acid is a weak acid. There may be some trace gases like H2S that would produce acid as well, but let us consider NO2 and CO2 only for the moment.

Suppose you had a bubbler of sorts -- like a bong. If you could divert a fixed proportion of TDI exhaust through such a bubbler, maybe you could use it as a crude NO2 analyzer. You could put a pH indicator and a strong base into the liquid and see how much cumulative flow is required to neutralize the base.

http://www.amazingrust.com/Experiments/background_knowledge/Nitric_Acid.html

Can anybody think of a way to make this selective for NO2 so that the CO2 does not also get counted?

Or, along other lines, is there a way to use a beam of light (maybe even ultraviolet or infrared) and a meter to take snapshots of the exhaust plumes from cars as they drive by on the road. (Naturally, one would have to figure in exhaust volumes and all sorts of other things.) Yet if we are talking orders of magnitude of variation in NOx emissions from one car to another, maybe ultraviolet or infrared vision in the right wavelength would make NOx visible to the naked eye.

Wouldn't it be something if somebody could rig up some sort of video camera that sees NOx?

How about this? Sort of like a video speeding camera, but for pollutants:

http://www.bar.ca.gov/FormsPubs/Fact_Sheets_and_Brochures/Remote_Sensing_Devices.html

CO2 absorber http://www.udmercy.edu/crna/agm/07.htm

I have got a bubbler setup for calibrating breathalyser. A fritted disc (like sintered glass) gas dispersion tube does the micro-bubble job for scrubbing gas (or in my case the reverse). There are different grades of frit for different purposes - somewhere there is a guide.

For example http://www.chemglass.com/adv_search.asp?mt=0&k=gas+dispersion&search=Search

ronaldleemhuis said:

How about this? Sort of like a video speeding camera, but for pollutants:

http://www.bar.ca.gov/FormsPubs/Fact_Sheets_and_Brochures/Remote_Sensing_Devices.html

Click to expand...

Those vans exist in VA already. I see them parked on off-ramps from time to time.

-J

To see how the latest ecu flash is working on my 2010 Jetta TDI, I logged EGR percentage and signal from the rear oxygen sensor (VCDS measuring group 38). This was for a round trip country drive 24 minutes and 15 miles each direction.

The EGR percentage (100 minus the signal in field 2 or 3) if often as high as 65%. Naturally it goes to zero when NOx regeneration is taking place.

Occasionally I will see a quick rise to 49.8 or so from the 12.6 plateau of the rear oxygen sensor signal. If this is a true reading and not artefactual, then I think it shows that the ecu is adjusting the amount of late and post-combustion diesel injection to cause the NOx catalyst to purge most or all of the NOx at least some of the time. One could presume that a NOx generation model in the ecu would be able to estimate the amount of stored NOx at any time and could probably even assess the storage capacity of the NOx storage catalyst.

This route is one that I habitually drive in the same conditions, the same time of year and the same time of day. My mileage on the MFI is running 5-10% less this year than it did this time last year on this same route.

These observations suggest to me that VW made significant changes in the recent ecu flash, and that they may be using a lot of EGR and much if not all of the NOx storage catalyst function.

Here is an abstract of a 2014 publication showing the advantage of rapid lean/rich cycling of the NOx storage catalyst. It describes the benefit of going from cycles of about 70 seconds to cycles of about 7 seconds, with about 1/6 of the time in a rich phase.

http://www.sciencedirect.com/science/article/pii/S0920586114000613

By watching the signal of the rear O2 sensor, it appears that my 2010 Jetta TDI with the latest flash is cycling rapidly as in this article. About 1/5 of the time the rear O2 sensor reads rich, and about half of that 1/5 is at the +12.6 plateau of deNOx-ing and a good fraction of the rest in rich spikes apparently trying to get the deNOx catalyst to start deNOx-ing.

All in all, it appears that the latest flash is doing these things:

1. Pushing EGR to 65% or more in highway driving.
2. Converted to rapid cycling technology of the NOx storage catalyst.
3. Rear oxygen sensor senses rich signal about 20% of the time.

I suspect, but cannot prove, that VW changed the timing of some of the fuel injections to move along an NOx - soot tradeoff curve.

They probably adjusted exhaust temperatures to avoid high gradients that would crack DPFs and lead to soot clogging the EGR cooler and fouling the NOx storage catalyst.

They may be using the air inlet flap to throttle (and reduce NOx) in some driving situations.

They pulled out all the stops, in other words.

It is possible that NOx performance would be good in real life except for degradation, sulfation and soot loading of the NOx storage catalyst.

I've seen filter-like devices that can enrich ambient air a few percent in nitrogen content. When such nitrogen-enhanced air is used instead of regular air, it reduces NOx production with everything else the same.

The good news:

BMW never cheated: http://www.bbc.com/news/business-34519184

http://pics.tdiclub.com/data/500/D3E802F8708-2_0L_TDI_Common_Rail_CBEA-CJAA.pdf

have you seen this listing of MB?

Now that list of Measured Value Block List for 2.0 L CR BIN5 looks interesting. If it is applicable to my 2010 Jetta TDI, then I have misinterpreted field 4 of measuring block 38 as an oxygen sensor signal. According to this listing, I should expect to find something in measuring blocks 138 and 139.

If these labels are correct, it will certainly help figure things out. Looks like a gold mine.