Modern Diesel Fueling Strategies
 

Having just bought my first diesel (2016 BMW 535d) and being a huge gearhead, I'm wondering about how diesels work with regards to fueling, A/F ratios, etc.

In a gas engine the "gas pedal" is basically telling the throttle how far to open and how much air to allow into the engine. Based on this, the computer calculates how much fuel to add based on how much air mass the engine is breathing and a number of other fine-tuning factors. The engine runs rich during cold starts to warm up the catalyst faster, and to keep combustion temperatures and knock under control under high load.

As far as I know a diesel runs lean all the time. The strategy is different in that the "gas pedal" actually controls the amount of fuel that goes into the engine, and in modern diesels there's a programmed delay in the rate of fuel added to allow the turbocharger to keep up with the air required to burn it efficiently (again, as far as I know).

My questions are:

1. Do diesels ever intentionally run rich, and if so when? I know there are EGR strategies to keep emissions in check, but does the Air/Fuel ratio play a part as well? Too rich results in excessive soot from unburnt fuel, but is there a strategy as to how rich they need to run or do they always run as lean as possible? There is also the regen cycle that puts additional fuel into the DPF to burn off the collected soot, but I don't consider that as it's all done after the engine.

2. Efficiency vs. load. A gas engine has a maximum efficiency (BSFC) load at any given speed, and is less efficient with more or less load applied. Do diesels work this way as well, or are they more efficient the more load they're under?

If there's a website that clearly explains all of this please feel free to direct me to it instead of regurgitating it all here. Thanks!

A diesel engine usualy has no throttle.
It always takes in as much air as possible.
Power is adjusted by injecting more or less fuel.
Less fuel = less power
More fuel = more power (up to a point)

Modern diesels use EGR to make the engine run richer without adding more fuel.
This results in lower NOx raw emissions, wich need to get converted to nitrogen and water by the SCR catalythic converter.
The SCR cat needs an urea solution to do that, so by needing to convert less NOx into N2 and H2O, it saves you DEF.

Downside of running close to stochiometric is that the engine produces more soot particles.
These particles get trapped and burned in the particle filter.
To not clog up the filter over time, it needs to burn the particles, wich happens when driving fast on the highway for a while or with a regeneration cycle.
In said cycle the engine injects additional fuel after the main combustion to get the exaust gas temperature up and burn off particles.
To avoid unnessesary regeneration cycles, the ECU avoids running stochiometric or even rich.

Diesels have a point where their BSFC is best, but as they have no throttle, they are more efficient under low load than gasoiline engines.
This results in great fuel economy.

Turbo-lag has not been much of an issue anymore, with the variable-nozzle turbochargers becoming mainstream on Diesels.

I disagree, it has become less of an issue, but it's still there.
A modern turbocharged engine doesn't have the throttle response of a naturaly aspirated engine with ITBs.

I said not much of an issue, but I didn't say it was not an issue at all.


Even a naturally-aspirated engine with just one throttle-body often has a better throttle response, no wonder Volkswagen makes naturally-aspirated versions of the T-Cross in Brazil for regional exports, as some countries such as Mexico and Bolivia are quite hilly, and the turbo-lag is felt harder in La Paz and Mexico City than in Brasília for instance.

Oh, I'm quite aware of that.
The response of my Toyota MR2 and a recent model Audi TT is a difference like night and day.
With the throttle response of an engine like the TT's, the MR2 would be undriveable at the limit.

The higher-revving nature of a gasser tended to dictate the need for a larger turbo. Didn't you notice how turbochargers for Diesel engines used to be proportionately smaller than those for a gasser? No wonder often the turbocharger of a low-revving tractor-truck would be used for adaptations to cars.

I believe modern turbodiesels have artificially sluggish response purely for emissions reasons. They only add fuel at a rate slow enough for the turbocharger to completely keep up, since running even slightly rich results in huge increases in particulates and soot.

This was evident when I drove truck for a lumber yard in my younger years. The old '90 (I think) International 466 flatbed 5-ton was a joy to drive, and the throttle response was very good - it was like the tach was simply attached to the throttle pedal with a cable. You could put that engine anywhere you wanted with speed and precision, making it an absolute pleasure to rev-match. (For those not aware, you only use the clutch to get it moving in the lowest gear, and it's all clutchless shifting from there on out).

A newer International 466 they leased, a '96 I think, had an electronically controlled fuel delivery system and felt completely disconnected compared to the older truck. It took me a long time to adjust to driving it, and even though it had air conditioning (and an advertisedpower increase) I still preferred driving the old '90.

Having said all that, the engine in my 335d is quite responsive. It's sitting in front of an automatic transmission, so it's pretty much blasphemy to speak about throttle response... if it was a manual, OK. :D

EDIT: Forgot to mention that the older trucks would spew black smoke on every shift as the turbo caught up to the fuel being dumped into the engine by the mechanically actuated fuel pump. The newer electronic engines had no such black smoke between shifts, but they seemed to take FOREVER to spool back up and make power.

Turbochargers are specified based on flow rates. 300 hp is 300 hp, regardless if its attached to a Honda or a Peterbuilt. It's not quite that simple when you get into pressure maps and what not, but people shopping the scrapyard for Powerstroke turbos to boost their budget projects aren't usually too fussy. :)

The old '90 International 466 I drove way back was only rated at 250hp if I remember right. That figure is nothing special for a gas four-banger 1/4 the displacement these days.

I'm very curious how these modern diesels control for adding power... if the air is available is it just a matter of adding more fuel? I know how to tune gas engines, but diesel is a completely unknown concept for me.

This car must have a very effective EGR system then, because I have put about 15,000km on it since buying it a year ago and have yet to top up the DEF tank. That, or the DEF tank is just huge?

So far I have been completely unaware of any regen cycles taking place. From what I've read it should be obvious when a regen cycle is happening as your mileage will suffer terribly. I've noticed nothing out of the ordinary so far, and I don't think I use the car on the highway more than most people would. My commute is actually very slow, travelling about 50MPH with the cruise on for about 15 min. which doesn't seem like much time/load to facilitate the burning of soot in the DPF? When I do have it on the highway I tend to drive it harder just for that reason, but it's not a regular thing.

...and is the reason I was so interested in this car. For a big heavy AWD sedan it's phenomenally frugal. I average about 8-8.5 L/100km (27.5-29 MPG), and do much better on long trips. I got 5.7 L/100km (41 MPG) calculated at the pump on a 3.5 hour highway trip in January with four brand new snow tires, averaging around 60 MPH.

One thing I'm curious about is cold weather operation. A gas car runs rich to warm up the catalyst, which can take quite a while with light-footed driving, or while idling, and can result in fuel contaminating the oil. I don't believe diesels have this issue and would be much more efficient in the winter because of that, no? I didn't notice much change in fuel economy in the cold last winter and I'm wondering if that's why?

The only way I know my golf is in regen mode is the idle climbs to 1300 rpm or there's slight bog on heavy acceleration and the cooling fans go to high speed. Cruising above 50, you'll never notice, but the flow and temperature might be high enough to break down the particulate size. On a schoolbus I drove there was a regen light that would go on every so often.

If you don't notice your oil turning black from unburned fuel, you're not looking at the engine oil

1. Modern diesels work about the same as oldschool ones in that regard, but they do add some EGR especialy under low load to keep NOx down

2.That's what the EGR is supposed to do, the filling intervalls can depend on driving conditions though.

3.You might not even notice the car initiating a regen cycle when driving on the highway as it just injects a little more fuel after combustion to get the exaust gas temps up enough.
Plus highway driving means you might not even need regen cycles at all as there your EGT goes up sufficiently on its own.

4. Yep, diesels can be very efficient, yours is actualy not that efficient by diesel standards.
Take a look at the Passat 1.9 TDi or a Lupo 3L

5. Diesels hate the cold.
As they are so efficient under low load and move that much air, they take a long time to warm up.
In idle they might not even heat up at all, so if it gets very cold where you live, consider a block heater.

No worries there... the oil is B L A C K when I change it! I've decided to change the oil in this car at 10k km intervals instead of the recommended 15k. The timing chain on this engine is a very very big job... one I don't plan to do! :eek:

It's usual for the tanks to hold enough DEF for such long mileages, but it's also worth to consider cars often resort to both EGR and DEF while heavy-duty trucks and buses usually have either one or another, so a car would use fewer DEF proportionately to the fuel consumption, even though the DEF tank of a truck might hold a volume proportionately closer to that of the fuel tank.

Good to know!

Makes perfect sense.

Right. I've read that a good highway blast is good for modern diesels from time to time, and that's probably why. This car enjoys the left lane anyway. :)

I could also ride a scooter. :)

I'm just comparing this car to its gasoline powered siblings, or other comparable cars with gas engines. It was only available in this AWD 3L six configuration in North America, but in Europe is was available with several four cylinder diesels that are all quite a bit more efficient than this.

The newest Passat TDi sold in North America was the 2015 model, and the most fuel efficient version was EPA rated at 30 MPG City, 42 MPG hwy, 34 MPG combined:
https://www.fueleconomy.gov/feg/Powe...=1&rowLimit=50

Compare that to the 535d at 26/38/30 combined:
https://www.fueleconomy.gov/feg/Powe...=1&rowLimit=50

Compared to the 535d the Passat TDi is smaller, 600 lbs lighter, only FWD (no AWD option), 118 hp less powerful, 3 seconds slower to 60 MPH, and doesn't even have an automatic transmission (the auto drops it to 33 MPG combined). For an additional $12.50/month in fuel compared to the automatic TDi, I'll take the 535d. Especially in Canada where AWD is important three months of the year.

A more even comparison to that Passat TDi would be the 2016 BMW 520d, the same car but with a 184 hp 2L diesel. Compared to the 520d XDrive the Passat TDi still has all the disadvantages listed above, except it's only slower to 60 by about 1/2 a second. Extrapolating from published European data the 520d would test about the same as the automatic equipped Passat TDi - 33 MPG combined.

All that said, if the Passat TDi is considered an efficient car, the diesel BMW 5-Series is a very efficient car since it seems to do more with the same amount of fuel. Shame more European diesel models were not made available here in North America.

I did quite a bit of research on this car before buying, and a good chunk of that was looking into winter performance. It didn't seem to be an issue, and based on my experience last winter it doesn't look like it will be.

Ottawa can get pretty cold, though last year was warmer than usual. We had a few nights down to -25C (-13F) and I'm pleased to report it started absolutely perfectly each and every time. It will ask that you wait 5-6 seconds for the glow plugs to warm up, then it will crank and start like a champ. It did sputter/vibrate once as if one cylinder wasn't firing, but a blip to 1500 RPM solved that quickly. If we have a more typical winter this year I hope the same remains true at -30C (-22F). If so, it gets a huge gold star for cold weather insensitivity.

With heated steering wheel and seats, engine warm-up time isn't as much of a concern. Having said that, I'm absolutely shocked at how quickly this thing warms up. I'm not 100% sure, but I think there must be a resistance heater in the HVAC because it's blowing warm air within a couple minutes of setting off - but that's not the impressive part.

What really gets me is how quickly the engine gets up to temperature. The dash doesn't have a water temperature gauge, it has an oil temperature gauge, and after my easy ~22 min 22km (14 mi) commute to work the oil temp is right in the middle of its operating range (roughly 110C or 230F). On really cold mornings the oil temp doesn't quite make it to the middle of the gauge, but since oil temp generally trails coolant temp I'm reasonably confident the engine is fully warm. That's with the HVAC set to auto to prioritize heat in the cabin, and fuel efficiency seems to be not far off what I see in warmer weather to boot.

Winter heat is just an absolute non-issue with this car, which is in total contrast to my last car - a gas 2012 Chevy Cruze Eco 6MT. That car would never warm up if driven gently on really cold days, even at highway speed (there are lots of complaints from people even with the less efficient automatic equipped versions). On my short and easy commute I never got any appreciable heat out of it unless I left it in 5th (or sometimes 4th!) gear... burn more fuel, make more heat. I designed an elaborate plug-in heater system for that car consisting of the 200W OEM oil pan heater, a few stick-on pad heaters, a 1000W thermosyphon coolant heater and a 12V Prius water pump. With all that going on the engine would be just OK for commutes to work, but the physics were obvious; there just wasn't enough waste heat coming out of that little turbo 1.4 unless you deliberately drove it inefficiently.

So to wrap this up, for what it is I'm seriously impressed with the economy of this car given its lack of apparent shortcomings. It's spacious and supremely comfortable year-round in a difficult climate, an AWD tank in the snow, powerful and quiet, handles well, and is impressively efficient (to me at least). Purchased used for less than 1/3 its original price with 88k kms it seems like a good value so far. Considering I put premium fuel in my Cruze it barely costs more to fuel this car with diesel.

Makes sense. This car was asking for an oil change when I bought it, so I wonder if it got an oil change and a DEF top-up when it was returned at the end of its lease and they forgot to reset the oil life monitor? Possible I guess.

I've had a jug of DEF sitting in my garage since I bought the car assuming it will run out. Now I'm curious to see how far it will go before needing a refill. Also curious if that 9.5L (2.5 US Gal) jug will even top up the tank? :D

OK, so a little Googling says BMW claims 10k miles on a fill, so it must have been topped-up right before I bought it. 15k kms is getting close to 10k miles.

I never drove any vehicle fitted with SCR, as in my country it was initially fitted to heavy-duty commercial vehicles and the Agrale Marruá in 2012, and only in 2017 or 2018 it became available on some Peugeot/Citroën commercial vans. The first vehicles mostly focused on private customers fitted with SCR here are the '22 Jeep Compass and the recently-released Jeep Commander, yet now I see more of the flexfuel versions of the Compass because most of its buyers don't seem so willing to deal with SCR. I must confess I'm also not so fond of SCR, yet I look at it from a perspective similar to what 2-stroke motorcycle owners look at those automatic oiling systems with a separate oil tank instead of relying on premix.

Diesels can be "tuned" by adjusting the injection timing as it happens before TDC like the spark ignition in a car. The effects are similar.

Regen burning isn't always provided by a fuel injector in the exhaust. The fuel can be added to the exhaust through the engine's injectors if they are activated during the exhaust stroke.

Diesels tend to be more efficient than gassers at low loads due to lack of throttle. However, as far as I've researched, diesels are most efficient at full throttle. The more you spool up the turbo and the more fuel gets injected the hotter the combustion. The hotter the combustion the more energy can be extracted from it.

There is no throttle so adding EGR doesn't help efficiency like in a gasser, because in a gasser the EGR allows the throttle to be open more.

The only reason to run rich is to "roll coal". Even at full throttle diesels don't reach stoichiometric because there's no good way of completely mixing all the fuel into the air and you wouldn't gain any power as a result.

Even without EGR, a diesel will normally run cooler than a gasser even at full throttle because they run lean enough that the extra air is effectively an inert gas just like exhaust gas is.

What about lean mixtures running hotter? Ever fry a piston top in a VW bug? In the diesel F250, you watch EGT on hard pulls because the aluminium stuff tends to melt.

That's somewhat of an old wive's tail.

In my 1972 1600cc VW Super Beetle with mostly stock air cooled engine (stockish cam grind, stock VW heads, stock 34-4 Solex carb, stock oil bath air cleaner, stock VW heat exchangers, stock VW muffler, etc. etc. etc. just electronic distributor and heads machined for higher CR, did adjust pistons for .030 quench),

I ran the compression ratio up to 10:1, ran lean, about 16:1 or leaner at cruising (with the aid of an O2 sensor and AFR gauge) and adusjusted the ignition as far advanced as possible (most power without pinging) with a digital timing map and drove it mostly over steep mountain passes at full throttle and never had any problems at all. No pinging, no knocking, no fried pistons. The engine ran like a charm.

Why? Because leaner doesn't always mean hotter. Hot is around stoichimetric. A little leaner, like 15:1, is still hot. Go richer, 13:1, 12:1, etc. and it gets cooler. But going leaner, 16:1, 17:1, 18:1, (at 18:1 my Bug started misfiring) also runs cooler. In fact I actually had a hard time keeping my oil temps up as 180°F was the hottest I could get, and head temps never wanted to go over 325°F either.

I did make sure I got at least 12:1 at full throttle, but I drove all over, both going up steep mountain passes and along long stretches of road at highway speeds and never had any problems. I also averaged 30mpg.

I'm not the first VW aircooled owner to do this. Lots of others have found that they can do the same thing. Many airplane owners have found their planes run better leaner and cooler at loads below 75%.

The lean = hot and rich = cool depends on the context. Are we building a race engine and are getting AFR's of 13:1? Then yes, leaner is hotter and richer is cooler. And maybe you could do 16:1 in a race engine, but you wouldn't have the power.

I don't recall problems with pistons unless a broken retainer clip let's a valve drop onto the piston.

Traditional failure mode is unequal cooling on #1 that crisps a valve or blows out the cylinder top.

Nope, the AFR gets so lean that there is not enough fuel to get it very hot in a diesel at it's most lean AFR, wich is usualy idle.
The more fuel you inject, the hotter it gets untill you reach the stochiometric AFR.
If you run it richer than that, you just produce lots of smoke.
So you'll get the highest EGTs under high load when the engine injects a lot of fuel.

On a sidenote:
The higher the combustion temperature, the more power you make.
But your engine doesn't like it when it gets too hot, it starts melting if your combustion temperatures get too high, and gasoiline engines start to knock/detonate when you push it too far.

The higher the difference between combustion and exaust temperature determines how efficient the engine is, higher expansion ratios can increase that.

Odd enough, now some Diesels are getting an actual throttle, mostly due to the emissions control devices, while some gassers got rid of the throttle. Yet EGR is still more effective on gassers...

Sorry I'm late to this thread, I just got the email.

Not exactly. The purpose of the EGR is to reduce cylinder temp, which reduces NOx emissions. Any resulting change in AFR is just a side-effect.

I own some IDI engines and a common-rail. All of them have black oil, despite a disabled EGR and oil catch can. AMSoil sells a bypass oil filter that claims to remove soot and it's on my wish list.

We have a large mechanically-injected Cummins generator at work and the oil always looks pristine. I'm not sure what the secret is. Maybe because it never runs above 16% load except during semi-annual maintenance.

I drove a 2015 school bus with a Cummins 6.7 for two years and I never once saw the regen light turn on. Probably because I drove a rural route. I don't recall if the oil was black in that one.

The oil in diesels turns black quickly regardless of EGR as the raw particle emissions are pretty high and contaminate the oil.
Interestingly the opposite can be observed in CNG/LPG vehicles where the oil takes a long time to darken.

The EGR is *not* lowering the EGTs as long as the engine isn't running rich.
Even cooled EGR is hot compared to ambient air/charge air after the intercooler.
The point of EGR in diesels is to get closer to stochiometric ratio to reduce the amount of oxygen that can react with nitrogen, reducing the amount of NOx produced.
However that's a tradeoff with particle emissions as these increase when you run less lean.

With the possible exception of throttle-fitted diesels, mentioned by another poster but completely unknown to me, a diesel engine is controlling idle by metering the fuel. Air flow is unrestricted. When the engine is cold, more of the fuel passes through unburnt (you may see white smoke on a cold start). This results in a less efficient burn and thus more fuel is required to maintain the idle. This natural feedback loop results in a richer AFR entering the cylinder, but this is not the direct result of any direct AFR control on the part of the ECM.

Some newer diesels have an exhaust choke that activates when cold to increase backpressure and thus temperature, to help warm the engine. This creates a very distinct and noticeable sound on some Powerstroke engines especially that sounds like a jet turbine even at idle.

As for cold idling, diesels are very efficient at idle, and this can lead to problems such as slobbering and glazing due to insufficient heat in the cylinder. This is why many bigger diesels are equipped with high-idle controls. Check out an informative video on YT by Adept Ape called "Is Idling Your Engine Bad? Does Idling Hurt Your Engine?". I'm not allowed to post links.

I'm not an EGR expert but Wikipedia ("Exhaust gas recirculation") appears to disagree with your first and third statements. Yes, the recirculated gas is hotter than intake air, but it is lowering cylinder temps by reducing available oxygen, and the reduced cylinder temp is the mechanism of reduced NOx emissions. If you can show me a good source to refute this I will happily update the wiki page.

I could dig out a german textbook on internal combustion engines, but you'll likely not understand it because it's in german.
Regarding the temperature:
As long as all the fuel combusts, EGR does not lower combustion temperatures in *diesel* engines as the same amount of fuel combustion produces the same amount of heat.
It is different in gasoiline engines as here it dilutes the air/fuel mixture while keeping it stochiometric and thereby reduces combustion temperatures, where as in diesels it replaces intake air and enriches the air/fuel ratio.

In gasoiline engines it also improves efficiency under low load as it reduces pumping losses as the engine doesn't have to work against as much of a vacuum.
Meanwhile in diesel engines the engine doesn't have to work gaianst a vacuum to begin with.

Enabling the engine to run for a longer time at higher RPMs decreases the soot buildup at the DPF, so there are fewer and shorter regen cycles. On a sidenote, I guess those rural routes didn't have any of those no-idling policies.

Quote: "...Some newer diesels have an exhaust choke that activates when cold to increase backpressure and thus temperature, to help warm the engine. This creates a very distinct and noticeable sound on some Powerstroke engines especially that sounds like a jet turbine even at idle...."

My 1996 F250 Ford pickup truck has a 7.3L diesel engine with such a warm-up valve; they call it an Exhaust Back Pressure Valve, or EBPV . It is actuated via a 12VDC signal and operates via oil pressure. A lot of users eliminate the EBPV if and when it begins to leak oil or otherwise malfunction rather than repair or replace it.

Others, like myself, put a $5 electric switch on the EBPV's 12VDC control wire so that we can turn it on and off via the switch. The reason: for all practical purposes, the EBPV acts as an exhaust brake, which is very helpful coming down steep grades when towing a heavy load. I've often wondered why Ford and International, the engine's actual maker, didn't advertise it as an exhaust brake in addition to being a fast warmer-upper. If they had done so, they would have had the first diesel pickup truck on the market with an exhaust brake, which has now become a standard feature for diesel pickups.

The clean oil in the Cummins is a function of a properly broken in engine that is running a light fixed load. Throttling up/down for load induces more overfueling opportunities.
I worked as a deck engineer on a derrick crane that had a 350 Cummins for the hydraulics, and it was the same. At 250 hours it was a slightly darker amber color. It got throttled some, but the load was relatively light. The throttle applications were usually for just speeding up what you were doing at the moment.
https://i.postimg.cc/SQHtvLhg/IMG-20170831-104331.jpg

In comparison, the 400 Cummins on the frictions had dark oil right after change, more the typical diesel pattern. It worked hard on a big load.
https://i.postimg.cc/kM2ps1kV/IMG-20170831-104314.jpg

Ironically the 350 was swapped out for a Tier 4 Cat engine for an emissions credit with the feds. It had to be destroyed to get the $$ towards the Cat.
Here is the Cat in the same spot.
https://i.postimg.cc/RVfb53C4/IMG-20190905-144150.jpg

Here is a video of the 350 doing a little throttling, while picking up another crane off the bank and putting it on a barge. This is a largish derrick with 185' of boom and a nice 140 ton chart. The Bucyrus Erie 30B with boom and no counterweight was about an 85,000 pound pick at 65' radius.
https://youtu.be/2vJz6_Qr1BE

Even though the emissions credit seems to be worth, destroying such an engine which could be still operating somewhere else, eventually not within the United States and mayve repurposed as a remanufactured truck engine.

It’s much more likely that someone would just put it in another piece of equipment in the US instead of paying to export it. So the EPA wants the engine permanently disabled.

Diesels with variable geometry turbos will add restriction by closing the turbo vanes during warm up and braking.

they did the same tier 4 engine conversion on a dump barge, which was really a waste. The dump barge engine runs a fraction of working time.

I was recently trying to find information regarding the suitability of Steyr Monoblock engines to some alternate fuels, and it was mentioned the unit-injection system would either require some fuel density detector or having the injection adjusted individually for each fuel used. At least with a common-rail engine, it's simpler to just do a remap instead... An associate professor from Illinois Tech had even proposed the usage of gasoline on turbodiesel engines with just a remap.

https://www.autoevolution.com/news/d...te-186175.html

- There is optimal injection timing for best efficiency
- There is optimal boost level for certain injection level at certain RPM
- There is optimal cam (or cams) timing

You need to find those out by doing sweep runs and tests.

In general the EGR is not boosting efficiency so its not in use. I should still try to test will it help at low rpm and injection levels as there is a reason why it migth help to boost little bit efficiency.

In my own ecu programs is a AFR limitation to set 19 meaning, I will never go richer than that. There is no need. At higher rpms AFR can go over 45 if you want to get best efficiency.

There might be only a handful of times when EGR is actually beneficial. When it comes to cam timing, even though most Diesels don't resort to a variable cam timing, it may even be suitable to provide some internal EGR effect similar to what Mazda applied to that SkyActiv-D, and that was not bad at all. At least it's way less likely to cause carbon buildups to the intake manifold...

Good question.

On a gasoline powered engine with a throttle the EGR makes the charge produce less power since it has less gasoline which in turn allows the throttle to open more to produce the same amout of power again. A more open throttle means less pumping losses.

But in a diesel there is no throttle, so how could EGR in a diesel increase efficiency?