Adventures in Hypermiling - Increasing MPGs

[Butcher]

Firstly, fuel and air mix in a petrol car must be in perfect union. Only in a diesel can you run excess air. As boost pressure rises, in a petrol car, petrol injection rises too. It has to to increase power.

Not strictly true. Lean burn is quite common in petrol engines these days, though only under restricted conditions due to the problems with NOx and heat build up.
There's a list of some examples here: https://en.wikipedia.org/wiki/Lean-burn

Also when you boot it most petrols go rich as it's safe to run a little excess fuel and you get better power (at the expense of CO emissions). If you watch the AFR on a car with on-board monitoring (most stuff made in the last 15 years) you can see the ratio drop as you open the throttle. Mine goes to about 12:1 from memory (14.7:1 is stoichiometric for petrol).

[Zak33]

Not strictly true. Lean burn is quite common in petrol engines these days, though only under restricted conditions due to the problems with NOx and heat build up.
There's a list of some examples here: https://en.wikipedia.org/wiki/Lean-burn

Also when you boot it most petrols go rich as it's safe to run a little excess fuel and you get better power (at the expense of CO emissions). If you watch the AFR on a car with on-board monitoring (most stuff made in the last 15 years) you can see the ratio drop as you open the throttle. Mine goes to about 12:1 from memory (14.7:1 is stoichiometric for petrol).

All true.. going rich yes. But that's not more efficient. It's just safer for no detonation and to prevent short term (seconds ) of underload overheating where pre ignition occurs. (long term running rich creates excess heat but that is yes another story)

Lean burn is a thing but it's irrelevent to this discussion about how a turbo makes a car more powerful, because its only frationally "leaner"

What I wanna ensure is that everyone understands what a turbo does, and how....no confusing the turbo theory with the idea of more air- but not more fuel. With more air, we always needs more fuel.
Only diesel run excess air

[kalniel]

I thought lean burn petrol was not possible in the UK as MOT specifies lambda of 1 (with v. small tolerance).

[Butcher]

All true.. going rich yes. But that's not more efficient. It's just safer for no detonation and to prevent short term (seconds ) of underload overheating where pre ignition occurs. (long term running rich creates excess heat but that is yes another story)

Depends what you mean by more efficient.
It gives better power for a given engine; somewhere around 12:1 gives maximum power. Also somewhat counter-intuitively the additional fuel helps cool the engine rather than heat it (vaporising fuel is a large heat sink in a petrol engine), hence why a rich mix is used under high load, it helps to prevent the engine from overheating. The actual risk for running lean long term is burning out the catalytic converter, as unburnt fuel tends to make them run hot.

Lean burn is a thing but it's irrelevent to this discussion about how a turbo makes a car more powerful, because its only frationally "leaner"

What I wanna ensure is that everyone understands what a turbo does, and how....no confusing the turbo theory with the idea of more air- but not more fuel. With more air, we always needs more fuel.
Only diesel run excess air

Lean burn has been used up to around 22:1 ratio in petrols, that's a fair bit more than fractionally leaner. Actual peak efficiency is usually at around 16:1 though.

Also, something else to consider with turbos is that they heat the air (a lot), which then makes engine temperatures higher. That's usually combatted by that old standby, more fuel. So if the turbo is working hard, not only are you sucking in more air, so burning more fuel, but the ECU is probably running a rich mixture to keep temperatures down, which means a double fuel hit. So yes, turbos are definitely not just more air, they are more fuel and then a bit more fuel on top of that.

I thought lean burn petrol was not possible in the UK as MOT specifies lambda of 1 (with v. small tolerance).

Under MOT test conditions, yes. Also lambda=1 is good for catalytic converters, so most cars tend to run close to it when possible (this is another issue that makes lean burn a bit pants).

[DanceswithUnix]

I thought lean burn petrol was not possible in the UK as MOT specifies lambda of 1 (with v. small tolerance).

It is possible, but just one of the tools the engineers have when designing an engine.

Alfa used to make "jts" engines where supposedly they used the direct inject to explode a small amount of petrol in the middle of the cylinder so there is a layer of cooling air surrounding the heat of the lean explosion. It was hard to make it work with multi-air though which is more effective, so they dropped the technology and stopped using direct injection.

Depends what you mean by more efficient.

More power out for a given amount of petrol going in.

People do seem to confuse it with bhp per litre, but that isn't efficiency.

[badass]

Prepare for your minds to be blown:

Petrol lean running burns cooler

Seriously. It does. The hottest burn is at the perfect stochiometric ratio. leaner than that and it burns cooler. i.e. over 16:1 and it runs cooler than at 14.7:1
So why does everyone say that lean running is hotter? Simple. Cars don't run 14.7:1 - they run rich almost all the time. So they run cooler. Make them run closer to 14.7:1 and they run hotter as the mixture gets leaner, then over 14.7:1 they burn cooler.

On the diesel turbo on the motorway thing: My completely standard Golf diesel produces about 8 PSI when maintaining 70 MPH. i.e. It's boosting when on the motorway. If it was identical in every way except minus the turbo and had the higher compression ratio to make up for it, it would use less fuel to maintain the same power to do 70 MPH. (idealised) Engine efficiency increases with increasing compression ratio.

In reality however it's probably better on fuel than a N/A 2.0 Diesel at motorway speeds simply because it's got taller gearing. Diesel fuel economy is hurt far more with extra revs than even petrol fuel economy because the higher compression results in more pumping losses.

[DanceswithUnix]

Combustion isn't that simple though is it. When I bought my kit car, the guy that built it had put the wrong spark plugs in as a performance tweak. Bigger spark, faster combustion, higher pressure in the cylinder (yay!) leading to higher temperatures and eventually a hole burnt in the piston (boo!). That's without any change to the fuel air mixture, spark timing or anything else. Just AIUI a slower burn rate means the gasses are cooling more from expansion during the burn.

[kalniel]

Prepare for your minds to be blown:

Petrol lean running burns cooler

Seriously. It does. The hottest burn is at the perfect stochiometric ratio. leaner than that and it burns cooler. i.e. over 16:1 and it runs cooler than at 14.7:1
So why does everyone say that lean running is hotter? Simple. Cars don't run 14.7:1 - they run rich almost all the time. So they run cooler. Make them run closer to 14.7:1 and they run hotter as the mixture gets leaner, then over 14.7:1 they burn cooler.

Again, I don't see how that's borne out at MOT - running rich = lambda <1 and MOT fail. As far as I can tell, cars only run rich at wide open throttle which they're allowed to for protection.

EDIT: To partly answer my own question, there's an appendix to the MOT guidelines which list a ton of cars who's manufacturers have specified safe lambda values for their specific catalytic convertors.

https://assets.publishing.service.go...th-edition.pdf

In almost all cases, it's still the usual 0.97 to 1.03, but you can see specific cases like BMWs N43 has a range of 0.7 to 4 - that's what I call lean burn!

[Butcher]

Prepare for your minds to be blown:

Petrol lean running burns cooler

Seriously. It does. The hottest burn is at the perfect stochiometric ratio. leaner than that and it burns cooler. i.e. over 16:1 and it runs cooler than at 14.7:1
So why does everyone say that lean running is hotter? Simple. Cars don't run 14.7:1 - they run rich almost all the time. So they run cooler. Make them run closer to 14.7:1 and they run hotter as the mixture gets leaner, then over 14.7:1 they burn cooler.

Seriously, it doesn't.
Allow me to present this image:

As mixture leans out above stoichiometric efficiency increases. Higher efficiency means higher heat (more or less by definition, if you're getting more energy out, you're getting more heat). Once you get past the trough of the curve it cools down again as you're then past maximum efficiency.

Stoichiometric just means there is exactly enough oxygen for the fuel. Running slightly leaner means that the fuel will tend to burn slightly faster; combustion relies on fuel and oxygen molecules colliding in the vapour phase and reacting, more oxygen means more chance for a fuel molecule to collide with one. However you also need proximity between reactions; too much air (i.e. very lean) leaves too much space between reactions which then slows the reaction down. This is also why EGR is so effective, you're adding excess nitrogen to the mixture which slows combustion reducing temperatures.

This gives the graph above, as you go past stoichiometric the increased oxygen increases reaction rate and give better efficiency (and heat) until the reduction in reaction rate from too much air overtakes it and reduces the efficiency.

Again, I don't see how that's borne out at MOT - running rich = lambda <1 and MOT fail. As far as I can tell, cars only run rich at wide open throttle which they're allowed to for protection.

Yes. Petrol cars run very close to stoichiometric most of the time (closed loop mode) because running rich or lean for too long will damage the catalytic converter or give poor emissions (or both). When you floor it the ECU will go into open loop mode where it just chucks in fuel to keep temperatures down, but this is inefficient. 30+ years ago badass would have been correct as carburetted petrols did tend to run slightly rich all the time.

The typical regime is actually to run very slightly rich and very slightly lean in an oscillation (typically 0.01 lambda difference or there abouts, so really very slight). This gives better cat performance than running bang on lambda = 1 all the time.

[badass]

Again, I don't see how that's borne out at MOT - running rich = lambda <1 and MOT fail. As far as I can tell, cars only run rich at wide open throttle which they're allowed to for protection.

EDIT: To partly answer my own question, there's an appendix to the MOT guidelines which list a ton of cars who's manufacturers have specified safe lambda values for their specific catalytic convertors.

https://assets.publishing.service.go...th-edition.pdf

In almost all cases, it's still the usual 0.97 to 1.03, but you can see specific cases like BMWs N43 has a range of 0.7 to 4 - that's what I call lean burn!

At MOT IIRC its only tested at idle/fast idle. i.e. almost no load. I believe this is alluded to by butcher earlier in this thread.

[Butcher]

At MOT IIRC its only tested at idle/fast idle. i.e. almost no load. I believe this is alluded to by butcher earlier in this thread.

To be fair from what I've seen it's representative of most driving in terms of AFR. It's only when you have the throttle wide open or close to it that it enriches the mixture. Typical cruising load will be running closed loop unless you're hammering down the autobahn at 150 mph.

[kalniel]

At MOT IIRC its only tested at idle/fast idle. i.e. almost no load. I believe this is alluded to by butcher earlier in this thread.

Yes, true - emissions is tested at fast idle ie just held at 2500-3000 rpm so it's not idle but equally not much load. But that's even more vital for a lean burn as you'd intuitively think low load condition is exactly when you would want to lean burn. I guess that's the reasoning for cylinder deactivation becoming a more common thing - you keep the cycle and mixture the same, just reduce the capacity of the engine.

[badass]

To be fair from what I've seen it's representative of most driving in terms of AFR. It's only when you have the throttle wide open or close to it that it enriches the mixture. Typical cruising load will be running closed loop unless you're hammering down the autobahn at 150 mph.

True. I believe that all VW petrols since the mid noughties now use Wide band lambda sensors. I presume this is so they can remain in closed loop mode at more boost/wider throttle openings. Probably because they went big on Turbocharging with Direct injection instead of larger engines - i.e. engines run under more load more of the time so they want them to be more accurately fuelling under a wider range of conditions.

[badass]

Yes, true - emissions is tested at fast idle ie just held at 2500-3000 rpm so it's not idle but equally not much load. But that's even more vital for a lean burn as you'd intuitively think low load condition is exactly when you would want to lean burn. I guess that's the reasoning for cylinder deactivation becoming a more common thing - you keep the cycle and mixture the same, just reduce the capacity of the engine.

I think so. My take on it is that 2 cylinders running in the Otto cycle with the drag from 4 cylinders but only the valves of 2 cylinders is more efficient than running 4 cylinders and the full valvetrain running in the Atkinson/Budack cycle

[badass]

My take on that image: With the normal imperfect mixing of any engine, 12.6:1 makes sure there's enough petrol to burn all of the oxygen. When running 15.4:1 there is enough oxygen to make sure all of the fuel is burnt. Side note: I'm not disagreeing at all BTW. The typed word can often lead to misunderstandings.

[peterb]

Seriously, it doesn't.
Allow me to present this image:

As mixture leans out above stoichiometric efficiency increases. Higher efficiency means higher heat (more or less by definition, if you're getting more energy out, you're getting more heat). Once you get past the trough of the curve it cools down again as you're then past maximum efficiency.

Stoichiometric just means there is exactly enough oxygen for the fuel. Running slightly leaner means that the fuel will tend to burn slightly faster; combustion relies on fuel and oxygen molecules colliding in the vapour phase and reacting, more oxygen means more chance for a fuel molecule to collide with one. However you also need proximity between reactions; too much air (i.e. very lean) leaves too much space between reactions which then slows the reaction down. This is also why EGR is so effective, you're adding excess nitrogen to the mixture which slows combustion reducing temperatures.

This gives the graph above, as you go past stoichiometric the increased oxygen increases reaction rate and give better efficiency (and heat) until the reduction in reaction rate from too much air overtakes it and reduces the efficiency.


Yes. Petrol cars run very close to stoichiometric most of the time (closed loop mode) because running rich or lean for too long will damage the catalytic converter or give poor emissions (or both). When you floor it the ECU will go into open loop mode where it just chucks in fuel to keep temperatures down, but this is inefficient. 30+ years ago badass would have been correct as carburetted petrols did tend to run slightly rich all the time.

The typical regime is actually to run very slightly rich and very slightly lean in an oscillation (typically 0.01 lambda difference or there abouts, so really very slight). This gives better cat performance than running bang on lambda = 1 all the time.

This can be demonstrated very simply on a carbureted engine like a lawn mower engine. Starting is accomplished by a rich mixture (set by restricting the air inlet - the choke)

Once started, the choke is opened as the engine warms up - if you don't do that the engine will stall. As the choke is opened, the engine speed will increase as the mixture becomes leaner.