This article was first published in 2008.
Do a web search under ‘BSFC’ and you’ll find
results including Birkenhead Sixth Form College, the Bouncing Souls Fan Club and
the Black Swan Folk Club. But my favourite is the Boston Society of Film
Critics. I mean what do they do - sit around together, criticising films?
However, you’ll also find under BSFC this term –
Brake Specific Fuel Consumption. And that’s what we’re going to talk about
Specific Fuel Consumption
When we use the word ‘specific’ in an engineering
context, we’re describing something in the light of a comparison. So, ‘specific
power’ is probably better known to you as ‘kilowatts per litre’, or ‘hp per
cubic inch’. Simply, it’s the peak power divided by the engine capacity.
Specific fuel consumption is similar. It’s the
amount of fuel consumed, divided by the power being produced.
So it could be expressed in litres of fuel divided
by the kilowatts developed. In fact, the fuel quantity is normally described as
a weight in grams or pounds, so Specific Fuel Consumption is expressed as:
Sometimes you’ll find other units being used as
well. However, don’t get too hung up on the units –
remember, they all express fuel divided by power.
Specific Fuel Consumption
what’s Brake Specific Fuel Consumption? As with bhp
(brake horsepower), this refers to the specific fuel consumption when the
power is measured by an external brake – in other words, a dyno. Most times, 'brake specific fuel consumption' and 'specific fuel consumption' are the same thing.
Power, Torque and BSFC
The easiest way of understanding Specific Fuel
Consumption is to look at an engine performance graph.
Here’s a good one from a classic engine – the
Jaguar V12 HE, the ‘HE’ indicating the use of high swirl heads.
The red line shows power – all the way to about
220kW in this graph. The green line shows torque (although here it is expressed
as Brake Mean Effective Pressure). And then we have the newie – SFC, shown by the purple line. As can be
seen, the SFC curve doesn’t initially appear as you might have imagined it
Firstly, at idle it’s about 280 g/kWh, then as
revs rise, it drops to be at its lowest at about 2500 rpm (at say 270 g/kWh).
From there it rises steeply to reach 350 g/kWh at 6000 rpm.
Firstly, why should the SFC be lowest at middle
revs? Or, to put this another way, what causes an increase in fuel used per kW
at both low and high revs?
At low revs, SFC suffers because there’s increased
time for the heat of combustion to escape through the walls of the cylinders and
so not do useful work. At higher engine speeds, the frictional loses of the
engine rise alarmingly (especially in this case with 12 cylinders!) and so the
energy of combustion is again being wasted, this time in heating the oil.
There’s another reason that SFC is lowest at
‘middle’ rpm. Because the engine is tuned to develop best cylinder filling (ie
to produce best torque) at middle revs, the engine’s breathing is at highest
efficiency at these speeds. But don’t fall into the trap of saying that SFC is
always at its best at peak torque – that’s not usually the case.
But the real trouble with diagrams like the one
above is that in many ways, they’re irrelevant to real-world fuel consumption.
Why? Because these graphs are drawn for full throttle! So if the
engine is powering a racing ski-boat travelling constantly at full load, then
yes, the shown SFC data is all well and good. But what happens at part throttle,
as occurs in nearly all normal car use?
Well, then, the situation is very different! And
the trouble is, the SFC figures are always much worse...
you’re having difficulties coming to grips with this (“What? He reckons fuel
consumption gets worse at small throttle openings?”), remember that we’re
talking about specific fuel consumption – the amount of fuel used per
if the BSFC gets worse at lower throttle angles, the power must be being reduced
at a quicker rate than the decrease in fuel consumption...
Light Load SFC
This graph shows what happens at lighter engine
loads – it’s from a Repco manual for “a typical four cylinder” engine. The SFC
is expressed this time in pounds per horsepower hour – but as we said earlier,
it doesn’t matter what units are used.
At 100 percent load (ie wide open throttle) this
engine has a minimum SFC of 0.43 – see the bottom curve. As we by now expect, at
both lower and higher revs that this, the SFC rises.
But have a look at what happens at 50 per cent
load! The SFC results at half load and 1000 rpm (ie idle) doesn’t matter much
(when would you be in that situation?) but at 2000 rpm, the SFC has gone up by
13 per cent. At 4000 rpm, it’s gone up by just under 30 per cent!
And keep in mind that in normal use, even 50 per
cent is a lot of throttle. A more frequently used load is 25 per cent. At 25 per
cent load, the SFC at 2000 rpm has risen by a massive 117 per cent over that
achieved at full load! You can also see from the shape of the 25 per cent load
curve, BSFC is even more heavily influenced than ever by the rpm being used.
So what accounts for this terrible decrease in SFC
at just the throttle openings the engine will be used at most often? ‘Throttle’
is the key word here – as the engine is increasingly throttled, it has to work
harder and harder at drawing air past the throttle blade. This is the reason
that there is a measurable vacuum after the throttle blade – the engine is
trying to drag in more air than it is being permitted to. Each time a piston is
descending on the intake stroke, it’s having to do this extra work. Working
internally hard as a vacuum pump means there’s less power available at the
This work against the throttle restriction is
referred to as ‘pumping losses’.
When You Close the Throttle
Remember above where we said that if the SFC gets
worse at lower throttle angles, the power must be being reduced at a quicker
rate than the decrease in fuel consumption?
These graphs clearly show the effect. They’re
taken from the 1976 edition of Oldhams New Motor Manual.
The top graph shows the power output at quarter
throttle, half throttle and full throttle. The bottom graph shows what happens
to the SFC at these different throttle positions – and doesn’t the
quarter-throttle SFC rocket, especially at higher revs!
So if SFC is much worse at lower loads, what
actually happens in the real world? Let’s start off by showing a SFC graph in a
slightly different way.
This diagram has engine revs along the bottom axis
and engine load (expressed in BMEP) on the vertical axis. The lines on the
diagram join points of equal SFC. This and the following two diagrams are
sourced from a 1999 report prepared for the Canadian government by Sierra
Research of California. The diagrams are based on a sample of 1995 model year,
naturally aspirated, EFI 2-valve engines.
Here colour has been added to the graph (good,
isn’t it?!) to show more clearly where the different ‘islands’ are located. The best BSFC is the red area centred around 2000 rpm and
And now we have the killer. Here each dot shows
the speed and load for a typical mid size car at 1 second intervals during the
US fuel economy test. Of the time the car takes to do the test, just 5 seconds
are in the island of best BSFC. Quite a few of the dots (the authors say that
they overlay) are at worst BSFC – idling at zero load with the car
full report can be seen at www.tc.gc.ca.
It has some more interesting BSFC ‘island’ diagrams showing the advantages of
turbo and multivalve engines.
Here are some interesting BSFC charts to look
The Toyota Prius petrol electric hybrid uses its
pseudo-CVT and strong low-rpm electric assist to keep the 4 cylinder internal
combustion engine working as much as possible in the area of low SFC (red
The Honda Insight hybrid uses a 3 cylinder VTEC
engine. This diagram, taken from a French engineering investigation of the car,
shows measured SFC for the engine. The testing was of the CVT transmission
Insight and was done on a chassis dyno.
Note how the blue/yellow island of best SFC is
achieved at relative high revs and load, and how there’s a second area of low
SFC at about half load and 1500-2000 rpm. I assume that this second area is
achieved through the VTEC mechanism, that in this car, at low revs shuts off one
of the two inlet valves for each cylinder, promoting better swirl.
(Apparently the CVT Insight does not use lean
cruise – that is, very lean air/fuel ratios at constant throttle. The BSFC map
for such a car would make very interesting viewing!)
Not every engine has its lowest BSFC at moderate
rpm. This quad rotor rotary Le Mans engine has minimum BSFC at 6000 rpm!
BSFC is a good concept to have in your mind, not
because it sounds impressive when you chuck it into a conversation about fuel
economy, but because it makes you think about things in a different way.
For example, anything that allows you to keep the
throttle open wider and the revs lower (like changing up to a tall gear and then
holding it) will reduce fuel consumption because BSFC will be improved. But
equally, recycling exhaust gas (ie EGR) might also achieve that same effect
because pumping losses will be reduced - not all the inlet charge needing to
come past the throttle.
Diesels, which we’ve not mentioned here, are much
more efficient at low loads because they don’t have a throttle restriction in
any type of driving – low loads are catered for by just reducing the fuel that’s
As you can see, BSFC – and how it changes with
load - is one of those fundamental concepts that helps explain a lot.
TC, et al, (1999) Alternative and Future Technologies for Reducing Greenhouse
Gas Emissions from Road Vehicles, Sierra Research Inc
(2004) Gasoline Engine Management
L (Ed), (1976) Oldhams New Motor Manual, Hamlyn Publishing Group
(1972) Repco Engine Service Manual
R, (1985) Introduction to Internal Combustion Engines, Macmillan
R, et al, (2003) Hybrid Light Duty Vehicles Evaluation Program,
International Journal of Automotive Technology, Vol 4, No 2