Even those with only a cursory interest in
automotive technology can see the changes looming from just beyond the horizon –
the revolution that will either (depending on your perspective) enlighten or
BMW currently has their hydrogen-powered 7-series
doing the world rounds of breathless journalists; Toyota’s hybrid
petrol/electric Prius has achieved a success that the company could only have
dreamt of; and electric cars like the Tesla Roadster look set to change the
On the fringes are bio-diesel and ethanol and CNG
And not for one moment to be underestimated are
conventional petrol engines - but boasting efficiency improvements like direct
fuel injection and turbocharging.
Listen to the hydrogen exponents and you’d think
there’s the answer. Listen to the biodiesel enthusiasts and there’s no doubt –
we all need to use that renewable fuel. But then again, the arguments of those
that believe in hybrids seem persuasive....
So how on earth do we make sense of all this? What
are really the best answers – the automotive technologies that will
give the optimal outcome in terms of energy consumption and greenhouse gas
A seminal paper written by Andrew Simpson draws
back the cloud of obfuscation, vested interests and often pure ignorance that
confuses these most important of issues. Produced when the author was working
for the Sustainable Energy Group at the University of Queensland, the paper in
one fell swoop undercuts many of the ideas we’ve started taking for granted.
The data is absolutely vital reading for anyone
interested in where automotive technology should be heading.
Making Sense of it All
For the first concepts to get across are
‘Well-to-Tank’ and ‘Tank-to-Wheel’ figures.
When we asses vehicle efficiency, we most often
talk about the fuel consumption of the vehicle. For example, an efficient car
might be able to travel 100 kilometres on 5 litres of petrol. We call the
resulting fuel economy 5 litres/100km - and reckon it’s pretty good. (And it’s
probably also going to be good in terms of greenhouse gas emissions, too.)
This figure can also be termed the
tank-to-wheel efficiency: that is, how much of the energy of the
fuel in the tank makes it to the wheels as propulsion for the car. In these
terms, a diesel engine car is more efficient than a petrol engine car; a 5-litre
V12 engine car is less efficient than a 2 litre 4-cylinder.
But the concept doesn’t apply just to internal
combustion engines. Electric motors are highly efficient (vastly more so than
internal combustion engines) and so the tank-to-wheel efficiency of an electric
car is likely to be very high. (In this case, the ‘tank’ is the battery.)
Another car with a high ‘tank-to-wheel’ efficiency
is a fuel cell car, a car that uses the chemical reaction in a special stack to
produce electricity that in turn powers an electric motor.
But in terms of overall efficiency,
the tank-to-wheel figure is often a complete furphy, misleading and sometimes
Why? Read on...
Clearly just as important as the Tank-to-Wheel
efficiency is the Well-to-Tank data.
In other words, how much energy does it take to
refine the fuel so that it is suitable for going into the car’s tank?
Think of a petrol engine car. Petrol is not able
to be dug straight up out of the ground and labelled with an octane rating.
Instead, it needs to be laboriously located, obtained using energy-intensive
giant oil rigs (either land or sea), piped and transported to plants where it is
refined into the substance that we know as 98 RON unleaded. Each of these steps
takes energy and releases greenhouse gas emissions.
Compare that with (say) a wind generator producing
electricity. The wind is free in terms of both energy and greenhouse gas
emissions. Capturing that wind energy as electricity requires a generator and
tower and rotor; but after those have been paid for, the energy is free of
encumbrance. Then, even if the wind generator itself is only 20 per cent
efficient, getting the electricity into the battery that powers the car (the
“tank”) will be very low in energy and greenhouse gas cost.
So the well-to-tank figure takes into account the
on-going cost of energy and greenhouse gas emissions for producing the fuel the
car runs on.
If any sense is to be made of what is best for all
of us, clearly it makes sense to include both the Tank-to-Wheel and the
Well-to-Tank figures – done by using a Well-to-Wheel figure.
The Well-to-Wheel figure is the whole box and
dice: the total energy and greenhouse gas emissions cost of turning the wheels
of your car. It is – to excuse the expression – the no bullshit evaluation of
the best approaches to powering the cars of today and tomorrow.
The Well-to-Wheel figures take into account not
only the efficiency of the device powering the car but also the way in which the
fuel to power that engine has been produced.
Without that background it is simply impossible to
evaluate the real situation. If hydrogen fuel cell electric cars produce little
emissions and run for a long time on their tank of fuel, does that make them the
panacea? Not if producing the hydrogen fuel in the first place takes enormous
energy... Do biodiesel cars make a lot of sense if the burning of the fuel adds
little total CO2 to the planet? That depends on how the fuel is produced in the
Think about it for a moment and you’ll realise
that the Well-to-Wheel figure is an absolutely vital parameter in evaluation of
alternative car technologies.
First Half of the Story
Let’s look at what will be most familiar – the
energy-efficiency (and greenhouse gas emissions) of the tank-to-wheel process.
That’s as simple as how much energy of the fuel in the tank (and remember the
“tank” might be a battery) gets to the wheels, assessed in equivalent litres/100km of petrol. (In other words, the total energy consumption over
the test expressed in terms of petrol per 100 kilometres.)
Andrew Simpson’s paper was first published in
2003, with updates in 2004 and 2005. We have communicated with the author and he
believes that the results are still valid today (2008). His benchmark was a 2003
model Holden VY Commodore. The Commodore, when assessed over the New European
Driving Cycle, had a fuel consumption of 10.1 litres/100km.
Andrew Simpson evaluated no less than 32 other
forms of propulsion (click on the graph to enlarge). Importantly, these alternative cars were modelled to
have the same performance and range as the Commodore.(The exceptions to this
are the nickel metal hydride and valve regulated lead acid battery electric
vehicles where it is not technically feasible that they have a driving range
matching the Commodore.) The fact that most of the vehicles matched the
Commodore in terms of performance and range is very important because this
‘level playing field’ is seldom adopted when comparing alternative technologies.
For example, this criterion meant that the battery electric cars were relatively
heavy – as they would be in the real world.
Rather than look at all the alternative
technologies modelled by Andrew Simpson for their tank-to-wheel performance,
let’s look at those most in the news.
As already indicated, the Commodore scored 10.1
litres/100 in the standardised test (we wonder if the current VE model would do
as well!). An LPG-fuelled equivalent internal combustion engine Commodore was
barely any better at 9.7 litres/100, while the best performance (8.1 litres/100)
using an internal combustion engine was when it was fuelled with diesel. Note
that biodiesel (8.2) was actually slightly worse than conventional petroleum
diesel. This makes sense when you consider that the energy density of biodiesel
is lower than that of conventional diesel.
Then come the hybrid internal combustion engine /
electric cars – one running on petrol turns in a 7.7 litres/100 figure and a
diesel fuelled hybrid would get 6.9 litres/100.
(It’s worth being reminded at this point that all
cars have the same performance and range.)
Fuel cell electric vehicles? Well, if they drink
petrol, they’re terrible. But on hydrogen they’re much better – as low as 5.1
But best of all are the pure electrics – remember,
their electric motors are very efficient and so nearly all of the battery juice
ends up pushing the wheels – an energy equivalent of as low as 3.5 litres/100km
for a lithium-ion battery electric vehicle (and the Li-Ion car has the same
range and performance as the Commodore).
So, in really general terms, the ascending order
of merit for the Tank-to-Wheel figures go:
Petrol internal combustion engine
Hydrogen fuelled internal combustion engine
Hybrid petrol electric
Hybrid diesel electric
Hydrogen fuel cell electric
But as we’ve already mentioned, that is not the
full story – in fact it’s only half the story!
The Full Story
Rather than looking at the well-to-tank figures
and then correlating them with the tank-to-wheel figures described above, let’s
jump straight to the vital stuff: the well-to-wheel figures. The figures that
really matter, that take into account both tank-to-wheel and well-to-tank
In this diagram (click on it to enlarge) the Commodore is normalised to “1”
– anything above that is worse and anything below that is better.
The first shock is there are approaches that are
worse than the benchmark - considerably worse. Using coal to produce
either liquid or gaseous hydrogen that is run in a fuel cell car has an energy
consumption that is between 15 and 46 per cent worse than the Commodore on
petrol. In greenhouse gas emissions they’re between 35 and 75 percent worse!
That’s a really vital point to keep in mind:
despite the high technology of gasifying coal and then running a fuel cell
hybrid electric vehicle, the result is worse than the status quo.
Using a coal-powered electrical generating station
to charge a lithium ion battery electric car also gives results which are at
best marginal – fractionally better energy efficiency than the Commodore but
about 17 percent worse greenhouse gas emissions.
So here’s another vital point. Even a battery
electric car, despite its high tank-to-wheel efficiency, is poor overall if it
uses a coal-powered generating station for its juice.
Another approach that is worse than the petrol
powered Commodore is using a methanol fuel cell hybrid electric vehicle, where
the methanol is chemically refined from natural gas.
So what is clearly better than the Commodore?
Hybrid petrol electric and diesel electric cars are better than the Commodore
(and very similar to each other), while hybrid LPG cars are better again.
But – as you would expect if you’ve been paying
attention all this time - it is the renewable energy fuelled cars that do best
overall. The winner is the renewable energy powered (eg solar or wind) lithium
ion battery electric car – it has no greenhouse gas emissions and the best
overall energy usage of the lot, using only 33 percent of the energy of the
This story is most important in highlighting what
doesn’twork as much as what does work. Despite impressive high
technology, some approaches (including those being touted as panaceas by some
major car companies) have appalling energy and greenhouse gas results. Others
that may look the same (eg battery electric cars) can have completely different
results, depending on how that electricity is generated.
Andrew Simpson writes (with our emphases):
The best way to utilise an energy feedstock is via
as direct a pathway as possible, avoiding unnecessary energy conversions. This
is an important conclusion in regard to synthetic fuels such as hydrogen. For
the pathways considered in this study, it is preferable not to use hydrogen
since the energy losses and emissions incurred in the production of hydrogen
outweigh the higher efficiency of fuel cell-based powertrains.
Use of coal as a feedstock for production of
vehicle fuels will result in extremely high levels of full cycle energy
consumption and greenhouse emissions.
Hybrid Electric Vehicles using conventional
fuels (petrol or diesel) offer significant near-term reductions in energy
intensity and greenhouse gas emissions.
Natural gas appears to be a promising
transitional energy feedstock for automotive fuels. Hybrid Electric Vehicles
fuelled with Compressed Natural Gas or Liquefied Natural Gas offer even greater
reductions in energy intensity and greenhouse emissions than Hybrid Electric
Vehicles using conventional fuels. Natural gas-fired electricity can also be
used to charge electric vehicles resulting in the lowest greenhouse emissions of
any natural-gas pathway.
Based on their energy intensity, biofuels may
not be the most practical method for reducing greenhouse gas emissions.
Well-to-wheel pathways based upon renewable
electricity generation offer near-zero greenhouse gas emissions. However,
renewable electricity should be utilised directly in an electric
vehicleto avoid the energy intensity of converting it into other fuels (i.e.
– unleaded petrol
– liquefied petroleum gas
– compressed natural gas
– liquefied natural gas
– compressed hydrogen
– liquefied hydrogen
– internal combustion engine
– fuel cell electric vehicle
– hybrid electric vehicle
– battery electric vehicle
– fuel cell hybrid electric vehicle
Ion – lithium ion
– nickel-metal hydride
– valve regulated lead acid
well-to-wheels study considers only energy inputs/emissions outputs used for
driving. But what about the energy used (and the emissions created) when the car
is being manufactured? And what about the emissions and energy for development
of the fuel creation infrastructure (whether that’s a power station or an oil
rig)? These latter factors are taken into account only in a ‘cradle-to-grave’
Simpson says: “I did not include the ‘embodied energy/emissions’ (as they’re
known) in my study.
you should include everything, but as a practical matter, you have to set the
system boundary somewhere. A rule-of-thumb that I have seen is that the
embodied energy in making a car is ~10% of the energy it consumes via driving
over its life. This of course varies by technology, and doesn’t mean you
can ignore the embodied factors, but I think my comparison was informative
the full paper below: