More on How Much Power You Really Need

Posted on January 29th, 2009 in Aerodynamics,Driving Emotion,Electric vehicles,Opinion by Julian Edgar

Back in this blog I mused about how little power is actually needed in a car. My benchmark was not acceleration or top speed. Instead, it was the ability of a car to climb hills at the open-road speed limit (here in Australia, 110 km/h).

Based on dyno tests and the hill-climbing performance of my diesel Peugeot 405, I decided an at-the-flywheel figure of 35 kW/tonne was about the right minimum.

I applied that idea to electric cars, where for reasons of lightness, battery power consumption and cost, an electric motor that errs on the side of smallness makes sense.

A number of comments were then made that this was completely wrong, that electric motors don’t work in that way (apparently, an electric kilowatt is different to a petrol motor kilowatt!) and so on. However, I saw no evidence that suggested a power/weight ratio of about 35kW a tonne was not the minimum for a car to be competent on the open road. (And a reasonably aerodynamically slippery car, at that.)

Recently, I’ve driven three cars that have a instantaneous power output display on the dashboard. These are all Lexus hybrids – the LS600hL, the GS450h and the RX400h. The latter’s display is shown above.

With this gauge I was able to see exactly how much power was being transmitted to the wheels, irrespective of torque curves, throttle position or anything else.

The actual power going to the wheels.

The RX400h weighs-in at 2040kg – say with my body mass, 2130kg.

Typically, I used in normal driving – even sporty urban driving – an indicated 50kW or less. That’s just a little under 24kW/tonne.

Getting into it more strongly, 75kW showed on the dial – but I stress, this was now going harder than most people would drive most of the time. That’s a power/weight ratio of 35kW/tonne.

To get 100kW (or higher) showing on the gauge, you had to be clearly pushing the car hard.

And 200kW? Full throttle and with lots of revs – completely unlike 99 per cent of daily driving.

I already know from the Peugeot that, if the car is being driven well, 35 kW/tonne is enough for open road driving – and now I know that it’s also sufficient for even quite sporty urban driving.

19 Responses to 'More on How Much Power You Really Need'

Subscribe to comments with RSS

  1. Ford Man said,

    on January 29th, 2009 at 8:35 am


    To extend the logic

    35kW @ 1000rpm requires 334Nm or ~3.5lt+ NA engine / tonne.
    35kW @ 2000rpm requires 167Nm or ~ 1.8lt NA engine / tonne.
    35kW @ 3000rpm requires 111Nm or ~ 1.3lt NA engine / tonne.

    So based on this power requirement, a trade off between relaxed accessable performance and a more revy experience can be acessed.

    Forced induction engines would come out well from a similar analysis. Same for hybrid electric.

  2. doctorpat said,

    on January 29th, 2009 at 2:09 pm

    To be totally nit-picky, Australia has some 130 km/h roads.

    To be more practical, this is power at the wheels isn’t it? So about 50 kW/t at the engine?

  3. Julian Edgar said,

    on January 30th, 2009 at 7:02 am

    Ford Man, that’s a very interesting way of looking at it

  4. Richard said,

    on January 30th, 2009 at 8:40 am

    It would be interesting to use your idea and apply it to something like a Commodore V8 with AFM. Use the logic that the engine should be in cylinder deactivation mode for the majority of the time unless full throttle is needed. I wonder what fuel economy could be seen then?

  5. Ford Man said,

    on January 30th, 2009 at 9:21 am

    The fuel savings with AFM are small. Look at the official consumption figures for the latest Honda Accord.

    The only savings by ‘deactivating’ a cylinder are pumping losses and some valve train friction. The friction of the piston rings in the cylinder bore remains.

    Also the engine runs like a spark plug lead has been disconnected! Very lumpy. Clever engineering hides the worst of it, also AFM is only engaged at higher engine revs.

    In the Commodore this can mean forcing the auto to run in 5th rather than 6th at hwy speeds. The small fuel saving from reduced pumping losses is almost completely lost by having to spin the engine faster.

    For an enthusiast who is willing to accept the NVH trade off there may be some F/E benefit in forcing more AFM activation. Perhaps another project for Julian?

  6. Henri said,

    on January 30th, 2009 at 11:18 am

    Julian – Did some work on this problem recently for a solar vehicle to compete in a cross-country competition from Dallas to Denver in ’09. Normally these cars compete on a flat track, but X-country includes climbing. A normal calculation of a base aero power requirement is V,expected plus about 13mph. Then- here in the US, the specification max grade “normally” encountered is 6%; any such normal grade seldom has a “slow” shoulder lane for trucks and the like. Any grade exceeding 6% requires special permits for construction and has a “slow’ shoulder lane. What then becomes interesting is that to climb this 6% standard maximum grade and maintain your flat road speed, it takes about 2X your normal flat power requirment +/- a little, and this holds true for bicycles all the way to fully loaded trucks.

  7. Peter Tawadros said,

    on January 30th, 2009 at 1:02 pm

    I would have thought AFM would also hold the exhaust valve open through 720 degrees? It makes little sense de-activating a cylinder by shutting an individual injector down if you are going to sink power into it by having it compress useless air.

  8. Ford Man said,

    on January 30th, 2009 at 2:11 pm

    The compressed air expands again as the piston retracts. The result is less net power consumption than pumping gas back and forth across an open valve.

  9. Ford Man said,

    on January 30th, 2009 at 2:17 pm

    Henri – the load factor for climbing a hill is only dependent on grade and vehicle mass. A truck will have very different performance on a hill laiden vs unlaiden. Are you sure about the factor of 2?

  10. Ford Man said,

    on January 30th, 2009 at 2:19 pm

    Peter, there is also the logistical problem of valve to piston clearance, which counts against leaving the valve open.

  11. Ford Man said,

    on January 30th, 2009 at 2:31 pm

    Sorry for the number posts.

    But I’m not 100% sure about the energy consumption of spining an engine with open vs closed exhaust valves, ignore the previous post.

  12. redhed said,

    on January 30th, 2009 at 2:36 pm

    Is there a direct relationship between engine flywheel & crank weight to the all up weight of the vehicle its driving?
    Lightening the fly wheel will give quicker throttle response no load, but loaded it would depend on the inertia of the vehicle?? or is there something else?

  13. Ross said,

    on January 30th, 2009 at 6:56 pm

    Ford Man, I think you are right, because with all the valves closed, the air inside the cylinder acts like a spring. When you compress it, it absorbs some energy, but it give it back when it expands again. Ignoring friction (which is the same whether you have the valve open or shut) you basically have a net pumping loss approaching zero.

    With the valve open, you need to overcome the pumping loss resulting from forcing the air out through the open valve. Then you have another loss with the piston on the way down as it tries to drag air in.

    You also have the problem of clearance to the piston, plus the problem that you have now formed a connection between the exhaust and inlet manifolds.

    This would make the flow of gases back and forth difficult to predict, and potentially interfere with the cylinders that are still operating.

  14. Mick said,

    on January 31st, 2009 at 8:50 am

    A return to the power to weight ratios seen prior to about the mid ’80s would do none of us any real harm. They might have been a bit slow on the hills doing it, but we used to be content towing trailers and caravans with six cylinder Kingswoods and Falcons back then. By Julian’s example, red hot performance cars (with maybe 90 kw (+ or -) lugging perhaps 1600kg for around 56 kw/t ), but still nothing on the contemporary Falcamrydore with around 110 kw/t…and that’s just the shopping cart models.

  15. Ford Man said,

    on February 2nd, 2009 at 7:47 am

    Hi Ross,
    Yes it makes sense logically, and all the websites make the same point. I don’t have the numbers to quantify the power used one way or the other. But kicking over a high compression 4 stroke motorbike engine without opening the exhaust decompressor is bloody difficult. Suggesting that not all the energy used to compress the gas is recovered. Of course valve timing would play a part in that.

    Also using the electric starter to crank a car engine with no heads spins the engine very fast indeed. So the compression losses with the heads on must be very high (there isn’t much air pumping going on at cranking rpms).

    I guess the answer is to do the experiment. Coast down a hill on the dirt bike with the rocker arms removed (valve closed – air spring). Then coast down the same hill with just the exhaust valve open and compare max speed. Higher speed = lower frictional losses.

    Trouble is I’m not sure that I can be bothered doing the experiment.

  16. Brandon said,

    on February 2nd, 2009 at 8:54 am

    ross and ford man: one thing that must be accounted for is energy lost through heat, compressing air generates heat, generating heat takes alot of energy

  17. Ross said,

    on February 2nd, 2009 at 1:44 pm


    Don’t confuse force with energy. Yes it takes a lot of force to kick over your lawnmower without the decompresser. When the engine turns over TDC, you get the same force in the opposite direction as the pressure is released again, unless you have held it at TDC long enough for the pressure to dissipate through leakage. At 3000RPM, that won’t happen as much.

    Actually what happens is that compressing the air raises its temperature. It doesn’t generate heat. The heat is already there. Expanding the air reduces its temperature again (that’s why a gas bottle gets cold). The amount of energy it contains does not change, apart from a small amount lost to the cylinder walls, friction and leakage.

    There is an equation for this:


    P1 & P2 are pressures, T1 & T2 are temperatures and V1 & V2 are volumes

    If you rotate the crankshaft and open and close the valves, without introducing fuel or spark, what you then have is a compressor, and that will get hot because the expansion is taking place elsewhere.

  18. TST said,

    on February 2nd, 2009 at 8:27 pm

    For more on AFM look here:

    Also Ford Man, I’m not sure your comments about how the AFM commodore drives are correct. I drove an AFM one and an older model back to back and couldn’t tell the difference. The auto works the same. In stop start traffic the fuel economy was a little better but on an 80km/h steady cruise the AFM car returned 7.6 L/100km vs 9.8 for the old car.
    Not outstanding economy next to a prius but somehow more satisfying when overtaking!

  19. Ford Man said,

    on February 5th, 2009 at 2:30 pm

    Hi Ross,
    Force x distance = work (energy)
    So more force over the same distance equals more energy.

    Hi TST,
    Sounds like you found a sweet spot in the operating range – did you get a chance to compare fill-up to fill-up range between the cars over a similar drive cycle?

    Interesting review: