Calculating power output

cframe

Is there any way of calculating (to a certain degree of accuracy) your power output without a power meter?

If you were to isolate as many variables as possible, ie do a set distance or time at a set resistance on a trainer could you work it out based on heart rates and/or speed?

I'm guessing probably not but figured there's no harm in asking the educated masses

DaveyL

Best way is to eliminate variables such as drag and wind resistance by climbing a relatively steady and steep hill which is sheltered from the wind. Then use http://www.analyticcycling.com/ to work out your power output, knowing your speed, weight and gradient of the climb.

Not sure how you'd do it on a trainer with no direct measuring of power.

Alex_Simmons/RST

cframe wrote:

Is there any way of calculating (to a certain degree of accuracy) your power output without a power meter?

If you were to isolate as many variables as possible, ie do a set distance or time at a set resistance on a trainer could you work it out based on heart rates and/or speed?

I'm guessing probably not but figured there's no harm in asking the educated masses

Only if you know the nature of the resistance forces acting against you.

Certainly not wrt heart rates.

cframe

@DaveyL: Cheers, I'll probably give that site a go sometime.

@Alex_Zimmons: Yeah, I kinda thought the heart rate thing was a long shot, ta

GeorgeShaw

I read somewhere that you need to find a hill and then use the formula:

Power = ((total_weight_in_kg x 9.81 x height_gain_in_metres) / time_in_secs) x 1.1

The 9.81 and 1.1 mean something that I can't remember.

Alex_Simmons/RST

GeorgeShaw wrote:

I read somewhere that you need to find a hill and then use the formula:

Power = ((total_weight_in_kg x 9.81 x height_gain_in_metres) / time_in_secs) x 1.1

The 9.81 and 1.1 mean something that I can't remember.

the 9.81 would be gravitational acceleration 9.81 m.s^2

The 1.1 is probably a rough factor to allow for other resistance forces other than overcoming gravity, which would be an approximation applicable only for reasonably steep gradients.

Jeff Jones

Alex_Simmons/RST wrote:

GeorgeShaw wrote:

I read somewhere that you need to find a hill and then use the formula:

Power = ((total_weight_in_kg x 9.81 x height_gain_in_metres) / time_in_secs) x 1.1

The 9.81 and 1.1 mean something that I can't remember.

the 9.81 would be gravitational acceleration 9.81 m.s^2

The 1.1 is probably a rough factor to allow for other resistance forces other than overcoming gravity, which would be an approximation applicable only for reasonably steep gradients.

I've just plugged in numbers for a 2.75km/140m climb and it seems to underestimate actual power output by a lot, I guess due to aero effects. I got 305W from the formula but I've measured it at ~375W.

Related to this is the reason why the Dr Ferrari-espoused VAM method is pretty useless, unless all the climbs you're comparing are the same gradient. As soon as it gets steeper, VAM goes up.

Simon Notley

As said above, I think the only time you could reliably estimate power using this method is on a pretty steep long climb... something like Alpe d'Huez... anyone got some power meter data from AdH and want to try it?

Essentially your power output is split four ways to overcome resitance from:
1) Gravity (as discussed above, quite easily measureable if you know your climbing rate)
2) Air resistance (not easy to measure at all)
3) Rolling resistance and mechanical friction (not so significant)
4) Inertia due to the mass of bike and rider (onl relevant when accelerating)

So to measure power without a power meter, you effectively need to make measurements of all the above elements, calculate the power required to overcome them and then add it up... not really easy without a plethora of instruments all over your bike. It's easier just to measure the power going through the pedals or cranks i.e. a power meter!

GeorgeShaw

Jeff Jones wrote:

I've just plugged in numbers for a 2.75km/140m climb and it seems to underestimate actual power output by a lot, I guess due to aero effects. I got 305W from the formula but I've measured it at ~375W.

That's cheered me up. You've just raised my power by 20%.

Alex_Simmons/RST

Jeff Jones wrote:

Related to this is the reason why the Dr Ferrari-espoused VAM method is pretty useless, unless all the climbs you're comparing are the same gradient. As soon as it gets steeper, VAM goes up.

Yep, which is why I said steep gradients. 5% a good solid climb but isn't steep, well not steep enough for VAM to make much sense.

Jeff Jones

GeorgeShaw wrote:

That's cheered me up. You've just raised my power by 20%.

Hugh A

Another way would be to use a consistent turbo trainer which has been calibrated using a power meter. If you use the same bike, wheels, tyres etc. then it should be reasonably accurate - with a non-variable type it should be possible to equate power directly to the speed indicated by a bike computer reading on the back wheel.

There have been some formulas published showing how to calculate this relationship but it varies between models of trainer (and of course wheels, tyre pressure, rider weight or tightness of clamp etc.)

Talking of which - anyone have any data for an Elite Fluid Crono?

Alex_Simmons/RST

Hugh A wrote:

Another way would be to use a consistent turbo trainer which has been calibrated using a power meter. If you use the same bike, wheels, tyres etc. then it should be reasonably accurate - with a non-variable type it should be possible to equate power directly to the speed indicated by a bike computer reading on the back wheel.

There have been some formulas published showing how to calculate this relationship but it varies between models of trainer (and of course wheels, tyre pressure, rider weight or tightness of clamp etc.)

Talking of which - anyone have any data for an Elite Fluid Crono?

Excepting that the resistance of some turbo units tends to drift with use as they warm up, and how hard you are going/how much heat build up there is. Some are better than others.

Anonymous

Or you could buy a power meter

(Actually the Tacx Flow trainer measures power, don't know how accurate it is but it's probably the cheapest way of doing it.)

bahzob

Simon Notley wrote:

As said above, I think the only time you could reliably estimate power using this method is on a pretty steep long climb... something like Alpe d'Huez... anyone got some power meter data from AdH and want to try it?

Here you go: 4 rides, same rider, on 4 different dates this year, guessing weight around constant, in varying levels of fatigue (distance differs because routes a bit different)

Mins Distance Power Real(Powertap)
Marmotte - 70 - 13 - 229 - 222
Rousse - 58 - 13.8 - 277 - 278
Grimpe - 63 - 13.8 - 254 - 261
Mars Tour - 60 - 13.5 - 268 - 282

So formula looks to get info in right ballpark, though not sure of what this tells you unless you already have a powermeter (in which case it can be used to plan pacing for unknown climbs)

Simon Notley

Cheers bahzbob, very interesting. I have to say that that is surprisingly accurate, I think the powertap is only accurate to with a few watts, so for average wattage over a long climb, the formula seems really quite good. Would be interesting to stick the power meter data in Excel and calculate power output for every ten seconds or so of the climb and compare to the powertap.