How many watts you need on the Alto de Patios

The cycling power equation is solved on every section of the profile: power is split between overcoming gravity, overcoming rolling resistance and overcoming the air.

Frontal area is estimated from your height and weight with the regression Bassett and colleagues (1999) built on the Du Bois body surface area, the usual procedure when there is no wind tunnel measurement. For a 176 cm, 72 kg rider it gives a CdA of about 0.33 m², the order of magnitude of riding on the hoods; the more upright position adds 8 %. Air density is0.89 kg/m³: what the pressure at the climb's mean altitude — 2,836 m — gives with the 9 °C of a morning on the plateau, not the sea level value. Using 1.225 here overestimates drag and with it the power required.

The profile comes from a GPX of the segment itself: 1,645 points with their elevation, summarised as the mean gradient of each 200 m block. Integrated in 20 m sections it returns the real 367 m of climbing, and that is how the time is calculated, not by applying an average gradient: between the 2.7 % of the relief and the 7.7 % of the hard part, the average lies.

The bike enters the equation, not as decoration

Three things about the bike change the result and here they are chosen one by one. The weight, which on a 6 % ramp weighs most. Thecomponent level, which sets the drivetrain's mechanical loss: 1.2 % with high end parts and up to 2.5 % with a dirty or worn chain. And the tyre, with rolling resistance from 0.0035 on a 25-28 mm racing tubeless to 0.0070 on a mixed gravel tyre. Picking your level is just a shortcut that fills all three with typical values; then you correct them.

On top of that comes the road surface, which multiplies rolling resistance and which almost no calculator asks about. It is not a detail: at 250 W, going from smooth asphalt to the patched surface this climb actually has costs around 25 seconds. That is why it defaults to “patched”, which is how the road is, and not to “smooth”, which is how the number looks prettier.

That mechanical loss was not there before, and it was why the model returned optimistic times. The power you type is pedal power — what your power meter reads — not what reaches the wheel.

What it is checked against

Against a Best Bike Split plan for this same segment with Egan Bernal's numbers: 357 W average, 62 kg, a 7.5 kg bike, high end components, 25-28 mm tubeless. Their sheet calculates a mechanical loss of 0.012, a rolling resistance of 0.00351 and a climbing CdA of 0.3442, and returns14:40.

With those same numbers and the patched surface the climb actually has, this calculator gives 14:42: two seconds on the slow side. Set to smooth asphalt, which is what their tyre coefficient assumes, it gives 14:31. The plan sits between the two, which is exactly where it should be.

The drag coefficient used here — 1.09 over the estimated frontal area — comes from matching that 0.3442 CdA; at 1.0, which is what was here before, drag came out 9 % low. And feeding their three exact constants into our equation returns 14:32, so the physics is the same: what changes is the inputs.

Why it does not ask for your sex

Because nothing demonstrated would change in this calculation. What decides the time is mass, power, frontal area and rolling resistance, and the area is estimated from your height and weight, which already carry the size difference. The frontal area regression was built on male cyclists, so for women it may overestimate drag slightly; inventing a coefficient by sex would not fix that. Where sex does change the calculation — body composition — it is asked for.

Where it stops working

• The calculated time is the ideal one. It assumes power nailed flat, still air and a clear road, and those three do not happen at once. That is why an expected range appears above: power that swings by 10 % costs half a point, a light 10 km/h wind costs another point and a half even blowing half in your favour, and on top goes whatever traffic you pick. The splits table shows the other side: what you gain by placing the watts well.
• Traffic is declared, not calculated. With a clear road the range sits almost on the ideal time; with buses stopping and slow trucks it runs from 3 to 7 %, which on this climb is one to two minutes. It is a named assumption, not a measurement: nobody has timed how many seconds each overtake costs on the La Calera road. Nor does it count the opposite, which also happens: sitting behind a vehicle for a while lowers drag, but that is not something you plan.
• There is no wind and no weather. The model assumes still air and a typical morning on the plateau. Putting real wind in means knowing where it blows from and where the road points at each moment: Patios starts north-east, straightens to the north around kilometre 1.5 and ends south-east, so the same wind is a tailwind low down and a headwind higher up. That, with the temperature and humidity of the day, is what gets modelled in the Race Plan, not in a free calculator.
• The power here is pedal power. It is what your power meter reads, whether it sits at the crank, the pedal or the hub — the last one already discounts part of the loss. What the chain keeps comes from the component level you pick.
• Performance at altitude is individual. Two cyclists with the same threshold at sea level do not lose the same at 2,600 m. That is measured, not estimated.