What power to ride on the bike in an Ironman 70.3, and how much time each watt is worth
70.3 bike
2.4 min
What 10 W more saves over the 90 km bike leg of a flat 70.3, for a triathlete riding at 220 W.
Model calculation: 80 kg with the bike, CdA 0.28 m², no wind. Lowering CdA by 0.01 m² saves 1.7 min at the same power.
The right power for a 70.3 bike leg is the highest your athlete can hold steadily for the whole segment while still running a good half marathon. In practice it sits clearly below FTP, because in studies FTP is held for about 51 minutes on average1, and the median drops to 35 minutes in recreational cyclists2. A 90 km bike leg lasts more than twice as long.
Why does a fixed percentage of FTP not work?
Training guides often give a percentage range of FTP for the 70.3, but we have not found any peer-reviewed study that validates it for that distance. What has been measured is that FTP describes an intensity whose sustainable duration varies a lot between people: 50.9 ± 15.7 minutes in 23 trained cyclists1, and medians ranging from 35 minutes in recreational riders to 51 in professionals2. Two triathletes with the same FTP can have very different ceilings for two and a half hours.
On top of that, thresholds measured fresh shift during long efforts. This has been called durability: when the physiological profile starts to deteriorate over hours of exercise, and by how much5. A 70.3 starts after a 1.9 km swim and ends with a 21 km run, so the same percentage of FTP means different things for a very durable athlete and for one who fades after the second hour.
So, as a coach, the useful reference is not a percentage but the power your athlete has held in long rides followed by a run, with race nutrition and race position, and with a run afterwards that did not fall apart.
How much time does each watt buy over 90 km?
To put numbers on the decision, Table 1 shows bike split times on a flat 90 km course with no wind, for two positions and five power levels. These are model figures, an example with reasonable inputs, not a prediction for a specific course.
| Average power | CdA 0.25 m² | CdA 0.30 m² |
|---|---|---|
| 180 W | 2:32 (35.6 km/h) | 2:41 (33.6 km/h) |
| 200 W | 2:26 (37.0 km/h) | 2:35 (35.0 km/h) |
| 220 W | 2:21 (38.3 km/h) | 2:29 (36.2 km/h) |
| 240 W | 2:16 (39.6 km/h) | 2:24 (37.4 km/h) |
| 260 W | 2:12 (40.8 km/h) | 2:20 (38.5 km/h) |
Table 1. Bike split (h:min) over 90 flat km with no wind. Model example: 80 kg for triathlete and bike, Crr 0.004, air density 1.20 kg/m³.
At 220 W and a CdA of 0.28 m², 10 W more saves about 2.4 minutes, and lowering CdA by 0.01 m² saves 1.7 minutes at the same power. The difference matters: extra watts are paid for later on the run, while the seconds from a better position are free if the athlete can hold it for two and a half hours. Before raising target power, it is worth reviewing the position (see what CdA is good for triathlon).
Constant or variable power?
On a flat course with no wind, near-constant power is the most efficient. In the example above, keeping the same 220 W average but alternating equal stretches of 250 and 190 W costs 21 seconds; with 270 and 170 W, a full minute. And every spike above sustainable power spends reserve that will be missing later. In a study of eight trained triathletes, only 20.9% of half-Ironman bike time was spent more than 10% above average power, compared with 43.8% at sprint distance3: over the longer distance, the self-chosen distribution is already more restrained.
Wind and hills change this: it pays to raise power slightly where speed is low and lower it where speed is high, because that reduces the variation in speed4. In an out-and-back example of 45 km each way with a 15 km/h wind and the same 220 W average, riding 231 W into the headwind and 203 W with the tailwind saves 42 seconds; 242 and 187 W saves 74 seconds. Beyond about 10% variation the saving grows little, while the risk of overcooking the headwind section grows a lot.
How do you set target power in practice?
- Start from what is proven: average power of the two or three long simulations in the specific block, on the race bike and in race position, followed by a run.
- Watch the run that follows: if run pace drops a lot after those sessions, the bike power was too high for that athlete, even if heart rate looked under control.
- Adjust for the day: heat, altitude and wind change the speed each watt buys. What does not change is the physiological cost: you do not raise power because the day is fast.
- Plan by segments: with the course and the weather at time of passage, the power for each segment adapts to climbs and wind. That is what SuperRecord does, keeping the CdA of the triathlon position separately; you can see it in the demo.
The same logic of pushing on climbs and into the wind is developed for time trials in how much to vary power in a hilly time trial. If you also coach the full distance, the difference in approach is covered in what power to ride in an Ironman.
What is known and what is not
It is well documented that FTP cannot be held for two hours1,2, that near-constant power is efficient in stable conditions and that varying it with wind and gradient saves time4. Also that the physiological profile deteriorates over hours, differently in each person5. There is, however, no validated percentage of FTP for a 70.3 bike leg, and no solid data on how much the half marathon suffers per extra watt. That is learned with each athlete, session by session.
Frequently asked questions
What percentage of FTP should you ride in a 70.3?
There is no validated percentage. It is clearly below FTP, because the bike leg lasts over two hours and FTP is held for less than one on average. The good reference is the power the athlete has sustained in long simulations followed by a run.
How much time do 10 more watts save on a 70.3 bike leg?
Over 90 flat km at about 220 W, roughly 2.4 minutes according to the model. A hundredth less CdA saves 1.7 minutes without spending more energy.
Should you ride at constant power?
On flat, windless courses, almost constant. With wind or hills it pays to raise power 5 to 10% where you are slower and lower it where you are faster.
Is heart rate useful for pacing the 70.3 bike?
As a check, yes; as a target, worse than power, because it rises with heat and hours even when power does not change. The useful thing is to watch both together.
References
- Borszcz FK, Tramontin AF, Bossi AH, Carminatti LJ, Costa VP. Functional threshold power in cyclists: validity of the concept and physiological responses. Int J Sports Med. 2018;39(10):737-42. doi:10.1055/s-0044-101546
- Sitko S, Cirer-Sastre R, López-Laval I. Time to exhaustion at estimated functional threshold power in road cyclists of different performance levels. J Sci Med Sport. 2022;25(9):783-6. doi:10.1016/j.jsams.2022.06.007
- Wu SSX, Peiffer JJ, Brisswalter J, Nosaka K, Lau WY, Abbiss CR. Pacing strategies during the swim, cycle and run disciplines of sprint, Olympic and half-Ironman triathlons. Eur J Appl Physiol. 2015;115(5):1147-54. doi:10.1007/s00421-014-3096-2
- Atkinson G, Peacock O, Passfield L. Variable versus constant power strategies during cycling time-trials: prediction of time savings using an up-to-date mathematical model. J Sports Sci. 2007;25(9):1001-9. doi:10.1080/02640410600944709
- Maunder E, Seiler S, Mildenhall MJ, Kilding AE, Plews DJ. The importance of "durability" in the physiological profiling of endurance athletes. Sports Med. 2021;51(8):1619-28. doi:10.1007/s40279-021-01459-0