The Critical Power concept is not new. In fact, it’s been around for nearly 50 years. It was first published by two scientists, Monod and Scherrer, in 1965 in an article entitled “The Work Capacity Of A Synergistic Muscle Group”.
Monod and Scherrer conducted experiments to measure muscular work in a series of exercise tests. The data was used to create a mathematical model which predicted the maximum amount of work a subject could perform during a given time period. The work was measured in ‘joules’. Given the relative ease with which work can be measured on a bike, using a power meter, the Critical Power concept has gained traction in the sport as a coaching, performance modeling and analysis tool.
Coaches and athletes may use the Critical Power model for a number of purposes:
1Prescribe an appropriate power output for a new interval session
The Critical Power model can be used to predict maximum average power for almost any time duration, up to 60 minutes. These predictions can be used to suggest appropriate average power outputs for new interval training sessions. For example, if you’ve never completed a 4 minute maximal interval before, the Critical Power model will help you to choose a target power output for this time duration.
Try inputting your test results on the Cycling Power Lab Critical Power calculator.
2Use shorter tests to predict the maximum sustainable power output for longer efforts
Even if a cyclist has never completed a particular effort, the Critical Power model can predict the maximum sustainable power output for this period with a fairly high degree of accuracy. For example, an athlete could predict their maximum average power for a 35 minute effort, to pace themselves more effectively during a climb of this duration.
Analysing my TrainingPeaks data for the last year reveals a maximum 60 minute power of 258 watts. However, this result is not a true maximum as I’ve never ridden intensely over this period during the last 12 months. The Critical Power model, based on my test results for shorter time periods, suggests that I should be able to maintain 313 watts for 60 minutes. If I was aiming to tackle a 25 mile TT, for example, I would target an average power in this region based on the prediction.
3Analyse changes in a rider’s fitness over time
By conducting regular testing at key time durations and applying the Critical Power model to the results, an athlete or coach can record changes and gain an insight into how improvements or deteriorations effect the rider’s capacity to produce power over the entire curve. For example, following a focussed block of low-intensity endurance training, what is the impact on their capacity to produce power over 2-5 minute durations?
4Compare rider’s performances
Contrasting two rider’s Critical Power curves can help to uncover their relative strengths and weaknesses. For example, the Critical Power curve for a powerful sprinter will likely exhibit a disproportionately high area below the curve at shorter duration, illustrating that they have a can carry out very high amounts of mechanical work over these times, relative to a more endurance focussed athlete, who will likely exhibit a greater area, and therefore capacity, over longer durations.
If you’re wondering how accurate the Critical Power model can be, have a look at this study from 2011. The researchers used Critical Power and Aerodynamic Drag to predict Time Trial performance in British Championship events. The difference between actual and predicted times averaged 0.2%, and was no greater than 1.1%!
5Identify strengths, weaknesses and areas for improvement
A coach can identify strengths and weaknesses in the rider’s power profile, where predicted power outputs for a given duration are significantly higher than the rider has achieved. This may be illustrated by a steep drop off in the curve based on actual values, recorded on a platform such as TrainingPeaks, using their Peak Power chart, relative to predicted values for Critical Power, which may indicate a disproportionately low value for a given duration. This may mean that the athlete has simply not recorded a representative average power for this time period, or it could represent a particular area of weakness.
6Predict the effect of changes in performance at one time duration, relative to another
If you achieve a new p.b. average power for 5 minutes, what effect would this have on your maximal sustainable power for 20 minutes? A significant improvement in performance over 5 minutes likely represents a physiological adaptation that would effect other time durations; Our energy systems are interconnected: the Critical Power model helps us model the wider impact created by changes in performance over specific time durations. You can find a discussion about how sprint training may impact endurance performance and how that may help ‘time poor’ cyclists, here.
I’ve been working to develop my Maximum Aerobic Power (MAP) so I wanted to predict what effect an improvement in my 5 minute power output would have on other time duration, specifically my 20 minute maximum sustainable power. According to the Critical Power model, increasing my 5’ max from 358 to 380 watts changes the 20 minute predicted average from 322 to 343 watts.
7Prescribe power training zones
Once a rider has recorded their best efforts at shorter durations, often 5 seconds, 1 minute, 5 minutes and 20 minutes, the Critical Power model can be used calculate power based training zones, based on percentages of predicted 60 minute power, for a wide range of time durations.
Last year, I carried out a range of Critical Power tests on myself during a weekend of riding in the Alps. You can see a video of the trip, here. Again, if you’d like to input your own test results into a Critical Power model, I highly recommend the CyclingPowerLab website’s online calculator.