[In-depth Analysis] How Triathlon Athletes Use Critical Power to Break Through Bottlenecks? Decoding the Core Scientific Mechanisms of Anaerobic Work Capacity (W') and Physiological Limits

In recent years, the triathlon coaching community has increasingly used “Critical Power” (CP) and “Anaerobic Work Capacity” (W’) to replace a single FTP or simple heart rate zones. The reason is straightforward: triathlon is not a single steady-state endurance event. It consists of three consecutive disciplines—swimming, cycling, and running—each with continuous loads, plus the interactive effects of track turns, gradients, chasing the pack, nutrition, and transitions. Relying on a single “power you can sustain for about an hour” cannot accurately describe when an athlete enters an irreversible fatigue zone.

The value of the CP/W’ model lies in the fact that it does not just tell you “where the threshold is approximately,” but instead breaks down high-intensity endurance performance into two key questions: First, at what stable aerobic rate can you sustain output without causing physiological homeostasis to collapse? Second, when you exceed this limit to chase the pack, climb hills, exit turns, counter-attack, or transition to running, how much high-intensity reserve do you still have available? For triathletes, this is closer to the truth of the race than simply looking at average watts.

More importantly, the cycling and running segments of a triathlon cannot substitute for each other. Recent studies have shown that the running CP of tested triathletes is significantly higher than their cycling CP, and cycling power explains only about one-quarter of the variance in running power. If you use cycling CP to prescribe running workouts, or use running critical speed to back-calculate cycling power, you will likely set your training zones incorrectly. True mature application involves understanding the model’s limitations and integrating CP, W’, running CS/D’, and transition fatigue into a multi-discipline load management system.

I. What Critical Power Actually Represents: It’s Not a Mysterious Number, but a Fatigue Boundary

From a sports physiology perspective, the most important meaning of CP is not “how long you can ride,” but that it approximates the boundary between heavy and very heavy intensity. When exercise intensity is below CP, oxygen uptake, lactate, phosphocreatine consumption, and intramuscular acid-base balance theoretically have a chance to reach a relatively stable state; once intensity exceeds CP, these physiological responses can no longer stabilize, oxygen uptake continues to climb toward VO2max, metabolic byproducts accumulate rapidly, until you are forced to slow down or stop.

Poole, Burnley, Jones, and others have organized CP as a “fatigue threshold” with genuine mechanistic significance. This differs from the traditional practice of using FTP as the core of training. FTP is practical, but it is often estimated by multiplying 20-minute average power by 95% or using plateau algorithms, which may not correspond to a clear physiological boundary; CP is based on a two-parameter model of power and sustainable time, allowing “steady-state capacity” and “limited reserve beyond steady state” to be viewed separately.

In the two-parameter model, the following formulas can be used to describe:

Tlim = W' / (P - CP)
W = CP × Tlim + W'
P = W' × (1 / Tlim) + CP

Where:

• CP is Critical Power, representing a steady-state boundary that can be approached for a long time but cannot be infinitely exceeded.
• W' is the limited work capacity above CP, with units typically in joules or kilojoules.
• Tlim is the time to exhaustion at a fixed power P.

This set of formulas is very practical because it clarifies a common phenomenon in races: two athletes with similar FTPs may perform very differently on short steep climbs, during chases, or with repeated accelerations, because one has a larger W' or recovers W' faster. In other words, in triathlon, you need not just an “endurance engine,” but also to know the size of your “high-intensity battery” and how fast it charges.

II. What is W’: It’s Not a Lactate Tank, but an Integrated Representation of High-Intensity Work Capacity

Many people simplify W’ to an “anaerobic energy tank” or “lactate bucket,” which is actually not precise enough. W’ is more like the limited work capacity you can utilize above CP; it integrates phosphocreatine reserves, anaerobic glycolytic capacity, hydrogen ion buffering, tolerance to metabolic byproducts, neuromuscular recruitment, and local metabolic stability. It is not a single substance or a single organ, but an observable overall performance parameter.

Studies show that W’ exhaustion is related to fatigue-related phenomena such as phosphocreatine decline, inorganic phosphate and hydrogen ion accumulation; W’ recovery is closely related to oxidative metabolism, phosphocreatine resynthesis, local blood flow recovery, and acid-base regulation, but recovery speed is not a fixed constant. This is why, when doing the same 30/30, 40/20, or hill repeat sprints, the difference between “blowing up at the start of the second set” and “still controlling quality in the sixth set” is very large among different athletes.

The practical significance of W’ is especially suitable for triathlon. Because short-distance and Olympic-distance events often contain a large amount of non-linear output, such as:

1. Immediately increasing speed after getting on the bike from the water.
2. Repeated acceleration and deceleration before and after technical turns.
3. Chasing the front group or taking a position.
4. Short-duration high power on hills, headwinds, or turnaround points.
5. Accelerating in the first few minutes after dismounting to maintain position before transitioning to running.

These actions almost all consume W’. If you do not understand your W’ capacity and recovery characteristics, you may think during a race that it is just a “short burst,” but in reality, after 10 to 15 unnecessary accelerations, you have completely overdrafted the high-quality rhythm needed for the transition run.

III. The Most Common Misjudgment in Triathlon: Treating Cycling CP, Running CP, and FTP as the Same Thing

For triathletes, one of the most dangerous concepts is treating all “critical” values as interchangeable numbers. In practice, this easily leads to three types of errors.

The first error is treating FTP directly as CP. Although the two are highly correlated, they are not identical. A study in Frontiers found on trained cyclists and triathletes that CP averages slightly higher than FTP, with an average deviation of about 7 W, and individual differences can be large enough to affect training and race interpretation. For high-level athletes, a 7 W difference is not trivial at 40 km or 90 km pace; if combined with environmental factors, fatigue, and nutrition errors, the practical gap is amplified.

The second error is using cycling CP to back-calculate running power zones. A 2023 study on trained triathletes indicated that running CP is about 20.2% higher than cycling CP, and cycling power explains only 26.7% of the variance in running power. This means triathletes must test cycling and running separately, rather than transplanting numbers from one to the other. The reasons for the difference are reasonable: pedaling is a closed-chain movement, while running is a semi-open-chain movement; sensors and algorithms differ; neuromuscular recruitment patterns are also different.

The third error is focusing only on CP and ignoring W’. Two athletes with both CP at 280 W, where one has W’ of 12 kJ and the other 18 kJ, will typically have higher error tolerance in group accelerations, bridging, and short hill handling for the latter on the same technical course. If a coach looks only at CP, they will mistakenly view both athletes as the same type, leading to oversimplified workout design and race tactics.

Therefore, the correct approach for triathlon is:

IV. How to Measure CP and W’: Not Just for a Pretty Number, but to Establish a Repeatable Estimation Process

Common CP measurement methods currently fall into three categories. The first involves multiple limit tests with fixed durations or fixed power, such as 3, 7, or 12-minute tests, or multiple all-out efforts within 2 to 15 minutes, followed by regression of power and time relationships to estimate CP and W’. The second is the 3-minute all-out test. The third is the single-session ramp all-out test. Literature reviews indicate that both the 3-minute all-out and ramp all-out tests can yield effective CP/W’ estimates when executed properly, but they are highly sensitive to the quality of test execution.

If you are a typical triathlete, I recommend establishing CP using “multi-duration time trials” rather than over-relying on a single all-out test. There are three reasons:

1. For endurance athletes with mature pacing skills, multiple time trials better reflect race effort patterns.
2. The 3-minute all-out test is highly dependent on familiarity; slight deviations in cadence or resistance settings can skew the CP/W’ estimates.
3. Triathletes inherently need high-quality self-pacing ability, and a byproduct of multiple TTs is pacing practice.

A practical and operable cycling test method is as follows:

Before each test, perform a full warm-up and allow at least 24 to 48 hours of recovery. Subsequently, use a linear work-time or power-inverse-time model for regression. Assuming an athlete finally estimates:

• CP = 275 W
• W' = 16 kJ

This means that when riding at 315 W, the portion exceeding CP is 40 W, and the theoretical sustainable duration is approximately:

Tlim = 16000 / 40 = 400 seconds ≈ 6 minutes 40 seconds

This is not a guaranteed value, but a theoretical approximation. If the athlete is already fatigued beforehand, the environment is hot, carbohydrate stores are insufficient, or pedaling efficiency is poor, the actual sustainable time is usually shorter. Conversely, if he only exceeds the limit for a brief 20 to 40 seconds, the focus is not on whether he will blow up immediately, but on whether W’ can recover sufficiently during subsequent segments below CP to ensure he can handle future accelerations again.

For the running portion, it is recommended to measure “critical speed” (CS) or running CP separately. If using a running power device, the protocol must be fixed under the same device and environmental conditions; one cannot compare numbers obtained today on an indoor treadmill, tomorrow in the wind outdoors, and the next day wearing carbon-plate shoes. For triathletes, test consistency is more important than flashy single-session maximums.

V. Practical Value of W’ Balance: You Are Not Just Consuming, But Also Recovering While Riding

Real races rarely involve sustained constant power; therefore, looking solely at “average power” is insufficient to describe fatigue status. The core concept of the Skiba series of models is to treat every segment of output above CP during a race as W’ consumption, while every segment below CP allows W’ to resynthesize at a certain time constant, known as W' balance.

The concept can be simplified as follows:

When P > CP: W' continuously decreases
When P < CP: W' gradually recovers, but the recovery rate is affected by recovery intensity, individual ability, and prior fatigue

Literature reviews indicate that higher CP is often associated with faster W’ recovery rates. Additionally, phosphocreatine recovery and W’ resynthesis are correlated, but the former is usually faster than the latter, and recovery slows down after repeated exhaustion. This is highly enlightening for triathlon races: you cannot treat every 20-second sprint as an independent event, because the recovery quality of the eighth sprint today will not be the same as the first.

Taking the Olympic-distance cycling stage as an example, if the course has 2 dangerous corners, 1 short hill, and 1 bridge per lap, you may need to use several P > CP accelerations per lap. JSSM 2024 analysis of world-class short-distance triathlons also indicates that the power profile of international cycling stages is highly nonlinear; the more technical corners, the more time spent above maximum aerobic power, and athletes who can continuously handle these intermittent high-intensity demands often perform better in the cycling stage. This is precisely where the W’ model comes into play.

However, extreme caution is needed here. W’ balance is not a crystal ball; it is affected by test errors, device errors, environment, muscle damage, nutrition, and model selection. The best way to use it is not as “absolute truth,” but as:

1. A post-race power zone review tool.
2. An individualized design tool for specific interval sessions.
3. A trend-tracking tool for whether an athlete can recover increasingly well under repeated high-intensity efforts.

VI. How to Convert CP/W’ into Triathlon Training Plans: Corresponding from Cycling to Running to Run-to-Bike Transitions

The true high value lies in the fact that CP/W’ is not just for “testing once to be cool,” but must be converted into actionable training plans. The following provides a structure practical for triathletes.

1. Cycling: Establish Stability Near CP

The goal is to enable the athlete to maintain stable output near the CP range without prematurely entering significant W’ consumption.

Example Plan:

2. Cycling: Train W’ Usage and Recovery

These plans are not just about sprinting, but about training “to survive after the sprint.”

Example Plan:

3. Running: Set Plans Based on CS or Running CP, Not Copying Cycling Numbers

Running plans can be arranged using the CS/CP framework of critical speed or running power:

4. Conversion Session: The One Ironman Athlete Cannot Skip

Solo bike CP and solo run CS alone are insufficient because the real race happens during the run after the bike. It is recommended to add a brick session every 7 to 10 days:

Bike 60 minutes:
20 minutes at 88% CP
3 x (5 minutes at 100% CP + 2 minutes at 110% CP + 3 minutes at 70% CP)

Run 20 minutes:
First 5 minutes at 92% CS
Middle 10 minutes at 96% to 100% CS
Last 5 minutes depending on condition, up to 102% CS

The focus of this session is not to blow yourself up, but to observe:

1. Whether running form breaks down after each CP exceedance on the bike when transitioning to run.
2. Whether the first 5 minutes of running are hindered from entering a steady rhythm due to excessive W’ depletion from the bike.
3. Whether nutrition and cadence strategies affect conversion run performance.

VII. How to Use Race Pace: CP/W’ Configuration Philosophy Varies by Distance

The application of CP/W’ cannot be one-size-fits-all for different distances. If the distance changes, the philosophy of W’ usage must change.

Short Distance / Olympic Distance

In these events, the bike leg often features high variability output, especially on draft-legal courses with technical corners and frequent attacks. Research has indicated that international race bike legs exhibit significant non-linear power profiles; therefore, the athlete’s goal is not to rigidly maintain a pretty average wattage, but to use limited W’ at key moments:

1. The transition from water exit to mounting the bike.
2. Key climbs or re-acceleration after corners within the main group.
3. Avoiding dropping out of the main group.

However, you must avoid repeatedly making heroic efforts on sections with no tactical significance, because this expenditure is often repaid at a much higher cost during the first 2 kilometers of the run.

70.3 / Ironman

The logic for long distance is the opposite. At this point, CP acts more like an upper limit warning line rather than a game of frequently using W’. You should keep most of your bike time well below CP, using W’ sparingly only for fill-in, short climbs, wind shifts, or safety needs. If you heavily exceed CP in the early stage of a 70.3, it usually means:

1. Early decline in muscle glycogen.
2. Increased core temperature and ventilation cost.
3. Difficulty initiating run CS/pacing after dismounting.

Therefore, the correct mindset for long-distance Ironman is to “preserve economy below CP,” rather than showing off maximum power on the bike. No matter how strong you are on the bike, if you drop out of the steady state zone below CP in the first half of the conversion run, your overall time will usually be dragged down.

VIII. Data Interpretation and Common Pitfalls: High-Depth Models Are Most Feared When Used Roughly

CP/W’ is powerful, but also easily misused. Below are the most common pitfalls for triathlon athletes.

Pitfall One: Inconsistent Testing Timepoints

Testing once before the season and then using the same set of CP/W’ for all training plans three months later often leads to errors. Especially when athletes go through high volume phases, heat acclimatization, tapering, or return from injury, both CP and W’ may change.

Pitfall Two: Ignoring Nutrition and Heat Stress

Literature reviews indicate that if sufficient carbohydrates are replenished during long-duration high-intensity cycling, CP maintenance time is better; however, W’ is not necessarily preserved in the same way. This means that improved nutrition cannot fully offset high-intensity reserve depletion. Heat environment, dehydration, low carbohydrate intake, and insufficient sleep can all make you think “CP dropped today,” when in fact the day’s conditions suppressed available performance.

Pitfall Three: Treating the Model as a Second-by-Second Prophecy

W’ balance can help you understand trends, but it is not sufficient to precisely predict “when you will blow up” in seconds across all courses, all recovery types, and all fatigue states. It is more suitable for answering:

1. Why does this athlete crash when encountering repeated speed changes?
2. Why do two athletes with similar CP perform so differently on technical courses?
3. Why does a certain 30/30 workout plan fit A perfectly but cause B to completely lose control?

Pitfall Four: Ignoring Technical Ability

International triathlon research shows that the more dangerous corners there are, the more time is spent above MAP. This is not purely a physiological issue; it also includes cornering, braking, hard acceleration out of corners, and judging group position. If an athlete has poor technique and is often forced to brake hard and then stomp on the pedals, their W’ expenditure will certainly be higher than that of technically mature opponents. In other words, CP/W’ cannot just be used to generate training plan spreadsheets; it must be corrected back to on-course technique.

IX. Practical Conclusion: Breaking Through Bottlenecks with CP/W’ Relies on Separate Testing, Separate Training, and Separate Pacing

For triathlon athletes, the value of critical power does not lie in having another trendy term, but in truly quantifying “stable sustainable output” separately from “limited but key changeable high-intensity capacity.” This allows us to more clearly answer: Is the athlete lacking steady-state ability, lacking W’ capacity, lacking W’ recovery speed, or lacking bike-to-run conversion ability?

To use this tool most effectively, remember three principles. First, test bike and run separately, do not substitute one for the other. Second, do not just look at CP, also look at W’ and recovery ability under course variability. Third, any model must be validated against race scenarios, especially conversion runs, technical courses, nutrition, and fatigue accumulation.

When you do these three things well, CP/W’ will not just be laboratory data, but become a very powerful decision-making system in triathlon: it can help you judge when to hold, when to surge, which workout plans can truly break through bottlenecks, and why you always lose speed in the latter part of the bike or the first part of the conversion run. This is the true value of critical power for triathlon athletes.

Related Reading

• 【Deep Analysis】How sub-max jogging and aerobic health athletes use Critical Power to evaluate breaking through bottlenecks? Exploring the core decryption of the scientific mechanism of Anaerobic Work Capacity (W’) and physiological limits
• Power Output Stability and Endurance Performance: Practical Application Research of Critical Power Model
• Critical Power and W’: Explaining How Long You Can Last with a Two-Parameter Model
• Bike W’ (W Prime) Model: Training Significance of Anaerobic Work Capacity