Saturday, January 10, 2009

Turbocharged Training

There's been a bit of discussion lately on various training forums about a topic that seems to crop up every so often. It's a perennial favourite. Certainly I'm not the first to write about it and I won't be the last.

Why is my power different when training indoors* compared to when I ride outside? And what can I do about it?
* Indoor training being training done on an ergobike, or with the bike locked into a turbo trainer or riding your bike on rollers. Often performed inside the house, in the garden shed or garage, on a balcony or at the local gym or training centre.

Usually people train in such a fashion because they either haven't the time or opportunity for a ride outdoors, they might be recovering from injury and/or need the controlled and safe environment an indoor trainer provides, or the riding conditions outside are not suitable (cold, rain, snow, darkness and so on). Certainly riding indoors is a safe and excellent training alternative when heading out the front door on your favourite steed is not possible.

For many riders though, they find generating power indoors much harder than when riding outdoors and end up riding at a lower power as that's all they can do (but this is not the case for all though, and some can actually produce more power indoors than outside, although that is less common).

So if I can't generate the same power, then am I getting the same training benefit?

And if power is significantly different indoors, should I use a different FTP for indoor rides (so training levels and ride data are adjusted accordingly)?

Well the answers are not straightforward but let's explore the solution(s).

The first thing to do is to understand why a difference in power production exists. Then the second thing is to take steps to address the differences between each scenario and "bridge the gap". Finally, one then needs to make decisions about how the data from their indoor training should be interpreted.

So why is it common for power to be different?

There are four main factors at play here:
  • Cooling & air flow
  • Inertial load
  • Motivation
  • Adaptation
I'll explore each of those in a bit more depth a bit further down.

OK, so what about the training benefit and setting of FTP?

Well power is power and if you are burning kJ at a lower rate, then the metabolic adaptations relating to that will be correspondingly different. So if turning out a lower power really concerns you, then the priority is to address the factors that influence indoor power production and reduce the gap so that training can still be done within the intended training level. Then the problem goes away.

Nevertheless, "hard is hard" and "alls you can do is alls you can do", so if you are unable to address/fix the key reasons why power is less indoors, then set your training at a level that is attainable for that scenario. It's better than staying on the couch. Rather than worry about what percentage of FTP or MAP that should be, just use previous indoor workouts as your guide. That really should be the guide anyway, irrespective of mode of training you are doing.

What matters is that you do the workout at around the right intensity for the right duration, rather than the precise wattage.

What about FTP, and the calculation of TSS and the other metrics that flow from it?

This really is an issue of what you are training for and where the majority of your riding will be during the course of that training period. If the trainer only represents a minority of your ride time and your power is say 10% less on the trainer, then it only represents a small difference in the calculation of overall training load. It is simply not worth the bother to have separate FTP values and calculations. The Impulse-Response model (aka the Performance Manager and the metrics CTL, ATL and TSB) is fairly robust. It is about the forest, not the trees.

For example, let's examine a common hour-long training ride and say, for whatever reason, your indoor power is 10% less than outdoors: 2 x 20-min at FTP + warm up and cool down.

Outdoors, this would accumulate ~ 85 TSS and indoors (with ~10% less power), ~ 70 TSS. A difference of 15 TSS (which is about the equivalent stress of 15 minutes of endurance level riding).
So if the difference in TSS calculated from a 2x20 workout (equivalent to about 15-minutes of basic endurance level riding) is concerning you, then sit on the trainer for another 15-minutes.

If however, the trainer represents (for that training period) a large proportion of your training time, then setting FTP according to that training mode makes sense. But where such rides are only occasional, then there really is no reason to worry about minor variations in the numbers, just move onto the next day.

The same principles apply if talking about training at altitude (occasional change in altitude vs. a lengthly block at a different altitude) or different bikes (occasional or lengthly training blocks on a given bike/position).

Read on for more details on the four elements of indoor training that affect our ability to produce power indoors and how you might do something about it.

People consistently underestimate the cooling needs when training indoors. There's some weird theory that a large pool of sweat forming beneath you is a good thing. All that tells me is that the air flow and cooling arrangement is perhaps inadequate for the task. A body that is under stress and not being adequately cooled will underperform.

Keep in mind that the typical cyclist operates at around 21-22% efficiency (give or take a couple of percentage points). Cycling efficiency is a measure of the ratio of energy reaching the cranks of the bicycle as a proportion of the total energy metabolised*.  In other words, to generate 100W at the cranks, our bodies are metabolising energy at the rate of 100W / 21.5%  = 465W.  

So of that 465W, 100W is converted to mechanical energy at the cranks, with the vast majority of the balance being converted to (waste) heat, with a bit used of course to run the rest of the body's functions.

What that means is that for every 100W we put through the cranks, roughly another 360W are generated as waste heat. How much heat exactly will vary depending on the individual's efficiency level (typical range is 19-24%).

So if for instance you are doing some intervals at 300W, you are in effect pumping out the heat equivalent of a 1,000-1,200W electric heater!  Now do you see why we heat up so quickly when training hard and an effective cooling system is required?  Especially if the ambient temperature is quite warm to begin with, and particularly so if the conditions are humid.

When you hop on your bike for an endurance ride, you have a ~ 30km/h wind flowing over your whole body constantly wicking sweat away and keeping you cool(er). So why would you expect to perform as well indoors with no air flow, or the piddling excuse of a breeze that comes from a domestic fan? Get real. If training indoors is going to become a sizeable chunk of your training time, then get some decent cooling happening and have some strong air flow over you. A large industrial strength fan costs much less than a trainer or rollers, so bite the bullet and sort it. But be prepared for the additional noise.

Some people do perform indoor training in quite cold environments, so of course they might be able to get away with less air flow than others.

* there are a couple of different efficiency measures (e.g. gross and delta efficiency), but for all intents and purposes, this basic definition will suffice in this context.

Inertial load is the next main differential factor when comparing indoor and outdoor training. Without going into too much detail, when we ride outdoors, we have the inertial load of a bike and rider moving at some speed, plus that of the wheels turning. If we stopped pedaling, our rear wheel doesn't suddenly slow or stop turning, we would coast for quite some time. On many trainers however, since we are not moving, the inertial load is much less and confined to the rear wheel spinning and any small flywheel that the trainer has attached to the roller. When you stop pedaling, the wheel slows and comes to a halt relatively quickly. Some are worse than others.

Now what happens is each scenario feels quite different to ride, muscle activation is different, the neuromuscular demands are different and these can be enough for some to make power production much harder. In general, low inertial load trainers tends to emphasise the "dead spots" in the pedal stroke (when the cranks are passing through the 12/6 O'Clock position), whereas riding with a higher inertial load enables one to breeze through (and not waste effort on) the dead spots and focus on the downstroke where the bulk of power is produced.

Fortunately there is a way to increase the inertial load of a trainer, and that's by having a flywheel attached to the trainer's roller (or even by adding mass to the wheel itself). How much mass is needed? Well to replicate the inertial load of a rider, it would need a very heavy flywheel spinning very quickly. Think of a 20-30kg flywheel spinning at 500-800 rpm. Yikes!!

Fortunately, for effective training, going that far is not really necessary and having enough rotating mass to help smooth out those dead spots is enough. I don't have one myself but trainers like the Kurt Kinetic Road Machine or the 1-Up trainer are an excellent example of this. They both have small but effective flywheels attached to the rolling mechanism.

These are ideal options for those that are looking to attach their existing bike to a trainer but also need some portability with their indoor unit.

The other option is a dedicated ergobike like the Schwinn or Saris indoor trainers (or other similar machines). These types of set up have the advantage of being able to incorporate a much larger and heavier flywheel than a turbo trainer. They are of course dedicated units and need a permanent place to live.

So what does one do? Well if cooling has been sorted, and power is still down, then consider the inertial load of your trainer set up. Does it have a flywheel? Can one be added? Should I look at an alternative trainer? Certainly I would recommend trying a trainer that has a decent flywheel to see how much better it is to ride.Edit note: I added the following paragraph in March 2011 as something I'd been meaning to do for quite some time, just had forgotten to do it.
I should add that the idea of inertial load on an indoor trainer affecting pedaling isn't actually backed by evidence other than anecdotal, from myself and many others I know that have used such trainers. As an example, this link to a study extract on PubMed indicates
that varying crank inertial loads has little or no effect on steady state pedaling coordination.

Motivation is a big issue in training and racing, and it is sorely tested when riding indoors. Many find training indoors mind-numbingly, excrutiatingly boring. Then there are others who really love it and are happy to spend hours tapping it out, sometimes preferring that to a ride outdoors. Each to their own.

If a lack of motivation is an issue, then it needs to be addressed, otherwise don't waste your money on a trainer you won't use. It'll just end up gathering dust in the corner of the shed.

There are many ways to overcome any motivational challenges you face:

Variety - there are lots of training workouts available, so keep the variety up. Dream up some of your own!

Duration - indoor riding is hard work, there's no let up or coasting, so don't make the workouts as long as you might ride outdoors. It is better to complete a shorter workout and want to come back for more next time, than to get off absolutely hating it and sitting out the next one on the couch or staying in bed.

Set Challenges - set yourself targets for the session and maybe have reminders of your goal event in front of you as well.

Music - this is a good one - having you favourite training tunes blasting away, or on your iPod to keep the neighbours happy.

Video - what about watching highlights of your favourite stage race or one-day classic. You can be smacking it up Ventoux with the Pros. Of course there is a big market out there for indoor cycle training videos, so if that floats your boat, then go for it!

Computer aids - there are lots and some of the favourites are heart rate monitors, spped and cadence measurement computers and of course my favourite - power meters. These are especially helpful so that training is focussed and performed at the right intensity.

Ergo controllers and virtual riding - there are many trainers that can automatically control the resistance level of the trainer and be pre-programmed to control a workout. Some can even display video of an animated figure or some real life video to provide a distraction from the effort and help to pass the time.

Of course the most obvious answer is simply to HTFU.

Adaptation is the last of the four key issues. Since there are differences in riding on a trainer to riding outdoors, some of which have been discussed already, then it stands to reason it will take some time for the body to adapt to training under different conditions. If you only ride the trainer occasionally, then you may never fully adapt to being able to generate power similar to outdoor riding.

However, if you ride on a trainer regularly and with sufficient volume, and you address the other three main factors, then you will adapt and improve your ability to produce power indoors and the gap to outdoor power will typically narrow.

What do I do?

Well I set about addressing all of the issues and descibe my indoor training set up here:
Turbocharged Training Thread on TT Forum

Have fun indoors!


Bill Black said...

Nicely said and true -- but I'm not sure that a man from "down-under" can ever truly embrace indoor training unless you are significantly further down under than Oz.

Bill Black

Bruce Diesel said...

Hey Alex,

As you may have seen, I have been very involved in the discussions around this topic. And there are a few things that bother me with these arguments.

Firstly, if testing my FTP in one environment (say on the road), then apply it to another (say indoors) is valid. Then surely the reverse must also be true, i.e. testing indoors and applying to outdoors. Assuming I can produce more power outdoors than I do indoors, then I am able to perform workouts of greater than 100TSS when training outdoors.

Secondly, if we correlate kJ expenditure to training adaption. Then how can 100TSS early in the season (when FTP is lower) be equivalent to 100TSS later in the season, when FTP is higher and overall kJ expenditure is higher for that worjout?

I know this sounds like nit-picking, and I agree that these things should only be taken into account when all other attempts to equalise environments have been exhausted. But I have experienced this issue with most of the athletes I coach, in fact, I have very seldom found athletes that are equivalent indoors, outdoors, on road bikes and MTB bikes.


Alex Simmons said...

Ha! True dat Bill. I have nowhere near the experience of indoor training that my northern hemisphere cousins do (we never experience the winter weather you guys do), but I at least I listen to and learn from the ones I interact with (and coach).

As I returned to riding again, I did do all of my earlier efforts on Thunderbird 7, not because of weather but because of injury. In the last month, road rides outdoors are now more frequent than ergo rides (but it is also summer here).

In any case, it will now be a bigger part of my life than ever before.

Alex Simmons said...

Hey Bruce

If we are talking 1/7 rides being in a different mode, then I wouldn't worry about the FTP/TSS difference, train according to the majority mode for that block of training. So if you are mostly indoors and get one outdoor ride/week, then the TSS will be a bit inflated that day. Someone might have an overall weekly CTL ramp of 4.3 instead of 4.0 No big deal.

The time course for adaptations is relatively long, so 1/7 inflated (or deflated) figures isn't going to change the overall pattern of what we see in the ATL/CTL plots.

I mean how many out there are estimating TSS for a proportion of their rides but still find the Performance Manager works well? Quite a few I suspect. That's what I meant by the model's robustness.

If the modes are equally shared, and as long as training is according to the effort level relevant for that mode, then if you really want to go with different FTPs, by all means go ahead.

Personally I'd just split the difference, maybe with a slant to FTP being closer to the higher number.

As for TSS and kJ wrt adaptations, yes what you say makes sense, although I suspect it is the difference between relative adaptations (TSS/min) and absolute adaptations (kJ/min). By relative I mean wrt your existing fitness level. By absolute I mean things like absolute oxygen uptake, ability to consume sufficient kJ etc.

Interestingly, of the riders I've coached that do a significant level of indoor work, I don't see big differences in power output.

For me lately, I recently switched from indoor threshold tolerance work to outdoor efforts and so far I haven't produced more power outdoors yet!

But it's all good discussion and thanks for your comments guys!

Nice to know people check it out every so often.

Jesse's Cycling Addiction said...

Now that I am getting into training with power and have to do most of my training indoors because of work schedule your blog entry will really help me understand some things that I have been struggling with.


Bruce Diesel said...

To be honest, my biggest concern in this area is in the differences between MTB and road bikes.

I am the coach of a Pro MTB team, and we are seeing a difference in power output between MTB and road. Unfortunately the guys spend up to 50% of their time on road bikes.

The only major positional difference is in Q-factor, so we are busy investigating. But the differences are certainly large enough to not sweep under the carpet - up to 30w at FTP in some instances.

Stomp tests and dynamic calibration on computrainer have all checked out fine.

rmur said...

"Nice to know people check it out every so often."

the lurkers are here.

Re the FTP bit, I keep mine at outdoor FTP year-round despite doing 90+ % of my winter training indoors. Power around FTP has shown to be somewhere around 5% lower indoors for me - though with a better fan (18" industrial puppy plus my old anemic 16" pedestal fan) I hope to close that gap this winter.

As I belive Alex was pointing out - do everything you can to close that gap (from the bottom-up of course) before crying "enough". Now, I'm talking about tracking weekly TSS, avg.IF, PMC etc -- NOT about what levels to attempt my workouts: those are strictly based on week-to-week training ... plus however delusional I am on any particular day :-)

The one factor I can't get around though is altitude ... not a factor here but if I was living at sea-level and racing at altitude or vice-versa with significant blocks of training/racing in each ... I'm not sure a single-FTP mantra would work very well. Aside from training with an oxygen cyclinder just can one do?


SteveP said...


What about an arbitrary rule of thumb that, if a training block in a specific environment/terrain/bike position were to proceed for equal to or greater than one's selected ATL TC, then using an FTP specific to that environment/terrain/bike position is probaly good? Forays between training environments within an ATL TC would be best be handled by selecting FTP by the "which training environment predominates?" rule, or if closer to a 50:50 split between training environments could be handled either by Bruce's separate FTP method, or Alex's split the FTP difference method.

Also, I have been hesitant to accept that the training adaptations are lesser indoors than outdoors (assuming lower average power indoor, but same average IF). However, there is evidence that when the same relationship exists between altitude and sea level, the training adaptations at altitude do not necessarily translate to more power at sea level (hence the greater popularity of LHTL). I'm just not aware of any other evidence WRT lesser adaptations in a lower AP / equivilent IF situation.

Another gem. Sound advice as usual.
Steve Palladino

Alex Simmons said...

Yeah, all good comments guys.

Bruce, I don't have an answer specifically for your dilemma other than it's probably a bunch of 5W things adding up. Not sure what PM you are using on the bikes but if it's a PT then there's probably a difference in drivetrain efficiency between bike types. And a positional difference. And maybe the neuromuscular differences from pushing the MTB/terrain is causing a few watts as well. Don't know, just speculating.

Rick, bang on and by doing that, you are just lowering your reported CTL ramp rate but as long as you are consistent in how it's applied and know the rate at which you can sustain, that's what matters.

Steve, interesting thought and certainly one to be considered. Even then, I kind of see it akin to transitioning to a focussed block of higher intensity work of Level 5+, where sustaining a CTL ramp can sometimes be difficult to achieve, yet actual fatigue is ever present even though TSB neutralises somewhat.

End of day, I expect the CTL to do certain things at certain times under given conditions, and just because it's not going up doesn't mean one isn't accumulating fatigue or not improving fitness.

It's performance that matters, not one's CTL. :)

So if one goes to altitude for a week or two of training, then I wouldn't have an expectation of CTL ramping up at the same rate. I would however if we did choose to adjust FTP for that block.

Point taken on the evidence re lower watts, same IF.

Bruce Diesel said...

My guys are riding powertaps on both MTB and Road bikes. We are busy checking everything out, because hey, if I can iron out some of these issues and lift some of these guys 20w - 30w then that is part of my job is it not :-)

Maggs said...

Just found this blog. This is a good post. Answers a few questions of mine. I seem to be one of the few who get more watts inside...especially in group trainer sessions. I'm doing my first 20 minute test outside this weekend, so my goal will be to get similar watts on the road.

Matthew said...

I ride road and trained this year for the Marmotte, a race in the Alps. I did my intervals outside this year for this reason - I was getting too hot indoors and was worried my power was down.

However, I struggled to get steady power on the lumpy hills in UK, and when I got to Alps for the race I was basically put back in conditions I had indoors: steady power, in very hot conditions, with no wind! As such my numbers were lower than expected and I thought I was really suffering and having a bad day. I think actually training indoors would have helped more.

otb4evr said...

WRT the Mt. bike FTP differences: Wasn't there a discussion a while back on wattage (it could even have been when it was hosted on Topica) about the lower wattages because of the nobbies on the tires and how they interfaced with the terrain?

Neil Cianci said...

Personally speaking, I prefer to train outdoors. The various weather conditions makes the ride more interesting aside from the fact that I take different routes. I do have equipment at home as not all times of the year can I go out and enjoy the fresh air.

Dan Bennett said...

Nice and informative post, thanks for sharing!

Dan Bennett
Progressive Cycle Coaching

Lee said...


Great article. I am still however not clear how a variance in "inertial load" would affect my ability to produce a certain wattage. Clearly (to me at least) it would affect the wattage required to maintain a particular speed (or cadence in a particular gear at a specific resistance). I agree that it would "feel" different and require less wattage to "re-accellerate" back up to a given speed, or to maintain a particular speed. I fail to see how a larger or smaller flywheel would translate to my body being able to produce more or less watts to the cranks. What am I missing?

Alex Simmons said...

I think the easiest way to answer your quetion is to get you to ride a low inertia magnetic resistance trainer, then ride a high(er) inertial load trainer, e.g. a Lemond revolution, under the same conditions. The difference required to pedal on each "normally" is quite pronounced.

Jim Scott said...

The way that I have always looked at it was that there is an additional torque spike needed to bring the wheel back up to speed on the low inertia trainers. This would be just like climbing a steep hill.