Posted by: Ian | March 28, 2013

Quantifying and beating the bike bonk

On Tuesday I rode up Haleakala for the second time, having sworn three years ago that I’d never do it again. It’s a 35.1 mile climb that rises 10,000 feet monotonically from sea level at an average gradient of 5.4%. It’s so different from anything else I ever cycle that it was impossible to set a sensible target but I felt I had a four hour time in me. If I do, I certainly didn’t find it on the day.

I set off with a simple plan of keeping my heart rate under what I felt was a sustainable level and the general idea of getting up quickly enough that food wasn’t going to be a major factor. I’ve developed a scepticism of the science of sports nutrition: I find that if I eat enough proper food I never bonk; gels and energy bars make me feel sick. I do, though, usually carb up at least one bidon with powder if I’m out on a ride that’s likely to be tough. However, I’d forgotten to bring any drink powder to Maui and all I had instead were electrolyte tablets. To compensate, I picked up a couple of Clif bars at a local health food shop.

At first my plan went okay and the ride was fun and easy. I had one of the Clif bars – a small “organic” number – that was fine but I found the second one – a larger, peanut butter slab – inedible.  By 21 miles I felt like crap and had to stop. I chowed down a sandwich and took a short rest then carried on. But now it seemed like a real grind and all the power and speed had gone from my legs. I seriously doubted whether I would finish the climb, not because I couldn’t but because I was reduced to spinning up at 7 mph, which felt miserable. I stopped again a few miles later and then for six or seven miles in a row I stopped on the mile point to let my heart rate recover in the hope that I could staunch the flood of lactate to my muscles. Eventually I reached the top in an elapsed time of 4:53 and a moving time of 4:26.

The bonk is very clear from this chart of my mile splits:

Haleaka bonking

Each point represents one mile on the ascent, showing the net gain in elevation versus my average speed. The green dots are my first 21 miles and the orange dots the next 14. As you can see, the green dots lie very neatly on a line and the orange dots lie on a different, slower line.

The bonk is just as clear from my heart rate chart:

Haleakala Avg HR

For the first 21 miles my average heart rate over each mile was tending to reach the sub-160 bpm limit I’d set myself. Then from mile 22 onwards it crashes down. There are two ways to look at this. Mechanically, it’s a simple consequence of me stopping more, especially when I adopted the routine of stopping at each mile point to let my heart rate revert to a recovery level of 134 bpm. On the other hand, the reason I stopped was because it was very hard not to: it was just too tough to keep my heart near 159 bpm.

Right at the end I managed a slight rally over the last three miles once the end was in sight but this was more in the nature of preventing further decline rather than finishing on a high flourish.

When I did the same ride three years ago my heart rate chart had the same shape, with the crash at the same point (although then my speeds were lower and my heart rate higher). It’s no coincidence that the bonk came when it did both times. Another chart indicates why…

Haleakala calsAlthough I’m distrustful of Garmin’s (or any other) measure of estimated calorie burn, the shape of the curve (if not the absolute level of the values) is probably about right. It shows that the calorific load rises to a peak at the 20 mile point, running me out of my energy reserves.

So that’s the bonk problem: at the critical point of the ride I ran out of resources and couldn’t sustain the effort.

I can estimate the cost of the bonk. The line of best fit through the green points in the first chart has an intercept of 15.4 mph, while the line of best fit through the orange points meets the y axis at 12.1 mph. Thus post bonk my equivalent speed on the flat fell by 21%.

Paula cycled up Haleakala the day before I did it and her ride indicates the solution to the bonk problem. She ate well beforehand, rode at a steady pace and made frequent short stops, eating at each one. Proceeding this way, she made her way to the top at a consistent speed and, unlike me, looked reasonably fresh when she got there.

My two concrete strategies if I do it again would be:

1. Fuel properly.

The night before I fixed myself some pasta pesto but it wasn’t nearly as big a meal as I usually have on the eve of a hard ride. Before I left I had a bowl of muesli with a banana chopped in it and between then and the bonk a few hours later my entire calorie intake comprised one small energy bar. If I try this again I’m going to eat far more beforehand, take energy drink and eat something every 30 to 60 minutes.

2. Set a lower max heart rate.

I had a metabolic test a couple of years ago that gave me, along with other measures, my Lactate Threshold Heart Rate. I assume that all of the riding and training I’ve done over the intervening time has caused this to rise and I set my max heart rate accordingly. However, the benefit of not being exhausted with 40% of the ride left is so great that I’d be more cautious if I did this again and drop my max by a further five or 10 bpm.

The bad news is that the arithmetic shows that if I had stayed on the same slope as the green points on the first chart for the last 40% of the ride I would only have saved just under 19 minutes. Sadly, this suggests that even if I’d also avoided all of the bonk-induced stops, I would still have fallen seven minutes to the high side of my 4:00 target. I know that on shorter rides of the same gradient I can easily go faster than this – Porlock toll road, for example, has virtually the same slope and over its paltry four miles I can average over 10 mph – but the evidence that I can sustain the effort needed over a Haleakalan distance just isn’t there. Irrationally, I stubbornly believe that I can. Dropping my heart rate even lower would inevitably slow me down further over the first twenty miles. I’d have to hope that this would leave me fresh enough to ramp up the speed towards the end.

From the top I was so achey that I kept my descending speed in the 30’s, making a pit stop at the visitor centre and then again for an ice cream. By the last third of the descent I was having fun again.

Haleakala summit



  1. […] just read Ian’s post about bonking, so I recalled my own little problem. About a month ago, I came across an article on, […]

  2. That’s a gorgeous R3!

    The stats make this ride sound almost as difficult as Hurricane Ridge, but quite a bit more drawn out. 😉

    I didn’t see this mentioned in your post, but the air thins out a bit at those altitudes, and I’m certain this contributed to the difficulty of the ride.

    • Thanks! The R3 is fantastic and takes away any bike excuses I might ever have.
      Haleakala is rated the 2nd hardest climb in the US on the two sources that I know of, at least one of which uses altitude as a factor in the difficulty calc, as you suggest. I wrote a post on this called “Difficulty” a few years ago that you can find by using the little search box if you’re interested, like I am, in how climbs are rated.

  3. 35 miles at 5%. That sounds incredibly tough and like nothing you can really prepare for. Chapeau for making it to the summit

    • Thanks. It is tough and a big part of the toughness is knowing that it’s a gradient that really shouldn’t be slowing you down as much as it does!!

  4. Would have been interesting to see what your blood glucose levels were. You can buy a tester from Boots.

    • It would be. Our neighbour has type 1 diabetes and he tests his glucose when he’s out riding. He now has a continuous glucose monitor – and a glucose pump – wired in like a battery pack.

      • That’s what made me think of it. I developed late onset type 1 last year. I thought I’d bonked on a bike before but nothing prepared me for the feeling that too much insulin causes.

      • I’m sure that’s true. Even after my “bonk” I managed to cycle up the remaining 4,000+ feet at a reasonable pace – just not the speed I hoped for.

      • Yes not taking anything away from that sensation, I was there this morning. Not a full on hypo but I had a sudden 20% drop in pace and I could do nothing about it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: