Giant power meter left/right balance

Question

I have a Giant TCR Advanced Pro 0 Disc 2021 which came with the Giant PowerPro power meter built into the cranks.

I've seen people posting things like, "Don't take too much notice of the L/R power balance" which I figure is because they are typically not that accurate.

But mine seem awfully consistent in telling me that my right leg is doing exactly 60% of the work on average.

I've tried updating the firmware, tightening the two crank screws super evenly in small steps to the recommended torque setting and calibrating, after which the readings are exactly the same, 60% on the right leg.

Is this normal or should I be looking at whether my pedalling really is that unbalanced?

Some additions, I went for a ride this morning and followed some advice from answers below.

1. Yes it's a two part power meter, one in the middle of the crank arm and one on the drive side spider.
2. I put the bike outside for 10 minutes before the ride then calibrated using the Giant app. First half of the ride I averaged 40-60 L-R.
3. At the turn I stopped and tried to calibrate using the Garmin 1050 plus. It said it was happy with the result but I'm pretty sure it did nothing. Power balance unchanged.
4. After another 10km I stopped again, turned off the Garmin, fired up the Giant app and calibrated again. Suddenly I'm averaging somewhere between 48-52 and 49-51 for the rest of the ride. Also, when applying 300W or more, the left power goes up to 52 to 54%. When I drop under 125W it goes down to 46 to 48%.

So my conclusions:

1. Calibration is tricky. Mostly it seems to get it wrong. Now that I have what seems to be a usable calibration I hesitate to change anything until the temperature goes up again next Summer.
2. I think that the left power meter may be reading higher than the right under higher load. But I suspect that my left leg is also being lazy under reduced load. The current calibration seems to balance those out a bit. I'm not going to bother doing anything more, unless I happen to find a group with a rider of the right stature (and preferably left handed) to check their power balance on my bike.

Answer 1

Typically, we don't need to pay too much attention to left/right balance because it's not clear that 50/50 is the optimal balance to aim for. Many of us have a slight imbalance. The advice isn't necessarily due to inaccuracy.

Unfortunately, it has been shown that the 2016-2020 generation of Shimano cranksets (including R8000, which you have), produce inaccurate drive-side power readings, as detailed here. This is because the crankset is asymmetric on the drive-side, which means that the strain gauges don't produce the same amount of strain in all orientations. However, as the testing I linked has shown, the typical result is that those cranks underestimate power on the drive (right) side. If this holds in your case, that means your actual imbalance may be greater than 60/40.

Again, it's not clear what to do if you have an imbalance and if you don't have any congenital issues with one leg or an older injury that you haven't rehabbed from.

---

Tackling the unanswered question of what to do now: If you feel like you have an injury, you could consider seeing a physiotherapist, who should be able to evaluate if you have a current strength imbalance. Some bike fitters will be able to evaluate this and will know corrective exercises, even if they aren't physiotherapists by training.

As @Carel's answer says, you should manually zero your power meter using your head unit. This is a basic calibration step. Your power meter may have auto-zeroing functionality, but this can't hurt. If you want to confirm the power imbalance, you could try to borrow a pair of power meter pedals.

You could continue to use your power meter for training as-is, especially if that is your only power meter. You could probably convert it to single-side (e.g. take the battery out of the DS meter, may have to unpair the meters from your setup app) and use the NDS power reading only. However, because we tend to produce power asymmetrically, it's not clear that the total power estimate you got from a left-only power meter would be more accurate than your current setup. If this isn't your only power meter, then having numbers that aren't comparable across your setups is a bit of a more complex problem.

---

Tackling the issue raised in the comments: because this bike has an Ultegra group including stock crankset, I believe that the power meter (which I believe is this model) is two independent units. Power meters like Quarq and Power2max are spider-based, i.e. they measure power only at the chainring spider. These are only compatible with direct mount cranks, so they physically can't fit on Shimano cranksets (although both companies make versions that take Shimano's chainrings). Spider-based power meters directly measure your total power. They can algorithmically estimate how much each leg contributes to power, but they don't directly measure each leg. (It isn't clear that algorithmic estimation is that inaccurate.) The same would be true of hub-based PMs like Powertap, except that Powertap doesn't make hubs anymore. True dual-sided PMs, like the Giant unit or power meter pedals, will directly measure each leg's power.

Answer 2

Power meters need calibration, even if they have been accurately factory calibrated. This is ideally done when they have reached the temperature of the operating environment. They are based on strain gauges, thin films of a component that changes electric characteristics when deformed. The electronics in the meter register these and and compare them to a built-in table to compute the power value.

Unfortunately not only deformation of the gauges changes the indicated values, the temperature of the device does as well. Therefore power meters include a temperature compensation circuit. When you take your bike outside you'll have to wait for a while until it has reached the temperature of the environment you're in. Repeat the recommended calibration process after a short warm-up ride. The values will certainly be more consistent.

Bottom bracket mounted devices work like torsion bars. They measure the torque on the BB axle, the tiny angle the axle is twisted by the load on the forward foot against the trailing foot vs. the unloaded state of the axle. Because they're enclosed they also need more time for temperature equilibrium.