It shouldn't have affected how hot the ballast got, but I can assure you it did. We needed a fan pointing at the fixture before adding PFC, but not afterwards. I don't recall the exact temperature figures, but it was significant.

Your low power factor is not directly metered. However, if "everyone" has a low power factor, the transmission cost of the electricity will go up. Because, as Rommie said, although the low power factor is not registered as power used, it does impact the current running through the conductors in the grid, in the transformers and in your home. Eventually that gets factored into everyone's power bill, so it's always best to compensate loads with a bad power factor.

We used to have a tanning bed with 10x 100w fluorescents, with conventional 230v choke ballasts, and no compensation capacitors. Although 1000w (4,3 ampere) is easily handled (at our 230v mains here) by any bog standard appliance cable, the cable on the tanning bed got seriously warm and soft. This was because the power factor of the choke ballast and lamp is together 0,53. that makes it look as if there was 8 ampere flowing through the cable.

It likely was a 3x 0,75mm² cable, which would be fine for 1000w appliances, but woefully insufficient for 8 ampere, or loads that 'look like 2000w'.  But i do not remember what cable was on it, sadly. 2000w heaters often run on 1,5mm² cords and not heat up, so they likely seriously miscalculated the load on the cord.

The appropriate capacitor for the 100w ballasts here is 8µF, and an 80µF mains rated capacitor would've added significantly to the cost, so likely left out for that reason.

There should be no effect on how hot a ballast gets, it does not influence the current running through the ballast if it's just a compensation capacitor parallel to the mains connection.

Just out of curiosity, what's a barium getter?

James, I have a Sylvania H39KC-175/W from 1978 that's still a clear-top...
Interesting info, BTW...

Not usually. If you are a normal customer, they are billing you for real power in kWh which accounts for power factor in the equation of voltage times current to get the *real power*.
But if you are an industrial customer, you can be being billed for apparent power in kVArh (which doesn’t account for power factor and just straight up multiplies the voltage by the current), then you will be charged more for a low power factor.

But chances are you’re not because apparent power is only Really charged to industrial customers, and not residential customers.

That's what I thought, but I wasn't sure. Thanks!

We noticed a big difference to how our Goldeye ran after installing the correct rating of capacitor to bring the PF up to .85 from .4 (ish) 
Before it was running really hot and needed additional cooling, but not after the work had been done. 

I'd never trust one of those el-cheapo adaptors, no matter what the rating of the lamp.!

It won't make any difference to your bill, but it will make a difference to the power used. The lamp will draw more current and if you're on the brink of the maximum capacity of the wiring, it could push you over.

Personally I always like to make sure the PF of my discharge lamp circuits is as high as possible. Here, that's usually done with a capacitor across the supply, it's often different on your side of the water.

Not usually. If you are a normal customer, they are billing you for real power in kWh which accounts for power factor in the equation of voltage times current to get the *real power*.
But if you are an industrial customer, you can be being billed for apparent power in kVArh (which doesn’t account for power factor and just straight up multiplies the voltage by the current), then you will be charged more for a low power factor.

But chances are you’re not because apparent power is only Really charged to industrial customers, and not residential customers.

According to the records I can find, the clear band design was introduced towards the end of 1976 for all powder-coated BT lamps irrespective of whatever the coating material was. This was an emergency quality project and by start of 1977 it should have been applied to all types.  Even though it was strictly only needed for certain quartz lamps it was applied to all, because it was the same coating machines that produced all lamp types.

In 1979 the phosphor adhesion problems were solved and full coating was resumed.  Since that was not an emergency project it seems to have taken longer to introduce, and it was not until 1980 that all types had been converted.  The BT56 line seems to have been the last to revert to full coating.  If anyone has the time, it could be interesting to make a study of all the Sylvania coated HID lamps on this site from the period eg 1975-1985 to identify which coating design was in use!