The low leakage AVX 10F 2.7V supercapacitor that I used was specified to have 30uA max leakage current, and that worried me because low power applications can have 20uA average current. Is my supercapacitor really so leaky? 

I have tested 3 different supercapacitors. One is from the well known brand AVX and two others are brandless supercapacitors from Aliexpress. 

After keeping the capacitors at a charged high voltage for 24 hours I simply measured voltage decay every day with a cheap multimeter and calculated i_leak=c[F]*(u_yesterday[V]-u_today[V])/t[s]

The tested supercapacitors

1. AVX SCCT20B106PRB 10F 2.7V, specified to have 30uA at 72hrs
2. Pixiuonline J469Y 1F 2.7V from Aliexpress
3. ZHIPU CXHP series 1F 2.7V from Aliexpress, specified to have 8uA at 72hrs

And the result is put in this chart

• Interestingly the low leakage 10F AVX supercapacitor had a lower leakage than the 1F supercapacitors from aliexpress, while I expected larger leakage for larger supercapacitors. So it seems that the AVX is really low leakage.
• the leakage after 72hrs is below 1uA for all supercapacitors which is much lower than expected (30uA for the AVX and 8uA for one of the others). Apparently the max leakage in the datasheet is much higher than actual leakage. Maybe because it is specified at the full temperature range and I tested at 25 degrees.
• the leakage in the first 24hrs is much higher than after 72hrs but still less than 10uA for all supercapacitors

I conclude that my tested supercapacitors have a leakage current (1uA/72hrs) much lower than many low power applications