Certainly beyond my understanding. Perhaps Bean can give some insight to such tripping.
@BeanAnimal
GFCIs have evolved a good bit over the last 30 or so years.
The units from the 1970's were somewhat terrible. They worked, but were very prone to nuisance tripping. Surprisingly, you will still see some of these in service. They had no fail-safe so still work as receptacles even if the protection has failed!
The 80's units were better, but still not overly reliable and also nuisance tripped. I have read that they were sensitive to moisture internally (think bathroom steam). Again, you would be amazed at how many of these are still in service. I don't think that these are fail-safe either and have no internal self-test circuitry. 5 mA was the common trip current, but time to trip varied by manufacturer.
The units from the 2000's were greatly improved with better surge filtering and moisture protection. I think this is when the trip threshold of 5 mA < 25 ms was established. These units were far less prone to nuisance trips.
In 2015 UL 943 was updated to require self monitoring and lockout (you can't reset the device) if the internal diagnostics determine that the GFCI protection circuit has failed. It is my understanding that they will also not "set" if miswired.
The most current generation of GFCIs ~2020 forward have far better filtering and are very stable with regard to nuisance trips.
Nuisance trips:
A GFCI compares the current flowing out on the hot conductor to the current returning on the neutral. Any difference means current is leaking somewhere else. So, possibly through you or water, and it trips.
UL 943 specifies that a GFCI must trip when the imbalance is between 4 mA and 6 mA.
Such leaks are called faults and may result from real insulation faults. They may also not be actual faults, but rather small capacitive or inductive leakage inherent to normal devices.
If a single device is connected, it can leak up to about 4 mA before tripping. Large transformers (think MH ballasts), capacitors (LED power supplies), or motors (pumps) with poor power factor can cause brief transient imbalances when energized.
When many aquarium devices share one GFCI, each may leak a small amount (often < 1 mA) of current. This is normal and expected. However putting say, ten devices on a single GFCI might total 3.5 mA of normal leakage. A short inrush event adding just 2 mA raises the total to 5.5 mA, exceeding the trip threshold.
So why did the UV in question trip the GFCI?
Possible reasons:
Older (and even new) UVs are typically driven by core and coil ballasts. The very nature of these ballasts produce leakage current by way of capacitance and inductance, especially on startup as the core and coil saturate with energy.
Several things could have been occurring.
1 - The UV had actual leakage current that exceeded the trip threshold or pushed it right to the edge.
2 - Other devices on the circuit contributed to the total leakage and from time to time pushed past the threshold.
3 - Either of the above with an older generation device that was more prone to nuisance trips.
These conditions are common with inductive loads and do not necessarily indicate a defective UV unit.
I know that the post was long winded, but the bottom line is that current generation GFCIs are much more reliable from both a safety and nuisance tripping standpoints. There really is no reason not use them. Personal safety is paramount and there is far less risk to your aquarium than there would have been even a decade ago.
www.reef2reef.com
