Current on a 34.5kv Concentric Neutral

[Desk Jockey]

I am attempting to close out a PV farm and have encountered a problem that I am unable to solve. I have a high circulating current at the concentric neutrals on all Photovoltaic and Battery Storage MV circuits at the (2) substations we have on site. Our 3rd party testing contractor was on site and recorded values from 30A to 100A at 60Hz. The current is not constant, it will hold for a short period of time and then disappear, “almost like a pulse” according to the NETA tech. Some MV circuits had current on all (3) phases and others only on individual phases, the specific individual phase with current varied from circuit to circuit. (5) MV junction boxes were tested and produced the same results, 19A to 85A. Measurements were also taken at (1) inverter but the results were less than 1A on each phase.

Approximately (3) weeks ago, an MV junction box suffered a fault at the C phase deadbreak connector, we believe it to be due to a loose connection. During the circuit inspection, it was discovered that multiple Cold Shrink tubes located at the base of the deadbeak connector in multiple MV junction box locations had failed. The Cold Shrink had either split up the side and hung from the cable or completely fell off. My concern is that this is related to heat created by the circulating current found on the concentric neutral and the potential to damage other terminations. Any ideas?

 

[ron]

Are any of the 480V inverters connected as 4 wire? What kind of transformer is being used, delta-wye with the delta on the 480V side?
Is the concentric neutral really a neutral connected to the X0 of a wye? Is the wye bonded?

Tell use more of the details.

 

[tcleghorn]

I assume all step-up transformers in the PV and battery storage systems are wye-wye configuration to meet the utility's effective grounding requirements. HV side is wye-grounded. LV side is wye-grounded.

I also assume the utility's substation transformer is wye-grounded on the distribution side.

The substation transformer concentric neutral will be tied to the center-point of each distribution transformers wye center point. This will create an opportunity for circulating ground currents if phase currents are unbalanced including on an instantaneous basis.

I recommend checking phase and ground currents with a PQ meter. Check for harmonic content at each order, especially triplen currents. Triplen currents will add on a neutral and wye-wye's will pass them thru.

Good luck.

 

[synchro]

I have a high circulating current at the concentric neutrals on all Photovoltaic and Battery Storage MV circuits at the (2) substations we have on site. Our 3rd party testing contractor was on site and recorded values from 30A to 100A at 60Hz. ... Some MV circuits had current on all (3) phases and others only on individual phases, the specific individual phase with current varied from circuit to circuit. (5) MV junction boxes were tested and produced the same results, 19A to 85A.

If the concentric neutrals (CN's) in a MV circuit are joined on one end of a run but are open on the other end, voltages will be induced at the open ends by the magnetic fields from the phase conductor currents, just like on a wye secondary of a transformer. But if the CN's on the opposite end of the cables are joined together, circulating currents will flow between them in proportion to the open circuit voltages, and inversely with their associated impedances. And so even relatively short runs may have non-negligible circulating currents, because even though the induced voltages would be lower so are their impedances.

In the design stage, one way to reduce such circulating currents is to break the cable run into 3 segments of equal length using 2 junction boxes in between. Inside the junction boxes the CN's of each cable are rotated between associated phase conductors, but the phase conductors themselves are not. By doing this, the phase of the induced voltage along each CN segment is incremented 120°. And so when these CN segments are connected in series, the voltages add vectorially as in a delta and therefore get back to the same voltage as at the beginning of the run. The net induced voltage across each CN run is therefore significantly reduced. As a result, the circulating currents when the CN's are joined at both ends of the cables are also substantially reduced.

This "cross-bonding" method is included on this webpage:
https://elek.com/articles/sheath-bonding-design-guide-for-hv-cables/

Bonding the CN's on one end ("single point" bonding) might be an option in some cases, as noted in the article.

As others have mentioned, more details from the OP are needed to get better feedback about what's happening, and any options there might be to make it better.