corrosion and voltage on neutral

[bitcoin miner]

My problem isn't happening with all my panels. The majority of the corroded panels is on the north side of the building but the problem started with a couple on the south side. It appears isolated to maybe 15 of the 90 panels we have. I have 5 switchgear inside the building and all but one have panels affected.
Wire from switchgear to panels was supposed to go in cable trays overhead until they realized that the building structure couldn't support the trays. They cut the concrete and placed underground conduits and dumped concrete right on top of the conduits. That concrete is sinking and borescope pictures show a couple of the conduits collapsed and full of water. The one switchgear that doesn't have any problem panels is the one that hasn't sunk yet. That may be completely unrelated to the water intrusion but it's all I can reason out.

 

[ptonsparky]

My problem isn't happening with all my panels. The majority of the corroded panels is on the north side of the building but the problem started with a couple on the south side. It appears isolated to maybe 15 of the 90 panels we have. I have 5 switchgear inside the building and all but one have panels affected.
Wire from switchgear to panels was supposed to go in cable trays overhead until they realized that the building structure couldn't support the trays. They cut the concrete and placed underground conduits and dumped concrete right on top of the conduits. That concrete is sinking and borescope pictures show a couple of the conduits collapsed and full of water. The one switchgear that doesn't have any problem panels is the one that hasn't sunk yet. That may be completely unrelated to the water intrusion but it's all I can reason out.

I think you have your answer.

 

[ptonsparky]

That is going to be an expensive repair.

 

[Fred B]

Haven't seen that level of corrosion outside of a "normally corrosive environment", pools and sea side
Is there any dehumidification on the site?

Wire from switchgear to panels was supposed to go in cable trays overhead until they realized that the building structure couldn't support the trays. They cut the concrete and placed underground conduits and dumped concrete right on top of the conduits. That concrete is sinking and borescope pictures show a couple of the conduits collapsed and full of water. The one switchgear that doesn't have any problem panels is the one that hasn't sunk yet. That may be completely unrelated to the water intrusion but it's all I can reason out.

Is the conduit used in the the slab, but sounds like under the slab, rated for use underground?
Sounds like you have a huge amount of subsurface water. Would be recommending eliminating subsurface installation re-routing. Also recommend some perimeter drainage around the whole building. This is likely the source of excessive moisture within the building and the conduit entries are bringing in the humidity. Temperature fluctuations will enhance the moisture being "pulled" into the building, along with intentional use of atmospheric ventilation (cooling) .

 

[MattG0311]

I could try setting the panels horizontally instead of vertically with the outlets facing up. That could be done with minimal cost using existing equipment.

408.43

 

[Fred B]

I could try setting the panels horizontally instead of vertically with the outlets facing up.

That still wouldn't change anything with the issue causing the problem.

 

[synchro]

My problem isn't happening with all my panels. The majority of the corroded panels is on the north side of the building but the problem started with a couple on the south side. It appears isolated to maybe 15 of the 90 panels we have. I have 5 switchgear inside the building and all but one have panels affected.
Wire from switchgear to panels was supposed to go in cable trays overhead until they realized that the building structure couldn't support the trays. They cut the concrete and placed underground conduits and dumped concrete right on top of the conduits. That concrete is sinking and borescope pictures show a couple of the conduits collapsed and full of water. The one switchgear that doesn't have any problem panels is the one that hasn't sunk yet. That may be completely unrelated to the water intrusion but it's all I can reason out.

Conduits that are partially filled with water might actually be the worst because conductor heating could significantly raise the absolute humidity of any air flowing thru the conduits. The temperature of the interior surfaces in the panels could be below the dew point of this humid air, which will then cause moisture to condense on these surfaces.

Perhaps you could measure the temperature of the panels themselves on their outside surface. Maybe the cooling system of the building is keeping the temperature of the panels at levels that may fall below the dew point of the air inside of them, at least on some occasions. So you might want to avoid having cooled air from the HVAC system directly impinge on the panels. If the temperature of the panel enclosure is relatively cool, you might even consider putting some insulation around it, at least on an incremental basis while monitoring its effect. Preferably, a temperature probe would be placed inside of the panel to make sure you don't exceed a 40°C air temperature, or if so, account for any derating that would be required.

Of course, the proper longer term solution is to correct the source of the excess humidity as Tom and Fred have indicated above.

 

[bitcoin miner]

Haven't seen that level of corrosion outside of a "normally corrosive environment", pools and sea side
Is there any dehumidification on the site?


Is the conduit used in the the slab, but sounds like under the slab, rated for use underground?
Sounds like you have a huge amount of subsurface water. Would be recommending eliminating subsurface installation re-routing. Also recommend some perimeter drainage around the whole building. This is likely the source of excessive moisture within the building and the conduit entries are bringing in the humidity. Temperature fluctuations will enhance the moisture being "pulled" into the building, along with intentional use of atmospheric ventilation (cooling) .

This site presents a number of challenges. I am used to switchgear being run at about 80% capacity, not here. We typically run our equipment close to 100%. More power=more hash rate=more money. Makes my job tough trying to keep things running knowing that everything can crap out at any time.
My biggest issue is heat. Miners produce a lot of heat which we direct to the "hot" aisle(actually a nice place to be this time of year). I can redirect the heat to help solve the moisture issue but only while it is cold outside. It's not terribly humid here(I'm born and raised in New Orleans so nothing is as humid as there). But any humidity in the air or created by hot wires steaming the water in the conduits hitting the cold metal will create problems but like I mentioned I only have a few panels that are real problems.
We have a similar building 50 yards away from this one but the wires are run overhead in cable trays and they are connected to PDUs instead of these panels. No issues with those except heat related, in fact I will be finishing the installation of water walls in that building to evaporatively cool the air before in gets to the machines. A trial of that last summer showed a 15 degree temperature drop with no issues from the added humidity. I wish there was money in the budget to convert this building to overhead but until it hurts them in the pocket I'll have to keep fighting the corrosion.

 

[bitcoin miner]

Conduits that are partially filled with water might actually be the worst because conductor heating could significantly raise the absolute humidity of any air flowing thru the conduits. The temperature of the interior surfaces in the panels could be below the dew point of this humid air, which will then cause moisture to condense on these surfaces.

Perhaps you could measure the temperature of the panels themselves on their outside surface. Maybe the cooling system of the building is keeping the temperature of the panels at levels that may fall below the dew point of the air inside of them, at least on some occasions. So you might want to avoid having cooled air from the HVAC system directly impinge on the panels. If the temperature of the panel enclosure is relatively cool, you might even consider putting some insulation around it, at least on an incremental basis while monitoring its effect. Preferably, a temperature probe would be placed inside of the panel to make sure you don't exceed a 40°C air temperature, or if so, account for any derating that would be required.

Of course, the proper longer term solution is to correct the source of the excess humidity as Tom and Fred have indicated above.

This building is only cooled by the filter walls, so basically open air except shaded and protected from rain.

 

[bitcoin miner]

That is going to be an expensive repair.

Yes it will but that's usually what happens when you compromise on the design at the beginning. I would've thought adding structure to hold the cable trays would've been cheaper than cutting and busting out concrete but I'm not an engineer. It is definitely cheaper than fixing it twice though.

 

[bitcoin miner]

408.43

That answers that!!

 

[bitcoin miner]

Is it possible that the steam is coming through the wires instead of around them? The foam would stop liquid water from entering but if it's coming through the wires then the only solution is to eliminate the water source. I know up on poles the wires usually face down to prevent capillary action from drawing water in. If my wires have a nick in the insulation I could be sucking water in. That would explain why only a few are having issues.
Would a megger show a nick in the wire? Would I need to have one lead on the wire and the other in the water in the conduit?

 

[zbang]

Just from a quick read and the photos, you gotta abandon the buried conduits and replace some of those panels! AFAICT there is no other solution.

 

[ramsy]

It appears isolated to maybe 15 of the 90 panels we have. I have 5 switchgear inside the building and all but one have panels affected.

If raceway drains and seal offs fail to fix moisture intrusion damaging equipment, much less fix sinking foundations damaging raceways, the electrician is only responsible to report uncorrectable listing violations per 110.12(B), unserviceable safety hazards to personnel that violates OSHA, and shut down each PDU pending permanent correction, or documented Mgmt. orders to the contrary.

An AHJ Inspector may red tag affected PDU's, switch gear, or condemn the building, especially with sinking foundations not disclosed on TDS, or commercial lease agreements, allowing your company a way out, and to recover losses.

 

[bitcoin miner]

Just from a quick read and the photos, you gotta abandon the buried conduits and replace some of those panels! AFAICT there is no other solution.

I agree but my "experts" say I am just using the wrong foam. They want me to try the Polywater purple foam. At $85/can it better do something but my guess it won't do anything.

 

[bitcoin miner]

If raceway drains and seal offs fail to fix moisture intrusion damaging equipment, much less fix sinking foundations damaging raceways, the electrician is only responsible to report uncorrectable listing violations per 110.12(B), unserviceable safety hazards to personnel that violates OSHA, and shut down each PDU pending permanent correction, or documented Mgmt. orders to the contrary.

An AHJ Inspector may red tag affected PDU's, switch gear, or condemn the building, especially with sinking foundations not disclosed on TDS, or commercial lease agreements, allowing your company a way out, and to recover losses.

I am waiting on OSHA coming in and shutting us down. They own the property and built the building so it's on them. The proper solution is to tear it out and do it right.

 

[Fred B]

I agree but my "experts" say I am just using the wrong foam. They want me to try the Polywater purple foam. At $85/can it better do something but my guess it won't do anything.

If underground conduit is compromised, Foam is not going to fix it. Also the compromised conduit and collapsing concrete above it will eventually compromise the wiring making for a potential fault condition that can be dangerous let alone inconvenient. Scheduled outage is far better and less expensive than an emergency situation.

 

[ramsy]

They own the property and built the building so it's on them.

Wow, how naive of me to observe residential remodels as a construction-industry train wreck, without looking at commercial & industrial construction.

 

[MattG0311]

That answers that!!

Just giving a notable reference.

That said, I work in a facility that has its own “eNgiNEERing SOOPERvisOn” and effectively act as their own AHJ. They do what they want. I do what I can to stay gainfully employed.
It’s fun.

 

[ramsy]

I am waiting on OSHA coming in and shutting us down.

OSHA may accept watertight PDU's with plug boot & connectors.

Leviton 6031 Weather Resistant Plug Boot 20 Amp & 30 Amp : Plugs - $12.64 EMI Supply, Inc

EMI Supply, Inc : Leviton 6031 Weather Resistant Plug Boot 20 Amp & 30 Amp - Boot for Locking Plug, 20 Amp & 30 Amp, 3-Wire, Weather-Resistant - BLACK...

www.emisupply.com

Leviton 2323-B Industrial Grade Connector 20 Amp, 250 Volt : Connectors - $28.05 EMI Supply, Inc

EMI Supply, Inc : Leviton 2323-B Industrial Grade Connector 20 Amp, 250 Volt - 20 Amp, 250 Volt, NEMA L6-20R, 2P, 3W, Locking Connector, Industrial Grade, Grounding - BLACK...

www.emisupply.com