Surge protection question

[Jay Ashworth]

What would you think if you looked at the panel schedule for a PAC, and all the panels had local, breaker protected surge protection devices on the side *except the one that fed the Sensor rack*?

Would that have something to do with dimmers and chokes, as we're discussing in a separate thread this week? Or is it just an oversight?

 

[JonCarter]

Jay, It would seem to me that the chokes in a dimmer rack would at least partially stop a surge from getting to the instruments connected to the dimmers, but would do nothing for any electronics in the rack itself. In order to protect the rack's electronics the surge protection would need to be on the power input o the electronic equipment in the rack. Protection for the entire rack would be nice, too..

 

[MNicolai]

Lack of an SPD at a specific panelboard has nothing to do with dimming. Possibly a coordination oversight, possibly outside of the scope of the lightning and surge protection system design (code requires SPD's in specific circumstances but not at every panelboard, so there could be a method to the madness depending on how the overall electrical system design was planned).

 

[FMEng]

It seems like an oversight, especially in a lightning prone area. The only way to protect the control electronics is to clamp over voltage ahead of the dimmer rack.

 

[jtweigandt]

I would wonder if the weird wonderful harmonics of the dimmers might not give some sort of inrush spike that is interpreted as a surge.
Are those like a trippable breaker, or just a passive surge protector? I have had a variety of equipment that would trip a household grade GFI recepticle.

 

[FMEng]

Most surge protection is a MOV, which acts as a brute force clamp with a fairly high conduction voltage, along with some circuitry to drive indicator lights. The dimmer load should not affect the operation at all. GFCIs are a completely different sort of animal.

 

[TimMc]

I would wonder if the weird wonderful harmonics of the dimmers might not give some sort of inrush spike that is interpreted as a surge.
Are those like a trippable breaker, or just a passive surge protector? I have had a variety of equipment that would trip a household grade GFI recepticle.

I wish I'd saved my little dog and pony GFCI ppt deck...

GFCIs/RCDs sense the relative balance of current present in the load side's neutral and line. If there is an imbalance of >5 mA (USA/Canada, UK is higher), the device opens the circuit. It is presumed that this current is leaking via a connected device, appliance, or their power cords/wiring devices and therefore presents a potential hazard to humans. If there is no leaking current these devices will not open; they are not over-current protection devices. Anything that trips a GFCI outlet or breaker needs to be carefully inspected - device, appliance, cable, extension cord, power strip (esp the "surge protection" kind, as a sacrificed MOV shunts current to ground, which will trip the GFCI) - to find what is causing the problem. Note that current leakage is cumulative within the circuit, so a device with 2 mA leak, a cord set with a 2 mA leak will not trip the GFCI, but when you plug in something else with a 2 mA leak, it trips.

Happy hunting!

 

[jtweigandt]

I wish I'd saved my little dog and pony GFCI ppt deck...

GFCIs/RCDs sense the relative balance of current present in the load side's neutral and line. If there is an imbalance of >5 mA (USA/Canada, UK is higher), the device opens the circuit. It is presumed that this current is leaking via a connected device, appliance, or their power cords/wiring devices and therefore presents a potential hazard to humans. If there is no leaking current these devices will not open; they are not over-current protection devices. Anything that trips a GFCI outlet or breaker needs to be carefully inspected - device, appliance, cable, extension cord, power strip (esp the "surge protection" kind, as a sacrificed MOV shunts current to ground, which will trip the GFCI) - to find what is causing the problem. Note that current leakage is cumulative within the circuit, so a device with 2 mA leak, a cord set with a 2 mA leak will not trip the GFCI, but when you plug in something else with a 2 mA leak, it trips.

Happy hunting!

Inductive loads can trip GFCI.. maybe the new GFI's are better, but Motors especially can do this with no true fault. The inductive capacitance will cause a momentary imbalance between load and neutral. (sucking and storing electrons if you will) Maybe the newer GFI's are better (or worse less sensitive) I was just wondering aloud if maybe the surge supressor could be degraded by the inductive shenanigans of the dimmer.. It would experience a "spike" when the dimmer coil is loaded.. and the MOVs will degrade over time when repeatedly cycled. when they do, the inline breaker could get trippy. Just a theory as to why there might not be a surge supressing breaker on the dimmer supply. But then I'm just a Dog Doc with an interest in things electical/electronic, and my conjecture might not be worth a warm bucket of pee.

 

[almorton]

Very little of that makes sense, I'm afraid. In a series circuit, the current in and the current out are the same, unless, as TimMc explained, there's another route involved (a ground fault). An inductor in and of itself won't cause a L-N imbalance, nor will a capacitor (or even a tuned circuit comprised of both), unless it has a connection somewhere else to ground. Over current, yes, quite possibly; ground fault, not unless there's a ground leakage fault.

As to inductive capacitance - I'm not sure what you mean by that. You do get "storage" effects from both inductance and capacitance, but they are caused by different effects.

 

[STEVETERRY]

What would you think if you looked at the panel schedule for a PAC, and all the panels had local, breaker protected surge protection devices on the side *except the one that fed the Sensor rack*?

Would that have something to do with dimmers and chokes, as we're discussing in a separate thread this week? Or is it just an oversight?

That separate thread covered short-circuit fault current ratings--totally different from surge protection.

Might the lack of a surge protector have something to do with the size of the three-phase feed to the Sensor rack? What was the current rating of the other surge-protected circuits?

Another possibility: did the surge-protected panels include emergency circuits, where the Sensor rack did not? See NEC 700.8 below.

700.8 Surge Protection.
A listed SPD shall be installed in or on all emergency systems switchboards and panelboards.


ST

 

[jtweigandt]

Very little of that makes sense, I'm afraid. In a series circuit, the current in and the current out are the same, unless, as TimMc explained, there's another route involved (a ground fault). An inductor in and of itself won't cause a L-N imbalance, nor will a capacitor (or even a tuned circuit comprised of both), unless it has a connection somewhere else to ground. Over current, yes, quite possibly; ground fault, not unless there's a ground leakage fault.

As to inductive capacitance - I'm not sure what you mean by that. You do get "storage" effects from both inductance and capacitance, but they are caused by different effects.

Sorry poor choice of words... I did a review.. as a result.. Confining ourselves to GFI for the moment. GFI can indeed trip due to inductive loads, particularly electric motors. I tried to look into the actual mechanism/circuitry of the GFI but was time limited.. so I still need a better understanding of how.. But in essence both inductive and capacitive components put the voltage and current out of phase with each other. Moving on to spike protection..... switching off an inductive load can indeed cause a large momentary voltage spike. Moving on to spike protection with MOV.. they can degrade over time depending on how severely and how often they are "spiked" https://iaeimagazine.org/2004/march2004/metal-oxide-varistor-degradation/ So even though my initial explanation was severely inadequate, it appears phase imbalance can trip GFI without a "true" current leakage. It also appears that repeated spiking from switching inductive loads can degrade MOV's over time. Am I totally off base here, or just a poor explainer?

 

[almorton]

Hmmm ... I can certainly believe the back emf from an inductive load can affect a varistor, and I know that LCR circuits lead to phase angle between voltage and current, but still the voltages and currents in each leg ought to be the same unless there's some other current path involved (leaky filter capacitor? Poor winding insulation?). Maybe some designs of RCD can be "fooled" by bad power factor?

If you have a RCBO (which is a combined RCD and overcurrent circuit breaker) that can easily be tripped by the inrush current from an inductive (or capacitive, really) load, with no ground fault, as the overcurrent part is tripping. I'm still not convinced that a supply with a big phase angle in and of itself would trip a pure RCD, but if any installation engineers can explain it, I'm all ears.

 

[jtweigandt]

And as to the MOV.. their purpose is to provide high resistance at normal voltage and low resistance in abnormal high voltage. Their degradation/failure trend is to move toward less and less resistance, with final failure being a defacto short. So what's going to happen to either the fuse, or (substitute breaker for fuse) in this diagram as the MOV's trend toward failure... they shunt to ground/neutral and the fuse blows or the breaker repeatedly trips. So back to my original conjecture... maybe someone was worried about this scenario.

 

[FMEng]

The clamp voltage of the MOVs are surprisingly high, which is why they survive for years of use. Consider that surge protectors are designed to be installed on panels with highly inductive loads, like HVAC equipment. The little bit of inductance from a dimmer is nothing compared to other loads that are quite common. I have installed a few, new surge protectors, and I have never seen anything in the instructions that say not to use it in panels with certain loads.

 

[jtweigandt]

The clamp voltage of the MOVs are surprisingly high, which is why they survive for years of use. Consider that surge protectors are designed to be installed on panels with highly inductive loads, like HVAC equipment. The little bit of inductance from a dimmer is nothing compared to other loads that are quite common. I have installed a few, new surge protectors, and I have never seen anything in the instructions that say not to use it in panels with certain loads.

hey hey... never let the truth get in the way of a good theory...

 

[Jay Ashworth]

Might the lack of a surge protector have something to do with the size of the three-phase feed to the Sensor rack? What was the current rating of the other surge-protected circuits?

Every interior terminal panel in the power riser has an SPD hanging off a 3-ph 30A breaker *except* the panel which feeds the Sensor rack and the Paradigm rack. There's one panel out in the yard that doesn't, but I think it's on the 480 side of the plant.

Another possibility: did the surge-protected panels include emergency circuits, where the Sensor rack did not? See NEC 700.8 below.

700.8 Surge Protection.
A listed SPD shall be installed in or on all emergency systems switchboards and panelboards.

That one would take a bit more looking into. I'm not sure what would qualify as an emergency circuit in this building; the Fire Pump is the only obvious candidate, and it's diesel powered.

 

[STEVETERRY]

Every interior terminal panel in the power riser has an SPD hanging off a 3-ph 30A breaker *except* the panel which feeds the Sensor rack and the Paradigm rack. There's one panel out in the yard that doesn't, but I think it's on the 480 side of the plant.


That one would take a bit more looking into. I'm not sure what would qualify as an emergency circuit in this building; the Fire Pump is the only obvious candidate, and it's diesel powered.

Emergency Lighting?

ST

 

[Jay Ashworth]

Hmmm. By "Emergency lighting", I assume you mean "lighting circuits dedicated to (or carrying) power sourced in a way that makes it emergency breakage resistant"...

not "circuits that charge battery powered emergency lights", right?

I didn't look that closely, but I didn't see much if any of the former.

 

[STEVETERRY]

Hmmm. By "Emergency lighting", I assume you mean "lighting circuits dedicated to (or carrying) power sourced in a way that makes it emergency breakage resistant"...

not "circuits that charge battery powered emergency lights", right?

I didn't look that closely, but I didn't see much if any of the former.

I simply mean Emergency circuits as defined by NEC Article 700. That includes a wide variety of equipment and distribution.

ST

 

[Jay Ashworth]

Well, since the mainstage auditorium house lighting comes off that panel, it would seem a good candidate, if that's why the rest of the panels fit. I guess I'll have to go read 700.