voltage sag location in circuit capture

[farmape]

Consider a typical 480V 3P distribution system. A loose / high impedence connection is causing momentary voltage sags during the intermittant arcing. (Not a line to ground or phase to phase fault) I know these voltage sags are detectable downstream of the loose connection, however, are they present upstream of the arcing? Could they be detected on the upstream bus of the loose connection causing the voltage sags?
Thank you.

 

[rbulsara]

Not withstanding your understanding of the voltage sag ( which is different than voltage drop) the impedance change not only will drop the voltage but also change the current and power.

There will be a change in the current upstream. Depending on the how much the current change, the voltage may change. Voltage change will not be same as what the load sees, as there is no "voltage drop" upstream per say.

By the way the hypothetical condition you describe will not last long and the loads will not operate if they don't get adequate power supply. Arcing generates intense heat that usually turns in to insulation failure and much more, pretty quickly. Loose connection without arcing is bad enough.

Do you have a real issue at hand?

Rafiq Bulsara

http://www.srengineersct.com

 

[farmape]

Thanks for the reply. Typically, I see fractional cycle sags at 60-85 % of nominal on one or two phases prior to a catastrophic failure. A specific example would be a 480V draw out MCC bucket "finger" or "stab" starting to fail. This is often accompanied with overheating, pitting and arcing, and eventual complete failure. Downstream of this connection the voltage sags are detected, and you're right, depending on the size of the service, time is limited. That being said, typically a few days is available before catastrophic failure. What I am really after is if there is research or published papers that discuss detection method on the common bus (MCC, bus duct, switchgear, panelboard, whatever the case may be), of these events, or, if the signature is too unique.

 

[rbulsara]

I can't answer our specific question on research or the paper. But the type of loose connections you are referring to do not happen overnight and are easily detectable via thermography (IR scan), which needs to be done at regular intervals and especially after any repair or modification work.

If the load itself is not affected (which should be most obvious alert), I doubt you could detect much upstream.

It would help if you can throw more light on your background and what you are after, someone else may be able provide more useful information.

Rafiq Bulsara

http://www.srengineersct.com

 

[rbulsara]

If there is indeed an actual arc, which means its is too late for the load, the optical sensors used for arc flash detection could provide you some clues. None of those are perfect.

Rafiq Bulsara

http://www.srengineersct.com

 

[waross]

The refrigeration industry may be able to help. They have some economical monitors available with adjustable alarm/shutdown points and event monitoring.
Typical features may be;
Voltage monitoring.
Shutdown on low voltage (adjustable set point)
Shutdown on voltage unbalance (adjustable set point)
Time delay after a power failure.
Contact monitoring and shutdown on the event of an open or high resistance contact.
A 25 event memory.
Phase loss protection.
There are several manufacturers to choose from and variations in features.


Bill
--------------------
"Why not the best?"
Jimmy Carter