Shortcircuit scenario and need of a crowbar device

[xj25]

Hi Guys,

We have an application in which there are a 750Vdc and a 24Vdc circuit connected by a normally open sectioning switch, and protected by a quick acting fuse (Protistor type).

The 24Vdc circuit is always energized, supplied from a 260Ah NiCd battery.

The 750Vdc circuit is not energized when the sectioning switch closes. There are electrical interlocks that prevent that.
Even so, in case of sectioning switch combined fault, it could happen that 750Vdc reaches 24Vdc circuit and I want to be as sure as possible that the fuse will blown up clearing the fault.

Shortcircuit current will be (750Vdc-24Vdc) / Shortcircuit impedance. I work with the requirement (self imposed) to be at least about 6kA to blown the fuse in <100ms. So about 121mOhms.

The main question is how the battery+paralel loads behave in this catastrofic scenario:
1- will the battery work as an efective low impedance path for the shortcircuit current? could fail in "open circuit" mode?
2- the 24vdc side loads are fundamentally 20% passive and 80% active power supplies. I don´t see very clear how to use them to stimate the "shortcircuit current share" that they get, so prefer not to assume nothing about them. (we assume that the "health" of the 24Vdc equipment after the fault will as better not good)

In case I can not trust in the battery+loads to cause a large enough shortcircuit current I have thought on using a crowbar device. The usual single component devices that I found are for small currents and short pulse transients. Anybody knows a supplier for this kind of application?
- Crowbar device:

​

750Vdc, must support a >=6kA shortcircuit >=100ms . Single use (it is no needed to be reseteable).

Thanks and regards!

 

[xj25]

Not any comment? Thanks anyway if reading up to here.

 

[ScottyUK]

Most crowbar devices are thyristor-based. Thyristors are commonly available up to about 4kV in blocking voltage and a few kA in forward current, with a non-repetitive peak of perhaps 30kA - 40kA for 10ms. These larger thyristors typically require a gate pulse of a couple of hundred mA for at least 100[&mu;]s. I would think most of the cookbook crowbar circuits would work, although you might need to increase the gate drive current for the larger device.

You'll probably find one of the larger stud-type devices will suffice to clear a protistor fuse up to a couple of hundred amps, after that you will be looking at hockey puck types. Have a look at Westcode and Powerex for possible thyristors.

 

[xj25]

Thanks Scotty, I would love to try that, but just for type testing it would be prohibitive for our budget.

Actually we have checked with the battery supplier, and it seems that the low impedance of the battery will behave practically as a short, so I think that we will live with that possibility and battery will do the function to blow the fuse up in case of fault.

Regards.