China 380 LL Sys & Receptacles 3

[BlackJackJacques]

I have an application in China where they are running 380 V 3-Phase for motors, etc and for the utilization receptacles, they are coming directly off the line phase to neutral for 220V without using a lighting transformer. The receps serve the standard items from coffee pots, office equipment, etc. I have a number of concerns, including short-circuit current performance since the supply is either through a 2 MVA 380V transformer or a 2 MVA emergency diesel generator, the utility transformer and EDG are relatively close to the switchboard and both are fed via low impedance bus-duct. The facility is a light commercial office setting and the motor loading is mostly all HVAC. My concern is that the branch OCP devices are all 10 kA AIC like here in the US. Has anyone seen such an installation?

 

[Mbrooke]

What is the short circuit current at the 10kaic devices? Earth fault loop impedance calcs should give you a good idea.

 

[Bill West]

2 MVA 380V is 3000A load current, so no way the SC current would "only" be 10 kA.
A 200A branch in 4/0 wire needs to approach 500ft in length for its resistance to get the current down to 10kA (I'm assuming the grounding system would be a much lower resistance return than the phase wire). So I think the question depends on the site layout. A sticky point I see is ensuring that no present or future branch panels can be placed too close to the principal switchboard.

Bill

 

[waross]

What is the practice for fault protection of receptacles in the UK?


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

 

[Mbrooke]

@Waross: Typically either a 32 amp ring done in 2.5mm2 or a 20amp radial done in 2.5mm2. Because of the ring system the plugs have fuses inside them up to 13 amps.

 

[waross]

Hi Mbrooke. I was wondering how they addressed protected against the available fault current when a 220/380 system is fed from a large transformer.
From the sound of this system IEC standards may be of more use than North American standards.

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

 

[Mbrooke]

Typically they rely on the impedance of the conductor itself going to the OCPD in question. For example, while in many parts of the world a 250kva pole pig or even 800kva+ ground mount transformer may feed hundreds of homes with a dozen businesses mixed in- the overhead LV run and service drop usually drops the short circuit current down to an acceptable level ie 6ka.

Further, most loop impedance testers are capable of giving a Ze along with a perspective fault current value. Once these values are known you can select your main OCPD.

For OCPD devices downstream, you take your listed conductor impedance, multiply by the length to get your R1+R2 and then add that to your ZE to get you ZS or max fault current that will be seen on the downstream OCPD.

The AC impedances for various common wiring methods will be listed in the back of BS7671.

The goal is 1) to select an OCPD that can interrupt the maximum anticipated fault current. 2) Make sure the short circuit current is high enough to meet the code listed disconnection time requirements for the circuit.

https://carelabz.com/about-earth-fault-loop-impedence-test/

 

[BlackJackJacques]

Thanks gang. A quick calc shows the SC current to the supply side of the 10kA branch circuit breakers to be less than 10 kA, despite the bus duct feeds. All the large motors (all HVAC) are fed through VFDs so there is no motor SC contribution. This entire project has been a cultural learning experience for me as the engineers performing the work are Chinese, and the way they handle interface with the Chinese AHJs is not at all like how we do here - where higher rank is given to older members and "losing face" is dreaded. In any case, compliance to short-circuit concerns appear OK. Thanks again!

Jim

 

[Mbrooke]

That and if it make you feel better most short circuits are not bolted faults but more sputtering meaning they have impedance which in of itself knocks down fault current.

 

[BlackJackJacques]

Thank you and concur. Might be my laziness, but just used to figuring bolted as worst case and putting it to bed. But yes, that does make me feel better.

Thanks so much!

 

[BlackJackJacques]

To Waross re: standards. Their Code is the GB code and is nearly identical to our NFPA 70, at least I determined by invoking NFPA70 provisions and they concur to be the same as theirs, and not by translation. However, they do invoke IEC standards. Their approach, however, is "nuts" when it comes to certain items. For instance, they think nothing of using multiple feeds from different transformers to service one customer such that end equipment can colocate with one another. They also are generous in the use of many ABT devices (likely because of the multiple feeds) and many swbd modules, in lieu of just one single centralized swbd). I understand that in the region of this project there is a power shortage and the "landlord" (what they call their GSA) will "lose face" if they request more power from the Power Bureau (as they miscalculated to begin with). Our bureaucracy is nothing compared to theirs. It is interesting to see the issues and conflicts for a very modern country emerging out of the Communist model.

 

[Mbrooke]

You should figure bolted- if someone miswires something it will be a bolted fault.