selection and erection (large installs 50ka+ psc)

[Lee321]

The thread that KME posted about his hotel with own transformer and sub main etc has got me on a few things regarding ka ratings of cpd's.

so my issue is how do we select cpd's when dealing with very high psc's due to local supply transformer etc

Inparticular the main mccb or switch ka rating then determining the ka rating needed of down stream cpd's

cpd=circuit protective devices

 

[Chr!s]

I suggest if their was a transformer as in kmes post then you may be better of calculating rather than measuring, reason being is that on large supply's we need to take into account the resistance and the reactance of the circuit which standard meters wont be able to measure accurately due to the large currents required to produce the reactance. Taking both into account this may reduce the pfc or pscc somewhat.

 

[Chr!s]

If the case is that the pfc/pscc is high then we then need to see what the energy let through values are for the protective device (A2S), with this value we can now use the adiabatic to see if the cpc is protected and to see if the live conductors are protected.

On large fault currents fuses will operate on the first half of the sine, most fuses and circuit breakers have good energy limiting properties, so although the pfc/pscc may be high the actual current that the fuse lets through will be greatly limited.

regards chris

 

[Sidewinder]

With pscc that large, you are as you suggest going to have an onsite HV/LV trasnformer.

The supplier of this, probably DNO will have to provide their quoted value of PSCC, they may have to calculate this for you from transformer manuf' data and their network design values.

Else they will quote you a default maximum value they will not exceed.

They have to do this under ESQCR foc.

Typically the front end devices would be MCCB's typical breaking capacities would be 25 or 35 kA for such items.

At the origin you would probably have an ACB, like the one in KME's example, that one can break 65kA as I understand it, I have not checked the abbreviations mind, just how I would read it from memory.

Thus your front line distribution is protected by the ACB, so they could be 35kA MCCB's, then downstream from there you should calculate the possible fault currents to see if you can gradually downsize to 25kA devices etc.

I would NOT design to fit a 10kA outgoing device even well downstream from the ACB unless I had a higher rating breaking capacity between the 2.

At these fault levels you probably need an MCCB at the incomer for each board, to offer siutable protection.

Often on such systems you have bus bar risers with REALLY negligible impedance thus losses, why they are preferred. This gives high values of pscc well downstream of the origin cpd.

You can do energy let through and discrimination calculations and modelling based on manufacturers data in some software packages.

I don't have one!

This can be quite involved and complex (i's & j's depending on your schooling!).

This sort of design calc is reallistically beyond most of us to get the fault levels, discrimination and cascading correct to ensure that there is minimum distruption maximum re-usability in the event of a fault and the minimum risk of collateral damage again in the event of a fault.

The thing is once the system is modified for the first time by the local handyman, or by a well meaning spark who is not given the full level of information required, it throws all the original design calcs out the window!

 

[Lee321]

would you say it was beyond the realms of a tester to ensure compatability?

I would say at this level the panel board would/should be bespoke and designed correctly at manufacture with the transformer details being provided to the panel builder from the outset. i read that control panels since 2005 should have sccr and ka capability name plates but not sure if this applies to distribution panel boards or just motor control etc?

i would just like to know that if i came across a large dis panel and had to install a sub main from it to a new db etc i had the know how to ensure i meet Ka ratings downstream

 

[green-hornet]

Excellent post

 

[Sidewinder]

would you say it was beyond the realms of a tester to ensure compatability?

lostit,

That would depend on the gear you carry and your competence with such large supplies.

Would you have a ductor tester and be able to isolate the bus bars to check the DC resistance?

It is not possible to measure the response to such high fault current levels with any "normal" test gear carried by "normal" sparks.

Most fault response verifications on such systems are by calculation.

Try to make friends with some of the engineers at your local DNO.

I'm lucky, I have a regional office local to me for my DNO and they guys are excellent.

I even have a mobile no. or 2 written down somewhere.

I beleive that you can learn a lot if you get them out for a site assessment and ask the right questions, show respect, and an interest in what they do and, why they do it.

I had an enquiry on a 1600A incomer 11kV sub 5m from consumers origin, tails were double insulated in trench, no armour anywhere.

Could not sort out the earthing, could have been TN-C-S or TN-S.

Spoke to local DNO engineer, he told me if I get the job he'll get the senior linesman over to go in the sub with me to see how they have installed it as it will be quicker and easier than looking up the job records, the building O&M manuals have gone walkies!

Think I've gone a little OT here!

:coat

 

[green-hornet]

lostit,That would depend on the gear you carry and your competence with such large supplies.

Would you have a ductor tester and be able to isolate the bus bars to check the DC resistance?

It is not possible to measure the response to such high fault current levels with any "normal" test gear carried by "normal" sparks.

Most fault response verifications on such systems are by calculation.

Try to make friends with some of the engineers at your local DNO.

I'm lucky, I have a regional office local to me for my DNO and they guys are excellent.

I even have a mobile no. or 2 written down somewhere.

I beleive that you can learn a lot if you get them out for a site assessment and ask the right questions, show respect, and an interest in what they do and, why they do it.

I had an enquiry on a 1600A incomer 11kV sub 5m from consumers origin, tails were double insulated in trench, no armour anywhere.

Could not sort out the earthing, could have been TN-C-S or TN-S.

Spoke to local DNO engineer, he told me if I get the job he'll get the senior linesman over to go in the sub with me to see how they have installed it as it will be quicker and easier than looking up the job records, the building O&M manuals have gone walkies!

Think I've gone a little OT here!

:coat

Very true I have some "mates" at my local dno and to be honest I do not think I could do half my work without what I have learned from them and what they do for me.

The excellent post remark was for sidewinders previous post sorry lostit

 

[Lee321]

i was hedging my bets on a simple evalution as in the tranny has nameplate with ratings and the switchgear i could reference there ratings so no testing at that stage?

so not a good thread than Gh O)

 

[Sidewinder]

Lee,

No this is a good thread!

The problem is when you get to these sorts of installs, the inductive and capacitive parts of the load can be significant.

You are so close to the transformer, BIG inductance, you have some big cables, significant capacitance, which the transformer secondary may also have.

You honestly have no idea of the load profile or the load impedance, nor the source impedance on the HV side of the tranny.

You can't, this is not a criticism, no one except the DNO & the original client design engineer will unless you have access to this data.

These sorts of supplies are serious s**t!

 

[Sidewinder]

Please remember most of these installs will have been designed with the "old" voltage levels and that there will be an affect on the pscc and other network characteristics will have an affect also.

 

[Chr!s]

Well if you use the above equation this will give u the Pscc at the terminals of the transformer.

Then if we equate the impedance of the condutors to the db we can arrive at the level of pscc at the db

regards chris

 

[Sidewinder]

Chris, I meant the DNO side, as I understand it they build in fault current limiting by design into their HV & LV sides

 

[green-hornet]

i was hedging my bets on a simple evalution as in the tranny has nameplate with ratings and the switchgear i could reference there ratings so no testing at that stage?so not a good thread than Gh O)

Its a very good thread, and some very interesting and informative replies.

You can calculate, however you need to have all the data and as sidewinder says most of this data is not available to most people and even if it was they would mean very little unless you know what your looking at.

A lot of the design is done DNO side and the people who do this are well versed in electrical engineering.

My uncle was before he retired one of these people, I remember as a child he had a terrible accident caused by a HV flash discharge which gave him 3rd degree burns to most of his body. He fully recovered and then emigrated to South Africa where he worked on large installations for the Gold mines.

The added risk of explosion meant that all the power had to be designed to the absolute ohm.

Any alterations to any of the installation would mean a full redesign because there was no tollerance built into any of the designs, for absolute safety reasons.

 

[kme]

Excellent posts ( & indeed thread guys).

I`ll be the first to admit that, although I am aware of the theory being postulated by Sidewinder & plumber, I don`t get much call to use it, so it becomes sort of rusty & unused..........to the point where I couldn`t have provided the descriptions above

Generally, it isn`t something you`ll generally have to be aware of ( unless you intend to go into installation design?)

With hindsight, a very low reading ohm-meter would have been an advantage:

for instance; a 200A submain (SWA/XLPE) was given a max Zs of 0.03 ohms. My reading at the DB on that sub was 0.05 - a definite fail. However, was that down to errors on the croc clip/ probe connections, or is the resistance of that cable that high?

Problem is, in that situation, you tend to calc r1+r2 from Zs-Ze; as the sub cannot be de-energised. The only test you can do is Zs. You can`t faff about cleaning the contact area etc, cos it is still on supply! (FYI - code 3 - further investigation )

KME