Amtech MSDB modelling question

[ibra123]

Hi guys,

With regards to modelling a multi service distribution board, how would you guys model it in Amtech?

I have tried using a TPN distribution board and each circuit on the DB represents one apartment protected by a BS88 80A fuse (within the MSDB).

The error I am facing is "The disconnection time of the upstream overcurrent protection is lower than that for apartment circuits"

In my example I have a 400A cut-out (315A fuses) that feed my main panelboard, which then feeds my MSDBS with micrologics via 160A MCCB and then 80A fuse to each apartment which also has an 80A fuse at the apartment cut-out (henley red head series fuse).

Thanks in advance!

 

[pc1966]

I'm not familiar with Amtech's design software but do they have options for setting the MCCB parameters? Do they offer a current-time plot of the various OCPD overlayed?

Depending on the fault current you could well find the 80A fuse fails to prevent the 160A MCCB for tripping with both going, but to know if that is a realistic risk you need to factor in the PFC / PSCC at the installation.

Finally, the red Henley cutout "fuse" is normally a solid link simply for local isolation when changing meter, etc. They are fitted in situations like this when there is already a 60A/80A fuse in the MSDB, if you are modelling two 80A fuses in series than clearly it will complain about a lack of selectivity!

 

[ibra123]

I'm not familiar with Amtech's design software but do they have options for setting the MCCB parameters? Do they offer a current-time plot of the various OCPD overlayed?

Depending on the fault current you could well find the 80A fuse fails to prevent the 160A MCCB for tripping with both going, but to know if that is a realistic risk you need to factor in the PFC / PSCC at the installation.

Finally, the red Henley cutout "fuse" is normally a solid link simply for local isolation when changing meter, etc. They are fitted in situations like this when there is already a 60A/80A fuse in the MSDB, if you are modelling two 80A fuses in series than clearly it will complain about a lack of selectivity!

See below current-time graph of the protective devices. The red henley cut-out fuse is not modelled in Amtech so just the 80A fuse in the MSDB is used in the calculation.

 

[pc1966]

See below current-time graph of the protective devices. The red henley cut-out fuse is not modelled in Amtech so just the 80A fuse in the MSDB is used in the calculation.

You can see at about 1kA fault current the MCCB will trip as the 80A fuse curve intersects with the MCCB curve. You might be able to fiddle with MCCB short-term delay to around 0.4s and maybe put it in 'I2t' mode to push that crossover point high enough so you have selectivity at the expected PFC/PSCC levels, but I think that needs the fancy electronic trip module.

Does the software model the supply characteristics and cable lengths to know what those current values are expected to be?

MCCB are good in they are adjustable, user-resettable, etc, but for simple selectivity goals having a 160A fuse for the MSDB feed would meet all goals (except non-skilled replacement).

 

[ibra123]

You can see at about 1kA fault current the MCCB will trip as the 80A fuse curve intersects with the MCCB curve. You might be able to fiddle with MCCB short-term delay to around 0.4s and maybe put it in 'I2t' mode to push that crossover point high enough so you have selectivity at the expected PFC/PSCC levels, but I think that needs the fancy electronic trip module.

Does the software model the supply characteristics and cable lengths to know what those current values are expected to be?

MCCB are good in they are adjustable, user-resettable, etc, but for simple selectivity goals having a 160A fuse for the MSDB feed would meet all goals (except non-skilled replacement).

I have adjusted the selective short circuit (Isd) from 10.0 to 8.0 and all the errors have gone, thank you so much:

 

[pc1966]

Do you have a few MSDB so need a MCCB panel to feed them?

If just one then a 160A fused-switch such as Schneider SQB1603K would do for approx same cost as MCCB & enclosure.

If a couple you might consider a bus-bar chamber and those fused-switches off it, but any significant number then MCCB panel is far better solution I suspect, even if you have to go to the fancy trip module.

 

[pc1966]

I have adjusted the selective short circuit (Isd) from 10.0 to 8.0 and all the errors have gone, thank you so much:

Good to hear!

My last post just happened before yours became visible.

 

[ibra123]

Do you have a few MSDB so need a MCCB panel to feed them?

If just one then a 160A fused-switch such as Schneider SQB1603K would do for approx same cost as MCCB & enclosure.

If a couple you might consider a bus-bar chamber and those fused-switches off it, but any significant number then MCCB panel is far better solution I suspect, even if you have to go to the fancy trip module.

I have 2No. MSDBS fed from separate MCCBS from the main panelboard, see attached PDF of the LV schematic that shows this.

 

[pc1966]

I have adjusted the selective short circuit (Isd) from 10.0 to 8.0 and all the errors have gone, thank you so much:

Actually looking at your plot it seems it has just moved the MCCB curve a bit to the right, but the reason it is no longer complaining is it only plots the fuse characteristics down to 0.1s, which is not terribly clever (but not unique). There is still some risk that at 1.2kA or so the MCCB would be tripped.

However, it is unlikely that would happen on final circuit faults as then the MCB should trip below the MCCB's short-term delay, at least up to a few kA and that might well be above the PFC/PSCC at that point.

 

[pc1966]

I have 2No. MSDBS fed from separate MCCBS from the main panelboard, see attached PDF of the LV schematic that shows this.

The diagram show a 125A fuse on the incoming side of the MSDB boards. That seems a bit low for having 4 * 80A flats on each 125A fuse but maybe that is considered acceptable diversity.

Given you already have the 160A MCCB protecting the board you might want to consider putting links in for those fuses.

 

[ibra123]

The diagram show a 125A fuse on the incoming side of the MSDB boards. That seems a bit low for having 4 * 80A flats on each 125A fuse but maybe that is considered acceptable diversity.

Given you already have the 160A MCCB protecting the board you might want to consider putting links in for those fuses.

Those are 3No. 125A fuses (TPN), which will supply 13No. flats. Should I consider dropping from 80A to 63A fuses (even though the calc works with 80A fuses)?

What is the difference between a fuse and a link that you are referring to?

 

[pc1966]

Those are 3No. 125A fuses (TPN), which will supply 13No. flats. Should I consider dropping from 80A to 63A fuses (even though the calc works with 80A fuses)?

Why was 125A chosen, is it a limit of the MSDB or someone's choice?

Dropping to 63A fuses would help of course, and also buy a bit more selectivity with the 160A MCCB, but really there should be a design aspect as to why 63A or 80A was chosen. Who has designed the flats and assessed the expected demands? If they are small and not using multiple electric showers, etc, then 63A might be perfectly adequate.

What is the difference between a fuse and a link that you are referring to?

You can get metal bars, etc, designed to replace a fuse if the protection is not needed, say to turn a fused-switch in to a simple isolator as other OCPD exists. Check with the MSDB manufacturer if they offer one without incoming fuses (or linked out) for a situation like this where the up-stream protection is adequate.

Having the 125A fuse where it is not needed is just another thing to blow if many flats max out demand at the same time and cause problems for restoring it, more so than someone who is able to flip the MCCB closed to reset it.

 

[ibra123]

Why was 125A chosen, is it a limit of the MSDB or someone's choice?

Dropping to 63A fuses would help of course, and also buy a bit more selectivity with the 160A MCCB, but really there should be a design aspect as to why 63A or 80A was chosen. Who has designed the flats and assessed the expected demands? If they are small and not using multiple electric showers, etc, then 63A might be perfectly adequate.


You can get metal bars, etc, designed to replace a fuse if the protection is not needed, say to turn a fused-switch in to a simple isolator as other OCPD exists. Check with the MSDB manufacturer if they offer one without incoming fuses (or linked out) for a situation like this where the up-stream protection is adequate.

Having the 125A fuse where it is not needed is just another thing to blow if many flats max out demand at the same time and cause problems for restoring it, more so than someone who is able to flip the MCCB closed to reset it.

If I'm honest, I selected 125A as the previous design did not show a fuse rating so you might be right that I should use solid links as I am already protecting that cable upstream with an MCCB.

I also selected 80A as the stage 3 design did not say what rating that fuse is. See below attached image of the loads for each apartment. I applied a manual diversity that is based on a ADMD calculation method from an energy supplier. The ADMD load for each apartment is calculated to be 22.87A, so it looks like I should go for 63A fuses.

Do you know if I can feed the EV charging feeder pillar from the MPB instead of the LLPB because it is difficult to discriminate the upstream and downstream MCCBs? Note that the whole current meter is at the LLPB and I have to think about how I would meter the EV feeder pillar.

 

[pc1966]

If I'm honest, I selected 125A as the previous design did not show a fuse rating so you might be right that I should use solid links as I am already protecting that cable upstream with an MCCB.

Those incoming fuses are needed when that sort of panel is used by a DNO and they form the cut-out for protecting the feed cable for overload, as well as a means of isolation.

Here you don't need that as the supply is already protected and an easy means of isolation by the MCCB.

I also selected 80A as the stage 3 design did not say what rating that fuse is. See below attached image of the loads for each apartment. I applied a manual diversity that is based on a ADMD calculation method from an energy supplier. The ADMD load for each apartment is calculated to be 22.87A, so it looks like I should go for 63A fuses.

That makes sense.

Do you know if I can feed the EV charging feeder pillar from the MPB instead of the LLPB because it is difficult to discriminate the upstream and downstream MCCBs? Note that the whole current meter is at the LLPB and I have to think about how I would meter the EV feeder pillar.

Normally MCCB - MCCB gives good selectivity, at least if the settings are sensibly chosen. What sort of limit does the software give for the selectivity limit? If it is above the expected fault currents then it is not really something to worry about.

More fundamentally though I would expect the EV charge points will need RCD protection You can do that with fancy MCCB add-ons, but does the EV pillar have RCBOs, etc, for protection of the charge outlets? If so the RCBO should be selective with the MCCB feeding them so your only real worry would be physical damage to the SWA cable.

 

[ibra123]

Normally MCCB - MCCB gives good selectivity, at least if the settings are sensibly chosen. What sort of limit does the software give for the selectivity limit? If it is above the expected fault currents then it is not really something to worry about.

More fundamentally though I would expect the EV charge points will need RCD protection You can do that with fancy MCCB add-ons, but does the EV pillar have RCBOs, etc, for protection of the charge outlets? If so the RCBO should be selective with the MCCB feeding them so your only real worry would be physical damage to the SWA cable.

With a 160A MCCB upstream and 100A MCCB downstream I have a selectivity limit of 2.4kA and my prospective fault current is 5.069 kA (see below error message). I was thinking of supplying the EV feeder pillar via the MPB so that only 1No. 160A MCCB is used but then I would have metering issues I think. I will have RCBOs for each charging point.

 

[pc1966]

With a 160A MCCB upstream and 100A MCCB downstream I have a selectivity limit of 2.4kA and my prospective fault current is 5.069 kA (see below error message). I was thinking of supplying the EV feeder pillar via the MPB so that only 1No. 160A MCCB is used but then I would have metering issues I think. I will have RCBOs for each charging point.

You might want to look at having one of the fancy electronic trip modules for the upstream 160A MCCB feeding that board.

At least in the Hager catalogue (I think you are Schneider which I'm not so familiar) the h250 LSI offers total selectivity with the TM version in that sort of combination.

Do you get selectivity errors on any other circuits on the landlord's board?

 

[ibra123]

Do you get selectivity errors on any other circuits on the landlord's board?

Yes, anything I put on that board with a 100A MCCB i get errors. At the moment only the EV charger and the Landlords DB, but for the landlords DB I am getting a different error:

 

[pc1966]

Yes, anything I put on that board with a 100A MCCB i get errors.

You might want to ask Schneider support for selectivity guidance. It might be you can change the 160A supply MCCB for a different model (e.g. 250A set to 160A 'In', or maybe going to one with 50kA or 70kA break capacity) or different settings to get enough for your MPB to LLPB requirement of over 5kA selectivity (on the computed PSCC current, more probably it would be L-E fault not 3 phase bolted fault so around half that PFC and OK).

At the moment only the EV charger and the Landlords DB, but for the landlords DB I am getting a different error:

View attachment 109194

For info - the "pre-arcing I2t" is the energy needed to cause the fuse to get to the point of blowing. It might just survive but will be seriously weakened, so you want down-stream protection to limit the fault energy to less than that. Once a fuse blows there is arcing as the wire evaporates, and another rating for a fuse is the total let-through I2t, which is what a fuse that rapidly blows will let-through. When cascading fuses if the downstream let-through is below the up-stream per-arcing all is well. For BS88 fuses in the same general series you typically get that at a 1.6: ratio (i.e. 2 steps on the nominal series, for example, a 63A fuse from 100A, from 160A, from 250A, from 400A, etc).

Its not clear how to match the notation there with your LV schematic, but the 300k A2s figure looks like the pre-arcing for a 250A fuse, what it is telling you is a hard fault on the landlord's board it going to take out fuses in the 250A fused-switch feeding it.

As it is fed from a 160A MCCB you could simply put in an isolator switch for that board, or drop it completely as you can isolate all with the landlord's incomer switch, or isolate the feed cable using the supply MCCB (suitably locked off).

But...your 315A DNO fuses might be at some risk as their pre-arc is about 500k. Make sure they are in the model just in case.

 

[pc1966]

Just to add my experience with MCCB is very limited, it might be someone like @Julie. has better suggestions.

 

[pc1966]

Also I would add that sometimes you can't achieve perfection in design at an affordable cost so you need to be pragmatic about what your threshold for acceptance is. Obviously there are some aspects you MUST meet as safety is immediately dependant on them:

• ADS times - to avoid excessive risk of exposure to dangerous voltages.
• Break capacity - so a MCCB/MCB/fuse does not explode in someone's face.
• Adiabatic / CCC limit - so cables (including bonding) are not at risk of fire, or the serious cost and down-time of replacing a damaged sub-main, etc.

Selectivity is a definite design goal and in the regs, but it is not always possible to meet for various reasons and often this is simply accepted as "good enough". A simple and common case is blowing a 13A fuse will normally take out a 20-32A MCB, but as that can be reset with little skill, and often nothing too critical to safety is impacted - we see that as perfectly OK.

So in your design you need to look at cases where the software tells you selectivity is not met and evaluate if that is an acceptable risk or not, and in discussion with the client. There are some points to consider such as:

• What are the implications of selectivity failure? I.e. what other healthy circuits are taken off-line in that situation, and how much risk or inconvenience does that cause.
• How long an outage is tolerable? Based on the impact (above) of such a fault.
• How easy is it to restore upstream power? E.g. a fuse requires electrically skill intervention to replace, but a MCCB/MCB can be reset by a caretaker, etc, with just basic training and written guidance, so time-scales for an authorised person to react looked in to.
• How probable is a fault that would trigger the situation? E.g. is it happening on the more likely L-E fault (PFC) or only on a 3P bolted L-L-L fault (PSCC), and are the related cables or accessories at high-risk of damage (e.g. public socket or flex of some appliance, etc), or low-risk (e.g. sub-main in SWA run in risers, etc, away from DIY drilling regions).
• Is any mitigation present? For example, dual power feeds or UPS/backup generator when failure has serious impact, emergancy lighting, etc. Failure of course might be a wider power outage and not just an installation fault!

So if you get to the point when only a few circuits are vulnerable to low-risk PSCC faults, mitigation is going to be very expensive or impractical, and rectification of the fault manageable by local staff in reasonable time-scales, it might be fine for the client to accept those aspects in writing, and your design is done.