IET logo
 
IET
Decrease font size
Increase font size
Topic Title: BS88 Part 2 fuses - let-through energy.
Topic Summary:
Created On: 27 January 2009 01:42 PM
Status: Post and Reply
Linear : Threading : Single : Branch
1 2 Next Last unread
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 27 January 2009 01:42 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Hi all,

I am trying to find out a bit more about BS88: Part 2 fuses and have come across a Total I²t let-through energy, specified by the manufacturer for different fuses and system voltages.

Does anyone know if this is the maximum let-through energy for the fuse, regardless of fault level? i.e. could you use this value to compare against a cable withstand rating, across a range of fault levels up to the max. interrupting rating of the fuse.

Thanks,

Steve

Edited: 29 January 2009 at 10:02 AM by svand75
 27 January 2009 02:43 PM
User is offline View Users Profile Print this message



mstaple

Posts: 324
Joined: 23 July 2004

Hi Steve,

Yes this is the total let through energy of the fuse i.e it is no longer in one piece. You could use this value to compare against a cable although with a cable we are not interested in total let through energy (we try not to melt the conductors) but rather we are preventing the conductors getting too hot and melting the insulation. This could lead to fire, electrocution etc.

Therefore the regs advise using

Ib < or = In < or = Iz

Where
Ib = design current of the circuit
In = rated current or current setting of the protective device (fuse is this case)
Iz = current carrying capacity of the conductor

Regards

Edited: 27 January 2009 at 02:49 PM by mstaple
 27 January 2009 04:02 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Hi mstaple,

Thanks for getting back to me. Yes, the regs advise using that check for protection against overload current. I'm fairly happy with that part of the regs but the protection against fault current section has been causing some confusion!

For very short durations (less than 0.1s), we are advised to compare k²S² with the I²t let-through energy of the device. I believe this is to ensure that the cable does not reach it's rated short-circuit temperature before the protective device opens the circuit.

So, going back to the manufacturer's total I²t value... If this value is given for a specific fault level, say 200kA, is it possible for the total I²t value to be higher for a lower fault level. As an example (not based on real data):

Total I²t (at 200kA, 690V) for a 16A fuse = 1800A²s

If a 20kA fault occurs, is it possible that the let-through energy could actually be higher? I would have thought so because the fuse might not rupture as quickly and there is a possibility that the actual (not limited) peak current could be reached. Also the duration of the fault would be longer and, hence, t would be too.

Thanks,

Steve
 27 January 2009 04:49 PM
User is offline View Users Profile Print this message



mstaple

Posts: 324
Joined: 23 July 2004

I2t is the total let through energy of the fuse before it ruptures or the energy required to rupture the fuse.

If you have a high fault current the time to rupture is reduced, if you have a low fault current the time to rupture is increased. The let through energy must be exceeded for the fuse to operate (for a simple rewirable fuse, it is the energy required to melt the fuse).

Once the fuse is ruptured the circuit is opened and hence no more current flows. Therefore you cannot exceed the I2t rating of the fuse as that is the nergy required for the fuse to no longer be a fuse!!

Hope that makes sense

What you may be confused with is the rupturing capacity of the fuse which is the maximum fault current the fuse can break safely (without exploding)



Edited: 27 January 2009 at 04:51 PM by mstaple
 27 January 2009 07:24 PM
User is offline View Users Profile Print this message



FerrazShawmut

Posts: 3
Joined: 27 January 2009

 28 January 2009 01:19 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Yes mstaple, completely agree with everything you've said above. However, Ferraz Shawmut has just provided a great example for me to use to demonstrate exactly what is confusing me (thank you Ferraz S.).

Let's take the fuse characteristics on page 8 and 9 of the document in that link.

For a 315M400A fuse, the total I²t let-through (at 415V) shown is roughly 1.5 million A²s. Yet, when we look at the time-current characteristic on the previous page, for that same fuse, a current of 5000A causes the fuse to rupture after 0.1 seconds. Surely this means that the energy let-through of the fuse, for a 5000A fault level, is:

I²t = 5000² x 0.1 = 2.5 million A²s.

So we have a case where the total I²t given on page 9 is exceeded? If we try the same thing at a lower value, 3000A, we end up with an even higher let-through energy of 8.1 million! Perhaps I have misunderstood, but it does look like the total I²t given is for a particularly high fault current (the fuse breaking capacity, 80kA) and it does not represent the maximum let-through energy of the fuse.

Thanks again,

Steve
 28 January 2009 01:57 PM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19747
Joined: 23 March 2004

Most protective devices have the ability to limit fault current to a lower value than that of the prospective fault current they are exposed to. This is achieved by a combination of energy absorbed into the arc and the fact that when disconnection times are less than one half cycle (ie 10mS) then the device does not see a "full half cycle" hence the energy is also less. (Power factor of the fault also has an effect)

Comparing A2s data aginst time current characteristics is not really valid.

For the BS 88 fuse, the breaking characteristic is pretty linear on a log/log scale so it can be assumed that the A2s is pretty constant over the whole range of fault levels (no moving parts, just heat to operate).

This is not true for circuit breakers where you need to evaluate A2s against PSSC for the particular case (as they can be pretty independant of fault current at the millisecond level as you have to overcome the inertia of all those moving parts first.

Does that help

Regards


OMS

-------------------------
Failure is always an option
 29 January 2009 09:19 AM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Hi OMS,

Yes, that is as I thought. The fuse disconnection time for high fault levels is extremely quick and, as you explained, the I²t let-through is restricted to a very small value because the peak fault level is never reached. I can also appreciate that energy is absorbed into the arc once the fuse element has started to rupture.

With regards to the use of TCCs not being suitable for comparison with energy let-through data, I would have thought that this is true for faults of a very short duration (i.e. those that are disconnected before a full cycle has past as described above). However, if the fault level is relatively low and the protective device takes longer to disconnect the circuit, then surely calculating the energy of the fault over the period before the protective device trips will provide a reasonable estimate of the energy that the cables/devices will need to be capable of withstanding?

I know that TCCs are generally for discrimination purposes and that I²t data is usually provided, or plotted, separately but for discrimination purposes the trip times on the TCCs have to be reasonably accurate and should therefore provide a rough guide as to how long it will take for a device to trip for a particular fault level. The fact that we are looking at faults that exist for much longer than quarter of a cycle should mean that the rms symmetrical fault level is also a reasonable indication of the current that will flow through the circuit. This gives us approximate current and time values to apply to the let-through energy calculation and a final I²t value to compare with cable data.

That's my understanding of it so far anyway. As you may have guessed, I'm pretty new to all this so any further advice will be very much appreciated!

Thank you,

Steve
 05 February 2009 01:12 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Does anyone else have any thoughts on this?

Thanks,

Steve
 06 February 2009 01:32 PM
User is offline View Users Profile Print this message



hileys

Posts: 9
Joined: 03 June 2003

If you consider energy let through I2t in A2 secs the values are only valid for very short time periods. The equation K2.S2 = I2t is known as the adiabatic equation, which assumes no heat loss from the system (basically to simplify the maths). Where long times are involed (Greater than a few seconds) the heat disipated becomes significant and results in inaccuracy of the equation.

If you consider the energy supplied to a load continuosly for several months the I2t would be many orders higher than the let through of the fuse.

In addition to TCC's, fuse manufacturers also publish cut-off current characteristicts which show the current limitation effect. They basically show on one axis the prospective RMS current and give a resulting cut off current which is given as a peak current value (due to the cut off current being non sinusoidal). The prospective RMS value is not achieved due to the interuption of current occuring in less than a quater cycle. On low values of fault current, no limitation occurs, (fuse operating time over 0.01sec), the max prospective fault current would occur. Heating effect to cables and equipment would then be based upon fault current squared times fuse operating time and not the I2t energy of the fuse.

Stuart Hiley
 10 February 2009 02:59 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Thanks for the response Stuart and sorry for the delay in my own reply.

I appreciate the clear explanation; it seems to make more sense now. So, the manufacturers provide the energy let-through curves for faults of very short durations, where the fuses disconnect the circuit before the first quarter cycle has passed. Faults of longer durations can expose circuits to higher fault energies and these can be calculated using the symmetrical rms short circuit currents to establish the temperature rise that cables will be exposed to.

That seems to make reasonable sense. However, I am now wondering if manufacturers' energy let-through characteristics should actually state that they only apply to faults of very short durations (as an example, see the ones provided by Ferraz previously)?

Perhaps it is obvious, to the trained eye, that the I²t let-through characteristics are only applicable to faults of short durations. As someone who does not have much experience in the selection of fuses, I might have been tempted to think that the total I²t (shown on the fuse datasheets) applied regardless of fault current and duration and that cables and equipment could not be exposed to higher fault energies. I now see, from the time/current curves that the energy let-through can be significantly higher for lower fault currents that exist for more than a cycle.

The manufacturers' data makes it easy to check that circuits are protected from fault of very short duration adequately. For fault of longer durations, I guess that the only way to ensure that cables are not exposed to fault energies that they cannot withstand is to plot both fuse and overload curves on the same axis and, from this, plot the I²t let-through for all fault levels (and durations)?

Edited: 10 February 2009 at 07:38 PM by svand75
 11 February 2009 10:38 PM
User is offline View Users Profile Print this message



peterridge

Posts: 36
Joined: 18 September 2005

Hey svand75

The above is getting very involved!

I think the important thing to appreciate is that energy let-through is essentially for discrimination purposes only.

Basically there are two types of protection - current limiting and non-current limiting.

For non-current limiting devices time-current curves only are required to appraise the discrimination.

However, for current limiting devices it is also necessary to consider the energy let through of the device. Time-current curves are still used in the first step of the discrimination study - as a fault of low magnitude would be time dependent.

Step 2 of current limiting discrimination is by ensuring the total let-through energy of a device is less than the pre-arcing of the upstream device.
Please note, the energy let through consists of two components, Pre-arcing and arcing. Pre-arcing is the energy the fuse can see before it is destroyed. When a fuse reaches its arcing zone, the fuse needs to be replaced before the circuit can be re-energised.

EXAMPLE see attached data sheet on Eaton BS88 part 5 (J type) fuses.

Using the graph, a fault of 3kA would trip a 250A fuse in just under 0.2 seconds, and a 400A fuse would take over 1 second. Hence it appears that discrimination is acheived.

However, what if there is a fault 10kA?

This is where step 2 of the discrimination study comes in for current limiting devices.

250A BS88.5 has total energy let-through of 480,000 A2S (at 415Volt), and the 400A BS88.5 has a pre-arcing of 475,000 A2S. Therefore discrimination is NOT guaranteed, and a fault of large magnitude will probably knock out both fuses.

In this case a 450A fuse should be selected to ensure discrimination (450A fuse pre-arc = 740,000 A2S which is less then the 250A total let-through of 480,000 A2S).

Hope this helps

Cheers
Peter
 11 February 2009 10:43 PM
User is offline View Users Profile Print this message



peterridge

Posts: 36
Joined: 18 September 2005

oops, how do you attached?
 12 February 2009 12:50 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

Hi Peter,

Yes, I agree with all that you said above, I think.

TCCs provide you with the information you need to do a Time Discrimination check. If the TCCs cannot provide you with enough detail then, with current limiting devices, you can also do an Energy Discrimination check to ensure that the device closest to the fault trips first.

Your note about using the pre-arcing current of the upstream device is very important too, as we don't want the upstream device to be affected by faults outwith their own protection zone.

What are your thoughts about cable withstand checks? If both the TCCs and the energy let-through data are for discrimination only, then how do we ensure that cables are not exposed to temperatures above their rated values under fault conditions? I think the cable withstand check is specified as a requirement (for faults of very short duration) in the wiring regs too.

Thanks for your help,

Steve
 12 February 2009 01:56 PM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19747
Joined: 23 March 2004

Cable withstand checks are undertaken (by convention) with actual operating time for the fault level when that operating time exceeds 0.1 seconds (ie it assumes no current limiting effect) - essentially application of the adiabatic expression

If your fault level is likely to result in an operating time less than 0.1 seconds then manufacturers I2t data is required for comparison with cable withstand

OMS

-------------------------
Failure is always an option
 12 February 2009 02:26 PM
User is offline View Users Profile Print this message



peterridge

Posts: 36
Joined: 18 September 2005

Hi Steve

TCCs provide you with the information you need to do a Time Discrimination check. If the TCCs cannot provide you with enough detail then, with current limiting devices, you can also do an Energy Discrimination check to ensure that the device closest to the fault trips first.


This statement is in fact incorrect.

For a current limiting device, the TCC check holds good for nearly all fault conditions. However it is necessary to verify discrimination using the energy let-through method. This is the ONLY way to guarantee total discrimination of current limiting devices.

I wanted to try and keep it as simple as possible, but the energy let-through method is only strictly necessary when evaluating discrimination under short-circuit conditions between two current limiting devices.

If you really want to get into the nitty-gritty, check out BS EN 60947-2 (LV current limiting circuit breakers are classed as category B, and non-current limiting circuit breakers category B.

See OMS response regarding cables.

Hope this helps

Cheers
Pete

Edited: 12 February 2009 at 02:31 PM by peterridge
 12 February 2009 02:36 PM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19747
Joined: 23 March 2004

but the energy let-through method is only strictly necessary when evaluating discrimination under short-circuit conditions between two current limiting devices.


You also need to evaluate let through for cable withstand puposes where disconnection time is "short" - see 434.5.2

Regards

OMS

-------------------------
Failure is always an option
 13 February 2009 01:18 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

For a current limiting device, the TCC check holds good for nearly all fault conditions. However it is necessary to verify discrimination using the energy let-through method. This is the ONLY way to guarantee total discrimination of current limiting devices.


I take it that the fault conditions that the TCC check does not hold good for are those that are near the maximum breaking capacity of the protective device? These would be of very short duration and TCCs tend to stop at between 0.1 and 0.01secs.

After reading it again I think that, yes, my statement above was a bit misleading. I meant that the TCCs can only be used for faults that are within the range shown on their Time-axis. The energy let-through curves or data will then allow you to check the discrimination for faults of shorter durations.

I will try and get a look at that standard.

Yes, OMS, that's the bit of the regs I was referring to previously. Thanks for confirming that.

Almost moving on to another, related, topic: the regs don't specify what length the cable withstand check should be carried out at. Would you say that good practice would be to do your volt drop check at the point furthest from the cable's source and your cable withstand check at the protective device? If you don't consider faults along the full length of the cable then I wouldn't think that the protection would really be adequate.

Thanks for all the advice everyone, I do appreciate it!

Regards,

Steve
 13 February 2009 01:51 PM
User is offline View Users Profile Print this message


Avatar for OMS.
OMS

Posts: 19747
Joined: 23 March 2004

Almost moving on to another, related, topic: the regs don't specify what length the cable withstand check should be carried out at. Would you say that good practice would be to do your volt drop check at the point furthest from the cable's source and your cable withstand check at the protective device? If you don't consider faults along the full length of the cable then I wouldn't think that the protection would really be adequate.


There would be no point in undertaking a volt drop analysis anywhere other than at the furthest point would there - for all other factors being equal, it is only distance that varies the value of volt drop (as resistance/impedance increases with distance hence voltage will depress with distance).

The cable withstand check needs to be undertaken again at the load end - the increased impedance will damp the fault current and thus the cable is more thermally challenged by the longer duration fault.

The close up supply end fault should also be analysed (usually to determine the rated withstand of that tier of the distribution system).

You need to stop confusing discrimination studies with cable withstand assessments - whilst linked they serve to inform totally different criteria.

Trust this helps to clarify a little

Regards

OMS

-------------------------
Failure is always an option
 23 February 2009 12:59 PM
User is offline View Users Profile Print this message



svand75

Posts: 31
Joined: 27 November 2006

OMS,

Thanks for your guidance above. After reading your comments, I dug out a few books and refreshed my memory on that cable withstand check.

I was forgetting that the cable withstand check is actually a temperature check so I can now appreciate why it needs to be done at the load end. The cable will be exposed to the higher temperature for a longer period of time before any protective devices trip. I think that the longer cable length will also mean an increased circuit resistance and therefore more heat losses in the cable.

You were right; I was getting a bit confused. Will continue reading up on the subject!

Thanks very much for your help.

Regards,

Steve
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.