the 'laws' of Ib <= In <= Iz and I2 <= 1.45 Iz re: direct buried cable

I agree with Andy that Cc=0.9 should be applied to ref method D, for much the same reasons.

Firstly it is entirely possible to not require overload protection for a buried circuit. Feeding fixed equipment that cannot by design create an overload (as opposed to fault current) is quite common in my neck of the woods. Therefore why should the creators of the table assume that it is

Indeed the Cc factor of 0.9 is not an absolute. The Commentary to the Regs goes into its derivation and 0.9 is explicitly stated as rounding down for a general value that is deemed to comply; if you know enough about the circuit, a different - indeed sometimes lower, if you try hard enough - value is obtainable for Cc in order to achieve the basic requirement of ensuring the protective device operates before the cable is affected.

There is nothing that says Cc should not be applied to buried circuits.

Other reference methods aren't appropriate for buried circuits. Otherwise they'd say they were. You might choose to use them to obtain values by analogy but it's not in scope of Appx 4 and I would argue that requesting information from the manufacturer, or using IEC 60287 would be more appropriate.

BS7671 Appx 4§4 explicitly says that it should be applied also "To achieve the same degree of overload protection where a cable is “in a duct in the ground” or “buried direct” as compared with other installation methods a rating factor of 0.9 is applied as a multiplier to the tabulated current-carrying capacity".

IET Electrical Installation Design Guide (4.3.5) says that it should be applied: "Buried circuit rating factor Cc The buried cable ratings in Tables 4D4A and 4E4A of BS 7671 are determined at a ground ambient temperature of 20 °C (compared with 30 °C for cables installed in air). Whilst these increased ratings will result in the same conductor operating temperature at full load (70 °C for Table 4D4A), under overload conditions the conductor may exceed the limiting temperature (115 °C for Table 4D4A). The use of the rating factor prevents the cable going over this temperature during an overload, see Regulation 433.1.204"

Once overload is present then the 0.9 seems to come into it ...  but its all clear as mud now to me  ! 

The way my head thinks of it is that that if In ≤ Iz then we presume that I₂ ≤ 1.45 Iz is also true - i.e. there is some assumptions/fudge-factors about both the cable current/temperature/time performance and the OPD time/current characteristics. We know the fudge isn't always true for all OPDs - e.g. not for BS 3036 fuses (or perhaps BS 1362 ones either). I think what they're saying that it's not true for cables rated for other than 30 degree ambient either - which sort of makes sense to me as at 20 degrees we can create more heat before hitting 70 degrees - i.e. can carry a larger current.  But the gotcha is that while the extra current to get the cable to its short term overload temperature might be 45% of the 30-degree rating, it's probably rather less than 45% of the higher 20-degree current rating.

Some made-up numbers to illustrate my thinking...

Say we had a cable that was rated at 10A at 30-degrees ambient, so will cope with an overload to 14.5A (for whatever short duration).

Stick that same cable in a 20 degree environment and we might say it's good for 11A continuous, the problem is that it's probably not then good for 1.45x11A = 15.95A for overload, probably something slightly larger than 14.5A since there's a slightly larger temperature difference, but not that much larger. The 0.9 factor then brings us back to something closer to 14.5A. (Numbers made up for clarity rather than accuracy).

My head might well have it wrong of course...

  - Andy.

hmmm 

right then, back to how I first did it (sort of ... lost track now) seeing as how there is growing opinion Cc 0.9 is indeed applicable to It (to comply with I2 <= 1.45 Iz which for stated devices will be  if In <= Iz)

10mm2 SWA 70DegC  In (BS88) = 63A   It = 60A  (from RefmD - but use 10% 'clause' to that table )  Cc = 0.9  Cd = 1.04    (all others unity or not applicable )

Iz = (It * 1.1) * Cc * Cd   =  Iz = (60 * 1.1 ) * 0.9 * 1.04  =  61.7      (without the 10%  its  only 56.1 !  )

Ib <= In <= Iz  =  32 <= 63  <= 61.7 (56.1)  [not satisfied

It's a shame for less than a couple of amps  (one might imagine it would be fine in reality even if not regulatory compliant); of course  one could go scraping that extra couple of amps in Cs [soil conditions] perhaps ... or cajoling the manufacturer  :-)

some solutions

- increase cable csa  (cost increase, but copper underutilised perhaps  though poss. less power losses)

- decrease fuse rating  (if feasible, no cost increase  but reduction/loss in selectivity - if any, or no issue if not important)

- design so no overload possible (if feasible 0.9 prob goes away etc)

As an aside, this kind of thing is why Graham and I and a few others were discussing design qualifications a few weeks ago. The underlying problem with all this discussion is the underlying assumptions on which these "rules" are based. You have already listed some of them, ground temperature, fusing factor, whether 70C for Tmax is reasonable etc. Most of these numbers have a tolerance of perhaps 10%, yet they are assumed to be worst case, or best case to suit the situation. Actually none of these numbers will by themselves cause the design to fail, at least in an understandable timeframe. Design is much more complex than just taking these numbers from tables, and the job is done, and those of you who have used a design program will realise this. Ok, you decide the cable is rated at say 68A. but then you look at the fusing factor (1.45 say) and decide it is only suitable to use at 61A. Here's the bit you need to understand, you have looked at a continuous overload situation at 61A x 1.45, what is the actual temperature rise if you choose 63A instead? Also how fast is the circuit broken at this higher current, and what is the corresponding temperature rise at your assumed ground temperature? What if the ground is more likely to be at 10 degrees, because maximum current will only occur with a winter load?

Realistically now, is there any evidence that this design by BS7671 to the letter is in any way correct for any actual installation? Where are the failures? BTW if the DNO used these methods you are unlikely to have a supply at all at most UK properties. I am not saying it is wrong, but it is probably pessimistic for many cases, and good designers know this and make proper judgement of the design.

The answer to Habs second question is clearly that this would only cause a possible max conductor temperature of 72 degrees or so for a short period. This is not serious, and are you sure all the corrections are accurate and in the right direction? That is where the skill is necessary. However just like the OSG the BS7671 numbers are safe as far as possible without any necessary judgement.

David, the conversation seemed to boil down onto whether the tabulated current for RefMethD included the 0.9 or it did not (have a read of Graham's contribution !) and whether it was required or not .. the maths was perhaps just a mechanism for exploring it (or perhaps not).

If I summarised your reply to my understanding of it, I would agree that experience and knowledge does indeed count for a lot. I am not sure anything specifically helpful is added by what you have said (which is unusual when I read most of your contributions).

I do realise after many years of 'listening' here there and everywhere, that designs and subsequent construction can be entirely suitable (and safe) despite not following Regs verbatim due to all sorts of affects.  Still, one might start from basics and go from there. One never knows what one might find.

Regards

Dear PW

I put in that bit because we were beginning to go down a dead end route in the use of BS7671. For example consider an electric motor as a load. Should I consider the starting current as Ib or some other value like the FLC? The answer to that depends on the number of starts per hour and their duration, which could both be considerable. Whilst this kind of problem has become easier with the use of electronic controls, it has not "gone away". This means that the blanket formula is not always right or useful.

Appreciated and thank you - same  to all who continue to contribute on here.  There is always something to learn !

Habs