What's inside an SPD?

[aptsys]

As far as I understand the 18th edition unless you can provide a risk assessment which says the value of goods within the property likely to be damaged by an over voltage is less than the cost of an SPD then one must be fitted in domestic.

Not quite, depends on the area, method of DNO supply etc. Besides, the list of equipment likely to be damaged is small. Any modern electronic equipment incorporating a SMPSU will have everything integrated already, so it doesn't leave you with many susceptible items.

 

[bernardgreen]

SMPSU will have everything integrated already

Good quality ones probably have.

 

[JohnW2]

There are criteria for when such a device is required however as with everything in the regs, many don't understand the nuances, therefore they're likely to get fitted to every board.

I fear you could be right, daft though it might be.

I've now had a quick look at the relevant part of the regs and it certainly does look as if you are right in suggesting that, in reality, there are probably few situations in which someone who actually understood would conclude that there was a need/'requirement' per the regs.

Kind Regards, John

 

[JohnW2]

.... Those "Over Voltage Arrestors" were fitted to that CU after the following "event" :- .... there was a brilliant flash of lightning, followed almost immediately by the clap of thunder ... Following this "event", these devices were fitted by my son, who is a licenced electrical contractor.

There will, of course, be some such anecdotes, because such things do happen, even if extremely rarely. I am tempted to invoke the common saying (nearly always used 'by analogy', not literally as in this case!) that "lighting doesn't strike in the same place twice"!

In other word, the probability of any of us experiencing such an event just once is incredibly small - so it could be said that the probability of it happening a second (or third, or ...) time, whilst theoretically 'not impossible' is so small as to not justify the effort and cost of taking any steps toi address that possibility.

Kind Regards, john

 

[DetlefSchmitz]

John- I'm sure you know that the probability of a second strike is still exactly the same as it was for the first. Even after the first has occurred.

 

[JohnW2]

John- I'm sure you know that the probability of a second strike is still exactly the same as it was for the first. Even after the first has occurred.

I do, indeed know that (and many of my past teachers would be turning in their graves if I didn't know ) - and I agree that I did not word my statement strictly correctly.

However, as I'm sure you realise, what I meant (and what matters in terms of the context I was talking about) is that the probability of it happening twice (within a given time period - like 'a lifetime') to a given household/ person/ whatever is very different (dramatically less) than the probability of it happening once to that household/ person/ whatever during the same time period.

Kind Regards, John

 

[EddieM]

That's not true, the probability of throwing 2 sixes on a die before the first throw is 1:36 if after the first 6 is thrown it's 1:6 on the second throw.

 

[bernardgreen]

I'm sure you know that the probability of a second strike is still exactly the same as it was for the first.

Ionisation of the air created by a lightning strike makes it easier for a second strike to happen, Air movement disipates the ionisation so the second strike has to happen before the ionisation has been dispersed..

 

[DetlefSchmitz]

Ionisation of the air created by a lightning strike makes it easier for a second strike to happen, Air movement disipates the ionisation so the second strike has to happen before the ionisation has been dispersed..

Yes. But of course I meant to imply that enough time had passed to decide to fit some extra protection, by which time I expect the ionisation to have dissipated!

 

[JohnW2]

Ionisation of the air created by a lightning strike makes it easier for a second strike to happen, Air movement disipates the ionisation so the second strike has to happen before the ionisation has been dispersed..

We (at least I) were talking about the probability of a second strike weeks, months, years or decades after the first (in relation to a judgement of the 'statistical worth' of installing protection against any such subsequent events) - not in the very short period during which the environment was affected by the first strike!!

Kind Regards, John
Edit: Typed too slowly again!

 

[JohnW2]

That's not true, the probability of throwing 2 sixes on a die before the first throw is 1:36 if after the first 6 is thrown it's 1:6 on the second throw.

Indeed - which is the point that Detlef was making - but that doesn't alter the fact that the probability of getting two sixes in two throws is 1:36.

In context, if the probability of the electrical installation being damaged by lighting during, say, a 50-year period were, say, 1:1,000, then the probability of it happening twice in a 50-year period would be 1:1,000,000 - and that's still true after a first event has happened. The probability of the second is, indeed, still 1:1000, even after a first event, but there was only a 1:1000 probability of that first event happening in the first place - so, overall, there's still only a 1:1,000,000 probability of two events happening during the period.

Kind Regards, John

 

[DetlefSchmitz]

We are talking about the case when the first strike HAS happened. So the probability of the two strikes has now collapsed to 1000:1.

 

[JohnW2]

We are talking about the case when the first strike HAS happened. So the probability of the two strikes has now collapsed to 1000:1.

Needless to say, we don't disagree about the mathematics of the probabilities - and I perhaps confused matters by introducing a probabilistic discussion in relation to a common idiomatic figure of speech ("lightning doesn't strike the same place twice").

Since I was talking about (per 'the phrase') was "lighting striking the same place twice", what I was saying was literally correct - i.e. if everything is random, then the probability of "lighting striking the same place twice" is, indeed, the square of the probability of one strike (i.e. 1:1,000,000 if the probability of one strike is 1:1000). In other words, I was not, at that point, talking about a situation in which one strike had already occurred - and that's probably what has confused things a bit.

The point I was trying to make was not really a mathematical one at all. My point was that (assuming everything is random) the fact that one has been incredibly unlucky in having already suffered from one lightning strike is not really a reason to worry further about "it happening again", since the probability of that happening is as incredibly small as was the probability of the first one having happened in the first place.

This rather illustrates a common problem with the practical interpretation (in relation to an individual person, building or whatever) of extremely small probabilities of discrete events. When such probabilities are larger, meaningful interpretation is much easier. . If, for example, the probability of something happening in any one year is, say, 0.10 (10%, 1:10), then one can say that the most likely number of occurrences of that event during a 50 year period (say, an 'adult lifetime') would be 5 - or if the probability were 1:25, then the most likely number of occurrences of the event in 50 years would be 2 etc. etc. One can base decisions on such information.

If the probability were 1:1,000, then by far the most likely number of occurrences of the event in 50 years would be zero. However, 'by far the most likely number of occurrences of the event in 50 years' would again be zero for any probability less than 1:1,000. Hence, 'by far the most likely number of events in 50 years' (which is what matters to an individual person/building/whatever) will be the zero whether the probability per year is 0.001 (1:1,000), 0.00001, 0.000000001 or whatever.

What matters in terms of the individual (person/building etc.) is the number (if any) of discrete events which they are likely to experience in a given time period (e.g. 'a lifetime') - an 'average number of expected events (for the person/building etc.) in 50 years' (which is easily calculated) is not really a very useful guide to anything.

For reference, the following indicates the probability of various numbers of events occurring in 50 years for a range of probabilities of the event occurring in a particular location in any one year. As you can see, even with a 1:1,000 probability per location per year (obviously very high for lightning strikes), there is still a 95% chance that one will suffer no strikes in 50 years - and when one moves to probabilities more realistic for lightning strikes, the probability of there being anything more than zero events in 50 years gets so small as to probably not be worth thinking/worrying about (regardless of whether or not any previous strikes have occurred)!

Kind Regards, John

 

[EFLImpudence]

View attachment 159632

That doesn't look at all probable.


...but what does a 23% chance of rain tomorrow mean? Anything?

 

[DetlefSchmitz]

...but what does a 23% chance of rain tomorrow mean? Anything?

I would assume they have a definition of that statement. It doesn't necessarily make it any more useful.