IET
Decrease font size
Increase font size
Topic Title: surge protection: positioning and type
Topic Summary:
Created On: 13 June 2018 12:44 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 13 June 2018 12:44 PM
User is offline View Users Profile Print this message



psychicwarrior

Posts: 579
Joined: 18 October 2010

Hello folks :-)

I have been reading (guides, manufacturer literature) about spd a bit more of late, having not had any experiences with the devices. There is a fair choice of kit and commentary that's for sure. I still have to re-read the Regs on the subject too (again, something I glossed over!).

Just taking a made-up simple example installation to seed thoughts: 3-phase installation (TN-C-S), a main db (fuses) with distribution circuits to final circuit db's (perhaps split load, or all rcbo configuration). In addition there is a lightening conductor setup.

Now then and tentatively commenting, it would seem from what I've read that generally, in a domestic setting especially, in the UK there is no demanding requirement to consider [atmospheric] surge protection (based on AQ1 assessment?)...but one could adopt a risk assessment approach and consider specific client installation requirements, which may alter that position of course....and definitely where a lightening protection is present on a building (although installations would seem to have done ok without all this spd guff over time - though I suspect there is more impact than 'we' hear about).

I realise it pays to think of what is intending to be protected, but if it were decided to build in spd, where would they go and what type ?

- type 1 in/aside the main db and type 2 in/aside each final db
- type 2 in/aside the main db and each final circuit db
- type 2 in/aside the main db and nothing else
- type 1+2 in/aside the main db (a combo device apparently) and nothing else
- type 2 in/aside each final circuit db and nothing else ?
- any other

I had read that type 3 (where sensitive/important equipment is deployed) should be considered close to supply point/terminals - but it has to be combined with other types... (perhaps this is the 2+3 type in a board).

What protects the operator's metering equipment (if that's an issue) ?

(then I guess the TV/data network would need to be considered!)


Thanks
Habs
 13 June 2018 01:41 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16718
Joined: 13 August 2003

(then I guess the TV/data network would need to be considered!)

Indeed - in my experience it's often the overhead phone line that 'draws in' the biggest risk (especially in rural areas where it can run on poles for miles).

What protects the operator's metering equipment (if that's an issue) ?

I believe that DNO equipment is required to have a higher withstand than that deeper into the installation - I suspect too that conventional meters, by not having an earth connection, are a bit less vulnerable.

As for the types and where, I'm not entirely clear myself yet, but my thinking so far is the ideal is: type 1 at the origin, then a type 2 at the DB and type 3 at the equipment. If you had a simple domestic with just a CU at the origin then a combined 1+2 would seem useful for that, or omit the type 1 protection and use a plain type 2 if the expected voltage stress is lower. Presumably the type 3 could be omitted for any equipment that isn't particularly sensitive or where damage to it, or its malfunctioning, is acceptable.

- Andy.
 13 June 2018 04:00 PM
User is offline View Users Profile Print this message



mapj1

Posts: 10599
Joined: 22 July 2004

The aspect not always clear too those coming into this for the first time, is just how many volts can pile up across things that are connected, but by an overly long route. As such long meandering earth interconnections can undo much of the good work - a metre length of typical earth bond, say 10mm single, will have a self inductance of perhaps 1 microhenry - negligible impedance on the timescale of 50Hz, but for events that occur in micro-seconds, so waveforms with the frequency components into the Megahertz, capable of several tens of volts drop per amp flowing.
Depending on the source of the of surges you are trying to trap, the layout of the service cables, ideally so they all sweep close to a common MET on the way in or out of the building, is at least as significant as selecting the right type of SPD. One of the reasons modern phones and answering machines etc, and now also ADSl modems suffer so much is that they often are the point of closest approach between the limbs of an accidental antenna, formed with the telephone line as one half, and the mains cables as the other.

-------------------------
regards Mike
 13 June 2018 04:38 PM
User is offline View Users Profile Print this message



psychicwarrior

Posts: 579
Joined: 18 October 2010

mapj1: "...Depending on the source of the of surges you are trying to trap, the layout of the service cables, ideally so they all sweep close to a common MET on the way in or out of the building, is at least as significant as selecting the right type of SPD.."

Is it possible to say that if control of the layout of service cables is not possible (they are already in situ and layout unknown perhaps), then spd becomes a bit more relevant/ a requirement, or is it a bit more involved to prove if its worth it ?

If so, In my simplistic example and (thinking lightening strike direct to rod or to ground within vicinity): use a type 1+2 device at the main db (origin) and not bother with anything at the final circuit db's, or to put in Type 1 at main DB (where will be MET close by too) and then stick in some type 2's at each final circuit db to divert surge ..which the latter is my thinking too ?

Yep, 'coming into it for the first time' is accurate :-) The flow-charts (haha) for selection etc seem to indicate that given any lightening protection on the building, then spd's should be considered....hence the type 1 at the main db/met (origin) location and type 2 at other boards down stream (if necesary) question.
 13 June 2018 04:46 PM
User is offline View Users Profile Print this message



arg

Posts: 211
Joined: 18 March 2015

Also worth mentioning that it's easily possible to make things worse rather than better by just adding SPD 'for good measure' rather than clearly understanding what the problem is to be solved and how you are solving it.

For example, if you bring in a new earth connection to your SPD, that's now a tempting destination for currents to come in through wiring (on either side) and exit through your SPD and new earth: if that point was previously not well connected to earth then the surge currents probably wouldn't have entered your installation in the first place. That leads you to think that if you want SPDs at all they should be at the boundary where conductors come from 'outside' to 'inside' the installation, so you don't bring surge currents 'inside'.

Unfortunately, installations often have more than one connection to the outside world, and if you think you've fixed things by having SPDs tie everything down to a local, low impedance earth at the point where the supply comes into the building, and again to another low impedance earth where (say) a circuit leaves the building to feed an adjacent building, you've then got the problem of voltage gradients in the ground between your two earth rods - you can have current enter up one earth rod, across your internal wiring and exit down the other earth rod, with your internal wiring and SPDs being a low impedance path in parallel with the mass of earth.

(I don't claim any great skill in getting this stuff right, but there are certainly plenty of ways of getting it wrong!)
 13 June 2018 04:48 PM
User is offline View Users Profile Print this message



mapj1

Posts: 10599
Joined: 22 July 2004

Well, all the SPD can do is change to a low resistance during a gross over-voltage, and you need to place it between two things you;d like not to drift too far apart. It is not possible to do that at a point where the things you wish to tie together are not physically close.
I''ll see if I can dig something up that make this a bit clearer - I'm in the wrong place now.

-------------------------
regards Mike
Statistics

New here?


See Also:



FuseTalk Standard Edition v3.2 - © 1999-2018 FuseTalk Inc. All rights reserved.

 
..