Help needed with understanding IR

[GBDamo]

I am showing my ignorance but where would I apply this figure.

I understand what it is and how it is calculated but where, say on an EICR, would you ever record it.

Also on a typical EIC it is not unusual to have readings of >200MOhms (instrument limitation). The true figure could be much higher. How would you calculate this?

Is it just a feel for the numbers? This looks close best calculate it?

Say we have IR results of 2MO + 3MO + 4MO The value for the installation would be 0.93MOhms. Whilst each individual figure is a pass, although dodgy, the overall IR for the installation is a fail.

Where is this recorded and by what criterion does it fail against?

Thanks OP for starting this, of all the tests its IR I find most tricky.

 

[ELECNEWT]

Sorry Tel.
I did not mean to give you a dumb. I was actually trying to give Pete a like. But I was doing it on my cheap smartphone with big fingers. I agree that the product over sum method is very cumbersome and prone to errors.

 

[johnduffell]

Say we have IR results of 2MO + 3MO + 4MO The value for the installation would be 0.93MOhms. Whilst each individual figure is a pass, although dodgy, the overall IR for the installation is a fail.

there's no such thing as an overall IR failure, if you do the IR overall and it's >1mohm then that guarantees the whole installation is alright (subject to 2 way switching etc) as you can't have any circuit <1mohm if the global is >1mohm. That is a big time saver given that many installs would be OK.
However the converse is not true, so if the global IR is 0.93 you have to break it down in order to know if there's a fault. Same as any fault finding.

 

[Risteard]

by what criterion does it fail against?

643.3.2

 

[Risteard]

there's no such thing as an overall IR failure, if you do the IR overall and it's >1mohm then that guarantees the whole installation is alright (subject to 2 way switching etc) as you can't have any circuit <1mohm if the global is >1mohm. That is a big time saver given that many installs would be OK.
However the converse is not true, so if the global IR is 0.93 you have to break it down in order to know if there's a fault. Same as any fault finding.

Not quite correct.

The insulation resistance for the entire installation in parallel must meet the minimum requirements of BS7671. For a larger installation this is subdivided into individual distribution boards with all final circuits connected.

See 643.3.2

 

[johnduffell]

The insulation resistance for the entire installation in parallel must meet the minimum requirements of BS7671

Thanks did not know that, sounds crazy to me but those standards writers must know that they're doing!

 

[Risteard]

Thanks did not know that, sounds crazy to me but those standards writers must know that they're doing!

No problem. Strangely it seems to be a commonly misunderstood part of the Regulations although the same or similar Regulations have persisted through the various Editions.

 

[Ian1981]

The 1Mohm requirement is actually for initial verification but stands to reason that it maybe, can used as a guide figure for carrying out an EICR

 

[Risteard]

The 1Mohm requirement is actually for initial verification but stands to reason that it maybe, can used as a guide figure for carrying out an EICR

Yes indeed there isn't a minimum value defined for periodic inspection and testing.

 

[GBDamo]

No problem. Strangely it seems to be a commonly misunderstood part of the Regulations although the same or similar Regulations have persisted through the various Editions.

Unless I'm missing something is it because there is nowhere to record this value on the standard forms?

Unless you had a good feel for the results, and was concerned, why would you calculate it other than to put it in a result box on the form.

Really hoping there isn't a box I've been missing.

 

[Risteard]

Unless I'm missing something is it because there is nowhere to record this value on the standard forms?

Unless you had a good feel for the results, and was concerned, why would you calculate it other than to put it in a result box on the form.

Really hoping there isn't a box I've been missing.

I either record it against an entry for meter tails or for the relevant distribution circuit, or simply list the figure alongside every box.

 

[alasdairp]

Firstly I presume you meant one Megohm (one million Ohms) as opposed to one Ohm.

Secondly the minimum values in the Regulations are for the entire installation (i.e. after you have done your calculation if you don't measure the whole installation in one go). (For a larger installation it can be subdivided.)

Thank goodness somebody commented!

 

[alasdairp]

As above you can't accept a lower value than 1mohm on any part of the installation whether combined or not. If you measure 20 circuits and get 0.1mohm combined that could be a fault on one circuit.
However if you break it down and they are all 1.5mohm then that's classed as ok.

You wrote 1mohm which is 0.001 ohm not 1 M ohm! Yes we all know what you meant, but it is important to get units correct. In this case your unit was 1000,000,000 times smaller that the 1M ohm.

 

[alasdairp]

Yes indeed there isn't a minimum value defined for periodic inspection and testing.

Really? I did not realise that. And then they recommend expensive Arc detection!
I would have thought that an overall IR test was one of the easier ways of checking for problems that could cause electric shock from metal appliances / wall lights. I have found live metal wall and ceiling lights have often been the cause of poor IR. Immersion heaters are the most common culprits in my experience.