Search This Blog

Wednesday 24 April 2013

Risk based PAT testing - a proposal

The IET Code of Practice in its Edition 4, recommends adopting a risk based PAT testing. ie decide on the frequency of inspection and test based on the actual risk posed by the appliance. This post presents a proposal on how a duty holder could work this out for themselves.

Note: This proposal has been updated on the 3rd of June 2013. 

Background

Deciding on the frequency of inspection and testing based on risk is not new to Edition 4. It has always been the approach of previous Code of Practices. However, there is more emphasis on this in Edition 4 following the Lofstedt Report. The aim of the Report was to actually reduce the amount of PAT testing that is done unnecessarily. For example, in low risk environments such as offices, there is no need to inspect and test PCs every year, but some companies routinely do this.


Notes on updated proposal: Risk factors reduced form 5 to 4. More examples presented. New frequency of "inspect every year & test every 2 years" has been added.

Risk Factors

In our view there are 4 factors that affect the risk posed by electrical equipment. These are:

1. Class of Construction
2. Type of Appliance
3. Environment it is used in
4. Failure rate (for the organisation) from previous PAT testing

Assessing the risk and deciding on the frequency of inspection and testing

The proposal is to assign the following numbers to these factors.

Class of Construction:  Class 1 = 4, Class 2 = 1
Type of Appliance: Handheld = 5, Portable = 4, Movable = 3, Stationary/IT = 2, Fixed = 1
Environment: Construction = 5, Factory = 4. Used by Public/customers = 3, School = 2, Office/shop = 1
Failure rate: >10% = 4, 1-10% = 2, <1% = 1

Note: The failure rate refers to the average failure for the whole site and not for that particular appliance. For example, if out of 200 appliances on site, around 5 fail when inspection and testing is carried out, then the failure rate would be (5 x 100)/200 = 2.5%. 

If previous test results are not known then use other factors to decide on your risk factor. For example, if this is a hair dressers, then a factor of 1 can be used to start with. However if this is a car body repair shop then a factor of 2 or 4 should be considered.

To work out the risk factor for any particular appliance, one would simply multiply these factors together and use the guidance below to work out the frequency of inspection and testing.

<10: Inspect every 2 years, test every 4 years.
10 to 25: Inspect every year and test every 2 years
26 to 50: Inspect and test every year
51 to 100: Inspect and test every 6 months
>100: Inspect and test every 3 months.

Example 1: A Class 1 (4) desktop PC (2) in an office environment (1) and previous failure rate is <1% (1). This gives a risk factor of 8 - inspect every 2 years and test every 4 years.

Example 2: In the example above if the PC was a laptop, then the factor for this would be 4 and the overall risk factor would be 16 - requiring inspecting every year and testing every 2 years.

Example 3: A Class 1 (4) kettle (4) in a shop (1) with previous failure rate of <1% (1). This gives a risk factor of 16 - inspect every year and test every 2 years.

Example 4: A Class 2 (1) drill (5) used in a school workshop (3) and previous failure rate is around 5% (2). Ie on average out of the 20 items in this workshop, at least one fails every time inspection and testing is carried out. Although the environment is a school, as it is used by students a higher risk factor of 3 is assigned here (As students should be regarded as customers). Overall risk factor would be 30 and this drill would have to be inspected and tested every year. 

Example 5: The same as above, but experience suggests that the failure rate is more than 10%. Ie on average out of the 20 items, 2 or more fail every time inspection and testing is carried out. In this case this factor changes from 1 to 2, resulting in an overall risk factor of 60. This drill will now have to be inspected and tested every 6 months.

In summary,  I have updated my proposal based on feedback. It should now be even easier to work out the frequency of inspection and testing using this guide that is practical and easy for most duty holders to implement. If this helps to demystify this area then that would be a good outcome.

I will be interested in hearing the views from readers of this blog on this updated proposal.


2 comments:

  1. The proposal as suggested is most sensible and seemingly straight forward, but probably only as a theoretical model, for the following reason:

    Who is going to class appliances in the following way? The client or testing practitioner? If it is the client, how will he institute the testing program? If working with a supplier, how is the engineer who goes round going to know at what point to test what item? It is impracticable for companies to invite outside contractors in at 3 or 4 points during the year to test odd items, or just kitchen equipment and tools, for example. this could actually be more costly for the client testing in this way.
    Also because of the cut price cowboys testing at 82p or similar, this system wont work with them, as they are just sticker happy and wont be interested in a system like this. The methods as described may probably be used best for in house testing by practitioners who are actually interested in what they are doing and keep proper records. For decent and knowledgeable outside contractors, who most have now been priced out of the market, how can they offer such diverse record keeping to their clients at the ridiculously low cost that the market has seemingly steeped to? Remember, clients will be having to give instruction: test this, but don't test this, etc etc which although sensible, may be rather difficult to institute in terms of practicable testing. As a responsible practitioner, I do advocate split testing of items, but sometimes clients see this as rather a hassle and opt for testing everything but at less frequent intervals. A good thing is that the cowboys will hate this method though.
    Also for test engineers visually checking appliance tagging to discover whether the items need testing or not, is time consuming and I would suggest, again that this cannot be done at current market price rate. Now this in itself is a good thing as it may mean the prices are unsustainable and have to go up - but I seriously cannot see it happening.
    But I support the fact that in theory all frequencies for different types of appliances should be different. It just seems impracticable for test engineers, clients and suppliers.

    ReplyDelete
  2. Hi Grant, thanks for your detailed reply. This method is expanding on information presented in EDition 4 of the IET Code of Practice and is not new.

    I agree with you that is ideal for the duty holder who is planning to do in-house PAT Testing and is too complicated for outside contractors.

    ReplyDelete