Independent technical advice
on roofing and cladding
Roberts Consulting
Back to the home page | Details of our company approach and the services we offer | Since 1992 we have written more than 80 articles published in the UK and North America | We work on a number of initiatives to improve the performance of roofing and cladding. | Links to Institutions and other Roofing Industry Organisations. | Contact Information.  


A common difficulty faced when working close to unprotected edges of a roof is finding a secure point to which a harness and lanyard can be fixed with confidence. In recent years there has been a significant growth in the number of new roofs where anchor points and running line systems are provided above roof level, to which personal protective equipment can be attached.

The trend to provide greater safety provision for roof work is welcome. However, recent site observations raise serious concerns about the security of some anchor points, and these are discussed in this Technical Note.

HSG 33

The Health and Safety Executive book HSG 33 entitled 'Health and Safety in Roof Work' was published in 1998 and gives considered and practical advice.

The safe performance of a fall arrest system depends completely on a suitable anchorage being provided. The adequacy of all anchorages, including the ability of the supporting structure to carry the anchorage loads, should be verified by calculation or by testing. Appendix 3 of the book concludes:

'Fall arrest systems are not foolproof and their safe usage is not always common sense. Without proper training and fitting, use, maintenance, installation and equipment limitations, all that a fall arrest system can provide is a false sense of security.'

Site Observations

During a recent roof inspection it was pointed out that an anchor post securing a running line system was no longer vertical, as shown in photo 1. On closer inspection it appeared that the baseplate securing the post to the concrete deck had come loose. The anchorage system was unsafe.

On a second roof seen a couple of days earlier the galvanised steel anchor post had been fixed to a metal deck, in advance of the insulation and membrane being laid. As seen in photo 2, four rivets were missing from one side of the baseplate, presumably because the plate was too small to sit onto three crowns of the decking profile. As a result, if the top of the anchor was pulled towards the right then the safe anchorage would be totally reliant on just two rivets.

From the subsequent desk study it was confirmed that the pull-out strength of the two rivets fixed into the 0.7mm thick steel deck was inadequate. Since the posts were not fixed directly above the steel support beams, a twisting type deflection could be expected in the lightweight deck. Also, the shot fired studs holding the deck down to the steel frame had inadequate pull-over resistance to overcome the local over-turning moment.

The example is a particularly alarming case as once the insulation and membrane had been laid the means of attachment would have been hidden whilst the anchor point would appear to be safe. It is critical that for fixing into the building structure, the design and installation of the whole system should be verified by a qualified engineer to be capable of sustaining the test force.


The relevant British Standard is BS EN 795: 1997 which defines the requirements and testing for anchor devices. For example, where there is a horizontal rigid anchor line there is a recommendation for a static test with a force of 10 kN applied in the direction in which the force can be applied in service. The force should be maintained for three minutes during which time the anchor device should hold the force. If more than one person may use the device at the same time, then the test load is increased.

10 kN or 1 tonne is a significant sideways force. It is of the same order of magnitude as the force in a rope pulled by a 5 man tug-o-war team! The engineer needs to ensure that this lateral load and over-turning moment can be transferred down into the structural framework of the building.


Thankfully there are several firms of well established specialists who design, install and test safety related equipment. The examples of bad practice are hopefully isolated, but they are a timely warning to our industry. Potentially a man's life is at risk if the anchorage is not secure.

If we cannot be confident that the anchor point can take the applied load, then we would be better off not having the anchor point at all.


1. HSG 33, 'Health and Safety in Roof Work', 1998 (to order: telephone 01787 881165)
2. BS EN 795: 1997 'Protection Against Falls from a Height - Anchor Devices - Requirements and Testing'.

© Keith Roberts, 2000

First published in Roofing Cladding Insulation, April 2000

RCI Technical Notes

If you require any further information about this topic please contact us.