We will try and answer most of the questions in Mark's OP. But we are not going to go back and check every detail so forgive us if some of the values are not precisely correct or we use 'around such and such date'. We are also not going to respond to some comments touching on matters which became 'political' for fear of fanning the embers and creating yet another distraction. And we can't comment on HKD as we have no experience with them. But if people want to do some work on them, then BCA's anchor puller can be made available. We will say that in our opinion E&T has been working reasonably well for the past couple of years; it still has ups and downs but has mostly been able to resolve differences without letting them become 'political'. But no doubt some people will consider us to be biased; in which case come and participate in a meeting, especially if you can help. Meetings these days are usually held on line so travel problems & time is minimised.
First Mark clearly comes from an industrial background with considerable experience. He cites EN795 which is where the first and major misunderstanding arises. EN795 applies to anchors used in work situations. The current version of EN795 (2012) clearly states in its Scope that the standard does not apply to any anchors used in sport. That is because EN959 was developed in 1996 (at the same time as EN795) and subsequently reissued in 2007 covering mountaineering anchors.
Yes we are trying to 'cover our backsides' by basing our adopted approach within a standard. Surely rightly so? Mark comments about compliance with health and safety regulation. Because as he says, it often uses the word 'should' it allows for alternative ways to achieve compliance from that identified in the regs. So the claim is anchors which meet EN959 are as likely to be as good as those which meet EN795.
There is a fundamental difference between the two standards. To paraphrase, EN795 requires anchors to be placed and tested (that is pulled to a certain level) and then subsequently examined (that is look at them) and tested on a 6 or 12 monthly basis. EN959 does not. To paraphrase, it only requires them to be placed and then left. This approach evolves from a totally different set of key criteria for the anchors. They are whilst an EN795 anchor should sustain 12kN load to demonstrate the design meets EN795, EN959 requires anchors to sustain 15kN for axial and 25kN for radial loads. Simply put EN959 anchors are required to meet a higher load so it can be placed and not routinely examined and not tested. (But it still MUST be examined pre use.)
We are afraid we don't recall us ever referring to a 10kN criterion. The values are 15kN (axial) and 25kN (radial). But the UIAA standard was change in 2013 to specify a 20kN axial test and a torque test to check for non rotation (see
http://theuiaa.org/documents/safety-standards/UIAA_123_Rock_anchors_March_2013.pdf ). We await that feeding through into the EN959 standard.
There is also a further significant difference between the two standards. EN795 requires the anchor to withstand a given load without distortion. EN959 makes no such requirement and actually states permanent deformation is permissible in testing. So the comment by Mark about the distortion of anchors they placed whilst testing results from the simple fact that that they are not designed to be tested at 6kN without distortion. BCA E&T purchased a small number of testers in the early 2000's to undertake testing but around 2007 issued advice to regions to stop testing because of this fact. Yes the surface resin will usually be slightly damaged by testing to 6kN. Fortunately the Eco, Peco and BP anchors usually start to show a permanent set above 10kN. If Derbyshire wishes to maintain post placement testing fine; but they must live with that decision and also cope with the simple fact that the anchors were not designed to be so tested.
In addition, we would observe that the original NCA / BCA scheme required an ongoing inspection program. The volunteer resources required for this became overwhelming so that around 2007 this was dropped in line with EN959. The only inspection requirement is by the user, prior to use. Hence the absence in the policy document to routine inspection (as opposed to pre use inspection) and testing.
One thing which Mark said which confuses me is the reference to SWL. So far as we can recall we have never used that term. SWL (these days better known as ?WLL? for ?Working Load Limit?) is a concept which applies in the mechanical lifting industry but, as Mark indicated, has little application in the PPE/roped access sector. We presume he used that term to refer to something we have adopted, namely the 5% fractile limit. This arises from something we are rather concerned about within EN959. That standard only requires ONE anchor to be tested to show compliance. (We have checked with persons involved in setting the standard to confirm this understanding.) That seems completely at odds with standard engineering practice (see BS EN 1990:2002 'Eurocode - Basis of structural design) which talks about testing an albeit small sample. Fischer also talk about testing a sample (see page 27 in
http://www.fischer.co.uk/PortalData/1/Resources/service/sales-documents/documents/fischer-THB-INT-110640-082013.pdf ). Both talk about using a value 'off set' from the mean or average value. Simply put when you test a batch of anchors you get a spread of results. Usually that set of results will conform to a normal distribution or a bell curve around a mean value. Citing the mean value means that 50% of the population will have values lower than that mean value which sounds inappropriate. The standard engineering practice is to use a value computed using an assumption of a normal distribution where only one in twenty of the population will have a value below it. That value has been taken as minimising the risk of failure to an acceptable level. It is not a SWL which indicates a safe usage value.
Reference has been made to the Collinox anchor as a suitable solution. The major problem with this links back into the debate on 316 v 304 stainless steel. That debate arises from a concern expressed some while ago over chloride stress corrosion cracking. Despite our efforts (read paper at
http://british-caving.org.uk/wiki3/doku.php?id=equipment_techniques:meetings and also meeting minutes) several regions (including Derbyshire and the Dales as well as Mendip) have decided to reject using 304 which the Collinox anchor is made from. We will not go into the detail but we would welcome someone to come along and present more technical evidence which will persuade the nay sayers to accept 304 for most locations. We admit we failed.
The choice of anchors has been a major problem for BCA. To review, DMM stopped making Eco anchors around 2004, so we bought the remains of their last batch, some 700 anchors as we recall. We then tried developing a copy of the Eco, denoted the Peco but whilst the trial batch appeared OK, the first (and only) production batch had several failures where the metal broke at relatively low values (around 12kN as we recall). Although the others which failed in the resin / rock had an acceptable strength, we concluded after much debate that we could not trust the metal. So we went back to square one. The Bolt Product (BP) twisted shaft anchor was introduced to us and on testing we found it had more than adequate strength. So we invested in a large batch of them. Regrettably the first delivery was incorrectly supplied by the manufacturer in 304 but we have had replacements in 316. (Although we have used a fair number of the 304 BP twisted shaft anchors for testing and training.)
The costs differences between BP and Collinox are substantial. In round terms and including resin costs, we believe we spend about ?7 per BP anchor compared to ?15 for the Collinox. We have not considered seeking a quote or a large batch of Collinox anchors.
One point Simon has brought out in developing the IC anchor is that strength is not the only factor. Although BP claim their anchors can be drilled out, the testing experience does indicate the potential for loss of 'location' and a conservation issue. We have started work to confirm an extraction process which does not loose location. As has been mentioned elsewhere, we could have gone for the BP 6mm twisted shaft anchor. A point we have just realised from testing work conducted in the past month. (Hindsight is fairly accurate.) In response to a request we now have a couple of expressions of help to find a manufacturer of the IC anchor which we are following up on. So whilst we have not solved the problem we are in a much better position.
A number of comments relate to what we would term 'exceptions'. So most anchors are placed such that they are seen before use. Hence the advice inspect before use does work. Whilst the advice on using two anchors is sound, there is the exception of single anchor rebelays. It was the failure of one such rebelay back in the '80s that resulted in one person sustaining three crushed vertebrae from what was a fairly short loop length of ?slack? rope.
Simon suggested that we should remove the +/-1mm movement criterion. He argues any movement is obvious. That was not the experience of those who followed up reports of loose anchors in the past. E&T has currently left the debate open and perhaps it might be resolved at its next meeting. We suggest the nub of the problem is, do we keep things simple (permit slight movement to allow for people confusing movement in their fingers with movement in the anchor) or do we either say, use a tool to check for looseness or do we accept reports which turn out to be due to people mistaking looseness?
Obviously exposing such discussions to all leaves us exposed to critical comment - but would the caving community prefer that we moved to the other extreme of debating behind closed doors and telling you what to do? One thing which is worrying is the number of loose anchors found by Simon in the Dales. We say 'worrying' not from the numbers as so far which we gather still represents a percent or so of those placed but from the perspective that users are not reporting loose anchors.
We agree correct installation is the key to a good anchor. However regions differ in their view of what is correct or perhaps a better phrase is 'best'. One example is Derbyshire insist on an 'on site committee' meeting to decide on the precise location of an anchor. (That sounds grander than no doubt it actually is but fundamentally they want several people to be involved.) Other regions permit a single person to decide on the location having agreed that the cave needs anchors. Mark makes an interesting point about the ease of use of Collinox resin. Although it is likely to be acceptable, we would like to check its performance under wet conditions. We did check the current recommended resin (Fischer FIS V 360 s) and found it drops a few kN when used on flooded holes. If that comes out OK, then we think Mark will have made a valuable contribution.
Nick Williams - E&T Chair
Bob Mehew ? E&T Secretary
