Pete,
Thanks so much for posting. You are absolutely right and you make some really interesting points. Apologies for the delay in responding. I have discussed your post with my colleague Paul Muir of Treework Environmental Practice (the Uk's leading expert in tree statics) and he has written the following response:
Pete Bannister is correct. A safety factor of 1.5 is what engineers would be likely to use when designing an aircraft, and higher safety factors are incorporated during the design of buildings, for example. While there is more information and a higher degree of certainty regarding the material properties applicable to aircraft construction, I would suggest that accepting a safety factor of 1.5 is more to do with cost considerations whilst also requiring the aircraft to be light enough to actually leave the ground. The engineers would prefer higher safety factors, but this would require more expensive materials, a heavier plane and greater running costs.
The aviation comparison is actually appropriate. Tree statics developed from a university study in Germany (Lothar Wessolly) that looked at lightweight structures in nature. Remember trees have to balance competitive height growth with available resources (costs), so there is an evolutionary advantage to not over-investing in safety margins. Trees only start to achieve high safety factors during the mature phase (because they can’t stop producing annual increment). Even then lower order branching will have lower safety factors – twigs break on an annual basis.
In terms of the static load test a safety factor of 1.5 is a desired minimum. I am comfortable that this is (generally) acceptable. There are a lot of margins for error within each element of the model. Most importantly, I suspect that the design wind load that the tree is required to carry is almost always an over-estimate in our calculations.
There is a further important consideration. Trees are structures that change. When considering a measured safety factor an appropriate interpretation will involve considering how much the safety factor is different from what would be expected. A safety factor of close to 1.5 for an early mature tree might be the best one can expect, even if defect free. For a late mature tree with a large stem diameter a safety factor of up to 10 or more, might be expected, so a result of 1.5 would perhaps be cause for concern. All balanced against the vigour of the tree. What potential is there for the situation to improve rather than deteriorate.
My experience is that trees with large stem diameters, like the cedar at Kingston Lacey, have very high (theoretical) safety factors (i.e. the safety factor one would expect if they were defect free), and that even though these might be extensively reduced through decay (say from a safety factor of 10, to one of 3 or 4), the residual safety factor is still higher than one might expect for an early mature defect free tree. In other words these old trees are less likely to fail. The implication is that strength loss formulae are fundamentally flawed in terms of being appropriate failure criteria. Which is stronger - a rope with a 10 ton breaking strength with half the fibres cut or a rope with a 1 ton breaking strength?