Quantified Tree Risk Assessment (QTRA) - Questions & Answers

[Acer ventura]

Perhaps this may be a little on the extreme side for those on here that are not that way inclined Jules

 

I couldn't agree more Jonny. No one want's to see premature quantification.

 

Cheers

 

Acer ventura

[daltontrees]

Oh dear, I see we are in innuendoland! I had thought the general thrust of my argument would have over-ridden everyone's individual urge to fixate on the nuts and bolts.

[Acer ventura]

Acer

Sorry I had not looked at the whole thread...

For me human life represents max target value and I always work from this assumption, does QTRA have a value order? Also would QTRA take into consideration that a motorcar travelling at 60 mph would susstain more serious damage in the unfortunate event of a collision with a falling tree or branch than it would at 30 mph, but actual contact for the faster car would be less probable?

 

I would like to take a closer look at QTRA but can not even find the PN, quite useless realy.

 

Hi Jonny

 

I'm just sorting out a worked example for btggaz, but popped into answer you becasue it's an easier question. I would recommend starting at the beginning of the thread. Break it up into small sessions. You may even find the PN, which answers your questions. In short 1 occupied car = 1 life. Whether the car cannot stop in time before hitting the tree/branch and/or if the tree/branch hits the car.

 

If you can't find the PN, you could always try this.

 

Let me google that for you

 

Cheers

 

Acer ventura

[JonnyVine]

Jules

I'm terribly sorry, I just felt that the thread might have benefitted from a charge of humour?

There appears to be a stalemate situation regarding Q&A, the Q's are in abundance but sadly the A's have dried up? I am always suspicous when answers are hidden or maybe don't even exist.

Perhaps it has got to a point where premature quantification may be better than no quantification at all.

 

Acer

Forgive me if the questions I asked have already been dealt with, if so could you direct me to the post.

 

Many thanks

Jonny

[JonnyVine]

Acer

We must have both been typing at the same time and crossed over.

 

Perhaps I failed to explain properly as your answer dosen't fit?

 

Lets try again

1/ Does your system have an order of value? For example, I would put a higher value on a diseased Aesculus overshadowing a childrens play area than if the same area was used merely to exersize dogs.

 

2/ I see the word "probability" as being quantifiable so the motorcar travelling at 60mph would spend half the time in the same place as a car doing 30mph but at 60mph it would be more likely to be fatal. What provisions exist inside the workings of QTRA to quantify this theory?

 

BEst Wishes

Jonny

Edited February 16, 2013 by JonnyVine
Spelling

[Acer ventura]

I've been following this thread with great interest. Although I can understand the reasoning and quantifying elements assessed by QTRA, I can't still have a moral dilemma with the entire process.

 

Hi btggaz

 

Glad to hear you’ve found the thread interesting, and you’ve raised some really profound stuff with your posting. So please bear with me whilst I explore the issues with you.

 

I think a particularly important benefit of quantifying tree risk, so we can relate it to a published level of tolerable or acceptable risk, is that the ‘moral’ side of societal risk, at 1/10,000 and 1/1,000,000, has been decided by higher authorities and experts whose specialist field is risk management.

 

If you think about it, if we don’t’ quantify risk relative to a published threshold we are taking a ‘moral’ decision into our own hands. We’re playing god, and deciding a level of risk from a tree which could kill someone, and worse still we’re not sure what that level of risk is. I think this is one of the reasons why arboriculturists are understandably hard-wired to be risk averse rather than risk aware, because the last thing we want to happen is that someone could possibly die from the failure of a tree we’ve assessed or worked on, irrespective of the level of risk. So naturally, we habitually look at what could happen, rather than what is likely to happen.

 

A couple of years back' date=' I was involved in an application and an appeal to fell a large ash. The tree had a history of dropping limbs and had a large limb over the road way with a diametric crack almost two meters long. The road was a private access to the rear of twenty odd houses.

So, the the odds of hitting someone was low, the risk of harm (death) was high as the limb was 4-500 dia, and the potential of failure was very high.[/quote']

 

Just so we’re all talking about the same thing, what you’re calling the risk of harm (death) here, because of the size of the limb, is called Impact Potential (Size of Part) in QTRA. The Risk of Harm (RoH), the probability of death or harm, is the end product.

 

It is possible to have a high PoF, a high Impact Potential (Size of Part), but a low RoH if the Target value is low enough. It’s difficult to comment in detail about this case without seeing the condition of the tree, but as I have been asked for worked examples, what I can do is show you is neat trick with QTRA that enables me to show you how 'fecked' that branch would have to be, in terms of PoF, before the risk is unacceptable. I’ll have a play with the QTRA manual calculator and software to show you and post this later.

 

The appeals inspector stated that it would fail.

The tree officer told me that the odds of failure were lottery winning figures that it would hit someone' date=' just as a neighbour drove past.

"Try explaining to her family just how unlucky she was" was my answer.[/quote']

 

It may appear a difficult, but how far do you take this. At one extreme we fell all trees in case someone is 'unlucky'. On the other hand we do nothing to manage the risk. Somewhere between these to two extremes is a reasonable level of risk, and we know what a reasonable level of risk is because of the many benefits that trees provide. It’s 1/10,000.

 

Just to give you some perspective on this. The risk of harm from a traffic accident is around 1/300 (2011 figures). Yet this is a risk parents are happy to subject their children to on a near daily basis. The risk of death from tree failure is around 1/10,000,000. The RoH from trees is so low Insurance companies don’t even have what’s called a ‘call code’ for it. In other words, they don’t get their actuaries to crunch the numbers for tree-related damage or injury from failures to see whether they are exposed to risk of pay outs themselves. It’s all bundled under weather related risk.

 

I do find it difficult' date=' in that particular situation, to accept a tolerable level of risk. Like the picture of the cars in the car park. If they'd been occupied, would 1:400,000 still been acceptable?[/quote']

 

I understand where you’re coming from because it’s a natural tendency to think ‘what if’. Fear is a really powerful emotion and once that button has been punched, it’s difficult to ignore. This is why such events are commonly reported in the news. Fear is drama, and drama is news. With tree failures what could have happened is often far more exciting than what did happen.

 

In reality, with most car park settings the value of the Target is the car because the time it’s exposed to the risk exceeds the comparatively short period in time of getting in and out of the car. Yet again, we should be looking at what is likely to happen, rather than ‘what if’, or could happen. We should thinking like Pooh rather than Piglet. Off the top of my head, I can’t recollect a tree-related death, or severe injury, to someone in a parked car, but I can easily recollect many photographs of parked cars damaged by falling trees and branches.

 

Interestingly, in this case one of the cars was occupied by two people who were leaving the car park at the time. They saw the tree begin to fail, and in their panic drove into its path. Neither was injured, though one of them did need to wait until some branches were cut in order to get out because she was elderly.

 

Does all that help?

 

Cheers

 

Acer ventura

[Acer ventura]

 

the Q's are in abundance but sadly the A's have dried up?

 

Hi Jonny

 

How very dare you. I am trying to answer the questions, time permitting, in the order in which I get them. I'll queue jump you because it's an easier answer than getting onto WorcsWuss's calculation of bending moments. However, though it's great that you're interested and are asking questions, I really recommend you take the time to have a look at the PN because it may contain the answers to your questions, and if it does it would help with the backlog on here.

 

Perhaps I failed to explain properly as your answer dosen't fit?

 

Lets try again

1/ Does your system have an order of value? For example' date=' I would put a higher value on a diseased Aesculus overshadowing a childrens play area than if the same area was used merely to exersize dogs.[/quote']

 

Yes humans are valued higher than dogs. But children aren't valued higher than adults because they are no more at risk than adults; ie their behaviour within the falling distance of a tree or branch doesn't increase their risk of being hit. Similarly, an adult isn't less of a Target value than a child. It’s the occupancy time that determines the Target.

 

2/ I see the word "probability" as being quantifiable so the motorcar travelling at 60mph would spend half the time in the same place as a car doing 30mph but at 60mph it would be more likely to be fatal. What provisions exist inside the workings of QTRA to quantify this theory? .

 

30mph is not half the time of 60mph because of stopping distances. We don’t try to reduce the likelihood of a fatality based on speed. A car driving at 100kph has an occupancy time of 3.95 seconds. A car driving at 50kph has an occupancy time of 1.85. If there is 1/1 Impact Potential within these time frames then the consequences are calculated as a fatality for both speeds. Even if the failure is in the 50kph zone and the branch/stem falls onto the road at the outermost extent of the stopping distance, so by the time the car hits it it’s almost stopped. In other words there’s a considerable safety margin built in because we’re calculating on the worst outcome.

 

Is that what you were getting at?

 

Cheers

 

Acer ventura

Edited February 16, 2013 by Acer ventura

[JonnyVine]

Acer

Unfortunately, in most cases of accidents involving trees and motorcars the driver rarely has time to apply the brakes for two reasons, Field Of Vision (FoV) due to the windscreen and roof and the fact that most car drivers are concentrating on the road so braking distances are only relevant if the object is already on the ground.

 

A car travelling at 30mph is covering 13.4112 meters pre second, the car travelling at 60mph is covering 26.8224 meters per second exactly double, so the faster car would be in the contact area for exactly half the time of the slower car.

 

I'm sorry if I fail to explain the value concept, do you have a value order? Differentiating between children and adults is futile, they are both human beings.

 

Simply, in order of priority (hypotheticaly)

Human life=1

Buildings=2

Race horses=3

Garages with lambouginis=4

Cows, chickens, goats etc ????

 

 

And I dare because this is a debate and all opinions are equal, no?

 

To quote Dr Lonsdale yet again "Philosophical ramblings are not much use to someone trapped underneath a fallen tree".

 

Enjoy your trip to Australia

Jonny

[daltontrees]

Hey JonnyVine, don't worry about offending me, I had just come in from the pub and thought I was being funny in my innuendoland reply. I really ought to use those emoticons so you knoew I was saying "thrust" in a Benny Hill tone of voice.

[Acer ventura]

A couple of years back, I was involved in an application and an appeal to fell a large ash. The tree had a history of dropping limbs and had a large limb over the road way with a diametric crack almost two meters long. The road was a private access to the rear of twenty odd houses.

So, the the odds of hitting someone was low, the risk of harm (death) was high as the limb was 4-500 dia, and the potential of failure was very high.

The appeals inspector stated that it would fail.

 

Hi btggaz

 

Here’s the worked example I promised earlier.

 

Target (T) - We always start with the T because it’s the most important part of the risk calculation. If we allow for at least one car journey there and back for each of the 20 odd houses every day of the year, and some more to allow for visitors and additional trips and suggest 65 car journeys/day, which is corroborated as reasonable by those who now the site better, we end up with the highest value in T, Range 4.

 

Impact Potential (IP) - Let’s go to the upper end of your estimate of branch diameter 400-500mm, so IP is in Range 1.

 

Probability of Failure (PoF) - Now we have T and IP, we can move the inner PoF component around to see how ‘fecked’ the branch has to be before we get an unacceptable level of risk. If the PoF is in Range 3, 1/100 – 1/900, which is a high PoF, the RoH = 1/70,000 (Risk Index 70).

 

 

The current manual calculator doesn’t have PoF Range 2, 1/10 – 1/90 on it because we didn’t have the space, but it’s easy to work out. A PoF Range 2, 1/10 – 1/90 is x10 higher than Range 3 because it’s the highest value in the range that we input. 10 is a x10 higher PoF than 100. Therefore, because the PoF is x 10 higher the RoH is x10 higher.

 

Back to the calculation. If the PoF Range is 2 then RoH = 1/7,000 (70,000/10), which would be regarded as an unacceptable level of risk.

 

What all this means is that from knowing the T and IP we can work out the branch has to have a PoF in Range 2 before the risk becomes unacceptable. For a branch of 400-500mm to have such a high PoF it would have to remarkably defective and be displaying incipient indications of failure with no significant adaptive growth.

 

This is a very straight forward skill after training because all you’re doing it aligning the components on the manual calculator. It’s much easier to talk through in the workshop, and do in the field than write out on here. All it takes is a bit of detective work to put in credible T values based on those who know the site.

 

Now if these are posh houses with two cars and we expect more than 65 car journeys/day, with the manual calculator we would move up to the next T range. With the software calculator we’re not constrained by the size of the manual calculator and have more space, and we could look to refine with say 100 vehicles/day, and see the RoH for the same branch with a PoF Range 3 is 1/50,000.

 

 

Cheers

 

Acer ventura