Trusting N values = RISK^100
I’m back!
After some weird issues regarding my account, I can now post blogs again albeit I’ve had to get a new account linked to my work email in order to do so. So when I’m finished at WYG we’ll have to try and set up another account on my personal email so I can continue to see the blog – one for Jim to ponder…
Anyway, about three weeks ago I managed to get out of the office for two days. One day I spent on the Thames on a boat – more on that later. The other day I spent in a field.
I’m designing a couple of crossings over high pressure gas pipelines to provide access for vehicles into a quarry. You may have seen my previous blog where I was tearing my hair out over the vehicle loading. A number of site investigations had been completed previously but had focussed on the agricultural or quarrying value of the site. None had collected any geotechnical properties and I need them to complete the design. So I organised a local Geotechnical engineer to go along and do some boreholes.
I got there to find two blokes (the rig operators) stood by the side of their vehicle, looking at the farm land which is the site, shaking their heads. I said hello and they said “We’re not going on there”. Good start. It was a bit wet, I accept that, but what was he expecting?! The engineer turned up shortly after and we walked the site while we waited for the National Grid bloke to turn up. In order to plant a daffodil within 15 meters of a high pressure gas line you need National Grid to send an engineer to observe. I wanted the bore holes as close to the location of the crossings as possible, obviously. The rig operators didn’t want to go onto the field. Their argument was that it was too wet and the rig would sink. I really doubt it would of, but I was never going to win that one, so we came to a compromise. We selected locations for the boreholes that were on the firmer ground at the edge of the field. This meant that the locations were now more than 15 meters away from the gas pipe and therefore the gas man had driven for 3 hours to get to us only to find he now no longer needed to be there. He checked the locations, gave me a piece of paper saying it was all ok and then left, shortly followed by the Geo engineer who let his rig team get on with it.
![IMG_0034[1]](https://pewpetblog.com/wp-content/uploads/2016/02/img_00341.jpg?w=595)
John previously said that once you see a drilling team do their thing you’ll never trust any of their results ever again. I now agree…
Let’s take N values as an example. The book (CIRIA R 143, among others) says that to get an N value from an SPT you hit the stick with a weight and measure the number of strikes to move the stick 300mm (very simplistically). The report also states that poor operator technique can produce disturbances in the soil below the head that then leads to unreliable results, especially in granular materials, the bigger the particle size the bigger the effect. I’m working in gravel. Oh good then!
![IMG_0048[1]](https://pewpetblog.com/wp-content/uploads/2016/02/img_00481.jpg?w=595)
I watched the crew conduct an SPT at around 4.5 meters’ depth. They drove the collar to about the depth they wanted and lowered a weighted tape down the hole to measure the depth. But they didn’t measure the amount of collar sticking up out of the ground and so don’t know the difference. The point with SPT is that the material is undisturbed. If you don’t know the difference in level between the bottom of the collar and the test level, how can you possibly estimate the amount of material that has fallen from the sides of the hole? And this is a particular problem in large particle rounded non-cohesive soils like gravel.
The operators then introduced the rod into the hole and drove it “for a bit”. Not the stated 150mm seating drive. Additionally, BS 1377 states that the seating drive should be limited to 25 blows. I must admit I wasn’t counting, but I’m not confident it was. They then got a stick with some lines on it, held it against the rod and drew chalk lines on the shaft (cue Brad giggling) where the lines were (at about 75mm intervals). They then drove the rod noting how many blows it had taken to reach each line – or there about. The lines weren’t exactly straight but at this stage that was the least of my concerns. The operator then wrote all this down in his book and made a cup of tea while his mate changed the head over.
The crew also took samples at various depths including both the gravel and the underlying clay, which they did pretty well from what I understand. They put the coarse-grained stuff in bags and for the cohesive soils the sealed it in wax to keep the water in and treated it pretty carefully.
![IMG_0071[1]](https://pewpetblog.com/wp-content/uploads/2016/02/img_00711.jpg?w=595)
So what?
So I’ve got some N values and I’m awaiting the lab test results. Do I weight the reliability of the lab test results against the N values since the method of collection looks dodgy? No. There are loads of other factors that affect the accuracy of the N value including the stiffness and exact diameter of the rod and the angle of drive, but I didn’t measure them. So who am I to judge? The factor of safety applied to the results is there for a reason and as long as I try to triangulate the properties with the lab results and anything else I can find I’ll crack on and use them. I’ve got nothing else so I need to use something!
Go on then John, I’m braced!
Stay tuned for… “Messing about on boats”
Hi Guz,
The first few weeks at Arup was all about specifying ground investigation requirements. One of the key points was how practically to site a cable percussion rig – the site being the Thames and the surrounding area being roads. It is therefore interesting to hear your team had issues with a big open field, even if it was muddy.
I assume you want to get phi’ for the granular material. Did you specify other tests – trial pits or in-situ shear tests? Although I imagine you would say they would have been even less reliable. What laboratory tests did you specify? Have you been able to get a reasonable angle of shearing resistance from the soil classification?
Was 4.5m the full depth? Or did you go deeper? Had you thought of any other means of ‘hitting the ground with a stick’? I have specified the use of dynamic probing but not sure that will be any more reliable that what you have described. How long did the task take, including setup and execution?
Belt fed questions!
Full depth was 15m. There was talk of a piled foundation but I’m hoping to avoid it since there’s no requirement to limit long term settlement since the load is short term.
In total it took them 4 days, but they only did two of the requested 3 holes since even I had to accept they’d have trouble reaching the 3rd location.
Effectively I’m planning to take this as a shallow foundation problem so I’m looking for phi-dash, c’ and that squiggly thing that represents the self weight. If I can get it to work within the sandy gravel layer then I’ll stick with that.
I got a couple of lab tests done but haven’t kicked the ass out of it since there’s not really any money in the budget for it so I’m just going off what I can.
Which is obviously a false economy, the structure will probably cost them more in the long run since the FoS will be larger than if you had better geo results. But its not my money…
And as for the access bit, the drilling crew were just that sort of people. Had an excuse for everything and it was all just a little too difficult.
Guz – thanks for this. Wow 2 days a hole and only 15m that is a useful benchmark. I love the comparison at the end, short term gains and all that! What was the outcome on vehicle loading for out of the normal scope of size vehicles?
I’ve used SV100 loading from the NA to the EC.