geotechguy1
Civil/Environmental
Hi all,
Curious to here opinions on the following problem, which seems to be non-standard as I've been unable to find any comparable cases in the literature.
I am tasked with calculating settlements induced by temporary groundwater drawdown for an open excavation. The excavation will be open for 30 - 60 days (this is actually an exageration as the trench is for buried linear infrastructure and will be carried out in 10-20m sections only open for a week or so). I'll present a simplified version of the ground model and excavation below, with some minor details changed
The ground profile consists of 6m of clayey silt / silty clay soil, underlain by 5m of "peat", underlain by 20m of loose to medium dense sands interbedded with sandy silts. Underneath this layer is a very dense sand extending at least to 100m depth. The peat is highly organic and typically consists of an upper high plastic / amorphous layer, then a layer of fibruos peat, then another layer of high plastic clay or amorphous peat (it depends who logs it - some of the logger seem to log anything colour black as 'peat')
The fibrous layers of peat are relatively permeable (at least laterally) while the high plastic layers above and below it tend to have permeabilities on the order of 10^-8 or 10^-9 depending on the correlation or test method used.
The excavation will be 5m deep and 2m wide with a bench at the top. We model groundwater drawdown around the excavation using either the simplified methods described in CIRIA's groundwater control manual or Martin Preene's book on the topic, or alternatively a hydrogeology team uses advanced groundwater modeling software. A key difference in this case from most of the cases I can find in the literature is that the cases in the literature involve groundwater drawdowns on a huge scale - for example, for a mine, or an entire city or region. In this case the drawdown curve is confined to a relatively small distance around the excavation (depending on how you calculate it, say, 10-20m at most) and presumably recover with time after backfilling.
THere are groundwater monitoring wells installed in the upper clay, the peat, and the lower sands. They show that the profile is not 'hydrostatic' that is, there is a different groundwater table in the lower sands which has a phreatic surface several meters beneath the contact with the peat. The groundwater table in the clay is about 1m below ground, and the phreatic surface measured in peat essentially sits at the contact between the peat and the upper silty clays / clayey silts. Also fluctuations in the groundwater levels in the upper units don't seem to be reflected in the lower sands.
Suppose you are tasked with calculating consolidation settlement at the ground surface caused by the temporary drawdown in the groundwater by the trench.
Where would you set the bottom of your model for calculating consolidation settlements?
Does lateral drainage of the upper groundwater unit into the excavation cause an increase in pore water pressure in the soil at 25m depth (i.e. 5m drawdown = 50 kPa increase in effective stress at depth)?
Does the fact that the wells show non-hydrostatic groundwater units change your answer to the above?
If you had to calculate consolidate settlement including the soils at depth, would you apply the full 50 kPa increase at the center of the excavation to the soil at great depth?
For long term groundwater drawdowns on a regional scale, or the scale of an open cut mine, the radius of influence is huge, and so obviously the full effective stress increase will end up being applied to soils at great depth, even if there is hydrodanmic lag due to the aquitards and aquicludes, but in my case I essentially end up with a small temporary drawdown cone centered around the trench.
THe different ideas that have been floated by myself and different people to solve this problem (with vigorous disagreement among all) are: A: Cut off the model at the bottom of the peat on the basis that the change in pore pressure in the upper groundwater isn't doing anything to the pore pressures in the deeper sands / silts; B: use the average groundwater drawdown within the radius of influence to calculate the increase in effective stress (so for example, instead of 50 kPa use 25 kPa); C: disagree with both of the previous statements and say that the pore pressure will change in the lower sands silts as a result of the trenching, thus causing settlement.
The difference in calculated settlements is huge - if you cut off the model at the bottom of the peat it makes the difference between 20mm of settlement and 100-150mm of settlement. This will make the difference to whether or not we can gain approval to do the work. We had the opportunity to do a previous stage of this work using Method A described above, which got us approval, and then carry out a monitoring program (surface settlement monitoring near excavations and adjacent buildings, and groundwater loggers) ; this data from the monitoring program essentially shows that nothing happened as a result of the work (ie. less than 10mm of settlement or heave which seems to be related to expansive soils rather than groundwater drawdown. Unfortunately I'm now being pushed by technical reviewers into Option C being 'correct'. Which leaves me with calculations that I know are wrong on the basis of field experience thus preventing an easily constructable project from being constructed.
(And yes the government agency and industry here is very pedantic so they're not going to let me handwave or 'experience' my way out of it)
Curious on thoughts. Can do up a sketch later if that helps
Curious to here opinions on the following problem, which seems to be non-standard as I've been unable to find any comparable cases in the literature.
I am tasked with calculating settlements induced by temporary groundwater drawdown for an open excavation. The excavation will be open for 30 - 60 days (this is actually an exageration as the trench is for buried linear infrastructure and will be carried out in 10-20m sections only open for a week or so). I'll present a simplified version of the ground model and excavation below, with some minor details changed
The ground profile consists of 6m of clayey silt / silty clay soil, underlain by 5m of "peat", underlain by 20m of loose to medium dense sands interbedded with sandy silts. Underneath this layer is a very dense sand extending at least to 100m depth. The peat is highly organic and typically consists of an upper high plastic / amorphous layer, then a layer of fibruos peat, then another layer of high plastic clay or amorphous peat (it depends who logs it - some of the logger seem to log anything colour black as 'peat')
The fibrous layers of peat are relatively permeable (at least laterally) while the high plastic layers above and below it tend to have permeabilities on the order of 10^-8 or 10^-9 depending on the correlation or test method used.
The excavation will be 5m deep and 2m wide with a bench at the top. We model groundwater drawdown around the excavation using either the simplified methods described in CIRIA's groundwater control manual or Martin Preene's book on the topic, or alternatively a hydrogeology team uses advanced groundwater modeling software. A key difference in this case from most of the cases I can find in the literature is that the cases in the literature involve groundwater drawdowns on a huge scale - for example, for a mine, or an entire city or region. In this case the drawdown curve is confined to a relatively small distance around the excavation (depending on how you calculate it, say, 10-20m at most) and presumably recover with time after backfilling.
THere are groundwater monitoring wells installed in the upper clay, the peat, and the lower sands. They show that the profile is not 'hydrostatic' that is, there is a different groundwater table in the lower sands which has a phreatic surface several meters beneath the contact with the peat. The groundwater table in the clay is about 1m below ground, and the phreatic surface measured in peat essentially sits at the contact between the peat and the upper silty clays / clayey silts. Also fluctuations in the groundwater levels in the upper units don't seem to be reflected in the lower sands.
Suppose you are tasked with calculating consolidation settlement at the ground surface caused by the temporary drawdown in the groundwater by the trench.
Where would you set the bottom of your model for calculating consolidation settlements?
Does lateral drainage of the upper groundwater unit into the excavation cause an increase in pore water pressure in the soil at 25m depth (i.e. 5m drawdown = 50 kPa increase in effective stress at depth)?
Does the fact that the wells show non-hydrostatic groundwater units change your answer to the above?
If you had to calculate consolidate settlement including the soils at depth, would you apply the full 50 kPa increase at the center of the excavation to the soil at great depth?
For long term groundwater drawdowns on a regional scale, or the scale of an open cut mine, the radius of influence is huge, and so obviously the full effective stress increase will end up being applied to soils at great depth, even if there is hydrodanmic lag due to the aquitards and aquicludes, but in my case I essentially end up with a small temporary drawdown cone centered around the trench.
THe different ideas that have been floated by myself and different people to solve this problem (with vigorous disagreement among all) are: A: Cut off the model at the bottom of the peat on the basis that the change in pore pressure in the upper groundwater isn't doing anything to the pore pressures in the deeper sands / silts; B: use the average groundwater drawdown within the radius of influence to calculate the increase in effective stress (so for example, instead of 50 kPa use 25 kPa); C: disagree with both of the previous statements and say that the pore pressure will change in the lower sands silts as a result of the trenching, thus causing settlement.
The difference in calculated settlements is huge - if you cut off the model at the bottom of the peat it makes the difference between 20mm of settlement and 100-150mm of settlement. This will make the difference to whether or not we can gain approval to do the work. We had the opportunity to do a previous stage of this work using Method A described above, which got us approval, and then carry out a monitoring program (surface settlement monitoring near excavations and adjacent buildings, and groundwater loggers) ; this data from the monitoring program essentially shows that nothing happened as a result of the work (ie. less than 10mm of settlement or heave which seems to be related to expansive soils rather than groundwater drawdown. Unfortunately I'm now being pushed by technical reviewers into Option C being 'correct'. Which leaves me with calculations that I know are wrong on the basis of field experience thus preventing an easily constructable project from being constructed.
(And yes the government agency and industry here is very pedantic so they're not going to let me handwave or 'experience' my way out of it)
Curious on thoughts. Can do up a sketch later if that helps
