Secondary Compression Reduction - Peat

[jcm1890]

Hi all,

I'm currently working on a highway realignment project. The general stratigraphy is peat (up to 2.8m thickness) overlying glacial clays. The contractor is keen for all peat to remain in situ, and therefore high settlements are anticipated. We are proposing preloading to remove post-construction consolidation settlements.

My question is regarding reducing or removing secondary compression effects post-construction. I have reviewed various literature including Alonso et al 2000, and Oliveira et al 2016, citing references such as Ladd 1971. These cover the effect of preloading up to 150% on the secondary compression coefficient ca (strain / log time), however are limited to oedometer testing up to approx. 15000 minutes. Particularly Oliveira indicates that the impact upon the ca coefficient reduces over time (approx. 95% ca reduction short term for up to approx. 500 minutes, to only approx. 50% reduction beyond this).

I was wondering if anyone has experience of preloading of peat and its impact upon secondary compression, or have any relevant references regarding the efficiency of preloading (or other pre-treatment measures) for removing or reducing these settlements.

Thanks,
J

 

[fattdad]

calculate how much settlement will occur during the time frame of the project - let's say 30 years. That settlement (let's say 4 ft) should include all primary consolidation and secondary compression. Design a surcharge to develop 4 ft of consolidation - instrument the fill. Place fill and surcharge wait for the consolidation to occur and remove the surcharge. Now the fill will, "Feel" as if it has settled more than it should under the existing load. That is essentially driving out the secondary compression for the design life.

That's how I approach the problem.

f-d

ípapß gordo ainÆt no madre flaca!

 

[oldestguy]

OG here with comments. I have done many projects where preloading has been used. Most of these are for buildings. One of the considerations is the pressure coming from footings. All the sites have had fill added to get up to a "suitable" grade. So we have both the weight of permanent fill as well as footings. While we may run consolidati0n tests to get a feel for what the settlements might be, all sites have settlement platforms and have periodic elevation readings taken. Sometimes some limited undercutting is done at footing locations. The thickness of organic soil as well, as sites with misc dumped fill has been roughly 20 feet to 90 at one site. The field readings are plotted on semi-log charts. Once settlements have been going for some time after loading, the semi log plot is a straight line, usually. The difficulty that each site has is is where do we consider a single instantly applied load occurs as compared to the long term required for placing the surcharge. With that difficulty, generally some conservancy is applied. As the surcharge sits there and readings are in a straight line, that line is projected out to some assumed "end date", something like 100 years. By assigning some long term pressure ratio of loaded situation to current surcharge conservatively, against this straight line at 100 years, a point of settlement is picked off the plot as to when surcharge can be removed. Due to the unknowns and the difficulty mentioned, some "judgement" has to be applied, such as what differential settlement of the building might be tolerated, etc. You will see no separation of primary and secondary settlement is involved with the settlement readings, but likely most of the straight line segment is secondary. Sites usually have some peat over organic silt. For a roadway with only a few feet of peat and better soil below, moving the peat off to the side, outside the fill for the road is a common practice in my experience. There usually isn't time for surcharging, although it may only require a month or so for thin peat.An added item: The ratio of permanent load to permanent plus surcharge is significant, so the higher the surcharge, the shorter the loading period. For a movable surcharge to be raapidly moved along, it has to be quite high, sometimes over 10 feet high.

 

[AnotherDirtGuy]

With deep peats, do you also need to consider long-term settlement from the decomposition of the organic fraction?

Another thought, if the concern with secondary consolidation is differential rather than total settlement perhaps this can be addressed with layering geogrid in your fill and/or subbase.

 

[Mad Mike]

I would agree with the decision to leave the peat in place and work over it, due to the layer thickness and presumably a water table somewhere in the mix.

The question of organic decay is interesting, but my gut feel is this would pale in comparison versus the compression/consolidation settlements.

DirtGuy's second point is spot on- our design guys nearly always go the geogrid route for embankments over such soils...geogrids with pioneer rock for basal drainage, then monitor settlements. I suspect with the time-frames involved in achieving consolidation of your peat layer, it would be more economic to do a staged construction with geogrids than to surcharge and wait. The geogrids might facilitate steeper fill batters, saving on material costs.

I imagine case histories with monitoring data would be your best bet, laboratory data is unlikely to yield anything reliable - sorry I can't provide any sources.

All the best,
Mike

 

[jcm1890]

Thanks for your input and advice!

Fattdad - I had heard this type of advice previously but was slightly uneasy with its adoption - my understanding would be that because primary and secondary compression are due to different processes (e.g. primary due to pore pressure dissipation and secondary perceived due to viscous effects/particle realignment etc.) that a level of secondary compression would still occur (albeit likely reduced)?

OG - We anticipate that surcharging for one month would be adequate to forego any further primary settlements, and 'live' monitoring of settlement data will be performed on site for verification. The surcharging height is up to only 2.5m above existing ground and therefore we hope that this can be performed fairly quickly, and reduce the issue regarding time of load application.

Mike/DirtGuy - In terms of differential and the use of geogrid - we are planning to adopt a geogrid layer at the top of peat, largely to act as a separation layer, but also to allow a steep batter where the embankment is adjacent to the existing, and live, carriageway which will then help to reduce the differential across the carriageway.
However, the proposed route requires an length of embankment, followed by a cutting, followed by another embankment, and therefore longitudinal differential settlement is more of a concern than lateral(also a maximum post-construction total settlement has been specified by the end client).

Thanks,
J

 

[fattdad]

Hi J,

Basically, the consolidation from the surcharge, "ages" the formation. Just as if the secondary did occur. When you remove the surcharge, you'd essentially be like 10 or 20 years of aging. To see what may develop during the performance life, you'd just make the adjustments in your secondary equation.

At least that's how I see it.

f-d

ípapß gordo ainÆt no madre flaca!

 

[AnotherDirtGuy]

Hey J,

At this point are you set on the pre-load approach or have you looked at other types of ground improvement? My understanding is in peat 50% or more of total consolidation can be secondary(depending on fiber content, etc.) And since it sounds like your peat layer varies in thickness under a range of cut/fill depths, it seems challenging to me to meet strict total and differential settlement criteria without extensive pre-load time and/or depth. But perhaps PVDs could accelerate the preload and/or lightweight fill in the embankments could attenuate the secondary?

What do you think about using aggregate columns or rigid inclusions to bypass the peat altogether under your embankments? Depending on you fill heights you might be able to use geogrid form a soil load transfer platform in these areas, so your column spacing doesn't get cost prohibitive. The approach would speed up construction considerably and help with your headache about the long-term settlement.

Just some thoughts! Interested to hear what you end up doing

 

[oldestguy]

J: Your preload and with detailed settlement readings my bet is you will not see ongoing noticeable settlement after one week. If so, run a test fill to prove this and you likely can use that as a movable surcharge, what they call "rolling surcharge" here. That saves a lot of hauling in and removing of fill. Just make sure that the contractor doesn't figure on using that fill for some permanent site without also surcharging that site.

 

[eg1776]

I have analyzed settlement for peats over my 40 years of professional practice and not one site could not be explained completely by primary consolidation. When drains or wicks are involved, people fail to recognize the massive soil disturbance and often mass movement caused by 1. bulldozing a drainage layer into place and 2. installing the drains. What people attribute to secondary, is really the two sources of remolding of the peat during construction. Furthermore, it is impossible to discern secondary by looking at field settlement data. According to Roy E. Olson, Professor Emeritus at The University of Texas at Austin:

"The common view of secondary effects as being simply settlement following primary settlement is an oversimplification. Secondary effects occur during primary consolidation, as indicated by permeability data, as well as afterwards. Secondary effects probably result from different mechanisms in different soils. Some simple mechanisms include:
1. Soils have void spaces of widely differing sizes. In some soils, water may drain from the larger voids in accord with primary theory and then water may more slowly squeeze out of smaller voids, producing a secondary effect.
2. In organic soils containing plant matter, water may similarly squeeze out of the voids in accord with primary theory and then water may squeeze slowly out of the individual plant cells, through the cell walls, at a slow rate, producing a secondary effect.
3. In more granular soils, the shearing stresses between particles may be of a viscous nature. When primary consolidation is occurring, the rate of movement between particles is a maximum and the shearing stresses will then cause a maximum resistance to volume change. The retarded rate of compression explains the observation of lower fitted hydraulic conductivity compared to the measured values. The fitted value must be lower than the real value to explain the retarded rate of consolidation. Creep following primary consolidation would also be expected.
5. Field cases that show large secondary effects sometimes involve the settlement of comparatively narrow embankments with low factors of safety against failure. In some cases, some part of the time-dependent settlement may be due to mass movement of subsoil out from under the embankment due to the high shearing stresses and overall soil viscosity.
6. Some case histories of settlement of wide embankments involve a shallow highly compressible soil and deeper less compressible soils. Apparent secondary settlement may actually represent delayed primary consolidation of the relatively incompressible soil which cannot drain until the overlaying, more compressible layer, has consolidated somewhat.
7. In the case of some organic soils, the hydraulic conductivity of the soil decreases by more than an order of magnitude during consolidation under a given load. Consolidation naturally proceeds more rapidly initially but then at a decreasing rate because of the reduction in hydraulic conductivity, thus producing an apparent secondary effect.
8. Highly non-linear stress-strain curves can produce settlement-time behavior that looks like primary consolidation followed by secondary consolidation.

The behavior of the laboratory sample and the field layer are apparently different in that there is a clear secondary effect in the laboratory but none in the field. The bottom line is that In the field, primary consolidation is slow because of the layer thickness. As a result, effective stresses build up so slowly on most soil elements that negligible resistance develops in the dashpot and thus the secondary delay is eliminated."