High railway embankment on weak supporting soil 2

[rgakalo]

Hello,
we are about to construct an embankment of about 14 meters high on approximately 2.0-2.5 meter clay layer, underneath the clay is a layer of gravel and a water level. The clay is contaminated with roots and the area is a part of a riverbed with very intense vegetation. Usually, we remove the soil which contains organic material but here considering height of the embankment our engineering company suggested that we don't remove it but just cover the supporting soil with 0.5 m - 1 m of rock to prevent vegetation from entering the embankment. We can wait approximately 6 months till we install the track, so their proposal is just to wait till the supporting soil under 14 m embankment settles and then proceed with the construction of the railway track.

Calculation shows that the total settlement is about 16 cm and in 6 months the embankment will settle 12 cm, so what they suggest seems to be acceptable, this proposal is also very appealing economically considering the cost of replacement of 1-1.5 meters of weak soil with rock on about 1 km distance. However, the presence of roots makes me very nervous, first, the calculation is based on clay and not organically contaminated soil, second the vegetation there is very aggressive.

Your advice on the situation is highly appreciated.

 

[geotechguy1]

The catch is your calculation is probably at best +- 50-100% for amount of settlement (assuming you have good site investigation data and previous experience in these soils), and charitably maybe +-300% for settlement time.

You need to start at gathering all the inputs; what are the design criteria for the railway? How much total and differential settlement can it take, and what are the implications if it settles two or three times that amount? What are the financial implications if it turns out placing the fill and letting it settle takes 3x or 5x longer than you thought?

What data do you have on the ground conditions - how many boreholes / cpts / dmts / MASW lines do you have, do you have a ground model, what are the soil properties and how were they derived, what comparable experience in these soils exist? What's the geology / geomorphology / site history etc.

Piled embankments are an option; there are a hundred and one variations of design methods and ways to construct.

SUrcharging is an option (apply a larger fill than your design embankent and then remove) but you need a fill surplus / borrow for that.

Wick drains are possible to accelerate settlement but depends on soil conditions and available equipment, also experience globally seems to be mixed on effectiveness of these technologies.

You can follow your suggestion, but how confident are you in your assessment of settlement amount and duration and what are the implications of being wrong?

 

[TigerGuy]

The amount of roots shown in your images would not concern me. Any further consolidation due to decay of root matter is not likely to manifest to the top of a 14 meter embankment.

When you say a layer of rock, are you referring to gravel, or a rock fill material including cobbles and boulders for a significant portion of the material? If this is an area subject to periodic flooding, I would think you want a significant layer of clean boulders and cobbles to allow water to flow back and forth beneath the embankment.

You can also consider establishing some benchmarks on the top of the embankment, and periodically surveying them to monitor the rate of settlement. I'm guessing your settlement time estimates are conservative, they usually are. By monitoring the settlement as it progresses, you will likely save time on the project schedule. Don't forget, that settlement starts as soon as you start placing material. Much of your settlement will have taken place by the time the embankment is completed.

 

[rgakalo]

Thank you both very much for your comments. I've decided to proceed with replacement of 1 meter of existing material with crashed stone of approximate dmax of 400 mm. I know that it is probably too conservative but I prefer to minimize the risk of significant and prolonged settlement.