Temporal Variation of Coefficient of Consolidation and its Effect on PVD Consolidation

The installation of Prefabricated Vertical Drain (PVD) with preloading has become one of the most preferred and efficient technique for ground improvement of soft clay. A number of oil storage tanks of varied capacities were proposed on soft clayey subsoil at Budge Budge, (20 Km South of Kolkata) South 24 Parganas, located quite close to river Hooghly in West Bengal, India. Installation of PVD followed by preloading was adopted as a ground improvement technique, which was appropriate for the site to achieve, the required bearing capacity and settlement within a short time. Progress of the consolidation process was monitored periodically and at regular intervals. The settlement and pore pressure dissipation were measured with the help of settlement plates and pore pressure transducers. The observed time-settlement data have been analyzed to obtain the coefficient of consolidation in the horizontal direction. The variation of the Coefficient of Consolidation (Ch) with respect to time has been plotted and a relationship has been established to study the variation of Coefficient of Consolidation with time. It has been found that Ch values vary with time as well as settlement. Ch varied from 0.00442 m /day to 0.191128 m/day, when applied surcharge load varies from 21.6 kN/m to 115.0 kN/m, in stages, over a period of 30 days. This paper highlights the nature of variation of Coefficient of Consolidation (Ch) under a given surcharge over time.


INTRODUCTION
Prefabricated Vertical drains (PVD) along with surcharge loading are being extensively used for ground improvement in soft clay. The process is eco friendly, less expensive and easy to operate. A vertical drain consists of a slender, synthetic core with vertical grooves functioning as drain. Geotextile sleeves are provided around the core, which protects the grooves from clogging by acting as a filter. Installation of PVD is accomplished by pushing a steel mandrel into the ground up to the required depth. Excess pore water pressures caused by preloading finds an easy passage in the horizontal direction towards PVD and then along the vertical drain into the permeable drainage layers. Insertion of PVD shortens the drainage path, thereby accelerating the consolidation process. A desired degree of consolidation, which normally takes years, can be achieved within a few weeks.
Bharat Petroleum Corporation Limited (BPCL) in their effort to augment storage capacity at their Budge-Budge installation proposed to construct a number of oil storage tanks within their existing tank farm site, located quite close to river Hooghly. Jadavpur University was engaged by BPCL to provide solutions for ground improvement so as to complete the project within the available time. Eight Tanks were installed out of which one tank (T4) has been identified for detail monitoring. It has been often reported (Chung, 1999;Bo et al.,2003;Arulrajah et al.2004) that the main difficulty faced for ground improvement using PVD is reliable prediction of settlement and time for consolidation. One of the possible reasons for inappropriate prediction of time and settlement can be attributed to the improper assessment of the horizontal coefficient of consolidation (Ch), which is one of the most important factors for determination of rate of settlement. This hypothesis can be substantiated by the fact that variation of Coefficient of Consolidation (Ch) with time and has not been considered for estimating the time and settlement. In construction loads are placed on the ground over a period of time and as a consequence of that consolidation of soil takes place. Permeability and compressibility of the soil are functions of void ratio therefore as the void ratio changes with time during the process of consolidation both the parameters are expected to change ( Rowe,1968;Indraratna et al.2005;Hsu and Liu 2013) . Change in coefficient of consolidation with applied surcharge has been noted by Terzaghi et al.(1996); Seah and Juirnarongrit (2003). Barron (1948) presented the basic comprehensive theoretical solution to the vertical drain assisted consolidation of soft soil, and since then many theoretical, empirical and approximate modifications (Hansbo 1979,1981,Holtz 1987,Bergado et al.1991,1993, Indraratna and Redana 1997,1998Basu et al.2005 for predicting performance of PVD on soft soil have been developed. For estimation of coefficient of consolidation the graphical method proposed by Asaoka (1978) is extensively used, however the method is affected by individual skill set, and it underestimates the ultimate settlement depending on chosen time interval (Arulrajah et al.,2004,Edil et al.,1991. In this paper an attempt has been made to study the variation of coefficient of consolidation (Ch) with time from observed field settlement data.

FIELD SUB SOIL PROFILE
The subsoil of BudgeBudge, which is a part of Indo-Gangetic plains of West Bengal and similar to Normal Kolkata soil, consists of layers of normally consolidated, soft alluvial clay deposit of low shear strength. A detailed subsoil investigation was made at the site. 14 boreholes up to a depth of 20 meters were made at different locations. Disturbed and undisturbed soil samples were collected and tested at Jadavpur University soil mechanics laboratory. Details of average subsoil profile are shown in Table 1. Hence, Ground Improvement was considered necessary. Estimation of settlement of untreated soil (Based on 1 and 2) has been presented in Table 2.

Estimation of Settlement of untreated soil.
Estimated total settlement 978 mm and time required for settlement =31.88 years.
The preloading was applied in two stages. Loose sand and sand bags, were used for preloading. Coarse sand (unit weight 18.00 kN/m 2 ) was used up to a height of 2.3 m, beyond that, due to non-availability of coarse sand, fine sand (unit weight 15.60 kN/m 2 ) was used for the rest of the 4.7 m. total Height of preload being 7 m, Total load of 115.00 kN/m 2 (2.3x18.00+4.7x15.6), was envisaged to be applied in two stages. In the first stage pre load of 57.00 kN/m 2 and in the second stage pre load of 58.00 kN/m 2 was applied.

Change in shear strength after consolidation by preloading:
The soil is normally Consolidated (N C). First two layers will undergo maximum changes.
It is known that, ∆ ∆ 1 ⁄ = 0.25 ( Som,N.N. and Das,S.C. 2003) Where, Settlements in stage I and stage II were 333mm and 347mm respectively Settlement during Stage I of the preload ( Pc = 57.00 kN/m 2 ) has been presented in Table 3.
Settlement during Stage II of the preload ( Pc = 58.00 kN/m 2 ) has been presented in Table 4. Total settlement considering both the stages of consolidation =680.00 mm.

The design of PVD
From the calculations it is evident that the first two layers will undergo maximum settlement, hence band drain will be required in these layers. Both the layers can be considered as an equivalent single layer of thickness 11.0 m and average Cu of 23.00 kN/m 2 .
Time required for 90% consolidation using Band drain is given by Barron (1948): Installation was started by laying a 400 mm thick sand layer, to make the ground suitable for movement of crawler crane. Locations of each of the PVD were carefully marked on the ground .A 40 ton capacity crawler mounted crane operated band drain installer was used to push a rectangular mandrel (Size 100x35 mm) on the ground. PVDs were installed to a depth of 15 m from the ground. PVDs were cut off leaving 600 mm above the ground. A drainage layer of 300 mm thick coarse sand, with a gentle slope towards the outer periphery, was placed. Arrangements were made to drain the discharged water into the nearby brick lined surface drain. On top of the drainage blanket a woven geotextile was spread over the entire treated area. Preloading had started after installation of band drains, placement of drainage blanket and laying of geotextile. Sand and sand bags were placed in a systematic and staged manner up to the designed height of 7 m.

Field Instrumentation:
After installation of PVDs,(and before preloading,) four Settlement gauges (one in the middle and 3 in the outer periphery at a distance of 10 m from centre and two Piezometers , at depths, approx 2.5 m and 9.1m were carefully placed. Settlements were measured with respect to a fixed bench marks, located outside the tank firm area. An attempt has been made in this study, to determine coefficient of consolidation from the analysis of temporal variation of Ch.

Settlement Vs Time curve
On completion of installation of PVDs, placement of drainage layer and installation of necessary instruments, preloading with sand bags and sand were accomplished in staged manners. Settlements were recorded twice and pore pressures were recorded four times daily.
It can be seen from the time-settlement curve (Fig 3.) that settlement started almost immediately and continued at a slow rate .Whenever additional surcharge loads were applied the rate of settlement increased to some extent, This slow rate of settlement continued till 32 nd day. The rate of settlement increased when the surcharge load was increased to 115.00 kN/m 2 from 68.00 kN/m 2 (approximately). Preloading was paused for approx 11days and settlements observed during this period were 54 mm and 48 mm. From the time vs settlement curve (Fig 3) it can also be seen that the rate of settlement was slow till 59% of the surcharge load was applied. After placement of final load on the 55 th day, settlement at much faster rates were observed, this was further substantiated by the observation that when preloading started after the pause, rate of settlement increased again. Final settlements of 607 mm and 503 mm in centre and peripheral gauges were observed. Settlement nearly stopped after 88 days, that was, after 33 days from the final loading .
It may be mentioned here, that river Hooghly is at a distance of about 500m from the site. On close monitoring it has been observed that, tidal variation did not show any effect on in the pore pressure.

Estimation of Time of consolidation with temporal variation of Ch
Time Settlement data from the site has been analyzed for determination of coefficient of Consolidation.
Degree of Consolidation (U) for equal strain condition is given by This equation can be used to find out the coefficient of consolidation for a particular time and subsequent Degree of consolidation for that period .
The variation in Degree of Consolidation (U%) considering Ch as constant and Ch as variable can be seen in Fig 6. The rapid change of Ch ,if considered ,will result in more realistic prediction of both degree and time of Consolidation.

CONCLUSIONS:
Main findings from the Study may be concluded as follows: 1) Ch does not remain constant over the period of consolidation .Variation of Ch depends on the extent of surcharge load and time.
2) Change in the increasing trend in the value of Ch indicates end of primary consolidation. Variation of Coefficient of consolidation after application of full surcharge, can be approximated by a, best fit, quadratic equation: Ch (10 -3 ) = 2 − 55.5 + 653 ('t' is Time in days).