GeoCongress 2012 ASCE 2012

1928

High Vacuum Densification Method for Soft Soil Improvement R. Y. Liang1 and S. L. Xu2 Distinguished Professor, Department of Civil Engineering, University of Akron, Akron, Ohio 44325-3905, Tel: (330) 972-7190, E-mail: rliang@uakron.edu 2 Chairman, Shanghai GeoHarbour Group, E-mail: hvdm01127046@vip. sina.com ABSTRACT A soft soil improvement method was recently developed by integrating and combining the use of vacuum dewatering and deep dynamic compaction techniques. This method utilizes several runs of vacuum dewatering and deep dynamic compaction to generate negative and positive pore pressure, respectively, which help facilitate pore pressure dissipation and achieve both consolidation and densification effects. A case study involving field monitoring of the soil response to this soil improvement method is presented. INTRODUCTION The application of vacuum to facilitate consolidation in saturated fine grained soils has been used either alone by means of PVDs or in combination with the static surcharge load using fill materials (Kjellman, 1952 and Holtz, 1975). The effectiveness of vacuum consolidation with or without surcharge loading is highly dependent upon soil permeability and the efficiency of vacuum system. The desired degree of soil improvement and the allowable time duration for completion can also play an important role in determining if vacuum consolidation can be a viable soil improvement method for the project. The use of deep dynamic compaction technique in saturated fine grained soils has not been widely accepted, even though Menard (1975) has shown its working mechanisms. In recent years, the combined use of vacuum consolidation/dewatering and the dynamic compaction, however, has been used in practice to improve the fine grained soils. In this paper, the working principles of this ground improvement method are described. The distinguishing features of this method and its advantages and limitations are elucidated. The mechanisms of undrained strength gain due to this method are also illustrated. Finally, a field case is presented at the end of this paper. It should also be noted that this method is limited to treat soft soils to the depth of up to 8 to 10 meters. For treating the soft soils with depth greater than 8 meter or for achieving much higher improved strength, the conventional surcharge loading with prefabricated vertical drains (PVD) is needed. DESIGN STEPS AND CONSTRUCTION PRINCIPLES
1

GeoCongress 2012 ASCE 2012

1929

This in-situ ground improvement method utilizes the combination of active drainage, consolidation, and densification principles. To be successful, this ground improvement method needs to be executed in a controlled manner based on the feedback from the on-site monitoring data. Figure 1 provides a schematic drawing of the method involving the use of vacuum consolidation and deep dynamic compaction in repeated runs. The planning and preliminary design steps as well as the construction sequence of this ground improvement method are discussed below. Step 1: Conduct detailed geotechnical investigation at the project site. Evaluate and determine soil profile at the site to gain detailed knowledge of the depth and thickness and distribution pattern of fine-grained soils requiring the improvement. Obtain important basic soil properties, including particle size gradation curves, Atterberg Limits, water content, hydraulic conductivity, compressibility, and coefficient of consolidation. Conduct in-situ tests, such as CPT, to establish baseline values of the untreated soil conditions. Understand and establish performance criteria of ground improvement.

Figure 1. Schematics of the vacuum consolidation and dynamic compaction method. Perform preliminary design to provide plans for spacing and depth of vacuum pipes, determine different energy levels of deep dynamic compaction and number of drops for each pass and the grid spacing of the tamper, time needed for vacuum consolidation between cycles of dynamic compaction, etc. However, it should be emphasized that the initial plans will generally need to be modified based on the onsite monitoring data and the required final performance criteria. Typically, the spacing between the vertical vacuum pipes could be from 1.5 m to 3 m, depending upon soil permeability and the dimension of the tamper. The dynamic compaction tamper can be from 1 m to 2 m in diameter, while the impact energy could range from 400 kN-m to 2000 kN-m per drop. Step 2: Install vertical vacuum pipes and horizontal drainage pipes. The vertical vacuum pipes are made of steel, typically 2.5 to 3.0 cm in outside diameter, and 3 mm in thickness. The vacuum pipes contain perforated holes along the length of the pipe and are wrapped around on the outside by a geotextile fabric for prevention of soil particles from migrating into the pipes. The vacuum pipes can be

GeoCongress 2012 ASCE 2012

1930

installed using several different methods, such as a vibratory hammer and a mandrel, or a hydraulic system to directly push vacuum pipes into ground. The horizontal drainage pipes are typically PVCs, which are connected to the vertical vacuum pipes through an elbow connector. Figure 2 shows an array of horizontal drainage pipes connected to vertical vacuum pipes at a project site.

Figure 2. Array of vacuum pips and horizontal drainage pipes. Step 3: Apply first stage of vacuum dewatering and vacuum consolidation to lower ground water table and to reduce water content of the soil in the influence zone. Generally, the net effect of this stage of vacuum dewatering/consolidation is a lowered ground water table and an increase of effective stress of the soil by up to about 50 to 80 kPa, depending upon the efficiency of vacuum consolidation. It is noted that the highest vacuum pressure that can exert on the pore water in the soil is one atmosphere pressure, 100 kPa, without causing water cavitations. The undrained strength gain of the normally consolidated soft clays corresponding to 50 to 80 kPa effective stress increase is roughly 15 to25 kPa. This is based on a rule of thumb empirical correlation between the undrained strength and the effective vertical stress for the normally consolidated clays being equal to 0.3. Therefore, this stage of vacuum consolidation/dewatering is primarily for making the site accessible for heavy construction equipment to carry out the next stage of work, i.e., deep dynamic compaction. The time for completing this stage of vacuum consolidation is dictated by the spacing of vertical vacuum pipes and horizontal hydraulic conductivity of the soil. Smearing effects due to installation of the vertical vacuum pipes can cause the horizontal permeability of the soil in the smeared zone to be reduced, thus impeding the rate of radial pore water drainage. Nevertheless, this stage of vacuum consolidation is usually completed within 7 days before proceeding to next stage. This is due to the main objective of this stage being to increase effective stress in the soil to the point so that the dynamic compaction equipment could be brought on site. Monitoring of pore pressure variations and CPT reading could be used for determining the time needed for this stage of work.

GeoCongress 2012 ASCE 2012

1931

Step 4: Apply deep dynamic compaction to densify the soils and to generate positive pore water pressure. The direct impact by the heavy tamping creates craters, resulting in the displacement of soils and the corresponding reduction in void ratio (direct densification), while generating positive pore pressure in the influence zone. Previous studies by Menard (1975) indicated that deep dynamic compaction in cohesive soils can cause rapid increase of both pore water pressure and gas pressure, whether the soil is fully saturated or not, due to the presence of micro air bubbles. The important controlling parameters of dynamic compaction are the weight and dimension of the tamper, drop height, grid spacing, and number of tamper drops per spot. The decision of these parameters is usually based on experience and field trial with appropriate site monitoring data. The main criteria are that the soils underneath the bottom of the crater do not suffer the undrained shear failure or the so called rubber soil phenomenon, i.e., the impact energy is carried by the water rather than the inter-particle contacts. The typical dimension of the tamper is about 1 to 2 meter in diameter, and the weight can vary from 20 to 70 tons. The tamper drop height varies from 10 meter to about 20 meter. A study by Mostafa (2010) provides useful correlations between crater depth, soil properties, influence zones, and tamper energy. With two to three dynamic compaction machines on site, the duration of this phase of work could be accomplished within 7 days for a typical 10,000 meter square coverage area. The monitoring of pore pressure variations and ground water tables, as well as measurement of water content, are suggested to ensure that this stage of work is progressing according to the original assumptions. Step 5: Apply the second stage of vacuum consolidation to facilitate rapid dissipation of pore pressure and to further reduce water content and void ratio of the soils in the influence zone. The combined effects of vacuum generated negative pore water pressure and the deep dynamic compaction generated positive pore water pressure can create a very high pore pressure gradient, which in turn help facilitate accelerated dissipation of pore water pressure, resulting in reduced water content. The duration of this stage is generally 7 days or less before proceeding to next stage. Step 6: Evaluate the soil properties after completing Step 5. In particular, the water content, pore water pressures, ground water elevation, ground subsidence, and in-situ test results such as cone resistance of CPT, need to be determined to assess the results of the first run (Steps 4 and 5) of the vacuum consolidation/dewatering and dynamic compaction process. Evaluation of the outcome of ground improvement at this stage would allow for adjusting the operation parameters (spacing and depth of vacuum pipes, dynamic compaction energy level and grid spacing of the tamping points, etc.) in the next stage of vacuum consolidation and dynamic compaction process. With each successive stage of vacuum consolidation and dynamic compaction process, the energy level of the dynamic compaction is increased to achieve further reduction in void ratio and a deeper influence zone. Step 7: Repeat Steps 4 to 6 until the ground improvement performance criteria are satisfied. It should be pointed out that in general three runs of vacuum consolidation/dynamic compaction are generally used to achieve the results that meet the general required performance criteria. Of course, the actual number of runs would depend upon the starting soil conditions and the required final soil conditions.

GeoCongress 2012 ASCE 2012

1932

In summary, the aforementioned method is a repeated process of vacuum consolidation/dewatering and dynamic compaction, with each successive cycle involving the use of increased tamper impact energy to achieve the desired density and the depth of treatment. In addition, the on-site monitoring plays a key role to ensure that not only the soil properties before and after ground improvement are monitored but also that the operational parameters (e.g., the spacing of vacuum pipes, the energy of the tamper, the duration between each stage of vacuum consolidation, etc.) are optimized. ADVANTAGES AND LIMITATIONS Some of the interesting advantages of using the combined vacuum consolidation and dynamic compaction includes: (a) enhancing the vacuum well drainage techniques in the fine grained soils with relatively low permeability due to the enhanced pore pressure gradient created by combining the dynamic compaction induced positive pore pressure and vacuum induced negative pore pressure, (b) overcoming the common obstacle that dynamic compaction could not be applied to the saturated cohesive soils due to the ability to lower the ground water table and the creation of unsaturated soil zones through vacuum dewatering/consolidation, and (c) expediting pore pressure dissipation due to the creation of high pore pressure gradient as the result of the combination of negative vacuum pore pressure and positive pore pressure generated by the dynamic compaction. The anticipated results include the following: (a) the creation of a highly over-consolidated clay layer near the ground surface with thickness in the range of 5 to 8 meters depending upon the deep dynamic compaction efforts and the influence zone (see explanations in the next section), and (b) eliminating the post-treatment horizontal drainage path as a result of withdrawal of vacuum pipes from the ground after completion. The limitations of the vacuum consolidation/dynamic compaction method include that the treatment depth is generally limited to 10 m due to the limit of influence zone of deep dynamic compaction and the loss of vacuum efficiency when exceeding that depth. In addition, the fine grained soils that contain a large portion of organic materials would not be suitable for this method due to the pronounced secondary compression (creep) that could be treated. The range of permeability of the fine grained soils that can be suitable for this treatment method is limited to a minimum of about 10-6 cm/sec. MECHANISMS OF UNDRAINED STRENGTH GAIN The mechanisms of the combined vacuum consolidation and dynamic compaction in improving the soil strength and reducing water content can be illustrated by the e-log p plot in Figure 3, where e is void ratio and p is effective mean stress. Assuming the fine grained soil is normally consolidated and the state prior to improvement is at point A. With dynamic compaction, positive excess pore pressure is generated, moving the soil state from A to B1 (ignoring the apparent reduction of void ratio due to dynamic compaction). Subsequent to dynamic compaction, high vacuum is used to dissipate excess pore pressure rapidly to bring the soil state from point B1 to point D1. The reduction of the void ratio is due to

GeoCongress 2012 ASCE 2012

1933

dissipation of excess pore pressure during the accelerated consolidation process. With the repeated cycles of vacuum and dynamic compaction, both density and undrained strength of the fine grained soils can be improved. The undrained shear strength of the improved fine grained soils can be estimated by relating the undrained shear strength to the apparent over-consolidation ratio (OCR) as follows. p (1) OCR = Ei p Ai

( Su )OCR ( Su ) NC

= OCR

(2)

where = an empirical constant.

Figure 3. e-log p plot under many times of dynamic impact and vacuum consolidation. QA/QC PROCESS The success of the aforementioned ground improvement method depends upon the utilization of field monitoring data to allow for adjusting and optimizing the operational parameters, including the heavy tamping energy (mass of tamper, height of drop, spacing and number of drops per spot) and the vacuum consolidation parameters, such as vacuum pipes spacing and depth, among others. Field monitoring typically includes measurement of pore water pressure, ground water level, crater depth, ground subsidence, water content, and CPT. CASE STUDIES This is a land reclamation project at Ningbo Port, China, with an objective to provide a site for coke storage with the intended storage up to 5 million ton per

GeoCongress 2012 ASCE 2012

1934

hectare of area. The soil profile at the site consists of a 2 m of clay, underlain by a 2.0 m of hydraulically filled fly ash, and underlain by a 4.0 m of mud clay layer, and then underlain by silty sand (SM). The strength of the original soils measured by CPT is shown in Figure 5. The first phase of treatment area is about 300,000 square meters with the requirement that the improved site provides the bearing pressure up to 30 to 40 kPa. Pictures of site conditions before and after ground treatment using the presented ground improvement method are shown in Figure 4. Typical comparisons of CPT cone resistance results before and after ground improvement are shown in Figure 5.

Figure 4. Photos of site conditions before and after ground improvement.

Figure 5. CPT cone resistance readings between before and after ground improvement.

GeoCongress 2012 ASCE 2012

1935

As part of this project, a test program was conducted in four subdivisions labeled as B11, B12, B21, and B22, respectively. The vacuum pipes and PVDs arrangements in each zone are summarized in Table 1. The deep dynamic compaction controlling parameters for each subdivision are presented in Table 2. It should be noted that in all subdivisions, the third stage of dynamic compaction was carried out with 800 to 1000 kN-m impact energy, with the number of drops equal to two and with no spacing between the drop spot. This final run of dynamic compaction is commonly referred to as the final ironing. Table 1. Arrangement of vacuum pipes and PVDs. PVD (8m long) Vacuum Pipes (6m Areas Spacing in meter deep) Spacing in meter B11 1.11.1 3.55.0 B12 1.11.1 3.55.5 B13 1.11.1 3.55.0 B14 1.11.1 3.55.5 Vacuum Pipes (3m deep) Spacing in meter 3.52.25 3.52.75 3.52.25 3.52.75

During the experimental program, several items of the site response were monitored. The surface elevation was measured with a 5 m by 5 m grid. Pore pressure sensors were installed at the depths of 3.5 m and 6 meter. A groundwater observation well down at 4 m below the ground was installed and monitored twice a day. Water content in the soil was measured before and after each stage of dynamic compaction. Monitoring of vacuum pressure was also performed at the test site. As part of evaluation of the soil properties, static cone Penetration test was conducted. Table 2. Dynamic compaction (DC) controlling parameters. 1st Stage of DC 2nd Stage of DC Spacing Energy per Energy per Areas Drops Drops (m) dropt (kNm) drop (kNm) B11 4.04.0 800 3 1200 2 B12 4.04.0 1200 2 1600 2 B13 4.04.0 800 3 1200 2 B14 4.04.0 1200 2 1600 2 ANALYSIS OF MONITORING RESULTS AT THE TEST SITE Surface settlement At the end of each of three stages of dynamic compaction, the settlement for Sub-divisions B11 and B12 was 42 cm, 17 cm, and 6 cm, respectively. For Subdivisions B13 and B14, the settlement was 35.3 cm, 29.5 cm, and 8.9 cm, respectively. As can be seen, in all cases the first stage of dynamic compaction had contributed the most settlement toward the accumulated total surface settlement.

GeoCongress 2012 ASCE 2012

1936

Pore pressure monitoring results Representative pore pressure response is shown in Figure 6. It can be seen that pore pressure dissipation rate is very high when vacuum is applied. With the pore pressure gradient generated by vacuum and dynamic compaction, dissipation of pore pressure occurred very rapidly. For the 1st and 2nd DC/vacuum run, within 7 to 8 days, 90 % of the generated pore pressure had been dissipated. Both vacuum pipes and PVDs are used for drainage.

Figure 6. Pore pressure response with vacuum pipes and PVDs. Water content In the hydraulically filled fly ash layer, the water content was reduced from 54.7 % to 39.9 %, with an average of 15 % reduction. In the clay mud layer, the water content was lowered from 53 % to 36 %, with an average of 17 % reduction. In the silty sand layer, the water content did not change. Therefore, it was concluded that the method can reduce water content down to 5 to 6 meter from the ground surface. The method would not affect the water content in the soil layer that is 10 m or deeper from the ground surface. Evaluation of Improvement Results Sub-divisions B11 and B13 were subjected to higher impact energy and less number of impact, while Sub-division B13 had PVD installed. Sub-division B12 and Sub-division B14 were subjected to lower impact energy but with larger number of drops. Sub-division B14 had PVD installed. The average improvement of the entire

GeoCongress 2012 ASCE 2012

1937

site B is as follows. In layer 2, cone resistance was increased from 0.74 MPa to 2.51 MPa, with an improvement ratio of 3.37. In layer 3, the cone resistance was increased from 0.21 MPa to 0.35 MPa, with an improvement ratio of 1.66. In comparing B13 to B11 or B14 to B12 (i.e., zones with PVD and zones without PVD), it can be seen that the improved cone resistance can be on average 10 to 20 % higher in zones with PVD than in zones without PVD. CONCLUSIONS In this paper, recent advances in ground improvement techniques for fine grained soils involving the combined use of vacuum consolidation/dewatering and deep dynamic compaction were described. Specifically, the working principles of this method and its advantages and limitations were discussed. A case study was presented at the end of the paper. The case study presented the monitoring and evaluation results of a pilot testing program at the Ningbo Port Project in China. The site monitoring data confirmed the working principles of the described ground improvement method. The water content was effectively reduced and the undrained shear strength indicated by cone resistance was increased as a result of the application of this ground improvement method. REFERENCES Holtz, R. D. (1975). Preloading by vacuum consolidation: current practice, Transportation Research Record, No. 548. Transportation Research Board, National Research Council, Washington, D.C., 26-29 Kjellman, W. (1952). Consolidation of clay soil by means of atmosphere pressure. Proc. Conf. on Soil Stabilization, MIT, Cambridge, 258-263. Menard, L. (1975). Theoretical and practical aspect of dynamic consolidation. Geotechnique, 25(1), 3-17. Mostafa, K. (2010). Numerical modeling of dynamic compaction in cohesive soils. Thesis (PhD). University of Akron, Akron, Ohio.