Research Interests

¡¡

PILE ENGINEERING THEORY AND PRACTICE IN YPC-BASF PROJECT

Keyword:  pile engineering, YPC-BASF Project, bearing capacity, negative friction, precast-pile

1. BRIEF INTRODUCTION OF THE PROJECT

    YPC-BASF integrated petrol-chemical (IPS) project is one of the biggest petrol-chemical projects under construction in P.R.China. The total investment is USD 2.9 billion. The ethylene unit is the largest contract package of the IPS project, which is assigned to Stone&Webster, a world leading engineering company. HQCEC is the foundation & underground piping design subcontractor. The design began in Sep, 2001 and ended in Sep, 2002. Due to the poor geotechnical condition of the soil, cast-in-situ (CFA) concrete pile and precast(PC) concrete pile were adopted for the equipment, piperack and buildings¡¯ foundations. Totally 3760 precast piles and 463 CFA piles were installed by the end of the civil work.

2.  CHOOSING PILE TYPE AND DETERMINING BEARING CAPACITY

2.1 Application of different pile types

       The new ethylene unit is located on the Yangzi river flood plain alluvial area. Ground deposit material mainly consists of pain filling soil on the top layer, silty sand and clay layer which is 30~40m in depth and highly compressive, mid sand layer and mud rock in the bottom. Because some of the heavy and tall equipment, such as cracking furnace, hot vessels transfer remarkably large compressive, tensile and even lateral force to the foundation top. The natural ground can¡¯t provide enough bearing capacity to sustain theses equipment. So the cast-in-situ §¶800x60000mm pile was selected to meet these high bearing capacity requirements. Design bearing capacity of §¶800x60000mm cast-in-situ pile is shown in table 1.

                Table 1 Bearing Capacity of §¶800x60000mm cast-in-situ pile

      Besides these very heavy equipment, there are still large amount of heat exchangers, tanks, pumps and piperack foundations which are not very heavily loaded in compression. But some of them are subjected to lateral bundle force and pipe thrust force. It¡¯s apparently not appropriate to adopt CFA pile in these light equipment foundations. So 450x450mm² squire precast pile was designed to meet the more diversified requirements of light equipment foundations. The actual installed PC piles are 48m long in average and the pile toe goes into the mid sand layer by 0.7m.  Design bearing capacity of 450x450 precast pile is shown in table 2.

                        Table 2   Bearing Capacity of 450x450  precast pile

2.2  How does failure load become design compressive bearing capacity

Since the 400x400 mm² ,36m PC piles have been serving well in the old YPC factory foundations for nearly 20 years. The same type of pile was selected in the conceptual design. Totally 12 PC piles were tested for their vertical compressive, tensile and lateral bearing capacity. And PDA driving test, high strain (capwap) test and dynamic test were also done to these piles. The failure load & design compressive bearing capacity of the 400x400 mm² ,36m PC piles is shown in Table 3                   

 Table 3   Bearing Capacity of 400x400 mm²  precast pile

According to Pile Foundation Technology Code (PFTC) of China, the failure load should divided by safety factor which is based on the probability theory to become the design compressive bearing capacity. The lateral capacity is defined as the lateral load that happened when the pile top has 6mm displacement.

2.3 Comparison of different determining method

       It¡¯s regulated in the PFTC that class I & II (very important) pile foundations¡¯ compressive failure load can only be decided through load test. But experimental formula method, which is based on the relationship between soil property and the pile resistance, can be used to estimate the class III pile foundations failure load. The experimental equation presented by PFTC is

                    ( 2.3.1 )

Qisk---pile skin friction

Qpk---pile toe resistance

u----pile perimeter

Li---- depth of each soil layer

Ap---area of pile section

According to 2.3.1, the failure compressive load of 400x400 mm² ,36m PC pile would be

       We can see that the experimental formula result is a little higher than the load test result shown in table 3 and the pile toe resistance accounts for 45% of total bearing capacity. Since the load test is the most reliable way to determine the bearing capacity, we can draw a conclusion that the experimental formula method is not conservative in some cases, it¡¯s precision depends on the designer¡¯s experience to choose the Qisk & Qpk factor. But this method can be used in conceptual estimation. Compared with the capwap test results (shown in table 4) which reveal the actual resistance distribution along the pile body, the calculated pile toe resistance accounts for a much higher percentage than the actual situation.

  Table 4  CAPWAP TEST RESULT OF 400x400 mm² ,36m PC piles (kN)

2.4   Negative skin friction

         Because there exists about 20 meter highly compressive mucky silt 7 meter under the ground and the whole ethylene unit was backfilled with 3 meter plan clay only 1 year ago. The ground settlement is far away from stable. It¡¯s necessary to take the negative friction into consideration. After consulting to URS, an geotechnical engineering consultant, Stone&Webster decided to use longer piles instead of 36m piles. The main concern is the negative friction. According to URS estimate, the bearing capacity will be reduced by half and the 36m pile will have little bearing capacity left. So 450x450 mm² ,48m long squire precast piles were finally chosen. The failure load obtained through load test is shown in table 5.

Table 5  LOAD TEST RESULT OF 450x450 mm² ,48m PC piles (kN)

The experimental equation analyzed failure compressive load is: (according to 2.3.1)

        The same problem is revealed here that the experimental formula failure load is higher than the load test result. But the pile toe resistance accounts for 26% of the total resistance, which is much closer to the actual situation of the resistance distribution.(see table 6)

  Table 6  CAPWAP TEST RESULT OF 450x450 mm² ,48m PC piles (kN)

Comparison of  450x450 mm² ,48m & 400x400 mm² ,36m precast piles failure load are shown in chart 1

                                                   Chart 1

3  CHOOSING HAMMER TYPE AND DRIVING CRITERIA

3.1    It¡¯s important to choose an appropriate hammer for a good pile driving performance. A heavy hammer would produce too much striking energy, which will damage the concrete, especially result in concrete spalling or cracking. PFTC recommends monitoring the striking compressive strain and tensile train for piles requiring very strict crack control. But this check is not mandatory in this project. A light hammer may result in the difficulty in driving the piles down because insufficient energy can¡¯t produce a substantial displacement between the pile and the soil. The following table 7 is the code recommended hammer parameters.

                             Table 7         Hammer Parameters 

      Combining the code recommendation, piling contractor¡¯s experience and the piling rig manual, we finally decide to use 6t hammer. The total weight of a 450x450 mm² ,48m PC pile is 25t ,a 6t hammer can transfer enough energy on the pile to produce a plastic displacement.

3.2     The more complicated problem than choosing a hammer is the driving criteria. To make the explanation more clearly, some terms need to be clarified first.

        The refusal in the last 3 consecutive 10 strikes: the penetration measured in each 10 strikes just before stopping driving. This value should be less than a preset number (usually 4~8 cm for 6t hammer), which means the piles toe has reached the hard supporting course and obtained enough resistance from it, so the desired bearing capacity can be insured.

       The blows in the last 3 consecutive meters: although the refusal in the last 30 strikes can disclose where the pile toe goes, the pile sometimes doesn¡¯t hit the hard course suddenly, but goes through a rather long way in the sand layers before it actually stops. The blows in a standard distance (eg.one meter) can be used to measure this. If a pile undergoes more than 100 blows in each meter of the last 3 consecutive meters, that means it has gone deep enough into the sand layer and enough resistance is obtained.

      Another way can also be used to judge if the pile has reached the supporting course by comparing the pile toe elevation with the supporting course contour map. The pile toe elevation can be calculated by following equation:

                Toe Elev = Top Elev ¨C pile length                   (3.2.1)

      This method is also recommended by PFTC. But as we know, the bearing course contour map was generated by tieing the borehole point with curve lines. It can¡¯t be very accurate in every pile location. So the principle in the PFTC to stop driving is: take refusal as priority and elevations as reference for piles that penetrate through hard soil layer. The driving criteria applied in this project are shown in chart 2. We may notice that there is a category classified as Q70 in this chart. This type of piles was sorted out after consulting to URS, because they didn¡¯t show very good resistance from the supporting course but were qualified for light loading.

Another experience obtained from this project is the refusal criteria shouldn¡¯t be made too stringent to the test piles, because the driving performance of test piles is the most important basis to determine the driving criteria. Excessively strict driving criteria can cause much trouble in the construction. Many piles will have to be classified as ¡°unqualified¡±, though they actually are good to take light or medium load. Relatively loose driving criteria can also reflect the obtained resistance and bearing capacity if combined with pile testing.

4.   CHOOSING A PILE LENGHTH

      Usually the design institute wouldn¡¯t point out the individual pile length. The drawing will only show how deep the pile will be in the supporting course. Therefore, it¡¯s the site work to decide the pile length.

A pile length can be calculated by changing equation 3.2.1

into         Pile length = Top Elev ¨CToe Elev                   (4.1.1)

       Because of the same reason mentioned above, the calculated pile length couldn¡¯t be very accurate. But the site engineer should try to make the pile length more reasonable than leaving piles sticking up in the air. If the over pile length continue to happen in one area, the information should be transferred to the site engineer ASAP for him to make a quick response.

5            PILE QUALITY CONTROL

5.1    Can a precast pile that hasn¡¯t reached the 28 days curing date be driven? The answer is negative. This operation is banned by the PFTC. The main reason is the hydration heat can¡¯t discharge completely from concrete interior before 28 days, thus the pile tends to be crisp and easily broken. The pile breakage usually happens as the form of refusal suddenly change, which means middle or bottom section broke under the ground. And the crisp concrete also contributes to the top section breakage, crack and concrete spalling. In some periods during the piling construction, the pile breakage rate rose significantly. The investigation on the pile delivering records shows the majority of piles driven in these periods haven¡¯t reach 28 days. The relation between pile breakage rate and insufficient curing date is shown in chart 3.

Besides the curing date problem, the pouring sequence, jolt ramming and concrete mixing are also the vital causes to the poor concrete quality. We actually saw clay was mixed up with the rubbles in a broken pile top, which unclose the poor concrete quality very clearly.

6            CONSTRUCTION CONTROL

       Considering many factors will affect the piling work, it¡¯s necessary for the site engineer to set up a piling information database for a better construction control.

6.1 Collecting the data

        According to the PFTC, each piling crew should record such information as pile NO., blows in every meter, total blows, difference between design elevation and actual elevation, the refusal in the last 3 consecutive 10 strikes and abnormal situation in uniform tables. This information will help us make judgment on each pile. The worker¡¯s records must be collected from the crew on a daily basis.

6.2  Setting up a database

       Having owned all the driving records, a site engineer should classify the pile into different status, such as ¡°OK¡±, ¡°FAILED¡±, ¡°TEST¡± and ¡°Q70¡±. The discipline of classification is shown in the chart 2. According to my experience, it¡¯s better to set up an EXCEL database than sorting out original records all the time. A typical pile records database is shown as chart 4

                                          Chart 4  Example of Pile Database

By using the automatic filter tool, we can easily find out the desired result and give quick response to field request. Another benefit for a computer database is that different departments in a construction team can share the information through Intranet. So more convenience would be provided to the whole team.

7            PILE TEST

7.1    There are two phases of pile test. The first one is before the design commencing. Compressive load test, tensile load test, lateral load test, PDA driving test, PDA high strain test (one form of CAPWAP test) and dynamic test will be carried out as we discussed in paragraph 2. The purpose of this phase is to provide bearing capacity information for design. The second phase happens during and after piling construction. The high strain tests are usually employed to check the working pile integrity and bearing capacity in this phase. Although these two phases are separated from time sequence, they are related internally. For example, the PFTC requires that if the compression test wasn¡¯t done in the first phase and the bearing capacity was determined by experimental equation, then it must be done in the second phase to check the actual bearing capacity.

7.2    In the second phase of pile test, dynamic tests (high strain test and low strain test) are the main methods to check the working pile integrity and capacity. One thing that should be emphasized here is the principle in choosing the piles to be tested. The PFTC recommend the selection of test piles shouldn¡¯t be based on a random scope but should be based on a filtered scope. So the selected test piles should be neither the ¡°OK¡± nor the ¡°BROKEN¡± ones, but those whose status can¡¯t be identified from the driving record database. The relation between the actual tested capacity and the driving performance is shown in the chart 5.

¡¡