 |
 |
Development of an Instrumented Borehole Drilling System for Ground Investigation
Assistant Professor, Civil Engineering Department, and Development Director, Soletanche Bachy |
Abstract
The drilling of holes for grouting tubes creates the potential for obtaining supplementary ground information, which can be valuable to the succeeding tunnel construction. Based on the offshore exploration technique, an onshore instrumented drilling system that uses a set of pressure transducers placed at various locations of the hydraulic circuits of the machine has been developed and tested on site for ground investigation in order to examine its power to discriminate between ground strata. The method can theoretically be used in nearly every soil type for qualitative interpretation of soil formation changes while grout holes are being drilled. Because of its relatively small sampling interval - at every 5 mm - it can offer to assess local variability of soil formation, which can sometimes be missed by conventional site assessment methods.
In this study, the performance of this system was validated by testing at Kennington Park and Jubilee Line Extension site, in London. The result was then compared with the soil formation obtained from cores samples. Certain components of the drilling system and drilling method have been studied and standardized so that meaningful and globally applicable results can be obtained in the future. The effects of inconsistent drilling system and methods on the quality of data have also been quantified by means of cross-correlation analysis for groups of standardized and non-standardized tests.
Keywords: instrumented drilling; borehole; ground investigation; site characterization.
Introduction
The concept of instrumented drilling for subsurface investigation has long been applied in the oil and gas industry (Somerton 1959), but it is still a comparatively new concept in on-shore geotechnical engineering. This concept arises from the observation that an experienced driller can actually feel the type of the materials being drilled. If the drilling rig is instrumented and monitored via a computer, these drilled holes can be used to qualify a general view of the soil formations and soil properties through certain drilling parameters prior to a detailed ground investigation. Alternatively, a few conventional ground investigations can first be carried out prior to carrying out further monitored drill holes. The first on-shore application was via a system called ENPASOL (Hamelin et al. 1982) that was capable of recording various drilling parameters such as bit torque, bit downthrust and drilling speed. It is typically used on rotary destructive drilling rigs and can sometimes be use on coring rigs. In the latter case, the main use is to complete the core description in case of poor recovery.
One motivation for the development of such system is to capitalize on the “fine” data, which is available whenever holes are drilled. In tunneling projects, for example, it has become common to install ‘tube-a-manchettes’ grout tubes during the pre-construction phase of the project, so that grout injections can later be conducted to suppress settlements caused by ground loss. The drilling parameters obtained from the drilling of these grout holes may then be used either to confirm local stratigraphy or to determine the appropriate grout to be used (Guillaud and Hamelin 1992; Chun et al. 1999)
The quality of the data obtained from the drilling system varies depending on the drilling system and drilling method used, and interpretation of the drilling data remains largely qualitative. With the expansion of the drilling data base as a result of its deployment in the underground and grouting works, we hope to improve the understanding of the drilling data quantitatively. However, in order to obtain a better correlation, certain components of the drilling system and drilling method needed to be standardised so that the correlation derived from different sites can be applied globally without any modification.
The main purpose of this paper is to report on the findings of a study on instrumented drilling system from the improved technologies concerned with compensation grouting during underground excavation (Buchet et al. 1999; Soga et al. 1999). This paper will firstly discuss the development of an instrumented drilling system and describe the characteristics of the measured drilling parameters. It will then suggest a standardized drilling system in term of hardware and drilling method before good quality data can be achieved. To examine its application in the field, a brief description of the field tests conducted at Kennington Park and Jubilee Line Extension sites in London is also given. Finally, the repeatability and reliability of the drilling data obtained from the two sites will be mathematically quantified using the cross-correlation analysis.
background
The drilling of holes for grouting tubes creates the potential for obtaining supplementary ground information, which can be valuable to the succeeding tunnel construction. Based on the offshore exploration technique, an onshore instrumented drilling system for ground investigation has been developed and tested on sites in order to examine its power to discriminate between ground strata. Empirical correlation between soil types and drilling parameters could enhance the data previously acquired from site investigation, but drilling data varies with drilling equipment and the precise way it is used. Attempts to correlate soil information with drilling results using various types of drilling rig or drilling bits have been found to be unsatisfactory (Girard, 1985). In order to obtain better correlations, certain components of the drilling system needed to be developed and standardised. By developing and standardising the drilling system and the drilling method, drilling correlation derived from different sites may then be applied globally without any modification.
Description of Drilling Parameters
There are seven drilling parameters that can be recorded during an instrumented borehole drilling. They are: drilling fluid or mud pressure, torque, down-thrust, holdback pressure, rotation speed, drilling speed, and time to drill 5 mm. Each of these drilling parameters has certain obvious correlations with soil characteristics. For example, if other parameters remained unchanged while the drilling speed increases, it indicates that a looser or softer material is being encountered. A brief description of each parameter is given below (Gui et al, 2002):
Development of an Instrumented Drilling System
Drilling machine
Drilling parameters are strongly affected by the advancing speed that is dictated by the driving power. It is believed that if the driving power is too high, it will be impossible to identify the layering of the site formations. Furthermore, the machine size must be compatible with the need for it to work in especially the relatively constricted grouting shafts.
The drilling process must be of the rotary method instead of the roto-percussion method (Hamelin et al, 1982). Thus the most suitable drilling machine is of rotary type (see Figures 1 and 2) and that it has to be hydraulically driven because all the pressures are measured through the hydraulic circuit which is impossible via an electrical system. It is also recommended that the system should adopt drilling rigs with a minimum stroke of 3 m together with a high-torque rotary drill head. The drilling rod and bit diameter are generally depend on the diameter of ‘tubes-a-manchettes’ but, in most cases, API rods of 89 mm diameter or the ‘Tricone’ bit of 114 mm in diameter be adopted (Figure 3).
Figure 1. A typical instrumented rotary drilling machine that can be used to drill at various drilling angle.
Figure 2. Instrumented borehole drilling in progress.
Figure 3. Drilling rods and drilling bits used for drilling.
As mentioned earlier, there are seven drilling parameters that may be recorded during an instrumented borehole drilling. These parameters are drilling fluid or mud pressure, torque, down-thrust, holdback pressure, rotation speed, drilling speed, and time to drill 5 mm. In general, the drilling fluid, torque, down-thrust and holdback pressures can be measured using pressure transducers; rotation speed can be measured using an electromagnetic proximity sensor; while drilling speed and time can be easily measured via a movement transmitter sensor and an internal clock, respectively (Table 1). The locations of the sensors and transducers can be either down-hole or close to the bit or on the machine itself (Tani, 1998).However, because of the mechanical friction along the rod and the leakage at the rod joints, it is desirable to install the instrumentation as closer as possible to the drilling bit. Compromise must therefore be made for the limited spaces and unfavourable conditions for accurate measurement around the bit (Tani, 1998). For example, fluid pressure may be best measured at the bit, but because of the impracticality of placing a transducer near the nozzle the pressure is measured adjacent to the pump at the ground surface. The accuracy and locations of each instrumentation installed on a drilling machine are shown in Table 1.
Table 1. Typical details of the instrumentation used and their locations on the drilling machine
|
Parameters |
Instrumentation | Accuracy | Locations |
| Fluid Pressure | 35 bars transducer | ±3 % | As near as possible to the mud pump to avoid fluid pressure losses. |
| Torque | 250 bars transducer | ±3 % | As near as possible to the rotation head to avoid head losses and possible interference with hydraulic fittings (valves). |
| Down-thrust | 250 bars transducer | 3 % | As near as possible to hydraulic cylinder to avoid head losses and possible interference with hydraulic fittings. |
| Holdback | 250 bars transducer | ±3 % | - ditto - |
| Depth,
Drilling Speed, Time to drill 5 mm |
Movement transmitter sensor and internal clock |
±3 % | Fixed at the top of the drilling mast and connected to the rotation head by a rope (or a wire). It transmits continuously the head position as well as the feed speed at any time. |
| Rotation Speed | Electromagnetic proximity sensor | ±3 % | Located on the gear of the rotation head. |
Data Acquisition System
A data acquisition, processing, and displaying system have been developed to acquire and display data on screen in real time. Analogue signals from the measuring devices (Table 1) are first converted into digital signals via a junction box fitted on the drilling rig. The junction box is connected to a portable data acquisition system that is called Enpasol (Figure 4). The data acquisition system is a specially designed real time data logger which monitors, measures and stores the drilling data; producing a record at 5 mm intervals of the characteristics of the formation being drilled. Drilling data can also be processed directly and printed in real time on a built-in printer on this data acquisition system. Because of its flexibility this type of system has been selected for use by the U.S. Army Engineer Waterways Experiment Station (Smith 1994) and used in many other projects such as the Storebaelt Eastern Railway Tunnel in Denmark (Pazuki and Doran 1995).
Figure 4. Enpasol real time data acquisition system that has a built-in printer.
The data can be also be stored onto a PCMCIA type storage, which can then transfer these data onto a desktop computer for further processing. To retrieve the stored data, a computer program called “JOE” must be used (Soletanche-Bachy 1999). JOE is a database management software written to retrieve and process the drilling data. It is a Window based program, which incorporates a user-friendly graphical interface that allows the user to select any particular parameters (Figure 5), perform noise filtering and statistical calculations, and subsequently obtain hard copy of the selected drilling parameters. It also allows the digital data to be exported to other Window based software such as EXCEL and AutoCAD for further data manipulation (Figure 6).
Figure 5. An example of JOE software display.
Figure 6. Importing and exporting data between Window-based softwares.
Drilling method
The drilling method used has been found to be very important as it affects the drilling process and the quality of the data. The drilling method studied here was mainly concerned with the control of the drilling fluid flow rate, rotation speed, and down-thrust by the operator. It must be emphasized that throughout the drilling process a relatively constant drilling fluid flow rate, rotation speed and thrust on bit must be provided in order to obtain consistent data. There are a few notes that the drilling operator must be aware of, i.e.:
It is difficult to decide the standard values for these three parameters. In fact, most of the time they are determined on site to suit the drilling condition (e.g. the lifetime of the drilling bit) and the soil conditions. The values of these three parameters are generally adjusted at the beginning of the drilling work in order to obtain the best compromise between the output and the instrumented drilling investigation.
Needs for calibration with borehole information
Before performing any instrumented borehole drilling for subsurface investigation, it is generally necessary to calibrate the drilling results with existing geological and /or geotechnical information (borehole records, SPT, pressuremeters, etc.). This exercise allows a preliminary analysis and the result can be used:
The nature and the amount of borehole calibration depend on the type of the information required as well as the heterogeneity of the geology.
Fields Test and test results
The instrumented drilling system was tested and validated by comparing the soil formation obtained from conventional borehole investigation in London. For field calibration it is essential to conduct tests at sites with consistent and homogeneous soil deposits to minimize the effects of soil variability on the measured drilling data. Two series of field tests have been performed at the Kennington Park and Jubilee Line Extension (JLE) sites, both in London. Tests performed at Kennington Park formed part of the ground investigation work for the risk assessment of the condition of existing old tunnel lining (Gourvenec et al, 1999). Tests performed at the JLE site were mainly the grouting contractor’s own initiative to correlate drilling data, obtained during the drilling of ‘tube-a-manchettes’ holes, with grout properties.
Kennington Park
London Underground Limited (LUL) had arranged to investigate the condition of the existing lining of their Kennington Loop Tunnel on the Northern Line and required an understanding of the ground conditions surrounding the tunnel. The 3.6 m diameter tunnel was constructed in 1924 and is contained within London Clay with its axis about 21 m below ground level. The ground investigation included rotary coring, self-boring expansion pressuremeters, self-boring load cells, self-boring permeaters, vibrating-wire piezometers, and the instrumented drilling tests. Details of the in situ testing can be found in Gourvenec et al. (1999). The general soil formation obtained from the two high quality rotary core boreholes, which had been sunk to a depth of 30 m and 54 m below ground level, is shown by the dotted line in Figure 7.
Six instrumented drilling tests have been conducted on this site. All the tests were vertically drilled to a depth of 30 m, except for one test which extended to a depth of 50 m, using the drilling rig and standardized drilling methods described earlier. It was necessary to place a casing through the Terrace Gravel layer to maintain the stability of the borehole, beyond which the rods were advanced to the full depth.
Jubilee Line Extension site
The Jubilee Line Extension (JLE) is the extension to the Jubilee Line from its old termination at Green Park. It provides vital links to southeast London, Docklands and Greenwich, and extends the existing Jubilee Line eastward for almost ten miles. During the construction of the underground tunnels compensation grouting was used at the location of large underground stations, such as Southwark Station and Bermondsey Station, and at the location of an interconnection of several tunnels, i.e. Red Cross Way. Compensation grouting had been designed to compensate excessive ground settlements. A total of 3854 ‘tubes-a-manchettes’ holes were drilled at the JLE Contract 103 site in 1994 for this purpose. Out of these, a total of 376 holes were logged by the drilling system because the grouting contractor had decided to design their grout properties based directly on the drilling parameters. Due to the nature of the compensation grouting work, these holes were drilled using different types of drilling machine, e.g. KLEMM 804 and SM 305, non-standardized drilling methods, and at various drilling angles between 90o and 50o to ground level.
Test Results
Figure 7. A typical Enpasol result
(first 8 m was drilled with casing; soil layers were confirmed using rotary coring samples.)
A typical drilling parameters output obtained from the instrumented drilling test at Kennington site is plotted in Figure 7. The soil formation of this site revealed by cores samples is also annotated in this figure. In general, the following observations can be made:
It was found that the “noise” in the raw data could be used to quantitatively distinguish soil formation, in particular, sandy gravel made noise in both torque and speed whereas stiff plastic clay created noise in mud pressure. These mechanisms are created by bit-soil interactions – jamming and clogging – respectively. In this sense “noise” is a paradoxical description of the variability of the data, since it can be used for ground characterization. However, proper filtering of the noise perturbations, correcting data corrupted by faulty equipment, and compensating for any environmental effects such as temperature and humidity may still required before extracting the best information from the raw signal for quantitative soil characterization.
Reliability test of Drilling Data
Instrumented drilling is a relatively new ground investigation tool, therefore it is vital to quantify the repeatability and thus the reliability of the drilling data. There may be several ways to check the repeatability of a test. Perhaps, the most direct way is to perform a few tests at a homogeneous site such as the Kennington site and then compare the test data. However, as shown in the previous section, the data obtained is highly susceptible to random noise, which makes such comparison questionable. Thus a more scientific method of comparison is required. Cross-correlation function (CCF) is capable of indicating similarity of two signals as a function of the delay between them. Here, the CCF will be used to find out the similarity of the drilling data obtained from the Kennington Park and JLE sites. The tests performed at Kennington Park were so-called ‘standardised’ test because these tests were performed using the same testing procedures, instrumentation and equipment (see Gui et al. 1999, 2002 for details). On the other hand, tests performed at JLE site are ‘non-standardised’ because they were performed using unspecified but generically similar testing procedures, different instrumentation and equipment.
The cross-correlation of two signal sequences x[n] and y[n] is a third sequence rxy[j], defined as
(1)
where subscript x is a drilling parameter and subscript y is the same drilling parameter but from the other test. For example, the torque parameter of test ENP1 is compared with the torque parameter of ENP2, ENP1 with ENP4, etc. The second sequence y[n] is delayed by j units relative to the first sequence x[n], and the sum of the product terms is then evaluated. This is done for all values of j. To obtain a better comparison, the cross-correlation result is normalized against the auto-correlation values of both signals being compared as follow:
(2)
A good cross-correlation of any two tests is defined by a triangular shape of CCF profile and a close to unity CCF value at zero lag.
Standardized test at Kennington
Cross-correlation was performed for each of the drilling parameters of tests ENP1, ENP2 and ENP4 in Figure 8. Qualitatively, the overlapping or near-overlapping of CCF profiles in mud pressure, torque, and net-thrust in Figure 9(a), (b) and (c) show that there is consistency in the quality of the data between each of these standardised tests. From the quantitative point of view, the normalised cross-correlation value for mud pressure, torque and net thrust is about 0.7 which means that these results are reasonably correlated and thus there are reliable. However, the correlation between the drilling speed data, Figure 8(d), is poor where it only has a maximum normalised CCF value of about 0.4. In this case, the data is not very consistent.
Figure 8. Cross-correlation function to examine the repeatability of Kennington data.
Non-standardised test at JLE
The drilling tests performed at the JLE site were non-standardized ones. The machine, the instrumentation and the drilling procedures were not standardized like the Kennington test. Again, the quality of these non-standardized data can be compared using the cross-correlation function. Three vertically drilled tests, JQ05, JQ07 and JQ09, which were only a few metres away from each other, were randomly selected for this analysis. If these were standardized tests, a high degree of similarity between these tests would have been expected, as for the Kennington test data. However, from their normalized CCF results plotted in Figure 9, except for the torque and net-thrust in Figure 9(b) and (c), the CCF profiles for each of the drilling parameters do not overlap each other, i.e. inconsistencies exist in the mud pressure, drilling speed and time parameters. The non-overlapping and non-symmetrical envelopes are more severe for the drilling speed and time parameters, Figure 9(d) and (e). These data are thus less unreliable than they need be. The unreliability and inconsistency must be caused by the non-standardised method of drilling.
Figure 9. Cross-correlation function to examine the repeatability of JLE Contract 103 data.
Conclusion
An instrumented borehole drilling system, which uses a set of pressure transducers placed at various locations of the hydraulic circuits of the machine, was developed for subsurface exploration. The method can be used in nearly every soil type for qualitative interpretation of soil formation changes while grout holes are being drilled. Because of relatively small sampling interval, it can offer to assess local variability of soil formation, which can sometimes be missed by conventional site assessment methods.
In this study, a standardized instrumented borehole drilling system, in terms of hardware and drilling method, was studied and its capability in discriminating soil strata examined. The performance of this system was validated by testing at Kennington Park and Jubilee Line Extension (JLE) site, in London and the result was then compared with the soil formation obtained from the cores samples. The consistency of the data was mathematically quantified using the cross-correlation technique and it was found that the results obtained from the standardized Kennington tests were more consistent than the non-standardized JLE tests. This shows that proper standardization such as the one described in this paper needed to be adopted for any instrumented drilling data from subsurface investigation to be useful for subsequent data interpretation.
References
 | |
| © 2004 ejge | |
