Dan Brenton, Andy Condron, ERS Remediation and Callum Whitelaw, Geosonic Drilling
Arup Tech Paper logo (NB)
Sonic drilling uses the theory of sonics, which was initially presented by Romanian scientist Gogo Constantinescu (Constantinescu, 1918) and was developed in the 1940s in the oil industry in Russia and the US (Bruce and Despres, 2004). Modern sonic drilling was developed in the 1980s by Canadian mechanical engineer Ray Roussy (Clancy, 2006).
A sonic drilling rig appears very similar to a conventional rotary coring rig. The primary difference is the hydraulically powered drill head. Inside the drill head two counter-rotating weights produce adjustable high frequency vibratory forces. The frequency of vibration can be varied to suite the ground conditions and varies between 50Hz to 150Hz (BSI, 2015). A pneumatic or mechanical isolation system prevents the resonant energy from transmitting to the drill rig and preferentially directs the energy down the drill string.
The sonic rig employs vibratory force to provide velocity and localised displacement to shear and penetrate the formation. Slow rotary force is also applied to the drill string to enhance vibratory actions and an axial force provides a steady push or pull to aid advancement or retraction.
A typical drilling sequence in soils involves advancing the core barrel into the formation in 1m to 3m runs without the use of a flush medium. A length of casing is advanced over the core barrel, often with addition of water. The inner core barrel is retrieved for sample extrusion and the process is repeated.
The sonic borehole advances by different processes depending on the material encountered. In cobbles, boulders, fill and bedrock the sonic core barrel advances by fracturing the material. Cohesive silts, clays and argillaceous rocks are sheared by the core bit into the core barrel. Boreholes are advanced in granular strata by vibration and displacement of the particles into the core barrel. While drilling is possible in hard rock it is generally more efficient to utilise rotary coring in these materials due to excessive wear on the equipment.
2.0 Sonic drilling in the UK
Drillcorp became the first UK-based drilling contractor with sonic equipment in 2003 when it acquired two Eijkelkamp Sonic Samp Drill sonic rigs. These rigs used the aqualock sampling system mostly to undertake geo-environmental sampling and install arrays of remediation wells.
The largest supplier of sonic drilling rigs in the UK, Geosonic Drilling (formerly Boart Longyear), established an office in Stirling in Scotland in around 2006 (Lennon and Burn, 2008).
Sonic drilling grew in popularity in the following years and, in 2012, the technique was incorporated into the ICE Specification for Ground Investigation (Site Investigation Steering Group, 2012). This has led to an increasing awareness of sonic drilling and, as of 2018, there are nine UK ground investigation contractors with sonic drilling equipment as shown in table 1. Sonic drilling services are also routinely subcontracted in by other non-listed ground investigation contractors.
3.0 London 2012 Olympic Park
The sonic drilling case study presented in this paper is the London 2012 Olympic Park ground investigation.
The Olympic Park, now known as the Queen Elizabeth Olympic Park, is a 256ha site in east London. The site had over a century of mixed industrial land use with significant potential for contaminated land. At the time of acquisition by the Olympic Delivery Authority, land use included car breakers, chemical works, factories, warehouses, railway sidings and many more potentially contaminative site uses.
|Table 1 - 2019 Availability of sonic drilling rigs in the UK|
|Contractor||No. Drilling Rigs|
Geosonic Drilling (formerly Boart Longyear)
Hughes Exploration & Environmental
M and J Drilling Services
Raeburn Drilling and Geotechnical
As part of the enabling works contract extensive ground investigations were undertaken to determine the extent of contamination and provide geotechnical data for design of the stadia and associated infrastructure.
The Olympic Park ground investigations consisted of 3,500 exploratory holes on an approximate 25m grid. The investigations were phased and began under the London Development Agency from around 2005 onwards.
Significant problems, such as slow progress, difficulties pulling casing and poor core recovery were encountered by conventional drilling techniques (cable percussion and rotary coring) in the geology of the site.
The generalised geology of the Olympic Park is summarised in table 2.
|Table 2 - Summary of Olympic Park geology|
|Made ground||Up to 15 m||Highly variable|
|Alluvium||1 to 3 m||Soft silty clay|
|River terrace deposits||3 to 5 m||Sand and gravel|
|Lambeth Group||20 to 30 m||Interbedded clays, silts, sands and pebble beds (Upnor Formation)|
|Thanet Formation||10 to 20 m||Very dense, very silty fine sand|
|White Chalk Group||Unproven||Recovered as soft white putty chalk. Grade Dm (Lord et al, 2002)|
The river terrace deposits and White Chalk Group are aquifers at the site (Ellison, 2004). The presence of two aquifers at the site necessitated the use of clean drilling techniques in accordance with Environment Agency guidance (Fretwell et al, 2006).
In practice this meant cable percussion boreholes commenced in 300 or 250mm casing, reducing casing in the river terrace deposits and in the Lambeth Group to a final hole of 150mm diameter at depth. Commencing boreholes in large diameter casing involves manual handling health and safety issues. Geotechnical boreholes on the Olympic Park were frequently required more than 35 or 40m depth with piling for major structures typically extending into the Thanet Formation. This produced challenging conditions for cable percussion and slow progress through very dense Thanet Formation and stiff clays, dense sands, calcrete and pebble beds of the Lambeth Group.
Historical rotary coring on the site achieved very poor core recovery in the Lambeth Group and is difficult in the Thanet Formation. Problems with core loss in the Lambeth Group are well documented (Hight et al, 2004).
3.1 Sonic drilling trial at the Olympic Park
dan brenton fig 1
A trial of sonic drilling on the Olympic Park commenced in May 2007 with a Boart Longyear Delta Base 320 (DB320) rig undertaking investigations around the proposed main stadium.
The sonic rig was employed to perform the same frequency of geotechnical and geo-environmental sampling and in-situ testing regime as the cable percussion boreholes.
Standard penetration testing (SPT) was undertaken in geotechnical boreholes alternating with undisturbed U100 samples at metre intervals to 10m, 1.5m intervals to 20m depth and 3m intervals thereafter.
Granular soils were extracted from the sonic core barrel by gentle vibration from the core barrel into a 1.5m cylindrical clear plastic bag (colloquially named “sausage bags”) as shown on figure 1. The bag samples were split, logged and photographed in the same manner as a dynamic sample and discrete bulk and small disturbed sub-samples were taken (figure 2).
dan brenton fig 2
Where undisturbed U100 samples were specified in cable percussion mode the sonic rig undertook U86 sampling. The U86 samples are 1.5m long, 86mm diameter with rigid PVC liners (figure 3) that are extracted from the sonic core barrel in the same manner as a rotary core liner is extracted. The U86 samples were then used for strength (triaxial) and consolidation (oedometer) testing along with cable percussion U100 samples.
dan brenton fig 3
Source: Geosonic Drilling
After a number of successful holes, the sonic rig was retained on the Olympic Park until the end of February 2008. A total of 79 boreholes were undertaken using the DB320 rig, at the time this was the largest case study of sonic drilling ever undertaken in the UK.
From the trial, 43 holes were for contamination purposes only, with average depth of 27m. The remaining 36 were for geotechnical and geo-environmental purposes. The average rate of daily progress over this period was 15m a day and the maximum depth achieved in one day was 30m, in a borehole with geo-environmental sampling only. All but three of the boreholes were undertaken on the southern Olympic Park, delineated by the Hackney Wick to Stratford London Overground line. Three boreholes up to 70m depth were undertaken on the northern Olympic Park at the velodrome to investigate a suspected periglacial pingo feature (drift filled hollow). The detailed geology of this case study is described in Lee and Aldiss (2012).
The disturbed nature of the ground in this area with thick sequences of granular materials would have been almost impossible to continuously sample using conventional cable percussion and rotary coring techniques. A similar drift filled hollow was subsequently encountered during the construction of the Lee Tunnel and was also investigated using sonic methods (Bellhouse et al, 2015). This suggests that sonic drilling is a suitable method of investigating such features.
3.2 Sample quality
Limited information on sonic drilling sample quality or disturbance is available in the literature. Fretwell et al. (2006) describes vibration of the drill bit causing heat that can lead to the volatilisation of volatile organics. Baldwin and Gosling (2009) describe trials that have indicated sonic samples exhibit significant disturbance and have margins dried by heat generated during the drilling. Bellhouse et al (2015) describe the pros and cons of sonic drilling, including continuous sampling without water of coarse granular materials and alteration of clay strength by high temperatures and breaking of calcite bond in chalk. No laboratory test data is included in these references.
On the Olympic Park the sonic rig produced near continuous high-quality samples in all strata. The skill and experience of the operator and the ground conditions themselves remained a factor. For the Olympic Park trial the same drilling crew and equipment were used throughout.
The fabric of laminated layers in the Lambeth Group was retained and observed both in U86 samples and in non-rigid samples after subsequent splitting and logging. For some sonic locations the continuous samples were retained in wooden core boxes for visual inspection by the designers (figure 4).
dan brenton fig 4
This practice enabled retention of fine detailed lithological information in cohesive and granular strata that is typically lost after sampling has occurred and boreholes are completed (figure 5).
dan brenton fig 5
Geo-environmental test results from sonic boreholes were included in the datasets for detailed quantitative risk assessments undertaken for human health and controlled waters at the Olympic Park. Geotechnical results were also incorporated in datasets used by geotechnical designers for structures at the park.
The evidence of disturbance from sonic drilling would be most significant in cohesive strata. At the Olympic Park these were the superficial Alluvium and the Lambeth Group. The natural moisture content of free draining granular soils is not typically considered an important geotechnical parameter as described in BS1377-4 Cl.3.1.1 (BSI 1990).
The available geotechnical test data from the southern Olympic Park (approximately 1,500 exploratory holes) has been analysed and is presented (figures 6 to 8).
A reduced natural moisture content would be expected if drilling induced heat was a significant factor. The available natural moisture content for the Alluvium (figure 6) and undifferentiated Lambeth Group (figure 7) from both sonic and cable percussion boreholes is presented.
Data from cable percussion boreholes was from a variety of contractors and laboratories from a mixture of bulk, disturbed and U100 samples. The Olympic park GI was carried out before implementation of Eurocode 7 or development of the UT100 sampler. The sonic data is all from the same contractor and laboratory and there is no noticeable reduction in natural moisture content observed in the sonic datasets, the average values are remarkably similar to cable percussion data as shown in tables 3 and 4. A large maximum value recorded in the Alluvium is likely to be from a peaty layer which has been interpreted as Alluvium. The data from the Lambeth Group is presented together. The individual member beds (Woolwich, Reading and Upnor Formations) have been interpreted and show the same pattern albeit in smaller datasets.
Another likely result of sample disturbance would be a variation in undrained shear strength of a cohesive soil as described by Bellhouse et al (2015). The characteristic undrained shear strength being an important parameter in the design of deep foundations in cohesive soils (Tomlinson and Woodward, 2008).
The available Olympic Park undrained shear strength data from sonic and cable percussion boreholes for the Lambeth Group (figure 8) is presented. The undrained shear strength dataset from the Alluvium was too small to make any conclusions and is not presented. The undrained shear strength data from the sonic boreholes in the undifferentiated Lambeth Group (and constituent formations) does not show significant variation from the cable percussion data. There is significant natural scatter in the undrained shear strength of the Lambeth Group as described by Hight et al (2004). However, the range of values and average from both cable percussion and sonic drilling is similar and within the range described from other large projects in the Lambeth Group such as the Channel Tunnel Rail Link and the Jubilee Line Extension. The geotechnical laboratory test data from the Olympic Park is summarised in tables 3 and 4.
|Table 3 - Summary of Olympic Park Alluvium geotechnical laboratory test data|
|Natural moisture content (%) |
|Natural moisture |
|Table 4 - Summary of Olypic Park Lambeth Group geotechnical laboratory test data|
|Natural moisture |
|Natural moisture |
|Undrained shear |
Disturbance in a cohesive soil should be apparent in oedometer test results from undisturbed samples. An increase in specimen disturbance results in a slight decrease in the slope of the virgin compression line on a voids ratio vs log effective stress plot (Craig, 2004). The coefficient of volume compressibility (mv) represents the decrease in voids ratio for increasing effective stress. The mv data from the Olympic Park sonic and cable percussion samples of the Lambeth Group are presented in table 5.
|Table 5 - Summary of Olympic Park Lambeth Group oedometer test data|
|Stress range (kPa)||Coefficient of volume |
|Coefficient of volume |
|0 to 100||0.02-1.02||0.36||0.39||10||0.18-0.19||0.18||0||2|
|100 to 200||0.005-0.91||0.14||0.16||118||0.01-0.19||0.09||0.05||19|
|200 to 400||0.01-0.50||0.09||0.09||243||0.02-0.59||0.09||0.08||61|
|400 to 800||0-0.18||0.06||0.04||126||0.01-0.17||0.07||0.04||37|
|800 to 1600||0.01-0.10||0.04||0.02||33||0.03||-||-||1|
|1600 to 3200||0.02-0.04||0.03||0.01||7||-||-||-||0|
The dataset was limited to one sonic sample in the Alluvium and therefore the results are not presented. The mv values from the Lambeth Group sonic dataset are on average remarkably similar, particularly at the stress range representing the in-situ conditions on the stratum at the site (approximately 250 to 350kPa). From the above data it is concluded that sonic drilling can provide results similar to that of cable percussion boreholes.
Current codes of practice for geotechnical investigation and testing namely BS EN ISO 22475-1 (BSI 2006) state that sonic drilling can only achieve sample categories in accordance with BS EN 1997-2 (BSI 2007) of category B class 4 in cohesive soils and category C class 5 in granular soils.
BS 5930 (BSI 2015) notes in clause 25.9 that arguably classes 3 and 4 are more realistic, the former being on the proviso that undue heat is not generated by the drilling equipment.
The BS EN 1997-2 (BSI 2007) sample classes imply that in cohesive soils only the sequence of layers, broad strata boundaries, particle size distribution, Atterberg limits, particle density and organic content of the sample can be determined. In granular soils only the sequence of layers can be determined. It is noticeable that the standard is dated 2006 which is prior to the vast majority of sonic drilling that has been undertaken in the UK to date.
The Olympic Park ground investigation observed excellent recovery of granular deposits including the fine detailed lithology often obscured by cable percussion methods. A study for the aggregates industry (Jeffrey et al. 2011) concluded that particle size distributions from sonic boreholes were more representative of quarry face samples to those from flight auger, reverse circulation rotary drilling and cable percussion.
Most of the Olympic Park data presented are in heterogeneous Lambeth Group materials which present challenges to traditional ground investigation techniques. Sonic drilling is often employed when these techniques fail to progress through obstructions or poor recovery occurs. A more robust comparison of sample quality could be made in a more heterogenous material such as London Clay. However, sonic drilling would be unlikely to be specified in these materials on a commercial project where traditional methods are usually successful and far more economical. Therefore, perhaps further work in academia is needed on sample disturbance from sonic drilling.
The case study from the Olympic Park is comparing data from prior to the implementation of the Eurocode sample classes. It could be argued that this case study is comparing one non-compliant dataset with another. However, it is common on sites with phases of historical ground investigation data to make use of it for current designs using engineering judgement. A designer would not dismiss historical data simply because it did not comply with modern codes.
The Olympic Park case study has demonstrated sonic sample quality not be significantly different to traditional methods. The case studies presented suggest a category A class 1 in certain cohesive soils (Lambeth Group) and category B class 3 in granular soils could be achieved.
Given the time that has elapsed since the publication of the Eurocode sample classes, more data from sonic drilling in the UK may be available that could be used to compare with other techniques. Perhaps a review of the current sonic drilling sample classes is required. With increasing awareness that undisturbed U86 liner samples (figure 9) are retrievable by sonic drilling, designers may be more willing to specify its use and further data may become available.
dan brenton fig 9
Source: Geosonic Drilling
Thank you to David Dennis (Geosonic) for his continual support and encouragement.
Baldwin, M. & Gosling, D. 2009. Technical Note BS EN ISO 22475-1: Implications for geotechnical sampling in the UK. Ground Engineering, August 2009, 28-31.
Bellhouse, M.R., Skipper, J.A. & Sutherden, R.N. 2015. The engineering geology of the Lee Tunnel. Preceedings of the XVI ECSMGE Geotechnical Engineering for Infrastructure and Development. ICE Publishing, London.
BSI. 2006. BS EN ISO 22475-1:2006 Geotechnical investigation and testing – Sampling methods and groundwater measurements – Part 1: Technical principles for execution, 16-17. British Standards Institution, London.
BSI. 2015. BS 5930:2015 Code of practice for ground investigations, 87-88. BSI, London.
BSI. 1990. BS1377-4:1990 Methods of test for Soils for civil engineering purposes – Part 4: Compaction-related tests, 1-2. British Standards Institution, London.
BSI. 2007. BS EN 1997-2:2007 Eurocode 7 – Geotechnical design – Part 2: Ground investigation and testing, 33-34. British Standards Institution, London.
Bruce, D, A. & D, L. Depres. 2004. Drilling and Sampling of Embankments Using the Sonic Drilling Method, ASDSO Annual Conference, Phoenix, AZ, September 26-29, 9 pp.
Clancy, L. 2006. Aviation engineering heads down to earth. Geodrilling International, December 2006, 40-42.
Constantinescu, G. 1918. Theory of Sonics: A Treatise on Transmission of Power by Vibrations. The Admiralty, London.
Craig, R. F. 2004. Craig’s Soil Mechanics, 7th edition, 381-382. Spon Press, London and New York.
Ellison, R. A. 2004. Geology of London. Special Memoir for 1:50,000 Geological sheets 256 (North London), 257 (Romford), 270 (South London) and 271 (Dartford) (England and Wales). British Geological Survey, Keyworth, Nottingham.
Fretwell, B. A., Short, R. & Sutton, J. S. 2006. Guidance on the design and installation of groundwater quality monitoring points, Environment Agency Science Report SC020093, 10-11.
Hight, D. W., Ellison, R. A. & Page, D. P. 2004. CIRIA C583 Engineering in the Lambeth Group. CIRIA, London.
Jeffrey, K., Hill, I. & Hameed, A. 2011. Deposit Knowledge for Efficient Production. MIST Project MA/7/G/5/002, University of Leicester.
Lee, J. R. & Aldiss, D. T. 2012. Possible Late Pleistocene pingo development within the Lea Valley: evidence from Temple Mills, Stratford, East London. British Geological Survey Commercial Report, CR/11/033. 25pp.
Lennon, D, J. & Burns, C. 2008. Driven Piled Foundations Through Buried Dock Walls Proceedings of the BGA International Conference on Foundations, Dundee, Scotland, 24 – 27 June 2008. IHS BRE Press, 2008.
Lord, J, A., Clayton, C, R, I. & Mortimore, R, N. 2002. CIRIA C574 Engineering in chalk. CIRIA, London.
Tomlinson, M. & Woodward, J. 2008. Pile design and construction practice, 7th edition.151-154. Taylor & Francis, London and New York.
Site Investigation Steering Group. 2012. UK Specification for Ground Investigation 2nd edition. ICE Publishing, London.