An issue faced by young geotechnical professionals is when to terminate the Standard Penetration Tests (SPT) when advancing a borehole during a geotechnical investigation. This issue commonly arises in the soil-rock interface zone where the ground mass is transitioning from a soil to a rock. Commonly in New Zealand, a cut-off figure of SPT ‘N’ value >50 is used. However, depending on the end use (what will be constructed), some clients (and colleagues in different disciplines, e.g. structural engineers) want higher ‘N’ readings to provide stronger evidence of a full transition to rock strength. Some weathering profiles include hard inclusions (core-stones) and recovery in such horizons can often be poor. This paper examines the theories and realities behind what is actually being measured. A soil mass behaves differently from a rock mass, and as such the strength of each is measured and characterised in a different way, applying different criteria. This paper sets out approaches developed to apply SPT results to different situations, drawing examples from both soil strata and very weak rock units. The paper concludes with some simple guidelines that have been found helpful in developing engineering judgement with regard to SPT use in the soil-rock interface zone.
Young geotechnical professionals (0-5 yrs) are often sent out to do geotechnical ground investigations, with little or no understanding of how a Standard Penetration Test (SPT) works, or what it actually measures. Without this grounding in the theory, it is difficult to make a decision about the termination of the SPT on site. The SPT is primarily a test for soil and soft rock (UCS 0-20 MPa), and as such the termination of the test is usually in the vicinity of the soil-rock interface. This interface is highly variable in different environments and is primarily a factor of weathering and rock type. Weathering affects the depth and sharpness of the interface, and different rock types can weather in different ways, under a similar climatic regime (Fooks, Pettifer, & Waltham, 2015). The variety in site conditions and engineering requirements means that there can be no universal criteria for the termination of the SPT.
This paper presents a brief outline of the theory behind the SPT and how it works. Various correlations and corrections have been made over the years to standardise the penetration test and allow comparison between different soil types and measuring apparatus. The importance of understanding the difference between soil and rock is explored and how the strength of each is ascertained for design. This paper aims to provide some simple guidelines that have been found helpful in developing engineering judgement with regard to SPT use in the soil-rock interface zone for a variety of rock types.
2 The standard penetration test
The Standard Penetration Test (SPT) is an in-situ dynamic penetration test (Clayton, 1995) that measures the resistance of soils to a split spoon sampler (Figure 1) or solid cone (Figure 2). The SPT is an easy way of gaining additional information from geotechnical investigations at very little additional cost. Because of this advantage the SPT, while reasonably unsophisticated, is likely to remain in use for some time.
In New Zealand the test is conducted in situ down hole by driving a split spoon sampler or solid cone into the ground in increments of 75 mm for a total of 450 mm using a 63.5 kg weight, falling from a height of 760 mm. The test is divided into two parts, the seating drive is the initial 150 mm of the test and the test drive is the remaining 300 mm. The blows recorded for the seating drive are not included in the reported result (N), which is the combined number of blows recorded for the test drive.
The raw N-value or NSPT, is the uncorrected blow count and is subject to several influencing factors, including method and equipment used (Clayton, 1995). It is common practice in New Zealand to use the automated method of hammer release, which reduces much of the variability. However, in order to be correlated with any other test, including the equivalent SPT N given in CPT results, the N-value must be converted into N60 (Look, 2019). Every rig or apparatus used for undertaking the SPT should have a current calibration certificate which allows the correction to be calculated.
Skempton (1986), published an equation that allowed for the use of hammers of varying efficiencies and corrected for varying field procedures and apparatus.
Where N60 is the corrected SPT N-value, EM is the hammer efficiency, CB is the borehole diameter correction, CS is the sample barrel correction, CR is the rod length correction, and N is the raw SPT N-value recorded in the field. The corrected N-value is given as N60 because the original SPT apparatus had a hammer efficiency of 60%, and all SPTs are referred back to this original test to allow a comparison between results.
In 1986 Liao and Whitman published a correction for the overburden pressure which may influence the SPT result.
In this correction the overburden pressure is removed from the N60, where σ’z is the vertical effective stress from the depth of the test.
The solid cone was developed for use in gravelly soil in an attempt to overcome the effect of gravels larger than the mouth of the split spoon sampler. The solid cone is less likely to be damaged by the gravels and is more able to go through or around the gravel clasts. The solid cone is deemed inappropriate to use in anything other than a gravel as the increased surface area can result in up to 100% increase in the penetration resistance (Clayton, 1995). Despite this, it remains common practice in New Zealand to switch to the solid cone when the ground begins to reach SPT N values above 40. The SPT is now one of the most widely used in-situ tests for gauging the density and compressibility of granular soils and is also used to check the consistency of stiff or stony cohesive soils and weak rock (Clayton, 1995).
3 Soil Mass Properties
The engineering behaviours of soil are governed by many factors including effective stress, true cohesion, and internal friction (Skempton & Bishop, 1950). Soils can be broadly divided into two basic groups, granular and cohesive. The SPT procedure is more reliable in sands than in cohesive soils (Peck, Hanson, & Thornburn, 1953), because silts and clays exhibit different driving resistance when dry or moist (Rodgers, 2006). The strength of the ground is difficult to predict because the soil moisture can change, thus changing the strength of the ground.
The behaviour of granular soils is controlled by the effective angle of friction which according to Clayton (1995) is a function of:
- stress level,
- grain size distribution,
- void ratio (relative density)
- Cementation – Strength can be increased by particle interlocking and/or cementation.
- Ageing – density and interlocking tend to increase with age, and also penetration resistance above and beyond that expected as a result of the density increase.
Even in granular soils, the rate of testing means that excess pore water pressure will be generated and the conditions will be somewhere between ‘drained’ and ‘undrained’ (Clayton, 1995). The action of penetration leads to shearing, dilation and collapse.
The penetration resistance in cohesive soils is largely a function of undrained shear strength (Clayton, 1995) The relationship between the SPT-N value and the undrained shear strength is influenced by a number of factors including (according to Clayton 1995):
- Sensitivity – SPT values are a combination of side friction and end bearing resistance. The end bearing resistance is undisturbed and represents the undrained shear strength, where the side friction is remoulded. Sensitive soils therefore have a lower N-value due to the lower side friction.
- Fissuring (fabric) – If the fissuring aligns with the direction of penetration, then the soil will offer less resistance.
4 Rock Mass Properties
The strength of the rock mass is governed by the intact rock strength and by the density and condition of discontinuities. Intact rock strength can be measured with uniaxial compressive tests, and point load tests can be correlated with geology specific test data to gain an unconfined compressive strength. The estimation of the strength and deformation properties of jointed rock masses however is a more difficult and somewhat empirical exercise. Rock mass classification schemes are one such way of predicting the strength and deformation properties of rock masses. This is achieved by degrading the intact rock properties to rock mass properties (Deisman, Khajeh, & Chalaturnyk, 2013). These include but are not limited to, the RQD (Deere & Deere, 1988)the Q system (Barton, Lien, & Lunde, 1974), the RMR (Bieniawski, 1989) and the GSI (Hoek & Brown, 1997).
SPTs are often carried out in weak and weathered rock. The reported N value is affected by fracturing and weathering of the rock and is not a measure of the intact rock strength (Clayton, 1995). Although calculations have been developed to relate the N-value to UCS, it has been pointed out that the N-value varies with the geology (Look, 2004; Look, 2019). This is largely due to the rock mass fabric including jointing (Look, 2004).
The majority of Auckland, New Zealand, is constructed on a suite of soft Tertiary rocks named the East Coast Bays Formation (ECBF). These rocks exhibit soil-like properties throughout the weathering profile, and an empirical relationship is being developed to classify the weathering of the rock mass, where visual clues are difficult to discern. This classification does not use SPT to directly measure weathering, rather, it is inferred through the general strength of the ground mass. It is important to note that while the ECBF is a flysch deposit, it has never been ‘hard’ rock, can be uncemented, and is much weaker (e.g.(Hodgson & St George, 2008)) than other, older sandstones elsewhere in the world (e.g., Triassic Hawkesbury Sandstone, Sydney Basin (Pells, 2004). In weak and weathered rock, many factors influence the SPT and as such, the interpreted results will be uncertain and imprecise (Clayton, 1995).
5 SPTs in the soil-rock interface
As with all robust geotechnical investigation work, a preliminary geological model should be established prior to undertaking site work. SPT N-values cannot define the soil-rock interface alone and must be put into the context of the logged geology. It is important to differentiate if the ground is behaving in a soil-like or rock-like way and make the interpretation of the SPT based on that information. The following sections provide examples of the soil-rock interface in the common rock types found in Auckland.
Basalt (lava flows)
Basalt lava flows in Auckland can come in a variety of configurations and it is important to consider these when interpreting the SPT results. For example, an SPT of N=50+ has different design implications if it is located in the rubbly margin of a flow, or if it is in solid basalt. It is also useful to remember that lava flows can have soft material underneath or between flows (i.e. ‘stiff over soft’ conditions), and this material may need to be tested too. The weathering profile is often quite sharp, and the ground transitions from soil to rock rapidly. The rock mass can be highly fractured, or can be fully intact, with the intact rock strength usually in the vicinity of 50–100 MPa. As a general rule, SPTs are not undertaken in basalt. This is because conducting an SPT in strong rock damages the equipment and no relevant data is obtained. The rapid transition to competent rock can mean that there is a possibility of not reaching N=50+ at all before terminating the test. The SPT can be terminated if competent rock is identified in the previous run or if the hammer double bounces. Deeply weathered lava flows may leave core stones, so it is important to check the next run or two to confirm the rock head.
ECBF (soft tertiary sandstones and mudstones)
As stated above, ECBF is a soft rock that underlies much of Auckland. The weathering profile is not usually deep and fracture density varies, though is typically low. The unweathered intact rock itself reaches strengths of >20 MPa at its strongest and is a function of the degree of cementation. Caution needs to be exercised when interpreting the N-value in relation to rock strength because the density of the ground does not necessarily translate to the strength of the material. An example of this is the uncemented sandstone which can be found at depths well below the weathering profile. These unweathered sandstones often have an N-value of 50+ but a UCS of <1MPa. SPT refusal is usually around the moderately weathered to slightly weathered interface, and indeed is often used to classify that very transition, where the behaviour becomes rock-like rather than soil-like.
Northland Allochthon (sandstone/mudstone melange)
The Northland Allochthon is a very general term for a suite of rocks classified as a melange resulting from a submarine landslide or thrust sheet (Winkler, 2003). The Northland Allochthon is contemporaneous with the earlier parts of the ECBF and is not a very strong rock due to its young age and material composition. In these rocks even if an N-value of 50+ is reached 3 times, it can be useful to keep going. There have been examples where 3 x N=50+ was reached in one borehole, yet the adjacent borehole never achieved three in a row. This is due to the highly heterogeneous nature of the rock unit which includes concretions that can reach up to 2 m diameter. It is important to bear in mind the purpose of the investigation and consider whether SPTs will assist or not in determining ground conditions. For example, if the purpose of the investigation is to identify a landslide slip plane, then SPTs would be inappropriate as the crucial information could be obscured by the test.
Greywacke (indurated sandstone and mudstone)
Mesozoic-age Greywacke in New Zealand is highly indurated, usually highly fractured and faulted, and commonly veined with quartz. These rock mass and structural defects lead to an undulating and highly variable weathering profile. It is common to find core stones within this material, typically close to the soil-rock interface. Unlike the ECBF, the soil-rock interface in Greywacke is usually at the completely weathered to highly weathered boundary. It is important to note that even in completely weathered greywacke relict rock mass fabric can still control failure modes, even though the strength and behaviour is soil-like. An SPT in moderately weathered greywacke will show that the rock mass is fractured and that the intact rock strength is weaker than unweathered greywacke but will not quantify the degree to which this has occurred. Degree of fracturing and strength loss are better seen in core that has not had an SPT attempted in it. When terminating the SPT, satisfy yourself that you have reached rock head and not a core stone. If there is moderately weathered rock at the base of the run, do not undertake the test in the following run.
6 Guidelines for terminating the SPT
In the New Zealand standard NZS4402.6.5.1-1988, it is specified that the sampler be driven until it penetrates 450 mm into the ground or for a total of 60 blows, whichever comes first. The current rule of thumb is to terminate SPTs after 3 successive readings of N=50+, whether in the seating drive or the test drive. While this practice is convenient and easy to remember, it is not always the most appropriate course of action, and engineering judgement must be exercised. As demonstrated in the examples above, there are situations where fewer than three or greater than three N=50+ are required. Figure 3 is a breakdown of the questions to consider when terminating the SPT.
The SPT cannot be used as a predictive model to determine the weathering grade, rather the SPT needs to be interpreted in the context of the geology. It is vitally important to get an accurate understanding of the local geology, because characteristics other than strength are important. In an Auckland case study the rock head was set at SPT N=50+, which is adequate for pile design. However, the geology was misinterpreted, and the foundations were set in an aquifer. This had major implications for the project, that could have been mitigated if the geology had been correctly identified.
The standard penetration test has been in use since the 1920’s and is one of the most common in-situ geotechnical tests. Despite this it must be used with caution and within the same scale of accuracy as the test itself. Of fundamental importance is the conversion of N to N60. Without this conversion the SPT results cannot be compared to other N-values from other tests.
Soil and rock exhibit different engineering behaviour, and different strengths. The SPT in granular soils can be related with a degree of confidence to a number of empirical equations. With cohesive soil and weak rock, the SPT is more of an indicator of consistency. Examples from Auckland demonstrate the complexity of the soil-rock interface and how the SPT can be interpreted in different situations. A breakdown of the steps to follow and questions to consider when terminating the SPT have been presented. These may prove useful to the young geotechnical professional embarking on ground investigations.
Finally, the SPT must be interpreted in the context of the local geology and not be relied on independently of other contributing factors.
The author would like to thank Dr Simon Nelis for all the feedback and support in writing this paper, your input has been invaluable. Evan Giles, your encouragement and support to undertake this work is very much appreciated. Thank you to the external reviewers for reviewing my work.
4402.6.5.1:. 1988. “Soil strength tests – Determination of the penetration resistance of a soil – Test 6.5.1 Standard penetration test (SPT).”
Barton, N, R Lien, and J Lunde. 1974. “Engineering classificationof rock masses for the design of tunnel support.” Rock mechanics 6 (4): 189-236.
Bieniawski, Z. T. 1989. Engineering rock mass classifications: a complete manual for engineers and geologists in mining, civil, and petroleum engineering. London: John Wiley & Sons.
Claton, C. R. I. 1995. “The Standard Penetration Test (SPT): methods and use.” CIRIA Report 143.
Deere, D., and D. Deere. 1988. “The rock quality designation (RQD) index in practice.” Symposium on Rock Classification Systems for Engineering Purposes. Cincinnati, Ohio, USA.
Deisman, N., M. Khajeh, and R. J. Chalaturnyk. 2013. “Using geological strength index (GSI) to model uncertainty in rock mass properties of coal for CBM/ECBM reservoir geomechanics.” International Journal of Coal Geology 112: 76-86.
Fooks, P. G., G. Pettifer, and T. Waltham. 2015. Geomodels in Engineering Geology : An Introduction. Whittles Publishing, Caithness.
Hodgson, A., and J. St George. 2008. “An Appraisal of Tunnelling in the Auckland Region for Infrastructure Development.” 13th Australian Tunnelling Conference. Melbourne. 361-366.
Hoek, E., and E. Brown. 1997. “Practical estimates of rock mass strength.” International Journal of Rock Mechanics and Mining Sciences 34 (8): 1165-1186.
Liao, S. S. C., and R. V. Whitman. 1986. “Overburden correction factors for SPT in sand.” Geotechnical Engineering 112 (3): 373-377.
Look, B. G. 2019. “Geotechnical Basics – Common Practices to Avoid.” Edited by Acosta-Martínez and Lehane. 13th Australia New Zealand Conference on Geomechanics. Sydney, Australia: Australian Geomechanics Society. 1247-1252.
Look, B. G. 2004. “Rock Strength at the Coring Interface.” Australian Geomechanics 39 (2): 105-110.
Matsumoto, T., L. T. Phan, A. Oshima, and S. Shimono. 2015. “Measurements of driving energy in SPT and various dynamic cone penetration tests.” Soils and Foundations 201-212.
Peck, R. B., W. E. Hanson, and T. H. Thornburn. 1953. Foundation Egineering. New York: John Wiley & Sons.
Pells, P. 2004. “Substance and mass properties for the design of engineering structures in the Hawkesbury Sandstone.” Australian Geomechanics Journal 39 (3): 1-21.
Rodgers, J. D. 2006. “Subsurface exploration using the standard penetration test and the cone penetration test.” Environmental and Engineering Geoscience XII (2): 161–179.
Skempton, A. W. 1986. “Standard penetration test procedures and the effects in sands of overburden pressure, relative density, partical size, ageing and overconsolidation.” Geotechnique 36 (3): 425-447.
Skempton, A. W., and A. W. Bishop. 1950. “The measurement of the shear strength of soils.” Geotechnique 2 (2): 90-108.
Winkler, G. E. 2003. “Geotechnical engineering of the Northland Allochthon.” Institution of Professional Engineers New Zealand. 91.