THE IMPORTANCE OF ENGINEERING GEOLOGY INTERPRETATION IN CREATING GROUND AND GEOTECHNICAL MODELS – A CASE STUDY OF A SITE ON THE SWAN RIVER, WA

1???? ?INTRODUCTION

Geotechnical investigations are fundamental for almost every engineering project in order to categorise the subsurface conditions on which structures are built on, in or with. This concept is widely accepted and, for the most part, investigations are completed. However, engineering geology is not always considered an important part of the investigation and is not included.

The International Association of Engineering Geology and the Environment (IAEG) defines engineering geology as “the science devoted to the investigation, study and solution of engineering and environmental problems which may arise as a result of the interaction between geology and the works or activities of man, as well as of the prediction of and development of measures for the prevention or remediation of geological hazards.” Geological processes form all-natural material used in geotechnical engineering. Whether they are present beneath structures; whether structures are placed within them, or whether these materials are used as part of the structure as earthworks or fill materials, understanding the geological processes that form the materials is the only way to gain a complete picture of their composition and potential response to engineering interactions.

This paper presents a case study of a geotechnical investigation on the Swan River that highlights the importance of ensuring engineering geological input is used as part of the geotechnical assessment.

2???? ?PROJECT DESCRIPTION

Whilst details of the site and specific development remain confidential, the basic information for the site is described herein and provides sufficient information for the purposes of this paper. The site is located within the Swan River, near its banks, less than 5km from the Indian Ocean. The published geological conditions for the site are typical of the conditions formed by the cutting of the Swan River in this region. A high-level snippet of published geological conditions of the area is provided in Figure 1.?

This region of the Swan River is categorised by its incision through Tamala Limestone (S7), with some deposits of Alluvial Sand (S14) on the bank and deposits of Swan Estuary Deposits (silt - M7 and sand - S22) within the Swan River itself.

The project itself requires the design and construction of an overwater structure. The foundations for the structure are likely to be driven piles.

3???? GEOTECHNICAL INVESTIGATION

A geotechnical investigation for the project was carried out in June and July. The scope of the fieldwork comprised the following aspects, all completed using submerged plant and machinery. Some example images of these techniques are shown in Figures 2 to 5.

·???? Five machine boreholes (BH01 to BH05) were advanced using HQ3 diamond coring drilling techniques to depths of up to 19.8m to determine the subsurface ground conditions and obtain samples for visual logging and sampling for laboratory testing.

·???? Two Standard Penetration Tests (SPTs) were undertaken within BH03 but were subsequently discontinued due to time constraints.

·???? Seven Cone Penetrometer Test (CPT) probes, including pore pressure (CPTu01 to CPTu07) were advanced to depths of up to 17.5m to define the subsurface ground conditions through the underlying zone of influence of pile foundations and to provide parameters for pile design.

Standard Penetration Tests (SPTs) were carried out in accordance with AS1289.6.3.1-2004. It should be noted that all SPT tests for the marine boreholes were carried out with the SPT hammer underwater. At these locations energy losses associated with the hammer falling through a water column would occur and account needs to be taken when interpreting the SPT N values.

The energy losses purely due to the buoyant weight of the steel hammer amount to approximately 13%. A further 20% over and above that which occurs on land-based testing is estimated due to energy losses from hydrodynamic drag. Assuming a 60% efficient SPT hammer if operated in air the same hammer would be only approximately 40% energy efficient when operated in seawater. A correction factor Co of approximately 0.6 to 0.7 must therefore be applied to correct SPTs conducted with a submersed SPT hammer to an equivalent N60 value. In addition, normal effective stress correction must be applied when assessing relative density particularly at shallow below seabed depths.

In addition to the field investigation described above, rock strength laboratory testing was undertaken on the samples of limestone obtained during the borehole drilling. In total 22 samples were tested, and the results indicated that the strength was generally high to very high.

It is noted that no uniaxial compressive strength (UCS) tests were able to be completed. As such, no direct correlation between the point load index test results and the rock UCS was possible. The cementation and strength of Tamala Limestone can vary significantly as does the correlation between UCS and point load index. AS1726:2017 indicates the requirement to use UCS testing for strength profiling but provides a normative reference to strength categories based on a correlation factor of? 20. In the absence of any site-specific correlation, the strength determination provided in AS1726:2017 was used. It is also worth noting that sampling bias is prevalent within material like Tamala Limestone. Often the drilling process causes core loss and recovery of weakly cemented and cobbly gravels which are not conducive to sampling and testing, even for point load index tests. Consideration of the effect of only sampling higher quality (and higher strength) material must be made.

?4???? GROUND MODELS

4.1????? INVESTIGATION DATA

Both the CPT and borehole information were used to categorise the subsurface conditions. The use of the CPT was preferred due to its continual measurements and associated accuracy of interpretation.? The boreholes were used to gain further information and to provide samples for visual logging and, importantly, logging and sampling of the underlying rock. Figure 6 depicts one of the CPT probe results with a simple interpretation of the basic data obtained.

Using this simple approach four distinct units can be identified, as follows:

·???? Unit 1 with low tip resistance and higher and variable friction ratios

·???? Unit 2 with a marked increase in tip resistance and a marked decrease in friction ratio to a consistent value <~0.5

·???? Unit 3 with another marked increase in friction ratio, but variable in nature with a variable sleeve friction and tip resistance.

·???? Unit 4 below where the CPT probe refuses.

Using Robertsons (2009) a Soil Behaviour Type (SBT) Index can be derived from the CPT results. This can be completed to further define the soil types and assist in interpretation. Figure 7 depicts the SBT Index Plot for the same CPT

The SBT Index plot supports the definition and interpretation of four separate units. Using? both the CPT results and the boreholes drilled, the ground model components encountered can be summarised as shown in Table 1.

It is noted that the interpretation of Unit 3 based on CPT results alone would indicate a more fine-grained soil. However, the boreholes recovered a gravelly cobbly material which is more typical of a weathered/leached limestone. The significant variability in the tip resistance and sleeve friction, which leads to a variable friction ratio is also typical of CPT probing through variably cemented soil and weak rock characterised by some parts of the Tamala Limestone. The sharp rise in tip resistance followed by significant drop that occurs when the cement bond is broken can make the interpretation of the soil difficult from CPT’s alone.

The distribution of these units was variable across the site in terms of both horizontal and vertical extent, depths/elevations and thicknesses. This is depicted in Figure 8.

Using the information provided above from the geotechnical investigation, a 2D cross section and representative ground model can be developed. Here, two separate models have been presented using the same 2D cross sections from shore to overwater.

4.2?? ???INTERPRETATION 1

The first ground model that can be derived from the investigation results is depicted in Figure 9.

This ground model has simply been created by interpolating between the investigation points to derive surfaces between each of the units. This is perhaps a suitable ground model as it seems to suggest a deepening channel from shore into the river and there are some adequate correlations between the elevation of the units from the borehole and CPTs to fit rather well. It is noted that not all investigation locations are shown on Figure 9 for ease of understanding.

4.3?? ???INTERPRETATION 2

However, the importance of engineering geological knowledge and experience should not be understated. Simply looking outside the confines of the site, some other important factors can be taken into consideration that might influence and change the ground model presented.

Parts of Fremantle are known to be underlain by a palaeochannel that was cut through the surface geological conditions and into the underlying Mesozoic formation (Osborne Formation).

The palaeochannel is known to be present within the harbour area of Fremantle but has not been confirmed as being present in other locations. However, it is possible that it would roughly follow the meandering of the existing Swan River. In addition, the cutting action of the river created a steep sided channel with high angle cuts within the Tamala Limestone. It is possible that the steep limestone cliffs immediately adjacent to this site were formed by the palaeochannel development. Furthermore, there is evidence of historical palaeochannel slumps within the steep sides occurring within the channel in the harbour area. The slump is a result of the failure of the steeply cut limestone. A look outside the site bounds finds the presence of limestone cliffs immediately on the shore. It is possible that these steep limestone cliffs were formed by the palaeochannel.

These couple of interpretations together with the knowledge that the limestone is typically deeper than other areas surrounding the site, it is possible to interpret a different ground model, as shown in Figure 11.

5???? DISCUSSION AND CONCLUSIONS

The second ground model that has been presented in this paper could present some significant geohazards and associated risks to the proposed development. It is possible that the site is underlain by a slump of the steep sided palaeochannel. This may be the reason for the variable and lower than typical limestone elevation. The slumped limestone may have resulted in a variable strength and thickness unit. The risks associated with this scenario include the potential for punch through of a thin limestone/low strength limestone layer that are subject to high axial loading (from pile foundations) and a variable depth/elevation of the limestone leading to construction risks.

Whilst this specific ground model interpretation has not been proven during the course of this investigation, clearly, the addition of engineering geological knowledge and experience has provided valuable insight into this assessment. This case study reinforces the importance of using engineering geology inputs to provide geological content, ideas, hypotheses and potential geological situations, processes, environments and hazards that may not have been encountered at discreet investigation locations. In this instance, the geological process that caused a palaeochannel and the knowledge that one exists in similar conditions nearby. In this case the second interpretation was used for the pile design and construction methodology. Unfortunately, specific details of the construction outcomes are unknown by the author.

This case study also reinforces the importance of using all aspects of investigation and all sources of data and information. In this case, the CPT SBT index interpretation seemed to conflict with the borehole information and the knowledge of the engineering geological characteristics of the Tamala Limestone assisted in providing an appropriate interpretation. Similarly, the sampling bias and subsequent interpretation of limestone variability, weathering/leaching and strength was founded on geological knowledge.

Geology is paramount in the successful geotechnical application to any project and understanding the key geological concepts, principles and processes is critical. AS1726:2017 is a good starting point as it requires input from an engineering geologist to be used to form a ground model that considers geological, hydrogeological, geomorphological and topographical elements. The importance of this should not be underestimated.

6???? REFERENCES

Standards Australia (2017). Geotechnical site investigations. (AS1726) [Standard]

Standards Australia (2004). Methods of testing soils for engineering purposes, Method 6.3.1: Soil strength and consolidation tests — Determination of the penetration resistance of a soil — Standard penetration test (SPT). (AS1289.6.3.1) [Standard]

Geological Survey of Western Australia (1986). Environmental Geology Series, Fremantle, 1:50,000 scale.

Robertson, P.K., 2009. CPT interpretation – a unified approach. Canadian Geotechnical Journal, 46, 1-19

Tutton, M. (2003). Engineering geology of Fremantle harbour (an historical perspective). Australian Geomechanics Vol 38 No 4 December 2003 – The Engineering Geology of Perth Part 2, 91-102.