The Dark Art of Geotechnical Investigations

This is indeed a dark art. You need to keep your wits about you, train your eye, question everything, be cognisant and understand the local geology, have some idea what to expect in terms of soil types, and above all be thorough in your approach.

Desk study

Someone may pay you to do this, and if so then you will have a bit of time to scratch out what is available and go through the historical documents pertaining to the site. A desk top study can mean a number of things to different people. Working in the UK we had a duty of care to lay our hands on every map, every drawing and every document that we could to ascertain the history of the site. This was time consuming and expensive but would often reveal a wealth of information on the development and geological history of the site. However the circumstances are slightly different as there can be a 300 year development history attached to some of those sites, and one needs to get an understanding of what has gone on over the centuries to be able to assess contamination levels, the location of fills, water courses and so on. That said here in South Africa we are often dealing with green field sites (i.e., no developmental history) or sites where there has been limited human impact and in this light the desk top study might not be as critical.

Whether you are getting paid or not for a desk top study, it is always worthwhile trying to get an insight into what underlies the site before actually carrying out any fieldwork. The first stop is the geological map – these are one of the greatest repositories of geological information available and we must take our hats off to the geologists of the Geological Survey who have braved heat, dust, rough camps, political instability, buffalo bean, mosquitoes, flooded rivers, wild drillers and probably every affliction that you can list to produce those wonderfully coloured and amazingly informative sheets of paper. But I digress. Get the geological map out, find your site, find out what the geology is and you are part of the way to building a good geological model of the area. Google Earth is a wonderful tool, so have a look at the satellite imagery if you can. Then of course there are orthophotos and topographical maps available from the Surveyor General should the need arise or you need contour data. Garmaps also will sell your contoured mapping data if you need integration with your GPS handset.


There are different levels of fieldwork, ranging from a cheap and nasty walkover survey to a full blown test pitting and borehole drilling exercise that can last for months in some instances.

Walk Over Survey

Geologists will be familiar with the walkover survey – it is in essence a mapping exercise in which the following issues need to addressed:

· Geological boundaries

· Thicknesses of soil and soil types

· Location of hard rock outcrop and boulder fields

· Presence of surface water

· Presence of wetland areas

· Areas showing signs of previous instability, mass wastage or potential rock falls

· A very South African concern – the location of graves

· Location of houses, buildings or other infrastructure

· The nature of neighbouring properties and the state of the infrastructure

· Slope gradients

· Vegetation

· Access

· Presence of services, servitudes and overhead powerlines

Test Pit Investigations

The next step up, and perhaps one in which you will be consistently involved, is the test pitting investigation. You may not have the luxury of walking over the site beforehand so you will need to address all the issues listed under the Walk-over survey above, as well as dig, log and sample the test pits. Please have a look at the advice on test pitting, as much has been written on this already. However a few points need to be discussed here. Once on site with the TLB, get the operator digging as soon as possible. The location of the first pit is not that critical, so once he is set up you can catch your breath and get organised. The number of test pits that you need to dig will depend on the proposed budget. If there is a day’s fieldwork set aside for the job, then aim to dig 12 to 15 pits at least. Judicious use of a scale rule will quickly allow you to pencil in provisional pit positions across the site based on the site dimensions divided by the number of pits required. It is perhaps better not to number them at this early stage as some of the provisional positions might have to be bombed due to access issues, lack of time or other unforeseen circumstances.

Plan on at least 2 pits per hour, and if you put your mind to it 3 is very attainable. If you really get into your stride, and if the geology is simple, access is easy, and the level of pit logging is not critical, then it may be possible to increase this number even more. But it is not a case of quantity over quality – far better a properly logged pit with good and thoughtful sampling than a slap-dash approach with all its attendant errors. For example when test pitting for a development it is important to spend time logging the pits. However if the investigation is for a pipeline where the bearing capacity is not so critical but depth to bedrock, pit stability and the water table are important, then a little less time can be devoted to the logging and the pits can also be logged from the surface.

It is perhaps best to postpone the sampling until say the third test pit of the morning, for it is only at this stage that a feel for the geology is beginning to develop. Bear in mind that for roads the action is all going to take place in the upper 0.5 m or so, so this is where you need to sample. However cognisance also needs to be taken of the fact that there will be cuts to accommodate the road and at times you will need to sample the deeper subgrade material. Take a minimum of two large sample bags of material, with say 10 to 12 shovels full in each bag. In terms of housing developments, heave is going to be one of your issues so take an indicator sample between 1.0 and 1.5 m as it is in this zone that heave will take place. Send the sample to the laboratory for a full indicator test including hydrometer. Pay attention to slickensiding, high clay content and pinholing. A fissured and slickensided soil is a good indicator of the presence of highly expansive clays. Bring this feature out in the logs – I sketch a skull and crossed bones to remind of my discovery when I have a pile of undifferentiated soil logs on my desk back in the office.

Try not to renumber log sheets back in the office. The field sheet may correspond to a label in a sample bag in a laboratory or a list of photographs of the test holes or a GPS position. If you really have to renumber, then make a list of all the old pit numbers and in a separate column list the corresponding new number. It may be worthwhile adding samples and/or photo numbers etc to this data base. Put this in the file with all the original data. When trouble then comes calling you always have the original data base to fall back on. It all sounds simple now, but it is exceptionally easy for things to go wrong when the house keeping isn’t in order.

GPS technology is fantastic and we often wonder how we managed without it. But it can also be a curse for it makes us lazy. From bitter experience, put the test position onto the site plan – whether it is a borehole, test pit, DCP position or percolation test. It is all too easy to forget to mark the waypoint, only to find a log for a test pit in the file, but with no waypoint position. Now if it is marked on the site plan then your problems are pretty much solved.

Perhaps some more needs to be said with regard to sampling. I never believe that there is enough budget for this. However you will no doubt be working with a budget produced by others so you will need to try and get representative samples of the various geological materials. Rather sample too much than too little – it is far cheaper to throw samples out than to have to go back to site and resample. Try and get a representative bulk sample of all the various soil and rock types which occur on the site. Colluvium, residual soils and highly weathered bedrock are all worth bagging. Make sure the samples are labelled and it is good practice to write on the outside of the bag too, as this assists the laboratory in finding your samples. Otherwise they have to scratch around in the soil itself for an often soggy, barely legible label. Cable ties are a good idea – it stops sample falling about in the back of the vehicle and maintains the moisture content until such time as the testing can be done. In this regard then the bulk sample can double up as an indicator sample. Indicator samples comprise generally around 1 kg or so of soil sealed up on a plastic bag. Self sealing plastic bags have made our lives a lot easier in this regard. Indicator samples are great if you are uncertain of the geology, feel a need to bring something home for the boss to see, if the material is highly slickensided or very clayey and you suspect that it is expansive and a test is required. They are cheap, quick and easy to take and give everyone peace of mind. Take those indicator samples – rather too many than too few.

Be careful with your log sheets. Some write their logs into a field notebook which they photocopy and put into the file and pass on for typing. I am not a fan of this method as it increases the chance for errors. It is infinitely preferable to make a pro forma sheet which you take to the field. The MCCSSO system can be listed, along with boxes in which the test pit number, descriptions, samples, founding levels and the occurrence of water are written. This forces you to tick the boxes so to speak. If there is no water, then say so, if there are no samples, then say so. It is removes uncertainty and inspires confidence in the logger’s abilities and thoroughness. For a copy of a pro forma logging sheet click here.

GPS technology aside, it is in most instances important that you keep a rough track of where you are on site. Setting out test pits at 200 m intervals can be a bind if you are alone, you don’t have a tape measure and the GPS is not helping. Get some idea of the length of your pace, and then pace things out. It can be remarkably accurate provided that you try to make it consistent. By knowing that you have walked approximately 300 m from a known position along the contour for example can be of great assistance in keeping track of where you are on site. So get used to pacing, and get used to counting.


The bane of all our lives. If you dig test pits, get used to the idea that you are going to rip up power lines, sewer lines and water lines. It is always great entertainment when it happens, and generally speaking it is always late in the day when everyone is winding down and thinking of going home. So reduce the risks. Firstly try to get a drawing of where the services are. This is generally not possible, so take care of where you dig to reduce the chances of hitting services. Here are some tips.

  • Don’t dig underneath power lines, as they often run pipes along the same servitude but below ground.
  • Don’t dig on a line between manhole covers or other obvious markers
  • Look out for pipeline markers – often white concrete beacons set in the ground at 100 to 200 m spacings. Don’t dig on these lines
  • Water line off-takes run at 90 degrees from the main bulk water line to a water meter or tap in the front of a property. Needless to say, stay away.
  • Sewer lines are often laid along the valley inverts as they are gravity driven, so avoid digging near the valley floor in built up areas
  • Pipes are generally bedded in a sand pipe bedding, so pay attention to any subtle shift in colour within the soils or the appearance of sand in the excavator bucket.
  • Hard digging when not expected may indicate a concrete water pipe, so get the operator to stop, get in the hole with a spade and find out what is going on down there.
  • And finally, try to go to site armed with the local authority’s phone numbers. This is very useful when water starts to spray everywhere and you need to let someone know that you have just deprived the local community of their water supply for the night.

Good luck!

, , , , , , , , , , ,