Standards, tips and the tricks nobody writes down.
The codes that govern geotechnical work worldwide, translated into plain language, plus the field and laboratory habits that separate a defensible report from an expensive guess.
The four code families you will meet.
Most specifications in Pakistan reference one of these, and occasionally three at once. Knowing which is calling the shots avoids a lot of argument on site.
ASTM International
The American standards, and the default reference for laboratory and field testing across Pakistan. Prescriptive and procedural: they tell you exactly how to run the test, not what to do with the answer.
- D1586 Standard penetration test and split barrel sampling
- D5778 Electronic friction cone penetration testing
- D2487 Unified Soil Classification System
- D4318 Liquid limit, plastic limit and plasticity index
- D1194 and D1196 Plate load testing
- D2166 and D2850 Unconfined compression and triaxial
Eurocode 7 and BS EN
The European approach, and increasingly the reference for international clients. Where ASTM is procedural, Eurocode 7 is a design philosophy built on partial factors, characteristic values and ground investigation categories.
- BS EN 1997-1 Geotechnical design, general rules
- BS EN 1997-2 Ground investigation and testing
- BS EN ISO 22475 Sampling methods and groundwater measurement
- BS 5930 The UK code of practice for site investigations
- Key idea: pick a characteristic value, a cautious estimate of the mean, not the worst single result
AASHTO
Transport focused. If your project is a road, a highway embankment or a bridge approach, the pavement and subgrade testing will usually be specified to AASHTO rather than ASTM.
- T 193 California Bearing Ratio
- T 180 Modified Proctor compaction
- M 145 The AASHTO soil classification system
- T 245 Marshall stability and flow for asphalt mixes
Pakistani practice
The Building Code of Pakistan governs seismic provisions and, through them, how much your site investigation has to prove. Local authority requirements sit on top of it and vary by city.
- BCP SP-2007 Seismic provisions, including site classification
- Seismic zones 1 to 4, with Karachi in zone 2B and much of the north in 3 or 4
- Site class from Vs30, which is exactly what a downhole seismic test measures
- Authority approvals often ask for a signed report from a recognised laboratory
Twenty things that go wrong, and how to avoid them.
Collected over a decade of fieldwork. None of these are in the standards, which is precisely why they cost people money.
Check the hammer energy
An SPT N value is meaningless without knowing the energy ratio of the hammer that produced it. A donut hammer and an automatic trip hammer on the same soil can differ by 40 percent. Correct to N60 before you use any correlation.
Record the water table twice
Once on striking it, once after the borehole has stood for 24 hours. In clay the first reading is nearly always wrong, and the stabilised level is the one that matters for design.
Do not wash out the sample
Excessive water pressure during rotary drilling flushes the fines away and turns a silty clay into what looks like clean sand. Control the flush rate and log what comes up honestly.
Seal undisturbed samples immediately
A Shelby tube left open in the sun for an hour is no longer undisturbed. Wax the ends on the rig, label both ends, and keep them upright and out of the heat.
Locate boreholes where the load is
Holes placed for rig access rather than under the actual column grid produce a tidy report about the wrong piece of ground. Move the rig, not the borehole position.
Scan before you core
A rebar scan takes ten minutes. Cutting a main bar in a slab because you skipped it costs considerably more, and it is not a repair anyone enjoys explaining.
Backfill every hole
An open borehole is a hazard and, where it links two aquifers, a contamination route. Backfill and grout as you withdraw, not next week.
Photograph the core, in order
Core boxes get moved, dropped and reordered. A dated photograph with a scale in frame settles arguments months later.
Ask what was here before
Ten minutes with a neighbour or an old satellite image will tell you about the pond, the quarry or the dump that the site plan does not mention.
Go deeper than you think
Stopping at 6 metres because that is what the quote said, when N values are still falling, saves nothing. Agree a provisional depth allowance up front.
Respect the soaking time
Atterberg limits on a soil that was not properly wetted and matured give a low plasticity index, which then hides an expansive clay. Twenty four hours means twenty four hours.
Match the triaxial to reality
Unconsolidated undrained for short term stability, consolidated drained for long term. Running the convenient test rather than the correct one produces a number that does not describe your problem.
Dry at the right temperature
Organic and gypsum rich soils lose structural water above 60 degrees. Oven drying them at 110 overstates the moisture content and understates everything that follows.
Cure cubes properly
Concrete cubes left on site in the sun instead of in a curing tank will fail, and the pour will be blamed. Curing conditions are part of the test, not an afterthought.
Test sulphates before specifying cement
Sulphate resisting cement costs more. Ordinary cement in aggressive ground costs the structure. One chemical test settles it.
Settlement usually governs, not bearing
On clay especially, the allowable pressure is set by tolerable settlement long before shear failure becomes relevant. Design for movement, then check capacity.
Use a cautious mean, not the worst value
Designing to the single lowest N value in the whole site is not conservatism, it is waste. Eurocode 7 asks for a characteristic value, a cautious estimate of the mean of the governing zone.
Check differential, not just total
A building that settles 25 mm uniformly is usually fine. One that settles 25 mm at one corner and 5 mm at another is cracking. Angular distortion is the criterion.
Found below the active zone
In expansive clay the seasonal moisture change reaches a certain depth. Found above it and the building rides the seasons. Below it, the problem largely disappears.
Read the report, all of it
The recommendations section assumes the limitations section was read. The caveats about fill depth, groundwater and the tested area are not boilerplate, they are the conditions on the numbers.
Rules of thumb worth memorising.
Useful for a first pass and a sanity check. Never a substitute for testing the ground you are actually building on.
Borehole depth
1.5 to 2 B
Below founding level, where B is the foundation width, or until the stress increase drops under 10 percent of the overburden.
Borehole spacing
300 to 500 m²
One hole per that area for a building, minimum three, tightened wherever the ground looks variable.
SPT to bearing
q ≈ 10 N
Very rough allowable bearing in kPa for sand from N60. A starting sanity check only, never a design value.
Tolerable settlement
25 mm / 1:500
Typical total settlement limit for framed buildings, with angular distortion kept under about 1 in 500.
Our tools do this arithmetic for you.
Bearing capacity, borehole layout, USCS classification, concrete mix proportions and unit conversion, all free and all in the browser.
Open the toolkit