Site investigation
Site investigation finds out what lies beneath a site before anything is designed. Most foundation failures, claims and cost overruns trace back to ground that was different from what was assumed. A good investigation is cheap insurance: it typically costs a small fraction of a percent of the project but controls the design of foundations, retaining structures, excavation and drainage.
It proceeds in stages. A desk study reviews geological maps, old borelogs, satellite images, flood and groundwater records. A walkover survey looks for springs, cracks in nearby buildings, fill, rock outcrops and signs of landslides. The main investigation then uses boreholes with standard penetration tests (SPT) and undisturbed tube samples in clays; static cone penetration tests (CPT) for continuous profiles; trial pits for shallow strata; and rock coring with RQD where rock is met. Groundwater levels are recorded and sometimes monitored with piezometers. Geophysical methods such as seismic refraction and electrical resistivity fill gaps between boreholes.
Depth and spacing depend on the structure: boreholes should go below the depth where the added stress becomes small, typically around one and a half to two times the footing or raft width, and into rock for piles. Samples go to laboratory tests for index properties, strength, consolidation, chemistry and permeability. The output is a geotechnical report with borelogs, a soil profile, design parameters and recommendations.
Log groundwater level when first struck and again after 24 hours; the two often differ.
SPT N corrected for overburden and energy (N60, N1)Area ratio of a sampler: Ar = (Do² − Di²) / Di² × 100 (keep ≤ 10–20 percent)RQD = sum of core pieces ≥ 100 mm / run length × 100Investigation depth ≈ 1.5 to 2 × foundation width below founding level
Problem. A 1.5 m core run recovers pieces of 180, 60, 120, 90, 250, 40 and 300 mm. Find core recovery and RQD.
- Total recovered = 180 + 60 + 120 + 90 + 250 + 40 + 300 = 1 040 mm; recovery = 1 040 / 1 500 = 69 percent.
- Pieces ≥ 100 mm: 180 + 120 + 250 + 300 = 850 mm.
- RQD = 850 / 1 500 × 100 = 57 percent.
- An RQD of 50 to 75 is usually described as fair rock quality.
Answer. Recovery 69 percent; RQD 57 percent (fair)
Quick quiz
1Name the stages of a site investigation.
Desk study, walkover survey, ground investigation, testing and reporting.
2What is RQD?
The percentage of a core run made up of pieces at least 100 mm long.
3Why limit a sampler's area ratio?
To reduce disturbance of the sample.
- IS 2720 (series) (India): Methods of test for soils: water content, limits, compaction, shear and others, each in its own part.
- BS 5930 (UK): Code of practice for ground investigations.
Clauses and documents to read
- IS 1892:2021 (subsurface investigation for foundations)
- IS 2131:1981 (standard penetration test)
- IS 4968 (Part 3):1976 (static cone penetration)
- IS 2132:1986 (thin-walled tube sampling)
Cited by number and clause only. We do not reproduce clause text, tables or figures (BIS Act 2016, s.10(5) and s.11). Check the current edition and amendments before use.
References and further reading
Standard texts, listed in no order of merit. We do not copy text, figures, tables or solved problems from any book; our explanations and examples are original.
- B. M. Das, Principles of Foundation Engineering, Cengage, latest edition.
- J. E. Bowles, Foundation Analysis and Design, McGraw Hill, 5th edition.
- M. J. Tomlinson, Pile Design and Construction Practice, CRC Press, latest edition.
Free and official sources
Found an error? Write to indiacongroupofengineers [at] gmail [dot] com. For corrections and feedback. We never ask for payment or personal details.