Soil Test Before House Construction: What to Ask For

The cracks show up in year three. A hairline over the door, then a longer one across the sit-out floor, then a door that stops closing in July and closes again in March. By then the plaster is painted and nobody can see what caused it. A soil test before house construction is the one day of work that would have told you, before a single rod was tied, what the ground under your plot was going to do.

Most people building in Hyderabad or Goa skip it. The plot next door built fine. The contractor says the ground is hard. So the footing gets copied from the neighbour and the risk gets buried under the raft.

This is what a real investigation contains, what the Indian Standards require, and how to read the report you are handed.

What a soil test before house construction actually is

Two very different things get called a soil test. One is agricultural — nitrogen, phosphorus, pH. The other is geotechnical, and it decides whether your walls stay straight. A geotechnical investigation puts a hole into the ground, records what changes at each depth, and pulls out samples a lab can load, wet, dry and shear. The output is a borehole log and a recommendation. The log is the honest part: it shows filled earth, then soil, then weathered rock, then hard rock, with the level at which each begins. If your report has a recommendation but no log, you have a conclusion without the evidence, and no way to check it.

When we plan the sequence of construction stages for a house, the investigation sits before design, not after it. The foundation drawing is downstream of what the borehole finds.

How deep, and how many boreholes — the numbers your report must meet

This is the single most useful pair of numbers a homeowner can hold, because it is where corners get cut quietly. Clause 2.3.2 of IS 1892 states that the depth of exploration should normally be one and a half times the width of the footing below foundation level, and its Table 1 repeats that rule for isolated spread footings and rafts, extends it to one and a half times the length where adjacent footings sit closer than twice their width, and applies one and a half times the width of the structure for pile and well foundations. The same clause adds a floor that most reports ignore: the depth to which seasonal variations affect the soil should be regarded as the minimum depth of exploration. On a plot with any clay in it, that seasonal line is the one that matters, and it is deeper than the trial pit somebody dug with a JCB.

Read that against a villa plot. An isolated footing 2 metres square needs 3 metres of exploration below its own founding level, not 3 metres from the road. A report that stops at 1.5 metres on a plot with cotton soil has not met the code it claims to follow.

The SPT N value, in plain words

IS 2131 defines the test exactly, and the exactness is the point — it is what makes one lab’s number comparable to another’s. Clause 2.3.1 of that standard specifies a 63.5 kg weight with a 75 cm free fall. Clause 3.3.3 sets the procedure: the sampler is first seated 15 cm with the hammer falling through that same 75 cm, then driven a further 30 cm or to 50 blows, and the total blows required for the second and third 15 cm of penetration is termed the penetration resistance, N. The first 15 cm is thrown away deliberately, because it is disturbed ground at the bottom of the hole and counting it would flatter the result. When a report shows a suspiciously smooth run of identical numbers down the log, something was estimated rather than counted.

Two practical checks. Ask what happened at refusal — the depth where the sampler stopped advancing is your rock line and it belongs in the log. And ask whether the water table was recorded, because a saturated sand at N of 10 behaves nothing like the same sand dry.

Safe bearing capacity — the number everyone quotes and few read correctly

IS 6403 separates them cleanly. Under Clause 2.1.3, net safe bearing capacity is the maximum intensity of loading the foundation will safely carry without risk of shear failure of the soil, irrespective of any settlement that may occur. Clause 2.1.5 defines the allowable value as the net intensity of loading the foundation will carry without settling more than the permissible amount for that structure, but never exceeding the first figure. Clause 6.1 of the same code arrives at the allowable value by dividing the net failure load by a suitable factor of safety, and points to IS 1904 for the permissible settlements. That cross-reference is where the real limits live. Table 1 of IS 1904, set out under its Clause 16.3.4, permits for isolated footings on sand and hard clay under an RCC structure a maximum settlement of 75 mm, and, in the same table, IS 1904 sets the differential settlement limit as an angular distortion of 1 in 666.

Read those two words in your report — safe and allowable — and see which the engineer used. A number quoted with no settlement check behind it is half an answer.

Wide, lightly loaded footings on clay settle more than the arithmetic suggests, which is one reason we prefer one accountable team running design and execution together — the structural engineer sees the log, not a summary of it.

Black cotton soil, and why it breaks light foundations

You can measure the tendency. IS 2720 Part 40 gives the free swell index test, simple enough to picture: Clause 3.1 takes two 10 g oven-dry specimens passing a 425-micron sieve, Clause 3.2 puts one in a 100 ml cylinder of kerosene and the other in distilled water to the 100 ml mark for not less than 24 hours, and Clause 4.1 reports the index as the difference in settled volumes divided by the kerosene volume, times 100. Kerosene is the control because it does not make clay swell. IS 2720 Part 41 measures the quantity a designer needs, swelling pressure — what the soil pushes with when it is stopped from expanding. For scale, IS 9451 records swelling pressures for expansive soils of 50 to 300 kN per square metre or more, with design tables running to 500 kN per square metre.

The Indian Standards answer this two ways. IS 1904 Clause 7.5 says the foundation should bear at a depth where seasonal change is not important, or use a flexible form of construction, or a rigid foundation designed to ride the movement. IS 9451 takes the other route and specifies a cohesive non-swelling cushion between structure and expansive soil, the material defined in its Clause 4.2.5 as having a liquid limit above 30 but below 50 and a plasticity index above 15 but below 30. Two legitimate strategies, neither of them guesswork.

One more clause worth knowing, because it costs nothing and gets ignored constantly: IS 1904 Clause 5.3.3 advises that trees over 8 metres in height should not be planted within 8 metres of building foundations, precisely because root drying shrinks clay. A neem in the wrong corner of a compound wall can crack a boundary in five years.

Ground in Hyderabad, and why the plot next door proves nothing

The published geography of Hyderabad records the city’s rock as grey and pink granites around 2.5 billion years old, and the soil as mainly red sandy with areas of that expansive clay, with elevation swinging from about 456 m in the Old City to 672 m at Banjara Hills. Old, hard granite under a shallow and inconsistent soil blanket is why two plots on the same street can need different foundations. In Jubilee Hills and parts of Banjara Hills you are often working close to rock, and the question becomes rock excavation and levels rather than bearing capacity. Out along Tellapur, Mokila and Shankarpally, and in stretches around Kokapet and Narsingi, the cover is deeper and more mixed, and cotton soil pockets and old filled tank beds turn up more often. Gachibowli sits between the two, plot by plot.

Then there is the climate loading the clay. Hyderabad’s mean daily maximum in May reaches 38.8 degrees Celsius, and the city receives about 859.6 mm of rain a year with more than 75 percent of it falling between June and September. That is the shrink-swell engine written as a weather statistic: a long hard dry followed by four months of concentrated wetting, every year, for the life of the house. It is the same seasonal cycle that makes waterproofing in Hyderabad homes a sequencing problem rather than a product problem.

Goa argues for the same test differently: laterite, high coastal water tables, and slope plots where the fill from cutting the platform becomes the ground your compound wall stands on.

What it costs, and what actually drives the number

We are not going to quote you a price, because rates move by city, by season and by how deep the rig has to go, and any figure printed here would be out of date and wrong for your plot. What you can do is understand the shape of the bill. Boring is charged by the metre, so depth drives it. Sampling and each named laboratory test are charged separately, so the test list drives it. Number of holes multiplies both. Get two or three quotations from investigation agencies directly rather than through the contractor, and compare them on those three lines.

The useful way to think about it is not what it costs but what it decides. Knowing early that you need a raft rather than isolated footings moves a number far larger than the test fee, in either direction. Ask for the quotation itemised by metres of boring, number of holes and named tests. A lump sum cannot be checked against IS 1892.

How to read your soil report in ten minutes

Open the log and confirm how many holes were sunk and where they sat — a plan of hole positions should be in the report. Read each termination depth against the footing width your architect is assuming, using the depth rule in IS 1892. Find the water table entry; if it is blank, ask why. Scan the N values for the depth where they jump, because that jump is usually your founding level. Then check the recommendation names three things: foundation type, founding depth below existing ground level, and bearing pressure with its basis. If it gives a bearing pressure but never says whether that is a safe or an allowable value, send it back with that one question. And check the date, because a plot reads differently in September than in April.

What we do with the report, and when we are not the right fit

On our projects the investigation happens before the structural drawing, and the log goes to the structural engineer as a document, not a WhatsApp summary. Steel and founding levels are photographed before the pour, because once concrete is in nobody can check either. You can walk the finished work at Kingston Park and Ridhira Zen. Golecha Ghar is on site now, still under construction, and more useful to see if you want a foundation stage rather than a finished floor.

We are not the right fit for everyone. If you want the lowest quote per square foot, another builder will beat us and you should take it. If you have already poured a foundation and want an opinion on cracks, you need a structural consultant, not a design-and-build firm. And if your plot is a straightforward infill with rock close to the surface and neighbours on both sides who built without incident, an honest engineer may tell you one centre borehole is enough — which is what IS 1892 Clause 2.3.1 already allows for smaller buildings. We would rather say that than sell you five holes.

If you are choosing a partner to run the whole build, our construction work in Hyderabad and our approach to house construction across the city both start at the same place: know the ground first. The materials that go above it only perform if what is below stays still. When you are ready, bring us your plot and your soil report and we will read it with you.