Raft on Soft Clay: Bearing Capacity Versus Settlement and the Ground Improvement Decision
A six-storey RC framed residential block, 30 m × 20 m in plan, is to be founded on a raft at 2.0 m depth. The profile is 3 m of recently placed sandy fill (γ = 18 kN/m³) over 8 m of soft-to-medium clay (SPT N = 4–6, w = 42%, e = 1.15, Cc = 0.32, cv = 2.0 m²/yr, cu ≈ 28 kPa) resting on dense sand with N > 50. The water table is at 1.5 m. The net foundation pressure from six floors plus raft is 80 kPa. One-dimensional settlement computed at the raft centre is 15 mm in the fill plus 80 mm in the clay, i.e. 95 mm, against a permissible limit of 75 mm. The client wants occupancy in 24 months and asks whether ground improvement is really needed, or whether the raft can simply be made thicker. Bearing capacity calculations give a net ultimate capacity of about 145 kPa.
1. Check bearing capacity and settlement against the usual criteria and state which one governs. What happens if the raft is simply thickened?
Read the worked answer
Bearing capacity: FS = qnu/qnet = 145/80 = 1.81. For a permanent structure of this class the acceptable factor of safety on net ultimate bearing capacity is 2.5–3.0, so the capacity is marginal and would fail a code check, but the shortfall is small and cu will rise as the clay consolidates.
Settlement: 95 mm against 75 mm permissible - a 27% exceedance, and this is the mechanism that actually damages the building (services, drainage falls, partition cracking, differential movement between the raft and the external works).
Which governs: settlement. The raft is stiff, so differential settlement across the 30 m is small; the problem is the absolute 95 mm, which accrues mostly in the clay and therefore develops slowly over years - the building would keep settling after occupation.
Thickening is counterproductive. Every 0.5 m of extra raft adds 0.5 × 25 = 12.5 kPa, i.e. 12.5/80 ≈ 16% more contact pressure. Settlement scales with the increase in net stress, so 95 mm would become roughly 110 mm, and bearing capacity FS would fall to about 1.6. The raft should get lighter, not heavier - e.g. a cellular/voided raft or removal of the 3 m fill and replacement with lighter granular fill.
2. Quantify the time needed for 90% consolidation untreated, and with prefabricated vertical drains at 1.0 m triangular spacing.
Read the worked answer
Untreated (vertical drainage only): the clay is 8 m thick, drained by the sand below and the granular fill above, so it is doubly drained and Hdr = 4 m. t90 = T90·Hdr²/cv = 0.848 × 4² / 2.0 = 6.8 years.
That is incompatible with a 24-month programme - and most of the 80 mm would still be coming on after handover.
With PVDs at 1.0 m triangular spacing: equivalent diameter de = 1.05 × spacing = 1.05 m, radial drainage path = de/2 ≈ 0.525 m. Ignoring smear and well resistance: t90(radial) = 0.848 × 0.525² / 2.0 = 0.117 yr ≈ 1.4 months.
Add about 6–8 weeks for drain installation, 2.5 m of surcharge fill placement and its own consolidation, giving roughly 5–6 months of treatment before raft construction - comfortably inside the 24-month programme.
Caveats to state: real performance is degraded by smear and drain spacing tolerance (allow 30–50% extra time), the residual settlement at 90% is still ~10 mm, and the surcharge must be at least 1.2 times the design pressure with settlement plates and piezometers to confirm 90% before de-surcharging. The 3 m uncontrolled fill must also be assessed - if it is loose, it should be removed and recompacted before drains are installed.
3. Recommend a foundation solution, comparing PVD + surcharge, stone columns and piles, and justify the choice.
Read the worked answer
PVD + surcharge: cheapest per square metre, treats the whole clay mass, raises cu so the bearing capacity FS improves with time, and fits the programme. It does not, however, remove the residual settlement risk and needs a surcharge load and time; it also does nothing about the loose fill. Recommended as the base option with monitoring.
Stone columns: attractive because they improve capacity and accelerate consolidation in one operation, but at cu ≈ 28 kPa the lateral confinement is only just adequate; bulging and a low replacement ratio limit the achievable settlement reduction to perhaps 30–50%, so 95 mm would fall to about 50–65 mm - within the limit but with less certainty than preloading.
Piles to the dense sand: bypass the clay entirely, hold settlement to under 25 mm and carry the full six storeys with no waiting period. They solve both capacity and settlement definitively, at perhaps three to four times the cost of PVD + surcharge, and are the correct answer if the building is settlement-sensitive or the drains' performance cannot be verified.
Recommendation: PVD at 1.0 m triangular spacing + 2.5 m surcharge for 5–6 months with settlement monitoring, retaining the raft (reduced in thickness to lower contact pressure), with a pre-agreed trigger for converting the perimeter or the whole footprint to bored piles if monitored settlement exceeds the predicted rate. Strip and recompact the 3 m fill as a condition in all cases.