Pile Load Tests Below the Specified Factor of Safety
Three static compression load tests were carried out on 750 mm diameter bored piles, 26 m long, founded through dense sand into stiff clay, for a single-storey warehouse. The design working load is 2,300 kN per pile and the specification requires a factor of safety of 2.5 on ultimate capacity, which means a minimum interpreted ultimate capacity of 5,750 kN. Each test pile was loaded in increments up to 5,400 kN or plunging. Interpretation by the Davisson offset-limit criterion gives ultimate capacities of 4,400 kN, 4,900 kN and 5,150 kN for TP-1, TP-2 and TP-3 respectively. Settlement at the working load was 6 mm, 9 mm and 7 mm, and no pile head distress was observed. The test piles were installed three weeks ahead of the production piles and by a different drilling subcontractor. The structural engineer needs a decision within two weeks; a redesign adds about six weeks and roughly 9 % to the foundation cost.
1. Evaluate compliance with the specification and quantify what the test results actually show.
Read the worked answer
Convert each result into the mobilised factor of safety at the working load of 2,300 kN:
- TP-1: 4,400 / 2,300 = 1.91
- TP-2: 4,900 / 2,300 = 2.13
- TP-3: 5,150 / 2,300 = 2.24
All three are below the specified 2.5, so on the stated criterion the design does not comply and cannot be signed off as tested. That is the simple part.
The more important information is the spread: 4,400 to 5,150 kN is a range of about 17 % around a mean near 4,800 kN. For nominally identical piles in a layered sand-over-clay profile, that spread indicates real construction sensitivity - likely base cleanliness, drilling method, or concrete placement - not merely testing noise. In that light the mean is not a safe design value; the lowest result is the one that governs, because the pile that behaves like TP-1 is the pile that controls the foundation.
Note also that the criterion (Davisson ultimate) is a capacity measure, not a serviceability measure. The 6–9 mm settlement at working load is small and tells us the piles are stiff well below their ultimate capacity - which is the argument the contractor will make, and it deserves to be examined rather than dismissed.
2. The contractor argues the factor of safety may be reduced to 2.0 because settlements are well within the 25 mm limit. How would you respond?
Read the worked answer
Partly agree with the logic, disagree with the number, and disagree strongly with applying it on this evidence.
The logic is sound in principle: a factor of safety on ultimate capacity is a proxy, and where the actual serviceability performance at working load is known and comfortable, a settlement-based acceptance criterion is technically more meaningful than an arbitrary ratio. Here, 6–9 mm at 2,300 kN against a typical 25 mm limit is a wide margin, and the load–settlement curves show no plunging at working load.
But three things block acceptance of FS = 2.0 now:
- Sample size and selection. Three tests, with the worst at 1.91, is not a basis for lowering a global factor. Reducing the required FS below the lowest tested result would mean accepting a foundation with an unverified margin on every untested pile.
- The test piles are not the production piles. Different subcontractor, different date, and a construction-method sensitivity already visible in the 17 % spread. Whatever the tests show may not transfer.
- The structural consequence. If capacity is relied upon at FS 2.0, any single defective pile under a column line has a much thinner reserve, and the pile cap and column design must be re-checked for the resulting eccentricity.
A defensible position: retain FS 2.5 as the governing criterion, but use the measured load–settlement behaviour to argue that the shortfall is about capacity interpretation rather than serviceability - which opens the door to targeted verification rather than a blanket reduction.
3. Recommend the course of action you would put to the engineer, with the reasoning behind each element.
Read the worked answer
A single global decision is worse than a targeted package:
- Re-test with the production method. Install and test two additional piles using the production subcontractor's plant and sequence, instrumented at the base and shaft. This directly addresses whether the spread comes from method; it costs far less than the 9 % foundation uplift and takes less time than the six-week redesign.
- Separate base and shaft resistance. If the shaft friction is consistent and the base is variable, base-grouting or a modest lengthening is a cheap, targeted fix; if the shaft is variable, the design is genuinely short and the piles must change.
- Design on the lowest credible result, not the mean. For the current layout, size the foundation on the TP-1-type capacity or reduce the working load to about 1,900–2,000 kN, and confirm the pile cap and column for the resulting reactions.
- If the programme forces a decision before re-testing, accept a reduced global FS only if it is paired with (a) 100 % integrity testing (low-strain) on production piles, (b) a static re-test, and (c) the structural designer's written acceptance of the settlement and eccentricity consequences.
Reasoning: the risk here is not that the piles are weak - the settlements say they are stiff - it is that the variability is unexplained. Buy information before buying concrete.