How the two differ in kind
When the ground at foundation level will not carry the building, there are two routes, and they are opposites rather than variants.
PILING bypasses the problem. Piles transfer load through the weak material to something competent — end bearing on a firm stratum or rock, friction along a long enough shaft, or both — and the weak ground in between is simply passed through. The foundations become a pile cap on the piles rather than a footing on soil.
GROUND IMPROVEMENT changes the problem. Stone columns, dynamic compaction, deep soil mixing, preloading with surcharge, or vibro-compaction all make the existing ground stiffer and stronger in place, so that conventional shallow foundations become adequate. Nothing bypasses anything; the soil does the work it was previously unable to do.
The first question is therefore not cost but whether there is anything to reach. A deep soft deposit with no competent stratum within economic depth removes piling's premise — the piles would be relying on friction through material that has little to offer. A thin soft layer over rock removes ground improvement's, since it is cheaper to pass through a few metres than to treat them. Between those, both work, and the comparison becomes one of cost, programme, vibration, groundwater and what the structure is sensitive to.
One technical fact deserves stating before any pile count is estimated: a GROUP of piles does not carry the sum of what its members carry individually. Each pile stresses a zone of soil around it, those zones overlap when piles are close together, and the group's capacity is reduced accordingly by an efficiency factor that falls as spacing tightens. Designing a group by multiplying a single-pile test result by the number of piles overestimates capacity, and it overestimates it in the direction that settles.
The factors that actually differ
Show
| Piling | Ground improvement | |
|---|---|---|
| What it does with the weak ground | Passes through it. The weak material carries little or nothing. | Improves it so it carries the load itself. |
| Precondition | A competent stratum within reach, or enough shaft friction to develop capacity. | Ground of a type that responds to the chosen technique — granular soils to compaction, soft cohesive soils to columns or mixing. |
| Resulting foundation | Pile caps and ground beams — a different and more expensive substructure than a footing. | Conventional shallow footings or a raft, which is much of the saving. |
| Group behaviour | Capacity is NOT the sum of individual capacities; overlapping stress zones reduce it by an efficiency factor. | Treated as an improved ground mass with a design bearing capacity and settlement characteristic. |
| Settlement control | Excellent where end bearing is on a firm stratum. Predictable and small. | Good but larger, and the design is usually settlement-governed rather than bearing-governed. |
| Vibration and noise | Driven piles are severely disruptive; bored and CFA piles far less so. A real constraint near existing buildings. | Vibro techniques and dynamic compaction are also vibration sources — dynamic compaction especially so. |
| Groundwater | Manageable, and casing or support fluid deals with unstable bores. | Often helpful for stone columns, which also act as drainage paths and accelerate consolidation. |
| Spoil and disposal | Bored piles generate arisings, which on contaminated ground is a significant disposal cost. | Compaction methods generate little; stone columns import material rather than export it. |
| Programme | A specialist rig, a piling mat, then testing. Quick per pile, and the mobilisation is the fixed cost. | Fast over large areas — but preloading takes MONTHS of waiting while consolidation happens, which is programme rather than cost. |
| Verification | Pile testing — static load tests, dynamic testing, integrity testing — on a sampled proportion. | In-situ testing of the improved ground: penetration testing, plate loading, and settlement monitoring. |
Which one, and when
Choose piling when…
- There is a competent stratum within reach and the loads are heavy or concentrated.
- Settlement must be small and predictable — a sensitive structure, a stiff frame, a building connected to something that must not move.
- The soft deposit is thin enough to pass through economically.
- Loads arrive at columns rather than spread over an area, which is what piles suit.
Choose ground improvement when…
- The soft deposit is deep with no competent stratum within economic reach.
- Loads are spread over a large area — a slab, a yard, a tank farm, an embankment — rather than concentrated at columns.
- The ground type responds well to the technique and moderate settlement is acceptable.
- Conventional shallow foundations are wanted for cost and buildability, and this is what makes them possible.
Now run your own numbers
This page holds no prices on purpose — a national average is wrong for almost every real project. Quantify both options with your dimensions and your local quotes.
Frequently asked questions
- Why doesn't a pile group carry the sum of its piles?
- Because each pile stresses a zone of soil around and beneath it, and when piles are close together those zones overlap — so the soil between them is being asked to carry load from more than one pile at once. The group therefore behaves partly as a single large block rather than as a set of independent members, and its capacity is lower than the arithmetic sum by an efficiency factor that falls as spacing tightens and as the group gets larger. The same overlap makes group SETTLEMENT considerably greater than a single pile's at the same load per pile, because the stressed soil volume beneath a group is much deeper than beneath one pile. Both effects push the same way, and both are why a group is designed as a group rather than as a multiplication of a single-pile test result.
- How do stone columns actually work?
- By replacing a proportion of the weak soil with compacted stone, so the composite behaves better than the original ground — and the governing number is the area replacement ratio, the fraction of the treated area that is column rather than soil. The columns are much stiffer than the surrounding soil, so they attract load away from it, which both increases the bearing capacity of the treated mass and reduces settlement. They rely on the surrounding soil to confine them laterally, which is why very soft soils with little lateral resistance limit what the technique can do — a column that can bulge outward under load is not carrying much. A useful secondary effect in cohesive soils is drainage: the stone columns are vertical drains, so consolidation under the new load happens much faster than it otherwise would.
- Which is cheaper?
- Ground improvement, usually, where it is applicable — because it keeps conventional shallow foundations and avoids pile caps and ground beams, and it covers area quickly. The comparison is not simply cost per square metre, though. Piling has a substantial mobilisation cost, so a small job carries it over few piles and a large one spreads it thin. Ground improvement over a large footprint is efficient and over a small isolated foundation is not. And the real comparison has to include the substructure that follows: piles bring a pile cap at every column and beams spanning between them, which is a significant concrete and reinforcement cost that the shallow-footing alternative does not carry. Where both are technically viable, that downstream difference is frequently what decides it rather than the specialist's rate.
- What does a site investigation need to tell me?
- Whether there is a competent stratum, at what depth, and what the material above it is — which is precisely the information that decides this question, and precisely what is missing when the decision gets made on assumption. That means boreholes deep enough to prove the bearing stratum rather than stopping in it, in-situ testing to characterise strength and stiffness, groundwater levels including seasonal variation, and enough coverage to show how the strata vary across the site. Variation matters as much as depth: a competent layer that rises and falls across a footprint changes pile lengths considerably and can leave part of a building on improved ground and part on piles. The investigation is also the only place contamination shows up, which changes the disposal cost of bored pile arisings substantially and can rule a technique out.
- Can vibration damage neighbouring buildings?
- It can, and it is one of the main reasons a technique gets ruled out in an urban setting. Driven piling is the most severe — a hammer imparting energy into the ground repeatedly, transmitting through the soil to adjacent structures — and dynamic compaction, which drops a large weight from height, is in the same category. The risks are cosmetic damage, structural damage where a building is already weak, and settlement of loose granular soils beneath neighbouring foundations, which can do more harm than the vibration itself. The mitigations are to choose a low-vibration technique instead — bored or continuous flight auger piling, or stone columns installed by a method suited to the setting — and to monitor: a condition survey of adjacent property before work starts and vibration monitoring during it, with agreed limits and a plan for what happens if they are exceeded.
- What is preloading and why does it take so long?
- Placing a surcharge — usually a temporary fill mound heavier than the eventual structure — on soft ground so that it consolidates before the building goes on it, then removing the surcharge and building on ground that has already done most of its settling. It works, it is conceptually simple, and it is often the cheapest option in materials. What it costs is TIME: consolidation of a soft cohesive deposit is governed by how fast water can drain out of it, and that can be months or longer depending on the thickness and permeability of the layer. The usual accelerant is vertical drains — wick drains or stone columns — installed to shorten the drainage path, which can reduce the waiting substantially. Whether preloading is viable is therefore a programme question rather than a technical one, and it is decided very early or not at all.
- Can the two be combined?
- Yes, and it is common on sites where the ground varies or where different parts of a structure have different sensitivities. A frequent arrangement is piles under the heavily loaded columns and improved ground under the slab and the lightly loaded areas, which puts the expensive solution only where the load demands it. The detail that needs care is the transition: a slab on improved ground will settle more than a pile cap on end-bearing piles, so the connection between them has to accommodate differential movement rather than resist it — a slab tied rigidly to a pile cap cracks at the joint as the ground settles away from it. Designing for that difference is straightforward once it is recognised, and it is the thing to look for when reviewing a mixed scheme.
- How is the result checked?
- Differently for each, and both are verified rather than assumed. Piling is tested: a static load test on one or more piles to demonstrate capacity, dynamic testing on a sample to check driving behaviour and capacity, and integrity testing across a much larger proportion to detect defects in the shaft — necks, inclusions, or a pile that did not reach its designed length. The proportion tested is specified rather than left to judgement. Ground improvement is verified by testing the improved ground: in-situ penetration testing before and after to demonstrate the change, plate loading tests, and settlement monitoring under the actual structure or a trial load. In both cases the specification should say what will be tested and what happens if a test fails, because that conversation is much harder after the rig has left.