How to Determine Bearing Capacity: Every Method Compared

If you ask five engineers how they got their bearing capacity number, you will often get five different answers, and all five can be defensible. Bearing capacity is not measured directly; it is derived, and there are many ways to derive it depending on the project, the ground, and how much risk you are willing to carry.

This post works through every family of method actually used in practice, analytical equations, field tests, laboratory tests, settlement-based checks, numerical modelling, rock methods, and the codes that set the safety margin around all of them, so you can match the method to the project instead of defaulting to whatever you used last time.

Quick Comparison

1. Presumptive (Deemed-to-Satisfy) Bearing Values

For standard residential construction, the National Construction Code lets you skip a project-specific bearing capacity calculation altogether. Pad footings, strip footings and edge beams not connected to the slab can be founded in natural soil at a deemed allowable bearing pressure, provided the site is classified A, S or M under AS 2870 and the bearing is reasonably uniform. Under AS 2870, design bearing capacity (including uplift) is generally taken as no more than roughly a third of the assessed ultimate bearing pressure, which is effectively a built-in factor of safety of about 3.

This is the fastest and cheapest way to get a number, and for its intended use case (light, uniform, low-consequence residential footings) it is entirely appropriate. It stops being appropriate the moment the site is a P-classified problem site, the loads are heavier than standard residential, or the footing bears on fill, sloping ground or variable strata that the presumptive table was never meant to cover.

2. Analytical Bearing Capacity Equations

This is the classical approach taught in every geotechnical course: take shear strength parameters and foundation geometry and put them through a bearing capacity equation. Terzaghi's original 1943 equation started this approach:

qu = c·Nc + q·Nq + 0.5·γ·B·Nγ

where qu is the ultimate bearing capacity, c is cohesion, q is the surcharge pressure at founding level, γ is the soil unit weight, B is the footing width, and Nc, Nq, Nγ are bearing capacity factors that depend on the friction angle.

Several later methods extend this same basic idea with additional correction factors:

The output depends entirely on the strength parameters going in, cohesion, friction angle, and unit weight, drawn from laboratory or field testing, and on whether the analysis is run as a drained (effective stress) or undrained (total stress) case. For structures governed by AS 5100.3 (bridges) or AS 2159 (piles), the calculated ultimate capacity is then cut down by a geotechnical strength reduction factor before it becomes a design value, sometimes as low as 0.45 depending on the assessment method and consequence.

This method family is transparent and defensible, and it is the workhorse for most engineered footings. Its weak point is that every one of these equations models an idealised failure surface in a homogeneous soil; real ground with layering, fill, or unusual groundwater conditions needs engineering judgment on top of the formula.

3. In-Situ Field Tests

Field tests estimate bearing capacity directly or indirectly from testing the ground where it sits, rather than in a laboratory.

SPT correlations (Meyerhof's among the most widely used) are most reliable in cohesionless soils like sand, where the test was originally developed and validated; in clays the correlation is weaker and SPT is better used to guide sampling for proper strength testing. CPT's advantage is a near-continuous profile with depth instead of a data point every 0.5-1.5 m, which matters in layered or variable ground where a borehole alone can miss a weak layer between test intervals. DCP/Scala testing is the cheapest and fastest option, useful for screening and light structures but shallow and unsuitable for gravel or rock. A plate load test measures real load-settlement behaviour directly, but the plate (typically 300-750 mm across) is much smaller than the zone of influence beneath a real footing, so scaling the result up needs care. Pressuremeter and vane shear tests are less routinely used in Australian practice than SPT/CPT but are genuinely useful where those two don't suit the ground, particularly vane shear for very soft, sensitive clays where sampling for lab testing can disturb the soil too much to be reliable.

4. Laboratory Test-Based Methods

Where in-situ testing alone isn't enough, samples are tested in the laboratory to obtain design parameters directly.

The triaxial test is the most versatile of these: run as UU (unconsolidated undrained), CIU or CID, it can produce either total-stress or effective-stress strength parameters depending on which design condition you're checking. Direct shear is a simpler, cheaper alternative, commonly used for sands and granular soils, though the failure plane is forced rather than found naturally. The unconfined compression test is quick and specific to cohesive soils, giving undrained shear strength as roughly half the unconfined compressive strength (su ≈ qu / 2). Consolidation testing doesn't feed the bearing capacity equation directly, it feeds settlement calculations, but it's listed here because allowable bearing pressure is very often governed by settlement rather than shear failure, which is the next method family.

5. Settlement-Based Methods

In many practical cases, especially for sands and soft clays, the pressure a footing can actually carry is controlled by how much it will settle, not by when the ground will shear. A footing can pass every bearing capacity check and still be unacceptable if it settles, or settles unevenly, more than the structure can tolerate.

Common settlement methods include:

• Elastic (immediate) settlement analysis
• Consolidation settlement analysis, using oedometer data
• Schmertmann's method for settlement of footings on sand, using CPT data
• Janbu's tangent modulus method
• Burland and Burbidge's method, an empirical settlement estimate for sands based on SPT data

The allowable bearing pressure is then selected so the calculated settlement, immediate plus any longer-term consolidation, stays within the project's acceptable limits, which may govern the design well before the shear-failure bearing capacity is reached. The site already has worked calculations for the two most common cases: Immediate Settlement of Shallow Foundations and Consolidation Settlement and Time-Rate Calculations.

6. Numerical Modelling (FEM, FDM and Limit Analysis)

For complex geometry, layered ground, nearby structures, unusual loading, or where the failure mechanism is unlikely to match a textbook bearing capacity equation, finite element methods, finite difference methods, limit equilibrium analysis, or advanced constitutive soil models can represent the problem directly rather than relying on a closed-form formula.

These tools can model conditions the analytical equations can't easily capture: layered soils with contrasting properties, nonlinear soil behaviour, groundwater effects and seepage, staged excavation and construction sequencing, raft foundations, embankments, and ground improvement systems. This is standard practice for larger or higher-consequence projects, and it is often paired with the analytical method as a cross-check rather than used entirely on its own.

The output is only as good as the ground model and parameters that go into it, so numerical modelling doesn't remove the need for good site investigation data, and it needs an experienced analyst to build and interpret the model sensibly; the software will happily produce a confident-looking number from poor inputs.

7. Rock Bearing Capacity Methods

Foundations bearing on rock are a different problem from foundations on soil. Bearing capacity is typically governed by the rock mass, not the intact rock, since discontinuities, weathering and structure usually control behaviour long before the intact rock strength does. Inputs typically include:

• Uniaxial compressive strength (UCS) of intact rock
• Rock mass classification systems such as RQD, RMR or GSI
• Discontinuity spacing, orientation, and condition (infill, roughness, persistence)
• Plate load testing carried out directly on rock
• Empirical code recommendations for rock-founded footings
• Weathering profile and depth to sound rock

In practice, a highly weathered or closely jointed rock mass can have an allowable bearing pressure far below what the intact rock strength alone would suggest, which is why mass classification and defect mapping matter as much as the UCS number. See Rock Mechanics Testing for how the underlying strength data is obtained.

8. How Codes Set the Safety Margin

None of the methods above exists in isolation from the applicable design code, and it's worth being clear about what codes actually do here: outside of presumptive tables, codes don't usually hand you a bearing capacity number directly. What they do is set the investigation basis and the safety margin applied to whichever method you use.

In Australia, AS 1726 sets the expectations for how the site investigation itself should be planned and executed, AS 2870 provides the presumptive pathway for residential footings, and AS 5100.3 (bridges) and AS 2159 (piles) apply geotechnical strength reduction factors to calculated ultimate capacity rather than a simple factor of safety. Internationally, Eurocode 7 uses a partial factor approach applied to actions and material properties, and AASHTO uses a load and resistance factor design (LRFD) framework; both are conceptually similar to the Australian limit-state approach even though the factors and design cases differ. Road authority specifications add another layer of project-specific requirements on top of whichever base standard applies.

Ultimate, Allowable and Design Bearing Capacity

The terminology here trips people up as much as the methods do:

The traditional relationship is:

q_allow = qu / FS

where FS is typically around 2.5 to 3.0 for conventional allowable stress design, though the appropriate value depends on the project, the method's reliability, and the applicable standard. Limit-state approaches (AS 5100.3, AS 2159, Eurocode 7, AASHTO LRFD) replace this single factor of safety with method-specific reduction or partial factors instead.

Construction-Stage Verification

None of the design-stage methods above eliminates the value of checking what is actually in the ground once excavation starts. Footing inspections, proof rolling, and, for piles, load testing all confirm, or challenge, the assumptions made at design stage against real conditions.

So Which Method Should You Actually Use?

In practice, most projects combine several of these-

• Standard residential footing on a good site: presumptive AS 2870 values are usually sufficient on their own.
• Typical commercial pad or strip footing: SPT or CPT data feeding an analytical bearing capacity equation, checked against settlement, with a DCP as a quick supplementary check across the footprint.
• Soft or compressible clay sites: vane shear or triaxial testing for undrained strength, with settlement analysis often governing the final design pressure more than the bearing capacity equation itself.
• Bridge or major structure footing: analytical or numerical methods governed by AS 5100.3, informed by CPT and/or SPT data, with the appropriate geotechnical strength reduction factor applied.
• Footings on rock: rock mass classification and defect mapping alongside UCS testing, sometimes backed by a plate load test on rock.
• Unusual, layered, or high-consequence ground: numerical modelling alongside a plate load test or pile load test for direct verification.
• Every project, regardless of method: construction-stage verification (footing inspection, proof rolling) to confirm the founding conditions match what the design assumed.

Related Resources

• Bearing Capacity of Soil and Foundations
• Shallow Foundation Bearing Capacity Calculation
• Shear Strength of Soil
• Standard Penetration Test (SPT) Procedure
• Cone Penetration Test (CPT)
• Dynamic Cone Penetrometer (DCP) Testing
• Plate Load Testing
• Immediate Settlement of Shallow Foundations
• Consolidation Settlement and Time-Rate Calculations
• Rock Mechanics Testing
• Pile Capacity
• Geotechnical Site Investigation