Underpinning Existing Foundations: Methods and Sequencing
Underpinning is what happens when an existing foundation needs more capacity, or a deeper bearing level, than it currently has − because an adjacent excavation will remove the soil currently supporting it, because the building is settling and the original bearing stratum has proven inadequate, or because a new addition or added story increases the load on a foundation sized for a lighter original structure. In every case the challenge is the same: transfer the existing load to a new support system without the building noticing, ideally without taking it out of service at all.
Mass Concrete Pit Underpinning
The traditional method excavates a sequence of pits beneath the existing footing, one at a time in a specific non-adjacent order (commonly numbered so no two simultaneously open pits are next to each other), down to a new, deeper, adequate bearing stratum. Each pit is then filled with mass concrete built up in successive lifts, with the final lift dry-packed tightly against the underside of the existing footing to ensure full contact and load transfer before the pit is backfilled and the next pit in sequence is opened. The non-adjacent sequencing is the load path's real design element: at any moment during construction, the existing footing must still be supported by unexcavated ground on either side of the open pit, so opening two adjacent sections at once would remove too much of that temporary support and risk a local foundation failure exactly like the one the whole project is trying to prevent.
This method is labor-intensive and slow because it typically proceeds one small pit at a time by hand or with confined excavation equipment, but it needs no specialized piling rig and works in confined basement spaces where larger equipment cannot maneuver, which is why it remains common for older masonry buildings with shallow strip footings needing modest additional depth or capacity.
Underpinning design always has to answer a question ordinary new-foundation design does not: how much of the existing structure's actual load is really reaching this footing right now, and how will that load redistribute as adjacent pits are sequentially loaded and unloaded during construction? A poorly sequenced underpinning job can transfer more load than expected onto a still-unsupported section, or induce differential settlement between pits completed early and those completed later, even when every individual pit's final design is adequate.
Beam and Pile / Mini-Pile Underpinning
Where deeper support is needed, or where mass concrete pits would be impractical because the adequate bearing stratum is too far below the existing footing, a needle beam or continuous reinforced concrete beam is cast beneath (or through) the existing footing, transferring its load out to piles installed on either side rather than directly down through a pit. Mini-piles (small-diameter, typically drilled and grouted, often under 300 mm) are especially suited to this work because they can be installed with low-headroom, low-vibration rigs that fit inside an occupied basement or beneath an existing structure without the ground-borne vibration that driven piles or larger drilled shafts would create close to a building already showing distress.
This approach transfers load in a genuinely different path than a mass concrete pit: rather than simply extending the existing footing's bearing down to a new stratum directly below its own footprint, the beam picks the load up and carries it laterally out to piles that may be positioned specifically to avoid an adjacent excavation, existing utility, or a property line, giving the designer much more flexibility in where the new bearing points actually land relative to where the old footing sits.
Monitoring and Sequencing Control
Every underpinning project needs a monitoring program independent of the design calculations − typically precise leveling of the structure at multiple points, sometimes supplemented by crack gauges on masonry, tracked at defined intervals throughout construction and compared against trigger levels that halt or slow work if movement approaches a threshold the design assumed would not be exceeded. This monitoring exists because underpinning, unlike ordinary new foundation construction, is performed on a structure that is already carrying its full service load throughout the work, with no ability to simply stop and re-check calculations the way an engineer could before a new building's foundation is loaded for the first time; the building keeps settling, deflecting, and responding to every stage of the underpinning sequence in real time, and the monitoring program is what tells the contractor and engineer whether the sequence assumed on paper is actually behaving the way the design predicted.
Adjacent excavation is one of the most common triggers for underpinning, since removing soil next to an existing shallow footing can undermine the bearing support the footing's design assumed was there, closely tied to the same bearing capacity principles used in original footing design; excavation safety and support requirements near existing structures are addressed in OSHA's excavation and trenching standards, published at osha.gov/trenching-excavation, which many underpinning sequencing and shoring plans are built around even though the standard itself is a worker-safety rather than a structural design document.