Underpinning & Foundations

Underpinning covers the strengthening, deepening and replacement of foundations under structures that remain standing, together with the closely allied work that surrounds it - structural repair of the cracked masonry above, retrofit basement formation where the underpinning becomes a retaining wall, and engineered foundation solutions on difficult plots. That grouping is not this page's invention: it is how the trade organises itself. ASUC, the Association of Specialist Underpinning Contractors, describes its membership as covering subsidence repair and underpinning, retrofit basement construction and engineered foundations, and admits contractors only after a technical, health and safety, insurance and financial audit. The association's publications page carries the trade's public library - three guideline documents, technical guidance notes, and a set of method fact sheets that this page draws on throughout - and its members offer an insurance-backed guarantee on completed work, with a 10-year latent defects insurance arrangement announced with a specialist broker on the association's site.

The reference document for the work itself is recent and free to download: "Underpinning: Good practice guidance", document TWf2023:02, published in June 2023 by the Temporary Works Forum jointly with ASUC, written by a working group that included specialist contractors, temporary works designers and a Health and Safety Executive inspector. It defines underpinning as the downward extension of existing foundations by staged excavation and replacement of soil with concrete, reinforced or unreinforced, and aims itself at the people involved when the idea first appears on a project - planning consultants, architects, consulting engineers, developers, principal designers and lay clients, with contractors, temporary works coordinators, party wall surveyors and building inspectors named as secondary readers. Its scope note is candid about limits that matter: the guidance covers load-bearing walls at least 3m long under relatively uniform load, which are considered to have enough resilience to tolerate staged excavation beneath them, and it does not cover underpinning individual columns or piers, where that resilience is absent and specialist engineering advice is the stated route.

The regulatory position is unusually clean for a construction activity. The Building Regulations 2010 list work involving the underpinning of a building as a category of building work in its own right, in the definition regulation that also covers erection, extension and material alteration - which means the building control machinery applies to underpinning as such, not merely to whatever alteration prompted it. LABC, the body representing local authority building control teams in England and Wales, states the practical consequence plainly in its guidance article on underpinning: the work is notifiable in all cases, without exception, and an application to a building control authority is made before it starts. The structural requirements themselves sit behind requirement A1 and Approved Document A, the statutory guidance for structure in England, whose GOV.UK page also records that responsibility for keeping the Approved Documents under review has sat with the Building Safety Regulator since April 2023 under the Building Safety Act 2022.

On the regulator, the current position needs stating precisely because it changed recently and most published material predates the change. The Building Safety Regulator has been a standalone body corporate since 27 January 2026, when the Building Safety Regulator (Establishment of New Body and Transfer of Functions etc.) Regulations 2026 - SI 2026/20 - came into force and transferred the building safety functions previously exercised through the Health and Safety Executive. The BSR is sponsored by the Ministry of Housing, Communities and Local Government. It is no longer part of HSE, and descriptions that place it there are out of date. HSE remains the enforcing authority for construction work itself: underpinning sits under the Construction (Design and Management) Regulations 2015 like all construction, with the client, designer, principal designer and contractor duties the TWf guide walks through, and with the excavation-specific duties - supported excavations, competent-person inspections at the start of every shift, protection against falls and falling material - that HSE's construction guidance describes.

Around the statutory layer sits a small institutional ecosystem worth knowing, because it publishes nearly everything a practitioner can read on the subject without buying a standard. ASUC and the Temporary Works Forum carry the trade's own guidance. The Federation of Piling Specialists covers the piling side, including mini-piling, with audited members and published guidance on plant safety zones. LABC publishes building control commentary. The Institution of Structural Engineers publishes the standard book on the subject - "Subsidence of low rise buildings", in a second edition from 2000 - which the industry's own bibliographies cite as the practical text. The Clay Research Group publishes monthly research newsletters for the insurance side of the industry. The British Geological Survey maps the shrink-swell hazard that drives the workload. And the insurance institutions - the Association of British Insurers and the Financial Ombudsman Service - publish the claims-process material that shapes how most domestic underpinning is commissioned, because most of it arrives through an insurance claim.

The design references are the foundations codes rather than an underpinning-specific standard. BS 8004:2015+A1:2020 is the UK code of practice for foundations; BS 5930:2015+A1:2020 governs ground investigations; geotechnical design sits under Eurocode 7, where BSI's records show both the established BS EN 1997-1:2004+A1:2013 and the second-generation BS EN 1997-1:2024, published 30 September 2024, with the two generations coexisting during a transition period; and micropile execution is covered by BS EN 14199:2015. Temporary works control sits under BS 5975-1:2024 - the management-procedures half of the split that replaced BS 5975:2019 - and BuildPedia's refurbishment reference page covers that framework, the needling and propping of openings, and a comparison of underpinning methods in its section on structural alteration; this page goes deeper on the trade itself rather than repeating that material.

Finally, what this page is not. It is not a design guide, and it does not answer which method fits a given building, because the published guidance does not answer it either: the TWf guide's planning flowchart routes that question through ground investigation, structural appraisal and specialist advice, not through a table. Where this page names a document, the document is the authority. Where it reports a number - a bay width, a monitoring period, a claims figure - the number belongs to the source named beside it, and the section at the end of this page lists what could not be verified.

The TWf guide lists four situations in which underpinning is typically required, and they organise the whole demand side of the trade. First, ground levels are being lowered next to existing structures - light wells, lift pits, a neighbour's basement - in a way that would undermine or destabilise nearby foundations. Second, deep drainage or other services are being installed close to existing foundations. Third, a building has subsided, or settled differentially, far enough to need remedial action. Fourth, load is being added - an extra storey, a heavier roof, a change of use - that requires founding in a stronger stratum. The guide adds a fifth, market-driven observation: the growth of retrofit basements has made underpins common as structural retaining wall elements, a use covered later on this page. The first, second and fifth of these are planned events on someone's programme; the third arrives uninvited, usually through an insurance claim, and accounts for most of the domestic workload.

Behind the remedial workload sits the geology. Shrink-swell clay is described by the British Geological Survey as one of the most costly and widespread geological hazards globally, with worldwide costs it estimates in the billions of pounds annually, and it behaves exactly as the name says: clay soils swell when they take on water and shrink as they dry, and a foundation sitting in the zone of seasonal moisture change moves with them. The Financial Ombudsman Service's guidance to insurers - written to standardise how subsidence complaints are handled - puts the mechanism in one line: clay is made up of roughly 30% to 35% water, so soil with high clay content loses volume as nearby vegetation draws the water out, particularly during long hot summers, and the foundations follow the shrinking soil down. The hallmark of subsidence, as the same guidance describes it, is differential movement - parts of the building moving at different rates, which is what opens the tapering diagonal cracks around doors and windows that the ABI's consumer guidance teaches homeowners to recognise: sudden in onset, wider at the top than the bottom, thicker than a 10 pence coin.

Trees are the multiplier on clay. The ASUC publications page, describing the Institution of Structural Engineers' book on subsidence, records that trees and their management are responsible for over 65% of all instances of subsidence damage to domestic properties on clay subsoil, and the RICS consumer guide quotes a Which? estimate of about 70% of all subsidence cases being tree-root related. The mechanism is straightforward - roots draw moisture, high-demand species draw more, and a tree close to a shallow foundation on shrinkable clay dries the ground beneath it to depths that a season's rain does not refill - but the remedies are not, and the published claims practice reflects that. Insurers typically instruct an arboriculturalist before touching a tree; the options run from crown reduction and thinning to removal; root barriers are described in the ombudsman's guidance as viewed sceptically by many experts; and removal is not free of risk, because a clay soil rebalancing after a thirsty tree has gone can swell rather than settle. That upward movement is heave, and it does structural damage of its own - which is why heave protection runs through every underpinning method described later on this page, and why the RICS guide notes that removing a tree occasionally adds to the problem instead of solving it. Where a tree is protected by a Tree Preservation Order or stands in a conservation area, the planning regime on GOV.UK adds a consent step before works to it.

Water does the same damage by a different route. A leaking drain or water main washes fine particles out of granular ground - the ABI names sandy and gravelly soils as the most vulnerable - or softens the ground until it can no longer carry the building, and the RICS guide extends the washout mechanism to chalk. The claims machinery treats the cause distinctly: the ombudsman's guidance describes claims that begin under the escape-of-water peril, which typically carries a lower excess, and move to the subsidence peril - with its higher excess, commonly around £1,000 per the ABI - where the policy excludes subsidence damage from the water peril. The same guidance covers solution features, where percolating water erodes a cavity in chalk or limestone that eventually finds the surface, often first announced by a small hole opening in the ground. In washout cases the repair is frequently to the drainage rather than the foundations: the RICS guide records that fixing the leaking pipework is usually enough to stabilise the property without underpinning, which is one reason the underpinned share of subsidence claims stays small.

Mining legacy is the specialist corner. Buildings over historical coal workings can subside decades or centuries after the coal was worked, as voids migrate and shafts collapse, and the compensation regime for it sits outside ordinary home insurance: the Coal Mining Subsidence Act 1991 places remedial obligations on the responsible body, and claims run through a statutory damage notice rather than a household policy, a route the ABI's consumer page spells out. The public body administering the coal legacy was renamed in November 2024: GOV.UK now presents the Mining Remediation Authority, an executive non-departmental public body sponsored by the Department for Energy Security and Net Zero, with the former Coal Authority page redirecting readers to it. Property-level screening is a public service - GOV.UK carries a check for whether a property sits in a former coal mining area, and mining reports are part of conveyancing in coalfield areas - but the RICS guide adds the caution that very old local workings in some areas were never documented at all.

The remaining causes cluster around what was built rather than what the ground did. Settlement - the ombudsman's guidance defines it as downward movement from soil compressing under the weight of a building within roughly ten years of construction - is not subsidence, is commonly excluded from cover, and is distinguished by its uniformity; the guidance also lists the pretenders that mimic ground movement, from sulfate attack in floor slabs to lintel failure, thermal movement and missing wall ties. Made ground, poorly compacted fill and decomposing organic fill each produce downward movement of floors and foundations that the guidance treats on its own terms, noting that fill beneath a floor slab is generally not part of the insured site. And load increase is the quiet, planned case: an extension founded shallower than the house it joins, a loft conversion adding weight to walls on Victorian footings, a change of use - the ombudsman's guidance names extension foundations designed without regard to a site's subsidence potential as a recurring poor-design case, and the TWf guide's fourth situation covers the deliberate version, where new load simply needs better ground than the old foundations reach.

The money confirms the pattern. The ABI's data for 2025 records £6.1 billion paid in property claims - its highest annual total since collection began in 2017 - of which domestic subsidence payouts were £307 million, up 10% year on year and the highest on record, in the year the Met Office reported the UK's hottest summer. Against those numbers sit two published correctives that frame the whole trade. The RICS consumer guide estimates that fewer than 5% of properties suffering subsidence need underpinning, and describes the Institution of Structural Engineers' position as treating underpinning as a last resort; and the claims process is built around that hierarchy - investigate, monitor, manage the vegetation, fix the drains, and underpin the residue of cases where the ground itself cannot be left as found. The methods that residue relies on are the subject of the next four sections, and the table below routes between them as the trade's own fact sheets describe them.

TABLE:

Caption: Where the trade's own fact sheets place each method family (ASUC fact sheets and the TWf guide - the documents linked beside the method sections)

Head: Method family | Where the fact sheets place it | Access and disruption profile | Cautions the fact sheets give

Row: Mass concrete (hit-and-miss) | Shallow underpinning in cohesive or granular ground, widely used in shrinkable clay; partial schemes; retrofit basement assistance | Usually workable from outside the building; minimal plant; large volumes of spoil out and concrete in | Generally uneconomic beyond about 3m deep; difficult in unstable or waterlogged ground; wholly dependent on bay sequencing

Row: Beam and base | Partial or whole-building schemes where loads are moderate and bulk excavation is unacceptable; deep footings; sites where piling is not viable | Fewer, larger excavations; still spoil-heavy; external walls workable from outside | Same 3m-depth and groundwater cautions as mass concrete; needs experienced hands on poor brickwork

Row: Bored mini-piles | Deep competent strata; clay desiccated by trees, with heave sleeving or full-length reinforcement; vibration-sensitive sites | Rigs pass through domestic doorways and work in normal headroom; power pack stays outside | Design leans entirely on the site investigation; heave uplift on shafts has to be engineered out

Row: Driven mini-piles | End-bearing in granular strata; restricted access and headroom | Lightweight sectional tubes, welded starter-and-follower installation | Driving to a set is unreliable in clay because pore pressures give false sets; thin tubes are sacrificial

Row: Piled raft | Whole-building schemes; deep-seated problems; any soil, including through backfilled tips; combined subsidence and heave | Internal working throughout the footprint; new structural floor slab as part of the scheme | Anti-heave voids must be continuous; a detached building is never wholly on temporary props at one time

Row: Resin and grout injection | Void filling and stabilising sunken floor slabs; ground improvement in granular soils | Small drilled holes; days not weeks; no excavation | ASUC's published opinion is that grouting alone is not structural underpinning of load-bearing walls and is not guaranteeable without piling or excavated works alongside

The trade's own data makes the case for investigation better than any principle could. ASUC's first technical guidance note, TGN 01, opens with the results of a survey of its members: once on site, 74% of all underpinning projects increase in scope and value, the average increase is 27%, and smaller projects suffer proportionally larger increases - a pattern the note ties directly to less adequate site investigation. The note's stated purpose follows from the numbers: to define a bare minimum standard of investigation that gives loss adjusters the information to reserve a claim properly and contractors the information to price the work accurately, on the blunt reasoning that the money saved by skipping investigation is repaid with interest through variations. Its companion note, TGN 02, adds the finding that up to 10% of projects in the members' survey were technical failures needing rework, and observes that many jobs go ahead without the cause of the subsidence ever being established - the single omission the whole investigation exists to prevent.

TGN 01's minimum investigation is specific enough to be worth setting out in substance. It starts with a walk-round survey: identify the distressed areas, ask whether the pattern says subsidence or heave, log the drainage runs, basements, trees and topography, and treat the property itself as a full-scale test model, because the building has been load-testing the ground for decades and its crack pattern is the result. Then trial pits: approximately one per 8m of distressed wall, rounded up, plus a control pit against an undistressed area for comparison. Each pit exposes the foundation profile and the bearing surface and extends at least half a metre below it, with the pit log recording the concrete and brickwork quality, damage to the footing, the soils, tree roots, and groundwater strikes and standing levels; internal pits add the slab thickness and reinforcement. Pits extend downward by hand auger where possible, sampling with depth, and the note's risk assessment line is absolute about locating gas, water, electricity, communications and drainage services before any digging. In clay, the note asks for pocket penetrometer readings and moisture content samples at half-metre intervals, with Atterberg limit tests - the index tests for shrinkability - from immediately below the footing, from at least one representative depth, and again wherever the clay lithology changes; and it closes stage one by tying the pit findings back to the geological drift map. If the six questions the investigation exists to answer - ground and water conditions, cause and extent of movement, existing foundation types, scheme options, a suitable bearing stratum, and the drainage layout - cannot yet be answered, stage two is simply to keep investigating until they can.

The formal codes sit behind the trade note. Ground investigation practice is codified in BS 5930:2015+A1:2020, foundations in BS 8004:2015+A1:2020, and geotechnical design in Eurocode 7, and on anything beyond the simplest domestic scheme the investigation is specified by an engineer against those documents rather than assembled from a checklist. The TWf guide's planning stage asks for the same things at project scale: a full ground investigation establishing stratigraphy and water levels, a structural appraisal establishing form, condition and load paths, dilapidation surveys of the existing structures, a site survey of neighbouring buildings, services, utilities and transport assets, and further research where the risk register points to it - the guide names underground tunnels and unexploded ordnance as the kind of item that surfaces in urban ground.

The insurance industry runs a parallel investigation with its own published shape, because in the remedial cases someone first decides whether this is subsidence at all and what the policy answer is. The ombudsman's guidance describes the thorough investigation it expects an insurer to have made before deciding a claim: shallow trial holes to establish foundation depth and what the footing bears on, and whether nearby drains leak; hand-augered boreholes to establish the ground at depth and recover samples for analysis; CCTV inspection of the drains close to the property; and crack monitoring to establish whether the building is still moving, by how much, and whether the movement is seasonal - the signature of tree-driven desiccation. The same guidance lists the evidence file it asks insurers to produce in a dispute, which doubles as a description of good practice: the loss adjuster's reports, the site investigation report, laboratory soil results and DNA root testing where clay shrinkage is alleged, CCTV reports where drainage is alleged, and the monitoring record. Monitoring over 12 months is described as sometimes reasonable precisely because it covers all the seasons; the RICS consumer guide gives the homeowner the same expectation - measurement and monitoring for perhaps as long as 12 months, longer in extreme cases, inside a process that can take up to two years to investigate and put right - and adds the calibration that sudden cracks wider than about 3mm are the ones that justify concern.

Trees get their own investigative discipline because they are the largest single cause and the easiest to mishandle. The published claims practice is for an arboriculturalist to advise on species, water demand, influence distance and remedy - crown reduction, thinning, pollarding or removal - before anything irreversible happens; root identification by DNA testing appears on the ombudsman's evidence list because ownership and species arguments turn on it; and the legal machinery is arranged so that most domestic tree claims never reach court, through the ABI's Domestic Subsidence Tree Root Claims Agreement, under which signatory insurers do not pursue recovery against domestic tree owners unless the owner knew of the problem and took no reasonable preventative step - an agreement that does not extend to local authorities or non-domestic landowners. Protected trees route through the planning system first. And the caution recorded earlier bears repeating in the investigation context, because it is a design input: the ombudsman's guidance notes that over-enthusiastic tree management can saturate and swell the soil it was meant to stabilise, so the remedy itself is engineered against heave, not assumed to be harmless.

For the householder inside this process, the ABI's page describes the typical journey: notify the insurer as soon as subsidence is suspected; the insurer appoints an engineer or specialist; minor damage with an established cause and no ongoing movement is repaired directly; severe or ongoing movement is monitored while a long-term answer is designed; and in extreme cases the property is underpinned. A subsidence claim typically carries an excess around £1,000, damage to garden walls, drives and patios is commonly outside cover unless the house is damaged in the same event, and alternative accommodation is covered where the property becomes uninhabitable. Two pieces of market machinery sit behind the process. Where the policyholder has switched insurer, the ABI's Domestic Subsidence Agreement allocates the claim by notification date - the previous insurer inside eight weeks of the switch, shared costs between eight weeks and a year, the current insurer after a year - and the ombudsman's guidance applies the same split while warning that awareness of cracking is not the same as awareness of subsidence when non-disclosure is alleged. And on the question every claimant eventually asks, the ABI records its members' commitment under its guidance to work with policyholders to maintain cover after a subsidence claim, while acknowledging that continuation is not always possible and that specialist brokers exist for the hard cases.

Monitoring, finally, is not only diagnosis - it is the control that carries through the works, and the TWf guide places its design before the detailed design of the underpinning itself. The movement monitoring regime is considered and agreed with interested parties in the planning stage; it is usually a prerequisite of party wall awards; and it typically uses trigger action levels in a traffic-light arrangement with contingency measures pre-agreed for each level, so that an amber reading triggers a rehearsed response rather than a site meeting. The guide is direct about both purposes: all underpinning carries ground movement risk, regular monitoring of the affected buildings through the works is the control that keeps it acceptable - and the record it produces is valuable evidence when contesting spurious claims afterwards. The refurbishment reference page covers the instruments; what this page adds is the sequence: cause established, ground proven, neighbours surveyed, triggers agreed - and only then a design.

Mass concrete underpinning - the trade also calls it traditional underpinning, and site language calls the individual bays pins - is the method most UK underpinning still uses, and ASUC's fact sheet describes it as probably the most common form undertaken for residential properties. The concept has not changed in a century: excavate a short length of ground from beneath the failing footing down to a stratum with adequate bearing, fill the excavation with concrete, transfer the wall's load onto the new concrete, and repeat along the wall in a sequence that never undermines too much at once. The fact sheet places it at relatively shallow depths, in cohesive or granular soils but most widely in shrinkable clays, notes its use against both subsidence and heave, and records its supporting role in retrofit basement construction, where the pins are designed to act as retaining structure - a use covered in the basements section below.

The cycle for a single pin is where the craft lives, and the published descriptions agree on its shape. A bay is excavated beneath the footing - ASUC's sheet gives the usual maximum length as 1200mm, and the TWf guide's good-practice figures allow up to 1.5m where the structure can stand it while giving 1.0m to 1.2m as the norm - down to the designed stratum, with the depth of individual bases varying as the stratum itself rises and falls. Before any concrete goes in, the excavation is inspected: the fact sheet describes supervisory staff and local authority building control officers checking that the correct stratum has been reached and that the formation is free of soft spots and tree roots, which makes each pin an inspected, recorded event rather than a hole filled on trust. Shutters are set and the base concreted to a level short of the underside of the existing footing, leaving a controlled gap. When the concrete has cured enough to take load - the TWf guide gives 24 hours as the minimum - the gap is packed with a semi-dry sand and cement mix rammed home hard; the guide specifies the gap at between 75mm and 150mm, the mix at one part cement to three parts sharp sand with an anti-shrink additive, and a further 12 hours minimum on the dry pack itself. The alternative, in the right circumstances, is to flood the concrete up to the underside of the footing and vibrate it to expel trapped air. Adjacent pins are keyed together - joggle joints in the fact sheet's description, with reinforcement continuity through couplers where the design calls for it, and the TWf guide preferring cast shear keys or proprietary continuity systems over projecting bars, which are an impalement hazard in the next-door excavation. In shrinkable clay, anti-heave precautions - polythene sheeting and low-density polystyrene in the fact sheet's description - line the pin so that future swelling spends itself against a void former rather than the structure, and LABC's guidance makes the same point from the building control side: where the cause was heave, the scheme incorporates heave protection as a matter of course.

The sequence across the wall is the method's defining discipline, and the numbers published in the TWf guide are the ones the industry teaches. Bays are dug hit-and-miss in the numbered order 1-4-2-5-3 - never two adjacent bays in succession - so that every open excavation is flanked by ground or completed pins; no more than 20% of the wall's length is unsupported at any one time; and access pits are sized for safe working, typically the pin width by about 1.5m measured out from the wall. The guide's other dimensional rules exist for the same reason and are worth recording as the trade's published norms: pins not normally deeper than 2.5m; a minimum depth of 1m, because a shallower pin cannot be safely accessed to pack; a minimum stem thickness of 450mm, the practical minimum for a person working in the space; the formation taken at least 300mm below any adjacent reduced ground level so the completed pin keeps passive restraint; earthwork support to all faces of the pit, with sacrificial boards left behind the pin where the ground demands it; pits re-propped or backfilled with compacted granular material on completion; and no excavation in extreme weather without precautions, because a pin full of water is a pin whose bearing surface is being destroyed. Where pins would need to be wider than a metre, the guide has the depth match the width, and its method-statement expectation runs the other way - excavation methods chosen so that nobody needs to work beneath the footing at all, with upturned buckets and suction excavation named as the ways to dig the undercut without putting a person in it.

The reasons the method persists are practical, and the fact sheet lists them. The work is usually done from one side of the wall - normally outside - so the household stays in occupation; the soils are examined at close quarters in every pit and can be tested in place with hand penetrometers or vane testers, so the design is verified as it goes; the technique is simple enough for competence to be trained; plant requirements are minimal, which suits tight urban plots; a completed run of pins doubles as a root barrier; the base can be widened to spread load in weak ground; and the same pins, designed for it, retain soil for basement formation. Its limits are equally plain in the same document: spoil out and concrete in make it logistics-heavy and access-sensitive; open excavations do not survive unstable or waterlogged ground, and the TWf guide bars underpinning in water-bearing granular soils without a properly designed dewatering or ground treatment scheme; and beyond about 3m of depth the economics and the safety case both fail, which is the boundary where the piled methods in the next sections take over.

The safety spine of the method is excavation safety, and the published material treats it as the controlling risk. HSE's construction guidance states the two facts every induction repeats - no ground can be relied on to stand unsupported in all circumstances, and a cubic metre of soil can weigh more than 1.5 tonnes - and requires a competent person to inspect the excavation supports at the start of every shift and after any event that could affect stability, with written reports. The TWf guide names the key risks as ground collapse, structural collapse and unacceptable movement, notes that failures of poorly planned or executed underpinning still reach the press and have included fatalities, and adds the hazards particular to this trade: old clinker and brick spreader footings that become unstable the moment they are undermined, backfilled service trenches beside the pit that collapse into it, perpendicular wall foundations crossing the excavation, and the loads that change mid-works as the building or its neighbours are altered. ASUC's fact sheet adds the routine controls - services located and hand-exposed before digging, the structure's stability assessed and propped before excavation, shoring designed for the actual soils and continuously reappraised, contamination considered, the working area fenced by day and the pits covered outside working hours - and points to the association's guidelines on safe and efficient underpinning and mini-piling operations for the full treatment.

Two records close every mass concrete job, and they preview the records section later on this page. The building control file - the application LABC describes as required in all cases, and the inspection of each pin's formation - is the statutory record that the new foundation reached the ground the design assumed. And the contractor's own pin schedule - bay numbers, dates, depths, strata found, concrete and dry-packing dates - is the practical one, because a hit-and-miss sequence only demonstrably happened if someone wrote it down.

Beam and base underpinning is the designed middle ground between hand-dug pins and piling, and ASUC's fact sheet describes it as combining the simplicity of excavated underpinning with reinforced concrete beams spanning between discrete bases. Instead of supporting every metre of wall on its own pin, the scheme concentrates support at engineered points: a small number of mass concrete bases taken down to competent ground, with beams between them carrying the wall's load across the gaps. The fact sheet places the system where structural loads are not particularly high and bulk excavation is unacceptable, describes its normal use as partial underpinning of a failing structure, and allows whole-building schemes where appropriate.

The design comes first in a way the traditional method does not force. The fact sheet's sequence starts with a load assessment of the affected part of the building and of the ground, sizes each base for the load it collects directly and through the beams, and settles the economics base by base - plan area against depth of excavation - so the scheme lands the load where the ground is best rather than wherever the wall happens to be worst. That positioning freedom is the system's quiet advantage and the fact sheet says so: bases can be placed to avoid drains and services, to spare tree root systems, and to keep out of groundwater; the beams then bridge whatever needed avoiding.

The construction narrative in the fact sheet rewards close reading because the load-transfer choreography is the difficult part. The bases are excavated at their specified positions, checked against the design for dimension and founding stratum, lined where required with an anti-heave compressible board, shuttered and concreted to a level that will receive the beams. The beams are then formed - cut into the wall, constructed below it, or replacing the old footing entirely - with the wall meanwhile carried on temporary support jacks placed systematically as the masonry is cut away, generally at not more than 1m spacing, on concrete padstones where the ground needs them. The reinforcement cage is fabricated in place around the live jacks, with spacers holding cover; shuttering is fixed against movement during the pour; and the concrete is either flooded up to the wall - with a head of concrete and air-release provision at the rear so the contact is real - or stopped short and dry-packed once cured, exactly as in traditional work. Where the scheme includes anti-heave protection, the sheet's closing warning is the one that matters: every jack connecting ground to structure is removed at the end, because a single forgotten prop is a heave path through an otherwise continuous void. The sheet also records a variant used in the right conditions - reinforcing the existing brickwork itself, so the wall becomes the beam spanning between bases.

The advantages the fact sheet claims are concrete and mostly logistical: fewer excavations than deep traditional underpinning; less spoil and less imported concrete; schemes that thread between services; reduced disturbance to roots and groundwater; deep footings underpinned without a deep pit under every metre of them; ground conditions still verified visually and tested in place at every base; a route for sites whose space or environment rules out a piling rig; and external walls worked from outside with the household in residence. Its limits are the familiar ones for anything excavated: there is still real spoil to barrow out and concrete to bring in, and long barrow runs punish that; bases resist construction in unstable or water-bearing ground; depths beyond about 3m raise the same health and safety issues as traditional pits; and - the sheet is specific about this - poor or unstable brickwork demands experienced operatives, because the method cuts into the wall it is saving.

Two design notes from the trade's other documents complete the picture. TGN 02's design section, written for exactly these reinforced concrete schemes, asks the designer to check that the new beam genuinely acts compositely with the structure above and to watch the point loads generated by the columns of brickwork between openings - the failure mode where a beam designed for a uniform wall meets the concentrated legs of a wall full of windows. And BuildPedia's refurbishment page carries the one-line placement - beam and base suits variable ground and heavier loads and reduces the number of excavations under the wall - which this section has unpacked: it is the method of choice where support can be concentrated, provided the concentration is designed rather than assumed.

Piling takes over where excavation gives out, and the published triggers are consistent across the trade's documents: competent bearing strata too deep for economic pits - the TWf guide asks designers to consider a piled solution wherever very deep pins would otherwise put people in deep hand excavations; groundwater that open excavations cannot survive; clay desiccated by trees to depths no pin reaches economically, where the pile founds below the influence zone entirely; made ground and backfilled tips whose base sits metres down; and vibration-sensitive or access-starved sites where conventional plant cannot work. What makes underpinning piling its own discipline is the working envelope: the rigs operate inside standing buildings, through domestic doorways, under domestic ceilings, next to occupied rooms - and the entire kit is designed around that fact. BuildPedia's refurbishment page tabulates the structural arrangements - pairs of piles with needle beams through the wall, cantilever pile-and-beam where access is one-sided, and the pile-and-needle family the trade associates with the Pynford name, whose stooling technique the TWf guide's planning flowchart lists among the non-excavated options; this section covers the piles themselves as ASUC's fact sheets describe them.

Bored mini-piles are the clay workhorse. The fact sheet describes sectional continuous flight augers on tracked mini rigs that pass through a domestic doorway and work in normal room headroom, powered hydraulically from a diesel pack parked outside so the fumes stay there too. In dry, stable clay the augered hole stands open and is concreted; in stable but wet ground the pile is grouted by tremie from the bottom up, the denser grout displacing the water; and in unstable ground - fills, sands, gravels - the bore is either cased with sectional threaded steel tubes advanced ahead of the augers or drilled with hollow-stem augers that are grouted under pressure as they withdraw, so the ground is never asked to stand unsupported. The design lean is stated flatly: piles are installed to predetermined lengths calculated from anticipated soil properties, so the whole system stands on the site investigation. And because augered piles are generally used in clay, the fact sheet spends its longest passage on heave: the depth of clay affected by tree activity is established from the investigation, that length of shaft is ignored for load-bearing, and the uplift that swelling clay can apply by gripping the shaft is engineered out - either with proprietary anti-heave sleeving that limits the grip, or with full-depth reinforcement anchoring the pile head down into the working length below the heave zone, or with both. Friction piles of this family also develop useful tension capacity, which the sheet notes is essential for cantilever-beam underpinning schemes, and their vibration-free installation is why they substitute for driven piles wherever the neighbours or the structure cannot take hammering.

Driven mini-piles are the granular-ground counterpart, and the fact sheet is candid about where they do and do not make sense. A displacement pile - typically a thin-walled steel tube in short lengths - is driven to a set: a measured penetration per hammer blows, or per time under a fixed-frequency hammer, that demonstrates capacity directly. Driving to a set works where the founding stratum is granular and the pile is end-bearing; in clay it is described as slow and unreliable, because driving raises pore water pressures that produce false sets before they dissipate, and where driven piles are used in clay at all they go to a predetermined depth on a static design, calculated like a bored pile. The installation kit is built for restricted access: lightweight tube sections sized to the headroom, a starter length with a crimped point and a swaged top, followers slid into the swage and welded, and driving repeated - by winched drop weight, or by a pneumatic mole working inside the tube itself, which is the low-headroom answer - until the set is achieved, after which reinforcement goes in and the tube is filled with concrete or grout. The tube is treated as sacrificial against corrosion in aerobic ground, which the design accounts for.

The piled raft is the whole-building version, and its fact sheet describes the trade's answer to deep-seated problems that partial schemes cannot reach: a designed, piled, reinforced concrete slab extending into or under the existing walls as a complete replacement foundation. It handles any ground - cohesive or granular - and carries buildings through significant thicknesses of unsuitable material, backfilled tips included, to bearing strata at depth; it deals with subsidence and heave in the same scheme; and it leaves the property with a new suspended structural floor and a building tied together horizontally at ground level, which the sheet counts among its selling points. The construction narrative is a compressed course in load transfer: floors lifted and levels reduced inside the property; piles installed internally by mini rig, with occasional external piles where a heavy element like a chimney demands them; the walls then carried while the raft is formed - on reinforced concrete nibs let into pockets in the masonry where no heave provision is needed, or on temporary props with the slab continuous beneath the walls where loads or wall condition demand it; and where heave protection is specified, the void former made continuous and uninterrupted across the entire raft, bridging between needles or removing props entirely, so no path exists for swelling clay to push on the building. Load transfer is completed by dry pack or flooded concrete exactly as in the other methods, only after the raft has strength, with a hold period where props carried the walls. The pour sequence is described as a matter of judgment - concrete volumes, joint positions, weather, and how much of the building is on temporary support at once - governed by one absolute the sheet states outright: the whole of a detached building is never supported on temporary props at any one time.

Jacking deserves its own mention because it is both a method and a control. The TWf guide describes pre-loading with hydraulic jacks - of the ground, the structure or the propping - as a way to compress the load path deliberately: seating the structure, pre-consolidating the soils, and pre-deflecting steelwork so that the movement which would otherwise arrive as settlement is taken out of the system before load transfer. The guide pairs this with its observation that movement happens at every transfer of load between structures, that the aggregate of many small transfers can exceed reasonable expectations, and that the number of transfers is a design quantity to be reduced - jacked schemes, pile-and-needle arrangements and the stooling family all exist substantially to make load transfer a controlled, measured event rather than a hopeful one.

The framework around all of this is thin but real. Micropile execution is standardised in BS EN 14199:2015, which covers drilled piles formed with tools under 300mm diameter - the definition that captures most underpinning piles; foundations design sits under BS 8004:2015+A1:2020 and Eurocode 7, where BSI's records currently show both generations of BS EN 1997-1 during the transition; and the contractor side has its own institution in the Federation of Piling Specialists, whose members are audited and whose published guidance includes the plant restricted-zone material referenced below - the red-zone discipline that keeps people away from operating rigs, spoil handling and suspended loads in exactly the confined working areas underpinning piling occupies. The trade's own safety text remains ASUC's guidelines on safe and efficient underpinning and mini-piling operations, and the TWf guide's key risks - ground collapse, structural collapse, unacceptable movement - do not change because the excavation became a borehole.

Two different operations share the word resin in this trade, and conflating them causes real confusion, so this section separates them explicitly. The first is resin bonding by injection - repairing cracked masonry by injecting a structural grout into the fracture. The second is expanding resin injection into the ground - drilling through or beside a foundation and injecting a resin that expands and hardens in the soil. The first is an uncontroversial repair technique with an ASUC fact sheet of its own. The second is the most openly contested subject in the trade, with the supplier industry and the trade association holding published positions that do not agree, and this page reports both.

Masonry crack injection first, because it rides along with almost every underpinning scheme - a building that moved enough to be underpinned usually cracked on the way down, and rebonding the cracks restores the wall's ability to spread its stresses. ASUC's fact sheet describes the standard technique: holes of 8mm to 10mm drilled along the line of the fracture at around 75mm spacings, to roughly three-quarters of the wall or leaf thickness; the mortar along the crack cut back about 20mm; holes and fracture vacuumed clean; and a thixotropic grout - one that stays where it is put rather than running down the cavity - injected hole by hole under enough pressure to fill the fracture both ways, before the wall is repointed in a matched mortar. The sheet describes three grout families with different personalities: epoxy, the most widely used in subsidence repair, moisture-tolerant with a workable pot life; polyester, fast-setting but needing dry substrates and shrinking as it cures in larger volumes; and polymer-modified cementitious grouts, shrink-compensated and suited to bigger cracks and voids. Its stated advantages are low disturbance, speed and sealing against water; its limits are equally clear - the repair is only structural if executed correctly, it cannot be done at freezing temperatures, and a stitched crack redistributes stress no better than ordinary bonded brickwork, which is why the sheet pairs the technique with bed-joint masonry reinforcement for a strengthening repair. None of this is disputed by anyone, and building movement repairs use it constantly.

Ground injection is a different proposition, and the supplier description comes first because the technique is genuinely established in parts of the market. The dominant UK supplier, Geobear - which traded as Uretek and states that it invented resin injection ground treatment in the 1980s - describes the process as drilling small-diameter holes through or beside the foundation, injecting a geopolymer resin at engineered depths into the weak soil, and letting the material expand with force, compacting loose soil, filling voids and lifting bearing capacity, with the resin curing to 90% of its strength in about 15 minutes and most residential treatments completed in one or two days without excavation and with the household in residence. The company states that its materials and systems hold BBA certification, that it has treated more than 20,000 UK sites, that residential work carries a 10-year insurance-backed guarantee, and its site presents commercial and infrastructure divisions alongside the domestic one. Read as observed practice, the technique's commonly accepted territory is visible across the market: re-levelling and stabilising sunken floor slabs, filling voids under slabs and hardstandings, and ground improvement under infrastructure and industrial floors where taking the asset out of service for excavation is the cost that matters.

The trade association's published position draws the line at load-bearing walls, and it is worth reporting carefully because it is unusually direct. ASUC's opinion paper on epoxy resin grouting as underpinning - dated 28 July 2009 and still carried on its publications page - states the association's official position that the technique is not suitable for use as structural underpinning to load-bearing walls. Its arguments are specific. Definitionally, if underpinning means extending a foundation down to competent strata at a lower level, grouting alone does not do that unless paired with a complementary technique such as pin piling. Technically, the paper reproduces research findings from three member firms: one reporting that the injection process is hard to control because the chemical reaction continues after injection stops; one reporting no benefit in cohesive clays - the injection simply refuses early because the material cannot penetrate - and, in granular soils, voids remaining even where injection appeared successful, with a caution to CCTV-survey nearby drains before and after injection because the material can enter them; and one reporting that in clay the grout merely filled the hole its own lance had made, while in granular ground the densification achieved was real but random, impossible to direct to a specific location under a specific footing. The paper adds that a grouted solution in shrinkable clay carries no anti-heave element at all, and closes with the association's commercial position: grouting of any kind does not constitute a guaranteeable underpinning scheme under its member guarantee unless allied with mini-piling or excavated underpinning, while remaining, in its words at the time, entirely appropriate for void filling and stabilising subsided floor slabs.

The two positions are less contradictory than they first look, and the overlap is the practical takeaway. Both sides place slabs, voids and ground improvement inside the technique's competence; the disagreement is confined to whether injected resin, alone, is a demonstrated substitute for extending a load-bearing wall's foundation to competent ground. On that question this page referees nothing - it records that the association's paper is sixteen years older than the supplier's current certification claims, that assessment certificates, insurer acceptance and guarantee terms are all checkable documents, and that a reader weighing a resin proposal for a load-bearing wall is reading exactly the documents both sides point at: the current certificate and its scope, the guarantee wording and its backer, and the site investigation that says what the ground actually is - because the one point every source agrees on is that the technique's behaviour is soil-dependent.

A closing regulatory note, reported as the association's argument rather than this page's conclusion: the 2009 paper contrasts conventional underpinning - which is building work under the Building Regulations, with the inspection regime that brings - with resin injection as then practised outside that regime, and presents its members' work as carrying two independent checks, building control inspection and the guarantee scheme's technical audit. The regulations themselves list work involving the underpinning of a building as building work without defining the techniques that amount to it; how a ground treatment proposal is classified on a given job is a question for the building control body on the facts of that job, and the observed practice is simply that the question gets asked.

The retrofit basement boom changed what underpinning is asked to do, and the TWf guide states the change in one sentence: underpins are now commonly used as structural retaining wall elements. That is a different job from arresting subsidence. A pin supporting a wall against settlement fails, if it fails, by bearing - the ground beneath compresses. A pin that becomes part of a basement wall has the soil removed from one side of it, and the guide names the failure modes that arrive with the excavation: sliding and overturning, driven by the lateral earth pressures and surcharges that the retained ground applies, made worse by the water and moisture-content changes that alter soil behaviour mid-works. Mass concrete pins are heavy but unreinforced, with little bending capacity, so a run of pins exposed by a basement dig needs either extensive temporary propping - which the guide notes fills the working space with manual-handling and deflection problems of its own - or design measures that build the resistance in from the start.

The guide's design measures are the current pattern book for basement-formation underpinning, and they share one idea: use the permanent works to do the temporary works' job. Pins are given keels - deepened lowest sections embedded in undisturbed soil - or extended bases, to resist sliding without low-level props. Sections of the permanent floor slab are cast early to prop the wall's base; floor plates are installed top-down to prop the head; and the guide describes casting a reinforced concrete lining wall integrally with the mass pin so the composite section carries the bending that mass concrete cannot. Where reinforcement is added to a pin under or beside a boundary structure, the party wall dimension bites: the guide flags that a reinforced pin can be treated as a special foundation under the Party Wall etc. Act 1996, needing the adjoining owner's specific consent - one of the few places where a structural design decision requires a neighbour's signature. Deep formations are broken into multi-level underpinning - the guide's example turns a single 5m pin into two 2.5m stages - with lower stages centred on the joints of the stage above, bases blinded to protect formations from softening or drying between stages, bearing checked at every stage, and temporary toes spreading load until the full-depth pin exists. Reading the arrangement as a whole, hit-and-miss has become three-dimensional: the sequence runs along the wall and down it, and the 20% rule, the trigger levels and the records apply at every level.

The specialist reference for this work is ASUC's own: its guidelines on safe and efficient basement construction directly below or near to existing structures, in a revised second edition that the association's publications page records as endorsed by the Health and Safety Executive. The document's existence is itself a statement about the risk profile - retrofit basement work concentrates every hazard this page has described, deep excavation beside and beneath occupied buildings, groundwater, sequencing dependence and neighbour interfaces, into the smallest sites in the country - and HSE's short guide for smaller contractors on basement construction, CIS66, covers the same ground from the regulator's side. The TWf guide's contribution is the reminder that deep hand excavation is designed out wherever possible, with piled retaining solutions considered as the alternative to very deep pins.

The planning system adds its own layer in the places where residential basements concentrate. The documented example is the London Borough of Camden, where the Local Plan's basement policy requires a Basement Impact Assessment with the planning application - a technical statement covering drainage, flooding, groundwater and structural stability, prepared by chartered engineers or geologists, supported by the borough's basements planning guidance, and independently audited on the council's behalf by its retained engineers. Camden's published process, pro forma and audit arrangements are linked below as the worked example; comparable regimes operate in other boroughs, and the existence of an independent audit step is a fair measure of how much can go wrong when houses grow downwards next to their neighbours.

Waterproofing is the other half of every basement conversion, and the TWf guide places it in the sequence rather than leaving it to the end: constructing permanent works early changes the waterproofing strategy, so the guide has the underpinning designer consult the waterproofing side of the design team as the sequence is settled, under BS 8102:2022. The construction joints between pins, the interfaces between staged pours, and the junction between new slab and pinned wall are exactly the discontinuities below-ground waterproofing treats as its highest-risk details, and BuildPedia's structural waterproofing reference page covers that whole discipline - the three types of protection, the four grades, and the warranty rules that in practice require two systems under habitable basements - so this page simply records the interface: the underpinning sequence and the waterproofing design are settled together or they fight each other for the life of the building.

What the basement case adds to the trade's risk picture, finally, is concentration. The TWf guide's worked warnings - surcharge from the neighbour's building on the retained side, backfilled service trenches beside pits, moisture-content changes destabilising soils mid-sequence, the aggregate movement of many load transfers - all intensify when the excavation is deeper, the neighbours closer and the programme longer; and its monitoring regime, agreed with interested parties before design completes and run to trigger levels throughout, is the mechanism that keeps a basement-formation scheme inside the movement expectations its party wall award promised. Which is the cue for the next section, because on urban basement jobs the party wall machinery is not an administrative afterthought - it is where the neighbours' engineers, the monitoring thresholds and the right to dig at all get settled.

Underpinning lives inside the Party Wall etc. Act 1996 more than almost any other trade, because the Act's machinery catches it twice. First, the Act gives a building owner rights over shared structures themselves, and underpinning a party structure is among the works those rights cover - an owner may underpin a wall shared with next door, subject to the Act's procedures. Second, and more often engaged, the Act catches excavation near a neighbour's building even where no shared wall is touched: its adjacent excavation section applies where an owner excavates within 3m of a neighbouring building or structure and deeper than that neighbour's foundations, and extends to a 6m zone where the dig passes below a plane drawn down at 45 degrees from the neighbouring foundations - tests described here in substance, with the Act itself linked below as the authority. Since underpinning is by definition excavation below foundation level, and party structures and close-packed terraces are exactly where subsidence and basement schemes concentrate, most urban underpinning triggers the Act on one limb or the other. It applies in England and Wales; Scotland's arrangements are different, as the GOV.UK guide notes.

The machinery is notice-driven and, in the ordinary case, consensual. The building owner serves notice on the adjoining owners describing the proposed work - for adjacent excavation, the notice practice includes plans showing the excavation's depth and position relative to the neighbouring foundations; the neighbour consents or dissents; and dissent moves the matter into the Act's dispute-resolution track, where either an agreed surveyor acts for both owners or each appoints a surveyor with a third selected in reserve. The output is an award - a document settling what work may proceed, how and when, access arrangements, protection measures, and the condition evidence on both sides. The government's explanatory booklet on the Act, linked below, is the standard public walk-through of the notices, timescales and procedures, and this page deliberately describes the framework in its own words rather than restating the booklet's.

Underpinning gives the party wall process most of its hardest content, and the trade's guidance keeps colliding with it in specific, recurring places. The physical extent of the work is constrained by ownership: the TWf guide notes that the rear face of an underpin commonly aligns with the neighbour's wall face to avoid trespass, that party wall and trespass issues often govern how far pins can extend, and that the asymmetric pin cross-sections which result carry eccentric loading the design has to absorb - the guide's preference for thickening a pin's full stem over adding a rear heel is partly a collapse-risk point and partly a boundary one. Reinforcement is a consent point in its own right: a reinforced pin under or beside the boundary can amount to a special foundation under the Act, which requires the adjoining owner's specific consent rather than mere notice - the one place in the process where a neighbour can simply say no to a structural choice. And the monitoring regime is usually contractual before it is technical: the TWf guide records that agreed movement monitoring, with trigger action levels and pre-agreed contingency measures, is usually a prerequisite of party wall awards on underpinning and basement schemes, which makes the traffic-light system described elsewhere on this page a neighbour-facing commitment, not an internal control.

Condition evidence is the other half of the award in practice. ASUC's TGN 02 has the pre-contract meeting agree and implement a schedule of dilapidations of the client's and neighbours' property, with record photographs, before work starts - the same survey the TWf guide lists in its planning stage - and the reason is stated plainly in the guide's monitoring passage: movement claims arrive after underpinning jobs, some of them spurious, and the combination of a dated condition schedule and a continuous monitoring record is what separates a genuine claim from an opportunistic one in either direction. On terraces, where one owner's subsidence repair is physically continuous with three other households' foundations, that evidence file is frequently worth more than the concrete.

Time is the constraint that surprises people, and the published figures are worth planning around. The TWf guide records that serving and agreeing party wall awards can take several months, especially in urban settings with many boundaries, owners and tenants; that where the neighbour is an infrastructure owner rather than a householder - Network Rail, London Underground, Transport for London, a water company - the parallel instrument is an approval in principle, now commonplace in many London boroughs for works with the potential to affect public assets; and that a local authority approval in principle typically takes 8 to 12 weeks at minimum. On an insurance-driven scheme those clocks run inside a claim that the ombudsman's material already describes as long; on a basement scheme they run alongside the planning and audit process; and the observed pattern in the trade's guidance is early engagement - notices served, surveyors appointed and monitoring agreed while the investigation is still in the ground, so the statutory machinery and the engineering arrive at the start line together.

One boundary point completes the picture: the party wall process is separate from planning permission and from building control, a distinction the GOV.UK guide makes explicitly. An underpinning scheme in a terrace can therefore be running four consent tracks at once - a building regulations application in all cases, planning where a basement or protected tree is involved, the Act's notices and award with each affected neighbour, and an approval in principle where a public asset sits within influencing distance. None substitutes for another, and the records section that follows exists partly because four tracks produce four sets of paper about the same holes in the ground.

Strip away the methods and what remains is the discipline that makes any of them work: a sequence that never asks the building to stand on too little, a monitoring regime that proves it never did, and a paper trail that shows both. The trade's documents treat these three as one subject, and so does this section.

The sequence rules published in the TWf guide are few enough to memorise and strict enough to design around. Bays are excavated hit-and-miss in the numbered 1-4-2-5-3 order, so no excavation ever adjoins another open one; no more than 20% of a wall's length is unsupported at any time; completed pins are re-propped or backfilled with compacted granular material so finished work keeps restraining the wall; and excavation stops in extreme weather unless precautions are in place. On staged and multi-level work the same logic runs vertically - lower pins centred on the joints of the pins above, formations blinded between stages so they neither soften nor dry, bearing capacity checked at every stage because the redistribution during construction can overstress ground that the finished scheme never will, and temporary toes carrying stage loads until the full section exists. The guide's underlying observation is easy to miss: the sequence is not a programme convenience, it is the structural design of the temporary condition, and swapping bay numbers on site because access happens to suit is a redesign, made by whoever swapped them, usually without calculation. Its companion observation is about people - the risk assessment and method statement aim to keep workers out from under the footing entirely, with the digging done by machine reach and suction where the undercut demands it.

Load transfer is sequenced with the same formality because it is where settlement actually happens. The published controls run through every method on this page: minimum curing before load goes anywhere - 24 hours for pin concrete, 12 for dry pack in the TWf guide's figures; dry pack gaps sized between 75mm and 150mm so the packing can actually be rammed; flooded pours proved at their contact - the guide describes trial cores through the interface where a lower pin is flood-poured to an upper one, because a contact that was never verified is a gap that was never found; and the number of load transfers treated as a design quantity in itself, reduced where possible and pre-loaded with jacks where the aggregate movement of many transfers would exceed what the building above can absorb. Each of these is a checkable, datable event, which is why they reappear below as records.

The monitoring regime carries the whole works, and its published architecture is consistent across the documents: designed and agreed before detailed design completes; instruments and baselines in place before the first pin; readings at regular intervals throughout; and trigger action levels in a traffic-light structure with a pre-agreed response at each level - continue with attention at green, pause and investigate at amber, stop and implement the contingency at red, with the levels and responses agreed with the interested parties rather than declared to them. The refurbishment reference page covers the instruments themselves - levelling to fixed datums, crack monitoring, total stations and the rest - and the insurance context earlier on this page covers diagnostic monitoring's seasonal logic. What the works-stage regime adds is consequence: on underpinning, a missed reading is not a data gap, it is an uncontrolled interval in the exact period when the building's support is partly removed - and the TWf guide's blunt commercial point stands alongside the safety one, that the monitoring record is the evidence that answers movement claims, justified and spurious alike, after the job is done.

The statutory records are specified in detail, and they accumulate fast on an underpinning site. Excavation inspections happen at the start of every shift and after any event affecting strength or stability, by a competent person, with written reports whose required content HSE's guidance lists - place, date, findings, actions, further actions and the inspector's name. Building control inspections punctuate the pin cycle itself: ASUC's mass concrete fact sheet describes local authority officers alongside the contractor's supervision checking each excavation's stratum before concreting - which means the statutory file ends up containing a formation-by-formation record of what the ground actually was, the single most valuable as-built fact an underpinned building possesses. The consent file runs alongside: the building regulations application LABC describes as required in all cases, the party wall notices and award, planning consents where engaged, and approvals in principle where a public asset sat within range - TGN 02's pre-contract checklist has the statutory notices reviewed at the table before work starts.

The commercial records are the trade's own quality system, and TGN 02 describes them with unusual candour about why they exist - the association's survey put technical failures at up to 10% of projects and client dissatisfaction higher, and its prescribed process is aimed at both. The pre-contract meeting brings engineer, loss adjuster, contractor, householders, site supervision and, where appropriate, neighbours to one table, walks the job stage by stage - preparation, underpinning, reinstatement, repairs and decoration - agrees working areas, storage, noise, programme and payment, implements the dilapidation schedules with record photographs, reviews notices, guarantees, and health and safety, and records all of it with contentious items minuted and copied to all parties. Verification then runs through the works: the cause of movement, the actual ground and water conditions against those anticipated, visits by qualified engineers and skilled supervisors - because the site investigation was a sample, and the pins are the census. Completion closes the loop with the client walked through the finished work before the contractor leaves.

The guarantee file is the record that outlives everyone on the project. ASUC operates a contractor assessment scheme under which members' underpinning work carries an insurance-backed guarantee - LABC's guidance describes the 10-year defects insurance guarantee covering underpinning, mini-piling and foundation works offered by reputable contractors through the association, and ASUC's site describes the current 10-year latent defects insurance arrangement for members' contracted works. The guarantee's practical weight appears at the next sale: an underpinned house is a property with a history, and the documents this section has described - the pin schedule, the building control completion, the monitoring close-out, the award, the guarantee - are what convert that history from a suspicion into a file. The trade's 2009 resin paper made the same point from its own angle: its members' selling proposition was two independent checks on every job, building control's and the guarantee scheme's, and both checks are only as good as the records they leave.

A new foundation is specified into existence: the designer chooses the depth, the width, the concrete and the reinforcement, and the ground investigation tells the designer what to choose. Underpinning inverts every one of those relationships. The footing already exists, at a depth nobody alive chose, in a condition nobody has seen; the wall above has a strength that can be estimated but not specified; the loads travel down paths that alteration and decay have rerouted since the original builder set them; and the ground arrives as a series of discoveries, one pin at a time. BuildPedia's refurbishment page states the principle for structural alteration generally - the properties of the existing are established, not specified - and underpinning is that principle at its purest, because the element being replaced is the one carrying everything.

The practical consequences run through every earlier section of this page, and gathering them shows how deep the difference goes. Design proceeds on provisional information: ASUC's investigation note builds its trial-pit regime around exposing what the footing actually is - old clinker and brick spreader footings appear in the safety guidance precisely because opening them up is the first time anyone learns they are there - and its survey data shows what provisionality costs, with 74% of projects growing in scope once the ground is open and thin investigation driving the growth. Verification replaces assumption: the formation of every pin is inspected against the design intent, the actual ground is compared with the anticipated ground as a standing agenda item, and the design is expected to move when the ground disagrees with it. And the temporary condition dominates: a strip footing for a new house is designed for one load case, the finished one, while an underpinning scheme is designed for dozens - every stage of every bay, every prop, every transfer - which is why the TWf guide reads as a temporary works document and why the trade's good practice guidance was published by the Temporary Works Forum at all.

The regulatory identities differ too, in a way the documents make unusually clean. New foundations are approved as part of the erection of a building; underpinning is its own category - the Building Regulations' definition of building work lists work involving the underpinning of a building as a class in itself, and LABC's guidance converts that into practice: notifiable in all cases, without exception, whatever prompted it. The codes overlap but do not coincide: both activities sit under BS 8004 and Eurocode 7 - where BSI's records show the 2004 and 2024 generations of BS EN 1997-1 coexisting through the current transition - but underpinning adds the execution standard for micropiles, the temporary works framework of BS 5975-1:2024, and the party wall machinery that new-build foundations only meet at boundaries. Even the tree problem splits by tense: NHBC's chapter on building near trees - access-restricted to registered builders, like the rest of its Standards - governs how deep a new foundation goes where trees and shrinkable clay coincide, designing tomorrow's desiccation out at the start, while the underpinner meets the same physics twenty years later, as a claim, with the tree now mature, protected, or next door.

Movement tolerance is the quiet technical difference. A new foundation settles into place under a building that has never known another shape; published trade guidance for low-rise work notes that such settlements are usually small and rarely damaging, with design governed by bearing capacity. An underpinned building has already moved - that is why anyone is there - and is often cracked, patched, brittle and sensitive to further distortion; the scheme is therefore judged not on ultimate capacity alone but on how little additional movement it induces while being built, which is why pre-loading, transfer counting, trigger levels and hold points populate this page and barely feature in new-build ground works. The building is not the beneficiary of the operation; it is a live participant, monitored like one.

The market treats the two differently as well, and the insurance documents are candid about it. A house on new foundations carries a warranty; an underpinned house carries a history, and the ABI's guidance describes the consequences - insurers working with existing policyholders to maintain cover after a subsidence claim under the association's guidance, but continuation not always possible, new cover available at higher premiums or on different terms while risk remains, specialist brokers as the published route for hard cases, and the ombudsman's material handling the disputes that follow when insurers change mid-history. The trade's answer is the documentation described in the records section - the guarantee, the building control completion, the monitoring close-out - and the observed pattern is that a well-papered underpinned property transacts, while an undocumented one becomes a survey argument. The economics point the same way: the RICS guide's range for underpinning a typical house runs from £10,000 to £75,000 and more, against investigation and vegetation works costing a fraction of that, which is the arithmetic behind the guide's fewer-than-5% figure and behind the claims hierarchy that tries everything else first.

None of which makes underpinning a lesser discipline than foundation design - the dependency runs the other way. The methods on this page exist because buildings outlive their ground conditions: the clay dries, the drain leaks, the mine collapses, the city digs downwards, and the stock that results is standing on every kind of foundation the last two centuries built. New foundation design is the subject the codes were written for; underpinning is the subject that keeps the codes honest, because it is where the industry finds out, footing by footing, what the ground under the built environment actually turned out to be. The trade that does it - audited, guaranteed, published about, and small - is the one this page set out to describe.

BuildPedia would rather tell you where the evidence runs out than round it off. Underpinning's published base is unusually open - most of the trade's own documents are free downloads - but parts of it are dated, some institutional sites resist checking, and several widely quoted figures trace to sources this page could not reach. The following were the open points at the time of writing, and each is a place to check the primary source rather than this page.

BULLETS: