Digging a hole sounds simple until the hole becomes deep, wide and close to a city that must continue functioning around it.
Soil does not remain obediently vertical because a construction boundary says it should. Remove material from one side of the ground and the stress state changes. Nearby roads, drains, utilities and buildings continue to depend on the ground that remains. Water can move through the same soil that the excavation is trying to expose.
This is why excavation is an engineering problem, not merely an earthmoving activity.
An Earth Retaining and Stabilising Structure, or ERSS, is a system used to support and control ground while excavation and related works proceed. Depending on the project, it can involve retaining walls, walers, struts, anchors, slabs, temporary bracing and carefully sequenced excavation. The actual design is project-specific and must follow approved plans and Singapore’s building-control requirements.
BCA’s current Guidelines for ST Plan Applications, updated in September 2026, maintain a dedicated ERSS and slope section covering geotechnical requirements, excavation depth, groundwater control, impact assessment and observational methods. BCA’s 2026 Design and Engineering Safety Awards also highlighted an ERSS project as an example of how constrained urban excavation can be engineered safely.
This article explains how the system works conceptually. It does not provide a shoring design, strut spacing, anchor capacity, excavation sequence or safe-work procedure. Those details belong to the project’s Qualified Persons, Accredited Checkers where required, specialist builders, approved submissions and actual ground conditions.
For the ground information that comes first, read How Ground Investigation Works Before HDB Construction. For temporary construction-stage support more broadly, see How Temporary Works Keep HDB Construction Stable. For the full housing system, return to How HDB Works in Singapore.
Excavation changes the problem before the building changes shape
Before excavation, the soil occupies a continuous volume. Adjacent ground is supported partly by the soil beside it. Once material is removed, that geometry changes. The face of the excavation now needs another way to remain stable and to limit movement.
Imagine cutting a slice out of a stack of books that had previously leaned against one another. The books beside the gap now experience a different condition because one source of lateral restraint has disappeared. Soil behaviour is much more complex, but the analogy shows why excavation creates a new structural state.
The risk is not only collapse into the excavation. Even smaller ground movements can matter in a dense city. A neighbouring building can settle, a road can deform, a buried utility can be strained, or a drain can lose alignment.
Singapore’s Building Control Regulations require building works to avoid settlement or movement that may impair nearby premises. BCA’s current ERSS framework is built around that public-safety obligation.
The retaining wall is one part of the ERSS, not the whole ERSS
People often see a line of sheet piles, diaphragm walls or other retaining elements and call that “the retaining system”. The wall is important, but the entire system includes the forces and supports that keep the wall within its intended behaviour.
A retaining wall may need to resist pressure from soil and water while spanning between supports. Those supports might come from:
- internal struts;
- walers that distribute forces along the wall;
- ground anchors where permitted and appropriate;
- temporary or permanent floor slabs;
- other designed bracing systems.
The correct configuration depends on excavation geometry, ground conditions, nearby property, utilities, construction access and the permanent building design.
This is why a photograph of one wall tells very little about whether the ERSS is adequate. The wall’s material, embedment, support conditions, groundwater, sequence and monitoring all matter.
Struts do not hold back soil directly
A strut usually receives force from another part of the retaining arrangement rather than contacting the full soil face itself. Walers or similar elements can collect reactions from the wall and transfer them into struts, which then carry compression across the excavation to another support.
This creates a chain:
GROUND → RETAINING WALL → WALER OR CONNECTION → STRUT → OPPOSING SUPPORT → GROUND.
The exact load path varies by system. The educational point is that temporary excavation support is structural. Forces must continue somewhere. A member that looks oversized or inconvenient may be carrying a consequence that is not visible from the walkway.
This is also why removing or relocating a strut to improve access cannot be treated as an ordinary housekeeping decision. It can change the entire load path. Project-specific engineering review is required before any such change.
Walers turn a local support into a distributed support
A retaining wall experiences forces along a length, while a strut or anchor connects at particular locations. A waler helps distribute the wall reaction into those discrete supports.
Think of pressing a hand against a wide flexible board. One finger produces a local reaction. A stiff bar behind the board can spread that reaction across a wider width. Again, the analogy is simplified, but it explains why an apparently horizontal beam can be central to the excavation system.
The joints between walers, struts and walls also matter. A system built from strong members can still perform poorly if the interfaces are unable to transfer the intended forces.
The same principle appeared earlier in How Reinforced Concrete Works in HDB Buildings: components become a structure through reliable force transfer.
Anchors move part of the support outside the excavation
Ground anchors can support a retaining system by transferring tensile forces into suitable ground beyond the wall. Whether they are feasible depends on geology, site boundaries, surrounding structures, rights and the approved design.
The public explanation should stop there. Anchor length, inclination, bond-zone design, stressing force and testing requirements are specialist engineering matters and should not be copied from another project.
The important contrast is spatial. Internal struts occupy the excavation. Anchors can free internal working space but require reliable ground outside the wall and may face property or utility constraints.
BCA’s 2026 technology-alliance material notes that conventional braced ERSS can require complex wall, waler and strut systems and describes an innovative post-tensioned waler arrangement that reduced intermediate supports on one project. The lesson is not that one proprietary system should be copied. It is that excavation support is an active field of engineering optimisation within safety constraints.
The deeper the excavation, the more the surrounding city matters
An excavation in open land has different constraints from one beside occupied buildings, roads, MRT infrastructure, utilities or drains. As the works become deeper and more constrained, geotechnical impact assessment becomes more important.
BCA’s current framework includes performance-based impact assessment for ERSS and tunnelling works. The framework is concerned with potential effects on adjacent buildings and requires projects to assess settlement or movement risks in a structured way.
This is an important conceptual shift. The excavation should not be judged only by whether its own wall remains standing. It also has to be judged by what the work does to the environment supporting neighbouring assets.
A safe hole inside the hoarding can still be an unacceptable excavation if it creates damaging movement outside the hoarding.
Groundwater changes the forces and the movement pathway
Water in the ground affects excavation pressure, seepage, stability and nearby settlement. Lowering groundwater inside an excavation can also change hydraulic conditions outside it.
Singapore’s Building Control Regulations require adequate drainage for excavations and require provisions where water-table lowering could damage surrounding structures, roads or property. BCA’s current ERSS guidance separately links an observational-method framework for groundwater control in deep excavation.
The groundwater problem is important enough to own a separate article rather than being hidden inside this one. Read How Groundwater Control Works During HDB Excavation for that mechanism.
Here, the boundary is simple: an ERSS cannot be understood from soil pressure alone if water materially affects the excavation.
The excavation sequence is part of the structural design
A deep excavation is not normally created instantly to full depth. Work proceeds through stages, and support systems become active as the excavation state changes.
That means the structure must be adequate not only in the final excavated configuration but through the intermediate states that occur on the way there.
Imagine a hypothetical retaining wall whose final design includes several levels of support. Before a lower support is installed, the upper wall experiences a different structural condition from the completed bracing arrangement. Sequence therefore changes demand.
This is why an excavation method should not be simplified into “dig, then brace”. The exact sequence belongs to the approved construction method and specialist engineering. The principle is that support and excavation depth evolve together.
Temporary support can later hand its job to the permanent building
In some basement or substructure projects, permanent slabs or structural elements eventually contribute to retaining or bracing functions that temporary members carried earlier.
The transfer must be planned. A permanent slab cannot be assumed to carry a temporary support’s action before the slab, connections and surrounding structure have reached the required state.
This is exactly the principle developed in How Temporary Works Keep HDB Construction Stable: removal is a transfer of responsibility, not the disappearance of force.
A good sequence therefore identifies not only when a strut is installed, but which verified construction state allows it to be removed.
Ground loss is a different failure pathway from wall bending
An excavation can experience problems even if the main retaining members are not grossly overstressed. Soil or water can move through gaps, joints or defects in ways that lead to local loss of ground.
BCA’s current ERSS page includes specific guidance on maintaining sheet-pile-wall integrity to prevent ground loss and on mitigating utility gaps within ERSS systems. That detail reveals how local discontinuities can matter to a much larger excavation.
The public lesson is that an ERSS is not judged only by the nominal strength of its wall. Continuity matters. Water paths matter. Interfaces around utilities matter.
One small unplanned route through the retaining boundary can become a mechanism for material movement.
Utilities make the retaining boundary imperfect by necessity
Urban sites contain pipes, cables, drains and other services. An ERSS may need to coexist with existing or diverted infrastructure. That creates geometric interfaces where a perfectly continuous retaining wall may be difficult.
BCA’s current guidance includes specific requirements around utility gaps because the interface deserves separate attention. The design must preserve the necessary service while preventing the gap from becoming an uncontrolled path for water or soil movement.
This is a good example of system design under competing constraints. The excavation wants continuity. The city wants its utilities. Engineering has to create a controlled interface rather than pretending one need can simply be deleted.
The same reasoning applies throughout HDB construction: the best technical solution is often not the one that optimises one subsystem in isolation, but the one that keeps several essential systems compatible.
Monitoring turns excavation behaviour into evidence
Deep excavation design predicts movement. Monitoring measures selected aspects of what the ground and structures are actually doing while work proceeds.
Depending on the project, instruments can monitor wall movement, ground settlement, groundwater conditions or neighbouring structures. The specific instrumentation and trigger criteria are project-specific.
BCA’s observational-method framework allows projects adopting that method to compare observed behaviour with predefined expectations and respond within an engineered decision framework.
This is not “watching the wall and hoping”. It is a structured feedback system. The separate article How Construction Monitoring Protects Buildings Around HDB Sites owns that evidence loop.
The Observational Method is not permission to design casually
BCA’s framework on the Observational Method was developed to permit controlled optimisation where actual performance is better than conservative assumptions, without compromising safety.
That requires advance planning. Expected behaviour, monitoring, alternative actions and decision responsibilities must exist before the project depends on them.
It would be wrong to describe the method as “dig first and decide later”. The method is almost the opposite: decide in advance how measured behaviour will govern later choices.
The wider principle is powerful. Flexibility is safest when the boundaries of that flexibility were engineered before the need to use them.
Excavation creates a logistics problem inside the bracing system
Struts and walers occupy space. Excavators, trucks, cranes, pumps, workers and materials need access to the same site.
A highly braced excavation can therefore be structurally safe and operationally difficult. Modern ERSS innovation often tries to preserve safety while creating more open working space or simpler installation.
This is where structural engineering meets construction logistics. A support member cannot simply be removed because a truck needs to pass. Instead, the construction method should account for both the required support and the practical flow of work.
Productivity becomes an ERSS design input, but it remains subordinate to the safety gates that keep the excavation controlled.
Neighbouring buildings need a before-state and a during-state
BCA requires pre-construction surveys before demolition, piling, foundation, tunnelling or site-formation works to establish the condition of nearby existing buildings and structures.
The purpose is evidential. If movement or damage is later alleged, the project should not have to rely on memory about what existed before work began.
During construction, monitoring can then record whether selected points or instruments are changing as the excavation progresses.
These two layers answer different questions. The pre-construction survey establishes condition. Monitoring establishes change over time.
Failure mode: assuming a wall that is still standing is performing acceptably
An ERSS can remain upright while movement outside the desired range develops.
Performance includes deformation and impact on surroundings, not only collapse prevention.
Failure mode: treating every temporary member as removable inconvenience
A strut or brace may obstruct work because it is doing useful work. Removal should follow the approved transfer of load, not the preference of the next trade.
Failure mode: separating groundwater from ERSS design
Water changes pressure, seepage and ground movement. A retaining system that ignores the relevant groundwater condition is an incomplete model of the excavation.
Failure mode: monitoring without a response plan
A sensor value becomes useful only if the project knows what the observation means, who reviews it and what action follows when a trigger is reached.
Failure mode: copying a successful ERSS from another site
The neighbouring buildings, ground, groundwater, excavation geometry, utilities and permanent structure may all be different.
ERSS design is site-specific precisely because the forces are created by the local conditions.
A better way to read an excavation
- What ground was present before excavation?
- What material has been removed?
- What now restrains the remaining ground?
- Where do wall reactions go?
- How is groundwater being managed?
- What nearby buildings, roads or utilities could be affected?
- What monitoring evidence is being collected?
- What verified condition allows the next excavation or support transition?
These questions do not design the excavation. They reveal why a safe excavation requires more information than “the wall looks strong”.
Follow one hypothetical excavation downward
The project begins with ground investigation and design. Retaining elements are constructed according to the approved system.
Excavation begins. As soil is removed, the wall begins carrying a different lateral condition. The first designed support level becomes active.
Monitoring records wall and surrounding response. Groundwater controls operate as required. The team compares actual behaviour with the design assumptions.
Excavation proceeds to another stage only when the appropriate support and verification are in place. Additional bracing or permanent structural elements are introduced according to the project sequence.
Later, permanent substructure work takes over functions that temporary members once provided. Those temporary members are removed only through the approved transition.
The finished resident sees a basement, car park, plant room or foundation structure. The temporary city of walls, struts, instruments and working platforms has disappeared.
Its disappearance is evidence of completed work, not evidence that it was unimportant.
The deeper construction principle
Excavation is the act of removing support from the ground in order to create space. ERSS engineering replaces that lost support with a controlled structural system while construction proceeds.
The system succeeds when ground movement remains within the designed and monitored response, surrounding assets remain protected, groundwater remains controlled, and temporary support can be handed safely to the permanent works.
The hole is not the product. The controlled change in ground state is the engineering achievement.
The deepest answer
Earth-retaining systems work on HDB construction sites by creating a temporary structural boundary between the ground that must remain and the space that construction needs to remove.
Walls retain.
Walers distribute.
Struts or anchors restrain.
Groundwater controls manage another source of pressure and movement.
Monitoring compares the real excavation with the predicted one.
Construction sequencing ensures the support system changes only when another verified state is ready to take over.
A deep excavation is therefore not empty space. It is a carefully maintained structural condition inside a living city.
Sources and continuing through the HDB construction sequence
Singapore sources checked on 5 September 2026 include BCA’s Guidelines for ST Plan Applications, its Observational Method framework, performance-based impact-assessment framework, and the 2026 DESA engineering examples.
Continue with How Groundwater Control Works During HDB Excavation and How Construction Monitoring Protects Buildings Around HDB Sites.