Layers, Connections & Reversibility

The shearing-layers idea, popularised by Stewart Brand from earlier work by Frank Duffy, holds that a building is best understood not as one object but as a set of layers that change at very different rates - and that this difference in rate is the key to designing buildings that last and adapt. The classic list runs from slowest to fastest, shown opposite. Site is effectively eternal. Structure - foundations and frame - lasts from several decades to centuries and is the most expensive and disruptive layer to change. Skin - the external envelope - is renewed perhaps every twenty to thirty years as it weathers or as performance standards rise. Services - wiring, plumbing, heating, ventilation, lifts - wear out or become obsolete in roughly seven to fifteen years. Space plan - the internal layout, partitions, doors, ceilings, finishes - changes every few years to a few decades as occupants and uses change. And stuff - furniture, equipment, everything not fixed - churns constantly, over days and months.

The insight that matters for circular design is what happens when layers of very different speeds are bound together. If you fix a fast layer rigidly to a slow one - services cast into structure, finishes bonded to the frame, partitions built as structural walls - then every time the fast layer needs to change (which is often) you are forced to attack the slow layer (which is expensive, wasteful and life-shortening). A ten-year service life trapped inside a hundred-year structure means either the structure is damaged every decade to reach the services, or the services are left to rot and the whole building becomes obsolete. Conversely, if each layer is kept separate and connected so it can be changed on its own clock, the building can renew its fast layers many times over the long life of its slow ones - which is exactly what adaptability and long life require.

So shearing layers is not an abstract taxonomy; it is a design instruction. Keep the layers distinct. Never entangle a fast layer with a slow one. Make each layer accessible and replaceable without destroying the layers around it. Do that, and a building becomes a set of loosely coupled systems that can each be maintained, upgraded, adapted and eventually recovered independently - the physical structure of a circular building. Entangle the layers, and you get a monolith that must be demolished whole the moment any one layer fails or changes.

Turning the shearing-layers insight into design means treating the layers as a first-order organising principle - deciding, for every element, which layer it belongs to and detailing it to change on that layer's clock, independently of the others. The governing rule is a hierarchy of longevity: the slower a layer, the more permanent and generous it can be; the faster a layer, the more accessible, demountable and replaceable it must be. You design the structure to last and to carry an unknown future (the adaptability lesson's 'generous structure'), and you design the services and space plan to be swapped out easily and often, and you make sure the fast layers are attached to the slow ones in ways that let them be removed without damage.

Concretely, this reshapes many everyday decisions. Services should run in dedicated, accessible zones - raised floors, service voids, dropped ceilings, clearly defined risers - not be chased into structural walls or cast into slabs, so they can be maintained and replaced without touching the structure. Partitions should be demountable systems attached to floor and soffit reversibly, not built as permanent walls, so the space plan can change without demolition. Cladding should be mechanically fixed and accessible so the skin can be renewed or upgraded without disturbing the frame. Finishes should be fixed so they lift off rather than being bonded to the layer beneath. Each of these keeps a faster layer independent of the slower one it sits against.

This is where the previous two lessons converge. Designing layers to be independently replaceable IS designing for adaptability (the fast layers can change, so the building can change use) and IS designing for disassembly (each layer can be removed whole, so its materials can be recovered) - the layers idea is the organising frame that ties adaptability and DfD together. It also has a clear interior dimension: fit-out is the fast space-plan-and-stuff end of the spectrum, so designing interiors as independent, demountable layers - rather than gluing, bonding and building-in - is how interior designers apply shearing layers directly, and it is where the churn, and therefore the waste, is greatest. The judgement, as always, is to spend permanence and generosity on the slow layers and reversibility and accessibility on the fast ones - and to remember that how the structural layer is sized and loaded remains an engineering decision for the specialists and the codes.

Everything in this lesson, and much of the module, comes down to a single humble detail: the connection between one layer and the next. It is the smallest thing in the building and it governs the largest outcomes, because the interface between a fast layer and a slow layer is exactly where reversibility is won or lost. Get that junction right - make it demountable - and the fast layer can be changed a dozen times over the slow layer's life, and both can be recovered at the end. Get it wrong - make it permanent - and the two layers are welded into a single fate: to change or recover one, you must destroy the other.

The figure opposite shows the choice at its starkest. On one side, services are cast into the structure: a ten-year layer buried inside a hundred-year one, so that replacing the services means breaking the structure, and a sound frame is compromised by the obsolescence of the pipes inside it. On the other, the services run in an accessible zone, connected to the structure by demountable fixings: the fast layer can be swapped out entirely without the structure being touched, and each layer lives out its own life. Same two layers, opposite outcomes - and the only difference is the connection between them. This is why circular designers obsess over interfaces: the junction between layers is the true unit of reversibility, more decisive than the choice of material in either layer.

The design rules follow directly. At every interface between layers, especially where their change-rates differ most, make the connection reversible and accessible: fix the fast layer to the slow one with bolts, screws, clips or dry joints that can be undone, in a place that can be reached, so the fast layer lifts away cleanly. Never bond, cast-in or bury the junction between a fast and a slow layer. Sequence the connections so the building can be taken apart in the reverse of assembly, fast layers first. And record the connections - what is fixed to what, and how it comes apart - in the documentation that becomes the material passport. As with all of DfD, honesty is required: some interfaces must be permanent for fire, weather or structural reasons, and whether a given connection can be both reversible and structurally safe is an engineering decision, to be made with a qualified structural engineer and the governing codes - never assumed for a load-bearing junction.

Module 2 has moved from the biggest, earliest moves to the smallest, most detailed, and the layers idea is what unifies them. Build nothing, build less, build long set the strategy; design for adaptability keeps the building wanted; design for disassembly makes its parts recoverable; and shearing layers, resolved at the connection, is the organising principle that makes adaptability and disassembly physically true - a building of distinct layers, each on its own clock, joined by reversible connections, is simultaneously long-lived, adaptable and a genuine material bank. If you remember one structuring idea from the module, make it this: design in layers, and win the layers at the connections.

The practical pay-off is cumulative. A layered, reversibly connected building can renew its skin and services several times over the life of its structure (long life), absorb new uses by re-planning its fast layers without demolition (adaptability), be repaired and upgraded piece by piece during its life (maintainability), and finally be taken apart with every layer's materials recovered whole (disassembly and the material bank). One organising decision - keep the layers distinct and connect them reversibly - delivers all four. For interiors, where the fast layers dominate and churn is relentless, applying it is both easiest and most valuable: a demountable, layered fit-out is recovered and reconfigured rather than skipped.

The honest limits close the module as they opened it. Designing in clean, reversibly connected layers can cost more up front, take more space (service zones, raised floors, generous voids), and demand better coordination and workmanship than a cheap monolith. Some interfaces must stay permanent for performance and safety, so layering is a disciplined bias, not an absolute. None of it guarantees that a building is ever actually adapted or disassembled - that still depends on economics, a deconstruction industry and a reuse market that remain immature (Modules 7 and 8). And every binding decision the module has touched - how lean or generous a structure can safely be, whether a connection can be reversible and safe, whether a recovered element can be reused, whether an adaptation or reuse is code-compliant - is deferred throughout to qualified structural engineers, certified testing, and the governing codes (the National Building Code of India and local regulations), with all figures illustrative and context-dependent. The designer's power is real and large: to organise the building into independent layers, connect them reversibly, and design for a long, adaptable, recoverable life - and then to hand the binding questions to the people whose job they are.