Maintenance Access Is an Envelope Design Decision

On most of the envelopes described in the records, the part that ages first and needs attention is not the part that dominates the elevation. On a precast facade the joints, not the faces, are the maintained element, and they are reached from outside at height. On curtain walling the gaskets, seals and glazing unit edge seals age at different rates from the metal around them, so parts of the facade have a shorter working life than others.

The consequence is that a facade can be chosen for the durability of its surface and maintained through something else entirely. Whether joint renewal is a planned operation or one that waits until water appears indoors is described as depending on whether an access arrangement exists, which is a design outcome rather than an operational one.

A trafficked roof puts the working parts of the assembly under the surface people walk on. The outlets serve the concealed level and are reached through the surfacing above, and whether they can be inspected and cleared without dismantling the terrace is decided when the paving layout is set out. The record is blunt about the alternative: a terrace laid as a continuous bonded field over its outlets has been designed to be opened up destructively.

The same logic appears wherever a concept depends on one small component. An attenuation roof holds water back through a deliberately restricted outlet, and because that restriction is the element the whole concept depends on and the one most easily blocked by leaves, silt and debris, access to it is described as part of the design. A roof that hides it has designed out its own upkeep.

On some build-ups the surfacing is doing mechanical work as well as covering. In an inverted roof the ballast or paving is the restraint holding the stack down against wind and against the buoyancy of insulation surrounded by water, so it cannot be reduced, thinned or partly removed on aesthetic grounds without changing the roof. The record notes that removing surfacing from part of a roof to carry out unrelated work leaves the layers beneath it restrained by nothing.

That turns an access decision into a restraint decision. The terrace record makes the same point from the other side: lifting paving for maintenance over an inverted build-up removes restraint as well as access, which is not true of a terrace over a warm deck. Which areas are intended to be openable, and how restraint is restored afterwards, belongs in the design rather than in later improvisation.

Several of the assemblies described are explicitly unrepairable in place. A bonded tile or stone facade cannot be taken off and put back, and the record places the ability to detect units that have lost bond before they come off, together with access to a wall above ground for that purpose, in the design discussion rather than in the first maintenance year. A composite panel cannot be repaired layer by layer because the layers are bonded, so replacement usually means disturbing neighbouring panels.

Others are concealed for life. The waterproofing in an inverted roof cannot be seen, tested by eye or repaired without removing everything over the affected area. The components crossing a structural thermal break are concealed and cannot be inspected once the connection is complete. Where that is the case, the records treat the record of what was installed, and of any verification carried out before covering, as the only evidence a later owner or surveyor will have.

Where a facade contains something that grows or operates, maintenance stops being optional. The vegetated facade record describes planting, irrigation and drainage as needing attention that cannot be deferred without visible loss, and states that whether the whole face can be reached safely, and how, decides more about the facade's future than the choice of planting does. It describes such a facade as usually failing as a maintenance and access problem that first shows itself as a horticultural one.

An adjustable shading system adds actuators, controls and moving parts on the outside of the building, in the least accessible position it could occupy. The record treats the access strategy as part of deciding whether to have one at all, and notes that how each element will be reached determines whether it stays as designed or becomes a permanent fixture nobody touches.

Inspection regimes inherit the same limits. The parapet record notes that the outer face and the top of a parapet can generally only be reached from outside the building at height, while the inner face is reached from the roof, so an inspection regime covering only the roof side looks at part of the element. The eaves record notes that everything at an eaves is reached from outside and above, and that the components most likely to need attention, the gutter and fascia and guard, are the ones hardest to get to.

That is why the records ask how a part of the building will be reached once the scaffold has been struck. The question is not whether an inspection is intended but whether the position it needs to be carried out from will exist.

Read together, the records point at a small number of decisions taken early: what access equipment the building will accommodate, which surfaces lift and how restraint is restored, which components will be renewed on a cycle and from where, which layers are accepted as inspection-only, and whether matching replacement parts will still exist. Each of them is cheap while the elevation is still a drawing and expensive afterwards.

They also need an owner. The records repeatedly locate access at the boundary between packages: the facade designer, the roof designer, the drainage engineer and the landscape designer each stop at the same line. Whether anyone is appointed to answer the access question for the envelope as a whole is worth establishing rather than assuming.