Where Services Are Allowed to Run in an Insulated Envelope

A service zone is not a way of hiding cables. It is a space in front of the nominated control layer within which everyday electrical and plumbing work can be carried out without that layer being cut. The records for the framed and panel walls make this explicit: the purpose of the zone is that a fitting never needs to reach the layer behind it, which is why its removal changes the concept of the wall rather than only its finish.

That distinction matters because the two things look identical once decorated. A lining on battens over a sealed layer and a lining on battens over nothing present the same surface to the room. What differs is whether a later screw, a chased cable or a deeper back box reaches something the element depends on, and whether anyone will know that it did.

In a timber frame wall the control layer is a nominated sheet sealed on the warm side of the insulation, and its value comes entirely from its joints, edges and penetrations being closed. The record describes deleting the zone to gain floor area as quietly transferring the airtightness of the whole wall onto workmanship at each fitting. Nothing announces that transfer, because each individual seal is made correctly by someone doing their job as instructed.

The light steel frame record adds a consequence particular to a conducting frame. Where services pass straight through the control layer with no zone in front to take them, the layer is perforated at every fitting, and because the sections are metal and the sheathing is pressed against them, any moisture that arrives meets a surface which neither absorbs nor buffers it. It stays as liquid exactly where it forms.

The mass timber record is the clearest case in the family. Where the panel face is nominated as the air barrier, each socket, each pipe and each fixing driven into the panel is a penetration of the building's air control line and of the structure at the same moment. The record states plainly that the only arrangement avoiding that is one in which services are held in front of the panel and never reach it.

Two further consequences follow that no entity page can generalise. A hole here cannot be moved afterwards, so grouping penetrations and planning them before manufacture is a different exercise from sealing them where they happen to land. And where the panel's inner face is being left exposed, the decision about services, fixings and protection is taken along with the decision about the finish rather than after it.

In an internal insulation retrofit the same rule applies with an additional difficulty. The record puts it directly: services placed in front of the control layer keep it whole, and services chased into the wall behind it undo the purpose of building the lining at all. The problem is that the finished lining looks like an ordinary wall, so later occupants drill, fix and chase into it without knowing what lies behind.

That is why the record treats a drawing of where the layer runs and where fixings may safely be made as part of the assembly rather than as paperwork about it, and why where that drawing is held decides whether it is ever consulted. An insulated lining and an ordinary dry lining cannot be treated the same way later, because only one of them has a layer that ordinary fixing work can breach without any visible sign.

The roof records carry the same argument into the horizontal plane, and they are blunter about it. In a cold deck flat roof the ceiling is doing two jobs: it is a finish, and it is the plane limiting how much moist air can reach the void above. A recessed fitting, an unsealed hatch or a cable run opens that plane at the warmest and most humid face of the assembly, and warm air then rises into the bays under its own buoyancy.

A fitting recessed into that ceiling does something else as well. It is set into a bay whose depth is already shared between insulation and air space, so it displaces the ventilation path as well as opening the plane. The record's own conclusion is that where an item can be kept out of the roof zone altogether, that is a simpler answer than closing it well.

The ventilated pitched roof record identifies extract ducts as the sharpest case of all, because a duct terminating in the roof space delivers moist air into an unheated volume by design rather than by leakage. Where every mechanical extract in a building discharges is a question the record puts to the designer, not to the installer on the day.

A zone that is too shallow for the fittings actually planned fails in a particular way. The records for the framed wall and the internal lining both describe boxes and pipes beginning to press against or pass through the layer behind once the depth is squeezed, with the puncture concealed by the finish. The zone is then present on the drawing and absent in effect, which is the hardest kind of defect to find later.

This is why the published design questions ask not whether a service zone exists but whether it is deep enough for the fittings actually planned, and what happens if a fitting needs more depth than it offers. Both are questions for the designer and the installer together, settled before first fix rather than discovered during it.

Every one of these records ends at the same place: the arrangement disappears behind a lining or a ceiling and cannot be inspected again. What survives is a record of where the control layer runs, where fixings may be made and what was installed. That is what turns a later air test result, a damp investigation or an ordinary shelf into something that can be judged rather than guessed at.

It is also the only realistic protection against the ordinary work of occupation. Nobody hanging a cabinet, running a new cable or fitting a new light is trying to breach a control layer, and none of them can see it. What they can consult is a drawing, if one exists and if they are told where it is.