A Grade-II listed roof truss was sagging, but the brief was absolute: keep every original timber visible and the intervention reversible. That single constraint - reversibility, with nothing hidden - is what makes conservation engineering a different discipline from new design, and it shaped every decision that followed.
The constraint that ruled out the obvious answers
The usual ways to strengthen a tired timber truss were off the table. No member could be removed, so replacing the worst chords was out. Nothing could be hidden, so plating over the timbers - bonding or bolting steel onto the visible faces - was out too. And the work had to be reversible, meaning a future generation should be able to undo it and return the truss to its original state, which rules out anything glued, resin-injected or permanently embedded in a way that cannot be taken back.
These are not arbitrary heritage fussiness. The value of a listed structure is partly in its authentic fabric and its legibility; an intervention that erases or conceals the original defeats the point of keeping the building at all.
The solution: hidden steel, visible timber
The answer was to put the steel where it does the structural work but stays out of sight, and leave the timber to do the visible work:
- Slim steel flitches slid into routed slots cut into the timber members. The steel sits within the section, carrying load, while the original timber faces remain on view exactly as before.
- Countersunk, bolted fixings that can be undone - the connection is mechanical and reversible, not a permanent bond.
The division of labour is clean: the timber carries the eye, the steel carries the load. A visitor sees an original timber truss; the structure quietly relies on the flitches within.
Why reversibility drives the detailing
Every choice followed from "could this be undone later?" Routed slots can be re-filled or the flitch withdrawn; bolted fixings can be removed; nothing relies on adhesive that would damage the timber to reverse. Compare that with the tempting alternatives - resin-bonded plates, glued-in rods - which are structurally excellent but effectively permanent. In a new building they would be fine; in a listed one they fail the brief even if they pass the calculation.
The deeper lesson
Conservation engineering inverts a habit of new design, where you add whatever solves the problem most efficiently:
Conservation engineering is as much about what you choose not to do as what you add.
The skill is in restraint - finding the minimum, reversible intervention that meets the structural need while leaving the heritage value intact. The most elegant conservation solutions are often the least visible, doing their work without announcing themselves and without closing off future options.
Three principles travel from this to any heritage job:
- Reversibility first. Prefer mechanical, removable connections over permanent bonds wherever the structure allows.
- Hide the new, reveal the old. Put the modern material where it works structurally but does not compete with the original fabric.
- Minimum intervention. The right answer is usually the smallest one that works, not the strongest.
The related retrofit case study shows the same assess-then-intervene discipline on a concrete structure, where the constraint was keeping the building in service rather than preserving its fabric.