A sun tunnel goes through a roof covering, and on stone or slate that covering will not accept a standard flashing kit. What follows is about the tunnel: the aperture, the dome, the flashing detail and the sequence. The coverings themselves are set out elsewhere.
What this page assumes
That the covering is already understood, because it is covered in full elsewhere and repeating it here would help nobody.
Horsham stone, its graded courses, its weight, how the slabs are fixed and why no standard kit fits one, is set out on the Horsham stone slab roofs page. Slate roofs, their laying, their fragility and what a roof window needs from them, are on the slate roof page.
Read whichever applies first. This page is only what changes when the thing going through the roof is a sun tunnel dome rather than a rooflight.
The one advantage a tunnel has here
The aperture, and on a difficult covering it is a large advantage.
A rooflight needs an opening of several hundred millimetres in each direction, which on a stone or slate roof means disturbing a substantial number of units and forming a weathering detail around all four sides of it.
A sun tunnel needs a circular opening of a few hundred millimetres in total. On a covering where every unit disturbed is a risk and a cost, that difference is the whole argument.
Which is why tunnels get specified here
More often on these roofs than on ordinary tiled ones, and for reasons that have nothing to do with the room below.
Where a household would prefer a rooflight and the covering makes one difficult, expensive or unlikely to be permitted, a tunnel delivers useful daylight for a fraction of the intervention in the roof.
That is a legitimate reason to choose the smaller product, and it is worth saying plainly rather than presenting the tunnel as equivalent. It is not equivalent; it is achievable where the other is not.
The flashing problem, in tunnel terms
Standard tunnel kits are made for tiles and for slate of standard thickness and regular coursing. Neither describes a stone slab roof, and an old hand made slate roof is frequently outside what the kit assumes as well.
| Standard kit assumes | What these roofs give | |
|---|---|---|
| Unit thickness | Consistent, modest | Substantial and variable on stone |
| Coursing | Regular gauge | Graded on stone, variable on old slate |
| Surface | Flat or profiled to a pattern | Irregular, hand dressed |
| Fixing | Nailed to battens | Pegged or nailed, sometimes torched |
| Cutting a unit | Routine | Skilled, and often refused |
| Weathering method | Preformed skirt | Dressed lead |
Dressed lead, which is the answer
The same answer as for a rooflight on these roofs and it is easier at tunnel scale.
A lead slate, or a lead soaker detail dressed to the covering by hand, weathers a circular upstand into an irregular roof in a way no preformed skirt does. It follows the unit thicknesses, it takes the graded course changes, and it is a repairable detail rather than a sealed one.
At the size of a sun tunnel dome that is a modest piece of leadwork rather than a substantial one, which is another reason the tunnel is the more comfortable proposition on these roofs.

Choosing the dome for a lead detail
The product decision that follows from it.
A dome with a well defined upstand, or one supplied with a separate kerb, gives the lead something to dress against. A dome designed to sit tight to a tiled surface with an integral skirt is the wrong starting point, because the skirt is the part being replaced.
Which means the specification here starts from the flashing method and works back to the dome, rather than the usual way round.
Low profile domes
Where appearance decides, and on these roofs it frequently does.
A low profile or flat glazed tunnel top sits closer to the plane of the covering and is far less visible from ground level than a raised dome. On a listed building or in a conservation area, that difference can be what makes the proposal acceptable.
It costs a little light, because a raised dome collects from a wider part of the sky. Where visibility is the constraint, it is a trade worth making.
Where the dome goes on the roof plane
A siting decision with a heritage dimension as well as a technical one.
The tunnel can be offset within the roof space, so the dome does not have to sit directly above the diffuser. That freedom is worth using: a dome on a rear slope, out of the principal view, is a different proposition from one on the front.
On a graded stone roof there is a second consideration, which is where in the coursing it lands. That is a question for the roofer and it is worth asking before the position is fixed.
The roofer does the roof
The working arrangement on these jobs, and it is deliberate.
Opening and reinstating a Horsham stone or an old slate roof is specialist roofing work. The tunnel is our part; the covering is theirs. On these roofs we work with a roofer experienced in the covering rather than treating the opening as part of the fitting.
That means two trades on site and a sequence to agree, which is more organisation than an ordinary tunnel. It is also why these installations go right.
The sequence
- Survey the roof space, establish the run and the offset available.
- Agree the dome position on the roof with the roofer.
- Confirm consent where the building is listed or the area designated.
- Scaffold or access as the covering requires.
- Roofer strips carefully, sets aside units for reuse.
- Upstand set, lead dressed to it.
- Tunnel installed from the roof space downward.
- Ceiling opening formed, diffuser fitted and sealed.
- Roofer reinstates the covering around the detail.
Reinstating rather than replacing
The principle that governs the roofing side and it applies at tunnel scale too.
Units taken off come back on. On a stone roof they are irreplaceable in practice; on an old slate roof the existing slates match each other and a new one does not. Nothing goes in a skip until the job is finished.
At the size of a tunnel opening that is a small number of units, which makes it manageable in a way a rooflight opening on the same roof is not.

Cutting a stone slab or an old slate
Sometimes necessary and it is not routine work.
A circular opening rarely falls neatly between units. Where a unit must be cut, it is cut by someone who has done it before, dry or wet as the material requires, and with the acceptance that a proportion will break.
Which is an argument for setting the dome position to minimise cutting, and it is one of the things the roofer is asked at the survey rather than on the day.
Consent
Frequently required and it is straightforward to establish.
On a listed building, altering the roof covering needs listed building consent, and that is true whether the opening is for a rooflight or a tunnel. In a conservation area or inside the National Park, the visibility of the dome from public views is the question.
A tunnel is a considerably easier case to make than a rooflight because the intervention is smaller and it is reversible. That is worth saying explicitly in an application rather than leaving the officer to infer it.
Reversibility, as an argument
The point that carries weight on a listed building.
A tunnel opening is a small circular penetration through a covering, with no structural member cut at either end. Removing it means reinstating a few units and patching a ceiling.
Where a proposal is being judged on how much of the historic fabric it commits, that is a genuinely strong position and it is frequently what secures consent.
Underlay on an old roof
A practical matter that comes up as soon as the covering is off.
Old roofs frequently have no underlay, or a bitumen felt at the end of its life. Cutting into it around a new upstand can start a tear that runs.
The roofer decides what happens locally around the opening, and it is worth knowing before the day that a small area of repair or replacement may be part of the work.
Structure under a stone roof
Where the tunnel has an easier time than a rooflight, and it is worth being clear why.
A stone slab roof is carried on heavy timbers, frequently old and frequently irregular. A rooflight opening means cutting into them.
A tunnel of ordinary diameter passes between them, so nothing is cut and the structural question does not arise. Where the timbers are close set and a tube will not pass, a smaller diameter or a different position answers it, which is a specification change rather than a structural design.

Roof space conditions on an old building
Worth a survey note, because they differ from a modern loft.
Old roof spaces are frequently unboarded, poorly lit, dusty and irregular, with limited headroom at the eaves. That affects how the run is set and how long the installation takes.
It also affects the insulation and sealing detail, because these spaces are cold and often well ventilated, which makes insulating the tube along its run more important rather than less.
What it costs, in time
Not in money, which we do not publish, but in duration, which people ask about.
A tunnel on an ordinary tiled roof is generally a day. On stone or slate it is longer, because the roofing is careful work, there is access to arrange and two trades to coordinate.
It remains substantially less disruption than a rooflight on the same roof, and that is the comparison worth making.
Access on an old building
Frequently the item that decides how long the job takes.
A stone or old slate roof is not walked on. Reaching the dome position means scaffold, a tower, or a roof ladder set properly, and on a steeply pitched or brittle covering it means scaffold.
That is arranged around the roofer’s requirements rather than ours, and on a listed building the scaffold itself sometimes needs agreeing. It is worth establishing early, because it sets the date more than the work does.
Doing it during a re-roof
The best time to fit one of these, by a considerable margin.
Where a stone or slate roof is being stripped and relaid anyway, the covering is already off, the scaffold is already up, and the roofer is already on site. Adding a tunnel upstand and a lead detail at that point is a fraction of the work it is afterwards.
Which makes it worth asking any household on one of these buildings whether roofing work is planned, and worth advising them to wait where it is imminent.
When to say no
The cases where we advise against it.
A stone roof in poor condition, where opening it would start work nobody has budgeted for. A roof due for stripping and relaying within a few years, where the tunnel should wait and go in during that work. And a position where the dome would be prominent on a principal elevation of a listed building, where consent is unlikely and the application is not worth making.
In the second case the advice is simply to wait, because a tunnel installed during a re-roof is a far simpler and cheaper detail than one cut into a finished roof.
What we do
On stone and old slate we specify from the flashing method backwards: a dressed lead detail, a dome with an upstand it can work to, and a low profile top where visibility is the constraint. We work with a roofer who knows the covering, agree the dome position before anything is stripped, and set the offset in the roof space so the dome lands where the roof and the planning position want it rather than where the room does.
The small aperture is the reason a tunnel works on these roofs at all. Everything else follows from protecting the covering around it.
Thinking about this job? We survey across Worthing, West Sussex and the South Downs and give you a fixed price before any work starts, with a 10-year workmanship guarantee.