Where a walk-on unit makes sense, and where it does not

Where it earns its cost

A terrace over a habitable room. The classic case. The room below would have no daylight and the terrace cannot be given up.

A light well that is also a route. Where the only path to a garden crosses the well serving a basement.

Basement conversions. Where the alternative is a habitable room with no natural light at all.

A pavement or vault light, occasionally, reinstating an original feature in modern glass.

Where it does not

Anywhere the surface above is a roof rather than a floor.

We are asked for walk-on units fairly regularly by people who do not need one, usually because the flush appearance appeals. A flush glazed standard rooflight looks very similar from below, delivers the same light, and costs a fraction.

The other frequent case is “somebody might need to stand on it to clean the windows above”. That is a real requirement and the right answer is usually a pole system or a walkable margin beside the rooflight, not a walk-on panel.

The honest question

Will anyone genuinely walk on this, and how often? If the answer is once a year, that is not a walk-on case.

Half measures

On a large terrace, a walk-on panel over the part that needs light with standard construction elsewhere is usually better than a large glazed area. Light distribution does not improve much past a certain size and the cost keeps climbing.

What we do

Say which case you are in. Selling the expensive version to somebody who needs the cheap one is not a business worth building.

One situation it is for

A terrace, balcony or roof surface that people actually use, sitting over habitable space that has no other route for daylight.

There, walk on glazing is the answer and it is worth what it costs. The alternative is a dark room below, and no amount of clever artificial lighting substitutes for daylight in a space people occupy.

Everywhere else, one of four cheaper answers does the job better, and it is worth working through them before pricing structural glass.

Alternative 1: a non fragile rooflight

The right answer for most flat roof rooflights, and the one people most often skip past.

Non fragile is a safety specification, tested by dropping a weighted bag onto the unit. It means a person falling onto it will not go through. It is a life safety requirement rather than a traffic rating, and it costs a fraction of walk on glazing.

Where a roof is not a terrace and not a route, and the only foot traffic is somebody clearing gutters twice a year, the risk to manage is a fall rather than deliberate walking. Non fragile addresses exactly that.

It does not mean you can use it as a step. It means it will not kill you if you fall on it.

Alternative 2: moving the glazing out of the route

The one worth half an hour with the plan, and the one that most often saves the most money.

A terrace rarely needs glazing across its whole width. Putting the panel at the edge, or in a part of the terrace that is not the through route, turns an expensive engineered installation into an ordinary non fragile rooflight with a solid surface around it.

The room below still gets its daylight. The difference is only where the light lands, and that is frequently negotiable.

Alternative 3: lighting the room from somewhere else

Ask whether the room below could be lit from a different part of the roof entirely, or from a wall, or from a lightwell.

Where the answer is yes, a standard fixed rooflight does the same job at a quarter of the cost, with no structural work, no crane, and no perimeter drainage channel to maintain for the next twenty five years.

Alternative 4: access planning

Where the requirement is genuinely maintenance access rather than a floor, that is a question about how the maintenance is done rather than what the glazing is rated for.

A roof visited twice a year for gutter clearance needs a non fragile unit and a safe way to work around it: a defined route, edge protection where required, and a rooflight that does not have to be crossed.

Where none of the alternatives applies

Then walk on glazing is the right product, and the things to be clear about before committing are these.

The structure. A walk on unit is heavy before you add the people on it, and the opening has to carry both to a stated imposed load with tighter deflection limits than a roof would need, because glass has to sit on a support that does not move.

On a new build that is designed in and costs relatively little. On a retrofit it means an engineer, frequently steel, and it is usually the largest single line on the quotation.

Which way the joists run relative to the opening changes the answer more than the size of the unit does.

Getting it there. A large pane weighs several hundred kilograms and cannot be manhandled. Crane or glass robot access and a route to the position have to be established before anything is ordered.

A property with no side access and no frontage close enough for a lift is a genuine constraint on what size is realistic, and it is not something to discover after the glass is made.

The drainage. A walk on unit sits flush with the surrounding surface, because a raised kerb in the middle of a terrace is a trip hazard. That removes the usual 150mm defence against standing water.

Everything then depends on the surrounding surface falling away from the unit and a perimeter channel discharging into a working outlet. A blocked channel is a pond sitting against the seal, and that is the commonest cause of trouble on these installations.

The thermal detail. A flush unit over heated space is a cold bridge at the perimeter unless it is insulated and the vapour control layer is continuous and sealed to the frame. Get that wrong and there is condensation on the ceiling below, in a position that is awkward to correct.

What the loading actually assumes

A domestic terrace specification assumes people walking across: a distributed load plus a point load representing a person, applied and removed.

It does not assume a heavy planter, a water feature or a stone bench standing permanently in one place. A sustained point load in a fixed position is a different case and it has to be designed for rather than assumed.

It does not assume a commercial or public route, which carries a higher imposed load and different safety requirements.

And it certainly does not assume a vehicle. Wheel loads are an order of magnitude higher, applied dynamically and repeatedly, with turning as well as rolling. A driveway is an entirely different engineered product, and a pedestrian walk on unit must never be used where a vehicle will cross it, even occasionally, even a light one.

What you get, and what you give up

You get daylight into a room that would otherwise have none, and a terrace surface that reads as a designed element rather than a compromise.

You give up clarity. The anti slip treatment on the top surface, whether a fired ceramic frit or an etched pattern, scatters light. Add the thickness of a laminated stack over a sealed cavity, typically 50 to 70mm in total, and every interface reduces it further.

The result reads as a luminous panel rather than a window. You will see daylight, shapes and movement. You will not get a sharp view of the sky.

If a view is what you want, this is the wrong product and a standard rooflight elsewhere is the right one.

Ongoing commitment

More maintenance than any other rooflight, because it is being walked on.

The perimeter drainage channel needs clearing twice a year on a plot with trees and annually otherwise, and the outlet it discharges into needs checking. That is the item that matters.

The surface is cleaned with water and a soft brush: not wire brushes, not scouring pads, and not a pressure washer at close range, all of which damage the anti slip layer and drive water into the perimeter seal.

And grit needs keeping off it, because grit underfoot is what actually abrades an anti slip surface. A doormat at the door onto the terrace does more for the life of the unit than anything else.

What we will not do

Fit a standard rooflight where somebody intends to walk on it.

Certify an existing unit as walk on because it looks robust. The specification is either there in writing from the manufacturer for the loading it will actually see, or it is not.

Fit a walk on unit into an opening that has not been checked structurally.

Or sell walk on glazing to somebody whose situation calls for non fragile. The difference in price is substantial and there is no reason to spend it, and we would rather tell you which of the five situations you are in than default to the expensive one.

The test that settles most cases

Is the surface above genuinely going to be used?

Where a roof is becoming a terrace people will sit on, or a courtyard people will cross daily, the walk on unit lets one square metre be floor above and window below. On a site where outdoor space is scarce, that is a real gain rather than an indulgence.

Where the roof is accessed twice a year for maintenance, it is not, and a standard rooflight with an upstand does the daylight job for a fraction of the cost.

The four situations that justify it

A roof terrace over a habitable room, where the room below has no other route to daylight.

A basement or light well, where the glass is frequently the only natural light the space will ever get and the courtyard floor is the only place it can go.

A side return or courtyard where a raised rooflight would block the only route to the back of the plot.

A link between two parts of a building, where a walk on strip lights the space below without interrupting circulation above.

The reasons that do not justify it

Being able to stand on it to clean it. A water fed pole reaches most domestic rooflights from the ground and costs a tiny fraction of the difference.

Appearance alone, where the roof is not used. A flush finish costs meaningfully more than an upstand because the waterproofing and drainage detailing is far more demanding, and on a roof nobody crosses that money buys nothing.

Future proofing. Specifying a walk on unit against the possibility of a terrace one day, without designing the structure for it, gets you the cost without the capability.

What to weigh before committing

The engineering and steel come first and are frequently the largest line on the job. The drainage detailing is demanding, particularly flush. Lead times are weeks because every unit is bespoke. And access for lifting large panes may be the constraint that decides the design.

The middle option

Where the case is genuine but the budget is not, an upstand rather than a flush finish keeps the walk on capability and removes most of the drainage complexity.

It is a raised edge in the terrace, which is a compromise, and it is honest rather than a failure.