Overbanding or Joint Sealing: Which Does Your Surface Need?

Joint sealing and overbanding get quoted against each other as though they were two ways of doing one job. They are not. One fills a joint that was designed into the surface and designed to move. The other lays a band of material over the top of a crack. Overbanding or joint sealing is decided by what the feature in front of you actually is, not by budget or preference.

At Shepherd and Sons we have been cutting, sealing and repairing concrete and asphalt for more than 40 years, from our yard at Little Westerhill Farm near Maidstone in Kent. We are members of the Extruded Sealant Association. This article is a decision guide rather than two explainers bolted together. It covers how to tell a designed joint from a crack, what each treatment is physically able to do, the two traps that cost real money, and when overbanding is genuinely the right call.

Joint sealing puts sealant inside the joint. It sits on a backer rod, bonded to the two joint faces, finished flush with the surface or recessed below it. Its job is to accommodate movement while keeping water out.

Overbanding puts a band of bituminous material on the surface, usually hot applied, across the crack or joint, bonded to the pavement either side. Its job is to keep water out and to take the traffic wear. Its job is to keep water out and to take the traffic wear. It has a little give in it, and nowhere near enough to follow a moving joint.

That last line is the whole article in miniature. One treatment works with movement. The other works against water. Almost every wasted pound in this area comes from asking the second to do the first one's job.

Start here. Everything else follows from the answer, and you can usually settle it standing over the feature with a torch.

A designed joint is straight. It runs the full width or the full length of a bay, meets other joints squarely, and repeats at a regular spacing. It was sawn or formed when the slab went down. A crack does none of that. It wanders, it branches, it changes width along its length, and it pays no attention to the bay pattern.

• Straight, regular, full bay width. Almost certainly a designed joint.
• Wandering, branching, varying in width. A crack.
• Cutting diagonally across a bay corner. A crack, and usually a sign the slab has lost support there.
• Straight and regularly spaced, but in asphalt. A reflective crack, coming up from a joint in the concrete beneath. Treat it as a joint.
• Visibly wider in January than in July. It moves. That settles it, whatever else it looks like.

The last test is the one worth doing yourself. Measure the gap in cold weather, then again in warm. Concrete contracts as it cools, so a live joint is at its widest in winter. If the feature moves seasonally, it needs a sealant with extension. A band laid across the top will not survive it.

A concrete pavement is built in bays so that thermal movement has somewhere to go. A contraction joint is a deliberate weakened plane, sawn in so the slab cracks where you chose rather than where it likes. An expansion joint is a gap left so the slabs have somewhere to grow into. The sealant in it has one job above all others. It keeps water out while the joint opens and closes, year after year.

That demands extension capability, and it demands the correct installed detail. Grade comes first. Under BS EN 14188-1, which covers hot applied sealants, N1 is the elastic, high extension grade. N2 is the normal, low extension grade. Long bays, expansion joints and bridge approaches need N1. Short bays with genuinely modest movement can take N2. The two are set out side by side in the difference between N1 and N2 sealant.

Then the detail. A backer rod at the base of the slot sets the sealant depth and prevents three sided adhesion, which is a common cause of early failure. On trafficked joints the finished sealant sits recessed below the surface, so tyres cannot pull it out. Neither of those is available to a band sitting on top of the pavement. The method is set out in what is hot applied joint sealing.

This is the work we do most of. On the M20 between junctions 8 and 9 we reformed and resealed 23,304 linear metres of joint across roughly 20km of concrete carriageway. It ran over 20 night shifts, inside a 10pm to 3:30am possession. Nothing about that job would have been solved by a band over the top.

Overbanding has a reputation it half deserves and half does not. Laid properly, on the right defect, it is a legitimate treatment with a long service life. There are surfaces where it is clearly the better answer.

It belongs on asphalt. A cracked asphalt surface is letting water into the layers below. Water in the base is what turns a hairline crack into a pothole over a single winter. A band across the top stops that. It also takes the traffic wear that would otherwise fall on the sealant sitting in the crack.

It is quick, too. The compound is melted in a boiler, applied hot, and the surface comes back into service shortly afterwards. On a car park, a yard or an access road carrying a network of thermal and fatigue cracking, chasing and sealing the cracks and then banding over them is usually the proportionate answer. The band is the closing operation, not the repair. Grades are chosen by friction rather than by heat: the numbers in OBS 45 and OBS 60 refer to polished stone value and the skid resistance the finished band reaches, not to softening point. Both soften at roughly 85°C, give or take 15 degrees. What is overbanding covers the materials in more depth, and the overbanding service page covers how we do it.

This is the call back we see most often, and it is always the same picture.

A concrete pavement has a joint whose sealant has aged. Rather than cutting the old sealant out and resealing the slot, somebody runs a band of hot bituminous material over the top of it. It looks tidy on the day. The problem is that the joint underneath has not stopped moving.

Come the first cold snap, the slabs contract and the joint opens. The band has a little give in it, and nowhere near enough. It is bonded flat to the surface either side, so the joint's whole movement is concentrated in a thin strip of material never sized for it. It splits straight down the middle, exactly along the joint line. Or it holds in the centre, debonds from one side instead, and the loose edge starts to lift.

From there it goes downhill quickly. Traffic picks at the lifted edge. Water runs straight into the joint, which is the one thing the work was meant to prevent. The band then has to come off before the joint can be sealed properly, and that is a cost which did not exist before. You have paid for the band, paid to remove it, and paid for the joint sealing you needed at the outset.

In our experience the diagnosis is easy once you look at the failure pattern. A split running dead straight down the centre of a band, following the joint below it, is not a bad batch of material. It is movement the material was never asked to handle.

The second failure mode has nothing to do with movement. It is about what the band does to the surface it sits on.

Bitumen is slippery when wet. A band laid thicker and wider than it needs to be is a strip of low friction across a running surface. It is a particular hazard to motorcyclists. Highway authorities do care about this, and adopted highway specifications cap band width and thickness for exactly that reason. Band a whole network of cracks generously and those strips add up. A meaningful share of the running surface then behaves differently to the rest of it.

Thickness brings a second problem. A proud band is a trip hazard on a footway. On a container yard or an aircraft apron, turning traffic works at a raised edge until it lifts. The remedy is not complicated. Work to the specified width and thickness, and use a compound whose own aggregate restores the skid resistance, rather than relying on grit scattered over the top afterwards.

None of that is a reason to avoid overbanding. It is a reason to ask what width and thickness the gang is working to before they start. On adopted highway in England those figures are mandatory rather than advisory.

The choice is not really a two way one. There is a third route, and it is the one most often left out of the conversation.

Crack chasing cuts the crack out to a defined channel first. A saw follows the line of the crack and opens it to a consistent width and depth. That gives you something a raw crack never offers: a proper reservoir, cut to a consistent width and depth, with clean sound faces for the sealant to bond to.

The difference in outcome is not marginal. Sealant poured into a chased channel bonds to sound material at a known depth to width ratio. A band smeared over an unprepared crack bonds to whatever the surface happens to be, over a void of unknown shape, with dirt and fines still down in the crack. The first is a repair. The second is a covering.

Chasing is a sawing operation on a hard surface, so respirable crystalline silica has to be controlled at the tool. That means water suppression or on tool extraction, in line with HSE guidance on construction dust. It is worth asking which one a contractor intends to use.

For a designed joint rather than a crack, the equivalent preparation is joint widening. Where old sealant has failed or the joint arris has spalled, the slot is cut back to sound concrete and to the correct shape before anything is poured. Our crack chasing and joint sealing usually run as one operation for that reason.

Put the two questions together, the nature of the feature and the movement in it, and the answer falls out.

• Concrete, designed joint, moving seasonally. Joint sealing, with the grade matched to the calculated movement. Never a band over the top of it.
• Concrete, genuine crack. Chase it to a reservoir and seal it. A band goes over the top only where the surface needs the extra wear protection.
• Asphalt, isolated crack, little movement. Overbanding is proportionate, once the crack itself has been prepared.
• Asphalt, straight cracks at regular spacing. These reflect joints below. Treat the movement, not the surface.
• The defect is in the material either side, not in the joint. Spalled arrises and broken bay edges are not a sealing problem at all.

That last one catches people out regularly. A joint can look failed when the sealant is sound and the concrete shoulder either side has broken away. Sealing it again simply puts good material against bad. The shoulders have to be cut back to sound concrete first, or the new seal has nothing to hold on to.

Across airfield, port and highway pavements, we've found that a fair number of joints reported to us as failed sealant are really failed arrises. The sealant did its job. The concrete holding it did not. It is a five minute check with a screwdriver, and it changes what the work is.

A handful of questions separate a considered specification from a band over everything.

• Is the feature a designed joint or a crack? If nobody has asked, nobody has decided.
• Has anyone measured the movement, or looked at the joint in cold weather?
• Is the crack being chased and sealed first, or is a band going straight over a raw crack?
• What width and thickness is the band being worked to, and who is checking it?
• Is the concrete either side of the joint sound, or does it need cutting back first?
• Is the sealant grade matched to the movement, or is it whatever was on the wagon?

Get the first question right and the rest fall into place. Get it wrong and the best workmanship in the trade will not save the job, because the wrong treatment was specified before anyone arrived on site.

We work across Kent, London, Sussex and Essex from our own yard, and mobilise UK wide for longer programmes. We are SafeContractor and CHAS accredited and members of the Extruded Sealant Association. If you have a surface and you are not certain which of the three treatments it needs, get in touch and we will look at it properly before quoting.