Repointing a Parapet, and Deciding Whether It Should Stay Up

Half past eight on the third lift

The scaffold went up on Sunday night under a footway licence that runs eleven weeks. You are on the third lift with a tape, a plugging chisel, a torch and a camera, and the building is a 1908 retail parade with a brick parapet that somebody rendered on the street face in the 1970s. From the pavement it read as a pointing job. From here it does not. Two courses under the coping the render has debonded across four metres and rings hollow under a knuckle; where it has already dropped off, the bed joints behind it are weathered back the better part of thirty millimetres, and one perpend has nothing in it at all — you can see the roof side through it.

Two questions are being priced on this scaffold and they have very little to do with each other. One is a quantity: how many metres of joint come out, how deep, and what goes back. The other is whether the wall should be standing over a busy footway in the first place, which is geometry with a structural answer, and it is the one that decides whether the first question was worth asking. Repointing restores a joint. It does not make a slender wall less slender, it does not tie a free-standing cantilever back to a roof it was never anchored to, and an immaculately pointed parapet on a wall that a structural engineer would want braced is a lot of money spent on a hazard that now looks maintained.

So take them in the order that can stop the job, not the order the client asked for. What is holding the wall, and how much of it is unrestrained. What is genuinely dead as opposed to merely old. Then the joint, the depth, the mix, and the weather it has to go in during. The scaffold is already standing and it is the meter that runs whether anyone is on it or not, so anything you fail to write down this week gets paid for twice.

Sounding it, and marking what is actually dead

Weathered is not the same as failed, and the difference is worth a large fraction of the price. Mortar that has eroded back a few millimetres and gone soft on the surface but resists a chisel behind that is doing its job; raking it out gains the building nothing and loses it a week. Mortar you can pull out with hand pressure to a depth greater than the joint is wide has stopped transferring anything and is letting water into the core of the wall. The survey exists to draw a line between the two, elevation by elevation, and to come off the scaffold with a percentage rather than an impression.

Do it as a sequence, and do it the same way on both faces so the two sheets can be compared afterwards.

1. Take the outer face from the top lift down, then the roof face as a separate elevation. They weather differently, they dry differently, and they will not price the same.
2. Run a plugging chisel or a stiff key along a bed joint at every lift and at both ends of every straight run. Record the depth at which the mortar starts resisting.
3. Tap each unit with the handle. A hollow ring under render says the render has debonded and the joints behind it have been wet for years, which is a different job from the one that was quoted.
4. Mark and photograph every vertical crack, and note whether it steps through the joints or runs straight through the units. Straight through the units says the wall has moved, not eroded, and no amount of pointing addresses it.
5. Look for a damp-proof course under the coping and another at the base of the parapet. On buildings of this date there is usually neither, which is why the masonry stays saturated and why the joints failed here first.
6. Measure the thickness at three points along the run, on the parapet itself, and write down what you deducted at each for render, parge or coping overhang.
7. Close each elevation with the proportion of joint length you are calling dead, as a percentage, with the photographs numbered against the scaffold lifts.

Depth is the input nobody measures, and the one the order turns on

Feed the calculator the length you are repairing, not the length that exists. If the survey called sixty per cent of the joint dead on the street face and thirty per cent on the roof face, those are two separate lines: 465 m on each face, 279 m and 140 m respectively, about 420 m in total rather than 930. It is a number that will move again once raking exposes what is behind the face, which is why it belongs in the contract as a re-measured quantity rather than a fixed one.

Depth is where the money and the durability both sit. NPS Preservation Brief 2, Repointing Mortar Joints in Historic Masonry Buildings, sets out the working rule that joints are cut back to something like two to two and a half times their width so the new mortar has enough section and enough bonding surface to survive; a surface skim over sound-looking old mortar pops out within a few winters and takes the arris of the brick with it when it goes. On a listed building or in a conservation area the depth is not yours to choose — it is set by the specification and by whoever is advising on the consent, and BS 8221-2 Code of practice for cleaning and surface repair of buildings is the document behind the British approach to that. Note also that raking depth is capped at the other end by the wall: on a half-brick parapet leaf you are not cutting 40 mm into a 102 mm wythe.

Then add what the geometry does not see. Every raked joint has to be brushed and flushed out, and the joint has to be damped before mortar goes in or the brick will suck the water out of the mix before it can cure — a clay brick with a high initial rate of absorption measured to ASTM C67 needs noticeably more wetting than a dense engineering brick, and the same pointing gang on the same scaffold will get very different results on the two if they treat them alike. And keep a waste allowance on its own line rather than folding it into the depth, because droppings on the boards, the tail of every gauging and material that stiffens past use are real but they are not geometry, and the only honest source for that percentage is what your own last three jobs of this type recorded.

The mix is matched to the brick, not to a strength class

The single most common way to spend a client's money making a building worse is to repoint soft old brick in a strong cement mortar. The joint is meant to be the sacrificial element: softer than the units, so that movement is accommodated in the joint and so that water leaving the wall leaves through the mortar rather than through the face of the brick. Reverse that hardness order and the wall has nowhere to move and nowhere to dry, and the damage shows up as spalled arrises and blown faces on brick that was perfectly sound before the scaffold arrived. On a frost-exposed parapet, wet on three sides and colder than the wall beneath it, that failure arrives faster than anywhere else on the building.

So the mix is specified against what is already in the wall, and there is a test rather than a judgement available: ASTM C1324 Standard Test Method for Examination and Analysis of Hardened Masonry Mortar will tell you the binder type and give you an aggregate to match by grading and colour, and it costs a fraction of one lift of scaffold. ASTM C1713 Standard Specification for Mortars for the Repair of Historic Masonry is written for exactly this material, and is a different document from ASTM C270 Standard Specification for Mortar for Unit Masonry, which governs new work. In Europe the equivalents are BS EN 998-2 for the mortar and BS EN 459-1 for the lime itself, and Historic England's Practical Building Conservation volume on Mortars, Renders and Plasters is where the practice around natural hydraulic and non-hydraulic limes is set out properly.

Practically, the consequence for the programme is that a lime point is not a cement point with a different bag on the pallet. It wants damp curing under hessian for days rather than hours, it wants protection from driving rain and from sun for longer than that, and it wants a frost-free run measured in weeks before it can be relied on. Get that on the programme at the survey stage, because it is the reason the eleven-week licence is eleven weeks and not six.

Weather, and the fortnight nobody allowed for

Pointing is laid in thin sections on an exposed surface with wind on three sides of it, which makes it more weather-sensitive than the bricklaying it is repairing. Both ends of the temperature range have rules and they are written down: TMS 602 carries explicit cold-weather and hot-weather construction procedures, and BS EN 1996-2 Design considerations, selection of materials and execution of masonry with BS 8000-3 Code of practice for masonry does the equivalent work in the UK. Read whichever governs the job rather than carrying a remembered threshold across from the other one, because they are not the same numbers and a parapet is the worst place on the building to be wrong about frost.

The practical version is protection and patience: sheet the working lifts, keep hessian and a spray on the pointed work for as long as the mix requires, and accept that a hard frost forecast for the night stops the afternoon's pointing rather than merely worrying it. A joint that freezes before it has set does not recover, it just looks acceptable until the second winter.

What the price is conditional on, and what goes in the file

Come off the scaffold with a price that says what it assumes. The joint quantity is provisional and re-measured after raking, because nobody can see the back of a joint from the front of it. The mortar specification is subject to analysis of the existing material, not chosen from a catalogue on the day. The rebuild is a defined provisional item with a rate per metre of run and a stated height, so that turning it on does not become a renegotiation. And the whole of it is conditional on the slenderness check, because if the parapet has to be braced or reduced, the repointing scope changes shape entirely and some of it disappears.

One more thing worth writing down explicitly, because clients assume the opposite: repointing is not a waterproofing measure. It closes the joints, which is necessary, but a solid masonry parapet with no damp-proof course and a coping that sheds outward will still take water in and still stay wet, and if the complaint that started this was a damp ceiling on the top floor then the pointing may not be the cause at all. ASTM C1601 Standard Test Method for Field Determination of Water Penetration of Masonry Wall Surfaces exists precisely so that claim can be tested on the wall rather than argued about afterwards.

The file that leaves the scaffold should contain the two elevations as separate sheets with dead-joint percentages, the thickness measurements with their deductions written next to them, the unbraced height and what you concluded is restraining the wall, the photographs numbered to the lifts, and the mortar sample reference. That is a morning's work at the top of an eleven-week hire, and it is the difference between a variation you can evidence and one you absorb.