Crack Monitoring: Why One Measurement Is Never Enough

Every crack survey ends with the same question, and it is almost never “how wide is it?” It is “is it getting worse?”

That is exactly the right question to ask. It is also the one question a single visit cannot answer.

The film above is a crack survey in progress. A crack width gauge goes on an internal wall at high level, then at mid level, then lower down again. In the next room, a second crack running down the corner beside a window gets the same treatment, from the reveal behind the radiator down to the skirting. Nothing dramatic happens in a minute and a half of footage, and that is rather the point.

What is being built in those two rooms is not a repair. It is a baseline.

Width is a number. Change is an answer.

A crack width gauge is a strip of plastic with a stepped edge, and each step is a known width. You hold it against the crack, find the step that matches, and read the figure off. It costs a few pounds and it removes the guesswork, because “about the width of a pound coin” is not a measurement anybody can repeat.

So on the first visit you get a number. Several numbers, in fact, because the gauge goes on at more than one height. What you do not get is the thing that actually drives the decision: whether that number is the same today as it was last spring, and whether it will be the same next spring.

A 4mm crack that has been 4mm since 2009 is a decorating job. A 2mm crack that was 1mm in June is a structural problem. The widths tell you almost nothing on their own. The pair of readings, separated by time, tells you everything.

That is crack monitoring: measuring the same points, on the same crack, by the same method, at set intervals, and writing down the date every single time.

What the first survey can tell you

Plenty, as it happens. One visit, done properly, establishes:

• Where the crack runs — its full extent, floor to ceiling or corner to opening, not just the part you can see from the sofa.
• How wide it is at several points — because a crack that changes width along its length is saying something a single reading would miss.
• What shape it is — parallel-sided, tapering, or widest in the middle. We go into this in detail in how to read a crack in a wall, but the short version is that a crack which tapers has rotated, and rotation points at movement rather than shrinkage.
• Which way it steps — diagonal cracking that follows the mortar joints through brickwork reads very differently from a straight vertical split through the units themselves.
• What else is happening in the same place — damp, staining, a blown patch of plaster, a door that no longer closes. In the film, the lower part of that first wall carries brown water staining as well as the crack, and damp and cracking in the same place have to be separated carefully. Either one can cause the other, and the repair is completely different depending on which came first.

From that, an experienced engineer will usually have a working theory within the hour. Thermal movement. Plaster shrinkage. A lintel doing less than it should. Clay soil shrinking back from a thirsty tree. Settlement that finished decades ago and has been perfectly content ever since.

What the first survey cannot tell you

A theory is not a diagnosis, and this is where a lot of homeowners get sold the wrong thing.

Most cracking in British houses is not progressive. It moves with temperature and humidity, opens a little in one season and closes a little in another, and has been doing so quietly for years. Looked at once, in November, a seasonal crack and a subsiding crack can measure exactly the same.

The only way to tell them apart is to come back.

What crack monitoring actually involves

It is less elaborate than it sounds, and on a domestic property it is rarely intrusive.

Tell-tales

The familiar version: two overlapping acrylic plates, one fixed either side of the crack, with a scale printed on the top plate and a crosshair on the bottom one. As the crack opens, closes or shears, the crosshair moves against the scale and you read the displacement directly. Cheap, visible, and perfectly adequate for larger movement.

Their weakness is resolution. A tell-tale reads to roughly half a millimetre by eye, which is fine for a crack that is going somewhere and useless for one that is drifting by a tenth of a millimetre a month.

Studs and a demountable gauge

For anything that needs real precision, small stainless studs are fixed either side of the crack at a set spacing and a demountable strain gauge spans between them. The gauge is removed between readings, so there is nothing on the wall but two discreet studs, and it resolves movement to a few hundredths of a millimetre.

This is what insurers and loss adjusters generally expect on a subsidence claim, and it is what we would specify where the answer has to stand up to scrutiny.

The photographic and dimensional record

Every set of readings is accompanied by photographs taken from fixed positions, with the gauge in shot, and every one is dated. It is unglamorous and it is the part that carries the most weight later, because a photograph with a measurement and a date in it is evidence. A photograph without one is a picture of a wall.

Level monitoring

Where whole-building movement is suspected rather than a single crack, precise levelling across the floors or a set of external datum points will show whether one part of the structure is dropping relative to another. It answers a different question from crack monitoring and the two are often run together.

How long does crack monitoring take?

Usually between six and twelve months, and twelve is the honest answer more often than six.

The reason is seasonal. Shrinkable clay soils — which cover a great deal of the South East, the Home Counties and pockets of the East of England — lose moisture through a dry summer and recover it through a wet winter. Ground that dries out shrinks, and a foundation sitting on it goes down with it. Come the winter the clay swells again and the building lifts back. That is why a crack can look alarming in September and almost closed in March.

Measure across a single season and you will see movement and not know whether it is going to come back. Measure across a full cycle and the pattern resolves itself.

Readings are normally taken monthly or quarterly depending on what is suspected and how fast it appears to be going. Where there is any concern about safety, the interval shortens and the monitoring stops being a watching brief.

What the readings actually prove

Monitoring produces one of three outcomes, and all three are useful.

No measurable change. The crack is historic. Something moved once — a building settling after construction, a one-off change in loading, an extension bedding in against an older structure — and it stopped. Fill it, redecorate, and get on with your life. This is the most common result by a comfortable margin, and it is the cheapest piece of information you will ever buy.

Seasonal movement that returns to where it started. The crack opens and closes on an annual cycle but has no net drift. That points at clay shrink-swell, usually with vegetation involved. The answer is rarely underpinning; it is managing the cause, which may mean a programme of pruning, removing the right tree rather than the nearest one, or dealing with a leaking drain that has been softening the ground for years. We cover the usual suspects in shrubs, climbers and leaking drains.

Progressive movement. The crack is wider at every reading and does not come back. That is the result nobody wants and the one that justifies the whole exercise, because now there is a measured rate rather than an argument. Progressive movement moves the job on to a structural investigation: trial pits, drainage tests, soil samples, and a proper look at what the foundation is actually standing on.

When monitoring is the wrong answer

Monitoring assumes you have time. Sometimes you do not.

If a crack has appeared suddenly and widened noticeably within days or weeks, if a wall is bulging or leaning, if a lintel over an opening has dropped, if floors have gone out of level quickly, or if a crack is wide enough to see daylight through, the correct response is investigation and, where necessary, temporary support — not a set of tell-tales and a diary. The same goes for anything that follows an impact, a fire, a flood, or works next door.

A reasonable rule of thumb: slow, stable and inconvenient gets monitored. Fast, fresh or frightening gets looked at properly, now.

Crack widths and what they are generally taken to mean

Industry practice classifies visible damage in walls on a scale that runs from negligible to very severe, and the width of the crack is the usual starting point. Broadly:

• Under 0.1mm — negligible. Hairline. Almost always decorative.
• Up to about 1mm — very slight. Fine filler at redecoration.
• Up to about 5mm — slight. Easily filled, possibly some repointing externally.
• 5mm to 15mm — moderate. Brickwork may need cutting out and replacing, doors and windows may stick.
• 15mm to 25mm — severe. Sections of wall may need rebuilding, frames distorted, service pipes potentially affected.
• Over 25mm — very severe. Major repair, possible instability.

Treat that as orientation rather than diagnosis. Width is one input. A 3mm crack that is stable is a far smaller problem than a 1mm crack that has doubled since the summer, and classification scales exist to describe damage, not to establish cause.

Monitoring and an insurance claim

If subsidence is suspected and a claim is in play, monitoring is usually not optional. Insurers want a measured record before they will accept cause, and they will generally want it taken by someone independent, using a repeatable method, over a period they consider representative.

It is worth understanding that this works in your favour as well as theirs. A documented series of dated readings is what turns “we think the house is moving” into a position that can be defended. What goes into the report that follows is set out in structural report for an insurance claim.

If you are buying or selling

A crack discovered during a sale is a negotiation, and negotiations run on evidence. A seller with twelve months of readings showing no movement is in a far stronger position than one offering reassurance. A buyer with a monitoring recommendation in hand has a reason to retain funds rather than walk away.

If a homebuyer’s survey has flagged cracking and recommended specialist advice, that is the moment to get a baseline established rather than wait for the sale to stall.

What to do if you have a crack right now

• Photograph it with something for scale and note the date. A tape measure or a ruler held against the crack, and the date written down. If you do nothing else, do this — you are creating the first reading, and the clock starts from the photograph you take today rather than the survey you book in March.
• Note the season and the weather. A long dry summer behind you changes how a reading should be interpreted.
• Do not fill it yet. Filler hides the evidence and tells you nothing. If it reopens you have lost the baseline; if it does not, you still do not know why.
• Look for the pattern, not the crack. Is the same thing happening elsewhere? Do doors bind in the same corner of the house? Has a gutter been overflowing against that wall?
• Do not panic about age. Older buildings move, and movement that finished a century ago is history, not a defect.

Frequently asked questions

How often should a crack be measured?

Monthly or quarterly is typical, over six to twelve months. The interval is set by what is suspected and by how quickly the crack appears to be changing, and it shortens if the readings start moving.

Can I monitor a crack myself?

You can create a useful informal record with dated photographs and a ruler, and it is genuinely worth doing. What you cannot do at home is produce the repeatable, instrument-based readings an insurer or a purchaser’s solicitor will rely on, because those depend on fixed reference points and a gauge that returns the same figure in different hands.

Does a crack that has already been filled need monitoring?

If it has reopened through the filler, yes — and the fact that it reopened is itself a reading. Fresh cracking through a recent repair is one of the clearer signs that something is still moving.

Is a crack wider than 5mm serious?

It is worth a professional look, but width alone does not settle it. A 6mm crack that has been stable for twenty years is a repair; a 2mm crack that is growing is an investigation. Measure, date, and come back.

Will monitoring delay a repair I already know I need?

Only the structural element of it. Cosmetic work, redecoration and most non-structural repairs can proceed, provided the monitoring points themselves are left undisturbed. Where safety is in question, monitoring is not the route at all.

The point of all of it

At the end of the day, this is not about filling cracks. It is about finding out why the crack is there and dealing with that first, because a crack filled over a live problem is a crack you will fill again next winter.

Width, direction, position and a date. That is what a survey is actually building, and the date is the part that does the heavy lifting. Measured once, a crack is just a number. Measured again in three months, it becomes an answer.

Ora Labora Structural Engineers carries out crack surveys, monitoring programmes and subsidence investigations for homeowners, buyers, architects and insurers across Peterborough and London. If you have a crack you have been keeping an eye on, the useful first step is to stop keeping an eye on it and start measuring it.