Construction vibration and the limits in BS 5228 part 2

Construction vibration is measured as peak particle velocity in millimetres per second, and the number that matters is not the one people worry about. Cosmetic damage to plaster in an ordinary house needs peak particle velocities in the tens of millimetres per second. Complaints start at around one. The gap between those two figures is where most construction vibration disputes live: the works are nowhere near damaging anything, and the neighbours are certain their house is being shaken apart.

BS 5228-2 is the standard that sets out both scales, along with prediction methods for piling, compaction and other construction activities. Using it properly means reporting against the human response criteria and the damage criteria separately, because conflating them either alarms people unnecessarily or dismisses a real complaint.

What is measured and in what units

Peak particle velocity, PPV, is the maximum velocity of a point in the ground or structure as the wave passes, in millimetres per second. It is measured with a triaxial geophone, giving three orthogonal components, and reported either as the maximum component or as the resultant.

For human response inside buildings, a second measure is used: the vibration dose value, VDV, which accumulates vibration energy over an exposure period and is expressed in metres per second to the power 1.75. BS 6472-1 deals with human response to vibration in buildings and gives ranges of vibration dose value against probability of adverse comment.

Report which quantity you are using, at which position, and whether the measurement is on the ground, on a foundation or on a floor. A reading taken on a suspended timber floor can be several times the reading at the foundation of the same building, because the floor amplifies at its own resonant frequency. Comparisons between the two are meaningless, and that confusion appears in real reports.

Human response against building damage

The two scales are different by an order of magnitude or more, and stating both is what makes a vibration section credible.

For human response, the guidance in BS 5228-2 indicates that vibration around 0.14 millimetres per second may be just perceptible in the most sensitive situations, around 0.3 millimetres per second is perceptible in residential environments, around 1 millimetre per second is likely to cause complaint in residential settings, and around 10 millimetres per second is likely to be intolerable for anything more than a very brief exposure.

For cosmetic damage, the guide values in BS 5228-2, derived from BS 7385-2, are far higher. Reinforced or framed structures and heavy industrial buildings are given a value of around 50 millimetres per second at 4 hertz and above. Unreinforced or light framed residential and light commercial buildings are given around 15 millimetres per second at 4 hertz, rising to 20 millimetres per second at 15 hertz and to 50 millimetres per second at 40 hertz and above. For continuous rather than transient vibration those values may need reducing, and for sensitive or historic structures lower limits are usually adopted. Check the current edition of the standard for the exact values and their qualifications before relying on them.

The frequency dependence is the part that gets lost when a single number is quoted. Vibration at 8 hertz and at 40 hertz of identical amplitude present very different risks, and a limit expressed as a flat figure is either over-restrictive or unsafe.

Predicting vibration from piling and compaction

BS 5228-2 provides empirical prediction expressions for the main sources. They share a common form: peak particle velocity is estimated from a source energy term divided by distance, raised to a power, with a scaling factor that depends on the activity and the ground.

For percussive and vibratory piling the source term relates to hammer energy or to the vibrator's parameters. For vibratory rollers it relates to drum amplitude and width and to the number of passes. Dynamic compaction and rock breaking have their own expressions.

Three practical points about using them.

Ground conditions dominate the scatter. The same rig at the same distance produces very different results in stiff clay, in saturated sand and over shallow rock, so a prediction should present a range and state the assumption rather than offering a single figure to two decimal places.

Distance is the strongest lever available. Because attenuation with distance is steep, a modest change in the piling sequence or the position of the nearest working can bring a receptor from a predicted complaint level to a predicted imperceptible level.

Method choice changes everything. Replacing driven piling with a bored or continuous flight auger technique removes most of the vibration, and where receptors are close that is usually the answer rather than a monitoring regime.

Cosmetic damage thresholds and their basis

The guide values describe the onset of cosmetic damage: hairline cracking in plaster, opening of existing cracks, the sort of defect that is visible and inconvenient rather than structural.

Two caveats matter in practice. The values assume a building in reasonable condition, so a structure already cracked, an old lime-mortared wall with loose render, or a building with known settlement history justifies lower limits. And a building's own condition often explains what the occupier observes: seasonal movement and drying shrinkage produce cracks on exactly the timescale of a nearby construction project, which is why baseline evidence is so valuable.

An illustrative sequence. Piling is predicted at 3 millimetres per second at the nearest terraced house, well below the cosmetic damage guide value of around 15 millimetres per second at low frequency, and above the level at which complaints are likely. The contractor sets monitoring triggers at 6 and 10 millimetres per second, carries out a pre-works condition survey, and writes to residents explaining that they will feel the works and that the readings will be published weekly. Complaints are received, damage is not. Those figures are illustrative, and the real values come from the prediction for that site.

Condition surveys of neighbouring property

A pre-works condition survey is the single most useful document in a vibration dispute, and it costs a fraction of defending a claim.

It should record every elevation and every affected room with dated photographs, crack widths measured with a gauge, notes on existing distortion and previous repairs, and where possible level readings. It has to be agreed with the owner and carried out before any works start, and where access is refused, that refusal should be recorded in writing along with an offer to survey from the boundary.

On notifiable works, the party wall process produces an equivalent schedule of condition, and the two exercises should be coordinated rather than duplicated.

Monitoring during works

Monitoring converts an argument into data. A workable regime specifies:

1. Instrument type and positions, including at the nearest sensitive receptors and on any listed or fragile structure.
2. A baseline period before works start, to record background vibration from traffic and railways.
3. Continuous measurement with automatic alerts, rather than spot checks.
4. Tiered trigger levels, typically set well below the damage guide value, with an action for each: investigate, modify method, stop.
5. Who receives the data, how quickly, and whether readings are published to residents.
6. Reporting at intervals, with an end-of-works summary.

Trigger levels should be justified from the prediction and the condition of the receptors. Setting them at the damage guide value gives no warning margin at all.

Our note on railway noise and vibration for sites beside a line deals with the reverse problem, where an existing source affects a proposed development, and the piece on unattended noise monitoring duration, position and validity covers the equivalent methodological questions for noise.

What a planning condition can and cannot control

Planning conditions set the framework. They commonly require a construction environmental management plan or method statement to be approved before works start, restrict working hours, require a named contact for residents, and require monitoring where sensitive receptors are close.

The detailed control of construction noise and vibration sits elsewhere. Under the Control of Pollution Act 1974, a local authority can serve a notice under section 60 imposing requirements on how works are carried out, and a contractor can apply under section 61 for prior consent to an agreed method, which gives certainty in exchange for committing to the controls. BS 5228 is the reference standard in that process, which is why a well-prepared section 61 application and the planning submission should describe the same controls.

What a planning condition cannot do is guarantee that no damage occurs, or provide compensation. Damage remains a matter for the party wall award where one applies, and otherwise for civil law.

One product note for completeness: a standalone Vibration Assessment is not one of PlanScope's 66 report types. Vibration is normally addressed within a Construction Management Plan or alongside the Noise Impact Assessment, with the prediction and monitoring work carried out by a suitably qualified acoustician or engineer.

This is general guidance, not legal advice; check the position with your local planning authority, your environmental health team and a qualified acoustician before relying on it. For specialist technical reports, qualified professional input may be required.

Draft your construction management documents on PlanScope, then send them for professional review. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.