Speed surveys and how recorded speeds change a scheme

The visibility splay a highway authority asks for is derived from the speed vehicles actually travel, not from the number on the sign. On a lane signed at the national limit where the 85th percentile speed is 34mph, the required sight line is a fraction of what 60mph would demand, and a seven-day automatic traffic count is what turns that assertion into evidence. Without the survey the officer applies the posted limit, and on most rural accesses the geometry then fails.

A speed survey costs a few hundred pounds. Redesigning a site around a splay you never needed costs considerably more.

Why the posted limit is not the design input

Manual for Streets puts the point plainly: visibility should be based on observed vehicle speeds rather than the speed limit, because the limit is a legal maximum and drivers respond to the road they are on.

The mechanism is stopping sight distance. The driver needs enough forward visibility to perceive a hazard, react and stop, and that distance grows quickly with speed. The Y distance of a visibility splay is set equal to the stopping sight distance for the design speed, measured along the nearer edge of the carriageway in each direction from a point set back from it.

Check those figures against the current edition of the document rather than against a quoted table, because guidance is periodically revised. On roads where speeds are higher, Manual for Streets gives way to the highway design standards used for the strategic and higher-speed network, which produce considerably longer distances.

The X distance completes the splay. Two point four metres, measured back from the edge of the carriageway along the centre line of the access, is the standard figure for residential accesses, with 2.0m accepted on lightly trafficked, low-speed streets.

The percentile speed and the wet weather adjustment

The 85th percentile speed is the speed at or below which 85 per cent of vehicles travel. It is the standard design input because it represents most reasonable driver behaviour without being set by the fastest few.

Highway authorities generally want the wet weather value. Manual for Streets 2 records that speeds in wet conditions are lower than in dry conditions, and some authorities accept an adjustment to the dry-weather 85th percentile in place of a separate wet dataset. Practice varies between authorities and the adjustment is not universal, so ask which approach yours applies before commissioning the survey.

Where wet weather speeds are required directly, the survey has to run long enough to capture rainfall, and the data has to be matched against a rain record so the wet periods can be isolated. That is a reason to survey outside a dry spell rather than in one.

Mean speed is reported alongside, and it is useful context. A road with a mean of 28mph and an 85th percentile of 34mph behaves differently from one with a mean of 33mph and an 85th percentile of 35mph, and the spread tells you something about the consistency of driver behaviour.

Survey equipment, duration and placement

Automatic traffic counters using pneumatic tubes across the carriageway are the standard tool. Radar and video units are used where tubes are impractical, and the authority may have a view on which it accepts.

Duration matters. Seven continuous days is the normal minimum, covering both weekdays and a weekend, and fourteen days is better where the result is close to a threshold. A two-hour manual count with a speed gun is not a speed survey and will be rejected.

Placement decides whether the data is meaningful. The counter goes at or very close to the proposed access, on a length where vehicles have reached free-flow speed, away from junctions, bends, crests, parked cars and anything else that suppresses speed artificially. A counter sited 200m away on a straighter section records a different road.

Tell the highway authority where you intend to place it before you book the survey. Confirming the position in advance removes the most common objection to the resulting data.

Seasonal and weekly variation

Survey in a neutral period. Avoid school holidays, the weeks either side of Christmas, August, and any period when roadworks, a diversion or a temporary limit is affecting the route.

Traffic composition changes seasonally on rural roads in particular. Agricultural vehicles in harvest months, tourist traffic in summer and school-run peaks in term time all pull the distribution around, and a survey that captures an unrepresentative fortnight invites a challenge from the authority or from an objector.

Weather during the survey should be recorded even where a wet weather value is not being extracted, because an unusually icy or stormy week produces low speeds that are not representative either.

An illustrative example

A single plot on a rural lane signed at the national speed limit, with a hedge along the frontage. The council's initial comment applies 60mph and asks for 215m of visibility in each direction, which the frontage cannot deliver. A fourteen-day tube survey records an 85th percentile of 33mph westbound and 35mph eastbound. Applying the stopping sight distance for 35mph produces a requirement the frontage can meet with hedge reduction on one side and a small land transfer on the other. The access is approved with a condition requiring the splay to be kept clear above 0.6m. This example is illustrative rather than drawn from a specific case.

Converting speed into visibility requirements

Once you have the figure, the process is mechanical.

1. Take the 85th percentile speed for each direction separately, since they are rarely the same.
2. Apply any wet weather adjustment the authority accepts, and state which you have used.
3. Read the corresponding stopping sight distance from the current guidance.
4. Set the X distance at 2.4m, or 2.0m where the authority accepts it on a low-speed street.
5. Draw the splay on the site plan with the eye height and object height annotated, and show what is inside it.
6. Demonstrate that everything within the splay can be kept below the required height, including hedges, walls, signs, parked vehicles and the ground itself.

That last step trips people up. A splay drawn over land you do not own or control is worth nothing, and a splay over a rising verge is not deliverable without excavation. Identify the ownership and the levels at the same time as the geometry.

Forward visibility on the approach to the access, and visibility of pedestrians on any footway crossed, are separate checks that sit alongside the splay.

Challenging an authority's assumed speed

Highway officers apply the posted limit when they have nothing else, and that is a reasonable default rather than an unreasonable one. The way to move it is data, submitted courteously and early.

Put the survey in the Transport Statement with the raw data appended, the equipment and dates stated, the placement shown on a plan and the percentile calculation set out. Anticipate the two questions that always come: was the location representative, and was the period representative.

Where an objection relies on collision history rather than speed, that is a different argument requiring different evidence, and our note on using personal injury collision data properly covers how to handle it. Where the access is onto a road subject to the national limit, the specific considerations are set out in our guide to access onto a national speed limit rural road.

On larger schemes the same survey feeds the junction modelling in a Transport Assessment, so commission it once and use it for both.

Commissioning a survey that will be accepted

Four things make the difference between data the authority accepts and data it argues with.

Agree the location and the duration with the highway officer in writing beforehand. Use a surveyor who supplies the raw per-vehicle records rather than a summary sheet. Survey in a neutral period of at least seven days. And report both directions separately, with the mean, the 85th percentile and the sample size for each.

If the result comes back higher than you hoped, that is useful information rather than a problem to bury. It tells you the access needs relocating, the frontage needs opening up, or the site needs a different approach, and it tells you that before you have spent money on drawings.

This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.