Choosing comparable TRICS sites for a trip rate

Every number in a transport assessment traces back to the sites you chose from TRICS. Get the selection right and the trip rate survives scrutiny at committee and, if it comes to it, at appeal. Filter too hard, keep a survey from 2009, or leave Greater London sites in a set describing a market town, and the highway authority will attack the rate rather than the conclusions, which is a much harder argument to win.

Selection is a judgement exercise with a written trail. Treat it as the evidence it is.

Why site selection decides the assessment

Trip rates are averages of surveyed behaviour at real sites. The database does not know what your site is like, so the comparability of the set you build is the entire basis for applying its output to your scheme.

That has a consequence people skip. Before opening the software, write down what your site is: the land use, the size, the parking ratio, the settlement type, the distance to a rail station and a bus route with a useful frequency, and whether there is anything unusual about it. Those characteristics are the brief for the selection.

The alternative approach, which is to filter until the rate looks acceptable, produces an assessment that collapses when the highway officer asks why a filter was applied. Officers do ask.

Filters that change the answer

Land use category first. The database splits categories finely, and the difference between privately owned houses, privately owned flats and affordable or social housing is material, as is the split between food and non-food retail. Where a scheme is mixed, build separate sets rather than one blended set.

Region and area type do most of the remaining work. Including Greater London surveys in a set for a suburban site outside London depresses the vehicular rate because car ownership and public transport differ, and that is exactly the objection a highway authority raises. Exclude London unless your site is in it.

Location type, the classification covering town centre, edge of town centre, suburban area, neighbourhood centre and free standing, should match your site rather than your preference.

Population within one mile and within five miles is the filter most often left at the default. Applied sensibly, it separates a site in a dense urban area from one on the edge of a village. Applied aggressively, it reduces the sample to nothing.

Parking provision is the one to handle with care. Filtering on parking spaces per dwelling makes the set look like your scheme, and it also builds your proposed parking ratio into the evidence that is supposed to test it. Say what you have done either way.

Sample size and over-filtering

Each filter you add removes sites. The temptation, particularly when a rate looks high, is to keep going until only the convenient surveys remain.

Most practitioners aim to keep at least ten surveys in a set, and to be able to explain every filter applied. A set of three sites produces a mean with wide uncertainty and gives an objector an easy line: the rate rests on three car parks in three towns, surveyed on three days.

Where the honest position is that few comparable sites exist, say so and widen a filter you can defend widening, such as accepting a broader population band while holding location type fixed. Then show both sets, the tight one and the wider one, and use the more conservative outcome. An assessment that presents a sensitivity test is more persuasive than one that presents a single number.

Look at the spread as well as the mean. A set with a handful of outliers well above the rest deserves a sentence explaining them rather than silence.

Survey dates and what the pandemic did to older data

Date range is now a substantive judgement rather than housekeeping. Travel patterns changed after 2020, particularly for offices, and a set of office surveys from the mid-2010s describes a working week that no longer exists.

For residential schemes the change is smaller but real, with more midday movement and a flatter peak. For retail and leisure, delivery traffic has grown.

The defensible approach is to state the date range used, say why, and where older surveys are retained, explain what they still tell you. Highway authorities vary in how firmly they take the point, and the ones that do take it firmly will reject a set quietly built on pre-2020 data.

Also check the survey day. Weekday surveys support a weekday peak-hour assessment. If the authority is worried about Saturday servicing or a weekend peak at a retail unit, you need weekend surveys in the set, not an adjustment factor.

Recording the selection so it can be checked

Include the selection record in the appendix, not a summary of it. The record should list, in order:

1. The land use categories chosen and why.
2. Every filter applied, with the range or values used.
3. The number of sites remaining after each filter.
4. The final site list with survey dates and locations.
5. Any site removed manually, with the reason for removing it.

Manual exclusions attract the most scrutiny, so give a reason that stands on its own. Removing a site because it is a retail park with a drive-through restaurant attached is defensible. Removing it because it produced a high rate is not, and reviewers can see the effect of the removal on the mean.

State the trip rate unit clearly, whether it is per dwelling or per 100 m² of gross floor area, and carry that unit consistently through the assessment. Mismatched units between the rate table and the development schedule is a common error and an embarrassing one.

Where the highway authority pushes back

Four objections come up repeatedly, and all four can be headed off in the scoping stage.

The authority may say the sample is too small, which is answered by a wider set presented alongside. It may say the parking filter has assumed the answer, which is answered by showing the unfiltered result. It may say the surveys are too old, which is answered by a date range and a justification. And it may say the location type does not match, which is answered by matching it properly the first time.

Agree the methodology before you run it. A scoping note that records the categories, filters and date range, signed off by the authority in writing, removes the argument at the point it costs least. The same discipline applies to the fuller Transport Assessment methodology, and to the choice between that and a Transport Statement for a smaller scheme.

Specialist land uses need more care still, because the database categories fit them loosely. The problem is set out in the piece on trip generation for student and age-restricted housing.

Sense-checking the rate against local evidence

A trip rate that survives its own methodology can still be wrong for the site. Two checks are worth the hour they take.

Compare the modal split implied by your rate with census travel-to-work data for the ward or the middle layer super output area. Where the database set implies 80 per cent car driver and the local area records far less, either the set is wrong or the site is different for a reason you can name and you should say which.

Then compare the absolute number of vehicle movements with a local count if one exists, such as a traffic count on the access road or the authority's own monitoring. The question you are answering is whether the predicted flow is plausible against what the road already carries.

Where the scheme relies on a modal shift to make the numbers work, that shift belongs in a Travel Plan with monitoring and targets, not in an adjusted trip rate. Mixing the two is how an assessment ends up defending a reduction it never evidenced.

Start a transport document draft on PlanScope and send it to a transport planner for review before submission.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required.