Choosing TRICS sites and defending the selection

TRICS trip rate site selection is the single most contested page in most transport assessments, and the argument is almost never about the arithmetic. It is about which surveys you kept and which you quietly dropped. A highway officer who disagrees with your trip rates will go straight to the selection parameters sheet, count the sites, read the exclusion reasons and form a view about whether the sample was built to answer a question or to reach a number.

Settle the filters with the highway authority before you run the query. That one habit removes most of the disagreement.

What the database holds, and what it does not

The database contains multi-modal counts taken at real development sites across the UK and Ireland, recording vehicles, pedestrians, cyclists, public transport users and total person trips by time period, alongside a description of each surveyed site: its size, parking provision, location type, population within given radii and the date of the survey.

It does not hold a trip rate for your site. It holds observations at other sites that you argue are comparable, which is a different thing and worth saying plainly in the methodology.

Three limits are worth stating in the assessment rather than leaving the officer to find them. Coverage is uneven between land uses, so some categories have hundreds of surveys and others a handful. The sample skews towards sites that someone had a reason to survey, which tends to mean larger schemes with parking. And a count taken on a single neutral weekday captures that day, so variability between sites in the same category is real and should be reported rather than smoothed away.

Filtering: land use, geography and date

Work through the filters in a fixed order and record the count of sites remaining after each step. That sequence becomes your audit trail.

Land use category comes first, and the choice deserves a sentence of reasoning. A build-to-rent block is not a private housing estate. An out-of-town trade counter is not a retail warehouse. Where the proposal spans two categories, run both and explain which you have adopted and why, rather than picking the lower one silently.

Geography comes next. Excluding Greater London is standard practice for a scheme outside London, because parking restraint and public transport accessibility there produce vehicle rates unlike anywhere else. Excluding Ireland is equally routine. Beyond that, be careful: stripping out every region except your own usually leaves a sample too small to mean anything, and regional variation within England is smaller than location type variation.

Date range is where judgement bites hardest. A ten year window is conventional. Surveys carried out during the pandemic period and its immediate aftermath sit awkwardly, because travel behaviour at those sites was not typical, and there is no settled profession-wide answer on whether to include them. State your position, apply it consistently, and be ready to run a sensitivity test with the other assumption.

Location type is the filter that does the most work and gets the least attention. Suburban area, edge of town, town centre, neighbourhood centre, free standing and village describe quite different travel contexts, and matching the proposal's context matters more than matching its region.

Sample size and why three sites is rarely enough

A sample of three surveys will produce a mean trip rate, and it will not survive scrutiny.

The problem is not statistical purity, it is that a three site sample gives any objector an easy line: change one site and the answer moves substantially. Where the category is well populated, aim for a sample in the high single figures or better, and report the range and the standard deviation alongside the mean so the officer can see the spread you are working with.

Where the category is thinly populated and a small sample is unavoidable, say so and deal with it directly. Widen the date range or the size band and show the effect. Run a sensitivity test at the highest observed rate rather than the mean, and demonstrate that the junction capacity conclusions hold. An assessment that holds at the upper end of the sample is far harder to argue with than one that relies on the average being right.

Size banding deserves a word. Filtering to units within a narrow band around the proposal looks rigorous and often just shrinks the sample. Trip rates per unit tend to fall as schemes get larger, so a wide band with the relationship acknowledged is usually more honest than a narrow band with four sites in it.

Excluding sites honestly

Every exclusion needs a reason recorded against the site, and the reason has to be about the site rather than about its number.

Defensible reasons include a surveyed site with a materially different parking ratio, a site with an on-site facility the proposal will not have, a survey day affected by weather or roadworks noted in the record, a site whose location type does not match, or a survey period that does not cover the peak hours you need.

Indefensible reasons are all variants of the same thing: the site produces a high rate. Officers have seen this often enough to spot it, and the credibility cost extends to the rest of the assessment. Where a genuine outlier exists, report it, explain what makes it atypical, and show the calculation both with and without it.

Keep the excluded sites in the appendix. A selection sheet showing 22 sites considered and nine adopted, with a one-line reason against each of the thirteen removed, reads as method. A sheet showing nine sites and no exclusions reads as a query that was run until it gave the right answer.

Multi-modal rates and person trips

Vehicle trip rates answer the capacity question. They do not answer the policy question, and the policy question is increasingly what decides applications.

Use the multi-modal data to build a person trip picture: total arrivals and departures by mode, mode share by time period, and the implied number of walking, cycling and public transport trips. That output feeds three things at once. It gives the travel plan a measurable baseline and a defensible target rather than a round percentage. It gives the sustainable transport case real numbers. And it lets you test whether the vehicle rate you have adopted is internally consistent with the mode share you are claiming.

Where the scheme is deliberately low-parking, the vehicle rate from a well-parked sample is the wrong input, and the honest route is to use the multi-modal evidence to justify an adjustment rather than to quietly filter to the three lowest sites. The same discipline applies when evidencing parking provision above or below policy, and it matters especially for the specialised categories covered in trip generation for student and age restricted housing.

Agreeing the method before you run it

Send a short scoping note to the highway authority and ask for written agreement on six points.

1. The land use category and sub-category, with the reason for the choice.
2. The geographical filters, including whether London and Ireland are excluded.
3. The survey date range and the treatment of pandemic-period surveys.
4. The location types accepted and the size band applied.
5. The assessment years, peak periods and the growth factors to be applied.
6. Whether existing site trips are to be discounted, and on what basis.

Two weeks spent on this at the outset routinely saves two months later. Officers are generally willing to agree a method in the abstract and far less willing to accept one presented alongside a conclusion they dislike.

Presenting the audit trail

Append the full selection parameters output rather than a summary of it. The officer wants to see the filters exactly as applied, every site included with its reference and date, every site excluded with its reason, and the resulting rates with the range.

Then add a page of your own that the software does not produce: a short narrative explaining why each filter was set where it was, a table comparing the surveyed sites' characteristics against the proposal, and the sensitivity test. Cross-refer it from the main text of the Transport Assessment so nobody has to hunt for it, and make sure the mode share figures quoted in the Travel Plan come from the same run.

PlanScope drafts Transport Assessments, Transport Statements and Travel Plans structured to UK planning policy, with clear flags where survey data, trip rates and junction modelling still need to be supplied. Councils expect a suitably qualified transport planner to author and check this work, and professional review is strongly advised. Reports are AI-assisted drafts that should be reviewed before submission, and we do not guarantee planning approval. This is general guidance, not legal advice: agree the trip rate methodology with your local highway authority before relying on it.