Flood modelling inputs and what they add to a report

Hydraulic modelling is the reason one Flood Risk Assessment costs a few hundred pounds and another costs tens of thousands. The report structure is the same. What changes is whether somebody has to build a numerical model of the watercourse, and the flood risk assessment modelling cost is driven almost entirely by that single question.

Most sites do not need a new model. A surprising number of assessments include one anyway, because the scope was never tested against what already exists.

When a report needs modelling at all

A Flood Risk Assessment is expected for development in Flood Zone 2 or 3, and for sites of one hectare or more in any zone. That does not mean modelling is expected in all of those cases.

Modelling is needed where the assessment has to answer a question the published data cannot. Typical triggers are a site partly within Zone 3 where finished floor levels have to be set above a defined flood level, a proposal that changes the floodplain by filling or building within it, a watercourse the Environment Agency has not modelled in detail, a culvert or structure whose blockage behaviour matters, or a defended site where the consequence of a breach has to be understood.

Where the site sits in Zone 3 but the Environment Agency already holds detailed modelled levels for that reach, the assessment usually needs those levels rather than a new model. And where the site is in Zone 1 with a surface water risk showing on the published mapping, the answer is normally a drainage design rather than a hydraulic model at all.

Asking the Environment Agency first

The first call is always to the Environment Agency's data team, not to a modeller. The Agency releases flood risk data on request, including the detailed flood risk assessment data package generally known as Product 4, which typically contains modelled flood levels, flows and extents for a range of return periods, defence information, and records of historic flooding for the area requested.

Where the Agency holds a detailed model, it will also consider releasing the model files themselves under licence so a consultant can re-run them for your site rather than starting again. Charges apply to some products, and the schedule changes, so ask for the current one when you make the request.

Lead time matters more than cost here. Data requests are not instant, and a request made after the design freeze has been known to add weeks to a programme. Make the request at the same time as the topographic survey is commissioned.

Reusing an existing model

Reusing a model is by far the least costly route to a defensible answer. What makes it possible:

• The model covers the reach containing your site, at a resolution that resolves the site rather than smearing it across a 50 metre grid.
• The model is recent enough that the channel, defences and upstream development have not changed materially.
• The licence terms permit a third party to run it.
• The consultant has the software the model was built in.

That last point catches people. A model built in one package cannot simply be opened in another, and a conversion exercise can cost more than the run it was meant to save. Ask the question before appointing.

Where reuse works, the remaining effort is usually local refinement: adding the site topography, representing the proposed levels, and re-running the design events. That is a fraction of the cost of a new build.

Building a new model and its cost drivers

Where no usable model exists, the price is set by five things.

1. Hydrology. Deriving flows for an ungauged catchment takes a statistical analysis from the Flood Estimation Handbook or a rainfall runoff approach, and a gauged catchment with a long record is quicker than an ungauged one with none.
2. Survey. Channel cross-sections, structure dimensions, bank levels and floodplain topography all have to be measured. Open LiDAR terrain data covers the floodplain in most of England and removes part of this, but nobody surveys a culvert soffit from an aerial dataset.
3. Model dimensionality. A one dimensional in channel model is quick. A linked one and two dimensional model that routes water across the floodplain costs considerably more and is what most authorities now expect where the floodplain is active.
4. Extent and resolution. A longer reach, a finer grid or a shorter timestep all increase run times, and run times drive the number of scenarios that fit in a programme.
5. Structures and blockage. Bridges, culverts, weirs and defences each need representing, and each blockage scenario is another run.

There is no honest single figure for any of this. Ask for fee proposals that state the number of runs included, the software, the survey assumed and what a further scenario would cost, because that is where variations arise.

Sensitivity and climate change runs

The design events are only part of the work. A report that an authority will accept also includes sensitivity testing: roughness, blockage, defence condition, downstream boundary and inflow hydrograph shape.

Climate change allowances then multiply the runs. The allowances for peak river flow, peak rainfall intensity and sea level rise are published by river basin district and by epoch, with different allowance categories applied according to the vulnerability of the development and its lifetime. They are revised periodically, so check the current allowances rather than copying a figure from an earlier report. Using a superseded allowance is one of the most common reasons an assessment is returned.

Each allowance and each sensitivity test is a model run, a set of outputs and a paragraph of interpretation. That is why a scope of "model the 1 per cent annual probability event" always becomes something larger.

Getting the model accepted

The Environment Agency reviews modelling where it is a statutory consultee, and the lead local flood authority reviews surface water and ordinary watercourse work. Both will ask for a model log or audit trail: the source data, the schematisation decisions, the calibration or verification against observed events where any exist, and the parameter values used.

A model with no verification is not automatically rejected. A model with no explanation of why verification was impossible usually is.

Build review time into the programme. Where the consultee asks for changes, each change is a re-run of every scenario, not just the one queried.

Fixing the scope before you commission

The sequence that keeps the cost down is unremarkable and rarely followed. Check the published flood mapping. Request the Agency data. Establish whether a usable model exists. Ask the lead local flood authority what it expects on surface water. Only then write the modelling brief, and write it with the number of runs stated.

Doing the Sequential Test work early helps too, because a site that cannot pass it does not need a model at all. And the surface water side, set out in a Drainage Strategy, often carries more weight on a small urban site than the fluvial modelling does.

Programme risk is the cost nobody budgets. A data request, a topographic survey and a consultee review in series can occupy several months, in the same way that a missed ecological survey window does, as the note on survey seasons and the cost of missing an ecology window sets out. Run them in parallel where the evidence allows.

PlanScope can draft the assessment structure, the policy framework, the sequential and exception test narrative and the residual risk sections, then flag with a clear marker every place where modelled levels, survey data or an allowance figure has to be supplied. For specialist technical reports, qualified professional input may be required. See PlanScope pricing: one free report, £50 per report, or Professional at £150 per month or £600 per year.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.