Safe access depth and velocity thresholds in flood design

Safe access and egress is demonstrated with a hazard rating derived from flood depth and velocity, not by showing that a route stays dry. The rating used across England comes from the Defra and Environment Agency research on flood risks to people, which combines depth in metres, velocity in metres per second and a debris factor into a single number: the hazard rating equals depth multiplied by velocity plus 0.5, with the debris factor added. That number is then read against four hazard bands. Getting the distinction right between a dry route and a safe route is what usually decides whether the Environment Agency objects.

The reason it matters is that very few urban sites can offer a dry route in the design event. Plenty can offer a safe one.

What safe access requires

Planning Practice Guidance on flood risk and coastal change expects development to be safe for its lifetime, which includes people being able to leave the site, or remain on it safely, during a flood. Three things have to be shown:

• a route from the building to land outside the flood extent, or to a place of safety within it
• depths and velocities along that route in the design event, including the climate change allowance
• a management arrangement, usually a flood warning and evacuation plan, that makes the route usable in practice

The design event depends on the source. Fluvial sites are tested against the 1 per cent annual exceedance probability event with the appropriate climate change allowance, tidal sites against the 0.5 per cent event, and surface water against the relevant return period for the local flood authority. The allowances are revised periodically, so take them from the current Environment Agency allowances rather than from an older report on the same site.

Depth and velocity hazard ratings

The four bands are stable and worth knowing by heart.

The debris factor takes a value of zero, 0.5 or one depending on depth and the dominant land use, with built-up areas attracting the higher values because fences, bins and vehicles become projectiles. Including it is not optional, and reviewers check for it.

Two implications follow. Shallow fast water can be more hazardous than deeper still water, because velocity carries the multiplier. And a route can pass the test at one point and fail 20 metres along it, so the assessment has to report the worst point rather than an average.

Dry access against safe access

Dry access means the route is above the design flood level along its whole length, so the users never enter flood water. Safe access means they do, but at a hazard rating low enough for the user group in question.

The Environment Agency expects dry access for the more vulnerable uses: residential institutions, schools, nurseries, care homes and anything where occupants cannot self-evacuate. For dwellings the position is more pragmatic, and a route with a rating below 0.75 over a short length, supported by a warning and evacuation plan, is commonly accepted. For less vulnerable commercial uses the bar drops again.

Where a scheme needs the Exception Test, the safety limb of that test is the one safe access evidence serves. A scheme that fails on access is unlikely to pass the Exception Test however strong the wider benefits are.

Routes outside the applicant's control

This is the most common practical failure. The safe route leaves the site, crosses a neighbour's access drive or an unadopted lane, and then joins the highway. The applicant cannot deliver works on land they do not control, and the authority cannot condition them.

What works:

1. Identify the route in full and state the ownership of each length, with a plan.
2. Where works are needed on third-party land, show either a legal right already in place or a Grampian condition supported by evidence that consent is reasonably likely.
3. Where no works are needed, say so expressly. A route that is already above the flood level needs no agreement from anyone.
4. Where the route uses adopted highway, confirm the levels from survey rather than from mapping, because kerb and crown levels are what govern depth.

Avoid the fudge of showing a route across land the applicant does not own and leaving the question open. Reviewers look for it.

Vehicular against pedestrian egress

Separate the two in the assessment. Pedestrian egress is what the hazard rating was derived for, and it is the test that applies to occupants leaving on foot. Vehicular egress is a different question, usually asked about emergency service access to the site rather than about residents driving out.

Vehicles lose traction and begin to float at depths well below those that stop a walking adult, and the figures vary with vehicle type, so do not present a single depth as a universal limit. Say what the depth is along the access and let the emergency planning officer form a view. Where the authority asks for emergency vehicle access in the design event, the depths on the approach road are decisive, and that often means a level adjustment at the site entrance rather than anything on the plot.

Where refuge is accepted instead

Sheltering in place is accepted on some sites, usually constrained urban ones where no route exists. It needs more than a first floor. The authority will look for floor space above the design flood level with the climate change allowance and a freeboard, independent access to that level, services that remain functional or a plan for their loss, a means of signalling to rescuers, and an occupant group capable of using it.

Refuge is not available for the most vulnerable uses, and it should not be offered as a first answer. A flood warning and evacuation plan is a management measure that supports safe access. It does not replace it, and the guidance is explicit that an evacuation plan cannot make an unsafe site safe.

Evidence from the hydraulic model

The numbers have to come from somewhere defensible. State the model, its build date, the software, whether it is one-dimensional, two-dimensional or coupled, and who owns it. Where you are using the Environment Agency's model for a watercourse, name the study and the version, and say whether the model was run for your scheme or whether results were read from published mapping.

Report depth and velocity at identified points along the route, as a table with chainages or node references, for each event tested. Give the maximum depth, the maximum velocity and the maximum hazard rating at each point, and note that these maxima do not always occur at the same time step. Where results come from a blockage sensitivity run, the same care applies as when modelling a partially blocked flow control, and the rainfall inputs should be stated as clearly as when choosing a rainfall dataset for a drainage calculation.

One honest difficulty: velocities on floodplain routes from a coarse two-dimensional grid are sensitive to mesh resolution, and a reviewer who refines the grid can get a different hazard band. Say what resolution you used and test a finer one if the result sits near 0.75.

A Flood Risk Assessment that sets out the route, the model source, the tabulated depths and velocities and the hazard ratings answers the Environment Agency's standing questions in one pass. PlanScope drafts the policy and structural parts of that document and flags the site data it does not have, which is where the modeller's output goes. Reports should be reviewed before submission and PlanScope does not guarantee planning approval.

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