Assumptions registers in a basement impact assessment

Every ground movement prediction in a basement impact assessment rests on a list of assumptions, and the assumptions register is where you write them down so that a reviewing engineer can test them. A council asking for basement impact assessment assumptions to be set out is not checking your arithmetic. It is checking whether the inputs you chose are defensible, and whether you have said what happens if one of them turns out to be wrong. A register that names each assumption, gives its source, states how sensitive the result is to it and says when it will be confirmed is worth more to a submission than another twenty pages of modelling output.

Why the register exists

A Basement Impact Assessment is a planning document written before anyone has dug a trial pit on the party line or opened up a neighbour's foundation. The analysis therefore runs on judgement. Soil parameters come from a desk study and a handful of boreholes. Groundwater comes from a short monitoring period, often in the wrong season. The construction sequence comes from a contractor who has not been appointed.

None of that is a defect. It is the normal condition of planning-stage work. The defect is leaving it implicit, so that the reader cannot tell which numbers are measured and which are chosen.

Reviewing engineers appointed by local authorities tend to work through a basement report looking for exactly this. They want to find the sentence that says which undrained shear strength was adopted and why, or they want to establish that no such sentence exists. The second outcome produces a request for further information and a delay of weeks.

A useful register has five columns and no commentary.

Ground and groundwater assumptions

Start with the stratigraphy. State the assumed thickness and level of each layer, say which borehole or window sample it came from, and say what you have assumed between boreholes. On a narrow urban plot with two holes, the interpolation across the party wall is an assumption, not data.

Then the parameters. Undrained shear strength, drained strength, stiffness and the stiffness profile with depth all matter to a ground movement prediction, and all of them are usually derived rather than measured directly. Say which correlation you used. Say whether stiffness was taken at small strain or operational strain, because the difference moves the predicted settlement more than most other single input.

Groundwater is the assumption that most often fails. A standpipe read twice over three weeks in July tells you very little about the winter piezometric level, and a basement designed to the summer reading will leak or float. Record the monitoring period, the number of readings, the season, and the design level you adopted with the margin you added. Perched water in made ground sitting above a clay is a separate case and should be recorded separately from the main aquifer.

If the site sits over a permeable deposit that may be in continuity with a watercourse or with the deep aquifer, say so and say what you have assumed about drawdown during construction. A geotechnical interpretative report, which a Geotechnical Assessment draws on, is the right place for the derivation; the planning document should carry the conclusion and the uncertainty.

The construction sequence the analysis assumes

Sequence drives ground movement as much as geometry does. A top-down sequence with props cast as the slabs go in produces different wall deflection from a bottom-up sequence with temporary raking shores, and underpinning in short alternate bays produces different movement again from a secant piled wall installed from existing ground level.

So the register has to state the sequence actually modelled. That means the bay widths, the maximum unsupported excavation depth at any stage, the prop levels, the preload if any, and the assumed stiffness of the temporary works.

This is the assumption most likely to change, because it belongs to a contractor who will arrive after permission is granted. The honest treatment is to state the modelled sequence, state that the permanent works design assumes movement no greater than the prediction, and state that any alternative sequence must be shown to produce equal or lower movement before it is adopted. That sentence is what allows a condition to be discharged later without a fresh application.

Note: a sequence that nobody will build is worse than a conservative one. Modelling 1.2 m bays when the site has no room for the plant to work that way produces a number that will not survive contact with the site.

Adjacent structures, foundation depth and condition

The prediction is only meaningful against a known neighbour. Record the assumed foundation type and depth for each adjoining building, the assumed construction of the walls, and the assumed existing condition.

Where you have no access, say that. An assumed 450 mm deep brick footing on a Victorian terrace, inferred from the period and from one trial hole on your own side, is a legitimate planning-stage assumption, and it is also a material one, because founding depth relative to the excavation changes the movement a building sees.

Existing condition matters for a different reason. A building already showing category 2 cracking has less tolerance left than the Burland categories imply, and a pre-construction condition survey is the only way to establish the starting point. Where the scheme involves demolition or strip-out of the host building, the record you assemble for a pre-redevelopment audit will often turn up structural information that belongs in this part of the register.

Buried services are the neighbour nobody surveys. Sewers, water mains and cables under the pavement sit within the influence zone of most basement excavations, and their condition and depth are assumptions until the undertaker's records are checked and a trial hole is dug. The same exercise sits behind the utilities capacity evidence a residential scheme needs.

Damage category criteria and what they mean

Predicted movement is translated into a damage category, usually using the classification set out in BRE Digest 251 and the Burland and Wroth framework: category 0 negligible, 1 very slight, 2 slight, 3 moderate, 4 severe and 5 very severe. Categories 0 to 2 are generally treated as aesthetic, 3 and above as affecting serviceability or stability.

State which classification you applied and at which point on the building you assessed it. Hogging over a corner gives a different answer from sagging at mid-span, and an assessment that reports one figure for a whole terrace has averaged away the case that matters.

Where a local plan policy or a supplementary planning document sets an acceptable limit, quote the wording that applies to your site rather than assuming a number, because the limit and the way it is expressed vary between authorities.

Monitoring thresholds that follow from the assumptions

Trigger levels are not arbitrary numbers chosen for a monitoring specification. They should be derived from the prediction, so that green, amber and red thresholds correspond to movement consistent with the assumed behaviour, movement exceeding it, and movement approaching the next damage category.

Set out the derivation in the register: predicted movement at each monitored point, the amber level and what it triggers, the red level and what stops. Also state the instrumentation assumed, the reading frequency and the baseline period, because a trigger level without a baseline is unenforceable.

What gets revisited at detailed design

Close the register with a short list of the assumptions that must be retested once access and a contractor exist. Typically that means winter groundwater levels, the actual foundation depth and condition of each neighbour, confirmation of stratigraphy on the party line, and the contractor's own temporary works design.

That list is also the natural basis for the wording of a pre-commencement condition, which is why an officer reads it closely. Giving them the list rather than making them write it tends to produce a condition you can live with.

A Basement Impact Assessment is a report councils expect to see from a suitably qualified author, usually a chartered geotechnical or structural engineer, and professional review before submission is strongly advised. PlanScope can draft the narrative and the register structure, and it flags any missing site-specific input for confirmation rather than inventing one. Browse all 66 PlanScope report types and start a draft.

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