Limitations and assumptions sections that protect a report
Reviewers read the limitations section first, because it tells them what the conclusions are worth. Specific, short and uncomfortable in places beats a page of boilerplate.
Reviewers read the limitations section first. Not out of suspicion, but because the limitations section of a technical report tells you what the conclusions are worth, and a conclusion whose basis is not stated cannot be relied on by anyone. In a planning context that section is also where professional liability is defined, which is why a thin one is a problem for the author as much as for the reader.
A good limitations section is specific, short and uncomfortable in places. A bad one is a page of boilerplate that would fit any report on any site.
What the section is actually for
Three jobs, and they get confused.
The first is to record the factual boundaries of the work: what was inspected, when, by whom, with what equipment, and what was not reached. That is information, not protection.
The second is to state the assumptions on which the conclusions rest, so that a reader can tell whether the conclusion survives if an assumption turns out to be wrong.
The third is to define the scope of the author's duty, which is a contractual matter and belongs in a separate short paragraph rather than mixed in with the technical limitations.
Keeping those three separate is the single improvement most reports could make. A paragraph that slides from "the roof void was not accessible" into "this report is for the sole use of the client" in the same breath reads as evasive even where both statements are correct.
Access constraints and what was not inspected
Say what you could not see, where it was, and what difference it might make. The last clause is the one that is usually missing.
For a building survey that means naming the rooms and voids not entered, the areas with fixed finishes, the parts obscured by stored goods, and the elements below ground. For an ecological appraisal it means the parts of the site not walked, the buildings not internally inspected, the trees not climbed and the land in third-party ownership. For a ground investigation it means the positions that could not be reached because of services, hardstanding or access width, and the depths not achieved because of refusal.
Then add the consequence. "The roof void above the rear wing was not accessible and no conclusion is reached on the condition of the timbers in that area" is useful. "Inspection was visual and non-invasive" on its own is not, because it does not tell the reader which part of the conclusion is affected.
Where an access constraint can be removed by further work, say what that work is and recommend it. A reviewer asked to sign off a Preliminary Ecological Appraisal that notes an inaccessible outbuilding without recommending its inspection will add that recommendation themselves.
Seasonal and weather limitations
Season decides what can be found, and in ecology it decides whether a survey has any value at all.
Bat emergence surveys sit in the warmer months and a preliminary roost assessment outside that window cannot confirm absence. Great crested newt environmental DNA sampling has a defined spring window. Botanical surveys need the growing season. Reptile surveys work in spring and early autumn rather than high summer. Wintering bird surveys cover the colder months. Tree inspections can be done year round, but a crown condition assessment is compromised out of leaf.
In other disciplines the effect is subtler and still real. A noise survey carried out in a school holiday does not represent the term-time environment. A traffic count in August or late December is not a neutral month. A ground investigation in a dry summer may miss the groundwater level that governs the design. A daylight model is run at a stated date and the result depends on it.
State the season, the date and the weather, and state the consequence. Where a survey falls outside the optimal window, a precautionary conclusion is legitimate and far better than a confident one, as long as it is labelled as precautionary.
Data sources and their vintage
Every report leans on third-party data, and most of it ages. List each source with the date it was supplied or published, not the date you downloaded it.
| Source | What ages | Practical effect |
|---|---|---|
| Historic Environment Record extract | Records added since the search | New designations or events missed |
| Historic mapping and aerial photography | Survey dates, gaps between editions | Short-lived structures or workings not shown |
| Flood mapping and modelling | Model vintage and climate allowances | Assessment may not reflect current allowances |
| Protected species records | Records more than a few years old | Absence of a record is not absence of the species |
| Utility asset plans | Date of issue and plan accuracy | Position and depth are indicative only |
| Record drawings of a building | Whether as-built or as-designed | Structure may differ from the drawing |
The honest statement about biological records is worth making explicitly: the absence of a record means the absence of a record, not the absence of a species. Reports that treat a nil return from a records centre as evidence of absence get challenged, and rightly.
Assumptions that would change the conclusion
Separate the assumptions that are immaterial from the ones that carry the conclusion. Only the second group belongs in a prominent list.
A useful test: if this assumption were wrong, would the recommendation change? If yes, state the assumption, state what the conclusion would become, and say what evidence would resolve it.
Examples that recur. A structural appraisal assuming a foundation depth and type that was never exposed. A contamination risk assessment assuming an end use with a garden, where a later layout change removes it. A transport assessment assuming a trip rate from a database site that is not quite comparable. A drainage design assuming an infiltration rate from a single test pit. An energy strategy assuming an electrical connection capacity that has not been confirmed.
Each of those is reasonable to assume and dangerous to leave unstated.
Where a reviewer reads this section first
A reviewer signing their name to a document is accepting responsibility for its conclusions, and the limitations section tells them what they are accepting. In practice they check four things: whether the stated scope matches what the council asked for, whether the fieldwork was adequate for the conclusion, whether the assumptions are visible and sensible, and whether the recommendations follow from the limitations rather than ignoring them.
That last point catches a lot of drafts. A report that lists four access constraints and then recommends no further work is internally inconsistent, and a reviewer will send it back.
For AI-assisted drafts the same logic applies with more force. A draft that inserts a clear marker where a site-specific fact is missing is honest about its limits, and the author's job is to replace each marker with evidence rather than to delete it. The practice of taking ownership of a draft is set out in signing off an AI-drafted report as the named author, and where the work is split between firms, sub-consultants and where liability sits is the companion piece.
Write the limitations section before the conclusions, not after. It changes what you are willing to conclude, which is the point. For the report types where a named professional is expected, browse the full set.
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. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.
Frequently asked questions
What belongs in a limitations section?
Three separate things: the factual boundaries of the work, meaning what was inspected and what was not; the assumptions the conclusions rest on; and a short statement of the scope of the author's duty. Keeping them in separate paragraphs matters, because mixing technical limits with contractual wording reads as evasive.
Is it enough to say that inspection was visual and non-invasive?
No. That sentence tells the reader nothing about which part of the conclusion is affected. Name the rooms, voids, areas or depths not reached, and state the consequence, such as that no conclusion is reached on the condition of a particular element. Then recommend the work that would resolve it.
Does a nil return from a records centre prove a species is absent?
No. The absence of a record means the absence of a record. Local records are patchy and depend on where people have looked, so a nil return supports a precautionary approach rather than a conclusion of absence. Reports that treat it as evidence of absence are routinely challenged by consultees.
How should seasonal constraints be reported?
State the season, the dates and the weather, then state the consequence. Where a survey falls outside the optimal window, a precautionary conclusion clearly labelled as such is far stronger than a confident one. The same applies outside ecology: a noise survey in a school holiday or a traffic count in August is not representative.
Which assumptions should be listed prominently?
Only those that carry the conclusion. Apply a simple test: if this assumption were wrong, would the recommendation change? Where the answer is yes, state the assumption, state what the conclusion would become, and name the evidence that would resolve it. Immaterial assumptions can stay in the methodology.
Sources and further reading
This article is general guidance about the planning system in England and, where stated, the rest of the UK. It is not legal or professional advice. Check your local planning authority’s validation requirements and take professional advice for your specific site.