Designing an archaeological trial trenching evaluation

An archaeological trial trenching sample is normally expressed as a percentage of the site area, and county archaeologists in England commonly specify between 2 and 5 per cent, rising where the desk-based evidence points to dense remains. The percentage is a convention rather than a national standard, so the number in your brief has to be justified by the evidence rather than applied by habit. What decides whether an evaluation works is the design: where the trenches go, how they relate to the geophysical survey, and whether the layout can actually be dug given the services and access on site.

Why evaluation precedes the decision

Evaluation exists to answer a question the desk-based assessment could not: is anything there, what is it, how old, how extensive, how well preserved, and how much does it matter.

The Framework asks authorities to require developers to submit an appropriate desk-based assessment and, where necessary, a field evaluation, where a site includes or has the potential to include heritage assets with archaeological interest. "Where necessary" is doing real work in that sentence. On a site with strong cropmark evidence or a known monument nearby, evaluation before determination lets the decision be taken with the facts available and usually produces a narrower condition.

Deferring everything to a condition has a cost. If significant remains turn up after permission, the options are a layout change that the permission does not allow, or full excavation at a price nobody budgeted. The argument for front-loading is strongest where the desk-based evidence is positive and the layout still has flexibility.

A sound archaeological desk-based assessment is the foundation for all of this. It should pull the historic environment record entries, the national heritage list, aerial photography and lidar, historic mapping, geological mapping and any previous investigations nearby, and set out a statement of potential by period.

Sample strategies and how they are justified

The percentage is calculated on the area to be developed rather than the whole red line, which matters on a site with large retained areas.

Trench dimensions follow the machine. A standard trench is 1.8 to 2 metres wide, matching a toothless ditching bucket, and 20 to 50 metres long. Longer trenches cross more features and give better stratigraphic context. Short trenches scattered widely cover more of the site statistically and see less of any single feature.

Justify the figure in the written scheme of investigation by reference to the evidence, not to precedent. Where the brief specifies 5 per cent on a site with negative geophysics and no historic environment record entries within 500 metres, that is a conversation worth having with the archaeological advisor before the work is priced. The reverse applies too: a 2 per cent sample on a site with a Roman road along one boundary is unlikely to be defensible.

Targeted against systematic trenching

A systematic array, usually a staggered or herringbone grid, spreads trenches evenly and tests the whole area without preconception. It is the right approach where there is nothing to aim at, and it gives the clearest statistical basis for a statement about absence.

Targeted trenching puts trenches across identified anomalies to characterise and date them. It is far more efficient at telling you what a feature is and far worse at telling you what else is on the site.

Most briefs now ask for both: trenches targeted on the anomalies, plus a systematic element across the apparently blank areas. Insist on the second part. Geophysics misses a great deal, and a trenching exercise consisting only of targeted trenches cannot support any conclusion about the rest of the field.

Three design points are worth settling before mobilisation:

• Trench orientation. Trenches set at an angle to the expected feature alignment cross more features than trenches running parallel to them.
• Edge coverage. Features cluster on slopes, springlines and the margins of former watercourses, and a grid that stops short of the boundary misses them.
• Reserve trenches. Allow two or three in the scheme so that a feature found at the end of a trench can be followed without a formal variation.

Geophysics and its limits as a precursor

Magnetometry is the default technique on open ground in England. A gradiometer survey at 1 metre traverse separation with closely spaced readings along the line will pick up ditches, pits and kilns on chalk, gravel and most sedimentary geology.

Its limits are well known and routinely forgotten.

• It performs poorly on clay soils and over igneous or strongly magnetic geology, where the natural background swamps the archaeology.
• It is defeated by modern ferrous material, so a field edged with wire fencing, a former farmyard or a site with buried services produces unusable strips.
• It is weak on stone structures and on features with little magnetic contrast, including some burials and many timber buildings.
• It cannot see through alluvium, colluvium or made ground of any depth, which is why urban and floodplain sites need a different approach entirely.

Earth resistance survey complements magnetometry on masonry and on trackways. Ground-penetrating radar earns its place on hard surfaces and in urban contexts, and it produces depth information the other techniques do not.

On deeply stratified urban sites the useful tool is deposit modelling built from borehole and geotechnical data rather than any surface survey, and the approach taken on deep urban archaeology is the right starting point there.

A negative geophysical survey is not a reason to reduce the trenching sample on its own. It is a reason to shift the balance from targeted to systematic trenches.

Access, services and constraints on layout

Evaluation designs fail on practicalities more often than on archaeology.

Before the trench plan is drawn, overlay the service drawings and commission utility searches. Scanning with a cable avoidance tool is a minimum, and trial holes by hand are needed where a main is suspected. Overhead lines impose clearance distances on machine operation, and a 360-degree excavator with the boom up does not fit under an 11 kV line.

Then deal with the ground. Standing crops may mean waiting for harvest or paying compensation. Livestock need the trenches fenced. Topsoil and subsoil must be stored separately and reinstated in order, which needs space the plan has to allocate. Winter work on clay produces poor visibility in the trench base and a higher risk of misinterpretation, and a frozen or waterlogged site is a false economy.

Health and safety is a design constraint too. Trenches deeper than about 1.2 metres need stepping, battering or shoring, and the Construction (Design and Management) Regulations 2015 apply to the work. Deep trenching in confined areas can be the reason a trench moves.

Interpreting a negative result

Absence of evidence is not evidence of absence, and the report should say which of the two it is claiming.

A truly negative evaluation on a well-designed systematic sample across free-draining geology, with good trench visibility and no masking deposits, supports a conclusion that the site has low potential. The same result on clay, in February, with 2 per cent coverage and half the trenches displaced by services, supports nothing much.

Four questions to put to the report:

1. What proportion of the planned sample was actually achieved, and where were the gaps?
2. What were the ground conditions and the visibility in the trench base?
3. Was there truncation, and how deep was the ploughsoil?
4. Were there masking deposits, and was their depth established?

Where the answers are weak, expect the archaeological advisor to ask for additional work rather than accept the conclusion, and build that possibility into the programme.

Reporting and the next decision

The evaluation report should carry the trench plan as dug, trench by trench descriptions, section drawings, a finds assessment, environmental sample results and a statement of significance with a recommendation.

Three outcomes follow. No further work, with the condition discharged. Preservation in situ, achieved by moving buildings, roads or drainage away from the remains, which is the least costly outcome when the layout can still move. Or preservation by record, meaning excavation of the affected area before construction, priced from the evaluation results.

That third outcome is where an evaluation earns its fee, because excavation costed from known densities is a very different proposition from excavation costed from a guess. A heritage statement should then carry the evaluation findings into the planning balance, and an archaeological field evaluation draft gives the structure for the written submission. Councils normally expect a suitably qualified author, and professional review before submission is strongly advised.

See which heritage documents a site needs by browsing the report categories and starting a draft.

This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.