Ground is the only material on a project that nobody chose. It was there first, it varies between boreholes, and everything above it is designed on the strength of a few dozen samples and a professional judgement. That judgement is expressed as a handful of numbers in a table, and those numbers will be used for years by engineers who never read the report they came from.

We edit what geotechnical engineers produce — ground investigation specifications and scopes, factual reports and their limitations, interpretative reports and design parameter selection, foundation options appraisals and recommendations, slope stability and earthworks assessments, retaining structure design reports, settlement and heave assessments, ground improvement and piling specifications, contaminated ground and geoenvironmental interpretation, instrumentation and monitoring plans, construction stage advice and ground risk registers, and expert reports on ground-related disputes. Our editors work on the table everyone downstream copies.

Design parameter selection is where geotechnical engineering condenses into something other people use, and the risk is that a range of test results becomes a single number with the reasoning left behind. A cohesion value in a table is treated as a measurement by every engineer who inherits it, when it is usually a cautious estimate derived from scattered data by someone weighing the failure mode it will be used for. We write these so each parameter states what it was derived from — which tests, how many, their spread, and which results were set aside and why; so the basis of selection is named, whether a cautious estimate of the mean or a lower bound, because the appropriate choice differs between a settlement calculation and a stability analysis and a single table cannot serve both silently; so the stratum the parameter applies to is defined in terms someone can identify in a trial pit, rather than by a name that means different things to different logging engineers; so the depth range and any variation across the site are stated instead of averaged into a single value, given that a site with soft alluvium in one corner is not a site with uniformly moderate ground; so the report says explicitly which parameters are conservative for which use and would be unconservative for another; and so the limitations of the investigation are carried into the parameter table rather than left in a section at the front. Reports written this way are still safe to use in year seven.

Everything you send is treated in confidence, including site data, investigation results and client information. We are editors rather than geotechnical engineers, and we offer no view on parameters, ground conditions or design. What we can do is keep the reasoning attached to the numbers people will lift out of the table.