A geotechnical calculation report is the document that shows how a design number was produced: the method applied, the input values used, the intermediate steps, and the result. It differs from a summary report in one respect, and the difference decides whether anyone can confirm the work.
A summary states that the allowable bearing pressure is a certain value. An engineering calculation report states which bearing capacity equation was used, which factors were applied, which layer supplied each parameter, and what the intermediate resistance was before the factor of safety reduced it.
Both documents contain the same final number. Only one of them can be disagreed with.
A result without its derivation is not a calculation. It is a claim about a calculation.
What Has to Be Visible for a Result to Be Checkable?
Verification is not reading the conclusion and judging whether it looks reasonable. It is reproducing the path from input to output closely enough to find a mistake if one exists.
| Element | Why the checker needs it |
|---|---|
| Method and source | Terzaghi, Meyerhof and Vesić give different answers for the same soil |
| Input parameters with layer reference | A value is meaningless without knowing which layer it came from |
| Equation as written, not named | Shape, depth and inclination factors vary between formulations |
| Intermediate values | Where the error is, if there is one |
| Factor of safety or partial factors | Distinguishes ultimate resistance from design value |
| Assumptions and their justification | The part most often carried silently |
| Units at every step | The single most common source of order-of-magnitude error |
The list is short. What makes it demanding is that all seven have to be present at the same time, for each calculation, not once in a methodology chapter at the front of the document.
Which Calculations Need to Be Shown in Full?
Not every number needs a full derivation. The rule that works in practice is proportionality to consequence.
Bearing capacity, settlement and stability govern whether the structure performs, so they are shown completely. Parameter derivation is shown where a correlation was applied rather than a value measured, because the correlation is itself an assumption. Anything governed by a code check — embedment, anchor capacity, strut design — is shown with the clause reference, since the checker is verifying compliance rather than arithmetic.
Routine conversions and interpolations can be tabulated rather than derived. What cannot be compressed is the step where judgement entered.
How Does a Checker Actually Verify a Number?
By recalculating it with a different tool.
An independent check that reuses the original software reproduces the original assumptions along with the original arithmetic. If the input is wrong, or the method inappropriate for the soil, the second run confirms the first and both are wrong. This is why a hand calculation or a spreadsheet remains the standard verification instrument even in offices with substantial software budgets.
For that to be possible, the report has to expose the inputs in a form that can be typed into something else. A result presented only as software output forces the checker either to accept it or to rebuild the entire model, and in practice the first is what happens.
Why Do Summary Reports Pass Review and Still Fail Later?
Because review of a summary is a review of plausibility, and plausibility is a weak filter.
A settlement of 25 mm looks reasonable for a raft on medium-dense sand. It also looks reasonable if the compressible layer below was omitted from the profile, if the stress distribution stopped short of it, or if the modulus came from a correlation outside its valid range. The number passes because nothing in the document contradicts it.
The failure appears during construction or afterwards, when the assumption that was never written down turns out not to hold. At that point the report cannot be used to establish what was assumed, because it did not say.
What Does a Design Verification Report Add?
A design verification report is the separate document produced when an independent party checks the design. It records what was checked, by which method, and where the two sets of results agree or differ.
Its value depends entirely on the underlying calculation report. Where the original shows its equations, verification is a focused exercise: the checker recalculates the governing cases and compares. Where the original shows results only, verification degenerates into either a full independent design or a signature.
The cost difference between those two outcomes is large, and it is decided by how the first report was written.
Where Do Assumptions Hide?
| Assumption | Where it usually goes unstated |
|---|---|
| Drained or undrained condition | Implied by the parameter used, rarely written |
| Groundwater level adopted for design | A single figure, no seasonal or artesian note |
| Layer boundaries between boreholes | Interpolation presented as profile |
| Characteristic value selection | Table of values with no derivation |
| Rigid or flexible foundation behaviour | Determines whether pressure distribution is uniform |
| Construction sequence | Affects staged excavation and consolidation timing |
| Correlation validity range | Cited by author name, not by applicable soil type |
None of these is exotic. Each one changes a result, and each is easier to leave implicit than to state.
Is a Spreadsheet a Valid Independent Check?
Yes, and it is often the better one.
A spreadsheet built from the code equation makes different assumptions visible because they have to be typed in explicitly. It also forces the checker to read the report closely enough to extract the inputs, which is itself a review.
Its weakness is version control. A spreadsheet checked once and then reused across an office becomes an unverified tool the moment someone edits a cell, and the edit is rarely documented. Used as a one-off check against a documented calculation, it is sound. Used as the office standard method, it needs the same governance as any other tool.
When Is a Full Calculation Report Not Required?
For preliminary sizing, feasibility work and internal option comparison, a calculation set with its assumptions noted is proportionate. The document becomes necessary when the result leaves the office: when it is issued to a client, a checker, a contractor or an authority, and when someone else will rely on it.
The practical test is whether the number will be used by a person who cannot ask the author a question. If it will, the derivation travels with it.
What Does a Worked Bearing Capacity Check Contain?
A single check is enough to show what transparent design calculations require in practice. Taking the ultimate bearing capacity of a shallow foundation, the expression evaluated is stated before any number appears:
qult = c · Nc · sc · dc + q · Nq · sq · dq + 0.5 · γ · B · Nγ · sγ · dγ
where c is the cohesion of the bearing layer in kPa, q the overburden pressure at founding level in kPa, γ the unit weight below the foundation in kN/m³, B the foundation width in m, Nc, Nq and Nγ the bearing capacity factors, and s and d the shape and depth factors applied to each term.
Seven things follow from writing it this way. The source of the factors is fixed, because Terzaghi, Meyerhof and Vesić tabulate them differently. The cohesion carries a layer reference, so the reader can see which stratum supplied it and whether a drained or undrained value was used. The overburden term states the founding depth adopted and the groundwater level behind it. Each shape and depth factor appears with its own value rather than being absorbed into a single coefficient. The result is given as an ultimate resistance before reduction. The factor of safety or the partial factor set is applied in a separate, visible step. And the units carry through every term.
A checker holding this can reproduce the number in a few minutes with a spreadsheet. A checker holding only the allowable pressure cannot reproduce anything, and the difference between those two situations is roughly the difference between a review and a signature.
Common Errors in Calculation Reporting
| Error | Consequence |
|---|---|
| Method named but equation not written | Checker cannot tell which formulation was used |
| Parameters listed without layer reference | Values cannot be traced back to the ground model |
| Intermediate values omitted | An error can be detected but not located |
| Assumptions stated once in a methodology chapter | Reader cannot tell which apply to which calculation |
| Factor of safety applied without stating where | Ultimate and design values become indistinguishable |
| Units dropped in intermediate steps | Order-of-magnitude errors survive review |
| Software output pasted without inputs | Verification becomes a repeat of the whole design |
Where SETAF2018 Fits
SETAF2018 treats the calculation report as an output of the analysis rather than as a document written afterwards, and that is the reason the equations are available at all.
Three report types come from the same model. The general report gives the project-level summary with tables and graphs. The local report presents a single calculation in full, with the equation as written, every symbol defined, the input values with their layer references, the intermediate results and the final value. The templated geotechnical report assembles the complete document with company identification and cadastral data.
The local report is the one that matters for verification. Because it prints the expression rather than a method name, a checker can confirm a bearing capacity value against the same equation in a spreadsheet, or confirm a settlement against a hand calculation, without rebuilding the model. The same applies to deep soil mixing mixture design, where the water and cement quantities are issued with the equations that produced them, which is the form a contractor needs at the mixing plant.
Because analysis, drawings, quantities and reporting all read the same model, a revision propagates rather than requiring the document to be reassembled. Output is produced in Excel, Word and PDF, and the full set of calculation reports sits alongside the foundation analysis and excavation support workflows.
The design checks follow EN 1997 Design Approach 2, TBDY 2018, TS 500 and FHWA where each applies, and the clause reference appears with the check rather than in a separate compliance statement.
Frequently Asked Questions
What is a geotechnical calculation report?
It is the document that shows how a design value was produced: the method and its source, the input parameters with their layer references, the equation as written, the intermediate steps, the factors applied and the final result. It differs from a summary report in showing the derivation rather than only the conclusion.
What is the difference between a calculation report and a design verification report?
A calculation report is produced by the designer and shows how the design values were derived. A design verification report is produced by an independent checker and records what was checked, by which method, and whether the results agree. The second depends on the first being complete enough to check.
Why does a calculation need to show its equation?
Because named methods have multiple formulations. Terzaghi, Meyerhof and Vesić bearing capacity factors differ, and the shape, depth and inclination factors applied vary between references. A method name alone does not tell a checker which expression was evaluated.
Can a spreadsheet be used to check software output?
Yes, and it is often preferable, because a check made with a different tool exposes assumptions that a second run of the same software would reproduce silently. The spreadsheet needs to be treated as a one-off verification instrument rather than an undocumented office standard.
What should be written down as an assumption?
Drainage condition, design groundwater level, layer boundaries adopted between boreholes, how characteristic values were selected, whether the foundation is treated as rigid or flexible, the construction sequence assumed, and the validity range of any correlation used.
Are transparent calculations required by code?
Design standards require that designs can be verified, and verification requires access to the method and assumptions. The form is not usually prescribed, but a report that cannot be independently checked does not satisfy the intent even where it satisfies the checklist.
How detailed should a preliminary calculation be?
Proportionate to its use. Feasibility and option comparison need the assumptions noted rather than full derivations. The full report becomes necessary once the number leaves the office and will be relied on by someone who cannot ask the author a question.
Does showing the calculation make revision easier?
Yes. A report with visible intermediate values can be updated in the affected layers when loads or geometry change. A report presenting final results only has to be recalculated from the beginning, which is why summary reporting is expensive over the life of a project rather than cheap.
