Earthwork Volume in Excel

A typical earthwork spreadsheet has one row per station. Each row holds a chainage, a cut area and a fill area, and the volume between two neighbouring rows is the average of their areas multiplied by the distance between them. Sum the column and you have the quantity.

That is the average end area method. It is transparent, auditable, and specified by name in most linear-works contracts — anyone can open the sheet, check one row against one section drawing, and follow the arithmetic. None of that is a compromise, and none of it is replaced by software.

What the sheet cannot do is tell you whether its three inputs are right: that the stations are close enough together, that the correction for shape has been applied, and that the section template matches the ground. Each is a separate source of error.

Between two stations the spreadsheet assumes the ground changes linearly. Everything that happens in between — a gully, a rock outcrop, an old bench — is averaged away and never appears in the total.

This is a sampling problem, not an arithmetic one, so no amount of decimal places in the sheet will surface it. The only fix is to measure more sections where the ground changes quickly, which means going back to the survey rather than the spreadsheet.

The practical test is simple: halve the station interval on a stretch you already computed and re-run it. If the total moves materially, your original spacing was too coarse for that ground, and it was too coarse for every similar stretch on the job.

Average end area is exact only when the solid between two sections is a prism — that is, when the cross-section shape stays the same and only scales. Where the section is genuinely changing shape, the method systematically overstates the volume.

The classic case is a transition between cut and fill, where one area shrinks toward zero as the other grows. The correction for this is the prismoidal formula, which weights the mid-section rather than averaging the ends.

In practice many spreadsheets skip the correction entirely, either because the contract does not require it or because nobody added the column. That is a defensible choice — but it should be a choice, not an accident. If the sheet has no prismoidal column, its totals are slightly high wherever the section shape changes.

Cut and fill areas are usually built from a template: a design width, a formation level, and a side slope carried out until it meets existing ground. The point at which the slope daylights determines the area, and therefore the volume.

That intersection depends entirely on the existing ground profile you gave it. Take the ground from a coarse contour interpolation rather than measured points and the daylight point moves; on a steep hillside a small shift in ground level can change the section area substantially, because the slope runs a long way before it catches.

This is why two engineers can compute honestly different quantities from the same alignment and the same template: they seated the side slope on different ground.

All of the above concerns linear work, where an alignment exists and stations mean something. A large class of earthwork has no alignment: a platform, a pit, a borrow area, a stockpile, or the month-to-month difference between two surveys of the same site.

There is no natural place to cut sections through a stockpile, and slicing a pit into stations answers a question nobody asked. These jobs need a surface compared against a surface, cell by cell, which is a different calculation rather than a more precise version of the same one.

The second thing a station table cannot produce is a map. A total quantity tells you how much; it does not tell you where. On a platform that is over-excavated in one corner and short in another, the total can look correct while both corners are wrong — and only a cut/fill map shows it.

None of this argues for abandoning the sheet. The station table is often the deliverable — it is what gets checked, signed and paid against, and replacing it with a single number from a program is a step backwards in auditability.

The workable arrangement is to let the surface produce the sections and let the spreadsheet keep the arithmetic. Sections cut from a measured surface seat the side slope on real ground and can be spaced as tightly as the terrain needs; exported as a station table, they drop into the same sheet and the same checks.

For area work, use a surface-to-surface volume and report the reference you used. For linear work, keep the sheet — and make sure the ground under it was measured, not interpolated.