Earthwork Cost Per Cubic Yard
(2026) + Swell Factors

Three kinds of cubic yard

Every earthwork estimate lives or dies on understanding which cubic yard measurement you are using. The three states of soil — bank, loose, and compacted — are not interchangeable, and conflating them is one of the most common and expensive mistakes in sitework estimating.

A bank cubic yard (BCY) is soil measured in place, undisturbed. It is the baseline you calculate from survey data and grading plans. A loose cubic yard (LCY) is what that same soil occupies after you excavate it — because digging breaks apart the soil structure and introduces air voids, the volume expands. A compacted cubic yard (CCY) is what results after you place the soil as fill and run a compactor over it: the air voids collapse and volume shrinks below the original bank measurement.

The practical consequence is this: truck hauling is priced in LCY, because that is what fills the truck bed. Fill quantity specifications on drawings are in CCY, because that is what the compacted subgrade measures. Only the earthwork estimate that correctly converts between all three will price the job correctly.

• Bank cubic yard (BCY): soil in place, undisturbed — your takeoff source quantity
• Loose cubic yard (LCY): excavated and swelled in the truck — used for hauling cost
• Compacted cubic yard (CCY): placed and compacted fill — used for fill cost and spec compliance
• Truck hauling is priced in LCY; fill quantities on drawings are in CCY

Swell factors by soil

Swell factor — the percentage by which soil volume increases when excavated — varies significantly by material type. Getting this wrong means under-trucking a job (and paying for re-mobilization) or over-ordering haul capacity and absorbing idle equipment costs.

Sandy and gravelly soils are the most forgiving: they swell only 10–20% because the grains repack relatively tightly even when disturbed. Loam sits in the middle at 20–30% swell. Clay is the estimator's nemesis — 30–40% swell — because its plate-like particles trap air when excavated and are notoriously slow to dry out, adding both volume and weight complications in wet conditions. Rock is the extreme case, at 40–70% swell once blasted or ripped; a cubic yard of in-place granite can fill nearly 1.7 cubic yards of truck bed.

When soil borings are not yet available, the industry convention is to assume a 25% swell factor — meaning 1 BCY becomes 1.25 LCY in the truck. This is a reasonable middle-ground assumption for mixed urban soils, but flag it in your estimate and revisit once geotechnical data arrives.

Shrink and cut/fill balance

While swell governs hauling costs, shrink governs your fill budget. When excavated soil is placed as engineered fill and compacted to specification, it typically compresses to 85–95% of the original bank volume. In other words, a shrink factor of 0.85–0.95 means 1 BCY of cut yields roughly 0.85–0.95 CCY of usable compacted fill. The tighter the compaction specification, the more material you need to import to achieve a given finished grade.

Cut/fill balance is where these factors converge into real money. On a well-balanced site, the material you excavate from high areas (cut) equals — after accounting for swell and shrink — the material you need to fill low areas. Every yard you manage to balance on-site is a yard you are not hauling off to a spoil site or importing from a borrow pit. Both import and export carry trucking, tipping or purchase fees, and mobilization costs that can easily add $15–$40 per cubic yard on top of the base excavation price.

Net import or export is typically the single most expensive line item in an earthwork budget on sites with significant grade change. A preliminary cut/fill analysis, even one based on a rough digital terrain model, is worth doing early — ideally before the project goes to bid — so the design team can adjust finished floor elevations or parking grades to reduce the imbalance before it becomes a contract obligation.

• Compacted fill shrink factor is typically 0.85–0.95
• 1 BCY of cut yields 0.85–0.95 CCY of compacted fill
• Balancing cut and fill on-site avoids import/export haul costs
• Net import or export is the costliest earthwork line item on graded sites

Volume methods and accuracy

Two calculation methods dominate earthwork volume estimation. The average end area method — V = (A1 + A2) / 2 × L — calculates volume between two cross-sections by averaging their areas and multiplying by the distance between them. It is fast, easy to hand-check, and works well along linear corridors like roads, trenches, and utility runs. Its weakness is that it slightly overestimates volumes when cross-sections change shape rapidly, such as at the toe of a slope or around a pond.

For broad grading sites — parking lots, building pads, large industrial slabs — the grid method is more appropriate. Survey crews (or drone photogrammetry) establish a grid of spot elevations across the site; each grid cell becomes a prismatoid whose volume is computed from the difference between existing and proposed grades. Modern software automates this entirely from a surface model, but hand-checking a representative grid cell is still good practice to catch input or import errors.

Regardless of method, preliminary earthwork estimates carry an inherent accuracy of roughly ±10–15%. The uncertainty comes from soil variability, survey precision, and the fact that design documents at the bidding stage rarely show the full subsurface picture. Add a 10–15% contingency to any preliminary earthwork number, and document the assumption clearly so the owner understands the range. That contingency should shrink once geotechnical borings and final grading plans are in hand.

• Average end area method: V = (A1 + A2) / 2 × L — best for linear corridors
• Grid method: spot elevations across a site — best for broad grading pads
• Preliminary earthwork accuracy is typically ±10–15%
• Add a 10–15% contingency until survey and geotechnical data are confirmed

Questions estimators actually ask