The field density test on the pad passed at 96 percent of maximum dry density. The stockpile that was supposed to finish that pad is gone. The subgrade is still about 150 mm below the underside of the gravel bedding, it is Saturday afternoon, and the site engineer is phoning a hauler for six loads of selected fill that appear nowhere in the bill of quantities. On Monday the owner's representative will point to the BOQ line "backfill from excavated material" and ask why fill is suddenly a variation.
The cause is rarely bad compaction or a careless operator. The take-off carried one volume column when earthwork has three. Soil in the ground, soil in the truck and soil under the slab are three different volumes of the same material, and a bill that does not keep them apart balances on paper and fails in the field.
Every cubic metre of soil on the job exists in one of three states, each with its own volume:
The conversions come from dry density, because the mass of soil solids does not change when you dig. Swell is the bank dry density divided by the loose dry density, minus one; shrink is one minus the bank dry density divided by the compacted dry density. So loose volume = bank × (1 + swell) and compacted volume = bank × (1 - shrink). Equipment and estimating handbooks quote common earth at roughly 20 to 30 percent swell and 10 to 15 percent shrink. Those are starting points, not code values; a field density test on your own material supersedes them.
Three more facts compound the error. Concrete displaces backfill, so footing, pedestal and lean concrete come off the pit volume before you count what goes back in. Working space and batter are real dirt that must be dug, hauled and replaced, even where the contract pays only to neat lines. And the supplier sells fill by the loose truckload while the specification measures it compacted in place, so the two numbers cannot be compared without a factor between them.
A two-storey commercial building, footprint 12 m × 20 m = 240 m², on twelve isolated footings F1, each 1.8 m × 1.8 m × 0.35 m on 50 mm lean concrete. Underside of lean concrete is 1.50 m below the stripped datum. Pedestals are 0.40 m × 0.40 m from top of footing up to the datum. Finished floor level is +0.45 m above original grade; the slab is 100 mm on 50 mm gravel bedding. Topsoil strip is 150 mm over the footprint plus a 1.0 m margin all round.
Assumptions, all illustrative: 0.30 m working space each side (common practice; some specifications require 0.50 to 0.60 m where formwork is used); vertical pit sides, on the basis that the soil report and safety plan permit a short-open-time vertical cut at this depth (note that 1.50 m sits right at the threshold depth above which widely used excavation-safety rules require sloping, benching or shoring, so it is a competent-person determination on site that makes a vertical cut permissible here, not the depth alone; do not carry this assumption into deeper pits); silty clay spoil with 25 percent swell and 10 percent shrink, so loose = 1.25 × bank and compacted = 0.90 × bank; imported selected fill at 1.30 m³ loose per 1.00 m³ compacted; dump trucks rated at 12 m³ loose (ten-wheelers in the Philippines are commonly quoted anywhere from 10 to 14 m³ struck or heaped, so check your hauler).
Strip area = (12 + 2 × 1.0) × (20 + 2 × 1.0) = 14 × 22 = 308 m². Volume = 308 × 0.15 = 46.2 m³ bank, hauled off.
Pit plan = 1.8 + 2 × 0.30 = 2.4 m square, area 2.4 × 2.4 = 5.76 m². Volume per pit = 5.76 × 1.50 = 8.64 m³ bank. Twelve pits: 8.64 × 12 = 103.68 m³ bank.
Had a 1H:2V batter been required instead, the top of the pit would be 2.4 + 2 × 0.5 × 1.5 = 3.9 m square, area 3.9 × 3.9 = 15.21 m², and the frustum volume would be (1.5 ÷ 3) × (5.76 + 15.21 + √(5.76 × 15.21)) = 0.5 × (5.76 + 15.21 + 9.36) = 0.5 × 30.33 = 15.17 m³ per pit, about 75 percent more. That is why the side-slope assumption must be written into the estimate.
Lean concrete, taken over the full 2.4 m pit plan (state this in your own take-off; blinding only to the footing plus 100 mm each side would give 2.0 × 2.0 × 0.05 = 0.200 m³ and shift every figure after it): 2.4 × 2.4 × 0.05 = 0.288 m³. Footing: 1.8 × 1.8 × 0.35 = 1.134 m³. Pedestal: top of footing is 1.50 - 0.05 - 0.35 = 1.10 m below datum, so 0.40 × 0.40 × 1.10 = 0.176 m³. Total = 0.288 + 1.134 + 0.176 = 1.598 m³.
Per pit: 8.64 - 1.598 = 7.042 m³. Twelve pits: 7.042 × 12 = 84.504 m³ compacted.
Datum to finished floor = 0.45 + 0.15 = 0.60 m. Fill thickness = 0.60 - 0.10 - 0.05 = 0.45 m. Volume = 240 × 0.45 = 108.0 m³ compacted. The pedestals occupy 12 × 0.16 = 1.92 m², under one percent of 240 m², and are ignored. Fill beyond the building line is excluded because it depends on the plinth or edge-batter detail.
84.504 + 108.0 = 192.504 m³ compacted.
Assuming the geotechnical engineer accepts the silty clay for reuse: 103.68 × 0.90 = 93.312 m³ compacted. Deficit = 192.504 - 93.312 = 99.192 m³ compacted, to be imported.
Loose volume = 99.192 × 1.30 = 128.95 m³. Truckloads = 128.95 ÷ 12 = 10.75, rounded up to 11 loads, delivering 11 × 12 = 132 m³ loose. The rounding leaves 132 - 128.95 = 3.05 m³ loose of contingency, which rutting and over-excavation will absorb.
Loose volume = 46.2 × 1.25 = 57.75 m³. Loads = 57.75 ÷ 12 = 4.81, rounded up to 5 loads.
Spoil in loose measure = 103.68 × 1.25 = 129.6 m³. At 1.5 m average height the footprint is 129.6 ÷ 1.5 = 86.4 m², roughly 9 m × 10 m, kept off the pad. For the site soil, 1 m³ compacted needs 1.25 ÷ 0.90 = 1.39 m³ loose, which is why a stockpile that looks enormous makes so little pad.
All rates are illustrative placeholders for the arithmetic, not quotations.
The naive take-off says: total fill 192.504 m³, excavation 103.68 m³, so import 192.504 - 103.68 = 88.824 m³. Divided by 12 that is 7.40 loads, rounded up to 8 loads at ₱7,200 = ₱57,600.
The factored take-off says 11 loads at ₱7,200 = ₱79,200. The loose-volume ratio is 128.95 ÷ 88.824 = 1.45, so the one-column estimate orders about 31 percent too little loose fill, because the quantity actually needed is 45 percent higher than the naive figure: a ₱21,600 hole on a single line item, plus the extra spreading and compaction the naive bill never priced, plus the Saturday phone call.
Whatever tool you use, keep bank, loose and compacted quantities as separate line items with their own unit rates, and write the swell and shrink factors into the notes where the next person can see them. That is how we structure earthwork lines when building a BOQ in RHCES Estimator: excavation in bank measure, hauling by the load, placement in compacted measure, each with a rate that belongs only to that line.
Ask for a field density test on the bank material and a compaction curve on the material to be reused; the ratio of bank dry density to loose and compacted dry densities gives the factors directly. Failing that, use handbook ranges for the soil class, label them as assumptions, and correct the estimate as soon as the first density results come in.
Bank measure, because that is what can be checked against survey and drawings and what most contracts pay against. Quote hauling in loose loads and placement in compacted measure. If the contract pays excavation to neat lines only, keep the working-space and batter volumes in your internal cost build-up even though they never appear in the pay quantity.
Yes, on both sides. Wet clay spoil swells more, takes longer to dry back to a workable moisture content, and may need spreading and aeration before compaction, which changes cycle time and sometimes the reuse decision entirely. Groundwater also forces dewatering, wider working space for sumps and pumps, and often a thicker lean concrete or gravel blanket, all of which change the pit volume. Re-run the take-off with wet-season factors rather than padding a single contingency percentage.