Balancing Cut and Fill on a Wet Site

Shrinkage is a density ratio, not a property of the soil

The shrinkage percentage in an earthworks estimate is doing one job: converting an in-situ dry density into a specified compacted dry density. One cubic metre in the ground becomes the ratio of those two densities once it is placed and rolled. Which means the factor is not a fixed characteristic of the clay in front of you — it moves the moment the specification moves. A fill written to ninety-five per cent of a maximum obtained by modified effort consumes noticeably more bank metres per compacted metre than one written to ninety per cent of a standard-effort maximum, on the identical material.

The other half of the ratio is what the ground is doing before you dig it. Loose alluvium, soft made ground and anything recently placed and never compacted shrinks heavily, because there is a great deal of air to take out. A dense glacial till that has been consolidated under ice for ten thousand years is already close to the density your roller is trying to achieve, and shrinks very little. Push that far enough and the sign flips: a heavily over-consolidated clay, dug and remoulded, can occupy more space compacted than it did undisturbed, which turns an assumed shortfall into a surplus and a haulage bill nobody bid. The ten to fifteen per cent that gets quoted for ordinary cohesive fill is a planning assumption to open with, not a number to close a tender on.

A better figure costs one lab schedule. Take in-situ dry densities from the trial pits by sand replacement to BS 1377 Part 9 or ASTM D1556 Standard Test Method for Density and Unit Weight of Soil in Place by the Sand-Cone Method, and take the laboratory maximum from a compaction test at the effort the specification names — ASTM D698 for standard effort, ASTM D1557 for modified, or their AASHTO equivalents T 99 and T 180. The ratio between what the ground has and what the specification demands is your shrinkage factor, derived rather than borrowed, and it belongs in the file with the price.

Running the balance itself is then arithmetic with three careful inputs. Cut goes in as bank measure after the strip has been deducted and after any share of the cut you do not believe will make fill has been taken out. Fill goes in as the compacted design volume, which is what the drawing gives you directly. Shrinkage goes in as a negative percentage, because the convention here treats compaction as a reduction from bank. The sign is a convention and not a limit, though: the field runs from minus twenty-five through to plus thirty-five, so the over-consolidated clay that bulks on remoulding goes in directly as a positive figure and the cut volume stays the honest bank measure its label asks for. Both volumes are plain cubic metres in this one, so convert before typing if your takeoff is in yards.

Wet of optimum, and the roller that only polishes

For a given compactive effort, a soil reaches its highest dry density at one moisture content and falls away either side of it. That is the whole of the compaction curve and the reason a wet formation cannot be beaten into specification. Dry of optimum, the particles will not slide past one another; wet of optimum, water occupies the space the roller is trying to close, and no number of passes will remove it because the load path goes into pore pressure rather than into the soil skeleton. The reference tests are ASTM D698 Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort and ASTM D1557 for modified effort. BS 1377 covers the same ground in UK practice, though check which edition your specification names: the compaction tests moved into BS 1377-2:2022 when the long-standing Part 4 was withdrawn, and specifications written before that still cite the old part.

On site the failure announces itself before any gauge does. A formation that visibly deflects under a loaded lorry and springs back behind the wheels is pumping, and a proof roll that produces that movement is telling you the answer for the whole lift. Get a moisture number rather than a hand assessment: the oven method of ASTM D2216 is the reference but wants overnight, while ASTM D4944 using the calcium carbide gas pressure tester or the microwave method of ASTM D4643 will give you something to act on inside the shift. Where a UK earthworks specification is in force, acceptability may be set as a moisture content limit expressed against the plastic limit from BS 1377 Part 2, or as a band of moisture condition value, and the earthworks specification rather than the estimator decides which.

Field density then confirms what the proof roll suggested, by nuclear gauge to ASTM D6938 or by sand cone. Record moisture alongside every density result without exception. A density failure with no moisture reading beside it is an unexplained failure, and unexplained failures are the ones that become a conversation about competence instead of a conversation about weather.

Thin lifts as a drying strategy

Lift thickness limits are normally read as a compaction constraint — how deep the roller can reach before the bottom of the layer stops densifying. On a wet cohesive site they are also the only free moisture reduction available. A thin lift exposes its whole depth to wind and sun for the hour between spreading and rolling; a thick one buries its own wet half where nothing can get at it. Halving the lift roughly doubles the surface you are drying per cubic metre placed, and the plant is on site regardless.

Turning the material multiplies that effect. A harrow or an agricultural rotavator run through a spread lift breaks the clods that hold water in their centres and exposes new faces, and two passes of a harrow in a drying wind will do more than a day of rolling. Evaporation, not the plant list, sets the rate — which is why cohesive earthworks in a British winter effectively stop, and why the programme has to respect a season even though the specification only ever tests acceptability. Seal the formation before every shift end and ahead of any forecast rain: a pass with a smooth drum and a fall left across the surface sheds a night of rain that would otherwise undo three days of drying.

Match the roller to the material while placing. A padfoot or sheepsfoot kneads cohesive fill from the bottom of the lift upward and walks out as the layer stiffens, which is a visible progress indicator as well as a compaction method. A smooth drum on the same clay during placement seals the top and traps water underneath it, which is exactly what you want at the end of the day and exactly what you do not want at the start of one.

1. Scarify the surface of the lift below so the new layer keys into it instead of sitting on a rolled plane.
2. Spread to the loose thickness the specification allows for the plant on site, not the thickness the dozer finds convenient.
3. Harrow or rotavate the spread lift and give the wind an hour on it before anything rolls.
4. Take a moisture reading with a field method rather than judging by hand, and stop if it is above the acceptability limit.
5. Compact with the roller the material wants, padfoot for cohesive fill and smooth drum for granular.
6. Take density and moisture together at the specified frequency, recorded against lift and chainage.
7. Seal the surface and leave a fall on it before the shift ends.

Treating it rather than carting it

Every cubic metre of site-won material made acceptable removes two movements from the job, not one: a load out to a tip and a load in from a quarry, each with its own gate charge, its own haul cycle and its own lorry on the road. That arithmetic is why lime turns up on wet cohesive sites at all. It is not being used to build a pavement layer; it is being used to keep the balance the drawing already claimed, and the saving is measured in movements rather than in strength.

The chemistry is worth understanding because it decides what you can promise. Quicklime hydrates on contact with the water in the soil, consuming free moisture and releasing heat as it does, which drops the moisture content immediately. The calcium then exchanges with the cations held on the clay particles, flocculating a dispersed plate structure into crumbs and pulling the plasticity index down — measurable the same day by ASTM D4318 Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity Index of Soils, or by BS 1377 Part 2. That immediate change is modification, and it is what turns an unworkable formation into something a roller can act on. Given enough lime to hold a high pH, plus time and temperature, a slower pozzolanic reaction with the clay's own silica and alumina builds cementitious strength: that is stabilisation, a different job with a different dosage and its own strength testing, commonly a soaked CBR to ASTM D1883.

The dosage is a laboratory output, never a site guess. ASTM D6276 Standard Test Method for Using pH to Estimate the Soil-Lime Proportion Requirement for Soil Stabilization establishes the lime demand for the specific soil, and a full mix design confirms it against whatever performance the specification asks for. Mix designs commonly land between about two and eight per cent by dry weight, which is the band the calculator accepts, but the number for your clay comes from the study. Two things trip people up when converting that percentage into a delivery: the volume to enter is the treated layer only, area multiplied by treatment depth, and not the whole cut; and the dry density is your soil's measured value from the geotechnical report, because it scales the answer directly. Once you have a mass, divide it by the treated area to get the spread rate in kilograms per square metre that the operator actually sets on the spreader.

Sulfate is where lime treatment goes wrong, and it goes wrong late. In sulfate-bearing or sulfide-bearing clays, lime, sulfate and water together form expansive minerals, and the heave shows up weeks or months afterwards under a slab or a road that was signed off as passing. Total sulfate, water-soluble sulfate and oxidisable sulfide all get tested before any decision to treat, and the earthworks specification carries the limits and the mitigation — extended mellowing, a different binder, or a decision not to treat that material at all. BS EN 16907-4 Earthworks: Soil treatment with lime and/or hydraulic binders covers the process; on UK highway work the acceptability limits sit in Series 600 of the Specification for Highway Works.

Handling comes with its own constraints. Quicklime is caustic and reacts exothermically with moisture, including moisture on skin and in eyes, so it is handled to the supplier's safety data sheet and, in the UK, assessed under the Control of Substances Hazardous to Health Regulations 2002 with the PPE that assessment names. Spreading fine quicklime in wind puts dust on the operator and on the neighbours; shrouded spreaders and granulated products exist for that reason. Mellowing between mixing and final compaction is part of the design rather than slack in the programme — the material is re-mixed and compacted inside the window the specification gives, and a lift left mellowing over a wet weekend is a lift you are treating twice. The lime itself is specified: ASTM C977 Standard Specification for Quicklime and Hydrated Lime for Soil Stabilization in the US, BS EN 459-1 Building lime: Definitions, specifications and conformity criteria in Europe, and the class delivered has to match the class designed.

What still leaves, leaves loose and heavy

A surplus is a bank figure and lorries do not carry bank measure. Excavated soil bulks the moment the bucket lifts it, so the loose volume going up the ramp is meaningfully larger than the hole it came out of, and the swell factor between them varies with soil type, moisture and how the machine is loading. Wet cohesive spoil then fails the other way as well: it reaches the vehicle's weight limit while the body still looks capable of taking more, because gross vehicle and axle limits under road traffic law bind before the cube does. Every percentage point of moisture you did not drive off is water you are paying to lift, cart and tip.

Disposal is usually priced by the tonne while your takeoff is in cubic metres, and the bridge between them is a bulk density that moves with moisture. Convert the weighbridge dockets back through a density for that stream and compare them against the volume you predicted; a gap that runs consistently one way is your factor needing correction, not the haulier needing an argument. Do that reconciliation while the job is live, because the same clay is likely to be in the next tender.

The regulatory position is separate from the arithmetic and settles first. Material leaving the site is waste unless it moves under a recovery framework — in England, the CL:AIRE Definition of Waste: Development Industry Code of Practice provides one route, with classification and consignment otherwise following the regime in force. Anything a contamination assessment has flagged goes to a different gate at a different rate, and the classification has to be established before the disposal line is priced rather than discovered when the first load is turned away.