Excavation and Cut and Fill: Why Two Identical Houses Cost $80k Apart

Why two identical houses on the same street cost $80k apart

Same design. Same inclusions. Same builder. Two blocks four doors apart in the same suburb.

One comes in eighty grand higher, and the owner is convinced they've been had.

They usually haven't. The difference is almost always under the ground, and it's some combination of three things: how much dirt has to move, what that dirt turns out to be, and how far below the surface you have to go before you find something worth sitting a footing on.

This is the least glamorous part of a build and it is the single most common source of a blown budget. Nobody argues about the cost of a kitchen, because you can see a kitchen. Substructure is money you spend to end up at ground level, which feels like paying to get to the starting line.

So it gets underquoted, both by builders trying to win a job on headline price and by owners assuming a flat-looking block is a flat block.

Cut and fill, and why fill is the expensive half

If your block slopes and your house is flat — and houses are flat — the difference has to go somewhere. You cut into the high side, you fill the low side, and you build on the platform you've made.

Cutting is comparatively straightforward. It's machine time and truck movements. The variable that bites is what comes out: clean spoil goes one place at one price, and material that's contaminated or classified differently goes somewhere else at a very different price. On a knockdown rebuild, especially anywhere with a fibro history, that classification is not a formality.

Fill is where people get caught. You cannot simply push spoil into the low side and build on it. Fill under a slab has to be engineered — placed in layers, compacted, tested. Uncontrolled fill is exactly what an engineer will refuse to let you found on, and the fix is either removing it and doing it properly, or going through it with deeper footings to reach competent material below.

Which means the same cubic metre of dirt can cost you twice. Once to dig it out, and again to bring engineered material in because what you had wasn't suitable to build on.

And the deeper problem: fill depth drives footing depth. Rawlinsons puts additional excavation to twice the depth of a footing at $103.50 per cubic metre in Sydney, and to twice the depth of a foundation at $100.00 per cubic metre. Those are the add-on rates, on top of the footing itself. They look small per unit. Multiply them across a footprint where the engineer has just told you to go a metre deeper than the standard detail and the number stops looking small.

The real numbers, so you can sanity-check a quote

Here's where the substructure money actually goes, using the Rawlinsons Sydney rates as the reference point rather than a figure someone made up.

Strip footings — 20 MPa concrete, two layers of trench mesh, including excavation of equal depth and planking and strutting: $455.00 per cubic metre.

Column foundations — same concrete, 40 kg per cubic metre of reinforcement, formwork, excavation: $780.00 per cubic metre.

Foundation beams generally: $980.00 per cubic metre. Priced by the metre instead, a 450mm × 600mm beam runs $266.00 per metre, a 450mm × 750mm $331.00 per metre, and a 600mm × 750mm $394.00 per metre.

Notice what those last three tell you. Going from a 450×600 beam to a 600×750 beam is a fifty percent jump per metre, and that decision isn't yours — it's the engineer's, and the engineer is responding to soil classification and fill depth.

This is why I'm blunt with people about quotes that don't name a soil class. If a builder has priced your substructure without an AS 2870 classification in front of them, they've guessed. The guess is usually optimistic, because the optimistic guess wins the job. When the geotechnical report comes back Class H2 instead of the Class M they assumed, the beams get deeper and wider, and that arrives as a variation with your contract already signed and your leverage gone.

Run the numbers on your own site before you're committed — /tools/feasibility-check exists for exactly this.

Rock

Rock deserves its own heading because it's the one that turns a bad week into a bad month.

Parts of Sydney are sandstone close to the surface. The northern suburbs, the harbour side, plenty of the inner west. You can be digging comfortably at 600mm and hit material that a standard excavator simply will not move.

At that point you're into rock excavation, which means different plant, sometimes rock saws or hammers, sometimes engineering-controlled methods where neighbours are close. The cost per cubic metre is not slightly higher. It's a different order of magnitude, and the programme impact is often worse than the cost — rock breaking near neighbouring dwellings comes with hours restrictions and vibration limits.

Here's the thing that makes me genuinely irritated: rock is almost always excluded in building contracts. Read the exclusions schedule on any fixed-price contract in this city and rock excavation will be sitting in there, along with contaminated soil and unexpected underground obstructions.

That exclusion isn't dishonest, exactly. Nobody can price what they haven't seen. But it does mean the words 'fixed price' are doing less work than most owners think, and the honest response is not to be outraged at the exclusion — it's to reduce the unknown before you sign. A geotechnical investigation with boreholes at the right locations tells you whether rock is a risk on your specific site. It costs a fraction of one day of a rock hammer.

More on how those schedules work in /insights/fixed-price-contract-exclusions-schedule-2026.

What actually reduces this cost

Some of this is fixed by geology and you can't argue with it. Plenty of it isn't.

Design to the site instead of importing a design onto it. A house stepped to follow a slope moves dramatically less dirt than the same floor area forced onto one level. Split levels are not an architectural indulgence on a sloping block, they're a cost control. The flat-slab-everywhere approach is often the expensive one dressed up as the simple one.

Get the geotechnical work done before the design is locked. Not after. If the engineer knows what's down there while the layout is still fluid, footprint and orientation can respond to it. Once the plans are approved, you're solving the problem with concrete instead of with a pencil.

Balance cut and fill on site if the material is suitable. Every cubic metre trucked off site is paid for twice — once to load and cart, once to buy and bring back whatever replaces it. Where the spoil is clean and can be engineered into a platform, that's real money saved.

Be honest about the driveway. Steep driveways are excavation too, and a driveway grade that doesn't work is a problem you'll be reminded of daily for as long as you own the house.

Price the unknowns as named provisional sums, not exclusions. This is the one I'd push hardest. An exclusion means the risk is entirely yours and unquantified. A provisional sum means it's been thought about, given a number, and gets adjusted to actual with the margin applied only to the difference. You still carry the risk. But you carry it with a figure attached, which is the difference between a budget and a hope.

If you're at the stage of comparing quotes and one is meaningfully cheaper than the others, look at the substructure before you look at anything else. That's usually where the difference is hiding, and it's the part you can't change your mind about later.