Bulk excavation moves large volumes to rough level without distinction — the method for mass basement digs, platform cuts and site grading where the material is uniform and the final surface comes later. Production defines it: a 30-ton excavator with matched tippers moves 600–1,200 m³ per day in soil, at KES 400–700 per m³ all-in. Selective excavation removes specific material while protecting what must stay — foundation formations in layered ground, services corridors, material segregation for reuse, the last 300mm above a sensitive formation. Production drops to 100–300 m³ per day and rates rise to KES 700–1,500 per m³, because the machine works to a line, not a level, and the operator's attention is the specification. The mistake is treating them as competitors: every real site needs bulk first and selective last, and the craft is drawing the line between them correctly. Our bulk basement practice runs the hybrid on every deep dig.

The selection test is what lies between the surface and the formation. Uniform soil, no sensitive interfaces, formation tolerance achievable later — bulk all the way: mass production rates, lowest cost per m³. Layered or variable ground, services or structures to protect, a formation that the next trade builds directly on, or materials that must be segregated for reuse — selective from the start of the sensitive zone, usually the last 0.5–1.5 metres above formation. The decisive question: can a rough cut be cleaned up cheaply, or does an over-dig cost structurally? If over-dug ground needs engineered backfill, compaction testing and re-trim, the 'cheap' bulk pass through that zone is the expensive one. Site-specific triggers we apply: switch to selective at the top of rock, at the water table, at any service corridor, and at 300mm above design formation on specification surfaces.

Over-excavation is the silent line item that separates tidy budgets from post-mortems, and the arithmetic is unforgiving. Example: a 2,000 m² basement formation over-dug by an average 200mm is 400 m³ of unplanned removal — KES 200,000–400,000 at bulk rates, before the real cost: the over-dug zone must be backfilled with approved material (not the spoil that just came out), compacted in 150mm layers to specification, and re-trimmed — adding KES 600–1,200 per m³ of replacement, or KES 240,000–480,000, plus testing and a week of programme. Total for a careless 200mm: KES 0.5M–1M and a delayed slab. On black cotton sites the penalty compounds: over-excavation into expansive clay below formation can force full replacement depths the design never intended. The controls are procedural: level survey control visible to the operator, dig-to levels marked on the drawings, bulk stops 300mm above formation by rule, and GPS machine control where tolerances are tight.

2026 bands, all-in with carting within 10 km: bulk excavation in ordinary soil — KES 400–700 per m³, driven by machine class and haul; bulk in rock — KES 1,500–2,800 with breaking; selective excavation near formation or services — KES 700–1,500 per m³ at the production rates above; final trim and formation preparation — often priced per m² instead, at KES 150–350 per m², since the value is tolerance, not volume. The premium of selective over bulk — roughly 50–100% — buys the formation quality the structure inherits, and it is almost always cheaper than remediating an over-dug one. Our rate breakdown itemises every line from mobilisation to trim, and the method savings analysis shows where the hybrid beats both extremes.

The tolerances follow what the formation receives. General platforms and landscape areas: ±100mm is workable — drainage falls matter more than absolute level. Formation under blinding or mass concrete: ±50mm, achievable with grade control and a skilled operator. Formation under raft or pile caps: ±25mm, verified by level survey before the engineer's inspection — anything outside is trimmed or built up with approved material before sign-off. Slab-on-grade industrial floors: ±15mm over long bays, which is where laser/GPS machine control pays for itself by removing rework entirely. The cost logic: tolerance is purchased twice — once in the trim operation, and once in what the next trade does not have to correct. Cheap trim with loose tolerance is expensive if it pours extra concrete; a tight trim with tight tolerance is cheap insurance. State the tolerance in the excavation specification, and hold the engineer's inspection at formation before anything covers it.