How Bilby works
How Bilby recommends a foundation
Compare four foundation systems
Bilby compares slab on ground, crawlspace, piers and basement using the current ground and build-area evidence. The visible table presents categories: Recommended, Suitable, Potentially suitable, Less favourable, Not favourable or Specialist assessment. The internal numerical rank supports those labels; it is not a public bearing-capacity certificate or the foundation price.
Site evidence and construction burden
The assessment uses soil family, weak-soil depth, selected-area slope/relief, expected rock, groundwater and construction complexity. It also compares the cut/fill, rock, retaining and material transport associated with each foundation option.
| Factor | Weight | Typical evidence |
|---|---|---|
| Soil family | 30% | Mapped engineering family |
| Weak-soil depth | 20% | Depth band and settlement/soft-soil flags |
| Terrain fit | 20% | Pad slope, relief, cut/fill and retaining |
| Rock burden | 10% | Expected rock volume, or depth fallback |
| Groundwater | 10% | Risk evidence and system sensitivity |
| Construction complexity | 10% | Earthwork, rock, retaining and transport per m² |
Site score, residential prior and meaningful savings
The weighted site score contributes 65%; a residential prior contributes 35%. Priors are slab 85, crawlspace 70, piers 50 and basement 45. A conventional competent site can add 7 to slab: average slope ≤8%, relief ≤1 m, rock ≤20 m³, retaining ≤8 m² and cut/fill ≤50 m³, without weak-soil issues.
Crawlspace and piers can gain for avoided rock (meaningful/major at 25/50 m³), cut/fill (30/60 m³), retaining (8/15 m²) and material transport (30/60 m³). Full advantage needs slope ≥8% and at least two meaningful savings; weaker cases get smaller bonuses. Rock interaction subtracts 0 through 10 m³, then 2/6/12/20 as volume crosses 10/20/50/80 m³. Flat conventional contexts cap piers at 69 and crawlspace at 79 when alternative benefits are small.
How the categories are chosen
Internal scores ≥85 and ≥70 both display Suitable, 55–69 Potentially suitable, 40–54 Less favourable, below 40 Not favourable. The selected top eligible option displays Recommended instead. The internal top score must reach 55, and core soil-family and terrain-position evidence must exist.
Specialist gates override ordinary categories: designated specialist soils, clay deeper than 3 m, organic soil deeper than 1.5 m, significant unknown/environmental fill, instability/very soft soil, average slope ≥25% or maximum ≥35%, or average slope ≥20% with relief >3 m. In these cases the normal recommendation is withheld and specialist assessment shown. Basement remains particularly sensitive to excavation and groundwater.
Fictional ranking with missing terrain
Consider standard moraine at 2 m depth, no listed soil risks and low groundwater evidence, but no usable selected-area terrain result or candidate burden. For slab the factors are soil 95, weak-soil depth 92, terrain 50, bedrock fallback 65, groundwater 90.25 and complexity 50. Weighted site score = 28.5 + 18.4 + 10 + 6.5 + 9.025 + 5 = 77.425. Applying the prior gives 85 × 0.35 + 77.425 × 0.65 = 80.07625, rounded to 80, ordinarily Suitable.
Despite that internal rank, the overall recommendation is uncertain and no foundation is recommended: core terrain evidence is missing. This is why the category table must be read with the evidence warnings rather than treating the largest number as approval.
Confidence and the selected estimate
Low ground confidence, missing terrain or combined missing depth/groundwater gives low recommendation confidence. High needs high ground and terrain confidence, known depth/groundwater and at least a seven-point winner margin. Soil uncertainty can reduce it; a rock-avoidance winner with a depth range at least 1 m wide is also downgraded.
The foundation used for the cost estimate is a separate selected option. Use the foundation control and Recalculate to compare its scope; a recommendation does not silently change your house contract. Geotechnical and structural design still establish the final solution. Missing depth is not zero rock, and a cheaper option is not automatically the recommended one.
Method and review
Method FOUNDATION_SUITABILITY_V2 and the public foundation-category table reviewed 18 October 2026. Priors, weights, advantages and gates are Bilby screening assumptions. SGU supplies geological evidence; it does not certify this ranking or a foundation design. SGU
The review date covers this explanation and its implemented method. Price and data dates are stated separately in the article.
