Contingency percentages by project type
New-build extension on greenfield rear garden: 10–12%. Lowest risk — new construction, known ground (unless clay or trees), standard details. Whole-house renovation + extension: 15–20%. Higher complexity, more trades, services upgrades, unknowns in existing fabric. Period property restoration (Victorian, Edwardian terrace): 18–25%. Heritage cost uncertainty, lath-and-plaster discoveries, original joinery condition, structural surprises. Listed building: 20–30%. Specialist trades, consent risk, LBC-instructed changes. Basement conversion: 25–35%. Ground conditions, structural complexity, programme uncertainty, specialist trades. Outer-London 1930s semi or 1960s suburban: 10–15%. Apply higher contingency where: party wall complexity, listed, CA, basement, structural alterations, services upgrades, occupied during build.
Risk categories with separate releases
(1) Design risk contingency (3–5%) — errors/omissions in drawings; release when design fully developed at Stage 4. Reduce by completing design before tender; QS review for omissions. (2) Construction risk (5–10%) — unforeseen site conditions, weather delays, supplier failures; release as risks discharge (ground risk releases when foundations complete). (3) Client change (0–5%) — client-initiated changes during build (spec upgrades, additional scope). Many renovations don't budget for this; result is contingency burn on changes that crowds out real risk allowance. (4) Exceptional risk (0–5%) — low-probability high-impact events (party wall dispute, planning enforcement, supplier collapse). For listed/heritage/basement allocate; simpler projects often nil.
Managing contingency through project
Hold contingency separately — not as line item to be 'used up' by contractor. Track weekly: starting balance, draws, current balance, projected end-of-project balance. Each VO references risk category and contingency draw amount; QS reports monthly on contingency status. Avoid: (1) Contractor visibility — keep total client-only. (2) Premature release — releasing mid-project when risks remain leaves no cushion for late surprises. (3) No tracking — informal burns invisibly; surprise at final account. (4) Replenishment — when client adds scope (£15k island), do not draw from contingency; treat as new budget item with its own contingency. (5) Bank vs spend — unused contingency stays with client at project end; should not roll into 'optional extras'.
When contingency runs out
80%+ of contingency consumed before 50% project completion → structural problem. Causes: incomplete design at start, unforeseen ground conditions, late planning conditions, specification creep. Response — pause and re-plan: (1) Re-tender remaining scope at fixed price. (2) Value engineer remaining scope. (3) Top-up from client reserves only if remaining risks clear. (4) Reduce scope (omit landscaping, secondary bathroom). Continuing without addressing erosion typically leads to project halt mid-build, contractor abandonment, disputes. Better to pause at 50–60% than push through to crisis.
