Free Float

Free float is always less than or equal to total float. Where total float measures how much delay can be absorbed before the project end date is affected, free float measures how much delay an activity can absorb before it affects the next activity in the chain. An activity can have large total float but zero free float if it is directly followed by an activity on the critical path: the delay would not reach the project end but it would immediately affect what comes next. In practice, free float appears most visibly on activities that feed into a merge point — a milestone or summary activity where multiple paths converge.

Free float matters when you need to manage subcontractor interfaces. If you are handing off from one trade package to another, the free float on the predecessor activity is the amount of slippage the successor can tolerate without being late to start. This has direct commercial relevance on NEC contracts, where a subcontractor delayed by late access from a predecessor has a potential compensation event claim. Free float is also the right metric to use when looking at activities that are not on the critical path but where delay would cause downstream disruption even though the project end date is unaffected.

Many scheduling tools display total float prominently and free float only on request, which can give a misleading picture of how much slack actually exists at an interface. Planners reviewing a schedule for handover risk should always check free float at key predecessor-successor boundaries, not just total float on the overall path. A clean-looking Gantt with healthy total float can still have zero free float at critical interfaces, meaning that one slippage cascades immediately through the programme even though the headline float numbers look fine.