Who owns the float? Programme float, plainly — and why it decides delay claims

Every programme has slack in it somewhere. An activity finishes and the next one does not need to start for a week; the contractor plans to finish a month before the contractual completion date. That slack — float — feels like nobody's property until a delay lands on it. Then it becomes one of the most contested assets on the project: if the employer's variation eats the float, does the contractor still get an extension of time? If the contractor planned to finish early, can the employer's delay quietly consume that buffer without consequence? The answer decides real money, and most project teams cannot say with confidence what their own contract does with it.

What float actually is — three kinds worth separating

The word covers three different things, and the ownership argument goes wrong when they are mixed up. Total float is the time an activity can slip without delaying the contractual completion date — slack against the critical path. Free float is the time an activity can slip without disturbing the activity immediately after it. Terminal float is the gap between the contractor's planned completion and the contractual completion date — the buffer a contractor builds by programming to finish early. Herbert Smith Freehills Kramer's construction team sets out exactly this three-way split in its Legal Terms Explained series, and notes there is "no universal approach" to who owns any of them: it depends on the contract.

Who can claim each kind is the whole dispute. Total float sits inside the programme and either party's delay can consume it. Terminal float is different in kind — it only exists because the contractor chose to plan an early finish — which is why the standard forms treat it differently.

What the courts say: the float belongs to the project

The leading English authority is Ascon Contracting Ltd v Alfred McAlpine Construction Isle of Man Ltd (1999) 66 Con LR 119. McAlpine's programme carried five weeks of float. Its sub-contractors, Ascon among them, used it up. When McAlpine tried to pass its delay losses to Ascon, the court refused to treat the float as McAlpine's private reserve: where the contract is silent, float is a project resource, available to whoever needs it first. Judge Hicks put the logic simply — if six sub-contractors each cause a week of delay against six weeks of float, none has caused loss; they share the benefit, so they share it when it runs out.

That "first come, first served" rule is the default position in English law where the contract says nothing. It cuts both ways. An employer delay that lands while float remains does not, by itself, move the completion date — so no extension of time follows. A contractor delay that lands first consumes the same float. Neither side owns the buffer; timing does.

What the SCL Protocol says

The Society of Construction Law's Delay and Disruption Protocol (2nd edition, 2017) adopts the same default and states it precisely. Core Principle 8: unless there is express provision to the contrary, where total float remains in the programme, an employer delay entitles the contractor to an extension of time only to the extent that the delay is predicted to reduce that total float below zero on the affected critical path. In plain terms: the employer's event must actually push completion past the contractual date, not merely eat slack. The Protocol is guidance, not law — but it is the reference tribunals and delay experts reach for, and it is the position parties should expect to be held to when their contract is silent.

Commentators are not unanimous that this is fair — HKA's paper Float as a Project Resource walks through the competing "contractor owns it," "employer owns it," and "project owns it" schools and proposes allocation by agreement instead of by silence. Which is the practical point: the Protocol's rule is a default, and defaults can be drafted around.

Where the standard forms land

JCT and FIDIC: silent — so the float must burn first. Neither family of contracts defines float or allocates it. Under both, the contractor's entitlement to an extension of time depends on the contractual completion date being delayed. The consequence follows automatically: available float is exhausted before any entitlement arises, whoever's delay consumes it. On a FIDIC job, the 1999 Red Book's Clause 8.4 (Clause 8.5 in the 2017 edition) grants an extension only where completion "is or will be delayed" — an employer event absorbed by float does not qualify. The default in Ascon and the Protocol fills the gap.

NEC4: the contractor keeps its terminal float. NEC4 works differently, and the difference is deliberate. Under ECC clause 63.5, a delay to the Completion Date caused by a compensation event is assessed as the length of time that planned Completion moves against the Accepted Programme current at the dividing date — not as the effect on the contractual Completion Date. If the Accepted Programme shows planned Completion three weeks before the Completion Date, and a compensation event pushes planned Completion out by two weeks, the Completion Date moves by two weeks too. The three-week terminal float travels with the contractor, intact. As Brodies' analysis of the point puts it, the employer cannot use the contractor's planned early finish to absorb delays that are the employer's risk. Time risk allowances the contractor shows in its programme under clause 31.2 are treated the same way: they are the contractor's own contingency, not a shared pot. Total float within the body of the programme, by contrast, remains available to the assessment in the ordinary way.

So the practical rule of thumb across forms: total float is the project's; terminal float is the contractor's under NEC4 and unprotected under JCT and FIDIC unless the parties draft for it. Heavily amended contracts scramble this constantly — a bespoke amendment defining float ownership, or striking the NEC4 assessment mechanics, changes the answer entirely. The only reliable move is to read what the contract actually says.

Why this decides claims, not just arguments

Float ownership looks academic until a delay event needs notifying. It is not. Three failure patterns recur. First, a contractor assumes float protects it, absorbs an employer delay without notice, and discovers months later that the time-bar has closed on an event that — once later delays consumed the remaining float — turned out to be the one that mattered. Under the Protocol's rule, whether an employer event grants an extension depends on the float position at the time of the event; if nobody was tracking that position, nobody can prove it. Second, the parties fight retrospectively about how much float existed and when it was consumed — a dispute that is only winnable with contemporaneous programme updates showing the critical path as it actually moved. Third, on NEC4 jobs, teams fail to keep the Accepted Programme current, and clause 63.5's protection of terminal float becomes impossible to operate because there is no reliable programme at the dividing date to assess against.

All three come down to the same discipline: know your float position continuously, and treat any event that erodes it as a potential notice trigger — before the deadline, not after the analysis.

Where Aven-AI fits

Aven-AI does not decide who owns your float — that is a question of contract wording and, sometimes, law. What it does is make the float position impossible to lose track of. It reads the actual contract — FIDIC, NEC4, or a heavily amended hybrid — including any bespoke clause that allocates float, so the rule that applies to your project is the one it watches. It ingests the programme from Primavera P6 or MS Project and tracks how total and terminal float move with every update. When an event starts consuming float that bears on entitlement, it warns the right person while the notice window is still open, and drafts the notice with every statement cited back to the clause and the programme. It is advisory only: it flags and drafts, a human decides, and nothing is ever sent automatically. The float argument is won or lost on what was known and notified at the time — that record is exactly what it keeps.

This article is general information, not legal advice. It cites its sources so you can verify each point; confirm the current contract wording and the applicable law for your project before relying on any of it.

Related reading

• Extension of time under FIDIC, plainly: how to protect your delay entitlement
• The NEC4 eight-week time bar: how compensation events quietly expire
• Which delay analysis method should you use? The six SCL options, plainly

Sources & further reading

• SCL Delay and Disruption Protocol, 2nd edition (2017) — Society of Construction Law
• Legal Terms Explained: Float in Construction Contracts — Herbert Smith Freehills Kramer (February 2026)
• Ascon Contracting Ltd v Alfred McAlpine Construction Isle of Man Ltd (1999) 66 Con LR 119 — isurv case summary
• Who 'owns' terminal float under an NEC contract? — Brodies LLP
• Float as a Project Resource — HKA