Guide · CPM fundamentals

Total float, free float and negative float explained

Float is the most quoted and least understood number in a CPM schedule. It is not spare time that somebody owns; it is an output of the network calculation that changes every time logic, durations or progress change. This guide sets out what each type of float measures and how to read it in a live programme.

Total float

How long an activity can slip before it delays project completion or a constrained date.

Free float

How long it can slip before it delays the earliest start of any of its successors.

Negative float

The network cannot meet a date imposed on it — the amount of catch-up required.

Critical path

The chain of activities with the lowest total float that drives the completion date.

How float is calculated

The forward pass through the network gives each activity its earliest possible start and finish. The backward pass, working from the project completion or an imposed date, gives the latest start and finish that still permit that date to be met.

Total float is the difference: late start minus early start. Free float is narrower — it is the gap before the earliest of the activity's successors would be pushed, so it belongs to that activity alone and cannot be consumed by anything upstream.

This is why float is never a fixed property of an activity. Add a successor, change a calendar, apply a constraint, or record progress, and the float on that path changes with it.

Reading negative float

Negative float almost always means a constraint or an imposed completion date is earlier than the logic can deliver. Minus fifteen days on a chain means the network says you are fifteen working days short of the date you told it to hit.

Before treating it as delay, check the cause. A 'must finish on' constraint left over from an earlier revision produces negative float that has nothing to do with performance. A genuine slip on the driving chain produces the same number for a completely different reason. The DCMA check for hard constraints exists mainly to make this distinction possible.

Suspiciously high float

Total float above the profile limit — 44 working days under DCMA, 60 under AACE and USACE, or whatever you set for your project type — is usually missing successor logic rather than real slack. The activity is not tied into anything downstream, so the backward pass gives it an implausibly late date.

The practical consequence is that the critical path is understated: work that will in fact drive completion is sitting off the path because the network never connected it. Fixing open ends typically shortens float across the schedule and reveals a longer, more honest critical path.

Who owns the float?

Contractually this varies. Many standard forms and the SCL Protocol take the position that float is a project resource available to whichever party needs it first, rather than something owned by the contractor or the employer. Some contracts allocate it expressly, and some require float to be shown and maintained.

The analytical point is separate from the contractual one: track float trend across updates. A path whose total float erodes from thirty days to five over four months is the early warning that the eventual delay claim will be about.

Tracking float in TideLine

The Schedule module shows total float, negative float and the driving chain in the Gantt, with near-critical paths distinguished from the critical one so you can see which chains are one slip away from driving completion.

The Quality module tests float against the limits you set rather than a single fixed rule: high-float and negative-float thresholds come from the active profile — DCMA 14-Point, GAO, AACE RP-90R-19, NDIA PASEG or USACE/NAVFAC — and each metric's weight and warn/fail limits can be tuned to the project type, so a two-year fit-out and a ten-year infrastructure programme are not judged against the same numbers.

For float erosion over time, compare two schedules and open the result in the Delay module. Each window shows how dates and the critical path moved between the snapshots, and the root-cause view follows a late activity back up its dependency chain to where the float was actually consumed — early warning while there is still time to act, and evidence later if there is not.

Frequently asked questions

What is the difference between total float and free float?
Total float is the delay an activity can absorb before project completion or a constrained date is affected. Free float is the smaller amount it can absorb before the earliest start of any successor is affected.
Does zero total float mean an activity is critical?
In an unconstrained network, yes — the lowest-float chain drives completion. Where imposed dates create negative float, the critical path is the most negative chain rather than the zero-float one.
Is negative float always bad?
It always needs explanation. It can indicate genuine slippage against a contractual date, or simply a stale hard constraint left in the schedule. Check the constraint before treating it as delay.
Why do some activities show hundreds of days of float?
Almost always missing successor logic. Without a downstream link the backward pass assigns an unrealistically late date, which inflates float and hides the activity from the critical path.