Float, Criticality and Schedule Sensitivity

Introduction

A network schedule does more than identify the sequence of project activities. It also indicates how much movement is possible before a date, successor activity, or project objective is affected. This movement is described through several forms of float. Correct interpretation of float is essential for schedule control, resource decisions, change evaluation, and communication with sponsors.

Float is not simply “extra time.” It is a property of a particular schedule logic, calendar, constraint set, and calculation convention. An activity may have total float but still be unable to move freely because its movement could delay a successor, consume another activity’s flexibility, create resource conflicts, or alter the critical path. Consequently, float must be interpreted as conditional schedule flexibility rather than as an unrestricted resource.

This chapter develops the calculation and managerial interpretation of total, free, independent, and interfering float. It then examines multiple critical paths, path convergence, negative float, and schedule sensitivity.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Calculate and interpret total, free, independent, and interfering float.
2. Distinguish critical activities from activities that merely have low float.
3. Analyse multiple critical paths and converging network paths.
4. Explain the causes and implications of negative float.
5. Assess critical-path sensitivity under duration, logic, resource, and date changes.
6. Apply float information without treating it as an unrestricted resource.
7. Recommend appropriate schedule-control actions in professional settings.

1. Schedule Flexibility and Float

In an activity-on-node network, each activity has a duration and is connected to predecessor and successor activities. The forward pass calculates the earliest times at which activities can start and finish. The backward pass calculates the latest times at which they can start and finish without violating the required project completion date.

For activity i, the principal notation is:

For a standard finish-to-start relationship with zero lag:

The formulas assume a consistent calendar and that duration is measured in the same units as start and finish times. If calendars, time zones, shifts, or leads and lags are used, those features must be included in the scheduling calculation.

Float measures the difference between early and late timing. Different types of float answer different managerial questions:

• Total float: How long can an activity be delayed without delaying the project completion date or an applicable schedule constraint?
• Free float: How long can an activity be delayed without delaying the earliest start of any successor?
• Independent float: How long can an activity be delayed when predecessors occur as late as permitted and successors still begin as early as permitted?
• Interfering float: How much of total float would affect successor timing if consumed, even though it would not yet delay project completion?

These definitions depend on the network model. The calculations below use an activity-on-node network with finish-to-start relationships and explicit treatment of the project completion target.

2. Total Float

Total float is the maximum delay of an activity that does not delay the planned project completion date, assuming the rest of the schedule is managed according to the network logic.

The equality follows because both start and finish times are separated by the same activity duration. Total float is measured in time units.

An activity is commonly labelled critical when its total float is zero, subject to the scheduling tool’s tolerance and the project’s calculation rules. In a constrained or delayed schedule, an activity may have negative total float. A negative value indicates that the current logic and dates cannot meet the required completion target without corrective action.

Total float is a network-level measure, not a guarantee that the activity can be moved by that amount without any operational effect. For example, an activity with 5 days of total float may have only 1 day of free float. A 3-day delay could therefore affect a successor even though the project finish remains unchanged.

3. Free Float

Free float is the amount of time an activity can be delayed without delaying the earliest start of any immediately succeeding activity.

For an activity with one or more successors:

where j represents each immediate successor of activity i. If the result is negative because of unusual constraints or inconsistent data, free float is generally interpreted as zero or reported as a schedule exception, depending on the scheduling system’s convention. Under a conventional unconstrained network, free float is non-negative.

Free float is especially useful for short-term coordination. It indicates whether an activity can move without changing the early dates communicated to the next work package. Unlike total float, it focuses on successor protection rather than final project completion.

For a terminal activity, free float is calculated relative to the project completion milestone or contractual finish target. The project model must make this reference explicit; otherwise, terminal-activity float can be misinterpreted.

4. Independent Float

Independent float is a more conservative and less frequently used measure. It represents the time available to an activity when all predecessors are assumed to finish as late as allowed and all successors are assumed to start as early as allowed.

For an activity-on-node network:

Here, j represents immediate successors and k represents immediate predecessors. For a starting activity with no predecessor, the predecessor term is replaced by the project-start boundary, often zero. For a terminal activity, the successor term is replaced by the relevant project-finish boundary.

The formula assumes finish-to-start relationships, zero lag, a common time basis, and compatible early and late dates. It is not directly transferable to every relationship type or to schedules with complex calendars without adjustment.

Independent float is usually less than or equal to free float and total float. It identifies flexibility that does not depend on predecessors finishing earlier than their latest permissible dates. It is therefore useful when a project manager wants to know whether an activity has genuine local flexibility under adverse timing of surrounding activities.

5. Interfering Float

Interfering float is the portion of total float that, if consumed, delays the early start of one or more successors but does not delay the project completion date.

Under the standard activity-on-node definitions:

where IntFᵢ is interfering float. Since free float cannot normally exceed total float in a consistent unconstrained network, interfering float is normally non-negative.

Interfering float is not “wasted” time. It indicates that a delay can be absorbed at the project level but will interfere with the early timing of downstream work. Such interference may have consequences not visible in the final completion date, including resource disruption, changed handoffs, overtime, resequencing, or loss of managerial confidence.

The relationship among these measures is therefore:

This identity applies when the float definitions are calculated consistently for the same activity and schedule status.

6. Worked Example: Four Forms of Float

Consider the following illustrative network. Activity A precedes activity B, while activity C is an independent predecessor of D. Activities B and D both precede E.

For activity B:

Its successor is E, so:

Therefore:

For activity D:

Its free float is:

Thus its interfering float is also zero.

Now suppose D has an alternative late schedule in which its latest finish remains 9 but its successor E is controlled by B, which finishes at 7. The relevant early successor date is still 9 because E must wait for both predecessors. This illustrates an important point: the successor with the latest controlling early finish determines free float. One should not calculate free float from an arbitrary successor.

To demonstrate independent float for D, assume its predecessor C has LF = 6 and E has ES = 9:

Although D has 3 days of total and free float, it has no independent float. Its flexibility depends on C finishing earlier than its latest allowable finish. This distinction is important when predecessor commitments are uncertain.

7. Multiple Critical Paths

A critical path is a continuous path through the network whose duration determines the project completion date under the current schedule assumptions. In the simplest case there is one critical path. Real projects frequently contain several paths with equal or nearly equal durations.

Multiple critical paths arise when:

• two or more complete paths have the same longest duration;
• parallel work streams converge on a common milestone;
• constraints or imposed dates create zero float on several paths;
• resource levelling changes activity timing;
• different calendars or contractual milestones produce separate controlling chains.

The presence of multiple critical paths increases schedule risk. A delay on any critical path may delay the project, and corrective action on one path may leave another path controlling the completion date. A project manager who monitors only one path may therefore report a misleadingly stable schedule.

Near-critical paths are also important. A path with one or two days of float may become critical after a small duration increase, logic change, or resource delay. The practical risk of a path is not determined only by whether its current float equals zero; it also depends on the likelihood and impact of change.

Path convergence

Path convergence occurs when two or more predecessor chains join at a common successor or milestone. If a successor cannot start until all predecessor paths are complete, the latest finishing predecessor controls its early start. Convergence creates a “merge-point” risk: delays from several paths may independently affect the same downstream work.

For example, if paths P₁ and P₂ reach a milestone at 20 and 21 working days respectively, the milestone occurs at day 21. If P₁ is shortened by two days, P₂ still controls the milestone. If P₂ is shortened by two days while P₁ remains unchanged, P₁ controls it. The controlling path can therefore switch as work progresses.

Path convergence also creates probabilistic risk. Even if each path has a moderate chance of delay, the probability that at least one path delays a merge point can be substantial. This is one reason deterministic critical-path reporting should be supplemented by risk analysis or schedule simulation for complex projects.

8. Negative Float

Negative float indicates that the calculated schedule cannot meet a required date under the current logic, durations, calendars, and constraints.

Suppose an activity has ES = 18 and LS = 16. Then:

The negative value indicates a two-day schedule deficiency relative to the applicable completion requirement.

Negative float can be caused by:

• an imposed completion date earlier than the network’s calculated finish;
• a missed contractual milestone;
• a mandatory finish-on-or-before constraint;
• an approved change that increased duration without revising the deadline;
• an incorrect calendar or time-zone configuration;
• missing, excessive, or incorrectly directed logic;
• progress data that reveals actual delay.

Negative float is a diagnostic signal, not a schedule solution. It should not be eliminated by arbitrarily changing activity dates or deleting logical relationships. A sound response is to identify the cause, validate the baseline and constraints, and evaluate recovery options such as scope reduction, resequencing, additional capacity, overtime, procurement acceleration, or negotiated date revision.

The ethical issue is significant. Manipulating dates to remove negative float can create a misleading report and impair governance. Schedule updates should preserve an auditable record of actual progress, approved changes, assumptions, and corrective actions.

9. Critical-Path Sensitivity

Critical-path sensitivity describes how likely the controlling path is to change when project conditions vary. A schedule may have one apparent critical path but still be highly sensitive if another path is only slightly shorter.

Sensitivity is affected by:

1. Duration uncertainty. A small increase in a near-critical activity may make its path critical.
2. Logic uncertainty. A change in predecessor relationships can create or remove a controlling chain.
3. Resource availability. Resource limits may delay activities that initially have substantial float.
4. Calendar assumptions. Holidays, shifts, weather calendars, and working-time rules can change path lengths.
5. External constraints. Contractual dates, regulatory approvals, and supplier commitments may impose new controls.
6. Progress status. As actual work replaces estimates, the critical path may move.
7. Path convergence. Several paths may compete to control the same milestone.

A useful sensitivity review compares the current critical path with near-critical paths, examines float erosion, and tests plausible scenarios. Scenario analysis might ask: What if a key procurement activity slips by five days? What if an approval is returned for rework? What if a resource is unavailable for one week?

For complex schedules, Monte Carlo simulation can estimate the distribution of completion dates. It can also identify activities and paths that frequently contribute to schedule delay. However, simulation results are only as reliable as the duration distributions, correlations, logic, and calendars used. A technically sophisticated simulation does not correct poor schedule data.

10. Float Is Not an Unrestricted Resource

Treating float as an unrestricted resource is a common but serious scheduling error. Float belongs to the network relationship, not automatically to a particular team or manager. Several limitations explain why.

10.1 Float is conditional

Total float assumes that other activities follow the calculated schedule. If several activities consume their float simultaneously, the combined effect may exhaust the flexibility of a downstream milestone.

10.2 Float is shared at convergence points

Two activities may each appear to have float, but both may depend on the same successor. Using float on both paths can shift the successor or create a new critical path.

10.3 Float does not create capacity

A five-day float does not mean that five days of labour, equipment, or budget are available. Resource constraints may prevent an activity from moving even when its logical float is positive.

10.4 Float may be consumed by uncertainty

Weather, approvals, rework, supply disruption, and productivity variation can consume float without deliberate managerial action. Float should therefore be monitored as a risk buffer, not promised casually.

10.5 Float can be calendar-specific

A stated float of three working days is not necessarily three calendar days. Non-working periods and multiple calendars must be made explicit.

10.6 Float changes when the schedule changes

Adding a relationship, revising a duration, recording progress, or changing a constraint can recalculate float throughout the network. Float is a current schedule property, not a permanent entitlement.

A responsible project manager allocates time flexibility through explicit control decisions. If an activity is permitted to move, the decision should identify the affected successors, resources, milestones, and risk reserves.

11. Applied Case: Data-Centre Commissioning Programme

The following is an illustrative case based on common project-management conditions.

Background

A data-centre commissioning programme contains two major work streams. The electrical stream includes switchgear installation, testing, and energisation. The information-technology stream includes equipment delivery, rack installation, configuration, and integrated testing. Both streams converge at systems integration.

The contractual commissioning date is fixed. At the start of month four, the electrical path is critical with zero total float. The IT path has two days of total float and one day of free float.

Problem

The IT manager requests permission to delay rack installation by two days because a specialist technician is needed on another project. The manager argues that the IT path has two days of total float and therefore the delay is harmless.

Analysis

The schedule review identifies three issues:

1. Only one day is free float. A two-day delay would move the early start of integrated testing by one day.
2. The electrical and IT paths converge at integrated testing. The electrical path is already critical, so any delay to IT would remove the remaining flexibility at the merge point.
3. A supplier warning indicates that electrical testing may slip by one day. The IT path is therefore near-critical and highly sensitive.

The schedule team tests alternatives. Moving rack installation by two days delays integrated testing by one day and creates a second critical path. Moving the specialist assignment by one day, while resequencing configuration work, preserves the commissioning date. A further option is to split the installation into zones, allowing configuration to begin in completed zones.

Approach and solution

The project manager rejects the assumption that total float is freely available. The team chooses zone-based installation and assigns the specialist for one additional day. The schedule is updated with explicit logic and resource assumptions. The risk register records the convergence point as a high-priority schedule risk.

Outcome

The IT path retains one day of total float, integrated testing remains protected, and the project has a documented contingency if electrical testing slips. The decision does not merely preserve a date; it protects the integrity of the converging network and avoids an unplanned critical-path switch.

Learning points

• Total float and free float answer different questions.
• A near-critical path can be more important than a currently critical path when uncertainty is high.
• Resource decisions must be tested against network logic.
• Path convergence requires joint monitoring of parallel work streams.
• Float consumption should be approved as a controlled change, not treated as an automatic entitlement.

12. Tools, Governance and Contemporary Practice

Modern scheduling systems calculate float automatically, display critical and near-critical paths, and can compare current forecasts with baselines. Dashboards may show float erosion, milestone variance, and paths with high sensitivity. Artificial intelligence and machine-learning tools can assist by identifying unusual progress patterns, forecasting likely delays, or ranking activities for review.

These tools do not replace professional judgement. Automated outputs may be distorted by incomplete logic, excessive constraints, incorrect calendars, out-of-sequence progress, or unreliable duration estimates. AI-generated recommendations should be traceable to source data and reviewed by qualified schedulers. Governance should include version control, baseline approval, update cut-off dates, documented assumptions, and an audit trail for changes.

Best practice includes:

• maintaining a logically complete network;
• using consistent calendars and clearly defined data dates;
• reviewing both critical and near-critical paths;
• monitoring float trends rather than isolated float values;
• testing resource and constraint effects;
• separating approved changes from progress updates;
• validating negative float before reporting recovery actions;
• communicating that float is conditional and schedule-specific;
• using risk analysis when uncertainty or convergence is material.

Chapter Summary

Float expresses schedule flexibility, but different float measures serve different purposes. Total float measures allowable delay relative to project completion or an applicable constraint. Free float measures allowable delay without moving the earliest start of successors. Independent float measures flexibility when predecessors finish as late as allowed and successors start as early as allowed. Interfering float is the portion of total float that may affect successor timing without yet affecting project completion.

Multiple critical paths and path convergence increase schedule exposure because several chains may control a common milestone. Negative float signals that the current schedule cannot meet a required date and should prompt diagnosis rather than cosmetic date manipulation. Critical-path sensitivity requires attention to near-critical paths, duration uncertainty, logic changes, resource constraints, calendars, and external requirements.

Most importantly, float is conditional flexibility, not an unrestricted pool of time or capacity. Professional schedule control requires monitoring float erosion, testing changes across the network, protecting merge points, and maintaining transparent governance.

Key Terms

Critical path: The longest controlling path that determines the scheduled project completion under current assumptions.

Critical activity: An activity with zero or effectively zero total float under the applicable scheduling tolerance.

Criticality: The extent to which an activity or path controls a milestone or project completion.

Data date: The date separating actual progress from forecast schedule information.

Earliest finish (EF): The earliest time an activity can finish according to network logic.

Earliest start (ES): The earliest time an activity can start according to network logic.

Free float: Delay available without delaying the earliest start of an immediate successor.

Independent float: Delay available when predecessors finish as late as allowed and successors start as early as allowed.

Interfering float: Total float that can affect successor timing without delaying project completion.

Late finish (LF): The latest time an activity may finish without violating the schedule target.

Late start (LS): The latest time an activity may start without violating the schedule target.

Merge point: A network location where two or more paths converge.

Near-critical path: A path with small positive float and a realistic possibility of becoming critical.

Negative float: A negative schedule margin indicating that a required date cannot be met under current conditions.

Path convergence: The joining of multiple predecessor paths at a successor or milestone.

Schedule sensitivity: The likelihood that changes will alter the controlling path or completion date.

Total float: Maximum activity delay that does not delay the project completion date or applicable constraint.

Float erosion: Reduction in available float as progress, delays, or changes occur.

Review Questions

Short Answer Questions

1. Define total float and state its standard formula.
2. How does free float differ from total float?
3. What does negative float indicate?
4. Why are near-critical paths important?
5. State two reasons why float should not be treated as an unrestricted resource.

Long Answer Questions

1. Explain the calculation and managerial interpretation of total, free, independent, and interfering float.
2. Discuss how multiple critical paths and path convergence affect schedule risk.
3. Explain the principal causes of negative float and recommend an ethical response to it.
4. Analyse the relationship between resource constraints and critical-path sensitivity.
5. Evaluate the usefulness and limitations of automated scheduling and AI-based schedule analysis.

Analytical / Application Questions

1. An activity has ES = 12, EF = 17, LS = 15, and LF = 20. Its only successor has ES = 19. Calculate total float, free float, and interfering float. Interpret the result.
2. Two paths converge at a milestone. Path A has a current duration of 30 days and Path B has a current duration of 29 days. Explain how a two-day delay on each path would affect the controlling path and milestone.
3. A project has negative float of four days because its contractual completion date is earlier than the calculated finish. Propose a structured recovery analysis, including data validation, possible schedule changes, and governance controls.

Multiple Choice Questions with Answers

Question 1

An activity has ES = 8 days and LS = 11 days. What is its total float?

A. 3 days

B. 19 days

C. −3 days

D. 8 days

Correct Answer: A. 3 days

Question 2

Free float is best defined as the amount of delay that does not delay:

A. The activity’s predecessor

B. The earliest start of any immediate successor

C. The project budget

D. The latest finish of the activity

Correct Answer: B. The earliest start of any immediate successor

Question 3

An activity has total float of 6 days and free float of 2 days. Its interfering float is:

A. 2 days

B. 4 days

C. 6 days

D. 8 days

Correct Answer: B. 4 days

Question 4

Which condition most directly indicates negative float?

A. The activity duration is zero

B. The latest start is earlier than the earliest start

C. The activity has several successors

D. The activity has positive free float

Correct Answer: B. The latest start is earlier than the earliest start

Question 5

Independent float assumes that predecessors:

A. Finish at their earliest times

B. Finish at their latest allowable times

C. Are removed from the network

D. Have unlimited resources

Correct Answer: B. Finish at their latest allowable times

Question 6

What is a principal risk at a path-convergence point?

A. Only the shortest path can affect the milestone

B. Several predecessor paths may independently delay the same successor

C. All activities automatically become non-critical

D. Free float becomes equal to project duration

Correct Answer: B. Several predecessor paths may independently delay the same successor

Question 7

A path has two days of float while the current critical path has zero float. Why should the path still be monitored closely?

A. It must already be longer than the critical path

B. A small delay may make it critical

C. Float can never change

D. It has no relationship with project completion

Correct Answer: B. A small delay may make it critical

Question 8

Which statement about float is correct?

A. Total float always represents available labour capacity

B. Float remains fixed after the baseline is approved

C. Float is conditional on network logic, calendars, constraints, and other activity timing

D. Free float is always greater than total float

Correct Answer: C. Float is conditional on network logic, calendars, constraints, and other activity timing

Question 9

An activity has EF = 14 days and its successor has ES = 18 days. Assuming one immediate successor, its free float is:

A. 4 days

B. 32 days

C. −4 days

D. 14 days

Correct Answer: A. 4 days

Question 10

The most appropriate response to negative float is to:

A. Delete the controlling predecessor

B. Move dates arbitrarily until float becomes zero

C. Validate the cause and evaluate documented recovery or date-change options

D. Ignore it because negative float is always a software error

Correct Answer: C. Validate the cause and evaluate documented recovery or date-change options