Using Free Float to Optimize Resources

We’ve all been in this situation. You are driving down a highway at a fairly decent rate of speed when, ahead of you about a half-mile, you see a distant traffic light turn from green to amber to red. What do you do? Do you continue barreling down the road and then screech to a stop just in front of the light? Of course not! Why waste the fuel or the brake linings? No, instead, you simply take your foot off the accelerator pedal and you allow the vehicle to gradually slow down. This is known as avoiding a “hurry up and wait” situation.

Understanding the Dynamics of Parallel, Converging Path Segments

So now let’s talk schedule; CPM Scheduling to be specific. Let me establish the scenario: four parallel Path Segments that converge into Milestone Z.

  • Picture four Activities (Activity A1, Activity A2, Activity A3, Activity A4) connected in a linear series from left to right. Let’s call this Path Segment A. Activity A4 is succeeded by Milestone Z, which is a zero-duration milestone that the Project Team is interested in maintaining at all costs.
  • Another four Activities (Activity B1, Activity B2, Activity B3, Activity B4) are also connected in a linear series from left to right, and Activity B4 is also succeeded by Milestone Z. We will call this Activity Path Segment B.
  • Now imagine Activity Path Segments C and D similarly configured, and also feeding into Milestone Z.

Okay. Do you have the image clear in your mind? Each of the four Path Segments is a linear series of Activities that collectively precedes Milestone Z. This means that Milestone Z cannot be achieved until all four Path Segments have completed.

Most Temporally Significant Activity Path Segment

So, of the four Activity Path Segments, which is the most significant to Milestone Z’s timely completion? Well, the answer to that question depends on which one(s) of the four Path Segments will complete latest. Let us consider the question with a few different sets of assumptions.

Scenario 1: All Path Segments Have the Same Earliest Finish

In this first scenario, all four Path Segments reach completion at the same time. Maybe the four Path Segments all start at the same time, and are of the same collective length (in terms of aggregate Activity Durations). For example, suppose that all four Path Segments start on Day 1 and each of the sixteen Activities have the same two-day duration. Then each Path Segment will be eight days long, and each Path Segment will complete, at the earliest, by the end of Day 8. In this scenario, all four Path Segments are of equal significance.

Scenario 2: One Path Segment Has a Later Earliest Finish than the Other Three

In this second scenario, one of the four Path Segments reaches completion later than the other three. Suppose that the four Activities of Path Segment B are three-days, each. Then Path Segment B will take twelve days (three days per Activity times four Activities) to complete, whereas the other three Path Segments will take eight days each. Thus, Path Segment B is four days longer than the other three Path Segments. In this case, Path Segment B is the most significant.

Scenario 3: Multiple Path Segments Have a Later Earliest Finish than Others

This third scenario is just a more complicated version of the second scenario. An example would be where Path Segment C is also twelve days long, the same length as Path Segment B. Meanwhile, Path Segments A and D are, each, eight days long. In this case, both Path Segment B and Path Segment C are equally more significant to Milestone Z than Path Segments A and D.

Free Float Compares Path Segment Completion Times

Given the concept of a Path Segment we can now better understand what Free Float is all about. Free Float is a numeric value that expresses the difference in the amount of time that any parallel and converging Activity Path Segment is expected to “arrive” at the point of convergence compared to the arrival time of the most significant Activity Path Segment.

Let’s think this through with a different example of Scenario 2. Consider the following table:

Path Segment
Path Start
Activity Duration
Activity Count
Earliest Finish
Free Float
Path A Day 1 2 4 Day 8 +24
Path B Day 1 3 4 Day 12 +20
Path C Day 1 5 4 Day 20 +12
Path D Day 1 8 4 Day 32 +0

The above table should be fairly self-explanatory. Take a moment to study it. In this example we assume that all four Activity Path Segments start at the same time, Day 1.  But we could also have different start times for each Path Segment and the calculation process would be the same. Consider the following table:

Path Segment
Path Start
Activity Duration
Activity Count
Earliest Finish
Free Float
Path M Day 15 2 4 Day 22 +10
Path N Day 11 4 4 Day 26 +6
Path P Day 5 3 4 Day 16 +16
Path Q Day 1 8 4 Day 32 +0

I hope that these two tables help you better understand that Free Float tells us how much earlier any Path Segment is expected to achieve completion status compared with the most significant Activity Path Segment. (The Most Significant Activity Path Segment always has a Free Float value of zero, since it is being compared to itself).

How to Use Free Float When Managing Project Execution

Now we can return to our opening example. If you are the Project Manager in charge of the work being performed by Activity Path Segments M, N, P, and Q, how would the above table inform your decisions regarding deployment and utilization of critical resources required by the Activities of those four Path Segments?

  • Delay the Start: Well, one option available to the Project Team is to delay the start of an Activity Path Segment. For instance, Path Segment P enjoys 16 days of Free Float. This means that we could delay the start of the Path Segment until Day 21, and it would have no negative impact on the achievement of Milestone Z. Or, in the upper table, we could start Path Segment A on Day 25 and not delay the earliest possible achievement of Milestone Z.
  • Work Slower: But the better use of this information might be to simply reduce the rate at which the work is being performed. This is akin to taking our foot off of the accelerator pedal. To illustrate this, look at Path Segment M. What if we reduced the labor force by 50% for the four Activities? That would result in each Activity taking twice as long — or, four days each. That means that the Path Segment would take sixteen days, not eight days. If the Path Segment still commences on Day 15, it would complete on Day 30 (15 + 16 – 1), still two days earlier than the most significant Path Segment, Path Segment Q, on Day 32. And now we would have those labor resources to deploy elsewhere!

Technical Criticism: Free Float Belongs to the Path Segment, not to the Activity

I will close with a bold criticism, admittedly technical, of how Dominant Project Management literature explains and defines Free Float. Without virtually any exceptions, prevailing definitions for the term Free Float consistently relate the numeric value to an Activity.” Here are a few definitions I found when I Googled on Free Float; I added the underlining.

  • The amount of time an activity can be delayed beyond its early start/finish dates without delaying the early start or early finish of any activity.
  • Free float is used to describe amount of time that spans from the completion of one previously scheduled activity and extends to the point at which the next scheduled activity is set to begin. Free float can be calculated by determining the amount of the time between the earliest start date of the initial activity and the earliest start date of the succeeding activity, and then subtracting from that total the amount of time that it is expected the first activity will take to complete.
  • Free Float can only occur when two or more activities share a common successor.
  • Free float is the number of days an activity can be delayed without taking float away from the next activity. Another way of saying the same thing is that free float is the number of days an activity can be delayed without delaying the early start date of the next activity.
  • The length of time, expressed as work units, that a specific activity may be delayed without delaying the start of another activity scheduled to follow immediately after.

While the differences in wording between them ranges from subtle to extreme, what all of these definitions have in common is a fixation on what is essentially a secondary, albeit technical, truism about Free Float — rather than on what Free Float means in the real world of practical application.

It’s like ten hotel guests trying to board an elevator, each one toting several pieces of luggage. Nine of them squeeze into the cab, while the tenth person remains in the lobby, unable to join the others. Is it fair blame the ninth person to board the elevator, since he is the one nearest the door and it was his bags that were preventing the tenth passenger from boarding? If only the first (or fifth, or eighth) traveler had had fewer bags, couldn’t that have spelled the difference for Guest Ten?

Do you get the point? Any of the Activities within a Path Segment have the potential to elongate (or shorten) it. And if you agree with this point, then you must also agree that Path Segment’s Free Float belongs to all of them collectively, not simply to the last Activity in the series of Activities. Yet, that is how the preponderance of Free Float definitions explain Free Float (see above).

Anyway, had there not been a negative practical consequence for such misguided wording I would not have written this blog about it. But there is a negative practical consequence; few Project Managers understand how Free Float can be used to maximize the use of critical resources.

I hope this stimulates some discussion where you work. If you would like to learn more about Free Float, read ICS-White Paper WPA-MF-28, entitled, “Schedule Acceleration Using Free Float and Multi-Path Analysis.”

1 Comment to Using Free Float to Optimize Resources

  1. daveddbDave says:

    The blog explains an interesting scheduling concept that varies in its definition depending on which expert you listen to or read. The author’s treatment differs from many and yet by using different scenarios illustrates and explains this relevant if not often used scheduling topic.

    The different scenarios demonstrate how free float actually works, by using four parallel activity path segments, which all precede yet converge on the same milestone, which we’ll call X. Therefore, milestone X cannot be achieved until the entire four path segments are completed. The objective of the exercise is to determine which path segment is the most significant to the timely completion of the milestone.

    Ultimately, the illustration tries to provide the information necessary regarding the deployment and utilization of critical resources necessary for the activities on the four path segments. Therefore any of the path activities can potentially elongate or shorten the path. I agree that few project managers understand how to use Free Float to maximize the use of critical resources.

What is your opinion? We'd like to know!