Float May Be Sinking the Project
Once upon a time, project scheduling was a core methodology used to develop, monitor, report and, most important, direct project execution strategy. However, financial and contractual stakeholders soon spotted the schedule's wealth of data. Things haven't been the same since. Financial departments started to require the schedule's level of detail, work breakdown structure and reporting processes. These demands weakened the schedule as a temporal tool.
But it was the contractual sector that drove the final nail into the schedule's coffin. There must have been some "Reese's Peanut Butter Cup" moment when some consultant accidentally realized a "new use" for the Critical-Path schedule's key variable: Total Float.
Float alone could be used to determine what aspects of a project schedule were "critical," the thinking went. Any impacts to such "critical" activities could be used to justify time-extension requests, owners argued. In theory, this made sense. In practice, it has proved difficult to administer fairly or consistently.
Float manipulation is today's poster child for project management's loss of perspective. Clearly obsessed with the financial and contractual implications, the schedule is often used to "manage the float," rather than the work.
At its core, float is—or should be—a simple measure of available wiggle room. Properly used, float should inform a project manager's decisions. ICS-Global studies confirm that most project delays stem from coordination issues that could have been mitigated or eliminated through responsible use of a competent schedule.
The question of "float ownership" is moot until we find a way to credibly dissect the individual contributions to float gains or losses. I'm not talking about forensic delay analysis—that only deals with major "events" and disputes, and even then after-the-fact.
 |
| Woolf |
Float is a group attribute. However, without incorporating causation into the equation, talk of float ownership is disingenuous. For example, a contractor, upon learning the approaching winter will be harsh, beseeches the subs to work longer days in order to "bank some time" as a contingency against the weather. But once the float appears in the schedule, the owner imposes a scope change while he points to the "shared float" contract clause.
How fair is that? The owner has no claim to this time; the owner did not pay for it. A contractor has a right—and a need—to pepper his schedule with wiggle room, both as a time contingency for unknowns as well as for "space between the dominos." It's like leaving six car lengths for safety and others cut in. It's unfair and unsafe.
In the absence of fair and practical rules, chaos has ensued. As a result, loopholes and double standards contaminate Project Time-Management processes and products.
For example, while a contract may strictly prohibit contractors from sequestering float, owners reserve the right to do just that for their activities. Consider the owners' demand of four-week durations for all reviews, regardless of submittal complexity. We expect contractors to assign durations commensurate with the work, but owners exempt themselves from this rule. Owners contend they must hedge against all submittals coming to them at once. Let a contractor try that argument and see what happens.
Tug of War
Increasingly, schedules are a rope in a tug-of-war between owner and contractor. A schedule's content is contorted so as to posture the competing parties for financial or contractual advantage.
The good news? Float is not the only game in town. There are other, completely credible options. But until owners are courageous enough to ask for them, those who benefit from the current dogfight will maintain their silence. As long as the schedule is the primary route to claim resolution and progress payments, contractors and owners will game the system. There's no one to blame but ourselves.
Eliminating float as a way of determining delay would be a huge first step toward returning Critical-Path Method schedules to what they were in the beginning: a model of Project Execution Strategy used to coordinate and direct the work. Until that happens, projects will continue to finish late and over budget because (a) a distorted plan is worse than no plan and (b) those who fail to plan, plan to fail.
Murray B. Woolf is president of ICS-Global, a project scheduling consultant. He can be reached at mwoolf@ics-global.com.
CPMdoes not even theoretically give the shortest schedule. About 50 years<br/>ago, Ronald Graham, a mathematician at Bell labs found that arranging computers in CPM fashion actually mad...
ago, Ronald Graham, a mathematician at Bell labs found that arranging computers in CPM fashion actually made the job take longer than if the computers were arranged some other way.Investigating why, he found that CPM could take up to twice the time needed by an optimal schedule.
Since CPM does not give the optimal schedule, some other principle or principle(s) must be sought for seeking either damages for delay or relief from damages for delay and in the usual contract there are ample reasons owners and contractors can use to press forward with or defend against delay claims.
how do you arrange computers in CPM fashion? You are talking about a manufacturing process not a scheduling method.
The problem Ronald Graham was trying to solve some 50 odd years ago was to speed up a bank of computations so there would be enough time to intercept Russian bombers. Some parts of the...
Graham arranged his bank of computers in CPM fashion. That is he drew up a network of the hierachy
of computations that depended critically on the completion of other computations with those computations taking the longest time choosen first. This is the "critical path". To his surprise the computations took longer than when he had arranged the computers in some other fashion.
Investigating why this was so, he found out that the CPM rule of always choosing that activity that takes the longest time is no guarantee that this will give the shortest schedule time.
Granted this is against "common sense" so it fooled even someone as brilliant as Ronald Graham, but it is a solid mathematical fact.
Incidentally many years ago,I was coding for punchcard input to a CPM program and the incomprehensible result at least to me was that there must have been some error in the input to the CPM program, for the "critical path" had painting of the outdoor housing of the electrical equipment on the "critical path" while testing especially stage wise testing was not on the "critical path". This certaintly made no sense, for painting could be done any time especially when the weather was fine and there was slack time, while testing, especially testing in stages was absolutely critical so that any error is not buried too deep. I think it was entirely possible that the input to the CPM program was O.K
but the CPM logic have this nonsense output.
Thanks for asking.
Good points Murray. Our industry would benefit by focusing more on actually managing the flow of work and less on who might be trying to steal the ether. Your analogy between float use ...
Mark Sanders
Continuing the weather example of the prudent contractor who makes extra provisions for an especially harsh winter, should a contractor on an incentive contract benefit if milder than e...
Mr. Woolf touches on a core problem with schedules on a construction project in that they are no longer used as a management tool, but instead as a rigid legal document. The problem wit...
Shakespeare has the "fool" in "Twelfth Night" say "But, indeed, words are very rascals since bonds disgraced them."<br/><br/>Beautifully said! written contracts instead of word of honor...
Beautifully said! written contracts instead of word of honor has not been an unmixed blessing. I suggest that engineering education incorporate a course in the Philosophy of law and engineers be encourage to write contracts and not lawyers. Furthermore anything in the contract that is not germane to the contract should be taken as clear evidence of incompetence.
It would seam that a contractor could retroactively address time lost due to weather related causes by requesting extension of time accordingly. However, most weather related extensions...
Hamid K. Toossi, B.S.C.E., M.B.A. - (previous comment repeated with additions and corrections) - While all efforts should be made to utilize Critical Path Method (CPM) schedules for its...
It would seem that a contractor could retroactively address time lost due to weather related causes by requesting extension of time accordingly.
However, most weather related extensions, if granted, become excusable but not necessarily compensable.
In addition, a contractor could include contingency for weather related delays in change order schedule (fragnet), by assigning longer durations due to anticipated lower productivity and higher down time due to unsuitable weather (prospective planning). Impact of such longer durations, as justified later by actual weather conditions and reflected in subsequent schedule updates (retrospective analysis), could then become grounds for excusable and perhaps compensable time extensions that would then be ratified via appropriate modifications to contract price and time.
Of course, this is always easier said than done. However, for our hypothetical "forward-thinking-contractor" assisted by a professional construction claims expert (i.e. Hamid K. Toossi, B.S.C.E., M.B.A) such contemporaneous scheduling and time impact analysis could become easier and more do-able...
Murray Woolf -- <br/><br/>Hamid, thank you for your comments, which I certainly appreciate, but I think they miss some rather important points that readers would be wise to consider:<br...
Hamid, thank you for your comments, which I certainly appreciate, but I think they miss some rather important points that readers would be wise to consider:
First, the weather example was introduced to demonstrate the unfairness of an Owner usurping an additional “bank of time” that a proactive and responsible general contractor (and its subs) paid for with their own blood, sweat, and tears. The obvious impetus behind taking proactive measures was to avoid a delay, not posture for one. Your response immediately gravitates to “retroactively address[ing] time lost due to weather.”
Second, the ultimate and full impact to a contractor of inclement weather is always far greater than can ever be recompensed through a delay claim for bad weather. Let us set aside, for the moment, that under most contracts a contractor is only able to pursue a weather claim for “excessive” inclement weather (that’s my next point, below). Every project has a momentum to it, and much of that momentum can be significantly slowed by inclement weather than hit intermittent activities across the landscape of activities in the schedule. The prudent and experienced contractor knows that it is better not to lose the time in the first place, than to try to recover some of the damage after the fact.
Third, as promised above, I must point out that only inclement weather in excess of what a contractor would be expected to normally anticipate is acceptable for consideration in a weather claim. In my example, the contractor was trying to avoid or mitigate the effects of such “to be expected” bad weather.
Fourth, the inclusion of weather contingency in a schedule is a very complicated topic, and there are surely many different schools of thought on the matter. First there is a question of WHETHER such contingency should be incorporated at all. Second there is a question of WHERE to put such contingency. Third there is a question of how much contingency to allow. This forum is not the place to discuss all of these variables.
One popular option is to allocate weather contingency in an automated calendar and then assign weather-sensitive activities to that calendar. In this method, decisions have to be taken as to how much contingency, and where to put it in the calendar.
Another popular method is to bump up activities durations for weather-sensitive activities. I may be misreading your comments, but it seems as if you may be advocating this approach, where you say that “a contractor could include contingency for weather related delays in change order schedule (fragnet), by assigning longer durations ….” I am strongly opposed to deviating from the basic principle that activity durations reflect the amount of work to be performed, the assumed productivity of the activity performance team/crew, and the conceived configuration of that performance team/crew.
The immediately obvious flaw in bumping up durations for weather-sensitive activities is what happens if the schedule gains or slips significantly. What was anticipated for winter may now occur in the spring, and what was expected in the summer may not hit in the winter. A scheduler would have to constantly revisit durations to keep them relevant to the ebbs and flows of schedule status.
Fifth, and as I talk about at length in my book, one must be very careful when injecting a schedule with time contingency (for any purpose .. be that weather, risk, etc.) because such injections can become self-fulfilling prophecies. Schedules, when properly used, are instrumental in coordinating and directing the work. If a schedule is inflated with contingency, then the “earliest” dates for a given trade are artificially later than they would be in the contingency was not in the schedule. A Project Manager, relying on the Earliest Dates of the schedule, would reasonably rely on the schedule to inform the trades (or fabricators) as to when their services (or products) are needed “at the earliest.” If the risks or bad weather (for which the contingencies were injected) do not materialize, then the “earliest dates” advance as “scheduling windfall.” Now the PM has to pay premium acceleration money to get the trade or materials to site “earlier than scheduled.”
Your comments illustrate a point I try to make to Owners whenever I have the chance. I tell them that they need to decide whether they want the Project Schedule to be a project management tool or a contract enforcement tool. It cannot do both effectively.
Muray Woolf
