Guild of Project Controls: Compendium | Roles | Assessment | Certifications | Membership

Tips on using this forum..

(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?

(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.

Most Important Issue In Scheduling: Contingency

10 replies [Last post]
John Reeves
User offline. Last seen 1 week 4 hours ago. Offline
Joined: 10 May 2013
Posts: 132
Groups: None

MOST IMPORTANT ISSUE IN SCHEDULING: CONTINGENCY.  Why? Estimates & Budgets have it quantified.  But how do you put it in a schedule? and if you don't you get issues like being "behind" on a 3 year project the first month - or of course you start chipping away those days from a whole bunch of places instead of "contingency".  And if you have your original schedule ending a month early, ie a month float - that is generally "not allowed" but if it is then you have the widely popular trend of forcing a 0 float path through the path of the critical even though it is ending early which destroys the whole purpose of "float", and make any monthly comparisons worthless because you are moving the end date a little bit each month to keep it 0.  My solution would require a change in the software allowing a contingency pool on some items - basically forcing you to predict where you will steal those days from, but contractors would not like having to provide that.  Having "a contingency item" at the end destoys the point of float.  All of your items are basically potentially "off" by that much.  I searched this issue and the last post was from 2008.  What is your solution?

Replies

Rafael Davila
User offline. Last seen 17 hours 39 min ago. Offline
Joined: 1 Mar 2004
Posts: 5029

But internal planning must be done for 12 resourcers. 10 resources plan can be used for setting contract dates.

  • Agree, the problem is when internal planner and PM do not understand the need for the two schedules.  If you use 10 workers for management when the 12 are available (no one on leave, sick or absent for whatever reason) you might delay the planning. For the short run schedule management use 12 resources schedule for the long run use 10.

But internal planning must be done for 12 resourcers. 10 resources plan can be used for setting contract dates.

Rafael Davila
User offline. Last seen 17 hours 39 min ago. Offline
Joined: 1 Mar 2004
Posts: 5029

Sometimes it is good idea to keep some resource contingency outside the schedule.  Like planning for 10 resources A but hiring 12 resources A knowing at times some will be on leave or absent. If you include the 12 resources A in the schedule you might lose the reserve.

Look at the presentation made 10 years ago at PMI COS conference.

It discusses Success Driven Project Management based on contingency management.

http://www.spiderproject.com/images/img/pdf/Success%20Driven%20Project%2...

Xavier Ader
User offline. Last seen 15 weeks 5 days ago. Offline
Joined: 9 Feb 2021
Posts: 3
Groups: None

Interesting thread! I agree, contingency is a significant issue in scheduling.

 

Before I go into how I usually address it, I'd like to start by explaining my understanding of contingency.

To me, contingency is a "time-risk allowance" (as stated in Rafael's response). In other words, it is a time allowance coming out of a risk assessment carried out on the project and reviewed on a regular basis. The implications are 2 folds:

  1. Assuming the risk assessment has been well informed and carried out, the whole contingency is likely (in the probabilistic sense) to be consumed by the end of the project;
  2. It is impossible to determine in advance which risk will occur and at what time (hence why Monte-Carlo type analysis are being used).

 

The way I prefer to apply contingency to schedules is similar to Method Two described by Santosh, but instead of applying non-work contingency days to calendars I use hours by reducing the number of available hours each day.

In P6, it means that I set the 'Hours per Time Period' at say 10 Hours/Day and the 'Calendar Weekly Hours' at say 9 Hours/Day. For a seasonal risk like wet weather, I leave the Standard/Calendar Weekly Hours at the same as the Time Period (e.g. 10 Hours/Day) but manually overwrite the Work Hours/Day for each day. This allows me to introduce a risk profile by reducing more the available hours during the wetter months.

Note that this method is practically not usable in MS Project, as the Time Period is set for the whole project, not specifically for each calendar. (One of the many reasons why I don't like MS Project...)

 

Hope this helps.

Rafael Davila
User offline. Last seen 17 hours 39 min ago. Offline
Joined: 1 Mar 2004
Posts: 5029

The NEC mandates some explicit contingency:

However, the other two types of float are not jointly owned and are solely the contractor’s. These other floats are ■ time-risk allowance ■ period between planned completion and the completion date (commonly known as terminal float). 

Contractors-float

John Reeves
User offline. Last seen 1 week 4 hours ago. Offline
Joined: 10 May 2013
Posts: 132
Groups: None

I appreciate everyones input, the kiss of death on the internet is long replies but here it is.  I had never heard of drag – evidently drag “represents the individual contribution of any single schedule element (e.g. a task) to the overall duration of a project (see wikipedia)” I have been scheduling 33 years and do not have a good answer for the contingency issue – this is why I consider this the biggest issue.  Note having a pool at the end is what I do NOT prefer, not for.  Example, if you have 20 days of float contingeny, and you use 3 days of it per month, even going negative 12 months later.  Then you need to compare any month vs any month, you have to work all of those moving parts into the comparison – on a large jobs with changes that is just too much.  And if you have ever been to claims court – simplicity is paramount.  I have had to do that on airport jobs – sometimes the spec said use a pool, lesser times it was through the calendar.  The jobs went fine but those were only 2-3 month runway type jobs.  The best answer I can see is showing prior completion with some float and have a critical path with float.  Having a pool at the end is what I am against.  This also implies using a 0 float path that is ahead of the real contractual date.  Here is why – when your end date is free to slide later, when you compare floats and variances from month to month that becomes obtuse or useless.  Also, if you use a pool at the end, it may be fairly easy to do the math of changes on the critical path, but the non-critical path effects and math become very tough.  I think the paths would be effected differently with a pool vs natural float, in particular regarding changes and milestones and complications.  I have worked on big jobs with tons of changes month to month.  Starting with float is the only way to compare any month to any month and have the math correct.  People do not like the idea of starting a job with float because they say its harder to identify the critical path, just use the longest path and make it red, not that hard.  When I work for contractors I loved going to 0 float, I could claim delays very easily, when on the other side I consider the contingency issue more. Often times the spec states the baseline must have 0 float, so I would be at odds with the spec I am enforcing with my preference.  My next step would be to produce an example of what I am talking about, I will do that.  Santosh Bhat made a good point in his article that Contingency and Float are not the same thing-even though I have compared them here.  Of the five methods he listed: I liked Method 0ne (because I am a pareto principle guy), but on project used Method 4 the most (increased overlap), 2nd is the pool but noting it is (partially) moved to the delay not removed, Meth.04 relationship overlap, Meth.03 Pool, Meth.05 lags Perhaps is the answer is the truth is the truth and if you are 3 days late after the 1st month you make your changes that are best and specific at that time instead of having to decide if an issue is pool worthy and who owns the pool.  Or use a little bit of both and just deal with the non-sensical variances.  Awesome article:

https://www.austprojplan.com.au/schedule-contingency-methods/

Zoltan Palffy
User offline. Last seen 8 hours 3 min ago. Offline
Joined: 13 Jul 2009
Posts: 2938
Groups: None

John 

you said that you wanted a contingency pool at the end of the project this would require that in the planning stage that work is scheduled to finish PRIOR to the contractual completion date.

Option #1

You can still do what you want but instead of having them all at the end of the project the contingency days would need to be dispursed along the critical path of the project. In some cases you may use these days and in some cases you will not. So as you progress along the critical path you using up some of the contingency that was put into those durations.

Option #2

Add additional non working days into the calendar but actually work those days.

Option #3 

Work Saturdays

Stephen Devaux
User offline. Last seen 12 weeks 5 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Hi, John.

I certainly agree that schedule contingency is an important issue, particularly on contract-driven projects. Not only is it needed to avoid pretty much every project being delivered late, but it also needs to be a specific and identifiable line item as opposed to being "activity padding", an invisible "slush fund" there to cover subcontractor inefficiency. Projects without contingency often take unwise risks, especially when recovery from slippage is needed.

Understanding of the need for contingency in the contract has improved considerably over the 30+ years I've been consulting. US DoD used to be particularly opposed to inclusion of contingeny--but that has greatly improved over the past 12-15 years.

However, there are still a LOT of issues that are neither well understood nor managed:

  1. There is no magic bullet for estimating the amount of contingency needed. Three-point estimating & Monte Carlo systems can be of help. Of even more value is a good historical database of similar projects. But estimating an appropriate amount is still more art than science.
  2. You mention that having a "0 float path" 'destroys the whole purpose of "float" '. I'm with Santosh when he says he doesn't know what this means. A good scheduler knows float totals are relative: If I started with TF = 0D and now have TF = -12D, I KNOW that's exactly the same as if I started with TF = 20D and now have TF = 8D.
  3. What IS vital is knowing the critical path drag and drag cost of any schedule contingency! Indeed, drag is always more "critical" than float! If the contingency is all at the end of the schedule/contract with nothing in parallel, then computing its darg is pretty simple: drag = its duration. But any contingency placed before key milestones in mid-project may have drag that is tricky to compute on large and complex schedules. (Especially if your s/w is not Spider or Asta Powerproject, which compute drag!)
  4. Even more important, what is the cost of that drag? Since the value of a project usually doesn't accrue till after it's completed, this is vital info for the customer. Will being 12D late mean that 12,000 more people will die of the virus? That the Olympics will have to be cancelled? Or that we'll miss renting our hotel rooms for 12D of the tourist season? In 1747 Ben Franklin wrote: 'Time is a whole bunch of Benjamins!" And yet all these years later, contracts are being written and projects scheduled & performed where there has been NO quantification of the cost of time! And without that at the project level, schedule optimization and wise scope/time/resources trade-offs at the activity level are impossible. And that's even if we DO take the trouble to compute critical path drag. 

Finally, a Monte Carlo system that estimates probabilistic critical path drag and drag cost would be helpful--but I don't know if any such package currently exists. Perhaps Vladimir Liberzon, Rafael Davila or Ben Taunt would know...

Fraternally in project management,

Steve the Bajan

Santosh Bhat
User offline. Last seen 1 day 5 hours ago. Offline
Joined: 15 Apr 2005
Posts: 331

Here's what I wrote about it years ago:

Schedule Contingency Methods | Australasian Project Planning (austprojplan.com.au)

I don't understand this part of your post: "Having "a contingency item" at the end destoys the point of float.  All of your items are basically potentially "off" by that much"

 

If the contingency bar is at the end, then it doesn't make any of the precedding activities "off" as such, only any tasks after the contingnecy bar.