Does negative lag work and or make sense in the case of a mandatory Red Zone meeting 60 days prior to the end of the project? I ran this scenario by Army Corp. and they approved the use of negative lag in this situation.

Anoon...

Apologies for my wariness!

I think this discussion needs to be carried out on another thread. I think it has already been hijacked by a discussion on negative lag and now on how much experience is required to make a competent planner.

All I might add is that one of the measures that goes to make a good project model is the competency of the person who puts it together. Whatever your plans are like, even if your experience is not as great as others, in my opinion it will be better than having no plan at all.

Little by little you will gain experience and your employer will have more and more confidence in what you do. Apart from taking every opportunity that comes your way and working to the best of your ability there is nothing more I can suggest to help.

For the record, my first planning/programming job was in 1977 and some people, especially those on the opposing side in a dispute, still say what I produce is rubbish - but again, they would wouldn’t they!.

David

Anoon

Thanks for your question.

I see a programme as a ‘project model’. So if you think of it as any other model it depends on the information you have, how detailed you make it, how realistic it is, the skill of the model maker, the resources and finance available and the time allowed to make it.

The poorest model is one that is made with unskilled model makers, with little information, no resources and no time. Ultimately the best model would be a full size working version of the real thing (a prototype maybe).

Whenever the project model is put together all I expect is that it is made using what ever information is available at the time, by skilled planners/programmers using a reasonable amount of time and resources. But you know all that anyway.

Perhaps you are trying to say that a project model will always be a compromise, and I agree. I’d like to think that the majority of project models I have made are reasonable (and reasonable is the key word) - but I would say that wouldn’t I? Whether they be just a single line statement saying something like, the average time to construct this project is X weeks with a 95% confidence limit of A and B, or a detailed ‘construction master programme’ put together with all the information available prior to construction starting.

I think what is important is that the project programme is never considered to be a static document and that it is updated as and when additional information, changes and so on become apparent. That begs the question, is the only true programme made at the end of the project when all the information is available? I don’t agree; an as-built programme is a record of the work carried out, not a tool for carrying out the work.

And… as you will have seen from my recent postings, if I have a difference of opinion with other posters I will say so. Thank-you for your courtesy, I am not used to it, I hope you do not have an alternative agenda!

David

David,

We appear to be singing from the same song sheet - maybe we’re just of that generation!!!!!!!!!!!!!!

To throw my 2 cents worth in - one of the problems, particularly in higher level scheduling, is the use of the SS with positive lag eg start concrete foundations in an area and so many days after the start of concrete the steel erection can start. If the conditions/duration assumed to be within the original lag from the start of concrete are not met then the logic may still be correct (ie steel must start after the start of concrete) but a delay/change in the conditions upon which the duration of the SS lag was based means the original schedule assumption is incorrect. Note that if the activity "concrete’ is expanded into several lower level activities, allowing simple FS relationship(s) to steel then a delay in one of the concrete activities will correctly model the actual situation & forecast a delay in the start of steel. Like any scheduling logic - its fine if you understand and have recorded the ’basis of schedule’. cheers, Omar Grant

Oliver,

This is the fundamental of planning.

We plan using the best logic that the our western education had given us, using our best experienced, our best reading materials, our very best all the way, we plan thinking in the most positive way that such plan is feasible.

We dont plan using emotions, negative hypothetical scenario that the foundation will be delayed that is why you use negative lags.

We plan based on our exeperienced, our understanding of the scope of work, the best logic on how activities will work believing it will be done as plan because we have people and resources to implement the plans and schedules.

Now in the event that the foundation will be delayed and the steel work contractor have mobilized, then, we have to revert to the contract provisions and determine who is culpable to the delays and cost incurred.

Take note that planning is different from monitoring actual site conditions, actual site progress. It may be possible that due to tremendous delay of the foundation work, the project management may instruct the steelwork contractor to delay the mobilization.

The above scenario will not justify the use of - negative lag at the planning stage.

I hope you are now enligthen on the difference between best planning practice and preconcieve hypothetical negative scenario that may distort logic.

still you are free to share your negative lags

Cheers,
Charlie

Oliver, your idea????
"If the scenario was the mobilisation of a steelwork contractor 1 week prior to the foundations being finished , how would you drive the mobilisation out if the foundations were delayed?

The mobilization of steelwork contractor can be done in a lot of ways. But the start of steel erection works will follow after the finish of foundation. The scenario maybe that the steelwork contractor have mobilize but cannot start steelworks installation because the foundation was delayed.

So no need for this: "The use of FS -5d working days lag would automatically push out the mobilisation of the steelwork contractor, thus telling the PM when he needs ready the steelworkers."

my preference is never to use - negative lags.

Cheers,
Charlie

David,

I don’t want to contradict someone who I otherwise agree with (which is pretty rare around here) but the tail doesn’t wag the dog.
The predecessor’s (early) start or finish determines the earliest date for the successor’s start or finish, as the case may be, not the other way round.
For FF with no lag "Task B can finish no earlier than Task A’s (early) finish".

Oliver, I think it is wrong because it requires an assumption that something will happen in the future at a certain time/date even though we know, from all experience, that it is unlikely to happen on that time/date. That is a lot different to waiting for something to happen and taking that as a signal for the next thing to happen. It may have been stated or asserted that -lag is ok many times but just repeating it doesn’t amount to proof.

Oliver...

Sorry if I appear to be teaching you to suck eggs but this is the best explanation I have for those that cannot envisage the problem (and apologies for not being able to cite the source).

Take a track runner. The logical sequence of events between the staring gun ‘bang’ and starting is F-S. When the gun goes off the runner springs from the blocks.
Imagine the starting gun being more like ‘baaaaang’. To get maximum advantage the runner must spring on the starting ‘b’ not the ending ‘g’. That means the runner has to anticipate when the gun is going to go off so he can start his spring before the bang has ended. Fantastic if he gets it right but massive risks if he gets it wrong.

Its just a matter of what is logical and not logical. The S-F link says to start something before you can know when to finish it. Your timing has to be precisely right and precise timing is just not possible.

See, this is again where I have difficulty, it is the recurring problem that learned planners seem to think S=F and negative lead/lag is good practice whereas everything I have learned and practised and understand says the opposite. Is it me at fault I ask myself or is our discipline doomed to failure?

David

Trevor

I think it is right, just draw it out to see…
Say the first activity (predecessor) starts day 10 and finishes day 20.
The second activity (successor) finishes day 10 (and starts sometime earlier).
That means that the predecessor always finishes (10 days in this case) after the successor
A good example why the Start to Finish dependency is so dumb.

David

Apologies to all as I have joined this thread near its end but I am always enlivened by the debate on negative lead/lag and S-F links.

I recently wrote a critique of an analysis carried out by one of the UK’s top consultants who had, in their repair of the master programme, used negative lead/lag and S-F links. In my report I suggested that this was not good practice and received a somewhat terse response suggesting I did not know what I was talking about. Again, I am happy to admit I might be in error but if I am correct I do fear for the future of our discipline if some of the country’s leading experts dare say that negative lead/lag and S-F links are acceptable good practice.

By way of illustration I submitted three references that you might like to use when faced with a similar problem:


Construction Scheduling: Preparation, Liability, and Claims
(Jon M. Wickwire, Thomas J. Driscoll & Stephen B. Hurlbut. Construction Scheduling: Preparation, Liability, and Claims. Aspen Publishers, December 2004.)

Negative Lag Relationships. Still another Corps of Engineers clause pro¬hibits the use of negative lag relationships where precedence diagramming and logic is used for the network analysis system. Recall from the discussion in §2.03[A1] that precedence diagramming allows the scheduler to avoid the use of dummy activities (to show logic relationships) through the use of lead and lag connectors. Thus, a lag relationship might allow for activity L to start once 25 percent of activity J has been completed. However, a negative lag relationship would provide that a given activity, such as activity S, could start at some minus date before the completion of another activity, say activity M. Thus, the diagram logic might provide that S could start 180 days before the completion of M. The problem here is that we have no absolute assurance that M will indeed be finished in 180 days such that S can start now - in fact, the negative lag logic does not make a lot of sense. The Corps of Engineers clause states: "Lag durations con¬tained in the project schedule shall not have a negative value." E. William East states, “In my opinion, negative lags have been included in scheduling software to assist contractors to obtain ’correct results’ without constructing a project model that is a reasonable representation of the planned construction process”36.

36 Letter from E. William East to Jon Wickwire (Sept. 11, 1995)


Managing Projects for Success
(Albert Hamilton. Managing Projects for Success. Thomas Telford, 2001.)

The term negative lag (lead) is used to situations that permit succeeding activities to begin before preceding activities have been completed. However, this type of relationship can he shown using an SS relationship with positive lag (simply reverse the arrow). SS relationships with negative lag are complicated and difficult to understand they should be avoided.
An FF dependency links the finish of the predecessor with the finish of the successor. It indicates that the predecessor activity cannot finish until the successor activity has finished. FF relationships are used to show the relationship between the finishes (or completion) of two activities. Again, negative lag is difficult to understand and increases the complexity of the calculations to determine the critical path. Use of negative lag should be avoided.


Network Scheduling Techniques for Construction Project Management
(Miklos Hajdu. Construction Scheduling: Preparation, Liability, and Claims. Springer, December 1996.)

Some schedulers like to use negative lags to make activities concurrent. They say, `Activity B can start 3 days before activity A finishes, so the right relationship is FS -3.’ However, this logic is false. Imagine that activity B has started before the scheduled finish of activity A, but due to some problems, activity A has stopped for a week. According to the network logic, we have a situation which is impossible. The lag time between the finish of A and the start of B is less than the prescribed value. This means that the network logic was false. The correct relationship would be that `activity B can start after x days of the start of activity A,’ or ’activity B can start after y percentage of A has been completed,’ etc. This is the correct logic to express these kinds of activity interdependencies. Negative lags can produce the same start and finish values with wrong logic. It is used to schedule an activity to a desired date without describing the real activity interdependencies.

A network logic containing negative lag values in case of minimal type of relationships is never good, therefore one must handle the results of this logic very carefully.
Summarizing the above mentioned, avoid dangling activities and never use negative lag in minimal type of precedence relationships.

Regards

David

fellow P & S Folk

Forgot to give the details of Bramble & Callahan book on Construction Delay Claims.

Which are as follows, note I am not promoting these gentlemen, but I have found them very useful.

I can already hear the whinges about the price of the book, but as I am on record as saying that P & S folk generally are not poorly paid and want to paid what they are worth then the same is fair for others !

However that is another subject, so the book is

Construction Delay Claims, Third Edition
by Barry B. Bramble, Esq., Michael T. Callahan, Esq.

List Price: $236.00
ISBN: 9780735505483
Page Count: 888 pages
Format: Paperback
Publication Cycle: Supplemented annually
Last Updated: 8/28/2007

Raf

Andrew

I would certainly be interested in seeing this.

Keith Pickavance and Fenwick Elliot solicitors wrote a suite of supplements to the standard form UK building contracts (called the PFE Change Management Supplements), which put the principles of the SCL Protocol into a contractually binding form.

These are also worth a read.

Oliver,

If we all did exactly the same thing, life would be very boring.

May I ask what is the aim of introducing such restrictions? Is it to make sure the contractor is "doing all the possible works; once they can be done"? (Early Track)

Well, in this case, I recommend the Client or Consultant to prepare a full detailed schedule; and includes it in the Tender Documents; so the contractor abides to it.

Contractual Milestones are used; and reasonably accepted; if you wish to have more control. However; the contractor must be free to plan the works according to his method statements.

Moreover, why the -ive lag exist in the planning world? Do you suggest to globally eliminate the SF and -ive lags from the planning world?

Andrew,

I agree in full with your last statement and personally feel there is little difference between each method.

However, i think the driving/dragging concept may be the reverse.

The use of ALAP is a constraint that ’drags’ tasks to the end of their ’free-float-path’? (I think i may have just made that description up!?)

Whereas, the use of FS with negative lag is just a delayed push from the front.

What does a fixed -(ve) lag do? Anything different? If you want to ensure the mobilisation is not left until the very last minute, (not recommended), increase the mobilisation activity duration:

OR

Put a +(ve) lag on the mobilisation FS link to the work

No, schedule the mobilisation activity on ALAP dates rather than ASAP and the effect will be the same.

The mobilisation activity is linked to the first work activity then with a FS link and the mobilisation activity will always move back in time if the first work activity does, (or any other programmed constraint).

Thats a proper way.....

SS+ x days with FF + x Days

Rather than having SF link

Plz correct me, if i m wrong

And thats why these contract people, always add:

OR as otherwise directed by the Employer’s representative."

For procurement, Clause is "With the exception of approval cycles and the procurement of material, no activity can have a duration of more than 21 days."

Neway, Ramadan and Summer time is mandatory in UAE at least and so, always considered even by the Contractor while bidding and so, its a part and parcel of calender in UAE.

Andy,

On teh current project i m working on (In Dubai, UAE), following r d restrictions on d programming part. We hav put this as a Schedule to teh Conditions of Contract.

- Negative lags not allowed
- SF relationships not to be used
- Activity Duration not to exceeed 21 days.
- Only 2 calenders r allowed
a) Calender 1, 8 hours per day with 5 day working week
b) Calender 2, 10 hours per day with 6 day working week.

And so on.

Normally, after award of Contract we have planning workshops wherein we define our WBS structure, Life cycle phases, S-Curve & Manpower Format and the layout of program and other sequence of works.

All to be submitted in 3 coloured hard copies with 1 full copy in native electronic format on CD.

Cheers,

Rav