Critical Path

Hi Omar



You have described a scenario where no one knows what they are building - in which case no one can plan it in any detail so you just have to muddle through and plan it as you go along.



That is OK in your situation but there is no excuse for doing the same thing in a BOG standard structure that every one knows how to build even at concept stage.



But every planner on this site - with a few noble exceptions - does just that.



They are just lazy or ignorant.



Best regards



Mike Testro

Hi Mike - that is exactly what I am saying!! At the present time I am working on a project for a type of plant which uses a process which is relatively new. Even after the concept stage, at pre-feasibility where, traditionally, scoping options are explored - not only is there uncertainty in some areas of the plant about which type of equipment will be used but what size any particular piece of equipment will be! Technical specifications that appear suitable early on may radically change as the maturing process design throws up new requirements. Project development in the processing/oil & gas/chemical industries is a process over a relaticely long time of a maturing design which consequently requires a maturing schedule/cost estimate. It may be theoretically correct to develop schedules from the bottom-up but in the project development I am involved in there is little point in ’Level 4’ schedules early on when you don’t have the ’Level 4’ information on which to base it.



cheers,



Omar

We call it the library of typical fragments.



This library is must part of PM methodology. It is created by every company where Spider Project is implemented.



We have one improvement - adding fragment to the schedule project planner enters the volume (quantity) of work in the fragment and the software adjusts everything else (volumes of separate activities, material consumption, durations).



Regards,

Vladimir

Hi Anoon



Precisely.



If there are a number of alternative forms of construction either plan for each as an alternative or chose the longest to start with and put in some buffers.



I do have a library of different types of construction that I can copy and paste into any new programme.



If I ever have to plan something new I always keep a copy in the library for when I may need it again.



And before anyone asks for a copy of my library the answer is no.



Best regards



Mike Testro

Hi Mike,



You may have an idea of what will comprise the project (roughly), but you will never know what would be the specifics (which is supposed to be bounded by a contract) at the concept stage. So how can you develop a level 4 schedule? By using templates, and change it once you got the right details?



cheers!

Hi Omar



You seem to be saying that at concept stage nobody knows what the project will comprise.



Whatever it is some one in the system will have enough detail to at least put a budget on it.



And somebody else will know how to build it.



If assumptions are to be made it is best made at level 4 with bottom up planning so that different options can be compared and evaluated.



In another thread I explained how I did a bottom up programme for a hotel based on an artists brochure.



I did three versions comparing the three potential structures - steel - insitu concrete - load bearing brick / block.



The developer could therefore proceed with realistic information.



Best regards



Mike Testro

Hi Mike - that sounds fine in theory but in the case of most major engineering projects they go through developmental phases such as concept, pre-feasibility, feasibility studies followed by the execution (EPC) phase. At the early stages of this process the degree of scope definition simply does not exist to prepare detailed shedules. Therefore EPC schedules are usually developed ’top-down’. Particularly at concept and pre-feasibility study stages any attempts at detail planning will carry no more confidence than at higher levels - scope, plant sizing etc. is simply not developed enough.



cheers,



Omar

Hi Khaled



I have just looked at your starting thread -



I build a Level 2 schedule with all activities related logically. To make the relationship between activities I had to add lags. Activities with zero Total Float are depicted on Critical Path. Which impact does the use of lags has on the accuracy of the Critical Path? Is it a prerequisite to have zero lags between activities to make a Critical Path?

I’m using Primavera ver. 6.



The use of lags completely distorts the critical path - never use them - same with SS FF links.



Your level 2 programe should be built up from a properly prepared level 4 programme.



If you don’t know how to build it - don’t try to plan it.



Best regards



Mike Testro

Vladimir,



Since it was your email that triggered my presenting at the Conference, I owe you one! I’ll send you all my Powerpoint slides as soon as they’re finalized.



I already am committed to a long cricket discussion with Pradip -- perhaps you and I can have a session of blitz chess together?



Fraternally in PM,



Steve D.

Steve and Dieter,

thank you both.



Dieter,

Linear Diagram (or Time-Location Diagram) is one of Spider Project standard graphical reports like Gantt Chart or Network Diagram.



Steve,

beer is fine. Please send me the examples you would like to show at the presentation.



Best Regards,

Vladimir

Vladimir



After Stephen’s hint I corrected this sentence. But I only considered time lag.



"Volume lag", that seems a really good feature of Spider - a combination between a scheduling software and e.g. Linea. I don’t know any other scheduling software with this feature - congratulations.



Regards

Dieter

I must confess, Vladimir, the more I hear about Spider, the more impressed I am! I’m looking forward to meeting you for a beer (a vodka?) at PMI’s College of Scheduling Annual Conference in Boston in mid-May. (I’ll be presenting about DRAG at 9:10 am on Wednesday, May 20th.)

Dieter wrote:



"Independent from software, if you’ll consider what a scheduling algorithm does with lags. They are regarded equivalent to activities."



No, Dieter, it depends of the software.

In Spider Project you can define time lags or volume lags.

For time lag you can define any calendar. And each time lag can have its own calendar.

Volume lags are measured in the quantities on preceding activities. SS with 400 meters Lag means that the following activity can start after 400 meters on preceding activity were done. And since the distance shall be kept it is not easy to present such lag as an activity. We can come to a lot of dummy activities that will make understanding the schedule too complicated.



Regards,

Vladimir

Was it really???



Well, they say that copying from a template is one of the most efficient ways to plan a project...

As it was copied from your post.

Dieter, perfectly phrased, I think! Thanks!



Fraternally in PM,



Steve Devaux

Stephen



Thanks for correction. "Regarded" I used in the sense of treated. If you’ll have a SS-relationship with a positive lag there is a start and an end, i.e. a certain time frame covered - no work of course. My concern was to focus on the calendar used for this time frame.

"Lags on FS and FF relationships are "equivalent to activities" in that they add time to the project (i.e., have their own DRAG) that is NOT incorporated in the work of either the predecessor or the successor." would be my idea in proper English.



Thanks

Dieter

Dieter wrote:



"Independent from software, if you’ll consider what a scheduling algorithm does with lags. They are regarded equivalent to activities."



Hi, Dieter. I just wanted to tweak one statement slightly. The above is completely true when talking about lags on FS relationships. And, of course, earlier in the thread, we had established that we were recommending using FS relationships exclusively.



But I just want to point out that IF we were talking about lags on complex (fancy-shmancy?) relationships, the statement that they are "equivalent to activities" would not be completely true. Lags on FS and FF relationships are "equivalent to activities" in that they add time to the project (i.e., have their own DRAG) that is NOT incorporated in the work of either the predecessor or the successor.



But in the case of lag on an SS or SF relationship, the lag typically represents work already accounted for in the predecessor (i.e., DIG INITIAL 400m. OF TRENCH as an SS+3 predecessor to LAY PIPE). Then the DRAG is associated with the predecessor’s work. Shortening the time taken to dig the 100m. from three to two days would (should!) reduce the lag to SS+2 and, if on the CP, shorten the project by one day.



Fraternally in PM,



Steve Devaux

Khaled



Independent from software, if you’ll consider what a scheduling algorithm does with lags. They are regarded equivalent to activities. Now it’s becoming dangerous for the critical path: which calendar should be the basis: Project-, default-, predecessor’s, 7-days per week,...? In P6 under the scheduling options you can select. But this will be valid for all lags.

So, for my opinion it’s the best you’ll follow Samer’s advice and use activities instead of lags.

But I understood that you just have a rough level 2 schedule with many parts not yet properly planned. So the critical path will not be very reliable anyhow.



Regards



Dieter

Dear Khaled,



I would suggest that you complete the link with using another activity instead of the lag. Call it curing time/ checking/commissioning/ of standby time.



Going down with the level of detail and using the correct sequence of work, is better than assuming how the critical path is being calculated or what is the meaning of your results.



Use simple logic and simple sequencing and you will get the logical answers that you are seeking and start using the Program more effectively to represent the actual progress on site.



Best Regards,



Samer

Dear Khaled,



You will need to make sure that your "lags" and schedule relationships are correct. If the logic of your Schedule is correct, then you can proceed with the resources and the durations, and the schedule development.



Subsequently, you will have a Critical Path. The Critical Path will have activities with Zero Float. It is the minimum time to complete the project depending on your stated activites and until schedule development.



Best Regards,



Samer

Hi Khaled,

The use of lags like you described can lead to periods without any activity happening and even with this window of no activities it is shown as critical (which is not understandable) Like Devaux mentioned is far better to replace the blind lags by activities allowing the understanding of what is happening in the schedule (project).



I suggest you develope your detailed schedule and than make summary activities (can use activity type "Level of Effort") to show a lower level of detail.



Hope it helped.

BR

JMFrade

Khaled,



I don’t know P6, but I’ll try to give you an answer about general usage, independent of software package.



Using lags, in and of themselves, should have no impact on the ability to define your critical path (provided all the lag and logic are correct, of course!). The biggest problem for you with lags will likely be the need to go in and manually change the lags as either optimization or work is performed.



NOTE: This does NOT address complex dependencies (SS, FF, SF), which CAN result in "strange" happenings in critical path computations if the user is not careful. However, you don’t mention using such links.



But you say that "To make the relationship between activities I had to add lags." I confess, I don’t fully understand this. A lag on an FS (or even FF!) relationship functions exactly like an activity (and can have DRAG if it’s on the critical path). Indeed (as Mike Testro would probably point out!), any complex or lag relationship can be modeled using a combination of FS relationships, additional activities, and milestones. And the result will often be a clearer network where it’s easier to identify the big DRAG activities.



Hope that helps, though perhaps an expert with P6 will weigh in.



Fraternally in project management,



Steve Devaux