Good software shall be flexible enough to simulate any situation in your projects. It does not mean that all software capabilities shall be used in every project. If the software cannot do something it becomes planner’s restriction.

Hi Trevor

There was no entry for "jhuggiuhu" in google but there is a couple of hits for "jhuggi uhu" I think you may have misspelt it.

Best regards

Mike Testro

Denni, just because software allows it, doesn’t mean that it is a good idea or that it is compulsory to use it, or that it should be splashed around at random.
There may be some special cases where it might be useful, but smack bang in the middle of a critical path network isn’t one of them. Careful judgement, which links to use where, and which not to use, makes the difference between good modeling and bad modeling, and you get to choose which you prefer. There are other things that software allows that should not be used except in very special cases maybe, if at all. MSP allows scheduling backwards from a finish. Someone must have asked for it, it might be useful in some cases, but is definitely nothing to do with the critical path method. MSP also allows date constraints. It even lets you type in "jhuggiuhu" as a task name but not a good idea.

Hi Denni

The only possible use for a FS+d lead lag is to allow for curing or drying out time.

The problem is that the extended link takes on the calendar of the preceding activity when curing and drying out takes place on a 24/7 calendar.

In powerProject you can set the start link to +?ed - ie calendar days but if the link ends mid holiday the critical path is removed upstream.

I used to set up a specific task between - say - laying a screed > laying stone flooring of 1 day per mm of screed - called "screed drying" and this is set to a 24/7 calendar.

This also broke the critical path.

I now set up a "buffer" using two milestones:

1. Linked to the end of screed
2. Linked to end of the lead lag link and linked to laying stone floor.

These are summarised into a bar called "drying out period" using the regular calendar and the critical path is maintained across holidays.

The PowerProject buffer task option is no use as it tends to go critical even if there is float all around it.

Best regards

Mike T.

Thank you Vladimir

That was the 1st real laugh of the day.

Being an arrogant Englishman I have no interest in other languages - thoug I am fluent in Rubbish.

Best regards

Mike Testro.

Hi All

Please tell me what is a ships corpse?

Best regards

Mike Testro

Hi Mike,
we did not discuss not only calendars but also Resource Critical Path. With limited resources traditional Critical Path method does not work.
We did not discuss simulation of working in shifts, creating and managing buffers and many other interesting topics.
So we have good themes for future discussions.
But consider using both SS and FF links with lags between two activities - it is a good replacement for FS links in certain cases. I hope that Powerproject supports multiple links between activities.
By the way Caspian project used 10 different work weeks.
Today we are involved in the development of Russian Pacific Area program. It is very close to Sakhalin Island.
Best Regards,
Vladimir

Hi All

Please calm down dears - its only a thread.

But then again I am delighted that the discussion is going so well.

I think that the contributors are begining to understand that unles there is a fully developed FS linkage showing a true critical path - with a real resource modelled input (we have not discussed this bit yet) - then there is no real control of progress in relation to events that may combine to delay progress.

Wait until we talk about calendars - that is another overlay.

I worked on a project on Sakhalin Island (calm down Vladimir - I stayed in UK) - where there was 8 different calendars for different daylight hours.

There are many other factors to consider before a project has been properly planned.

Watch this space.

Best regards

Mike Testro.

Hello Vladimir,

And how do you manage resources for such schedules ?

Best regards.

Hi Mike,
imagine the situation when for some reason the following task is executed faster than the preceding one. It will start with 500m lag (FS relationship that you suggested) but then the distance will become 400 meters, 300 meters and then the work will stop waiting until 500 m section will be done on the preceding activity. Then this race will repeat.
Besides the lengths of sections between different crossings or soil changes are different. Sometimes it is 800 meters, sometimes 1300.

Creating such schedules we use different approach.
Project Planner creates the schedules for typical schedule fragments like construction of 1000 meters of pipeline in certain type of soil. The project schedule is created by inserting these fragments using multiplication factor that reflects the actual volume (quantity) of work to be done (length in this particular case) in the real schedule at the certain places.
If you have ready to use typical fragments library you will be able to create any pipeline construction schedule easily and fast.
The same approach is used for creating any large schedules where the types of work are repeated.

I see you are holding the fort all by yourself over here.
I agree. If a project has 20000 Tasks then that’s how many Tasks have to be modeled. Whatever the software, there is fast ways to do the editing and data input, and there are also slow ways. But 20000 Tasks or so is a piece of cake and very fast to set up, if you know how to use concatenation and a few text editing tricks (batch replace etc). And if you don’t then you resort to insufficient detail glued together with weird links and negative lag.

The distance between different types of work is 500m. The sequence: Transportation, ditching, welding, Joint control, Joint insulation, joint coating control, testing, lowering, backfilling. Of course there are many other types of work like road and river crossings, Block Valves, etc.
Best Regards,
Vladimir

Hi Mike..

Am starting to like your style. ha-ha-ha..

Cheers..

Hi Nestor

Now you are just being silly.

The programme would be in much greater detail than the drawings.

Best regards

Mike Testro

Good Day Mike..

That was incredible. I can imagine the person whose doing the construction need not even to look at the drawings because all the details are already in the schedule.


Cheers..

Hi Patricia

How many activities do you have in a single project?

The answer is how many do you need.

Please follow my basic rules - I say my rules because most memebers of PP threads say I am wrong - you choose.

1. One activity represents one trade in one location with no interuption.
2. If an activity is more than 10 working days it is too long.
3. Only use Finish Start links.
4. Every activity must have at least one outgoing link - if only out to project completion.
Never Use:
1. Lead lag links
2. FF SS links.
3. Constraint flags.

This applies to whatever software you are using.

I prefer PowerProject but even in P6 or 5 the same bottom up planning rules apply.

I have just completed a programme for a 3 storey house with more than 2500 activities - it was the minimum necessary.

Actually if you have a copy paste facility in your software it can be done very quickly - 16 hrs in the last case.

Thanks for listening and best regards

Mike Testro

Hi All

Trevor is right - use of any lead lags whether positive or negative are a bad thing - they distort the critical path and the logic goes haywire when entering part progess that bridges the link.

Stop doing it now.

Break up the activities into constituant parts and use FS links for each activity.

In PowerProject this takes seconds but I understand that with all P3 users this is a labour of love that may take hours.

Best regards

Mike Testro

Negative lag is a very bad idea because the trigger for the start of the successor goes off in the future.
"Task B cannot start until 5 days before Task A has finished"
You cannot know when to start Task B because you can never know if you really are 5 days away from the finish of Task A.
You really mean that a certain part of Task A must be finished before Task B can start, so you must identify that part.
Good modeling, rather than lazy modeling, simply requires that you break Task A into two parts, A1 and A2, so that you can make a FS link from A1 to B and also from A1 to A2.