(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.
I need to update a schedule for a 6 MM$ project (18 months duration from engineering, procurement to construction), The project team is relatively small (1 lead engineer, 1 project engineer, 1 CAD operator, 1 purchaser/buyer). We issued the baseline last week. I am planning to update the engineering and the procurement portion of the schedule on a monthly basis and the construction phase (which hasn’t started yet) on a weekly basis. However, the project engineer wants to update the engineering portion of the schedule on a weekly basis. There was no planner/scheduler before I arrived to this group and the schedules were developed by a planner who is in the head office and sent the baselines and updates by e-mail.
I would like to ask the experienced guys what would you suggest as to how frequent should I update the different phases of the EPC schedule. Also, I would like to know –and this is more important- HOW do you this? I mean, I am planning the following:
The information that I will collect will be: Actual Start, Actual, Finish for activities that have started and/or finished during this time period or New Forecast Date for activities that have not been completed yet. Also, I will get % Complete.
Please let me know if I am doing it the right way. I appreciate very much your comments and/or suggestions. Thank you very much in advance.
Printer-friendly version
Send to friend
Raphael, such good questions! You're obviously interested in this approach to project management, so I urge you to get and read my book. I'll try to answer your specific questions here, but there's a lot more in the book.
"Although you can compress the schedule in many ways, I agree that DRAG and DRAG Cost will help you look for the road of least resistance in terms of costs as long as you do not lose perspective of other options. It is an important tool of your toolbox, at times handy."
All methods of compressing the schedule consist of shortening the critical path. Anything that delays the CP has drag (work, dependencies, lags, NET or ON constraints, slow handoffs, resource unavailability or insufficiency, quality problems, customer indecision). The packages like Spider and Sumatra that compute drag associate it with activities, but the things make an activity take longer could be any of the above, as well as work. In fact, knowing what causes the drag of an activity is crucial info (and should be a major focus of the ABCP for both lessons learned and forensic investigation!). It's a lot easier to do something about some causes of drag (e.g., resource insufficiency) than others (having to let the concrete set before working on it).
"Perhaps not of much interest to the Owner but to the Contractor is key, this business is to make money, to loose I rather stay at home."
And there you have it! All projects are investments, even though the PMBOK Guide doesn't ever use the word! And I'd suggest that (although different types of contracts can sometimes distort the true situation), it's also of great interest to the owner! A mall that opens after the Holiday Season, or a hotel that misses half the tourist season will cost the owner hugely! All projects are investments, and the purpose of any investment is to extract as much value from the work over and above invested cost as possible. The greater the difference (i.e., project profit), the better any investment is.
"I perceive Drag Cost as the slope of the cost curve, you got to select the activity(es) with less drag cost per unit time and start from there in this sequence. You mentioned bigger drag and drag cost, I guess you would be looking for lower drag cost (lower acceleration cost) while drag will tell you when you have to look for other activities or group of activities, am I wrong? What is your definition of DRAG cost."
Drag cost is simply the cost of time on a project, being caused by the above causes of delay on the critical path. As Benjamin Franklin wrote in 1747, "Time is a whole bunch of benjamins!" In almost every case, the scope of a project is more valuable if it is finished earlier. Yes, there are exceptions (such as launching a satellite to examine a passing comet) -- but most are a function of distorting contracts, where the "deadline" is set and there seems no advantage to finishing earlier. But if the contractor bid an earlier delivery date, he'd get the contract a much higher percentage of the time, and often be able to charge more, too! The customer wants the product earlier 97,283% of the time!
So drag cost is the sum of:
The "True Cost" of any project work is the sum of the resource costs to do it, plus the drag cost (which is zero if the taks is not on the CP and therefore has no drag cost). Any decision that decreases the drag cost of a task while increasing the resource cost by a lesser amount is, all else being equal, a good one. And when using drag to optimize a schedule, one will almost invariably find places where the drag cost can be reduced by ten or more times the additional resource cost.
"Because at times compressing the schedule means compressing several activities that run in parallel maybe drag cost shall include and identify the group, plase any comment will be welcomed. Maybe I am missing something and you can help me here.
Maybe there is room for a definition of Individual Activity DRAG and Group DRAG? Not sure but interested on the issue."
You are absolutely correct! You see the implications fully. I have long dreamed of a software package that would allow the user to highlight several tasks at once and get the drag for the bunch of them together! Alas, it doesn't yet exist. (Hey, I've been trying for over 15 years just to get people to understand what drag is and why it's important!)
"I also understand that under resource leveling the concept of drag is slippery. Even if this is the case I still beleive it is useful."
And you'd be right! On the resource-leveled CP, the drag is the sum of both the CP drag (caused by work duration or logic) plus the delay due to resource insufficiency. And the drag cost of this is crucial in justifying expenditures for additional resources that reduce drag! (This leads to what, in my book, is called the Cost of Leveling with Unresolved Bottlenecks, which of course is the CLUB you use to get the resources you need.)
Fraternally in project management,
Steve the Bajan
Stephen,
Although you can compress the schedule in many ways, I agree that DRAG and DRAG Cost will help you look for the road of least resistance in terms of costs as long as you do not lose perspective of other options. It is an important tool of your toolbox, at times handy.
Perhaps not of much interest to the Owner but to the Contractor is key, this business is to make money, to loose I rather stay at home.
I perceive Drag Cost as the slope of the cost curve, you got to select the activity(es) with less drag cost per unit time and start from there in this sequence. You mentioned bigger drag and drag cost, I guess you would be looking for lower drag cost (lower acceleration cost) while drag will tell you when you have to look for other activities or group of activities, am I wrong? What is your definition of DRAG cost.
Because at times compressing the schedule means compressing several activities that run in parallel maybe drag cost shall include and identify the group, plase any comment will be welcomed. Maybe I am missing something and you can help me here.
Maybe there is room for a definition of Individual Activity DRAG and Group DRAG? Not sure but interested on the issue.
I also understand that under resource leveling the concept of drag is slippery. Even if this is the case I still beleive it is useful.
Best Regards,
Rafael
Hi, Henk and Rafael.
Thank you so much for the concept of the lookahead report -- I think it's a great idea, and I will hereafter mention it in my PM classes and insist on it when I'm consulting on a project. (In the past, I have included it as part of the progress (update) meeting, but I like the idea of separating it out.)
Two more items that I think should be included in the lookahead:
The second is something that the subcontractors will not be aware of, and informing them that finishing a specific upcoming task in five days instead of ten days would accelerate the schedule by, say, four days (thereby either triggering an early delivery incentive or increasing schedule reserve) can have great benefit to all.
(Of course, if you don't have either Spider or the MS Project add-on that computes drag, you'll have to calculate the drag yourself manually.)
Fraternally in project management,
Steve the Bajan
Shareef,
That P6 have a separate field for each is good, very good, but because I do not use P6 I do not know how they interact within P6 and what setting options there are. Maybe they are completely independent or maybe you can set one to be linked to the other. Obviously, P3 was getting obsolete and provided a single definition for % complete, P6 developers noticed the need for this.
Because in P6 remaining duration is directly related to the duration % complete there is no need to keep an option to unlink duration % complete and remaining duration, in any case an option to keep the different % complete unlinked, still updating using duration percent complete is confusing, recommend you use remaining duration.
The formulas that relate duration % complete and remaining duration shall be the same just that you have separate field for cost and units, not sure about the difference between Physical % Complete and Units % Complete but your software literature shall be enough to understand the difference.
Some people when updating want to enter a single % as for it to equal all types, some software allows the user to make such link, but such link is a dangerous assumption, rarely true. Avoid linking the different types of % complete.
About the duration types in P6 I was always confused, when it was P3e (enterprise) at that time I had a preference of Old P3 over P3e (similar names but different software). These definitions are tied on to how the software performs resource leveling and allocation. It also affects how resource availability can drive the duration of an activity.
Within Spider Project we use the concept of productivity type activities, the duration of these activities is determined on how resource add to the production, it is determined to be when resource assignment can complete required production, we call volume of work. Of course, depending on your resource assignments some resources will be required to work together while others can be set to work on independent teams. All software has their own methodology but this is the core that will make a huge difference between them. It affects how the different functionalities interact with each other and affect how the resource leveling works. This concept is so simple and correct I no longer want to go back to the other approaches.
Best Regards,
Rafael
In P6 there are 3 types of ' Percentage complete ' types:
1. Physical
2.Units
3.Duration.
Then there are 4 Duration types:
1. Fixed Units
2. Fixed Units/ time
3. Fixed Duration & Units
4. Fixed Duration & Units/ time.
Which means there would be 12 result for a given updation based on your settings.....
I am actually working out the results of various combinations and have yet not understood (finalised) which settings would best provide the result I require ;as such came up with the question.....
I havent used Spider... But would certainly like to try it out some day....
Best Regards
Shareef A Azeez
Raul,
Be careful though before you start updating your programme, make sure that you assign a baseline first before you update the programme. Since your project is realtively small, You may need to update it on a weekly basis. Creating a turn around report based on your programme is a good way of doing it, but at the same time you need to visit the site to verify the progress being reported is true.
Make sure you have the following information:
Actual Start and Actual Finish Dates for activities completed
Actual start, % complete and Remaining Duration for activities in progress
After updating your programme you can set-up the baseline and make a comparison to measure performance
Regards,
Shareef,
The following are the basic formulas to estimate remaining duration when you set the software to ling % complete and remaining duration.
(1) % completion =[Actual duration/total duration]*100
(2) Total duration = actual duration + remaining duration
An activity with original total duration of 10 days if statused to be 50% complete after 1 week the software will estimate remaining duration of 5 days.
0.50 = 5 days/total duration
Total duration = 5 days/ 0.50 = 10 days
10 days = 5 days + remaining duration
Remaining duration = 5 days
If statussing as planned there will be no problem, but what if it did not happened as planned, as usual. Say the activity had an original total duration of 10 days statused as before but 5 days later and no progress reported on the activity after last update. You will status again 50% complete but the software will estimate remaining duration to be 10 days instead of five. Why? It shall be the same remaining duration of five days ago.
0.50 = 10 days / total duration
Total duration = 20 days
20 days = 10 days + remaining duration
Remaining duration = 10 days.
What if you know that instead of in 5 days it will finish in 7 days, what percentage you are going to use so the software correctly computes the remaining duration?
% completion = 10 days / total duration
Total duration = actual + remaining = 10 + 7 = 17
% completion = 10/17 = 0.58823 = 58.23%
Is easier to tell the software remaining days than doing such math for every % complete to calculate correctly the remaining duration. Just for fun, let us check the math, something you shall always do if using % complete.
0.5823 = 10 days / total duration
Total duration = 10 days / 0.5823 = 17 days
Remaining duration = total – actual = 17 – 10 = 7 ok
Now apply this to costs and consider that cost progress is not always directly proportional to time, you will get a mess. Cost loaded schedules are particularly susceptible to this scheduling pitfall, because the percent complete is tied to the earned value and cost to complete for each activity. Thus the percent complete must be accurately reported otherwise the cost reports will be inaccurate. In the above example in order to estimate correctly remaining duration you had to change % complete but no progress was done, your cost estimates if using this percentage will be wrong.
Therefore, do not use percent compete to estimate remaining durations and do not link duration % complete to costs. Depending on your software, you will be able to unlink these, if you are using P6 ask for the answer at a P6 forum and in no time they will tell you exactly where to go same if using MS Project. Please if using Spider Project do not post the question at Spider Forum as Spider Project do not recognizes such erroneous tie, remaining duration is simply equal to remaining duration, very simple. Makes no sense giving you an option to use a wrong and controversial formula.
Best regards,
Rafael
Well if using Spider Project how would you come with this question, as it is impossible to happen within Spider?
I would like to know, Rafael, from the procedure of progress updation you have stated, I could make out that you do not use %complete for updation rather use remaining duration to calculate % complete.
I totally agree with this procedure but have a doubt that in this case the actual resource usage gets updated automatically based on this %complete which might not be the "Real" actual usage (consder a case of low productivity) according to time sheets and material usage data.
How can we tackle this?
I do understand that you have already mentioned there would be only a certain degree of accuracy to do costing using scheduling softwares...
Best Regards
Shareef A Azeez
Henk,
You are correct sending the whole schedule when only a fraction is to be updated is the way to go, the lookahead is efficient, if I am updating on a monthly basis then I send a 2 months lookahead and only about half will be updated, the extra month is to make sure no activity tha was started or finished ahead of schedule is missed. Yours is good advice.
Your involvement in looking for a way to keep schedule is also as important as the updating, perhaps more, and yes ASAP as you suggest.
Best regards,
Rafael
Hello Rafael,
I usualy send out a looahead.
I collect the following data: % complete , remaining duration. Most of the time the planning is not used for costcontrol but to draw a good s-curve i asume that % complete x budget equals actual units and budget ad completion equals equals original budget.
After reschedule i compare the new schedule with the baseline. In case of delay in activities wath causes delay of the project i contact the responcible manager of the discipline involverd. When its possible i adjust the schedule.
Afther that i make my repor wat usualy includes
Lookahead
Level 1 planning (1 sheet) including basline, S-curve scheduled, actual, remaining.
Progress report per discipline (scheduled % complete <> Actual percentage complete)
Report of slipping activities
Findings in written
Maybe you can do something with this informatie
Regards
Henk
Raul,
Common to all scheduling software is that they update the schedule using a cutoff date known as data date. This is the very first thing you must set at the start of every update run, therefore, you will be required to update all phases for every update run to be correct.
One limited option is to set a constraint for expected finish date for long lead deliveries activities as soon as the expected date is available; this will calculate remaining duration for intermediate updates. This was an option available in SureTrak and do not know if available in your software. For the remaining activities or events, such as the start of an activity I see no reliable shortcut.
For basic updating, I collect the actual start and actual finish for the activities that occurred during the update period. This shall be done in regular intervals within the updating period if the job is large. If the PM have weekly meetings with the owner and subcontractors including the schedule progress as a minor topic can be useful and even can provide proof of your updating for validation of future claims. Then, just at the end of the updating period, I ask field personnel for the expected remaining duration in workdays for the activities in progress, ie activities started on any prior period but not yet finished, the remaining duration as estimated starting from data date. Use of percentage complete to estimate remaining duration is a no-no, be skeptical about this method, you will find in some PP forums why it is not advisable.
If you are updating cost and resource usage, these shall be updated simultaneously in most software, particularly within those that do not allow you to modify or correct data after posting period performance. In Spider Project, I do not have this issue but it is still bad practice if not done soon as not doing so and can be dangerously misleading, and always on the low side. Although we do not track costs using our scheduling software, as we use our accounting software, the modeling of projected or estimated costs is key for financial decisions. It is just that the accounting and job costing precision is not needed here, nevertheless even if your company do not track cost within the scheduling software I strongly recommend you consider modeling of costs in a simplified manner using budget amounts only. Remember that accounting precision is always late, is always lagging your schedule, your model using estimated values might be closer to reality until it catches up your accounting, most probably after the job is finished.
Schedule updating if done on regular basis shall be no problem, the interesting part is changing the schedule logic as needed in order for the model reflect the dynamics of the job, to reflect change in plans, to recover schedule when your job start lagging behind.
Best Regards,
Rafael
Email: PPadmin@planningplanet.com
Replies