Issues of time impact analysis

Hi Rafel,



sorry that i took a little long to get back.



The updating precision was a real problem for me for two main reasons



1. The horizontal nature of projects – the programmed logic was not really honored by the Contractor’s construction guys. Moreover when there was a delaying event, the construction guys shifted the workforce to a new available work front. Doing this had caused disruption to the Contractor, but it reduced the impact of delay no project finish. Because the Contractor was still able to continue with works on a different work front which was initially planned to execute at a later date.



But these changes in construction logic (which Contractor adopted at site due to the delaying event) were not reflected in their updates. Instead they maintained the original baseline logic and just updated with actual start and finish dates. So unless an activity is started (actual start) it showed expected start dates as per original logic. Since the client delayed activity was in the middle of the path, the updates always showed an inflated delay on project finish.



2. The large number of delaying events – Due to this if I had to consider correcting the program updates at the start of the events I had to practically re-do the entire updates.



I think shahul has impacted the updated program with the delaying event, in that case will you call it as ’As Planned Impacted method’. I think for ’As Planned Impacted’ method the Baseline program needs to be impacted with delaying events.

Hi Shahul



You have followed the correct procedure for what we in the UK call an Impacted as Planned analysis.



Just because there is no overall delay the loss of float may have financial effects which should be investigated.



Plant standing for longer than planned for instance.



When you are faced with a number of events they have to be impacted in strict chronological order and the key to accuracy is calculating the date of impact.



A written instruction is often issued retrospectively and the starting date for the event would have been earlier.



You must also account for the period between start of the event and the impact date which is the earliest date that you could have started the work after processing the changes.



If the event impacts on more than 1 task you have to impact the event on all of them.



Best regards



Mike Testro

Vladimir



Sure that FF helps when SS is broken. I would like to add it is not just redundancy, is more than that, it keeps logic. True ladders have equal steps (durations) and equal legs (SS and FF lags) but at times durations, SS or FF lags are or become not equal and one takes control over the other, at other times a change in remaining durations change the finish date of the successor and the FF relationship is a must to keep true logic, the FF relationship.



Shahul



I don’t think use of lags is going to invalidate your TIA, the wrong use of any functionality whatever it is might. But yes pay attention to lag, it is sometimes abused and can be used as a float suppression technique especially if your software is not good at filtering and reporting for float. Usually file comparison software like Digger for P3 is handy; unfortunately it is no longer available for sale.



When I was a Project Manager I had my character, at that time if a scheduler on my jobs would come with the idea of not using lags I would rather send him home instead of having in our team a scheduler thinking in the opposite way we at the jobsite plan our jobs. Before taking such a radical approach you should investigate how your team do their planning.



Best regards,

Rafael

Mike,

you remember that it is necessary to set both SS and FF links with lags. And FF link can help when SS is broken.

In Spider Project everything is easier - you can set volume lags like 500 meters distance between crews but this is specific to Spider Project.

So I think that both SS and FF can help with retain logic option.

Hi Vladimir



Can you answer the question that I posed to Rafeal.



What happens to the logic when progress on the bar goes over the lead lag link?



Best regards



Mike Testro.

Mike,

we returned to the same question:

Using SS and FF with Volume Lags we created Caspian pipeline construction schedule consisting of more than 8000 activities (with 1145 resources).

Using your approach we should divide most activities to the small sections and thus multiply the number of project activities.

Is it practical?

Besides it is too rough if the sections are not very small.

Rafael’s definition "continuously movable link" is true reflection of reality.

Best Regards,

Vladimir

Hi Rafael



In each location the trade cascades are each linked FS so the logic flow goes from trade to trade.



In a typical simple finishing sequence this would be something like:



Blockwork > BWIC > M&E Drops > Joinery 1st Fix > Plaster > Ceiling Grid > etc etc to Clean Clear & Lock up.



Between each location the trades are linked in sequence FS.



There is no "continuous moving link" just lots of FS links between individual trade sections.



If you want to call this a ladder then so be it.



It is very quick and simple when using copy - paste and trade filters.



Best regards



Mike Testro

Mike



You are tracking plastering vs. plastering, but what about CMU (Masonry Units) vs. plastering or any other group of activities where the continuous flow relationship is also hidden when using discrete activities.



How are you going to represent a continuously movable link? Sorry I don’t know CPM software that can do this, lag is even short, is an approximation. At times the best option is to multiply by two what you oppose, use two links with lag to create laddering. Even Microplanner many years ago created a special type of activity for this purposes, it was great.



Best regard,

Rafael

Hi Rafael



In my mind there are two ways to sequence activities.



1. You can have 1 activity spread over several locations.



> Plastering split into multiple floor levels



2. You can have several locations with all the activities in it.



> Level 4 with plastering in the cascade.



I prefer Nr 2 which suits my Basic Priciples of planning.



You can keep the logic flow of plastering through the levels by filtering and group linking FS to all plastering activities - takes 20 seconds.



Best regards



Mike Testro

Sreejish



You are right, but at the Data Date of a TIA, just prior to the impact, you just need to have the remaining work correct as TIA in a prospective mode will only consider the impact events over the remaining work.



You don’t have to be so picky about the past.



Updating precision is usually more reliable when you schedule with activity durations smaller than your update periods, this to determine with more precision the remaining duration of started but not finished. These you got to pay particular attention as they are part of the remaining work. A scheduler might be tempted to assign low remaining duration to in progress critical activities as to increase the effect of delaying events on the projected finish duration.



Best Practice guidelines are useful, regarding duration, use of lags and everything else. Of course they are not intended to be interpreted literally but to be applied wisely.



Mike,



“Progress does not necessarily equate to improvement.”- You are right just consider P3 vs. P6.



“Doing it properly the first time is not the lazy way.” – Again you are right I merely used the word lazy just for fun as in a few of our previous discussion on the lag issue you labeled as lazy those who use lag.



I believe use of lag or splitting an activity into two different activities does not eliminate the arbitrary time/duration factor but instead will increase the occurrence as at only the lag you will have it at the activities, with the effect that it will show twice, on the first and second activity.



P3 by keeping splitting the activity within the same activity record will show you the out-of-sequence event until you correct the logic in the right way. Asta by splitting the activity into two different activity records hides the out-of-sequence event which might be wrongly rescheduled as the fix is not always the correct one is a two sided knife.



When you eliminate lag through the use of discrete activities you hide true logic, the logic as a Delay Analyst you would like never happens. By making use of discrete activities you hide the linear flow that must be maintained between the activities. A dynamic relationship that is not fixed at a single point, fixing it at a single point is in error.



Vladimir,



There are linear projects where lag is the order of the day. In order to communicate, to keep them manageable and understandable you need software that provides the appropriate reports while keeping the rigor in the logic. Here I got impressed with the functionality of Spider Project.



Best regards,

Rafael

We discussed using lags previously. There are projects that could not be properly simulated without using lags.

One example - pipeline construction.

Regards,

Vladimir

Hi Rafael



Progress does not necessarily equate to improvement.



Doing it properly the first time is not the lazy way.



Hi Sreejish



You are quite right that the biggest obstacle to demonstrating cause and effect of delay is lack of records.



Best regards



Mike Testro.

When it comes to doing a TIA or even window slicing method of analysis, the toughest issue that i have always faced was the availability of a properly updated program. In most cases there were reasonably good baseline programs, but the updates were not reflecting the ’real updated intentions of the Contractor at each updated point of time’.



Unless there are updated programs which really reflect the changes in construction logics, it is very difficult to go for a TIA analysis, especially when it is tried to be used as a prospective method in the middle of the project.

Mike



The SCL should ban all software that have the functionality for lag and leads in favor of the software of the 60”s. Yes like those who would like to go back to the activity on arrow networks and the Dummy activity and prohibit PDM, with only finish to start relationships. Delay Analyst like the idea as it fit their needs, most are lazy and don’t want to get into the new functionality that makes their work harder.



On the other hand I know of no Contractor that would go back to the 60’s. Their business is making money where professionals manage their jobs and not rookies that after learning a few software commands pretend to manage million dollar jobs. They need serious but easy to use software that model their way of planning; they do not need others to tell them how to plan. Sorry ladies this is serious stuff as you said at another thread, if you don’t know how to build it you cannot plan it.



Even better to finally check mate on the schedule and completely take it away from the Contractor why not let the Owner play the Contractor and prohibit Lump Sum Contracts, only Cost Plus allowed and everyone will be happy. Prequalify your Contractors and at Bid opening just throw the dices and select the Contractor. Forget about the Contractor using the tool to model the schedule the way he thinks of it, let the Owner think about it 100%.



Mike, I love Planning Planet because here we are able to openly debate the issues and hope ORACLE will never take over and not let us debate as we have being doing it for so long.



You know I don’t mean every word it is just a debate, like politicians do. Throw me your best punch.



Best regards,

Rafael

Hi Mervyline



You wrote



For our TIA, I used the Clause 14 and add some fragnets activities then make a progress update..recording of 2 parties delays..(Clause 14 VS Progress Update)



What else should I show to our schedule?



Can you give me a step by step procedures?



No I can’t give you a step by step procedures without telling you everything that I have learnt over the last 12 years in delay analysis.



Best regards



Mike Testro





Thank you,

Hi Rafael



In my terms of reference any lead lag is execessive and should be avoided.



I come across the problem every time I reveiw a contractors flawed programme for delay analysis.



Do it right or get it wrong.



Best regards



Mike Testro.

Mike



Regarding Lag, if this is what you are referring to, I found SCL to be deficient as it leaves too much for interpretation when it says that excessive leads and lags should be avoided. Maybe 10,000 days is what they mean as excessive!



Our specifications and protocols do not prohibit the use of lead and lags as well but in addition provide more specific guidance with reference to the scale of the schedule as to what is excessive. It is common in our practice to consider lags and leads over 10 days as “suspicious” when the schedule scale is set for 20 days max activity durations.



Best regards,

Rafael

Hello Mike,



Pls tell me if its correct...



For our TIA, I used the Clause 14 and add some fragnets activities then make a progress update..recording of 2 parties delays..(Clause 14 VS Progress Update)



What else should I show to our schedule?



Can you give me a step by step procedures?



Sooooo confused....





Thank you,

Hi Rafael



Me too.



Now look at what the Protocol considers to be a "proper programme" particularly para 2.2.8.



Best regards



Mike Testro

Thanks Mike



I downloaded the Protocol from the SCL site, and found the following under section 3.2.11 :



The methodology described in this section is known as ‘time impact analysis’. The Protocol recommends that this methodology be used wherever the circumstances permit, both for prospective and (where the necessary information is available) retrospective delay analysis. The methodology will not be capable of being used contemporaneously unless a proper programme has been prepared, accepted and updated as recommended in Guidance Section 2 above.



Agree 100+%.



Best regards,

Rafael

Hi Rafael



SCL is the Society of Construction Law based in the UK but with branches in many countries.



They published the Protocol in 2002 and it is now out of date.



I am a member and I attend some very useful seminars to keep up to date.



Best regards



Mike Testro

Shahul



Time impact analyses have many definitions and seems to me the USA definition is quite different to the UK standard. In the US TIA is generally intended to be used before the work is completed although it can be used after. On the other hand the Windows method is to include all impact events for a short span period, usually a month, the commonly required updating period in our contracts.



The following links can provide you with some of the US definitions.



TIME IMPACT ANALYSIS – AS APPLIED IN CONSTRUCTION



From the above link:



The TIA procedure is performed while a project is on-going, and thus has a ‘forward-looking’ or a “prospective analysis” perspective in near-real time. Retrospective (hindsight) forensic research and analysis is not desired or required as a TIA is a forecast designed to facilitate a timely contract adjustment prior to the actual work being completely preformed.



Beware of the dark arts



From the above link:



I’ve devoted a greater part of this paper to the Time Impact Analysis method than the others because it is the method recommended by the SCL Protocol4, and it is presently enjoying extensive use in the various dispute resolution forums. Like the Impacted As-Planned method (discussed above) this is also both a ‘prospective’ and ‘dynamic’ method.



Is SCL in the UK?



Techniques and Methods for Assessing Delays



Best regards,

Rafael

Hi Shahul



Time Impact analysis is a very complex precedure that is used for forensic delay analysis after the work is completed.



It works best if there are good quality As Built records.



Every event has to be impacted into the programme in strict chronological order and then the impacted result has to adjusted to re allign with the As Built data.



If the work is in progress then an Impacted As Planned method will show the likely delay effect on the completion dates.



Again every event has to be impacted in strict chronological order and the cause and effect of each one recorded.



All types of events have to be impacted:



1. Employer risk

2. Contractor culpable

3. Neutral causation



So that concurrency can be applied.



Best regards



Mike Testro.