Earned value (EV) - My experience

"I suspect that you will go through several iterations before you finally settle on the ’correct’ model".



or before I get fired???

Dunno what’s happening guys, but I keep receiving the e-mail notifications of postings into this forum and yet I cannot see any of the new messages from anyone.

How can you expect to ’define’ a one-size-fits-all system? Different industries will have different sequences, different weightings.



Just decide on specific gates, reach a concensus of usual progress associated with those gates and award progress accordingly.



I suspect that you will go through several iterations before you finally settle on the ’correct’ model.



I know of no two companies that agree exactly on the gates I presented and this is within the process - refineries, petro-chemical - industry.



Might be interesting to compare the rankings by KBR, Fluor, Bechtel, Jacobs, etc.

i agree but "whatever works for anybody" is the thing that needs to be defined

The 5-10% is to account for sketches, research, etc. and the inefficiency that accompanies set-up. But feel free to use whatever works for you.



The point is to remove some of the subjectivity in reporting percent complete.



Is it a perfect system? Probalby not. Does it work? It seems to.

why only quite correct? Budget man-hours for Design Works is "quite" difficult to determine...Conceptual/Basic Design; Detailed Design; Specifications are only in the minds of the Designers and not the Planner (i supposed) and they can easily change it in a blink, so why the hell bother to plan with Design?

Anoon - you’re quite correct. How the hell anyone ever concluded that "start"ing something is worth 10% is totally beyond me.



Approx 90% of my planning is "design" planning. For EV purposes I use man-hours as the unit of mesurement. It works beautifully, is objective, is dynamic, easy to measure and more accurately extrapolates within the EV equations. The biggest difficulty is the mind-set of the people who insist on measuring the progress via the "physical" aspect of "where" in in the "process" the document resides. OK, the conventions are probably not too far out. But why not use a true measure of value eg. hours. With regard to EV, "where" in the process the document resides is of absolutely no consequence whatsoever!



James.

i think it’s very difficult to relate your mentioned standards with regards to time, "5-10% is earned when drawing is started" this may sound like a joke but what if they had started only the title block? i mean if there’s no specific definitions, it may still sound like a guess...and time is always running...

It may be different in the process industries, but we regularly measure design.



There are essentially four stages to drawing development - IFR, IRA, IFF/C, and As-bullts (which can be excluded from EVMS and engineering progress).



While the percentages will vary by company, they follow this pattern. 5-10% is earned when drawing is started. 65-75% when drawing is issued internally for review (IFR). 75-85% when drawing is issued for client review (IFA), and 90-95% when issued for fabrication or construction.



Generally 5% is withheld for as-builts. The other 5% can be reserved for late vendor data - see company policy. This means that the lead discipline engineer can claim progress up to the amount available for the release. That is up to 65% for IFR, but no more.



For MRs/EMRs, it is usually all or nothing. No credit until the material requisition is issued. Other deliverables, like equipment lists, BFDs, PFDs, H&MBs, etc. can have partial credit or not.



In my experience, we started using EV in design before we applied it to construction.

yes, i’m interested in EV with regards to Design, i’m not that experienced, but come across with such things and found out that there isn’t a standard or how can you come up with a standard?



you said there’s nothing to measure, why? is Design Deliverables immeasurable? or nobody really cares about Design/Engineering Planning (not including Procurement)!

Sounds like you boys got lost.



Firstly Earned Value is quite difficult to measure if you have nothing to measure i.e. design.



If you are monitoring design then your life just got a lot more complicated than sprouting a fanny (if you are female then my humblist appologies).



If you need to know more on EV & design then I suggest that maybe a few of us with more experience in this field get together and come up with an industry normal. I get the feeling from many jobs that if design is not controlled from the onset (which many projects are), then the costs of a project can multiply out of control. Which is great if you are the designee and not the client, for obvious reasons, but if you are the client it can be a nightmare. And this makes getting involved as early as possible for a planner crutial to the success of any given project.



It’s late & I canne go on.

Raviraj,



I hope that this would not be difficult for you to do.



Can you please rephrase or retract your example and for more future healthy conversations, please avoid such statement again? An apology in this international community that we are into with would be much appreciated.

E.g. Egyptian workers perform far better than workers from Phillipines?!!!... ha ha ha!

Maybe this would be true during the time egyptian laborers were working on the pyramids. Yep, any available database out there that we can look at and do the statistical analysis on the time and motion study on the skills (apple vs apple) between these two work gangs? I’m interested to see the exercise of this egyptian madness! Just an opinion, ok?

That’s why generalization and stereotyping will not work in a healthy conversation.

Ooooooo Anoon,



This is just an example. Basically, Egyptians are more skilled in doing the finishing works whereas, Phillipino workers are mostly used in general civil works.



What u r saying, might be true.



Keep Smiling



Cheers,



Ravi

I disagree that Egyptian workers performs much better than workers from the Philippines! it’s the other way around!

Not only in India and China is all of the work subcontracted out to suppliers who do not show any man-hours. This is the practice everywhere.

You are contracting and paying for the Task, not the hours, so you don’t care.

The number of hours that it takes is to the risk of the subcontractor.

It is vital that the Tasks be broken down, divided up (decomposed if you like) into small chunks. Then you can say that if progress of the small, short Task is anything other than finished it is counted as zero. A 30 day Task is too long. If the subcontractor wants progress payments, it must be broken down into 6 x 5 day Tasks which can be identified and measured and claimed. You could be generous and make a payment for the start of a task and another for completion.

Raviraj,



You are quite right in that, if your work is performed using sub-contractors only, you may not be given a breakdown of hours.....only a total cost. That is perfectly normal. You then use the budgeted quantities to monitor progress. You have no other option.



Have fun.



Cheers



James

James,



Very well said and written. I do agree to most of ur statements, but in my viewpoint, its very difficult to quantify anything in terms of manhours.



Skills and techniques to do the same job varies a lot within different gangs of workforce. E.g. Egyptian workers perform far better than workers from Phillipines. Here, in these type of projects, manhours will be more for philipino workers to do the same job than Egyptian workers.



Also, Management always want to keep a close look on MONEY than MANPOWER. Manpower cost will become a significant part but ultimately Management is more worried @ the final outcome.



One more thing is that, most of teh jobs in the fast-growing markets like GCC, India, China are all sub-contracted rather than comapnies having their own manpower supply.



Cheers,



Ravi

Dieter/Andrew



I appreciate your sentiments, and accept that you are not able to fully understand how the man-hours methodology functions. As stated in other postings, it requires a change in mind-set. Although the other cited examples of excavations are all perfectly good methodologies, the excavations are working with a well-defined unit of measuring the physical progress – ie. M3. However, everyone has failed to answer the question of how the cost of $1 per M3 was derived. There is only one way…and this being the hourly/daily cost of the resources (manhours/machinery). Remember, therefore, that every hour it takes to perform a function, regardless of whether it’s writing a document or excavating a building site, it is going to cost you $X per hour x quantity of hours. Therefore, in terms of the BCWP, the claimable value should be in the same proportions as the cost. If your excavations to-date have consumed 1000/3000 hours, but you have completed only 8000/30000 M3, the choice is yours as to whether you are 1000/3000 complete or 8000/30000M3 complete. If, after having re-forecast, your ETC figure is going to remain at $20000, then your overall EAC is still going to be $30000. You could argue that your % complete could be 8000/30000 (using M3 as the unit of measurement = 26.7%) or 1000/3000 (using manhours as the unit of measure = 33.3%) or 9600/30000 (using $ as the unit of measure = 32%). The interpretation is just that the remaining cost per M3 is forecast to be less than the to-date cost per M3 and also less than the initial cost per M3 (which seems unlikely). Ultimately, should you supply a complete set of Earned Value data, the % complete figure is of little relevance. The important element, from both the client’s and the contractor’s perspective, is the overall cost and end-date.



To take it one step further, and extrapolate the initial time or cost per M3 to the end of the activity, using the equation of $9600/8000 M3 you get a future cost of $1.2 per M3. Applying this 1.2 ratio to your existing % complete, as calculated using manhours, the result should have been 1000/[1000 + (2000 x 1.20)] = 29.4%. If you wanted to use $ as the unit of measure, then your % complete would be 9600/[9600+(20000 x 1.20)] = 28.6%.



Now, after all that, the major reason that we use manhours, is because our “deliverable” is usually a unit of 1. ie a single drawing or a suite of drawings (suite quantities unknown due to uniqueness of the product– therefore treated as 1) or a report or a uniquely manufactured product. To attempt to use a unit of measurement, other than manhours, becomes extremely difficult and excessively complicated – and it doesn’t improve the accuracy. Moreover, if you have compiled your estimates and costs using manhours, then it remains perfectly valid to measure the project on such. If you estimate 1000 hrs to produce a finished model and suite of drawings, then attempting to calculate the cost of each drawing serves no purpose, because a particular drawing could actually take 100 times longer than another.



I have always maintained the fact that there are numerous ways of doing EV, and most of them will provide reasonable results – assuming that integrity and consistency are applied. I find that the biggest barrier is often the PM’s “denial” of the results.



James.

Thanks Trevor for the explanation,



anoon,



Earned Value is basically Budgeted Cost of Work Performed and in this case BCWP (EV) will remain same, but AC will be doubled and so, it will show the loss value.

In above example, EV will be $8000 and AC will be $9600 X 2 = $19200 (assuming vol. got doubled)



Since, v r paying more money for disposal and hauling, so the same need to be considered in calculating the EV.



But, practically all our contract terms and conditions are back-to-back in any project. So, if we loose money from our client, then so our sub-contractor as well.



Hope it is clear.



Cheers,



Raviraj

Surely it goes without saying that if they dig it up they have to remove it?

And just because it goes from dense to loose doesn’t come into it either, does it?

The 2nd example is much better than the first because it includes the Status Date, end of Day 10.

The 1st example lacks a status date and therefore does not have enough information for even a progress report, or any EV calculation.

BCWP = $8000

ACWP = $9600

BCWS = $10000

CV = BCWP-ACWP

SV = BCWP-BCWS

Check.

Since excavation is being done at 800 m3/day instead of 1000 m3/day, and 22000 m3 remain, it is essential to re-estimate the Remaining Duration, ie 22000/800 = 27.5 days.

You can leave the RD at 20 days if you wish but it won’t happen. Observed progress always provides an opportunity for, and in any case necessitates, a new estimate for what remains to be done, how long it will take and what it will cost. Of course, if this task is on the critical path, the successors and the project finish date will now be delayed by 7.5 days. Even if not on the Critical Path, if the 7.5 days exceeds the total float then this Task will become Critical.

Ravi,



thanks for your time, I forgot to mention the hauling and disposal of excavated materials, which originally presumed to contain only the original estimated volume, and i don’t want them to leave the remaining volume on site.

Anoon,



In EV, actual work is compared with respect to time.



Say, in 30 days u r supposed to do 30000cum of excavation works with $1 per cum.



Say, after 10 days(Considering linear progress), u shud have done 10000cum of excavation with value of $10000



But, actually on site only 8000 cum of excavation work is complete giving value of say, $9600 (8000 cum of trench / pit excavated rather than the volume of loose material), then in that case, EV will be calculated in following way:



EV = 8000

AC = 9600

PV = 10000



So, CV = 8000 - 9600 = -$1600

SV = 8000 - 10000 = -$2000

CPI = 8000/9600 = 0.833

SPI = 8000/10000 = 0.800



This means we are behind schedule and over budget



This is what EV is all about



Hope Its clear!!!!



Cheers,



Ravi

may I lay down one example here: simple earthworks, say Excavation Works; you got an estimated volume XXX cubic meters (without considering density or factor of compaction), volume maybe based on basic calculation (X*X*X = V), now here comes actual works or actual volume that came out (please note that this will become loose material), the original estimated volume (XXX m3) maybe doubled by the loose material.



How do you apply earned value?

Hi James

I share Andrews opinion. 50% of manhours spend doesn’t say how much was reached at all. It is effort spent but not value earned. Which EV should be calculated if you’ll need double hours - against planned - for a deliverable? So this cannot be taken for earned value.

EV has to be related to results/deliverables.

Another important item is: The parameters used for EV must be agreed before the start of a project - bid, contract, latest kick-off-meeting - to receive a common understanding.

Regards

Dieter

James,



Was interested in the statement:



"If you have completed 50/100 hours, then you are 50% complete"



You may have used up 50% of the planned resources and possibly planned cost, but you might be only 25% complete.



Although using a % complete method is relatively easy, a much more accurate method, although somwewhat more time consuming, is "planned time remaining".

Carmen,



We use Earned Value for both small and large projects. It works equally as well for each. However, there are always "issues" regarding the methodology. This is especially relevant when calculating % complete, because calculating your progress is probably the MOST IMPORTANT calculation that you’ll make within the EV data. This is because all subsequent values (BCWP, CPI etc) derive from the initial equation of % complete.



We tend to use manhours as the progress calculator - as this retains a direct correlation between time and cost. Attaching a "% complete" to a document having been "issued" can be totally misleading and is open to manipulation. If you have completed 50/100 hours, then you are 50% complete. It makes no difference whether or not it has been issued/draft/final review. You still have "earned" 50% of its "value".



I have written a document that compares different ways of calculating progress. Each method gives different results for BCWP, CPI etc, and they were all wrong except when using manhours.



I’ve been using EV for almost ten years- and it CAN work extremely well. If used properly, it is entirely objective (apart from any re-forecasting)and it avoids subjective evaluation and interpretation. It is not difficult, it just requires a systematic approach and well-organised data. However,there are far too many people who still think in the "old way" - and they basically make life very difficult.



If you operate the full-circle, closed-loop process of EV calculations (Estimate > Actual > Remaining) the you will realise that you can never "lie", as the "lie" will be exposed in subsequent calculations.



Cheers.



James.

we use EVA for everything but we find it’s at its most useful when tracking large numbers of loosly or unconnected small jobs. Documents could be one such example. Small is the operative word that seems to be missing here in your case though, I suspect.



The problem with any sort of progress tracking system comes when the individual activities are overlong. Conventional wisdom is that an activity should last no longer than 2 reporting periods. Any activity which lasts longer should be broken down (decomposed I believe is the term du jour) and a deliverable (or at least a confirmed validation step) defined for each sub-activity.



Thus your activity is progressed by the issuance of the document itself, not by some voice reporting it’s done.

Breaking down the activities is not always easy, but talking to the people actually doing the work will usually produce some stop points that you can identify and track. Trouble is, some people will clam right up once they realise that you’re trying to decompose their job, that just needs percistence.



Issuance of incomplete documents is another example of how someone may undermine the progress reporting, but short of reviewing the things yourself I can’t see much to do about that except report it when you see it. If your own manager takes no action then they don’t deserve to know where they are.

To my believe,



it really DOESN’t matter, weather u r measuring progress for progress of documents or procurement or actual physical site progress. What matters most is how u report ur actual progress.



Baseline is always fixed and the sole purpose of using EVA is monitoring the as-on-date progress.



LIKE in ur case, progress was OVER - reported; its definitely going to give wrong results weather u r using EVA, P3, LOB or anything else.



EVA doesn’t establishes the baseline. It is used to measure progress against the set baseline. TECHNIQUES shud b used to give realistic picture. If ur input is wrong, then definitely ur output will b wrong.



:-)