The pitfalls in cost estimating that cause project overruns have been well known for decades, yet they continue to plague the CPI
When I first started writing this article, I set out to summarize the mechanics of project cost estimation. That was to be followed by a discussion of various management tools, for example Project Definition Rating Index (PDRI) , that are used to validate the input to the estimate. But it was apparent that there were other factors at play, given the anecdotal and very public records of numerous project delays and cost overruns, that went beyond simply not following a procedure and going through some management check-list gate. As one of the owners of my company pointed out in a recent meeting, to Manage With Certainty, one must focus on the input and not just the outcome.
About four decades ago, in my first job with Union Carbide Corp. (UCC), I was given an assignment to find out why many of the plant projects were plagued with delays and cost overrun. In the late 1970s, the Bound Brook plant was suffering from a very tight capital budget constraint. The era was in the mid stage of factory relocation away from the Northeast and mid-Atlantic regions. The conclusions that I drew, after interviewing many of my coworkers at the site, were rather simplistic and seemed to be just common sense. Being a newly minted engineer at the time, I thought it was because I did not fully appreciate the complexity of project execution; and I suppose, so did the senior managers.
Over the years, the problem of project cost and schedule overrun persisted in the chemical process industries (CPI) and the energy sector. It was apparently prevalent enough that many companies were willing to expend scarce resources to hire consultants to work on this riddle. A lot of brain power was being brought to bear on the problem, but it seemed to be more stubborn than that grass stain in the laundry commercial.
Fast-forward to 2018, in a recent news article, a large public utility company announced another 15% increase in a project’s cost, making the current estimate double from the project announcement just four years ago . What happened? Was there some kind of evil mastermind at work? Was it the dreaded tariff? How can a company that is more than fifty years old, working with well qualified contractors, make these mistakes? As most of this cost overrun will likely be borne by the public utility rate payers, the truth of the matter may eventually come to light after some years of investigation. Even then, the fog of project execution will make it difficult, if not impossible, to separate facts from fabrications.
In this article, I assume that the reader can access multiple well-written articles and websites on the Internet  that discuss various methods to generate a cost estimate. Instead of delving deeply into the mechanics, this article discusses some of the pitfalls that one may encounter while following those procedures (Figure 1). Hopefully, the reader will find the simple solutions to be easy to apply to everyday project execution. If nothing else, the article may help one sleep better, and it also helps explain how a kitchen remodeling can balloon from $10,000 to $20,000.
Class of estimate
The “class of estimate” is a generally accepted term to signify the claimed range of accuracy of cost estimates from one to five , the most accurate being one and the least being five. During the development of a project, the estimate accuracy improves as more information is generated to put into the estimating effort. Presumably, the project team is busily checking off those boxes specifying the required inputs. So, if all the boxes are checked, then the estimate must be accurate as defined. There is an often repeated phrase in the information technology (IT) world: “Garbage in equals garbage out.”
Beware, just because there is a document transmittal fitting the description, it does not necessarily mean that there is a real contribution to improving the accuracy of the estimate. In fact, it could be quite the opposite. Bad information is invariably worse than no information at all. It is incumbent upon the management team to make sure that the inputs are properly vetted.
The engineering disciplines of the project team must be encouraged to question and evaluate every assumption and calculation. Project management must be encouraged to foster an environment of honesty and openness so that team members can talk about any difficulties encountered. It is also necessary to be realistic and admit, if substantial input is lacking, that the desired objective of estimate accuracy is not feasible.
Wishing does’t make it come true
Human nature being what it is, I admit that I still tend to wish my project estimates to be lower and the project schedules shorter. Warning: self-delusion is a very bad way to run a project. It is beyond bad if one is successful in coercing a subordinate or a contractor to abet in that lie.
A friend told me about a meeting with a potential client for a very large project who was adamant about competitively bidding every phase of the work, from front-end loading (FEL) through construction and startup. The logic was that he would easily find the contractor most desperate for his business and therefore give him the lowest price with little or no contingencies and minimal profit margin. He was clearly choosing to ignore the lessons of the time value of money , as the schedule would have stretched out more than a year to allow for the bidding process. There was a definite lack of understanding of human nature. What would anyone expect from desperate businessmen trying make their commission or to save their company? If he was not secretly wishing his project to be plagued by endless change orders, he must have optimistically assumed that the contractor estimated correctly or equally unlikely, intended to make up for any overrun out of his own pocket.
The better approach for this client’s problem would be to do his homework. First, understand what costs his business plan can support. Second, evaluate the quality of inputs and assumptions that are going into the estimate. Finally, third, delete or revise the scope so that the project cost fits within the economic model. Along the way, he would have found the engineers that he could trust and who would be on board with his objectives.
Along that same line of thought, make a sound estimate and schedule a reality by clearly defining the scope of the project. This scope then drives the discipline details, which are verified with the cost estimator. Note: the ends are not achieved by reducing quality of the inputs and hoping that the estimator can come up with assumptions.
Unfortunately, when the project is driven into a competitive, fixed-cost bidding process, it is not likely that the contractors would reveal the salient information to a project manager. Without it, the project manager is left to scan through the qualifications and exclusions like a first-year law student, perhaps not wanting to believe that the submitted bid is just a book of fiction.
Understand your experience
“I have eaten more salt than the grains of rice that you have eaten,” goes a Chinese saying emphasizing the importance of life experience. However, another common saying goes, “If I had a dollar every time a project manager demands to know why an estimate did not match the expectations based on a previous project, I would be a rich man.” While a project manager’s experience is a prerequisite for the job assignment, he or she must be very careful to balance life experience with the facts on the ground. Building a project in Asia is not the same as building it in Europe, and is not the same on the U.S. Gulf Coast.
Amid the U.S. process industries depression in the mid-1980s, I had occasion to purchase a small pressure vessel that was to be made from Monel. The low bidder was able to deliver the vessel, upgraded to a high nickel alloy, in eight weeks. Due to the unforeseen circumstance, the shop was in possession of the needed alloy materials, which were promptly put to my use. If I were to continue to use that project as my reference, then all the subsequent estimates would have been unreasonably high with long extended schedules.
The combination of wishful thinking and selective memory make a potent brew. The desire to have a project at a certain cost can lead a project manager to question whether the estimator is correct. Instead of approaching an estimate by comparing it to the last project and asking why it can’t be cheaper, the first questions that should be asked are as follows:
- What are the inputs and the assumptions that generated these results?
- What are the relative impacts of those assumptions on the bottom line?
- What is the quality of the inputs?
Then, a detailed comparison to the previous project can be made only after ascertaining that the current facts are comparable to the relevant experience. Then, too, the manager’s experience should give guidance on what inputs or assumptions are relevant to the problem at hand. Remembering that every project is unique in its place and time allows the project manager to filter out the useful knowledge that can be applied to subsequent work. Otherwise, one could be guilty of comparing apples to oranges.
Don’t change your mind…
… unless you have to. We are bombarded daily from social media that this is a fast-changing world. Constant change is the mantra. As a project develops from the conceptual stage to increasingly detailed definition, the cost estimate and schedule are also progressing in the cone of uncertainty (Figure 2) . Assuming the inputs are valid (and this better be true), the probable range of error becomes smaller and smaller with increasing definition and the freedom to make alterations to the overall project scope approaches the vanishing point. Even if one were to be in possession of a perfectly accurate estimate and schedule, changes in personnel, site conditions, or state of the economy, may have a negative impact.
The objective of a project manager is to minimize the changes to the project unless there is an overwhelming reason to do so. The rule should be: “ No changes unless it is unsafe or won’t work.” But when conditions create that reason, the project manager must act aggressively to put everyone on the same page. Then, when all stakeholders agree , and the negative impacts of changes are understood, the scope change should proceed as quickly as practical. This implies, then, that certain inputs and assumptions that went into producing the earlier cost estimates must be revisited to confirm their validity.
So once again, we emphasize the importance of validating the inputs that go into the estimate. It is also important to periodically revalidate the critical inputs, whether technical, commercial or legal, that are the foundations for the justifications for the project in the first place. With the advent of many upstream and midstream projects in North America, the commercial inputs can quickly change the premise of a project. But at some point, in the execution, an executive management decision will be made to allow a project to go to completion while initiating a follow-up phase.
Know the project environment
Human relationship defies mathematical modeling. Thus, I do not claim to give adequate advice in the area. Nevertheless, in project management, it is necessary to be in alignment with the stakeholders. It should be noted that unless one is aware of the total history of the proposed project, and maybe not even then, care should be taken to avoid challenging the previous conclusions. Unfortunately, for those working on the later part of project development, it means that one may have to settle for less than optimal solutions; or some better idea may never see the light of day. The project manager is still the final arbiter to decide whether certain improvements are needed as the critical factors to maintaining cost and schedule. But these should be discussed with the project team and decided upon one at a time.
I put forward the argument that discretion should always be the governing rule unless the grounds are very familiar. I submit the analogy of a visitor coming into a home and insulting the homeowners. The visitor will most likely not receive an invitation to stay for dinner. On the other hand, it is always necessary to point out that the kitchen is on fire, as that will have greater impact on the overall scheme of things.
Another aspect of the environment is the general uncertainty that is introduced from the imposition of tariffs. In this global economy, there is no doubt that cost increases on material can be expected. The question remains to be seen how much it will be. A properly estimated project will have identified the various inputs based on clearly delineated material costs. The project can account for the uncertainty with a modified calculation of the usual escalation factors for each cost item. This number would be summarized below the cost total. Business management would have to decide what level of risk to accept in negotiations, either internal or external, for the project to move ahead.
The estimate got a great PDRI
Business schools teach the importance of measuring performance. As mentioned before, one of the accepted ways of measuring the accuracy of an estimate is by means of the Project Definition Rating Index (PDRI). The methodology is a comprehensive checklist of estimate inputs to which the project team give a completion score: five being not defined, to one being completely defined.
This can be a very effective tool if one is aware of two hidden flaws, both of which can give a deceptively low score. First, while the checklist identifies all of the activities (inputs) necessary to support an estimate, there is no practical way to know if every input is correct. And because the method does not give an absolutely true relative weight to each of the inputs, it is difficult for an outside manager to know which part to focus on. It is up to each discipline lead and the project manager to know whether the input was worth the paper it was printed on.
Second, the PDRI evaluation is done at the end of an estimate effort. It therefore becomes a de facto report card on the project team and specifically the project manager. So, it comes back to that business of human nature and relations. Not being an expert in such matters, I venture a guess that the average project manager would be loath to admit to management or the client (or both), that the time and money spent produced a less-than-perfect product. And at the same time, there will only be scant objections from the project team that does not want to confront their manager.
While discretion is advised on advertising any deficiencies in the input to the estimate, the project manager can utilize the PDRI table to validate the accuracy of these inputs. This evaluation of course, is required while the work is being done and not at the end of the estimate.
Resource allocated for estimate
In an article about the high failure rate of mega projects in the upstream oil-and-gas sector, Edward W. Merrow identified one of the main culprits as being inadequate front-end loading (FEL) being performed [ 8]. I submit that this problem can be just as detrimental to any size project. The lack of an adequate FEL, either because of budget constraints or lack of time, equals poor input to the cost estimate effort. While managers have attempted to plug that gap with “experience”, this replacement tool has a very dull blade and should be used with care.
If the FEL effort is bid out as a fixed-cost project, there should be a very detailed listing of the final deliverables, and a plan on how to reach them in the time allocated. The project manager should be careful that the team is not under duress to declare completion without finishing the work, either because of time or budget constraints. If that is the case, then the inputs should be carefully vetted and the claimed accuracy of the estimate appropriately downgraded.
There should be an open communication channel between the discipline leads and the estimating group so that inputs are submitted on a timely basis and the information is useful toward the production of an accurate estimate.
The more things change, the more they stay the same. It seems that while technology has revolutionized the way we live, the nature of project management remains the same. The four main reasons for cost overrun given by my colleagues 40 years ago at UCC were:
- There was not enough effort put into the development of the job scope. Frequently, preliminary process studies turned into full-fledged projects without further development. Subsequently, many project experienced major scope changes.
- The first point was a manifestation of capital shortage. Management increased effort to conserve cash for project execution. The obvious place to cut was “unproductive” project scoping efforts. This was deemed possible as most of the staff had many years of experience at this plant site. It was more important to get more project authorization applications in the pipeline.
- Subordinates were reluctant to question project sponsor decisions, even though they know it to be incorrect.
- In another manifestation of capital shortage and wishful thinking, cost estimates were arbitrarily reduced because “It just can’t cost that much.”
1. Project Definition Rating Index Overview, Construction Industry Institute (CII), Austin, Tex., September 1, 2018, www.construction-institute.org/resources/knowledgebase/pdri-overview
2. Hyman, L. and Tilles, W., Southern Company Just Raised Cost Estimates for this Megaproject Again, OilPrice.com, August 11, 2018., https://oilprice.com/energy/energy-general/southern-company-just-raised cost-estimates-for-this-megaproject-again.html
3. American Society of Estimators, The Ultimate Guide to Project Cost Estimating, www.aspenational.org/page/sep, October 26, 2018, www.smartsheet.com/ultimate-guide-project-cost-estimating.
4. ASTM E2516-11, Standard Classification for Cost Estimate Classification System, 2011, ASTM International, West Conshocken, Pa., www.astm.org/standards/E2516.htm
5. Chiu, A., The Time Value of Money, Chem. Eng., December 2017, pp. 47–49.
6. “Cost Estimating and Assessment Guide”, p. 38, Government Accountability Office, Washington, D.C., GAO-09-3SP. www.gao.gov/new.items/d093sp.pdf.
7. Chiu, A., Ten Tips for Smart Project Managers, Chem. Eng., January 2012, pp. 40–43.
8. Merrow, E.W., Oil and Gas Industries Megaprojects: Our Recent Records, Oil and Gas Facilities, April 2012, pp. 38–42.
Alfred Chiu is a project manager with S & B Engineers and Constructors Ltd. (7825 Park Place Blvd. Houston, TX 77087; Phone: 713-845-4156; Email: email@example.com). Chiu is a registered professional engineer in the state of Texas. He received his B.E.Ch.E. degree from the City College of New York, and an executive M.B.A. degree from the University of Houston. Chiu has 40 years of experience in petroleum refining, chemicals and water processing. Before joining S & B, he worked for Union Carbide Corp., Lummus Co. and Stone and Webster Engineering Corp.
The Process Synthesis Model presented here is a methodology that integrates technology and economics, thereby improving process design and reducing…
Deciding whether to put a project on the fast track entails risks. Some observations are presented here to help minimize…
The differences between industrial-scale facilities and pilot plants go beyond size, and these intricacies must be understood when estimating the…
Before any engineering or design work can take place, a project’s feasibility must be considered. Reasonably accurate costs can be…
By intensifying cooperation between process designers and cost engineers in the conceptual stage of plant projects, better-informed decisions can be…
5 ways to Optimize Production of Polymers and Intermediate Petrochemicals
7 Ways to Achieve Process Safety in Chemical Production
Five Reasons Why Chemical Companies Are Switching to Tunable Diode Laser Analyzer Technology
Simplify sensor handling and maintenance with ISM