Only 120 hours were assigned to investigate problems with a batch fluid-bed dryer. The scope included troubleshooting and preparation of a project scope document complete with process and instrumentation drawings (P&IDs), justifications, vendor estimates and a report. The client didn’t have a pipe standard let alone AutoCAD files. The new sales engineer thought he landed a whopper — I thought he should have thrown it back! As it was, I spent at least 40 hours off the books wrapping up the report and finalizing the scope, schedule and budget. There’s got to be a better way to prepare a solid estimate — right?
As with all budgets for engineering or construction, the soundest approach is to understand what the task is, how many people are required, and what are the goals and likely obstacles. Obviously, it’s best if the person who will do the work makes the estimate.
For construction projects the best reference is “R.S. Means Facilities Construction Cost Data 2009.” This almost-1,500-page paperback breaks down labor into manageable hours by trade. Once you know the billing rate and skill level for each trade, it’s easy to estimate labor costs.
There’s no such tool published for engineering labor, although some contract firms have assembled databases. R.S. Means provides some information but it’s not useful for process engineering. A few engineers have estimated projects based on the number of drawings but this takes experience with a particular job or type of work. The best solution I’ve found — and one that other engineers also have used — is to collect binders of old estimates.
Developing sensible estimates of engineering hours requires familiarity with the work and with the customer. Customers could be internal, e.g., your boss, or external, a client. With clients that want flexibility to make changes, agree on a time and materials (T&M) contract for at least the beginning of the work. It’s critical to understand the nature of the work before deciding on the type of contract.
The most common problem in doing an estimate is lack of a clear scope. “Scope creep” often stems from having a scope that doesn’t meet customer needs. Even a clear scope won’t forestall creep unless the customer understands project limits and will abide by an agreement. A team at Eli Lilly has had success in employing a “Just Say No List” to address scope creep — see www.ChemicalProcessing.com/articles/2009/181.html.
The second leading cause of poor estimating probably is inadequate understanding of the work. This can be especially dangerous if there’re only a few minor changes in work very familiar to the estimator. People tend to be more cautious in new territory. A common problem is misapplying a past estimate. Sometimes the simplest tasks can be more complicated than expected. Once I specified a relief valve for a slippery heating media: we used 300-psi flanges on a 150-psi-rated system to prevent leaking. Finding the right seal cost an additional 40 hours that were unplanned. Thankfully, it was a T&M contract.
Another common problem is task assignment: who does what. Often several disciplines are involved, so it’s best to have a flow diagram and hierarchy established. Just writing a memo won’t do. For example, after process flow diagrams are developed, the instrumentation group should provide input into the control scheme so the process team can create the P&IDs. Without appropriate attention, waiting for approvals and coordinating tasks can eat the whole budget.
Some engineers try to protect themselves by including a healthy contingency — this is a bad idea. It probably won’t fly with the customer. Contingency frequently is based on a percentage, which is why small (under-$100,000) projects often are over budget. Use a fixed contingency for these projects. Look at specific scenarios where the project can go sour. Be able to justify reasons for a contingency greater or less than 10%. Doing so will protect your contingency from being whittled away.
Now that I’ve discussed some of the pitfalls, let’s consider what an estimate of engineering hours should include. Consider adding these generic items: 1) estimating hours; 2) preparation for meetings — zero as a participant, a minimum of two hours per meeting hour as a presenter; 3) drafting/revising the scope; 4) writing request for proposal; 5) developing and correcting drawings; 6) vendor time; 7) permit processing; 8) training; and 9) closing the project.
Peer review is critical in assuring project success. Leave plenty of time for reviews with at least three attendees. Keep good notes and carefully file them. As a minimum, conduct a department review followed by another with a construction manager or project manager. Sometimes, a vendor can be useful at these meetings. At each step, provide sufficient hours to fix errors as well as to modify scope, schedule and, potentially, budget. Hence, pack as many meetings as possible in the beginning of a project. Always include at least one peer review before the customer review — unless you have a desire for professional suicide.
Dirk Willard is a Chemical Processing contributing editor. You can e-mail him at firstname.lastname@example.org.