Construction sector urged to go more granular in benchmarking of project costs
The construction sector must harness the power of benchmarking if it is to end a decade of stagnant productivity, according to the manager of capital-projects consulting at petroleum and chemicals advisory AP-Networks.
Productivity has been flat in the construction sector over the last 10 years, Shawn Hansen noted in a presentation to Petrochemical Update’s Refining, Engineering & Construction Conference 2016 in Houston last week.
Overall productivity has grown at about 1% per year over the past decade in the United States. Accountants and lawyers have improved productivity at a rate of 2% per year in that same period, Hansen noted.
“We (construction) have had less productivity days than lawyers. And this is substantial,” Hansen said. “Two percent per year over 10 years is more than a 20% difference in productivity that other sectors have been able to achieve or we have not been able to achieve.”
For productivity to improve, the construction sector will also need to embrace digital technology and advanced job planning, Hansen said. He argued there was a direct correlation between the legal sector’s high adoption of digital technology and high productivity, versus the construction industry’s low adoption of digital technology and low productivity. Regarding job planning, he said only about 60% of field labor hours are spent on tools. The remainder is spent waiting for materials, getting back to workstations after a break, and so on.
No more predictable than before
Despite greater availability of data, large capital projects are no more predictable today than 10 years ago, Hansen noted. According to AP-Networks’ database of projects worth more than $20 million, about 72% are failing to achieve cost, schedule or safety targets (with failure measured as 10% more than the original median estimate). One in four projects are “train wrecks” in which budget or schedule blow out by more than 25% or in which there are material delays during startup.
Hansen recommends beginning with an estimated Lang factor, a ratio of the total cost of installing a process in a plant to the cost of its major technical components. The Lang factor for fluid process plants should land between 4.74 to 6.21. If a project’s estimated Lang factor is higher than that that of similar projects, this means there is an opportunity for savings, Hansen explained; if it is low, there may be some risk associated with the project.
Lang factors and cost-capacity curves are good tools for Class 4 or Class 5 estimates, which are typically used in the early stages of projects to determine feasibility and get to preliminary budget approval. But in order to provide more definitive estimates, Hansen says far-more granular benchmarking data is required.
“Benchmarking performance on capital projects is challenging. You need to understand the regional differences and cost impacts of labor productivity, construction, regulation operating conditions, and even design,” he said, adding that one must also take into account the different requirements between newer plants and older plants in regard to automation.
It is not enough to use average project costs as a benchmark. To get a more accurate figure, one has to calculate labor hours in relation to scope quantities.
“We’ve developed a basket or set of benchmarking models aimed at looking directly at labor hours based upon the actual scope. We relate the design labor hours to scope, we relate the direct field-labor hours to the scope quantities," Hansen explained. “For instance, how many tons of steel? How many cubic yards of concrete? How many feet of pipe? Instrument count. Equipment count. Ways to quantify the size of the equipment. Relating this type of scope to the field labor hours. And then looking at the indirects as well, and having it measured to models around the indirects that you’d layer on top of that as well.”
What you shouldn't be doing
Historical data does not always apply to the future, Hansen further warned. Technology advances can change costs, as can importing of labor into the U.S. from other parts of the world.
“You need to have the ability to understand how you stack up on field labor hours in those types of cases,” he said. “We have regulatory changes that impact design, but also what’s happening out in the field: the design and engineering advances; rapidly changing engineering and construction market conditions; and the rapidly changing exchange rates and escalation.”
AP-Networks has been working with the industry for 15 years, and Hansen and his colleagues have seen plenty of owners take the wrong approach in that time. One common mistake made by organizations is to cut office costs rather than going back to the root cause of high projects costs. “One approach has been to not open the purchase orders soon enough for the engineering contractor,” he said. “We’ve seen that happen in several circumstances, and (the result has been) not getting the engineering work done on time to support turnarounds, and then of course you end up with engineering rework, and then problems during construction as a result of that.”
Another common mistake made by companies is to lower the cost of construction supervision. Again, using benchmarks to arbitrarily cut costs can mean you are setting over-aggressively targets that lead to overruns. Hansen explained, “At the end of the day the costs come back. We see that oftentimes “we’re focusing more on the symptoms than on the drivers of performance.”