”In-sha-la,” he said as he walked with his hands in the air in frustration. This poor contractor was the focus of my pestering about acid-proof coatings for the pump bases. It was hot, growing dark, and his family was waiting. The site representative, Osman, a Turk himself, smiled. “What the hell did he say?” I asked. “If Allah is willing,” Osman said with a smirk. I got it: if we were very lucky, we’d get our concrete coating before the start-up. Things were going badly. The PLC had no power, the lab and control room were empty and I had a start-up crew arriving in five days. This is a perfect example of how not to do commissioning— rushing from construction to the start-up.
A start-up is only as easy as the preceding steps: construction and commissioning. If I may, let me offer some suggestions to make this go smoothly: 1) daily meetings are critical; 2) don’t confuse ISO-9001 with punch lists; 3) develop the schedule after reviewing the construction — not before; 4) calibrate instruments and check-out equipment at the vendor shops; 5) train maintenance staff before commissioning; 6) review spare parts before, after and during commissioning; 7) lastly, do not rush start-up.
Scott, the best project manager I ever knew, held daily meetings during construction and commissioning. Minutes were posted for input from all the trades and were the topic of regular toolbox discussions. We developed contingencies and check-out procedures for each instrument and item of equipment, e.g. pumps. Every detail was considered by the group in the morning meeting. Engineers were assigned to study problems. Unlike his unfortunate predecessor, he was allowed to set his own start-up date. Construction and commissioning were allowed to proceed methodically. As a result, the second reactor was commissioned in three days, instead of 10 days— we scheduled four days. In addition, he cut costs by almost $1 million dollars by re-thinking many hasty decisions. What was missing from that project was the safety review. A daily review should reduce mishaps.
On a project in the Philippines we completed ISO-9001 data sheets. These are very helpful but they are only a checklist. What—is obvious from a sheet. Why—is something for a punch list. Based on the punch lists, which were issued daily, we were able to develop a reasonable schedule; the start-up and operational acceptance test (OAT) went relatively smoothly.
Perhaps the most critical step in commissioning is the check-out at the vendor’s shops. If this step is missed it could cause significant delays in start-up. For instruments and PLCs, an independent contractor can be useful but be careful in this case. During the reactor commissioning we found that the contractor had set the ranges were wrong in virtually all the pressure transmitters. It was a scramble but the manager had planned extra contractor help to support in-house maintenance: a good contingency plan.
One of the worst mistakes is failing to train mechanics and instrument technicians on new equipment: things tend to get broken. It is best to bring these people in early to help with the check-out. A trusting relationship must be established between the technicians and the engineers so that the engineers, when necessary, can do some of the hands-on work required to write realistic procedures.
The smartest construction manager I ever knew had extra parts for all critical items like packing for the absorber. Roger kept a daily tab on parts. Extra parts are crucial to any contingency plan.
A successful OAT is the product of a chain of events. A thorough and methodical commissioning is a link in that chain. Once the commissioning is complete, pause to take inventory and review procedures in preparation for the start-up to come.
By Dirk Willard, senior editor