Bob Sansone, Principal, The Power Gen & Construction Practice LLC, Bolton, Conn, primed the audience with an overview on outage planning and execution. His presentation was not meant to get into the nitty gritty of outage management, that would have taken days.
Rather, he posited a series of questions that users should be able to answer going into the outage (Does your MMS system contain accurate, reliable information?); during the outage (Who is playing traffic cop to choreograph the activities?); and coming out of the outage (Are you capturing the work and improvements made in electronic and written form?).
Sansone’s questions got the grey matter exercised and you could tell by the note-taking in the room he was hitting on points that attendees wanted bring back to the plant.
Perhaps the most interesting part of the consultant’s presentation was the mind-expanding thought that users consider “doing as little as possible” during the outage. You could hear the “huhs?” from the group, thinking Sasone didn’t mean to say what he had said.
But he did, and clarified the statement reflecting on his experience and using statistically significant data compiled from NERC GADS (North American Electric Reliability Council’s Generating Availability Data System) for base-load coal-fired plants rated 400 MW and higher. The latter had been published previously by Canadian Consultant James Reyes-Picknell of Conscious Asset Management.
Here’s what the NERC information revealed:
* There’s more than a 40% chance that a unit will have its first forced outage on the day it returns to service.
* The probability of having at least one forced outage in the week following a planned outage is more than 60%. Also, 60% of the forced outages that occur in the first week, regardless of cause, last less than one day.
* The ratio of forced outage hours to period hours decreases at a significant rate with each successive day during the first week. Major contributors to forced-outage hours during the first week are boiler tube leaks, turbine vibration, and boiler control system failures—in that order.
Significant improvements in these numbers are possible, Sansone told the group, with better planning, greater attention to detail and top-quality workmanship, and better closeout protocols.
He supported this conclusion with more data from the survey conducted by Reyes-Picknell, where 54% of the respondents confirmed that the cause of the forced outage was directly related to work done; another 23% were not sure.
More than 55% of the survey participants opined that forced outages were more likely after doing work not originally planned into the outage.
Finally, more than half of the respondents said that their plants always or often schedule work for the outage that does not require a shutdown to perform; it’s done for “convenience.”
Sansone’s suggestions:
* Minimize business impact by keeping shutdowns as short as possible. He asked attendees to think about how much downtime costs their companies.
* Plan outages more thoroughly than you normally would.
* Watch the critical path closely; lock down scope early and add new tasks only with the highest levels of approval.
* Regarding work “found” during the outage, he urged attendees to answer these three questions:
1. Does the additional work really need a shutdown to correct?
2. Is there any way to correct the issue on the run?
3. Will it extend the shutdown?
* In addition, ask yourself and answer these two additional questions during the post-outage critique:
4. Why didn’t you know the issue was there before the shutdown started?
5. What’s wrong with your predictive maintenance program?
