Finish on-time, on-budget with better planning, preparation, execution
Freddy Alvarez has managed his share of gas-turbine (GT) outages. The former plant manager, now director of operations for EthosEnergy, and his colleague Chris Wilkinson, VP major maintenance, told GT users participating the webinar “Delivering a cost-effective outage on-time and on-budget,” that a success-based attitude is what’s critical to dealing with the unexpected and still achieving project goals. “You know going into the outage that you may find something wrong,” Alvarez opined. “Having said that, the only question is, ‘How do we still achieve success?’”
Let’s say we’re ready to start a gas-turbine outage on a Monday morning, he continued, and want to be ready by the following weekend to perform a cold start on the unit. We know that managing schedule and scope are very important and that every hour translates to cost, and we must continually explore opportunities to improve efficiency and control spend. To achieve success, Alvarez suggested focusing on three areas: planning and preparation, execution, and discovery.
He began by identifying the real risk to any outage plan: It’s discovery during the execution phase of the project. Example: In the middle of the outage, the plant and contractor recognize they probably should do an extra item or two—that is, expand the outage scope. This presents the possibility that a workmanship issue may be encountered while performing the new work. What are the consequences of that with regard to inventory management? Do we have the parts?
What about schedule? Any schedule creep will have to be reviewed/coordinated with the other stakeholders: Power marketers/dispatchers, management, contractors, logistics, owner’s engineer, other contractors onsite that may be affected. What about cost? Will there be market availability consequences, possibility penalties over and above the cost of the physical work?
Points made, Alvarez circled back to the all-important planning and preparation (PAP) phase. “The soft stuff is the hard stuff,” he quoted a mentor as having said. Planning and preparation is not exciting, for sure, but single-handedly, it can impact efficiency and effectiveness the most. Invest the right resources and time into PAP and outage risks and potential cost impacts go down, Alvarez added.
Regarding the planning and preparation associated with work-scope development, he said, it’s a data-driven process first and foremost. In EthosEnergy’s experience, most plants are conducting the required equipment condition assessments and they rely on the computerized maintenance management system (CMMS) as their data repository. Both are very important, he stressed. Engage your contractors early, Alvarez urged. Providing them access to your previous outage reports, inspection reports, equipment database and work history, etc, is very important to creating a meaningful and complete work scope with the benefit being mitigation of discoveries during project execution.
The EE outage experts next introduced failure-mode effects analysis (FMEA), a risk management tool, into their presentation. Recognizing that everyone on the webinar does not use the exact same tools and acronyms, Alvarez explained: There may be other types of software or spreadsheets or processes that you follow, but what’s important is their ability to identify risks in advance.
More specifically, a planning engineer identifies the tasks, the steps, the procedures that the outage contractor is going to perform—whether it be on a valve, the removal of a gas-turbine case, etc—and uses the analytical software to identify potential risks in that work and the probabilities of their occurrence. This process identifies high-risk items and allows you to prioritize work scope. Plus, if done early enough in the planning process (a year or more prior to outage start), you can use the results to develop realistic budgets for capital expenditures, labor, and consumables.
Failure-mode effects analyses are missing in many outage plans because plant owner/operators often choose to rely on a “standard” scope. Customers who do not see the value in FMEA before the outage sometimes realize it afterwards when their expectations have not been met. The bottom line: FMEA is a valuable, proactive tool for mitigating operational risks.
Alvarez transitioned to “cost,” which he said requires an understanding of the “rules of engagement.” Whether you’re on the operations side or performing the work, you must understand what is, and is not, scope. If you have a term maintenance agreement, your knowledge must expand to understand what’s planned, unplanned, and extra work. Add to that vocabulary in and out costs, logistics, division of responsibility, plus knowledge of who is responsible for the cranes, scaffolding, etc, as well as who really is in charge of the project and how to contact that person.
Last point is particularly important when you discover something abnormal and have to contact, at the moment, the person who can put his or her signature on a work order for a $50,000 repair. Better still, establish an extra-work purchase order in advance. Sure, there can be boundaries—a not-to-exceed dollar figure, for example—but a firm understanding of the processes and financial limits of extra work are critical for avoiding project stall. If applicable, review time-and-materials (T&M) pricing in advance, too, and discuss and agree on terms before the outage starts.
Scheduling was not a significant part of the presentation because of the ready availability of software, such as Primavera and Microsoft Project, which generally are well known to station personnel and are acknowledged as beneficial in the planning and coordination of work. He noted the value of these tools for identifying resource constraints and for schedule optimization when dealing with multiple contractors.
However, Alvarez was quick to point out that the availability of such tools is not the end-all to scheduling. He said scheduling requires discipline and a dedicated champion. It’s not about appointing someone “outage coordinator” to update the schedule, it’s about managing the schedule. The downstream effects of not managing the schedule include adverse impacts on the completion of tasks on time and within budget. The most successful outages were said to be the ones having dedicated managers focused on nothing else except managing administrative functions.
Sometimes, the former plant manager went on, we see a senior control room operator, or an operations manager, in charge of scheduling. Sounds like a great idea because these people certainly are in the know, but the problem is they’re also needed elsewhere—such as for walk-down inspections, coordination of lock-out/tag outs (LOTOs), etc. Scheduling is a priority item; it demands a person dedicated to the activity.
Data trumps all arguments, Alvarez said. It reflects what the equipment is telling you and needs to be listened to. Of course, it’s your responsibility to aggregate in the CMMS, on an on-going basis, all applicable photos, videos, links, and commentary from inspections and maintenance activities for use in developing outage scope. Regarding the last point, what are you seeing during preventive maintenance? For example, is the valve you’re inspecting corroding? If so, enter that in your notes. What steps did you take to mitigate the corrosion? Put that information in your notes as well. The more detail, the better.
Predictive maintenance—for example, vibration monitoring, thermography, oil sampling, etc—contributes significantly to the accuracy of failure prediction. This information is reviewed during the outage planning phase to be sure you don’t overlook an important action item. It also assists in identifying what parts should be ordered in advance of the outage.
Condition assessments are especially valuable for use in developing a meaningful outage scope and for identifying capital budget requirements. A year or 18 months prior to a major inspection, Alvarez said, conduct comprehensive condition assessments of major equipment—including heat-recovery steam generators, high-temperature piping, condensers, gas turbines, steam turbines, generators, etc—to the degree possible without compromising the plant’s mission.
Operating and trip logs provide useful information as well, especially details on “abnormal” events that offer clues regarding the factors contributing to damage. Remote monitoring findings from the OEM or a third-party services provider is yet another source of data, particularly valuable for its trending of critical variables over time. And don’t overlook data from the combustion dynamics monitoring system (CDMS); it too tracks equipment behaviors over time.
Alvarez then offered an example of how data from the foregoing resources, received and evaluated well in advance of an outage using FMEA and other analytical tools, identified the root cause of a problem that in all likelihood would not have been found otherwise. Critical to this solution, he said, was the partnering between EE and the customer. Gaining access to all the data and sitting across a conference room table from the plant staff—an environment conducive to a free-flowing exchange of ideas—enables a capable contractor to understand in detail the plant’s problems and tap the thinking of deck-plates personnel who had been dealing with the issue, and trying to correct it, for years.
Here Alvarez stressed that there’s no substitute for having the “right” people in this type of meeting. “If we’re sitting across from a manager who just has a list of the trips, or a list of the information available, that’s not nearly as good as having subject-matter experts from both the plant and contractor collaborate on the issue at hand.”
This experience supported the notion of early contractor involvement. It’s not possible to review data, identify problems, extract plant knowledge on issues, propose solutions for consideration by the owner in one or two outage meetings. It’s an evolutionary process, he said. Plant personnel resources are limited for this type of collaborative investigative activity to be sure, but if you think of the effort as “discovery” in advance of the outage it becomes clear that the time constraints are far less onerous than when problems are identified during the outage.
The “ideal” outage profile is not the same for every plant; in fact, what’s considered “ideal” for a given plant today might change by the next outage. In some cases, the spend may be what drives the outage scope and decisions; for other outages it may be schedule. Perhaps, too, there’s a warranty constraint, a regulatory constraint. So, it’s important for the major-maintenance service provider to understand the plant owner’s priorities to assure an outage solution that meets expectations.
Alvarez stressed conducting inspections as soon as the outage starts—perhaps dedicating the first shift or two to inspections. The logic: If you’re going to find something, discovery needs to happen early to minimize potential schedule impacts. Make every effort to have LOTO, confined-space, and hot-work paperwork completed before the outage starts.
On the first day of the outage the control room can look like a goat rodeo, the former plant manager said, with two- or three-dozen contractors, several-dozen representatives of those firms wanting to coordinate their work. You want to avoid this to hold costs in check and maintain high productivity. There’s no reason for someone to be in the control room at 0700 and not starting work until after lunch.
Key is to get all the admin work done the day before and do walk-downs the night before. Having a plant subject-matter expert dedicated to the major-maintenance contractor contributes markedly to a smooth outage. That person is there when you conduct the inspections, Alvarez said, he or she helps you through the LOTOs, etc. It’s like having a personal plant consultant from the operational side on your team—invaluable.
Hold points are important to discuss in detail before the outage begins. Plants have different hold points for witnessing and observing critical welds, tests, etc. The contractor, inspector, and plant owner have to manage their respective schedules and lost time results when these people are not in synch.
Unplanned work is identified, so now what do you do? Alvarez said, “Find a deviation.” Having onsite engineering disposition enables this, he added. EthosEnergy tries to have an engineer onsite, at least during critical operations. Borescope inspections of gas turbines is a case in point. If the major maintenance contractor has responsibility for these repairs, you want an engineer available to make the call on a repair or replace decision, for example, so you don’t have to wait on conference calls. The ability to handle repairs in-situ and promptly can save shifts.
Contingency plans are important to addressing discovery issues quickly and effectively. Formulation of contingency plans is part of the planning and preparation step, of course. Preapproving of parts, suppliers, and vendors is critical to maintaining outage schedule, and possibly cost as well. Trying to qualify a supplier on the spot can strain an organization in an outage; circumvention of qualification processes is ill advised.
Understand your inventory requirements in advance of the outage so parts you expect to use are on-hand in advance of need. It’s a good idea to revisit sharing agreements with neighboring plants in case you become squeezed for parts. Discuss with your contractor its ability to provide a just-in-time parts solution, a rotable exchange, etc. You can’t do too much contingency planning, which should go beyond components to shop capabilities/time if needed. CCJ