In brief:
- An effective maintenance program is founded on a critical equipment list developed through a rigorous analysis of the probability of failure and the consequences of failure.
- Multiple sources of data are available for quantifying the probability and consequences.
- The details of the resulting maintenance program needs to be reviewed no less than annually to accommodate changes in equipment condition and technology.
Preventive maintenance, run-to-failure, zero tolerance for equipment failure, predictive maintenance, operator rounds, thermography and routine maintenance are the tools and programs that facilities use to ensure plant equipment performs effectively and efficiently.
Of course, we all know there’s much more to this picture than a simple question and subsequent answer. What would a perfect power-generating facility look like from an operations and maintenance perspective? What would be the perfect activity balance to maximize resources while operating safely and operating within budget? This can be broken into several categories.
The first category would be management perspective or management protocol with regard to maintenance. Different managers can have various mindsets when it comes to preventive maintenance versus run-to failure models. Some managers established hard lines for the preventive maintenance and corrective maintenance work activity percentages.
Corrective maintenance typically falls into one of two categories. The first category is when equipment that has been well maintained simply breaks or fails to some degree that it requires maintenance. Corrective maintenance to make any needed repairs is scheduled and performed.
The second corrective maintenance category is when equipment is run to failure deliberately before any corrective maintenance is scheduled and performed. This is what we’ll call the run-to-failure model. As indicated earlier, some managers have hard percentage numbers for the relationship between corrective maintenance and preventive maintenance. A typical scenario would be 60% of maintenance work hours are scheduled for preventive maintenance and the remaining 40% of scheduled hours would be for corrective maintenance.
Define the numbers
There are two basic approaches to determining these percentages. One approach holds that the more preventive maintenance performed, the less corrective maintenance will be needed. An example is changing the oil in an automobile. It’s generally accepted that if motor oil is changed regularly, the less engine maintenance will be required. Using the same vehicle example, a person wouldn’t consider changing certain pieces, such as a muffler, until it actually fails. That’s the run-to failure approach given that mufflers rarely fail and the cost to replace one is relatively low. To review the first item in our discussion, management philosophy normally determines the percentage for corrective maintenance work hours and preventive maintenance work hours.
[pullquote]
The preventive maintenance percentages are the leading indicators for the overall maintenance program. Managers who follow this percentage guideline would expect the corrective maintenance work hours to be 40% of the total work week hours, if 60% of the hours were spent on preventive maintenance. If either percentage fell outside a predetermined tolerance window, the maintenance program would undergo a review.
Now that we’ve defined the two subcategories of preventive maintenance and run-to-failure in a maintenance program, we need to determine the category in which each piece of equipment falls. Most maintenance programs begin by establishing a critical equipment list.
Make a list
The list can be delivered several ways, but most programs use some form of the probability and consequence table to determine the category. This involves a determination that quantifies the probability of equipment failing and the consequences associated with the failure.
Consider a 1-to-10 scale with 1 being a low probability and 10 being the highest failure probability. Management needs to assign the actual scale numbers to use for this determination. Most equipment in a power-generating facility has a low failure probability, or else it wouldn’t have been purchased and installed in the first place. Even so, there will be some equipment with a higher failure probability than others, and these receive a higher probability rating. Because probability failures are normally low, the consequence of a failure number assigned to the piece of equipment carries more weight than the probability.
But we need to use both numbers in our run-to failure versus maintenance performance equation. We’ll use the same 1-to-10 scale for the consequence, with 1 being low-to-no consequence and 10 being high negative consequence.
How do we quantify the equipment failure consequence? Most power-generating facilities use the safety, environmental, equipment and megawatts (SEEM) method. Safety should be the No. 1 criterion when determining negative consequences. Going through the acronym, the second most importance consideration is environmental impact, followed by equipment damage and finally megawatt loss.
Each item is important and should be considered when determining the equipment failure consequences. Determining the consequence level and the failure probability allows us to determine the maintenance frequency. The critical equipment list is the final list of identified equipment with the highest level of probability and consequences.
Data sources
There are many tools and programs that can help determine equipment failure probability and consequences. The plant’s historical records are a good starting point. Reviewing the equipment past performance, including corrective maintenance work activities performed along with documentation from any preventive maintenance records, helps establish a number for the equipment failure probability and consequence.
Company records of like-in-kind equipment can be a good information source for the determination. Consult with those at other power-generating facilities within your company. This is normally relatively easy.
Industry standards and operating experiences (OE) are another good source of information. In the digital age, searching for similar equipment types is as simple as a few keystrokes. Check for lessons-learned records from other power-generating facilities. If possible, find a point of contact and communicate with the site. Most facilities will share generic information about equipment failures and maintenance programs.
[sidebar id="1"]
Another source of information is vendor recommendations. No matter how good the working relationship with your company’s vendors is, they’re going to default to more conservative data when it comes to equipment maintenance. This isn’t necessarily a bad thing, but make sure you factor in other resources of information for your probability and consequence equation.
Individual crew member experience usually is a good source of information. Don’t limit your equipment questions to the most experienced crew members, but question anyone that has been exposed to maintenance on the equipment. Sometimes the newest crew member will have the best insight on maintenance. Along that same line, question other work groups, such as operations. Plant operators can certainly bring a different perspective to suggested equipment maintenance. Plant operators know which equipment gives them the most trouble. Ask operators what equipment failure they’d most hate to see. Other suggested work groups to question might be an environmental specialist and a regulatory specialist, if your site has them.
How often?
Once armed with a critical equipment list, it’s time to determine the frequency of maintenance needed to maintain the equipment properly. The tools described above also can be used to determine the frequency of maintenance needed for a piece of equipment. I’ve seen maintenance frequencies range from daily checks to a five-year inspection frequency, and everything in between. Establishing the initial maintenance frequency isn’t an exact science, but using the tools described above should get you in the ball park.
You should review your maintenance frequency list yearly, at a minimum, to ensure the frequency is still adequate. Sometimes maintenance frequencies need to be extended, and sometimes they might need to be shortened. For example, a weekly inspection on a bearing temperature might be extended to monthly if historical records indicate there has been no movement in the temperature reading. Be careful not to extend the frequency prematurely, but only after careful research and review. Another example might be yearly inspection of a motor that is getting some serious age on it. In this example, the maintenance frequency might need to be changed to semiannual or quarterly. Again, you don’t want to shorten the maintenance cycle prematurely, but if you’re going to make a change, this would be the more conservative approach.
Spec the maintenance
When you determine which equipment needs to be part of the preventive maintenance program and you establish the maintenance frequency, you need to determine the type of preventive maintenance to perform. Most maintenance activities fall into the intrusive and non-intrusive categories. Intrusive means there might be a high possibility that you’ll introduce problems when performing the maintenance. Some types of electrical checks might cause wiring insulation to break down if not performed carefully.
While we need to have confidence in the technicians performing the maintenance, we need to help eliminate any potential problems that could result from performing the maintenance. It might be necessary to perform the electrical check based on the research performed, but there might be a suitable alternative that can shift the maintenance activity into the non-intrusive category, while obtaining an acceptable test result. Predictive maintenance normally falls into the non-intrusive maintenance category.
It’s advisable to use as many of these tools as needed to ensure your equipment maintenance program is sufficient to meet the goals your management team established. No one tool gives all the information needed to set up a quality maintenance program. The program needs to be reviewed yearly, at a minimum, to determine if the overall program or any individual maintenance frequencies need to change.
A quality maintenance program integrates each step outlined above. An understanding of management team expectations with regard to the balance between corrective maintenance, run-to failure and preventive maintenance is a primary starting point. The expectations, or management philosophy, should include a basic approach to developing your critical equipment list. This usually includes some variant of the probability and consequent curve. Preparing a critical equipment list is needed to place equipment into its proper maintenance category. Then the equipment can be incorporated into a frequency-driven work activity, taking into account the degree of maintenance needed. If you can choreograph the steps outlined above, you’ll have a quality maintenance program.
Jeff Townsend is hydro maintenance and operations superintendent at Progress Energy in Raleigh, North Carolina. Contact him at [email protected] and (910) 439-5211.
