Are you actually measuring your downtime? Even if you are, you might be missing opportunities that help beyond the correction of individual downtime events.
It’s 2 o’clock on a Saturday afternoon. You’re the shift production supervisor, and you get a call from the press operator station. “We just had an overload trip on number four press pump,” the voice says. You respond: “We had this problem yesterday, too. Let’s get maintenance down there and look at the pump.”As you hang up, you’re thinking to yourself that the plant sure has a lot of problems with pumps.
Downtime information is essential to correct ongoing machinery problems and deficiencies, and to fine-tune maintenance and operations management systems. Many facilities, though, still do not measure downtime. Even if they do, they often miss opportunities that can help the plant in larger ways than correcting one downtime event.
Downtime costs plants millions of dollars each year in lost production, downgrade and loss of customers. That’s why it is so important to know what’s causing the downtime and how to use this information to correct the problem.
There are various methods used to track downtime. The simplest is where an operator merely fills in a log book, noting what happened, what was done about it and how long production was down. This is where many older plants started–and where some of them have remained. Many of them still are not measuring their downtime.
As industrial plants started to mature in the 1970s, many switched from using logbooks to adding downtime details on production forms that were collected at the end of the shift. These forms were kept on clipboards and made available for plant personnel to read.When a clipboard got full, the forms were filed.When using logbooks and forms, trending typically was not reviewed except for a month-end report that listed the total hours down. Sometimes, plants would separate the maintenance from operational downtime, maybe even by craft (electrical vs. mechanical) and, if they were clever enough, by equipment area such as press, former, drying, etc. Scheduled and unscheduled downtime would be tracked as well. It is important to evaluate both scheduled and unscheduled downtime to attempt to reduce each.
In the 1980s, the beginning of the computer era, plants started to use spreadsheets and databases to track downtime. In the first “computerized maintenance” years, many plants collected downtime from forms filled out by operators.Administrative personnel would fill in “electronic” spreadsheets and databases from these hand written forms. This allowed for misinterpreted information-which often resulted in misrepresentation of root cause.
By the 1990s, computers had become much faster and less expensive. The spreadsheets improved and some plants had operators inputting data into home-brewed downtime databases or were using software sold by various companies. It wasn’t until the mid ‘90s when computers were extremely fast and had large memories that plants really started to understand the importance of good downtime data.We then saw plants use more sophisticated databases to track downtime.
Many plants subsequently began evaluating Overall Equipment Effectiveness or OEE, which is the true cost to the plant. The overall performance of a single piece of equipment (or even an entire plant) is governed by the cumulative impact of the three OEE factors:
- Availability (or downtime)
- Performance rate (or optimum production rates)
- Quality rate (or downgrade)
OEE is a percentage derived by multiplication of these three factors.
Plants now can buy computerized maintenance management software, CMMS/EAM or other Web-based and HMI systems that can report real-time OEE or downtime information for instant management control. Software packages are available to connect to equipment controls to indicate precise time and device information. These same controls also can track OEE.While some maintenance software systems now have downtime tracking capability, some plants still rely on their own database packages. There also are plants that collect no downtime information at all.
Benefits of downtime analysis
Downtime tracking and analysis is reactive. Something happens, and we do something about it-after the fact. On the other hand, through the development of good maintenance and operations programs, downtime can be reduced.
Many industries, however, have not advanced their maintenance technology to the point where it is possible to operate without downtime.As an example, there are plants that are working toward a yearly goal of 97% for 365 days. For some industries and plants this may seem impossible, while for others 97% is terrible, and any unscheduled outage simply cannot happen.
Currently, a wood products plant is considered to be running well if it has 95% total uptime. That includes all scheduled and unscheduled downtime. A 95% rate is 22.8 hours per day. Even at 95% uptime, the lost time during the year represents a substantial decrease in possible profit margin. Short duration, repeat offenders will cause downgrade of product. Most continuous process plants are meant to operate all the time with scheduled, proactive maintenance. If the plant is up and down all the time, not only is there loss in production, but there also can be product quality, safety and environmental issues coming into play. Good downtime analysis will help both maintenance and operations in determining the root cause of nagging problems.
Effective downtime collection and analysis
For downtime information to be effective, the data must be easy to enter and understand, and must include enough detail to allow good root cause analysis. The latest automated systems will collect accurate information if enough effort is placed into monitoring the correct components and if the operations and maintenance personnel add their comments.
Be aware that the right information needs to be collected and entered. The operator needs to add the time that production stopped and started. If this goes past a shift, the next operator should enter the startup time. Items that should be recorded include:
- Stop/start times to the minute
- Operator name, shift and crew
- Plant area(s) affected, such as: Lathes
- Plant equipment shut down, such as: Lathe #1
- Equipment area such as: Lathe Spindle, and the equipment identification code or number
- Sub-equipment, if known, such as: Lathe Hydraulic Pump #2, and the equipment identification
- The component that failed, if known, such as: The pump itself or, better yet, the pump front bearing
- Failure code, such as: Tripped, Stopped and Jammed
- Reported problem, such as: “The pump overloaded and kicked out.”
- Action, such as:Welded, Replaced, Filled and Cleaned
- Shift maintenance review approval. Someone from shift maintenance during which the event occurred reviews the downtime entry and approves the details, or has further comments.
- Maintenance comment example: “After resetting this motor overload and restarting,we noticed high vibration from the front pump bearing.We checked the pump bearing temperatures and vibration level and they exceeded safe operating conditions, so we shut down and replaced the pump.”
- Shift supervisor review approval
- Shift supervisor comments
- Maintenance management approval
- Production management approval
- Work order number for this event
- Root cause: Lack of lubrication
- Root cause program failure: PDM
- Root cause program failure note: “This pump had not been identified as requiring vibration analysis.”
- Follow-up required: “Add pump to vibration analysis route. Repair pump removed.”
- Follow-up work orders: There may be more than one.
What a list! And to think we started downtime tracking by simply entering some details in a logbook.Not all of these listed items are required, but the more you document, the easier it will be to determine the root cause.When using a CMMS/EAM or database, drop-down choices can be selected to speed up the process of selecting the various options. Options should be parent/child driven, such that when you select “pump” as the component, there are limited choices for pump failures. The same holds true for equipment. When the press area is selected, only the press equipment and its sub-equipment should be listed as dropdown choices.
What to do with the data
Now that you have collected data, what do you do with it? Hopefully not what we did with the paper forms we collected before-when the clipboard got too full, we tossed the forms or, maybe, put them in a box to store somewhere.
It is good to review downtime daily and assign someone to correct the issue. Unfortunately, we often get so tied up in looking at the day-to-day issues and not finishing what we started yesterday that we lose track and never get back to solving the root cause of larger problems.With work orders, we at least have a better way of tracking these opportunities. But, what do we do with the history?
The key to preventing downtime lies in its history, as long as the right data has been collected.We need to know the following from the data collected:
- When it happened: Date and time
- How long production was down
- What plant area, equipment, subequipment and component failed?
- Who was involved?
- What was the root cause and solution?
- What type of program failed? Was it due to PM, training, management decision, improper engineering, improper installation or poor design? What caused the component to fail and forced the plant to shut down?
- Is this a repeat offender? Have there been multiple events of the same problem? How many times? Is there a trend?
- Is it happening at a certain time or season? Is there some typical frequency?
By using a CMMS/EAM or database, charts can be developed to show trends that can lead to root cause analysis and solutions. Don’t overlook scheduled downtime when analyzing downtime, either. Chart downtime as follows:
- By year
- By month
- By day
- By crew
- By shift
- By plant location, such as: Log Yard, Press Line #2, Finishing, etc.
- By equipment, such as: Press Loader, Core Flaker, etc.
- By component, such as: pump, motor, switch, gearbox, conveyor belt, etc.
- By failure code, such as: overload, tracked off, spark detect, etc.
- Root cause program failure: PDM, Training,Resources,Engineering/Design, etc.
With good information on downtime, problems can be solved and downtime reduced.When a downtime event happens, such as a pump bearing failure because of misalignment, you need to not only resolve that pump/motor issue, but also look at other alignment issues with similar pump/motors, and the alignment program for the complete plant. By identifying a potential program failure, such as the PM procedure of checking for misalignment, and correcting that for the entire plant—not just the one pump that failed—you will solve many more problems and reduce downtime quicker. It is far better to solve the overall maintenance program problem of this PM, than fire-fighting and chasing misalignments each time they occur.
You can’t work on everything at once, so don’t try.You must prioritize those items that cause the most downtime hours and the most events. Select the top three for each—then solve them. Select the top three downtime events by hours, and the top three for frequency for the plant, by plant area, by component and program failure. You may have other downtime problems you resolve right away to keep running, but you must have an ongoing list of priority downtime-related projects in front of you to reduce downtime. Then, when you have solved one, add another to the list.
Gaining a thorough understanding of its downtime can help your company increase uptime and profit margin. MT
George Meek began his career in the ’70s as an electrician. Today, he is a process specialist focusing on hands-on engineering and maintenance projects with Evergreen Engineering, Inc.Headquartered in Eugene, OR, Evergreen specializes in industrial engineering and maintenance consulting for industries worldwide. E-mail: firstname.lastname@example.org; telephone: (541) 484-4771; or Internet: www.evergreenengineering.com