Originally published in the September 2012 issue of Pumps & Systems.
When it comes to improving pump reliability, technology and people go hand in hand. In a follow-up to last month’s article (“Pump Reliability,” Pumps & Systems, August 2012), which focused on pump reliability programs, this month two of the most important elements of a successful program — well-trained people and accurate information — are discussed. Assembling a team trained in failure analysis and equipping its members with powerful software will help end users reap the benefits of improved reliability, including lower maintenance time and costs, a safer working environment, and increased productivity.
Establish the Reliability Team
Establishing a reliability team is the first step toward an effective program. Whether end users are building greenfield plants or looking to improve reliability at older plants, it is never too early or too late to create a reliability team.
Many equipment vendors offer resources, such as on-site personnel and data analysis, to help end users run their reliability programs. Depending on the size of the facility and the number of pumps and issues, an equipment vendor can place a dedicated engineer on-site full time to help end users manage reliability. Alternatively, the employees can work with the end user’s internal team on an as-needed basis. In addition to equipment vendor representatives, the reliability team should include maintenance engineers, operations personnel and procurement employees from the end user’s staff.
Since seal operation is a critical part of pump reliability, end users should also consider establishing seal teams as a subset of their reliability teams. This group can spot seal-related trends and issues, helping improve overall reliability.
Set Reliability Goals
Once end users assemble their reliability teams, the teams should establish short- and long-term performance targets based on the current equipment population and work order history. A review schedule should be created to measure the progress of the program.
While most teams focus on increasing the mean time between repair (MTBR), measuring the mean time between change (MTBC) can be an even more valuable metric. Any time a pump comes out of service, it is expected that the end user will change the mechanical seal before putting the pump back into service, which involves time and money. By measuring MTBC, end users capture every incident — not just those that involved repairs — giving them a better sense of their plants’ reliability and productivity.
Use Reliability Software
Providing teams with the right database software makes reliability programs significantly more effective. When information resides in silos throughout a plant, it can be difficult to see the big picture and identify recurring issues. Recording all pump-related information in a central database allows team members to spot trends and problem equipment more easily.
Reliability software can also be a time-saving survey tool. For example, if maintenance employees need to know what seal is installed on a certain pump, they can check the database quickly instead of checking the pump files or, worse, disassembling the equipment. Reliability databases should provide comprehensive information for each piece of equipment — including the model number, service parameters, part numbers, seal arrangements, seal materials and coupling details.
End users choosing a database software program should consider the following features:
Security access — Limiting the number of people who can input data helps keep the database relevant and uncluttered. End users should look for a program that provides customized security access. For example, some systems will allow organizations to give editing capabilities to certain users and read-only access to others. Large organizations may need a system that offers multiple security levels where users only have access to the unit they serve.
An Internet-based system — When people in multiple locations support an account, an Internet-based system allows all members to see updates in real time, which helps avoid confusion. Using an Internet-based program also means that end users do not have to install software on their servers, eliminating potential IT issues.
Exporting capabilities — End users should look for systems that allow them to export data easily to outside programs, such as the Microsoft platform. That way, team members can review data even when they do not have Internet access.
Flexibility — End users should choose software that can provide customized reports based on the information that is most relevant to their facilities. For example, some plants want to focus on pumps with the highest failure rates, while others want to focus on the pumps that cost the most to repair. The most effective systems offer custom report-generating capabilities and allow end users to select the metrics they want to track.
Establish a Bad Actor List
Pump reliability is a good example of the Pareto Principle, or the 80/20 rule — 20 percent of the pumps in a facility typically cause 80 percent of the reliability issues. End users should identify the bad actors — the pieces of equipment that pose the most problems — and then fix them. This is the most efficient way to improve pump reliability.
Database software allows reliability teams to create bad actor lists easily and accurately based on the criteria most important to them. Typically, end users consider failure frequency, failure cost and how critical the pump is to operations when placing equipment on the list. In addition to identifying pumps that meet bad actor criteria through the database, reliability teams should ask maintenance and operations personnel which pumps they think cause the most problems.
Reliability teams should monitor the bad actor criteria and subsequent lists continually. End users should keep in mind that pumps will not drop off the list just because they have changed their seals or made other repairs. With time, the list should evolve as overall reliability improves.
Failure Analysis
Once bad actor lists have been established, reliability teams can use their database systems to analyze what is causing each piece of equipment to fail. An effective software program should be able to track common causes of failure, such as flush flow interruption, process upsets, pump misalignment and high vibration.
Based on their findings, reliability teams can make recommendations for improvements, such as changing the flush plan or operating conditions or replacing bearings. End users will ultimately need to decide if the recommendations are cost effective and feasible. Recommended changes can sometimes be impractical, such as moving a pump several feet away from its current location. By determining the reasons for failures, however, end users will know their options and can make informed decisions about how to improve reliability.
Plant-Wide Improvements
Along with troubleshooting bad actors, database software allows reliability teams to spot trends across their plants and make general reliability improvements. For example, running a root cause report for all pumps in a plant may reveal that many failures are common to a particular seal type, prompting end users to switch to alternative methods or designs. If the database shows that human errors are causing a significant number of failures, reliability teams can implement training for maintenance and operations personnel on topics such as lubrication practices, mechanical seal selection, seal standardization and API practices.
When experienced reliability teams are coupled with database software, they form a powerful partnership for improving reliability. By investing in their teams and technology, end users can unlock the benefits of pump reliability programs.