Plant Engineering recently hosted a webcast “Enhancing pumping system reliability through predictive and preventive maintenance strategies” presented by Brian Kaiser, principal consultant at Life Cycle Engineering. After the webcast, Kaiser provided brief answers to a few attendee questions.
Question: Why should I still have scheduled preventive maintenance if I have online condition monitoring?
Kaiser: Preventive maintenance activities like infrared inspection, bearing lubrication, and invasive inspections provide additional insights that a condition monitoring program typically can’t cover and are still valuable as a validation check on the effectiveness of the condition monitoring program.
What investment might be needed to implement an advanced predictive maintenance/asset performance management system (APMS)?
Kaiser: The largest investment will oftentimes be in the infrastructure needed to collect equipment operating condition (e.g. flow meters, vibration sensors) and communicate that information in real-time to a location where the APMS can ingest and analyze. There may be human capital investment needed as well to have the right subject matter experts available to provide input during the APMS agent creation and to investigate and analyze when the APMS alerts that something may be wrong with equipment.
Question: It seems predictive maintenance (PdM) is the enablement tool for effective asset performance monitoring. The next trick is to do the work/repair/maintenance in a proper manner. Do you find that with existing labor challenges, organizations are keep maintenance in this CRITICAL asset category (pumps) in house or are there areas where outsourcing this function is perhaps considered a better option? For example, there is overall pump maintenance and then diving deeper there is also maintenance of critical and specialized components like mechanical seals.
Kaiser: Most corrective maintenance on pumps is done in-house by the organizations I have worked with, except for drive motor issues that require re-winding or similar activities. Even smaller sites maintain a skilled set of internal technicians given that pump issues are oftentimes production-halting events and need to be addressed ASAP.
Question: What maintenance activities are best practices on variable frequency drives (VFDs) for best motor/pump performance?
Kaiser: As most VFDs in operation today are microprocessor-based, they require minimal ongoing maintenance. We covered this briefly in the session, but for more detail:
| Activity | Detail | Condition |
| Visual inspection | Check the VFD readout for errors; verify the conditions surrounding the VFD to ensure that leaking fluids or structural failure won’t compromise the VFD | Running |
| Operating checks | Simulate operating conditions and verify that the VFD responds appropriately | Offline + LOTO |
| Configuration checks | Validate the loaded configuration in the VFD matches documentation | Offline + LOTO |
Question: Can you share a component-based failure distribution of a centrifugal pump?
Kaiser: I don’t have any resources with a failure distribution, but can list what I see as the most common failure modes that connect to components:
- Seal leakage
- Bearing issues
- Impeller cavitation
- Coupling misalignment and/or failure
- Power supply issues
- Operational error
Question: On data integrity from continuous equipment monitoring; how do you best distinguish between false data (failure imitation) and true data from equipment and conditions?
Kaiser: There’s no substitute for actual experience, so use failure signatures from past events whenever possible. To augment that, field verification with SME knowledge input will help to understand if the equipment is deviating from expected behavior. Incorporating data from the system surrounding the equipment will also help to isolate operating condition change (e.g. change in fluid viscosity, throughput change, differences in product specifications).
Question: In the context of facilities management, how do you configure trigger alerts to proactively prevent equipment damage—particularly ensuring that alerts are not set to activate too late in the production cycle? Do you incorporate metrics such as environmental conditions, usage intensity, and runtime data into your alert parameters to support predictive or preventive maintenance strategies?
Kaiser: This situation is a good application of an APMS; assuming that you can’t do that, start with a base maintenance strategy for an equipment type (e.g. all HVAC units), and then modify that with the variables that you stated (usage, environmental conditions). The base maintenance strategy should be calibrated to when signs of failure become apparent rather than when the equipment would be expected to fail (e.g. when a balance or alignment issue would be first noticeable, not when it would potentially cause bearing failure).
Question: What are the major things to consider in lifecycle costing?
Kaiser: All equipment will experience the full asset lifecycle at least once, so structured around the asset lifecycle:
| Design | Engineering and design, prototyping, simulation testing |
| Construction/Installation | Direct materials and labor, re-work and re-design |
| Commissioning | Direct materials and labor, documentation |
| Operation | Energy usage, operator labor, material consumption |
| Maintenance | Direct materials and labor, lost production |
| Refurbishment/Replacement | Direct materials and labor, remediation |
| Decommissioning | Direct materials and labor, remediation, scrap value |
There are many guidelines on this topic; one I recommend is the Asset Costing and Valuation SSG from the Institute of Asset Management.
Question: It is common to spec pumps that are driven using VFDs. Regarding pumps, is it also common for manufacturers to offer pumps with inverter rated windings, shaft grounding rings or insulated bearings (ceramic?) to protect wiring, bearings and pumps from the effects of high voltage/high frequency circulating currents (fluting)?
Kaiser: The flexibility in specifications varies widely by manufacturer, and there are many manufacturers that specialize in certain applications where hardening or specific materials are necessary. The question is what risks are being mitigated by these technologies/modifications, and is the risk mitigation significant enough to warrant the additional cost.
Question: Are there ways to apply condition-based monitoring (CBM) or predictive to a pump that doesn’t run continuously – like a fire system pump – but we need a high level of confidence that it will run when it is needed?
Kaiser: The same technologies can be applied to pumps that run intermittently, but the value of such monitoring should be evaluated as the likelihood of finding a failure before it occurs will be less than a continuously running pump. The best way to assess operational readiness (will it be able to run when needed?) is through regular inspection and run-testing when able. For the example you give (fire system pumps), many of those systems weren’t designed for run-testing, so adding the appropriate valves and piping to enable that would be a good option.
