Time for Maintenance

As the post-modern era adage goes, “If you can't maintain it, you can't afford it.” A simple statement conveying a very poignant truth, applicable as much to individuals and households in this consumerist era, as to manufacturing enterprises around the world.

Indeed, initial costs do matter, but lifecycle costs matter more. To justify the expenditure on the stateof- the-art and technologically advanced, one has to expend more – money, time and effort – to derive maximum value. This applies as much to white goods in households as it does to equipment and machinery in manufacturing setups or to infrastructure assets like pipelines and roads.

From simple cleaning, dusting and lubricating to sophisticated remote monitoring of machinery, “maintenance” is verily a family of a diverse range of activities, which is usually carried out by trained in-house personnel – operators or specialists, but may also be subcontracted out to the service engineers employed at the supplier-OEMs. It is usually a combination of both these approaches.

Industrial equipments may be assembled in less than a day. They may be dismantled, parts sent for reuse and/or materials for recycling, or crushed in a shredder, in a few hours. However, in the interim – the period between birth and death; the active, functional lifetime of the machine in other words, extends over some years. During this period, while value is derived from the equipments by the end-user, they expect to be tended to in return. The better this tending-to, the greater the value one derives from a machine.

Depending on the type of equipment/machinery, the operation-maintenance phase, apart from accounting for most of the machine's lifetime, may even contribute to the bulk of the lifecycle environmental impacts (an automobile for instance), and could also account for the lion's share of the lifecycle costs incurred on the equipment.

Be this so or not, it is expedient to expend an optimal amount of time, money and effort on maintaining equipments in manufacturing and processing facilities; while designing the maintenance methods and strategies with an eye on efficiency and effectiveness.

And users of capital goods in practically all the different types of firms in the manufacturing and process sectors, are availing of services from several productsuppliers in the instrumentation and controls sector – Fluke, GE Energy, Emerson Process Management, National Instruments, to name but a few.

Timing it right
As with manufacturing and production processes, the seemingly-mundane maintenance activities are facilitated and aided by hardware (sensing, measuring and recording instruments primarily), software (interpreting the signals recorded, and triggering or advising action) and humanware (committed managers and trained specialists and operators aware of the role proper maintenance plays in a setup).

The communication channels linking the hardware with the software could either be wired/cabled or wireless; transferring data signals over short distances within the plant, or over relatively longer distances to plant personnel or equipment suppliers controlling and monitoring the equipment remotely. A parallel can be drawn to geriatric care in countries like Japan, where the health of patients who stay alone is monitored remotely, so that aid can be rushed to the site, without the patients being expected to call for help.

The machines cannot “speak” for themselves – quite like the aged and sick – but devices embedded within them – just like sensors monitoring the health of patients – can speak up for them, so to say. Thereafter, a stitch in time saves nine – preventive or proactive maintenance, in other words. However, it is also well-known that the degree of wear of equipment is related directly to the usage rate. As preventive maintenance schedules are usually based upon the maximum usage rate, equipments which do not have 100 percent utilization may be replaced long before they experience significant wear.

Here, if the “in time” aspect in the afore-used adage is interpreted as exactly when it is needed or just-in-time, one moves over into the realm of predictive maintenance (which is in effect, both preventive and proactive, but avoids redundancy; by matching the supply of servicing exactly to the need for the same).

Taking this approach a step further, and defining maintenance actions as measures undertaken to maintain the functionality and reliability of machines and equipment, there may be instances when what the end-users expect from a piece of machinery becomes the guiding criterion behind the necessity of undertaking maintenance activities. This would bring one to the realm of what is termed as reliabilitycentered maintenance.

And then of course, there is the passive approach, which at times is inevitable owing to several constraints, or is often adopted for want of a well-developed, robust maintenance regime.

This goes by the name of corrective maintenance, or, in laymanlingo, repair. Any real maintenance regime essentially – especially when one considers a huge plant with numerous pieces of machinery and equipments – would be a mix of all the methods or approaches.

The total approach
Total productive maintenance (TPM) is a Japanese contribution which evolved from preventive maintenance in the automobile industry, to something more holistic for the entire manufacturing sector. Targeting (ideally) zero defects, zero breakdowns and zero accidents through effective housekeeping and technology usage, TPM takes maintenance to a realm of greater importance within manufacturing enterprises.

In a paper published in mid-2010 in the journal Global Journal of Management and Business Research, researchers from the academia in Bangladesh pointed out that TPM was indispensable for the Bangladeshi pharmaceutical industry which brings in much-needed foreign exchange earnings for the country.

Stressing on advocating the Japanese 5-S approach (Seiri, Seiton, Seiso, Seiketsu and Shitsuke) to ensure order, cleanliness, organization and methodicalness, the authors laid down guidelines for a planned preventive maintenance master plan in the spirit of TPM:
• Evaluate equipment and understand current status
• Restore basic condition of equipment
• Establish information management system
• Build periodic maintenance system
• Build predictive maintenance system
• Build spare parts management system
• Build a lubrication management system

It will be apt to add at this juncture that research and development in laboratories forms an indispensable component of the pharmaceutical industry and (according to Pharmatech.com) maintenance and service-related items are often the second-largest budget element in laboratories after salaries and employee benefits.

Maintenance, thus, incurs costs, but it is certainly not to be looked upon as a no-profit undertaking. Viewed from a different perspective, a holistic cost-benefit analysis of maintenance activities will reveal that they facilitate profits indirectly by avoiding losses of some categories and reducing others (unplanned shutdowns, drop in product quality and subsequent loss of consumer confidence, accidents calling for payments of compensations to victims, etc.).

Also, in the pharma-sector especially, being pulled up for noncompliance to quality guidelines which encompass equipment maintenance as well, is often something a firm would want to avoid at all costs.

In the pipeline
Serving as conduits for fluids, pipelines are also essentially pieces of equipment. Any major failure has its genesis in a defect or a set of defects (see Figure 1). If these defects are observed well in time, proactive maintenance activities would avoid deterioration. If defects go unspotted for a long time, and mature into deterioration which fortunately gets spotted, preventive maintenance activities are undertaken to avoid collapse.

Predictive maintenance is even more advanced – experience and historical trends enable one to understand exactly when a defect would worsen into deterioration, or for that matter, deterioration would worsen into a collapse. Action is taken just-in-time to avoid deterioration in the first instance, and a collapse in the latter.

However, one assumes that the reliability and functionality of the asset is not compromised while waiting for this right moment. If the deterioration does not affect the functionality of the asset in any way, reliability-centered maintenance would dictate that it be accorded a slightly-lower priority in the list of the to-beserviced components in the system/facility/plant. When a collapse occurs – with defects and deteriorations either being unnoticed or ignored – something needs to be done to get the component/ equipment/system back into working order. This is when corrective maintenance or repair is undertaken.

The damage in this case – socially, economically and environmentally; both direct and indirect – has already been done. An unplanned shutdown results. Remedying the damage calls for even more expenses, discomfort and annoyance to both the utility managing the pipeline network and the end-users (the consumers of water supply and sewage transport services).

In the oil & gas and power sector, equipment failures have often cost the companies in question, hundreds of thousands of dollars a day in lost production revenue and consequent maintenance expenses.

Figure 2 (sourced from Fluke Corporation) compares the per-hour revenue losses incurred globally in different process/manufacturing/generation sectors, owing to unplanned downtime. This is very crucial for the energy sector, as seen in the graph, while the pharmaceutical industry loses more than the chemicals, food & beverages, metals, electronics, construction and engineering sectors.

Maintaining information
Computerized maintenance management systems (CMMS) – web-based or LAN-based software packages – integrate the care and upkeep of entire plants, facilitating a hawk's eye view and thorough understanding of the goings-on inside and among equipments and machinery on the shop-floor.

It must be mentioned at this juncture that it is not just the production and manufacturing in the foreground which need to be monitored continuously, but also the health and condition of the enabling equipments and machines in the background! It is not just the “what and how much” that is important, but also the “how they are when they do what they are supposed to do” which needs to be accorded importance. For instance, the thermal imagers (referred to earlier) could be looked upon as input devices to a CMMS which records the status of equipments; and advises proper action at the right time.

A CMMS effectively is a huge data-store which maintains information about a facility's maintenance operations. The benefits include the following:
• Recording data about equipment and property including maintenance activities, specifications, purchase date, expected lifetime, etc.; and generate metrics like the Facility Condition Index, to measure the effectiveness of asset management.
• Facilitating the verification of regulatory compliance.
• Helping management make informed decisions – for instance calculating the cost of machine breakdown repair versus preventive maintenance for each machine, and thus leading to better allocation of resources.
• Producing status reports and documents giving details or summaries of maintenance activities (causes of problems, downtime involved, actions taken/not taken, and recommendations for future action).
• Enabling the management of spare parts, tools and the reservation of materials for particular jobs, recording where materials are stored, determining when more materials must be purchased, tracking shipment receipts and taking inventory.
• On the whole, enabling maintenance personnel to do their work more effectively.

According to Emerson's Spencer Kwan, the difference between an asset management system and a CMMS is that the former is event-based, scanning assets for alerts, while the latter, on the other hand, is transaction-based, typically generating work orders, scheduling maintenance and checking spare parts inventories.

“Asset management systems (AMS) typically monitor and collect information on the health of critical plant assets. Online, real-time information about assets, including health and process variables, is distributed to maintenance and operations staff, giving early warning of asset degradation. Asset information can generate alerts and give recommended action to the right people in the plant for fast response. Fast action on developing issues can mean downtime is limited, a shutdown is avoided, or a safety situation is prevented,” explains Kwan.

He also maintains that automated communication between the real-time asset management system and the CMMS mean that islands of asset information are integrated to give a holistic asset view that helps plant staff make the best decisions for improved asset performance and plant reliability.

“Emerson views its AMS Suite predictive maintenance software and the CMMS as complementary systems in a plant. They should be sharing information so plant staff have an integrated view of assets to support maintenance department planning and scheduling. By incorporating both real-time, online asset information with a CMMS, users have the benefit of event based information and transactional information.

“When used as complementary technologies, the AMS system serves as the event trigger to launch the CMMS transaction. The online AMS scans for an alert, which can automatically initialize the CMMS work order process. When both systems are integrated, users can effectively automate the entire maintenance process from point of condition alert through completion of the maintenance work order.” There are numerous options available in the marketplace for facilities keen on installing and availing of the aforesaid benefits of a CMMS. To name a few: Atlas 2000, eWorkOrders, Idhammar MMS Software, Intooligence, Maintenance Connection and MPulse Maintenance Software.

Pump it up
Against the backdrop of Emerson Process Management introducing its Integrated Pump Health Monitoring system recently, Tim Olsen, refining consultant at Emerson made the point that that over two-thirds of all essential pumps in industrial sites are not monitored online.

Analogous to the defects in pipelines depicted in Figure 1 are the warning signs given out by pumps in bad health – cavitation, excessive temperature and vibration, among others. These, at times working in concert and reinforcing each other, could damage the pump, and also other components of the circuit/network within the plant connected to the pump.

“Data compiled from multiple industry sources indicate that pump availability is a critical factor, and pump failure accounts for approximately seven percent of the total maintenance costs and 0.2 percent of lost production,” says Spencer Kwan, director, Asset Optimization, Emerson Process Management Asia Pacific. “Based on our study, we feel that annual savings for a typical refinery using pump health management can amount to US$500,000 or more.”

Kwan also points to the fact that reactive maintenance costs are 50 percent higher than planned and unplanned maintenance costs and that Emerson is focused on providing the best uptime to customers for the pumps they buy. “The return on investment (ROI) we calculate is based on reduced maintenance, increased uptime, and avoiding fines. The bottom line is more availability and a safer work environment,” he notes.

Damage to components apart, unplanned shutdowns will result – eminently avoidable in crunch situations when manufacturers adopt the just-in-time production approach. It is good to be wise after the event, but better to be wiser and avoid the event if and when you can. No doubt, Emerson Process Management will not stop with playing doctor-and-nurse to pumps alone, but will gradually extend its outreach to several other industrial assets.

Vibrations are early warning signs of trouble-to-come, on the shop-floor, not just for pumps, but for all machinery with moving parts which are translated, rotated, oscillated or reciprocated, by a prime mover (electric motor usually) consuming electrical energy. However, if one does not look out for warning signs, they may go unnoticed; and akin to cancer cells not detected early enough, wreak havoc.

GE Energy, thereby, would be very happy to have acquired Commtest, a New Zealandbased provider and designer of machinery health information systems and upgraded its portable vibration data collection and analysis capabilities.

In the absence of proper cooling, thermal stresses within equipments increase to unsafe levels, which trigger deterioration and eventual damage. Thermal imagers, like the ones supplied by Fluke, enable the inspection of a host of equipments across a variety of industrial sectors – transformers, transmission lines, capacitor banks, electric motors, generators, pumps, compressors, tanks and vessels, pipes and valves, reactors, evaporators, couplings and gearboxes, rollers, belts, switchgear, among others.

Industrial equipments apart, inspection of roofs, walls, chimneys etc., ensure that enclosures are given the degree of importance as the equipments housed within them. Detection is the first step – having done that, one would decide upon the best course of action to stem deterioration and avert damage.

And speaking for National Instruments, Chandran Nair, managing director, South East Asia, tells CE Asia that the company's LabView software has evolved from to becoming an enabler of predictive maintenance of equipments industry-wide.

“LabView is a full-fledged programming environment used to develop various kinds of applications. It can be used to program at a high level of abstraction, as well as to very low levels, depending on the skills and needs of the developer.

“In the early days, most of the LabView applications were based around instrument control and data acquisition. But as it has developed, and its programming community grown, the applications developed using this tool now span many different application areas, including machine control and predictive maintenance of machines on the shopfloor.

“While NI enables predictive maintenance and machine condition monitoring solutions, many of our Alliance partners engage in equipment maintenance and use NI tools to implement their solutions.” says Nair.

---

Terminology Time

To get an understanding what industrial machine/ equipment maintenance is all about it is important to get a hold on the terms used to refer to the various approaches adopted:

Corrective maintenance
Also called repair. It is reactive in nature and action is taken only after failure has occurred.

Reliability-centered maintenance (RCM)
This can be defined as a process to ensure that assets continue to do what their users require them to do, in their present operating context.

Preventive maintenance
Care and servicing (oil changes, lubrication etc.) by personnel for the purpose of maintaining equipment and facilities in satisfactory operating condition by providing for systematic inspection, detection, and correction of incipient failures either before they occur or before they develop into major defects.

Proactive maintenance
Maintenance strategy for stabilizing the reliability of machines or equipment. Its central theme involves directing corrective actions aimed at failure root causes, not active failure symptoms, faults or machine wear conditions.

Predictive maintenance
The condition of in-service equipment is determined continuously using non-destructive testing methods (infrared, acoustic, corona detection, vibration analysis, sound level measurements, oil analysis etc.) in order to predict when maintenance should be performed. It is more of a just-in-time maintenance. This approach offers cost savings over routine or time-based preventive maintenance, because tasks are performed only when warranted.

Total productive maintenance (TPM)
Targeting zero defects, zero breakdowns and zero accidents through effective housekeeping and technology usage, TPM takes maintenance to a realm of greater importance within the manufacturing enterprise.

---

NI LabView Goes Healthy

In October 2011, National Instruments introduced the Watchdog Agent Prognostics Toolkit to expand the capabilities of its LabView system design software in machine prognostics and health management (PHM) applications.

The toolkit is a product of collaboration between NI and the National Science Foundation's Industry/University Cooperative Research Center on Intelligent Maintenance Systems at the University of Cincinnati. It was developed with an eye on serving engineers' increasing demand for monitoring and reporting the health status of machines, structures, components and systems.

“Companies are looking for a systematic approach to rapidly develop and deploy prognostics for failure prevention, health monitoring and machinery prognostics,” said Dr Jay Lee, professor and director of the Center on Intelligent Maintenance Systems.

“The new Watchdog Agent Prognostics Toolkit for NI LabView provides an easy-to-use solution for predictive maintenance and prognostics.”

According to NI, the Watchdog Agent toolkit gives engineers a ready-to-run prognostics solution that can greatly increase engineering efficiency for developing any PHM application. The toolkit works with the signal processing capabilities of LabView and the analysis capabilities of the NI Sound and Vibration Measurement Suite. It provides a set of algorithms including logistic regression, statistical pattern matching, a self-organizing map, a support vector machine and a Gaussian mixture model.

Engineers can use the algorithms to create machine and component status descriptors of operating states and failure modes. The algorithms convert multiple field sensory readings into summarized health information values for efficient monitoring. The toolkit also includes a health radar chart that displays organized health values of multiple machine components on a single display.

In addition to integrating with the Sound and Vibration Measurement Suite, the Watchdog Agent toolkit can read history data collected from the NI measurement hardware including NI CompactRIO, CompactDAQ, PXI and PCI. It also integrates with the IOtech eZ-TOMAS Technical Data Management Streaming (TDMS) data plug-in, vibDaq from CalBay Systems or other sensory data acquisition systems based on NI TDMS data files.

---

‘Look for signs of failure to protect critical assets'

Complete online monitoring of the plant's most critical assets includes shutdown protection, predictive, and performance monitoring – all integrated with the process automation system, says Deane Horn.

Rotating equipment seldom fails without providing hints well in advance. Machinery health warning signs come in the form of vibration changes, process parameter changes, and performance changes to name a few. So imagine a form of machinery protection that allows you to anticipate and recognize these warning signs. A complete protection strategy can then be formulated based on the use of early information gathered from a combination of vibration, performance, and process data. Shutdown protection would be relegated to the last line of defense, and costly outages could be eliminated.

Advanced technologies, including online and wireless vibration monitoring and ASME calculations based equipment performance, can all be integrated with the process control system to nurture the health of machinery that is essential to maintaining uninterrupted production.

More than ever, it is important to utilize reliable information about the operating condition of critically important process equipment, not just a “trip” signal that comes only after significant internal damage has already occurred. It is no longer prudent to rely heavily on a vibration protection system for the most critical machines in your plant.

Machinery shutdown protection is only part of a complete online monitoring strategy to guard against events that can happen suddenly with little or no apparent warning. It is very important to have the right monitoring equipment, trained personnel, and a good analytical software package to pick up and identify those signs of failure long before a key compressor, turbine, gearbox, coupling, or even instrumentation fails unexpectedly. Timely maintenance is preferred over catastrophic failure and the costly repairs that will follow.

Mission: machine health
Studies indicate that more than 50 percent of industry maintenance man-hours is spent fixing equipment after a failure has occurred, whereas less than 18 percent of the time is spent determining when equipment might fail and acting accordingly. The numbers will improve only when maintenance departments establish the monitoring of machine health as a key mission, and go beyond shutdown protection that meets minimal requirements.

The four components of a complete online monitoring solution for a plant's top five percent most critical assets are:
• Shutdown protection monitoring
• Prediction monitoring
• Performance monitoring
• Integration of all the above with the process automation system

Shutdown protection monitoring
These systems typically function to prevent severe machine damage or even injury in the event of a totally unpredictable event, such as a turbine blade suddenly breaking away due a metallurgical imperfection.

In this case, the “trip” signal must be instantaneous, like the inflation of an automobile air bag, to minimize damage. Protection systems are still necessary for safety because unpredictable events can and do happen. However, implementing prediction monitoring with protection monitoring can allow “black box” analysis after an unplanned emergency trip.

The machinery health monitor captures information at all bearings simultaneously across an entire machine train so that a user can replay the event as it unfolds and determine the cause of the event. In this way, similar shutdowns can be prevented in the future.

Prediction monitoring
Newer technologies are enabling users to predict what was once thought unpredictable. In reality, most events actually can be predicted. As long as personnel have access to detailed diagnostic information, they can identify potential faults and gauge their severity months in advance. This allows maintenance planners to determine the optimum time to make repairs.

Prediction monitoring of rotating assets is intended to provide the information needed for accurate planning. Information is frequently obtained by acquiring machinery vibration data and analyzing signatures and levels, either periodically or continuously. Analysis of the results generally leads to a decision as to how long that piece of equipment can continue to operate productively before maintenance will be necessary.

While periodic monitoring of production equipment using a hand-held data collector has been a staple of maintenance departments for decades, the data produced by this method are valuable in helping analysts determine what was happening at the time readings were taken, but equipment that is critical to keeping a production process in operation should be monitored continuously. Indeed, some critical situations can be averted only if real-time data on equipment condition is available.

Online monitoring using the CSI 6500 Machinery Health Monitor, for example, provides a continuous flow of data, so a changing condition can be immediately recognized. When properly interpreted, these signals pinpoint the location, nature, and even the severity of developing problems. Personnel can use such data to predict with greater accuracy when a machine will need maintenance to prevent damage and avoid lost production Predictive maintenance of rotating assets also uses information gathered through oil analysis, infrared imaging, and ultrasonic detection. If the data indicates trouble ahead, a judgment can be made as to when a failure might be expected. With critical equipment, immediate repairs may be necessary. That's when a reliable early warning system will pay for itself. On the other hand, it may be possible to delay repairs until a scheduled unit turnaround.

Ultimately, technology helps plant/unit maintenance managers make business decisions about what to do, when, and how to do it. Gathering and analyzing of machinery information is far less expensive than reacting after something breaks.

Performance monitoring
With as much as 50 percent of machinery malfunctions caused by the process around the machine, real-time vibration information integrated with the process automation system can give operators key information to understand how their actions can have an impact on machinery health.

Performance monitoring is based on comparing the actual realtime performance of a major piece of equipment, such as a boiler or large compressor, with design specifications. As equipment performance deteriorates, energy usage increases and throughput decreases. Plant personnel may not even be aware that the performance of a piece of equipment is below normal or that it is consuming excessive energy.

Actual efficiency loss versus design for the given operating conditions can be determined by comparing a machine's actual performance with a thermodynamic model. Specialists are thus able to identify laggards and formulate actionable recommendations. For example, blade fouling on a compressor can degrade performance and simultaneously increase vibration. Online water washes can improve performance and minimize downtime needed for an offline wash, but most turbo compressors are being pushed to their design limits and sending a bunch of water into the compressor can be quite stressful. An online water wash can be observed from a performance perspective to ensure it is helping to increase efficiency – and from a vibration perspective to ensure the compressor is not damaged. If efficiency in downstream stages does not improve, an offline water wash may be necessary.

Without real-time performance feedback, degrading performance may go unrecognized by production personnel for a long time, but the rigorous mathematical routines embedded in this technology will highlight operating degradation very quickly and help identify root causes.

Complete solution
To obtain a continuous flow of vibration measurements, a large turbo generator might have more than 10 bearings with two sensors at each bearing plus other unique instrumentation – like speed sensors, thrust sensors, and eccentricity measurements. As many as nine different types of measurements might be needed at various locations down a machine train depending on the size of the machine.

The cables leading from these sensors are connected to online monitoring hardware that is the foundation for the complete online solution. By measuring for detailed vibration in addition to peak vibration, the complete turbomachinery protection system, which is intended as a retrofit on shutdown systems, has the ability to recognize developing conditions as well as a severe condition requiring total shutdown to protect the machine.

Ideally, data from continuous monitors are integrated with the plant's control system, so vibration monitoring becomes an extension of the central control system. When the position and the motion of the shaft inside the bearings are integrated with the control system, operators can see for the first time what is happening deep inside critical machinery – information of much greater value than just a trip signal once vibration has exceeded a limit.

When operators have real-time vibration information at their disposal, they can observe the impact of process adjustments on a machine's health and learn what steps actually improve performance.

New meaning to protection
Traditionally machinery protection systems monitor for high vibration and close a relay when a setpoint is exceeded to initiate shutdown of a machine. Looking for signs of failure rather than a setpoint excursion to protect critical assets gives new meaning to protection and even adds a new layer of protection, decreasing the need to actually call upon that relay to fire.

Complete online monitoring of the plant's most critical assets includes shutdown protection, predictive, and performance monitoring – all integrated with the process automation system. Technology is providing new opportunities to improve overall machinery health, increase efficiency, and optimize the process.