Start condition monitoring to stop machine failures

Acondition monitoring program collects data from a machine while it’s operating and detects changes in the machine that may cause a component to fail. By flagging problems at an early stage, the program tells the user when to perform maintenance, thus preventing unexpected shutdowns. Such programs cut downtime and parts inventory, boost productivity, improve quality, and ultimately, increase profits.

Before attempting to start a condition monitoring program in your plant, identify what you want to achieve, such as increase machine uptime, prevent failure of machines that are crucial to a major operation, protect workers from the consequences of machine failure, and enhance product quality.

Then conduct a comprehensive survey of your equipment. Review the operating characteristics and repair history of all machinery in your plant, including production and maintenance records.

If you have several similar machines, check the nameplate data on each one. Its components may have been replaced without updating the machine drawings or records. Document the equipment as it exists — not as it was originally installed.

Next, review each machine’s function and duty cycle, plus its probability of failure and the consequences of an unexpected shutdown. This will give a good indication of the type of monitoring required — either periodic or continuous.

Periodic monitoring. Machines that perform secondary or support-type functions generally require only periodic data collection. Examples include slurry pumps in cement plants, condensate pumps in power plants, and auxiliary fans. These devices are usually small (less than 50 hp) and often perform redundant functions. For example, when several pumps perform the same function, one can shut down while others continue to operate, or a standby unit replaces the failed one.

Periodic monitoring is usually adequate for machines that operate reliably or give advance signs of deterioration. The interval between measurements depends on the machine’s failure history. For example, a machine that normally breaks down within minutes after the first sign of trouble needs more frequent measurements than one that takes several hours or even days.

For this type of application, choose a simple vibration detecting pen or a handheld data collector with diagnostic functions (cost $800 to $10,000). An operator can easily take such a device almost anywhere in a plant.

Though the equipment cost for periodic monitoring is relatively low, the ongoing cost can be high (depending on the number of monitoring points and measurement frequency) because this method is labor-intensive.

Hand-held devices are not advisable in some cases, either because monitoring points are inaccessible (50 ft off the ground in a grain elevator) or in dangerous areas (near exposed rotating equipment or in toxic environments). For such cases, a better choice may be a simple online data acquisition system that monitors data periodically. Such systems cost $500 to $1,000 per measurement point.

Continuous monitoring. Prime candidates for continuous monitoring include machines that run 24 hours a day, perform functions that are crucial to a major operation, have high failure consequences, are expensive to maintain, or could pose a risk to personnel safety. Examples include machines used in continuous processes, such as paper manufacturing or chemical processing, where a service interruption could ruin an entire production run. Those with high failure consequences include power plant boilers, turbine generators, and equipment located in hazardous environments. Also consider machines that experience frequent failures because of a tough operating environment — common in mines, cement plants, and fertilizer plants.

If it’s absolutely necessary to prevent a machine from breaking down, select a device that continuously collects machine data such as vibration, temperature, and flow, then compares the data to alarm criteria and alerts the operator (to take corrective action) if any data reaches these criteria. Here, the cost can range up to $1,500 per measurement point.

Continuous monitoring requires a larger investment for on-line data acquisition and analysis equipment plus installation. Once a continuous monitoring system has been installed, however, the on-going cost is relatively low because the system functions automatically with little or no operator assistance.

Monitoring multiple parameters

With either type of data collection — periodic or continuous — the best way to ensure trouble-free machine operation is to integrate the measurement of several machine parameters in the monitoring system. For example, a complex paper-making machine may require measuring motor temperature, bearing temperature and vibration, and gearbox vibration — all at various points within the machine. Such an integrated approach increases the probability of detecting a malfunction at an early stage, so operators can diagnose the problem, prescribe an operational or maintenance cure, and perform the required correction before a failure occurs.

One company, KNP Paper in the Netherlands, keeps track of bearing and gear conditions in a large paper-making machine that operates 24 hr/day, Figure 1. In this machine, moisture-laden paper passes over a series of rolls that gradually reduce its water content. Roll size and temperatures vary within three main sections of the machine, as do bearing loads and speeds. Therefore, the company installed accelerometers and thermocouples to monitor vibration and temperature at several key points in each section. All sensors are linked to a computer that analyzes the data and detects conditions such as imbalance, misalignment, and bearing and gear defects. Identifying such defects at an early stage lets the company schedule timely maintenance and avoid unexpected shutdowns.

Outside help

Many companies don’t have the expertise or the resources to start a condition monitoring program alone. If internal resources are limited, or you’re uncertain where to begin, call on an outside source for help — either a manufacturer of condition monitoring equipment, an independent consultant, or a supplier of remote condition monitoring services. These sources offer services such as analyzing your needs and making system recommendations, preparing a cost-benefit analysis, training, and even implementing a monitoring program.

It’s important, though, to establish your goals and budget for the program before you call in an outside expert, and to communicate these clearly. By doing so, you’re more likely to achieve the expected benefits while preventing misunderstandings.

To find a good outside source, remember that experience is the key. Choose a company that:

• Takes a comprehensive approach to condition monitoring.
• Is knowledgeable about current technology in monitoring equipment.
• Is familiar with the type of machinery and plant processes used in your industry and is able to tailor a program to your particular needs.

Carefully evaluating the company’s success rate, references, and reputation will help ensure that you choose the right one for the job. Ultimately, the company’s experience and capabilities are more important than the type of company or its size.

Monitoring equipment manufacturers. Companies that manufacture condition monitoring equipment are usually a good source of information and assistance, especially if you’re just starting a program.

They are usually familiar with different types of measuring devices, and may offer a wide range of such equipment — from hand-held vibration pens to on-line data acquisition equipment and software. If failure of your machinery causes a hazard, the company can probably provide protective devices that automatically shut down the machine when critical parameters reach preset alarm levels, Figure 2.

Such companies frequently offer extensive customer support and training that focuses on how to use the measurement devices as well as how to apply data acquisition and diagnostic techniques in your particular application. Most of them can train your people to take over a monitoring program in-house.

These manufacturers usually specialize in certain industries such as pulp and paper, petrochemicals, power plants, or steel manufacturing. Generally, they are best suited for handling large, complex jobs, rather than small jobs involving only a few machines or measurement points.

Independent consultants. A consultant that specializes in condition monitoring can usually customize a program that meets your particular needs. Such companies can help you set up both large and small monitoring jobs — ranging from just a few measurement points up to several hundred.

Consultants can shorten the time in which you achieve a return on investment. And you can generally find one that specializes in your type of industry or machinery. Hiring a consultant with experience in your industry frees you from committing a large amount of labor to the program, and minimizes the cost of moving up the learning curve.

Because they don’t manufacture condition monitoring equipment, consultants can objectively recommend the best equipment to meet your needs. But small consultants may not have adequate resources to handle large jobs or provide training.

Remote monitoring services. If your company is small and has lots of machinery to be monitored, it may not have sufficient in-house resources to manage a condition monitoring program. For such cases, remote condition monitoring may be the answer.

With this approach, you can take measurements in-house, then send the data by modem or through telemetry to a remote monitoring service for analysis. The remote service handles data evaluation off-site, then makes recommendations based on their analysis.

A remote service can evaluate your operation and install measuring devices in as little as a few days. Moreover, start-up costs are reduced because such a program doesn’t require a major investment in data collection and analysis equipment.

In one example, a utility in Buffalo, N.Y. saved over $650,000 in materials, labor, and downtime during the first 18 months of operation with remote monitoring. The cost for this service, including hardware, was only $35,000 for the 18- month period, giving the operators an 18- to-1 return on investment.

Remote condition monitoring also lets you “test the waters” — to find out what such a system can do for your operation before buying expensive equipment and bringing the program in-house.

Evaluating the bottom line

The best way to determine what condition monitoring can do for your company’s bottom line is to perform a costbenefit analysis. For this, you need to determine last year’s costs for downtime, lost production, energy expense, and labor, then estimate what portion of these costs can be saved by installing a condition monitoring program. Here are some general guidelines for performing such an analysis.

Calculating costs. First determine the type of monitoring equipment you need and its cost (if you plan to purchase). Then estimate labor costs for installing and operating the equipment, plus operator training.

Equipment costs. These costs depend on the type of equipment needed, and whether you use a remote service that provides some of its own equipment. A hand-held vibration pen costs a few hundred dollars. On-line devices range from $500 per monitoring point for a basic system to $1,500 for one with more sophisticated capabilities. Your equipment vendor can provide specific cost estimates.

Labor costs. Again, these costs depend on how the program is implemented. An in-house program, for example, may require dedicating half of a technician’s time to the program. If so, include half of the technician’s salary (including benefits and overhead), in your cost estimate. Also add training costs.

Calculating benefits. Though there are many ways to quantify benefits, the most common way is in terms of dollars per horsepower per year. For this method, calculate the total horsepower delivered by the machine being monitored, and divide this amount into the actual or estimated costs for maintenance and related loss of production. Compare these costs for a one-year period before and after the condition monitoring program is installed. Four cost benefit categories should be considered: labor, material, electrical consumption, and down time.

Labor: The easiest way to calculate labor savings for a particular machine is to base it on the previous year’s repair records. The number of hours spent on unanticipated repairs gives a good indication of how much time you can save (and allocate to more important tasks) after implementing the program.

Material: Check the machine’s maintenance records to determine the cost of replacement parts such as bearings, gears, and couplings.

Electrical Consumption: This is a little harder to gage, because it’s usually not included in maintenance records. But improving machine efficiency can substantial reduce electrical consumption costs.

For example, a machine that is out of alignment with its motor operates inefficiently because of the resultant vibration, friction, and wear. Misalignment also induces forces on components such as support bearings, causing bearing wear and reduced service life. After misalignment is detected and the machine realigned, friction diminishes and energy costs go down.

Detecting and correcting misalignment on a rotating machine that operates 24 hours a day can save several hundred dollars per year, depending on the motor size.

Downtime and lost production: This cost is perhaps the most difficult to determine because many companies don’t directly track production losses. However, on average, reduced downtime is responsible for 60 to 70% of a company’s savings.

These savings will depend on the type of machine operation. Consider, for example, a machine that produces $10,000 worth of products per hour. By preventing a bearing failure on this machine, you could eliminate 5 hr of downtime and a $ 50,000 loss in production.

The rate at which companies recover an investment in condition monitoring depends on factors such as the type of products manufactured, amount of downtime experienced, and especially how well they implement the program. In some cases, a company can recover its investment in equipment, training, and labor within 6 months after starting a condition monitoring program. Within a year, they can obtain as much as a four to five times return on investment.

In other cases, there may be little or no return during the first few months. Moreover, maintenance costs may increase during these early months because many hitherto unknown problems are identified, diagnosed, and corrected in a short time period. Once these initial problems are corrected, however, maintenance costs drop dramatically and remain low. If the program is not providing a return after several months, review how it is being implemented — some procedures may need to be changed.

More benefits

Though a cost-benefit analysis is an effective way to estimate the financial benefits of condition monitoring, other benefits are difficult to quantify. For example, condition monitoring can help you meet:

• Federal and other industry standards, such as those published by ANSI, ASME, API, and INPO. This may include federal requirements for condition monitoring, such as those imposed by the Department of Energy on nuclear power plants.
• Safety and quality certification requirements mandated by agencies such as OSHA as well as many insurance companies. Documentation from a monitoring program can also verify conformance to customer quality standards.
• ISO 9000 certification requirements.

Doug Johnson is a training manager at SKF Condition Monitoring, San Diego, Calif.