Beyond Vibration: Maximizing the Value of Your Predictive Maintenance Routine

If your plant is like most, you have been doing vibration testing diligently for a long time. But, if you have been following the same routine for a year or more, then it’s likely time to freshen things up by incorporating other, less-obvious facets of equipment health.  You can easily do this without investing a lot of money – in fact, you probably already have the tools you need.

Mnet 165913 Motor Pic 0

If your plant is like most, you have been doing vibration testing diligently for a long time.  You are likely collecting data quarterly, monthly, maybe even weekly on critical machines.  Experience has shown that vibration testing is saving you a lot of money. 

If you have been following the same routine for a year or more, then it’s likely time to freshen things up by incorporating other, less-obvious facets of equipment health.  You can easily do this without investing a lot of money – in fact, you probably already have the tools you need.

Clues to machine health

Vibration is the most widely used indicator of machine health, and for good reason.  It’s non-invasive, a program can be set up in less than a week, and it is applicable to a wide array of equipment types.  However, vibration alone cannot provide the whole picture of asset health. A machine could be in very poor condition and close to imminent failure, yet vibration and oil analysis might indicate everything is running smoothly.  So, if you already have a good vibration monitoring program in place, how what are the next best steps to ensure that these underlying problems are identified in time? There are many simple ways to more fully assess the health of your machinery with minimal investment. You can likely utilize the data collector you already have, or work with your vibration testing vendor to improve your machine health monitoring strategy.

Simple process readings are the most under-utilized method of tracking machine health.  Almost all vibration data collectors have the capability of logging temperature measurements. Most can even perform simple calculations like a difference between two temperatures.  You (or your vibration monitoring vendor) are already walking down the plant with the data collector. It’s a very small effort to capture some key data while you are out there— a small effort with a potentially huge payout.  The biggest challenge is figuring out which data to take and how to use it - this is often equipment-specific. 

Let’s take a look at some common equipment types and what can be done to better understand the health of each.

Steam Turbines

This is the most important machine in your plant.  A sick steam turbine is very bad news.  Luckily, you can obtain deep visibility into the future health of your turbine with some simple checks.  The procedure below describes a very basic health check for a steam turbine that can easily be performed quarterly, even monthly.  For more insight, find an expert to do a more detailed evaluation at least once per year.

What you’ll need

  • Vibration Data Collector or other data logging instrument.
  • Pressure readings from the inlet and outlet of each turbine stage (HP, IP, LP).  Likely from plant instrumentation.
  • Pressure readings at each extraction and any other pressure readings that are available between the inlet/outlet of each stage.

What to do

  1. Ensure the turbine is operating at steady-state, and ideally at full-load.
  2. If possible, operate the turbine at valves wide-open, sliding pressure mode.
  3. Record all of the pressure readings listed above while the turbine is at steady state.
  4. Calculate the ratio between each consecutive pressure reading, following the flow of steam from one to the next.

How to interpret the data

What you are looking for is consistency.  The pressure ratios should not change over time.  A sudden change could indicate damage or other impact to efficiency.  If you are unable to operate the turbine in sliding pressure mode, then the ratio between the HP inlet pressure and the next available pressure will vary with valve position and this is normal.  If your pressure ratios change over time, then a more detailed evaluation can tell you why and what anything needs to be done.

Bearings

Bearings are in every rotating machine in your plant.  Without healthy bearings, your plant won’t run,   which makes checking vibration levels the first step in understanding bearing health.  This is the simplest test there is, but one of the most important.

What you’ll need

  • Vibration Data Collector or other data logging instrument.
  • Spot Radiometer (sometimes called a Raytek or temperature “gun”)

What to do

  1. Using the spot radiometer measure the casing temperature of each bearing.  Mark the spot for repeatability during future checks.
  2. Calculate the Temperature Rise - the difference between each bearing temperature and ambient air.

How to interpret the data

Bearing Temperatures will vary with the temperature of the ambient air, but the temperature rise should remain fairly steady.  A sudden change in temperature could indicate a change in loading of the bearing - perhaps from misalignment or damage to the machine.  Gathering the data and trending it over time will help you figure out what is “normal” for your equipment.

Heat Exchangers

Heat exchangers perform critical roles but are often neglected in predictive maintenance programs. Whether you have radiators, oil coolers, air coolers, condensers, feedwater heaters, or anything that moves heat from one fluid to another, following these simple steps can help to identify when heat exchangers need maintenance (and when they don’t).

What you’ll need

  • Spot radiometer
  • thermometer or thermocouple for measuring ambient air temp
  • Vibration Data Collector or other logging device

What to do

  1. Measure and record the temperature of each inlet and outlet to the heat exchanger.  If existing instrumentation is not available, use the spot radiometer to get a surface reading on the piping as close to the heat exchanger as possible.  Mark the spot where you take your reading with matte black paint to ensure repeatability with each test.  If your heat exchanger uses ambient air as the service fluid (e.g a radiator), then use the thermometer to measure the ambient (inlet) air temperature along with the temperature of the air immediately leaving the heat exchanger.
  2. Calculate the “Process Fluid Temperature Change (PFTC)” - the difference in temperature between the entering and exiting process fluid.
  3. Calculate the “Service Fluid Temperature Change (SFTC)” - the difference in temperature between the entering and exiting service fluid.
  4. Calculate the “Temperature Change Ratio” - the ratio of PFTC to SFTC.  Record this value.
  5. Calculate the “Approach Temperature” - the difference in temperature between the entering service fluid and the exiting process fluid.  Record this value.

How to interpret the data

If the calculated values don’t change from one month to the next, then your heat exchanger performance has not changed.  You can compare your data to the design specifications to see how well it’s performing.  If it’s meeting your service requirements then no further action is necessary.

If the Approach Temperature increases significantly (5 degF or more) with no significant movement in the Temperature Change Ratio, then you may have fouling or blocked tubes.  Clean the heat exchanger to correct this issue.  If the service fluid outlet temperature is considered acceptable, then maintenance can be deferred until the condition worsens.  Continue to monitor.

If the Approach Temperature and the Temperature Change Ratio both show significant deviations, then a flow condition may have changed (e.g. Process fluid flow has increased) OR you may have a significant problem.  Flow measurements can be difficult to obtain.  A heat exchanger expert can help you sort things out.

Pumps

I’ve never been to a facility that didn’t have pumps.  Chances are you have many pumps of all different types and sizes.  No matter what kinds of pumps you have, a simple health check can go a long way in understanding how they are performing and when they may need service.

What you’ll need

  • Two pressure gauges - one sized to measure inlet pressure, the other sized to measure outlet pressure.  Many pumps already have gauges in place which will work just fine if they are calibrated regularly.
  • An ultrasonic flow meter or other flow measurement device
  • Pump design curve for your pump (can be obtained from the manufacturer)
  • vibration data collector or other logging device
  • spot radiometer

 

What to do

  1. Make sure the pump is operating at steady state conditions
  2. If the pump does not already have instrumentation, then connect the pressure gauges to the inlet and outlet using appropriate fittings.  You may need your maintenance team to install appropriate taps and valves first.
  3. Secure the ultrasonic flow meter sensors to the outlet piping.  Make sure that you are measuring only the flow from this pump and that all T’s are downstream of your reading.  Read and understand the user’s manual for your flow meter to get accurate results.
  4. Record inlet pressure, outlet pressure, and Flow (usually in GPM)
  5. Calculate the Total Developed Head - the difference between outlet and inlet pressures expressed in feet of water (or the pumped fluid, if not water).

How to interpret the data

On the pump design curve, locate the operating conditions for your pump and mark that point.  If that point is on the design curve, then your pump is working as designed and no maintenance is necessary. 

If the operating point is on the curve, but is very far to the left or to the right of the Best Efficiency Point (BEP), then the pump may be oversized or undersized for the application.  A pump expert can help you determine if a different pump could save you a lot of money in operating costs. 

If the operating point is significantly below the curve, then the pump is not performing as it should.  There is a good chance the impeller is worn and needs to be rebuilt or replaced.  This maintenance can often be deferred as long as the service requirements of flow and pressure are being met.  Monitoring the performance over time will allow you to predict when you will need to do this maintenance in the future.

Communicating Your Findings

There are some final considerations in launching your new program.  Keep in mind your audience and how you will communicate the results and findings.  You may already issue vibration reports or you have a tracking system to which other stakeholders have access.  Ideally, you will have one source for all equipment health information regardless of the test methods and data.  If you use an outside vendor to perform your vibration services, you may ask them if they can perform these simple health checks while they are onsite.  Predictive maintenance vendors often offer a full suite of services.  Armed with a fresh predictive maintenance routine and the right tools, you will be on track for maintenance success.

Dave Geswein has filled various roles with Azima DLI, and has built and maintained the largest cloud‐based PdM programs in the world. Prior to joining Azima DLI, Dave was a Reliability and Performance Engineer for Duke Energy. He is a Vibration Institute certified Category IV Vibration analyst, registered as a Professional Engineer in the state of Colorado, and holds a B.S. in Mechanical Engineering from Purdue University.

More in Home