SPECIAL OFFER! Join ISA now and get the rest of 2023 Free.

  • By Tom Bass
  • System Integration
Steam trap monitoring enables predictive maintenance
Monitoring identifies steam traps and downstream equipment issues to maintain plant efficiency

By Tom Bass

Steam distribution systems support manufacturing across a huge range of industries, from craft brewing to oil refining and most everything in between. Steam can carry enormous amounts of energy, and it is valuable as a highly controllable heat source. On the other hand, producing steam is energy intensive, and an ineffective distribution system can be wasteful. Boiler designs can be highly efficient, but this efficiency can be rapidly lost with a poor distribution system.

Boiler design and efficiency factors have been the topic of countless articles, so here we will concentrate farther downstream and examine the distribution itself, especially steam traps (figure 1). They are the primary tools for separating condensate from steam. A steam trap failure can be predicted to some extent through nonintrusive ultrasonic acoustic event detection using data generated by a wireless acoustic transmitter. Fixing failed traps early can also help to prevent problems in downstream equipment caused by passing condensate slugs through traps.

Condensate is sent back to the boiler as feedwater, which makes it valuable for two reasons. First, boiler feedwater is heavily treated with expensive chemicals to avoid boiler fouling, so any that can be recaptured saves money. Second, condensate is usually hot, which reduces the amount of energy needed to turn it back into steam. Consequently, condensate collection is critical for overall system efficiency, and it depends on steam traps.


 

Figure 1. Steam traps are an important part of a distribution system, but often receive little attention.

Steam for heat transfer

We will explore more about transferring heat from steam in a moment, but steam is also used to capture heat. The term heat-recovery steam generator (HRSG) has become more common with the higher energy costs and carbon footprint concerns of recent years. Where older plants might have simply blown hot exhaust or process gases out a stack, they are now more commonly channeled into an HRSG and used to generate steam (see sidebar).

A prime example is a combined-cycle gas turbine power plant. Where a few years ago the hot exhaust from the gas turbine would have been blown to the atmosphere, now it is put through an HRSG and the steam runs a second turbine. This ability to recover what was regarded as waste heat contributes to the high efficiency of these generating units.

Steam is typically used to transfer heat to another fluid, such as a large-jacketed reactor or kettle used in a brewery (figure 2) or other food-and-beverage applications. Steam flows through passages and heats the product through the kettle walls or an internal coil. This can provide very even and carefully controlled heat, so the product is not burned. During the initial heating phase when the product is cold, steam condenses quickly, and the condensate collects in the lowest point. A steam trap separates condensate from the incoming steam and sends it back to the boiler as feedwater.

This is a critical point in the process and has a major effect on the efficiency of the kettle. To transfer all the heat possible from the steam, all of the steam should condense in the jacket. Early in the heating process, the temperature differential between the steam and product is at its greatest. The steam transfers its heat into the product quickly and condenses. If the process is aiming at the quickest heat up, this is the time when steam flow is highest. As the product temperature increases and heat transfer slows, steam flow has to be reduced or steam will be blown from the outlet, which wastes heat.

Condensate runs into a steam trap, which allows the liquid to escape and return to the boiler via a collection system, but the trap stops steam, trapping it in the jacket. A steam trap is actually a condensate separation device. It has an enormous effect on the efficiency of the application. If it does not remove condensate fast enough, the condensate backs up into the steam passages, which reduces heat transfer. If it allows steam to blow past, heat is wasted. If the steam trap is sized properly and uses an appropriate design for the application, its action should be automatic, provided it is functioning correctly.


 

Figure 2. Steam is a valuable way to heat products in the pharmaceutical and food and beverage industries.

Steam trap designs

The discussion so far suggests that all condensate needs to be removed from steam, but the situation is more nuanced. Small amounts of condensate in a high-pressure steam line will be at a high enough temperature that it will flash into steam if it reaches a point where the pressure drops. Condensate carries a great deal of heat itself, so removing it when it is not needed also wastes energy. Consequently, there are different steam trap designs that remove condensate under different circumstances. Again, there are many more in-depth resources available, but some types of steam traps only release condensate when its temperature falls below a specific threshold; whereas others are simply concerned with liquid volume. The application will dictate which is most appropriate.

Functionally, a steam trap is a valve that opens and closes automatically in response to its situation. All designs, therefore, have some moving parts and a seating surface. Thermodynamic traps are very simple with only a single moving part; whereas mechanical designs (e.g., float and inverted bucket) are more complex. Unfortunately, where there is a mechanism, there is an opportunity for malfunction, but these types of problems can be predicted.

Steam trap failure modes

Steam is not always clean. Although feedwater is heavily treated, it is still possible for scale, which can break free and be carried by the steam and condensate, to form in the system. Such particles have an uncanny ability to come to rest in problematic spots, such as valve seats or mechanisms. Similarly, if feedwater treatment chemicals get out of balance, excess corrosion can result. Operating conditions, such as water hammer and vibration, also take a toll on valves, fittings, and steam traps.

A steam trap can fail in one of two ways: it sticks open and releases steam, or it sticks closed and does not release anything. Inspectors on plant rounds checking traps generally classify them by diagnosis:

  • There is an obvious steam leak-a major mechanical failure.
  • The trap is too hot-it is the same temperature as the steam line, because it is releasing steam directly into the condensate line or to the atmosphere.
  • The trap is too cold-it is stuck closed, and no condensate is being released.
  • The trap is just right-it is releasing warm condensate.

Finding these "Goldilocks" units performing correctly, and tagging the bad actors for maintenance, requires an appropriate tool to evaluate temperature, such as an infrared viewing device. These can do the job, but a technician has to get to wherever the steam trap is and make the evaluation. Unless manual rounds by a very highly qualified and experienced technician happen regularly and frequently, one or many steam traps can malfunction for quite a while. A recent study suggests that 18 percent of steam traps in a large chemical manufacturing facility can fail in a given year, resulting in wasted energy costs up to $16,000 per trap.

One traditional approach to monitor a steam trap involves finding a way to mount a temperature sensor on the trap itself to measure the trap's condition. But this is an invasive solution, and the data it provides requires extensive interpretation and knowledge of what the just-right temperature should be under the operating conditions.


 

Figure 3. An acoustic transmitter mounts next to the steam trap on the pipe, so no shutdown is required for installation.

Hearing the solution

Most steam traps do not release condensate continuously. Although such situations are possible, under normal conditions and if sized correctly, all steam trap designs open intermittently and discharge condensate in slugs. The internal turbulence when this happens creates noise that transmits through the adjacent piping. Someone listening to a properly working steam trap should hear these periodic releases interrupting times of silence as appropriate amounts of condensate accumulate.

An acoustic transmitter mounted on the pipe adjacent to a steam trap (figure 3) can listen to the noise it makes. It is sensitive to ultrasonic frequencies, so it can hear the cycling, and an algorithm can be applied to learn the characteristic activity for each trap. Data can be sent from the transmitter via WirelessHART to a central data collection and analysis platform, where operators can see how the steam traps equipped with acoustic transmitters in all parts of the plant are performing.

Dashboards display (figure 4) which steam traps are working correctly and which are in one failure mode or the other. The software can estimate lost energy and resulting costs at any time. Maintenance can see at a glance which steam traps need attention, so they can plan activities appropriately and can predict and deal with small problems before they become serious issues.

Naturally the data may need some interpretation. For example, a steam trap reported as cold could be malfunctioning, or it could be cold because the equipment only operates intermittently and may simply be shut off. On the other hand, a steam trap attached to a process that runs continuously or at least regularly should develop characteristic discharge patterns. If these change, such as from a sudden increase in condensate volume, there may be some other cause for a deviation from the normal process operation.

Predictive maintenance can take many forms and provide many types of information. Just as measuring and watching process variables gives insight into what is happening with the process, the same applies to production assets. Steam traps are just one example. There are now many sensors and analysis tools available to use with a range of plant assets, such as pumps, heat exchangers, and pressure relief valves. When data is available, maintenance departments can find new ways to develop better predictive maintenance programs to treat problems sooner, or before they even fully surface. This saves maintenance costs, but more importantly reduces losses from unscheduled shutdowns, and often prevents serious incidents.


Figure 4. A dashboard supported by analytical software can capture and display performance and condition information for steam traps throughout a facility, enabling a predictive maintenance program.

Reader Feedback


We want to hear from you! Please send us your comments and questions about this topic to InTechmagazine@isa.org.



Like This Article?

Subscribe Now!

About The Authors


Tom Bass is the wireless product management director at Emerson Automation Solutions in Shakopee, Minn. Bass started at Emerson in 2004, holding many roles within the valve business before moving to wireless in 2016. Most recently, Bass led the wireless business development team with a focus on growing the Plantweb digital ecosystem. Currently, he is responsible for leading product management for wireless and manages the Plantweb pervasive sensing application portfolio. Bass has a BS in mechanical engineering from Iowa State University and an MBA from The University of Iowa.