Pervasive sensing solves Saltigo steam problems
In operation, steam traps don’t reveal much about themselves; software helps them speak a language relevant to plant productivity.
Pervasive-sensing technology means to monitor everything, everywhere. Wireless sensors make it economically feasible to monitor devices that previously were too remote or too expensive to monitor with wired instruments. However, when a plant monitors everything, it generates ubiquitous and huge amounts of real-time data that must be dealt with expeditiously.
Combining pervasive sensing with PC-based specialty software allows a plant to obtain data, analyze it for equipment problems, diagnose problems, and inform maintenance personnel, all without involving the main control system. This is sometimes called "computing at the edge." And that’s exactly what Saltigo, a chemical plant in Germany, accomplished with its steam traps.
Saltigo GmbH in Leverkusen and Dormagen, Germany, (Figure 1) is one of the leading chemical manufacturing companies in the field of exclusive synthesis. It produces pharmaceuticals, fine chemicals, agrochemicals, and products for crop protection.
Saltigo determined that defective steam traps were causing loss of steam and inefficient heat transfer, and therefore wasting energy. If they could identify failed steam traps as early as possible, they would save a considerable amount of energy.
The plant had been using WirelessHART transmitters for several years to monitor pumps, heat exchangers and other process equipment in three of its plants. Because the wireless infrastructure was already in place, it was no problem to install Rosemount 708 wireless steam trap monitors on critical steam traps. The results from the first installation were so good, they recently ordered more of the transmitters to monitor the remainder of their steam traps.
Surveying the situation
Saltigo had sent technicians into the plants to perform tests and inspect the steam traps, but the manual rounds weren’t always sufficient to find traps that failed open, failed closed, or were leaking. Sometimes the only visual clue of a leak or failure was a vent with steam escaping, but the vent might be attached to several steam traps, making it very difficult to identify the leaking trap.
In addition to visual inspections, Saltigo also used acoustic diagnostic tools and temperature sensors to identify and locate failed traps. One problem had been that a steam trap had to be in operation at the time of a manual test. A typical test took only a few minutes, and if a failed trap wasn’t leaking or failing during the short test because it wasn’t in operation, or due to some other reason, the technician would simply move on to the next steam trap.
Manual steam trap testing can be difficult because steam traps often are in areas that are difficult to reach (Figure 2), and testing can be hazardous to the technician because of high temperatures.
With many hundreds of traps in three areas at the Leverkusen plant, it was difficult to test them all on a regular basis. They might test each trap once a year. Because of this, Saltigo was only able to identify traps that had failed catastrophically-that is, those venting steam continuously, shutting down processes, or failing to provide enough steam to bring a process up to temperature.
Because Identifying failed steam traps was so difficult using various manual inspection techniques, the plant knew it needed a better method.
Saltigo first conducted a study to identify high risk and critical steam traps. This included steam traps that were known to fail on a regular basis, and traps that were critical to the successful operation of a process.
They already had several Emerson WirelessHART pressure, flow, temperature and vibration transmitters installed in the plants, so Saltigo asked Emerson Process to do a site survey to assess where they would need to place antennas and gateways to accommodate wireless steam trap monitors.
Installing wireless transmitters
Acting on the study results, Saltigo installed Rosemount 708 wireless steam trap monitoring transmitters (Figure 3) and three gateways—one in each of three plant areas. The gateways connect to the transmitters through the mesh network, and they connect to the control system via hard wiring.
Installation by supplier personnel and Saltigo technicians consisted of mounting each transmitter on a pipe, manifold, or relief valve. No cutting or entry into the piping was required as the acoustic transmitters are non-contact instruments and are easy to clamp on the outside of the steam pipe in front of the steam trap. Because the transmitters are wireless, no wiring, cable, conduit, or other infrastructure was needed that would have been required with conventional wired transmitters. And because the transmitters are battery-powered, no power wiring was required, resulting in additional savings.
The transmitters are certified for use in classified areas, so they are perfectly safe to use in hazardous areas, without any need for protective enclosures or other protection methods (Figure 4).
Installation took about 10 minutes per transmitter. Once the team mounted each transmitter, they installed the battery, powered up the transmitter, then entered the Network ID and Join Key. In about five minutes, the transmitter had joined the WirelessHART mesh network, was recognized by the gateway, and was transmitting data. As the initial study predicted, they experienced no obstructions or distance problems, a fact revealed after testing for connectivity and performance.
The team also installed Emerson SteamLogic software on a PC. The gateways connect to the PC via an Ethernet cable. The SteamLogic software (Figure 5) analyzes real-time data from the steam trap monitors. The Rosemount 708s measure the ultrasonic acoustic behavior and temperature of steam traps, and the software uses this data to identify existing and potential problems.
The left side of Figure 5 shows a low acoustic level with some peaks where the steam trap opens to let the condensate out. When the acoustic level is high, such as at the right side of the picture, the steam trap doesn’t close and is losing steam.
The transmitters and software were initially installed for a three-month test. The system immediately found several failed steam traps, all of which were repaired or replaced. During the three-month test the system also detected undersized traps, which were attended to by the maintenance department.
By repairing or replacing failed steam traps, the three plant areas immediately began to see substantial reductions in energy costs. Failed traps were no longer venting valuable steam, resulting in lower energy consumption to produce steam. Saltigo also reduced the number of process shutdowns because of steam-trap failures, and eliminated the need for maintenance technicians to make regular rounds, resulting in further substantial savings.
Now, when Saltigo sends a technician to a plant to check a critical steam trap, he or she knows exactly which trap to service, and what the software has detected. This not only guides the technician to the exact trap needing attention, but also provides information leading to quicker maintenance and repairs.
Saltigo’s experience with the steam-trap monitoring system was so good, they ordered additional transmitters for the remainder of the steam traps in the plant.
Saltigo also recognized that trap failures could be predicted early by noisy stats, before they turned into a full leaking failure. That gave Saltigo the chance to improve its preventive maintenance and change traps before they negatively impacted production.
Ralf Kueper is business development manager for wireless and pervasive sensing at Emerson Process Management, Haan, Germany.
For more information, go to www.emerson.com.