By Harry Forbes, Senior Analyst, ARC Advisory Group
Two major applications for wireless sensing in manufacturing are process measurement and equipment condition monitoring (ECM). While wireless field transmitters for process measurement have been on the market for several years now, equipment condition monitoring systems have used wireless sensing only sparingly. However, in the last few months the dam seems to have broken, as several major suppliers have commercialized new wireless ECM offerings that had been in development for some time. Let’s examine this new technology that has been recently introduced and think through what its implications are for the power generation industry.
Figure 1. Emerson’s wireless vibration transmitter
In today’s electric power utilities, critical baseload units are often large coal-fired units sized at 500 MW and higher. Problems or failures of critical bearings or shafts in such a large turbine-generator can take down the whole unit, potentially for extended time periods. Utility engineers confronted with equipment condition questions today have some data at their disposal. Such large machines are equipped with wired sensors that feed a condition monitoring system. The fortunate engineer may have a long history of equipment condition data and analysis results. He or she even may be able to work with experts in the company to get more than one head working on an important question. If the engineer is truly fortunate, collaboration can extend to an equipment monitoring service provided either by the turbine manufacturer or by a third-party firm providing ECM services.
Get beyond the main turbine-generator, though, and the equipment condition data usually becomes spottier. Problems with equipment such as feed pumps, booster pumps, ID and FD fans, coal pulverizes, PA fans and environmental systems can all cause unit outages or lost generating capacity though unit runbacks or firing limits. But how many plants are equipped with condition monitoring systems for all of this equipment? Very few, if any. The cost of installing and maintaining these systems in the harsh plant environment is viewed as too high and the analysis required too labor-intensive to merit installing permanent wired sensors. Utility engineers must rely on manual and intermittent data collections and hope for the best.
An Established Practice
Condition monitoring for rotating equipment is a well-established field that is decades old. The most basic practice is to take vibration measurements of machine shafts or bearing casings. The magnitude and phase relationships of this vibration are compared with historical baseline values to infer changes in equipment condition. In addition, practitioners examine the component frequencies present in the signal from the sensor. These may correspond to particular mechanical components of the machine. The objective is to identify the signatures of impending problems (and perhaps their root cause) in time to enable preventative maintenance action rather than running the equipment until a failure occurs.
Other sensor inputs besides vibration are also used to determine equipment condition. The other major data inputs are lube oil analysis, motor electric current, infrared thermography and ultrasonic measurements.
From an economic standpoint, continuous ECM has penetrated only a small fraction of the rotating shafts found in plants. The high cost of permanently mounted sensors and continuous monitoring systems has limited its penetration to only the most critical 1 percent to 5 percent of installed rotating equipment. Discrete manufacturing operations, which usually have many more (though smaller) rotating shafts are even less likely to have installed continuous monitoring. Instead, common practice is to use hand-held systems to capture periodic vibration signatures, which (hopefully) are quickly analyzed for problem indications.
New Wireless ECM Systems
Several suppliers have recently introduced new products intended to improve this situation. The rush began last September when Emerson Process Management introduced a wireless vibration transmitter at its Emerson Exchange user event. Emerson had been working on small-footprint condition monitoring solutions for some time with its CSI products and wireless sensing was a natural extension in that direction.
A few weeks later, GE Energy introduced wireless condition monitoring (the Bently Nevada wSIM system) at the 2007 ISA Expo. GE added the extra twist that sensors could be powered either by batteries or alternatively by vibration energy harvesters built into the sensor package. The use of vibration energy harvesting had been mentioned as one of GE’s visions for its wireless sensor research (which received partial sponsorship from the U.S. Department of Energy). At the same time, GE also announced that these new wireless vibration sensors had been tested and deployed at the Ormen Lange gas project in Norway, work done in partnership with Shell Global Solutions. Shell is keeping mum about the actual number of machines being wirelessly monitored at Ormen Lange. But Shell’s objective in using wireless sensors obviously was to expand the coverage of condition monitoring far beyond what could be economically justified using wired sensors and systems.
Figure 2. ABB’s small form-factor wireless vibration sensor for motors
GE was also mum about its technology partners for wSIM. However, there are not a large number of companies offering ATEX-certified vibration micro-generators, so ARC would give odds that the energy harvesting technology used in wSIM originated with PMG Perpetuum, a UK venture that was founded to develop such technology. In April 2007, GE Sensing announced a partnership with Dust Networks for wireless sensing. Neither party would provide confirmation, however, of what wireless networking was embedded in wSIM.
The deployment of micro-generator energy harvesting for the first time in a wireless sensing application is an important development for condition monitoring. Besides being a natural fit, this combination enables simplified retrofit of new ECM applications on existing machines. GE’s line featured a magnetic sensor mount so that sensor installation would not involve any drilling. The battery-free design was likely a big plus in the view of Shell Global Solutions. In past interviews with ARC, Shell was reticent about deploying large numbers of sensors with batteries in their plants. The printable parts of past ARC interviews with Shell on this subject have included statements such as “over my dead body”. The ability to deploy this application without batteries accounts for the big change in Shell’s attitude.
Figure 3. A vibration-powered generator(right) for wireless sensing applications
A few weeks later, Honeywell introduced a wireless Equipment Health Monitor (EHM) for its OneWireless product portfolio. This offering used sets of wireless sensors, which are networked to a Honeywell OneWireless mesh network. The analytical capability is provided by a partnership between Honeywell and SKF, the well-known ECM supplier. Honeywell also assembled a “starter kit” for its solution, consisting of the sensors, gateways and software needed to monitor four machines. Besides adding wireless sensing, the EHM package enabled Honeywell to promote the IEEE 802.11 compatibility of its OneWireless infrastructure and the incremental value of that infrastructure once installed.
Last but not least, ABB showed a new line of wireless condition monitoring sensors at this year’s Hannover Fair. These sensors were developed for the offshore oil and gas industry as part of an R&D program that included major offshore oil producers BP and StatoilHydro as sponsors.
Dubbed the “Wivib”, each wireless sensor has an accelerometer, a temperature sensor and a radio transmitter. These combination temperature and vibration sensors are quite compact, being packaged in a unit 10 centimeters long which is mounted directly on the motor. They use a battery along with some local digital signal processing to conserve battery power. The device will use the WirelessHART protocol and its block data transfer capability to provide data to a centralized ECM application.
Thus, within the space of a few months several suppliers have now wireless-enabled their ECM offerings. Does that do it? What more should utilities be looking for as these types of systems reach the market?
Besides getting data from sensors, effective condition monitoring requires analytics. When data sets arrive once a month from manual data collection, the processes required to manage the data and analytics can afford to be less than fully automated. However, with a much larger number of machines providing several equipment data sets each day, a deluge of data is created. This data deluge must be managed and maintained in order to extract the potentially valuable information it contains.
The most valuable ECM solutions will enable utilities to spend less time analyzing and managing data and more time focusing on exceptions and important findings that automated analytics uncover. In addition, collaboration and data sharing among plant personnel, in-house ECM experts, equipment suppliers and service firms will likely become an important future activity. The expansion of ECM brought about by wireless may enable structural changes in the business as well, since it may become possible and economical for many more people to collaborate and deliver value in this domain.
ECM in the Future Power Industry
Today’s utility engineer may be responsible for performance of a 500 MW or 750 MW unit, where problems with any of roughly 50 rotating shafts located somewhere on the unit can affect unit availability. Tomorrow’s utility engineer is going to confront an even greater challenge. For one thing, the low-cost baseload generation of future power markets will include far greater amounts of wind power. In the future, engineers will have to monitor the condition of a far larger number of machines, and the machines will not all be located within a few minutes walking distance from their desks.
Following the lead of Europe and with steady financial incentives in place, North American utilities are now building wind generation on a much larger scale. In 2007 over 5,500 MW of new wind generating capacity was installed in North America, nearly twice the amount installed just one year earlier. New wind farms are being developed as quickly as possible, stretching the ability of wind turbine suppliers to deliver equipment and of power grids to accommodate them.
As regional markets for power generation develop, the financial opportunities for wind power improve. After all. since these units often provide power on a “spot” basis, they can take advantage of the higher spot pricesprovided they can operate at capacity. Today’s higher fuel prices for all types of fossil-fired power generation improve the attractiveness of wind power even further.
Wind power growth will present major challenges for equipment condition monitoring. The size of today’s utility wind turbine is 3 MW to 5 MW, less than 1 percent the size of a baseload fossil unit. Utility-sized wind farms, then, consist of dozens or even hundreds of units. These units are located in remote areas that optimize high prevailing wind rather than easy access for engineers and maintenance crews. While many of these units will be of identical design, European experience indicates that over time large wind farms will expand to include several different turbine models and also different turbine suppliers.
Large wind farm operators will face the challenge of dealing with multiple generating equipment suppliers, which can create barriers to globally accessing real-time equipment condition information. Especially in a wind farm, this information is essential for improving operational performance. To succeed, utilities must be able to manage wide-area production networks. They will use these networks to connect their own experts and their partners with the large numbers of remote generating units. They must detect equipment condition degradation and perform the required analysis remotely.
Increased low-cost generation will be the most important benefit of this effort, but not the only one. The same data and infrastructure can be used to monitor other unit performance metrics. This can lead to better understanding of unit performance, and to better decisions about future equipment selection and location.
Operating and maintaining wind farms efficiently will be done best by larger operators who can exploit their operational and maintenance know-how and technology on a sufficiently large scale. The global trend in wind power is toward this model.
ARC believes that the cost, flexibility and other advantages of new wireless sensing technologies will expand the coverage of equipment condition monitoring to a far higher fraction of critical equipment in the power generation industry. At the same time, the power grid will be developing into a much more complex system that includes many thousands of new and much smaller units along with traditional large generating units. Tomorrow’s baseload generating capacity will include very large and very small units alike. This will make condition monitoring a greater challenge for utilities, who will rely on a much larger number of machines to deliver power (and profits) during peak hours.
What recommendations should a utility follow?
Figure 4. A wireless sensor packaged with a vibration generator
First, utilities should include automated analytics and diagnosis (not just automated data capture) as factors in their evaluation of new ECM offerings.
Second, they should use wireless sensing systems to expand both internal and external collaboration in the ECM area.
Third, greater levels of collaboration also will be essential for condition monitoring in the future, so start building this capability now.
And, finally, wireless sensing and vibration energy harvesting make a great combination. Utilities should expect their suppliers to deliver products that leverage this match.
Author: Harry Forbes is a senior analyst with ARC Advisory Group. He covers industrial networking technology and also covers power generation, power T&D and energy efficiency. Prior to joining ARC, he worked for Invensys where he served power industry customers in a number of engineering, sales and marketing roles. Forbes’s experience also includes work for the Detroit Edison Co. (now DTE Energy), where he was a Results Engineer and a control systems engineer at a nuclear generating station. He holds a BSEE degree from Tufts University and an MBA from the Ross School of Business at the University of Michigan.