Part One Part Two

Third in a series. Advances in wireless technologies can help address many common failure modes in mid- to low-level criticality assets and eliminate wiring costs for a range of asset types.

Scanning Systems and Wireless Scanning Systems

Where appropriate, wireless solutions can replace wired scanning systems, primarily where the failure mode detection methods match the technological capabilities of the wireless systems, such as casing vibration (seismic) and temperature data. Wireless scanning systems are only appropriate for essential measurements with half P-F cycles of greater than two hours, while scanning systems can detect failure modes with P-F failure modes of less than 30 minutes. A careful review of asset failure modes, failure cycles, detectability, consequence of an undetected failure and work process upon detection of a failure all need to be considered when deploying a particular technology solution. Failure modes with little warning time or that require automated relays for shutdown are not suitable for wireless applications.

Wireless Mesh Network Components

Wireless mesh topology can complement existing condition monitoring platforms on rotating and reciprocating machines in power generation, oil and gas (upstream and downstream), chemical and other process industries. Various components are used in the wireless mesh topology for condition monitoring.

Low powered casing vibration and temperature sensors are often used on these type assets, although other process variable measurements can also be captured using wireless technology. The sensors typically connect to the wireless sensor node (WSN) via a cable. The WSN is installed on or near to the monitored machine, and can connect to multiple sensors. The WSN powers the sensor, collects the data  and may do some minimal signal conditioning and processing. It also contains a radio transmitter/receiver that communicates on the wireless network. Each WSN must be in 50 meters line of sight distance of at least two or three other nodes to form a robust wireless network.

Repeater nodes are available for locations where the WSN density is insufficient or where there are long distances between groups of WSNs. Repeater nodes contain the radio-transmitter device, but do not contain the electronics needed to work with sensors. As such, they do not need to be near machines and are ideally placed in accessible locations where they can maximize the line of sight between nodes, such as in areas with high elevation (eliminating ground-based line of sight interference) or near doorways or building corners.

The wireless mesh network uses a combination of spread spectrum and frequency hopping technologies to provide signal reliability, including Direct-Sequence Spread Spectrum (DSSS) and/or Frequency Hopping Spread Spectrum (FHSS) technologies. Each transmission can transmit among 16 different available frequencies, which allows the network to limit any interference from wireless transmitters that are not part of the mesh network.

Mesh networks are simple to organize and install compared to wired scanning systems. After assigning the nodes to a network gateway, the network is self-organizing. The WSNs will form their own primary and alternate communications routes, depending on other nodes in their radio line of sight. Each node is automatically assigned an IP address. After installation, it may take a couple of hours for the network to fully organize, which is still less time than to install and configure a wired scanning system.

Because of the number of interlinking possible communications paths between nodes, mesh networks are self-healing; if one node no longer works or becomes temporarily blocked, the nodes will automatically switch to alternate nodes and continue to operate, providing a high degree of reliability. Each node only turns on and gathers data or communicates as necessary, reducing the power consumption per node to a fraction of that for continuously powered networks. They are scalable and can be changed to suit the plant needs.

Mesh networks can change their scalability and have tremendous flexibility. If a plant wants to expand its wireless network, it only needs to commission the new nodes and position them as needed in the plant. Similarly, a plant can move WSNs from one location to another. This would only cause a temporary loss in communications, as the WSN would identify that it lost communication with neighboring nodes and would automatically detect new neighboring nodes to rejoin the network.

The network gateway links the wireless network to the Ethernet and manages the network configuration and timing. While each network gateway can manage more than 200 wireless mesh nodes, additional gateways may be required for extensive wireless networks or to manage networks in separate areas of the plant. Each network gateway manages its network with unique join or authentication keys to ensure that only its nodes can access the network, with encrypted session keys to ensure that all transmitted messages are confidential. Additional gateways (to run separate networks) may be necessary if the distance of any node requires more than five "hops" to other nodes to send a message to the gateway.

Most gateways can connect to external antennas, which can then be placed in a location more desirable to maximize connectivity to nearby nodes. The best practice is to have redundant network gateways for each mesh network to enable the backup to continue to run the WSN in the event of a network gateway failure.

The mesh network system will ideally be compatible with the plant's CM software platform. This requires less training, and the similar capabilities between wireless and other platforms make it easier to configure similar alarm setpoints and align maintenance strategies between similar assets. It enables the plant to easily upgrade from a wireless scanning system to a continuous monitoring platform.

The CM platform must be able to collect data through a number of separate network gateways and correlate this data with other continuous, scanning and portable data collection systems. The CM software should be capable of displaying process variable data, temperatures and static and dynamic vibration data. A static value is a measurement that can be depicted by a discrete value, such as direct acceleration, rotor region (1X) acceleration, prime spike acceleration, direct velocity and direct enveloped acceleration. Enveloped acceleration is suited for early detection of roller element bearing failures, while direct and rotor region vibration measurements are suited for detecting other faults, such as imbalance, misalignment, fan blade and pump vane problems and cavitation.

Dynamic vibration waveforms can be captured for each channel. Because of the low power and limited bandwidth of the wireless mesh network, waveforms should only be collected once per day per sensor. An additional dynamic waveform should be captured in an alarm event. The primary purpose of collecting dynamic waveforms is to help diagnose machinery problems due to an alarm event.

For essential measurements on roller element bearing machines, collecting vibration data in one vibration plane (X or Y) is all that is necessary. Machine speed for constant speed machines can be accurately predicted based on the rated speed of the motor driving the process. While this does not provide true phase information, it is helpful in determining how closely the vibration data relates to the machine's speed.

Using "Green" Power

Improvements in battery technology and alternative power technologies continue to make wireless technology attractive. Extending battery life to more than a year, however, remains a challenge. To maximize and conserve battery life, scan times should be limited to once every 15 minutes for temperature and static (direct) vibration data for problem machines, and once every two to four hours for most machines. Dynamic vibration waveforms should be collected only once per day to conserve battery power and prevent bandwidth issues. Wireless nodes should provide a warning as battery life starts to deteriorate.

Alternative energy sources can be part of the solution. In areas with abundant sunlight, solar cells can provide energy almost year round. Energy harvesting technologies are improving, allowing each node to continue to operate on older batteries for extended periods. On rotating and reciprocating machines, energy harvesting devices can convert machinery casing vibration into power. Both solar power and energy harvester can provide the low power for nodes and repeaters, either directly or by recharging batteries, significantly extending battery life.

Combining CM Methodologies         

Wireless CM is not meant to replace existing CM platforms; rather, it is a complementary measurement suited for specific essential measurements. It should not replace highly critical or critical measurements where continuous monitoring is necessary due to the high consequence of failure. It can help supplement critical, online condition monitoring systems on highly critical machines; for example, temperature measurements previously taken by portable data collectors only after a failure seemed likely. In most cases, however, wireless CM can be used to supplement (raise the flag) and/or replace some periodic portable data collection routes to capture vibration and temperature data more frequently and safely. If the wireless system detects an anomaly, personnel can be sent to take additional data using portable instruments to help diagnose the problem.

Machinery Applications

Pumps

Pumps in tank farms are typically monitored with a walk-around portable data collector system. Scanning systems are not used since the distance between the control rooms and the tank farm would require long wiring runs. The consequence of failure, however, can be extensive. Failures can result in safety and leakage of process fluids. An environmental spill, fire or explosion may occur, depending on the process fluid.

Typical Pump Wireless Platform:

  • Two to three nodes per pump
  • Accelerometers on each motor and pump bearing
  • Thermocouples at each bearing
  • Thermocouple at high pressure seal face to detect temperature changes in the seal
  • Two redundant repeater nodes to transmit the signal from the tank farm to the network gateway

Fans

Fans can range in criticality from highly critical to low level criticality and are often good candidates for essential measurements. Many fans are monitored with either earthquake switches (which provide limited protection and no vibration data) or a portable data collector system. Many fans are difficult to access and are noisy when operating, causing potential safety and health problems for personnel relying on route-based collection methodology. Wireless condition monitoring platforms capture more frequent vibration and temperature data in a safe, cost-effective manner for these essential assets and can be used as raise the flag sets.

Cooling tower (CT) fans are mainly used in the power industry to cool process water. They have previously used earthquake vibration switches to shut down the cooling tower fan during high vibration events. However, earthquake switches are typically older and may not work. During operation, periodic portable data collector measurements can only occur on the motor and are typically not used on the gearbox for safety reasons.

Many CT fan owners have switched to a continuous monitoring platform. Continuous monitoring is the preferred solution in most cases, depending on the results of the criticality analysis. In some applications, however, some cooling tower fans are not considered critical assets (when more cooling assets are available and there are a number of hours between high blade vibrations and total blade failure). For these cases, a scanning system would be an improvement over the current monitoring system, and wireless condition monitoring would provide an effective, low-cost solution.

Fan Wireless Platform:

  • Two interface module nodes per fan
  • Accelerometers on each gearbox in line with output shaft bearing to detect bearing and rotor-related problems
  • Thermocouples measuring gearbox bearing oil and motor bearing temperatures
  • Accelerometer on motor bearings
  • Accelerometer perpendicular to output shaft for rotor- and fan-related problems
  • Two repeater nodes to transmit the signal from cooling tower to the network gateway

Similar to CT fans, fin fans cool liquids or gases. They are typically smaller than CT fans and are used as both forced draft (FD) and induction draft (ID) fans. Fin fans are more common in refineries and the petrochemical industry, but are occasionally used in place of cooling tower fans in power generation. When fin fans replace CT fans, more fin fans are needed in the same size power generation facility to cool water than with cooling tower fans. They have the advantage, however, of requiring a smaller area for installation.

In addition, the requirement for more fin fans makes them less critical than cooling tower fans; fin fans are usually less critical and rarely have a continuous monitoring system.

Pumps & Systems, September 2009