Wireless Sensor Network Selection for a Plant

[Cross Pot from Authors Blog ]

Wireless Network (WSN) technologies are not the same. There are very significant differences. The wrong technology for the plant will be unreliable; disruptive for the data users. Range and sensors per gateway may be the least important criteria in a plant. WSN is a key enabler for Digital Transformation (DX) including . And you can’t invest in separate WSN infrastructure for each use case like corrosion, vibration, and leaks etc. So how do you tackle a plant full of steel, with all kinds of advanced sensors from multiple automation vendors, integrating with existing automation, in an endless parade of new models and versions, with minimal battery replacement, learning the tools, effectively utilizing intelligence in the sensor, and avoiding incompatibility? Here are my personal thoughts based on technology requirements in the NAMUR NE124 Wireless Automation Requirements recommendation:

Risk of Wrong WSN Technology

If you are not careful you may end up with a WSN which has many limitations:

No alt text provided for this image

With the right WSN technology you can avoid these pitfalls.

The plant IT team may not be familiar with these issues since these are mostly related to plant environment, plant use cases, and integration with I&C systems. The I&C team must take the lead to guide the WSN selection.

WSN technologies have many similarities, but also important differences. For instance, several WSN are built upon IEEE 802.15.4 radio but apply it differently so the WSN technologies have very different characteristics.

Automatic data collection with advanced sensors is the foundation of digitization and Industrie 4.0; moving from manual and paper-based ways of working, to new automatic, digital, software-based, and data-driven ways of working; digitalization. Most plants when starting on their journey of digital transformation discovered that the process data they have collected over the years does not help them predict problems with corrosion, vibration, and leaks etc. To be predictive they need additional sensors throughout the plant and realize WSN is the most practical way to add sensors in an existing plant. The I&C department is responsible for deploying the WSN infrastructure in the plant making it openly available to every operational department in the plant. Every department can now deploy wireless sensors without participating in the costs of the infrastructure.

The WSN infrastructure includes the wireless gateway which communicates with the wireless sensors, a backhaul network to the automation system like a historian or DCS, and gateway power supply cabling. There are variations between WSN technologies. Some WSN technologies require many more gateways than others, and some WSN technologies divide the WSN up into smaller sub-networks using intermediate backbone routers with associated backbone network cabling and LAN switches in the field.

Availability and Reliability

You need 99% reliability or better for the WSN in your plant to make sure the data is there when the users around the plant need it. This number is not stated in NAMUR NE124, but based on experience from more than 12 years of WSN, 99% is what you need. But most WSN technologies do not state any reliability, there is no commitment. This is so because the reliability of most WSN technologies in a plant environment is very much lower than 99%.

Other WSN technologies do not commit any level of reliability

A plant is a challenging environment for wireless because radio frequency (RF) signals cannot penetrate steel, and a plant is full of steel: pipe racks, structural steel, and vessels. Even metal furnaces and storage tanks the size of multistorey buildings.

Moreover, there is a constantly changing RF environment in the plant as trucks, cranes, other heavy vehicles, ISO containers, and ISO tanks move in, scaffolding goes up, causing radio propagation failure, or when other wireless are brought in. That is, a WSN can be made to work initially, but as the RF environment changes over time; with the wrong WSN technology, part of it may suddenly fail.

Star Topology Limitations

Because radio waves do not penetrate steel, traditional radio communication requires a clear “line-of-sight” between communicating devices. That is, it must be possible to see the gateway from the sensor location for wireless to work. Most WSN technologies use star topology, meaning the gateway communicates directly with every sensor point-to-point. Many wireless technologies were created for open spaces like homes and offices or long range in large geographical areas like cities and agriculture where star topology works fine such as for smart grid water and power meters, smart city streetlights and parking meters. However, star topology requires the gateway to “see” every sensor directly which is not practical for sensors in a plant full of steel. In a plant the “line-of-sight” is limited, perhaps 50 m or less. You can’t see the gateway from the location of most sensors due to all the metal obstructions, and therefore radio waves also cannot reliably reach the sensor directly. Thus line-of-sight from the gateway to every sensor cannot be a requirement for the WSN technology used in a plant. Some star topology based WSN even suggest installing the sensor in a vacant space where there are no structures or other obstacles around – but that’s not practical since the sensor must be installed at or about the point of measurement. Further they suggest avoiding locations near the ground, instead installing the sensor far above the ground – which again is not practical since many equipment, like pumps, are located on the ground. Due to the limitations of line-of-sight, star topology means that many more WSNs, each with its own gateway, are required to cover an area of the plant. A complex infrastructure which would be very costly and disruptive to deploy in an existing plant.

No alt text provided for this image

Some WSN technologies use star topology with a backbone to create many sub-networks, each with an intermediate backbone router, to cover an area of the plant. This requires a backbone network and power in the hazardous area. A complex infrastructure which would be very costly and disruptive to deploy in an existing plant.

No alt text provided for this image

Most sensors in a plant are installed low; on pumps, motors, compressors, steam traps, and safety showers close to the ground. Because they are low, they are obstructed by multiple stories of metal also from above. So even a gateway mounted high up, like a mobile phone mast (cell tower), cannot connect with the ground-level sensors vertically, or sensors behind a tall furnace or column. Moreover, with the antenna installed in an elevated location, lightning-related risks increase. Sure, some sensors like those on relief valves are installed high, but you can’t have one WSN technology for sensors on relief valves and another WSN for all the others. You want a single WSN that covers all wireless sensors.

No alt text provided for this image

Lastly, with a point-to-point star topology the communication may initially work at the time of deployment, but when the RF environment changes, part of the star topology suddenly stops working. This explains why star topology WSN do not state reliability in a plant full of steel. And when a weak spot is identified in a star topology, an additional gateway or backbone router and associated cabling must be deployed in the vicinity of the sensors. Star topology is not the right WSN solution in a plant.

Mesh Topology

In a plant you need a WSN technology which does not require direct line-of-sight for every sensor. WirelessHART solves this using mesh topology so every sensor doesn’t have to “see” the gateway directly. A sensor which does not see the gateway directly instead communicates with a neighboring sensor which relays communication to the next sensor and so on in up to 7 “hops” or more until it reaches the gateway. This means the data gets routed around metal obstacles. Therefore a plant full of steel is not a problem for the WirelessHART technology.

No alt text provided for this image

If there is a weak spot in the mesh topology, drop in a repeater, no need to install an additional gateway or backbone router, backbone networking, or power cabling.

In a mesh topology. each sensor doesn’t just talk to one neighboring sensor. It talks to many. This forms a mesh of connections between sensors. This means there are multiple paths from a sensor to the gateway. That is, there is data path redundancy. This multiple data-path redundancy is key to data availability and reliability.

You don’t have to manually configure the communication paths between sensors. WirelessHART is self-organizing. When the gateway is turned on, the mesh network is formed automatically. Moreover, the self-organizing mesh topology adapts dynamically. If a sensor is moved, removed, runs out of battery, fails, or if an obstruction like a container arrives blocking the RF signal along one of the paths, the system dynamically choses one of the other paths rerouting the data. This is how WirelessHART achieves 99% reliability even in a plant full of steel. But, it is important to follow the WirelessHART and installation rules. The reliability is easily verified from the communication statistics in the gateway where you can also see signal strengths.

No alt text provided for this image

The more sensors there are in a mesh network, the more possible paths there are between these sensors and therefore the more robust the mesh network becomes. That is, because the WirelessHART network is shared between corrosion, erosion, vibration, acoustic noise, pressure, temperature, level, flow, position, and other sensors for various use cases you get a strong mesh of connections.

Other WSN technologies without mesh topology or which are not self-organization cannot achieve the same reliability.

Not Maximum Range

The most interesting fact is that NE124 does NOT specify range from sensor to gateway or how many sensors should be handled per gateway as a requirement; because it doesn’t make sense in a plant full of steel. Because radio waves do not penetrate steel there is no long range or wide areas in a plant. No radio waves can penetrate steel. It doesn’t matter which wireless technology it is. No wireless technology transmits through steel. It’s physics. Therefore, the real range of wireless in a plant is only as far as the next metal obstruction; and metal is everywhere in a plant. Thus “long range” and “Wide Area Network” (WAN) has no meaning in a plant full of steel. One WSN vendor claims longer range than the next but performance of a mile or more in an open space without obstruction using a high gain antenna doesn’t matter in a plant. Due to all the steel the range between any two devices for any wireless technology inside a plant is often only 50 m or less.

If vendor claims 1.6 km range in a plant full of steel you know they are not experienced

And on that same note, a gateway handling 2,000 sensors also has no meaning inside a plant because there aren’t that many sensors in range anyway due to all the steel.

Since the range in a plant full of steel is limited, you will need to install the gateway close to the process area. In many industries the process area is a hazardous area so you will need a zone 2 or zone 1 hazardous area certified gateway.

Although metal obstructions limit the distance of individual hops, multiple hops between WirelessHART sensors add up from sensor to sensor totaling long distance from sensor to gateway even in a plant full of steel. So when comparing WSN technologies, ask not what is the maximum range from gateway to sensor, instead ask what is the maximum number of hops in one communication path. Point-to-point topology is a single hop. Some WSN mesh topologies are limited to 2-4 hops. WirelessHART can do 7 hops or more. This capability allows WirelessHART to reach sensors deeper inside plant units than other WSN technologies. As a result, the multi-hop capability gives WirelessHART the longest effective range and greatest coverage in a plant.

In a plant, range may be the least important criteria, what matters in a plant is the number of “hops”

Real-Time Capability

Wireless sensors are mostly used in non-process applications which are not time-critical such as for vibration, corrosion, leak detection, and manual valve position monitoring. DX use cases are very much about fully instrumented equipment. For these applications an update period of 1 minute, 1 hour, or even 1 day are enough. But in every plant there are a few applications that require a faster update period such as 1 or a few seconds for process pressure or flow. However, it is important to select update period thoughtfully. Do not select unnecessarily fast update period because it will drain the battery sooner. This is true for any WSN technology. Most wireless sensors are set to 1 min update period to enjoy multi-year battery life. Vibration may be set to 1 hour. I&C engineers know what update period is required depending on the use case. Apply Management of Change (MoC) procedures if you want to change the update period later. Also, upload of large data sets like vibration spectrum/waveform should also be done thoughtfully.

Many other WSN technologies only support update period of a minute or longer meaning there are some applications they cannot handle, so you must wire those sensors instead.

WirelessHART solves this with real-time capability with update periods of 1 second (theoretically as fast as 100 ms but not yet implemented in products) in some sensors such as pressure and flow. So if the use case really requires such a fast update period, it is possible. The result is real-time updates. However, do not select unnecessarily fast update period as explained above. And, you cannot have too many sensors with fast update period on the network. The latency can be checked from the gateway.

WSN use security mechanisms like those used for Wi-Fi. This includes encryption, authentication, verification, key management, and sequence number. Authentication means for the wireless sensor to join the WSN and connect to the gateway, the wireless sensor must be “provisioned” (configured in advance) as part of commissioning with the network identifier (like Wi-Fi SSID) and a secret key (password) to join the network. To make “provisioning” easy and secure, this configuration cannot be done over the WSN. The network ID and join key is configured “out-of-band” meaning by other means. This is like the wired “console port” on Wi-Fi and other wireless telecommunication equipment.

For several WSN technologies the commissioning process; to configure the network ID and key is very time-consuming involving multiple steps and special tools. Even for the same WSN technology, the tools and procedures differ from one sensor manufacturer to the next. For instance, for one WSN technology, after configuring network ID and join key from a laptop using an infrared (IR) adapter that must be aimed precisely, you must generate a provisioning information file for each sensor commissioned, copy the files, bring them to the wireless system management software, load the files onto the computer, and then download into the WSN. And keep the files. Another WSN technology requires you to use an app on a phone to write the gateway tag, network ID (gateway MAC address from label on gateway), and join key into a key card. Then you again use a phone to read that network information from the key card into that phone, and then use that phone to write the network information into each sensor using NFC. Other solutions include scanning QR code on each sensor to get its MAC address and then write it into the network management software. For the initial installation the system vendor will do this for you, but for future additions and replacement you have to do it yourself so it is important this is easy.

WirelessHART solves this by using the same HART field communicator and HART modem used for 4-20 mA/HART instruments which the plant already has. Often many sets. Instrument technicians are already familiar with these HART tools. The result is sensors easy to commission with security credentials. Other WSN technologies cannot use these tools.

Coexistence

WSN operate in the unlicensed frequency bands such as the ISM bands just like Wi-Fi and many other wireless technologies. The 2.4 GHz band is most common as it is accepted globally. Just like Wi-Fi, WSN technologies use spread spectrum modulation to enable coexistence with the other wireless technologies by not interfering with the others and be more resistant to interference from them.

WirelessHART is based on the IEEE 802.15.4 radio standard which uses Direct-Sequence Spread Spectrum (DSSS) modulation which is both tolerant and unobtrusive to other spread spectrum wireless. Additionally, WirelessHART is also channel hopping to be even less interfering and more resistant. Lastly, Clear Channel Assessment (CCA) and channel blacklisting are also supported as additional options which can come in handy when dealing with non-wireless sources of noise. The result is coexistence.

Interoperability / Interchangeability

There are risks associated with proprietary technologies. For instance, non-backward compatible changes to the technology forcing upgrades, or changes to subscription models, license agreements, and fees that ultimately affect and spare part availability as well as their cost.

Proprietary WSN Technologies

There are WSN technologies which are not an international IEC standard. These WSN technologies are owned by a single manufacturer; proprietary technology. They license the technology to other , but ultimately the technology is controlled by a single vendor. There may be a trade alliance of manufacturers using the WSN technology in their products, that can be used together, but nevertheless the technology is owned by a single manufacturer and is not an international standard. Radio modulation and protocol information is not publicly available. Only obtainable through Non-Disclosure Agreement (NDA).

IEC Standard

WirelessHART is an international standard; IEC62591. The technology is not owned or controlled by a single manufacturer; it’s an open technology. This means you can buy WirelessHART products from multiple automation vendors and they will work together: interoperability. WirelessHART has universal and common practice commands used by all vendors meaning you can also replace a failed sensor with a sensor from another vendor and it will work with no system reconfiguration: interchangeability. Since WirelessHART is a standard, not owned by a single vendor, the technology, licenses, and fees cannot suddenly change. Standardization dramatically reduces the risks of the technology. Therefore I&C engineers specify IEC standards for all electrical and electronic equipment including telecommunications, not just WSN. Since there are WSNs available that are an IEC standard, do not invest in other WSNs which are not an IEC standard.

Transparent Integration

Just like wired sensors, wireless sensors need to be configured for the use case. And you need to monitor the self-diagnostics in the sensors. From time to time sensors may need calibration.

Vendor Specific Sensor Configuration and Calibration Software

Most WSN technologies do not have a common backhaul network application protocol for sensor configuration and diagnostics data. Therefore these WSN cannot be integrated into the plant’s existing Intelligent Device Management (IDM) software. Even the sensor manufacturer’s own IDM software used for their wired 4-20 mA/HART and FOUNDATION fieldbus devices does not work for their own wireless sensor product lines based on certain WSN technologies.

Note that although you could map sensor data to Modbus or OPC to bring that into the historian, analytics, or the control system, it would be impractical to manually map all the configuration and diagnostics information in each and every sensor that way. Too much work to initially deploy and then maintain. It is only practical to map the measured process variables.

Standard HART-IP for Intelligent Device Management Software

WirelessHART solves this with the HART-IP application protocol over Ethernet and other IP compliant media. HART-IP is used for the backhaul network to bring sensor configuration and diagnostics data into the plant’s existing Intelligent Device Management (IDM) software without any data mapping thus saving time and preserving the message format (how the information is logically organized in the message) and data structures (format of the data) for all information in the sensors. This is critical to make EDDL and FDI work. Therefore it is important to use a WSN with an associated backhaul protocol that does not change the organization or format of the data. The result is transparent integration with existing IDM system. WirelessHART also supports both the EDDL (IEC61804-3) and FDI (IEC62769) standards. That is, WirelessHART sensors can be managed from the same EDDL and FDI based Intelligent Device Management (IDM) software used for the plant’s existing 4-20 mA/HART and FOUNDATION fieldbus as well as the future Ethernet-APL instrumentation. The I&C engineers that deploy and manage the WSN are familiar with Intelligent Device Management (IDM), HART, and EDDL/DD. Even IDM software based on FDT/DTM can be used to manage WirelessHART sensors by installing the appropriate commDTM, gatewayDTM, and deviceDTMs. Other WSN cannot achieve the same result because they either don’t support EDDL and FDI, or they don’t have a standard backhaul protocol like HART-IP preserving the message format and data types for the configuration and diagnostics information.

Version and Lifecycle Management

Over many years of plant operation a never ending parade of new models and versions of wireless sensors will be coming into the plant.

Most WSN technologies have heterogenous software solutions with a sensor configuration and diagnostics software for every sensor manufacturer. Whenever a manufacturer introduces a new model or new version of wireless sensor, that software must be upgraded to support the new model or version of wireless sensor. This becomes a big burden to maintain software for pressure sensors from one vendor, another for the temperature sensors from another vendor, a third for level, another app for position, and yet another for toxic and flammable gases a.s.o. up to date. Whenever a new brand of wireless sensor is introduced in the network, yet another configuration software must be installed and managed. In some cases software update may require operating system upgrade which in turn may require computer change. And what if a software app for one of the sensors is not yet available for that new operating system version, then what do you do?

WirelessHART solves this with EDDL and FDI. Whenever a new model or new version of WirelessHART sensor comes into the plant, you load the EDDL file or FDI package for that sensor on the IDM software and this makes the IDM software compatible with the new model or version for easy version and lifecycle management. There is no need to install a new version of IDM software. As a result the IDM software is kept up to date with new models and version of sensors, and able to access all the features in both old and new sensors; backwards and forwards compatibility. I&C engineers are very familiar with EDDL/DD, and EDDL is at the core of FDI as well. Most other WSN technologies do not support EDDL and FDI and therefore the administration burden is greater.

Power Supply (Battery Life)

Sensors are often installed in hard to reach locations; high on the pipe racks so you need scaffolding for access which is time consuming and labor intensive. Some sensors are installed below grated flooring. Frequent battery replacement would become a burden. The battery life depends on the type of sensor and selected update period. Some types of sensors consume more power than others and therefore results in shorter battery life. The WSN technologies have very different battery life so it is important to select WSN technology carefully. Faster update period means shorter battery life. As explained above, do not select unnecessarily fast update period because it will drain the battery sooner. This is true for any WSN technology. Most wireless sensors are set to 1 min update period to enjoy multi-year battery life. Vibration may be set to 1 hour. I&C engineers know what update period is required depending on the use case. Apply Management of Change (MoC) procedures if you want to change the update period later. Also, upload of large data sets like vibration spectrum/waveform should also be done thoughtfully.

Most WSN technologies have a battery life of only a few years, even with update periods of several minutes or even once a day. This means frequent battery replacement. Taking the cost of erecting scaffolding into account, the cost of battery replacement is higher than the of the battery itself. A low-price sensor, even with a low-price off-the-shelf battery, but with short battery life has a higher total cost of ownership.

WirelessHART solves this with time-synchronized publisher-subscriber (pub/sub) communication. Every sensor has an internal clock and measure on a precise schedule (interval) set depending on the application. For instance a corrosion sensor may only need to measure and publish the data twice a day, a vibration sensor once an hour, a steam trap sensor once a minute, but a flow sensor once a second – or anywhere in between. I&C engineers know how to select the update period as slow as the use case permits to maximize the battery life. The sensor is in fact sleeping most of the time thus conserving battery. The sensor wakes up briefly to make a measurement. Then goes back to sleep again. As a result WirelessHART can achieve 5-10 year battery life for sensors. So you seldom have to replace the battery. This means lower total cost of ownership. Many other WSN technologies are not time-synchronized and therefore have shorter battery life.

Note that even with this sleep mode the sensor is still able to relay data from other sensors in a mesh topology because when it needs to relay data from the others, the radio, and only the radio wakes up to relay the data. To relay it does not turn on the sensor and signal conditioning electronics. The radio is extremely low power so you still get the long 5-10 year battery life. Also note that even if a sensor runs out of battery, this doesn’t mean the whole WSN goes down. Remember, thanks to the WirelessHART mesh topology, there are multiple data paths from the sensors to the gateway. If one of the sensor routing data from other sensors goes down, that data automatically gets routed through another sensor to the gateway instead.

Lastly, because sensors are installed in hazardous areas you need sensors using an intrinsically safe “power module” instead of the bare battery, such that you can easily replace the power module in the hazardous area, without the need for gas sniffing and hot work permit.

Self-Monitoring and Diagnosis (Network Management)

You cannot troubleshoot a WSN with a test pen or a multi-meter.

WirelessHART sensors have a standard way of reporting communication statistics to the gateway including signal received strength (RSSI), percent availability (reliability), latency, number of joins, and number of data packets lost which can be seen for each individual sensor from the gateway or software. This makes it easier to identify any weak spots in the network that may need to be improved, for instance with a repeater.

Equipment and Components (Tools and System Integration)

Just like wired sensors, you must be able to setup the sensor on the bench in the instrument workshop. This may include sensor type selection for a temperature transmitter or tank height settings for a level transmitter, to function test before installing it in the field. And you must be able to calibrate the sensor in the field or on the bench using a portable calibrator. Since there is no WSN in the instrument workshop, there must be an easy way to do this by other means; out-of-band. This is over and above configuring the network identifier and join key.

Digital transformation of integrity management, reliability and maintenance management, energy and emissions management, occupational safety and health management, as well as production and quality management work practices as part of 4.0 requires all kinds of advanced sensors such as vibration, corrosion, erosion, acoustic noise, pressure, temperature, level, flow, discrete contact, position, toxic and flammable gases. The plant will likely have wireless sensors from multiple manufacturers. Wireless sensors and apps cannot be isolated islands of automation; they must integrate with the existing automation such as the plant historian, control system, and IDM system.

Specialized Sensor Configuration and Calibration Tools

For several WSN technologies there is no common way for out-of-band configuration and calibration. That is, even for the same WSN technology, manufacturers have different ways of configuring their sensors before it is brought to the field. This means, different specialized tools for each vendor.

No alt text provided for this image

One requires a laptop with IR adapter, another by USB cable, a third using NFC, and a fourth from the LCD local operator interface on the sensor. Managing a variety of tools in such a heterogenous environment will cause delays and the system could even fall into disuse. Direct interface of a portable calibrator is not supported at all.

Familiar Sensor Configuration and Calibration Tools

WirelessHART solves this with a wired HART console port which all WirelessHART sensors have.

The 4-20 mA/HART tools almost all plants have today can be connected to this console port: HART field communicator, HART modem for laptops, and portable calibrators. This is ideal for configuration, function check, and calibration on the workbench or in the field because it is the same tool used for all WirelessHART sensors regardless of manufacturer, and the same familiar tools instrument technicians are already using for wired instrumentation.

Courtesy: Beamex

(Photo courtesy: Beamex)

Instrument technicians in the plant have been using these tools for the past 20 years so they are very familiar with them. The procedures for WirelessHART sensors are almost the same as for 4-20 mA/HART sensors.

No alt text provided for this image

As a result, the learning curve to adopt WirelessHART is less steep than for other WSN. Calibrators don’t support other WSN.

Heterogenous Systems

No alt text provided for this image

Many WSN technologies were designed for non-industrial applications like smart city, smart grid, or agriculture. Such WSN technologies don’t define a standard application protocol with a common message format or common data types. Instead, each sensor vendor has its own message format and data formats for a specialized software app which is the only way to use the data; either on a computer on premises in the plant, or in the sensor vendor’s cloud. Because of this, the gateway for some WSN technologies does not support Modbus or OPC for sensor measurement data. Indeed it is more of a base station (like a wireless access point) than a gateway because it doesn’t convert from one application protocol to another. Indeed they are often called “base station” rather than gateway because it is more of a tunnel or conduit for the measurement from the sensor to its dedicated software. The sensor data does not integrate with applications like the existing plant historian, control system, or IDM. This in turn means it becomes hard to put that data to good use.

Buyer beware: A WSN gateway may support Modbus, but only for gateway and network status information, not supporting access to measurement data from the sensors

Integration with Existing Systems

WirelessHART solves this by using the standard HART application protocol which has a common message format and common data types for all WirelessHART devices. This enables WirelessHART gateways to be true gateways, also converting standard data from a mix of sensors from multiple automation vendors to common industrial protocols such as Modbus/RTU, Modbus/TCP, EtherNet/IP, OPC Classic, and OPC-UA for integration with existing systems. As a result, data can be brought together into applications like the plant historian, control system, or IDM without custom programming. From there it goes into analytics, dashboards, reports, and notifications. I&C engineers are very familiar with these protocols and how they integrate with their existing systems. Multiple simultaneous use cases share the same network, and data is marshalled into common applications. Many other WSN technologies lack standard application protocol and therefore do not integrate with existing automation systems.

No alt text provided for this image

Think long term, beyond initial deployment of the first use case. The first use case will need only one type of sensor and one app and could be deployed stand-alone. But there are dozens of use cases in a plant and as they are introduced one by one, the wrong WSN technology will show its limitations.

Think beyond the first use case

Configuration / Commissioning (Sensors)

As part of commissioning of a new wireless sensor, the various sensors for vibration, corrosion, erosion, acoustic noise, pressure, temperature, level, flow, discrete contact, position, toxic and flammable gases need to be configured for the individual applications they are used in; such as selecting sensor type, engineering unit, and travel type etc.

Most WSN technologies do not have a common solution for configuring sensors from multiple vendors, not locally at the sensors, nor centrally from another location. Other WSN solutions offered by vendors often tend to be a single application; pressure, level, or valve position. So even with the same WSN technology you end up with multiple heterogenous systems with one software to configure pressure sensors from one vendor, another for the temperature sensors from another vendor, a third for level, another app for position, and yet another for toxic and flammable gases a.s.o. Each software app having its unique look and feel so they all work differently. This mix of software tools and procedures makes the mix of sensors in the plant harder to manage.

WirelessHART solves this with EDDL and FDI. This means that all kinds of WirelessHART sensors from multiple automation vendors can be configured and commissioned from the same single EDDL and FDI based software or the same handheld field communicator. The result is a common look and feel across all sensors, so they work the same way, with very consistent procedures, and therefore easy to use. IDM software centrally store the configuration of all devices including wireless sensors so the sensor configuration can be restored into a new sensor when an old sensor is replaced. I&C engineers are very familiar with EDDL and FDI so they can manage this. Many WSN technologies do not support EDDL and FDI and therefore do not provide a uniform user interface.

Service and Maintenance (Sensor Diagnostics)

Sensors are installed in the harsh plant environment with high or low ambient temperature, humidity, dust, and flammable atmosphere. Many sensors are in direct contact with the process with high pressure or vacuum, high temperature or cryogenic, abrasive, corrosive, and subject to vibration. Sensors need self-diagnostics to detect, and where possible, predict sensor failure. A mix of wireless sensors for vibration, corrosion, erosion, acoustic noise, pressure, temperature, level, flow, discrete contact, position, toxic and flammable gases from multiple vendors must be monitored.

As explained, with most WSN technologies you end up with multiple heterogenous systems with one software for each sensor vendor. Each vendor’s software app having its unique look and feel; different icons, color coding, and terminology for the severity of sensor problems detected or predicted. This inconsistency among vendors make sensors and software for such WSN technologies harder to use; both for device diagnostic alarm management and prioritization for maintenance scheduling.

NAMUR NE107

WirelessHART solves this by supporting the NAMUR NE107 recommendation for four standard status signals. Sensor diagnostics is mapped to these four status signals. As a result, diagnostics display becomes consistent across automation vendors and sensor types, making service and maintenance easy. Many I&C engineers are already familiar with the NE107 status symbols so they will welcome this.

No alt text provided for this image

Most other WSN do not support NE107 so they do not display diagnostics consistently.

Certification

Wireless sensors from multiple vendors must be able to join the WSN. It must be possible to identify, configure, and calibrate the sensor, and to get the measured variables and self-diagnostics status regardless of vendor. That is, the sensor must comply with the many details in the WSN technology specification.

Some WSN technologies do not have associated product conformance testing and registration by an independent organization. That is, without a test certificate, there is no telling if the wireless product will work well on the WSN.

WirelessHART solves this by conformance testing and registration by the independent FieldComm Group (FCG) organization. Products which pass the test are given a test certificate which gives some level of reassurance the product will work well on the WSN. As a result, you get peace of mind. I&C engineers are already familiar with the FCG test and registration since they are also using registered 4-20 mA/HART and FOUNDATION fieldbus devices. Other WSN technologies which do not have independent testing, do not provide such reassurance.

WSN Technology Attributes

The Fourth Industrial Revolution (4IR) is upon us. A WSN is a key enabler for this paradigm shift in a plant. This study is purely looking at the WSN technology requirements. NE124 also includes product and engineering requirements. The WirelessHART technology meets these technology requirements and there are WirelessHART products meeting the product requirements too. In summary, you need a WSN technology with these attributes:

No alt text provided for this image

WirelessHART is the ideal WSN technology in a plant. No other wireless technology meets these requirements. For other applications like smart grid, smart city, and agriculture there are other wireless technologies. No single wireless technology is best in every application. Every new plant being built shall be built with WirelessHART infrastructure from the very beginning. Many plants already have a WirelessHART network, but outside the I&C team it may not be widely known. If you plant doesn’t already have WirelessHART infrastructure, now is the time to lay that foundation for digital transformation. Schedule a (virtual) meeting with your I&C department today to kick-off the WSN project. If the plant has already done a digital transformation workshop to identify digitalization use-cases, look to those use cases to see what wireless sensors are required. If not, a digital discovery session is the place to start. Forward this essay to your plant manager to get the workshop organized. And remember, always ask for product data sheet to verify reliability commitment, technology attributes, and product features. Well, that’s my personal opinion.

 

Author :Jonas Berge

Leave a Reply