1 July 2005
Coming of age
The economic case for large-scale use of wireless sensors is overwhelmingly favorable.
By Clifford Lewis
Instrumentation engineers have successfully applied wireless communication technology in process plant applications for many years.
SCADA systems routinely send data through cellular telephone links.
Tank farms forward liquid level reports to central control rooms hundreds of yards away.
Offshore oil platforms transmit continuous streams of data at high speeds to land-based stations.
Wireless sensors are now giving plant engineers unprecedented access to plant data previously unavailable or unattainable with wired sensors.
Wireless sensors represent a significant cost savings, particularly for parameter monitoring, and provide a quickly implemented, secure link from the process plant areas to the control room.
Advances in digital technology have made large-scale installations of wireless sensors practical and increasingly common. With no external field-supplied power, wireless sensors can run continuously for over five years on an internal replaceable battery while giving multiple updates per minute for the entire period.
These large-scale wireless sensor installations can network and link to SCADA systems, distributed control systems (DCS) or operate independently.
Wireless sensor economics
The economic case for large-scale use of wireless sensors is overwhelmingly favorable.
In most industrial applications, the purchase price of a sensor and its associated electronics is generally less than 10% of the total installed cost of the device.
Many plants allocate the cost of a wiring pairs at over $8,000 per pair. Other plants estimate the cost of running wires by the foot, which run $40 to $100 per foot, depending on the electrical classification of the area.
Of course, wiring needs to run in cable trays, in conduit or underground and it never goes the shortest route between two points. It is not uncommon to run wires many hundreds of feet just to get a signal across even a single plant roadway.
The wiring costs add up very rapidly, so rapidly in fact, that most plants currently monitor fewer than half the measurement points they should be monitoring. The cost of running the wires is just prohibitive.
Wiring cost is just one dimension to the wireless sensor economics matrix. Ease and speed of installation are other important factors. Once a base radio sets up in a plant or on an offshore platform, it covers an area roughly 1,000 feet in radius. The actual signal reception can be easily determined in a quick site survey as noted in the sidebar. Installing a wireless sensor anywhere in the coverage area is as quick and easy as installing a gauge.
Installation time is far less. A typical installation can be up and running in less than fifteen minutes. The installation is so quick that wireless sensors can install in the field on a temporary basis, even by supervisory or engineering personnel. With this rapid installation time, instrumentation becomes flexible. Sensors can move about, test in one location, and then relocate to a more optimal spot if necessary. The low initial cost and speed of installation make it possible to put the first wireless sensor into a plant without the need for capital approval. After an initial trial, further justification is easy.
Tie into existing control
A large-scale installation of wireless sensors in a plant is only good if it can easily tie into an existing control system. Modbus digital output and easy conversion of the base radio digital data stream into 4-20mA or switch closure signals makes inputting large amounts of sensor data easily achievable. These tie-ins are understandable by plant personnel and reduce implementation time.
Most industrial process plants, offshore oilrigs, or oil producing locations already know numerous places where they would like good, solid data. This may be something simple, like continuous monitoring of relief valves to verify releases to a flare or to the atmosphere. Many tanks or loading facilities need overfill protection devices to prevent unpleasant and often costly situations. Pump efficiencies need periodic monitoring. Employee health and security measures, or safety shower uses need to be improved with alert signals.
All of these applications are applications for use of large-scale installations of wireless sensors. Such installations can typically take place with wireless sensors for less than 20% the cost of a comparable wired solution. With savings like this, coupled with a long battery life and the availability of rugged industrial grade sensors make a very compelling case for wireless sensors.
Behind the Byline
Clifford Lewis (email@example.com) has a B.S. in chemical engineering and an MBA from Dartmouth. He is a vice president at Accutech. He also worked for Celanese and General Eastern during his career.
Here is a typical installation in the oil and gas industry. The first photo shows a wireless pressure transmitter mounted on the outlet of a producing oil well.
Small quantities of natural gas often leak from the top of the well, making this an electrical Division I environment. In a typical field, there may be several hundred such wells.
Monitoring of the well casing and well tubing pressures along with a number of other parameters is possible. Running wires to these locations is difficult and maintenance on buried cable wiring is high, making this an ideal application for a sensor requiring no wires.
Oil from the well pumps to gathering point tanks, where water separation takes place and flow measurement registers. A SCADA system, such as the one in the second photo often locates near the gathering point.
This SCADA system gathers the data on the well production and forwards it to a control point, often hundreds of miles away.
In this shot, there are actually several radios. One of these is a base radio "receiver" from the wireless pressure sensors located on the wells. This base radio can receive data from as many as fifty wireless sensors.
Up to sixteen base-radios, each one talking with 50 field devices can work in collaboration on a single oil-producing field.
Relief at last
Here is a wireless sensor mounted on a pressure relief valve in an unmanned gas-pipeline compressor station.
If a relief valve vents, the pressure in the exhaust line increases and the pressure rise registers to the wireless pressure sensor. The pressure release data then transmits to a control system.
One base radio, communicating with up to fifty wireless sensors can monitor multiple parameters in an unmanned compressor station.
Many wireless sensors have a receive signal strength indicator (RSSI).
Just as on one's cell phone, more bars represents a stronger signal.
To conduct a site survey, a base-radio can power up for several hours using a pair of nine Volt batteries, or the base radio can operate in the desired location. Then simply walking around and watching the RSSI on the wireless sensor will tell you where in your facility you can conveniently place a field device.
Here is the topology for the wireless sensor networks that we're discussing.
The base radio is the network gatekeeper and the controller.
The wireless sensors and base radio communicate over a spectrum of frequencies, 26MHz in bandwidth from 902 to 928 MHz. This frequency range does not require a user site license if the devices conform to a number of rules issued by the Federal Communications Commission (FCC).
Conformance with the FCC rules ensures that the wireless sensors do not interfere with other communications and that any radio interference in the surrounding area does not influence the performance of the wireless sensor.
Frequency domain range
The communication frequencies in the 26MHz spectrum delineate to 50 spaces.
Each wireless sensor communicates in only one of these 50 sections at a given time. With 50 independent communications bands available, up to 50 different devices can be sending or receiving signals at precisely the same time.
In actual practice, each communication is contained in a small packet of digital data. This message transmits quickly, allowing sufficient time for other devices to use the same sections of the spectrum without interference.
The base radio synchronizes all the field devices assigned to it, giving each field device a specific time to communicate. The frequency of the communication changes with each communication, a technique known as frequency hopping. This technique of communications was the heart of all secure military communications for many years.
Because each wireless sensor sends only the small amount of data associated with the sensor reading and its internal diagnostic data, each field device needs only a small slice of time to send its message.
Synchronization of all the field devices assigned to a given base radio allows one base radio to talk with many multiple field devices. This communication synchronization allows for large-scale installations of wireless sensors. It is not necessary to have one "transmitter" assigned to one "receiver" as is common with many high-speed, high throughput data links that often associate with a dedicated radio communications link.
By properly allocating the available spectrum and controlling the communication time, multiple base radios can co-exist in the same physical space in a plant without interference.
Smart sensor standard sees commercial success—IEEE 1451 update
For years, it was unclear if the IEEE 1451 standard would make its way into live sensing systems. Indeed, it wasn't even clear that the smart sensing standard was even necessary.
ElectronicsWeekly reported that smart sensors commonly used in digital instrumentation and measurement systems now have their own industry standard-based plug-and-play capability to simplify and widen their use in a range of applications.
The IEEE1451.4 standard interface for smart sensors and actuators platforms on the mixed-mode communication protocols and transducer electronic data sheet (TEDS) format. The aim is to allow installed analogue transducers to easily connect to digital measurement systems.
According to Torben Licht, 1451.4 working group chair and product manager at Bruel & Kjaer, "the protocols it contains replace the diverse transducer solutions manufacturers had introduced with limited success. It will have a major effect by dramatically expanding the pool of network-compatible transducers and the use of control networks."
The standard requires non-volatile EEPROM – electrically erasable programmable read-only memory, a memory chip that retains data content after power has stopped. EEPROM is erasable and reprogrammable within the computer or externally – embedded in sensors to hold and communicate details needed for plug-and-play capability.
The necessary interface identifiers for chips are available to manufacturers via the Internet.
A complete TEDS may contain sections for identification and properties for a type of sensor, such as accelerometer, microphone, strain gauge, thermocouple, thermistor and many others. The TEDS may also contain a calibration template for the sensor.
"The standard, by adding self-identification at the transducer analogue interface, has the potential to make any measurement system, analogue or digital, easier to set up, configure and maintain," says David Potter, working group vice chair and platform manager at National Instruments.
According to supplier Honeywell, interoperability between hardware and software should be straightforward, because the standard is on a common platform, called Sensors Plug&Play, which retains its compatibility with the LabVIEW graphical programming software to program its sensors.
A smart TEDS sensor contains an integrated EEPROM chip which carries scaling, calibration and user information, facilitating digital interrogation of the sensor's configuration data. The customer's data acquisition system reads both the smart TEDS sensor analogue signal as well as the EEPROM configuration information.
One of the aims of the standard sensor interface is to simplify the often time consuming process of system set-up after the hardware and software is acquired.
According to Martin Armson, director of marketing for Sensotec sensors at Honeywell, the TEDS approach should provide more than improved accuracy and productivity. "This technology also eases the logistics of managing test and measurement systems that can comprise hundreds or even thousands of sensors," says Armson.
As a result, adds Armson, the designer can spend more time refining the measurement system, and much less time looking for sensors in the system.
Retrofitting the TEDS platform to existing sensors in legacy systems is also possible according to Armson: "Designers can retrofit any of our sensors by burning an EEPROM and mounting it in the connector, the sensor itself, or a special adaptor attached to the sensor cable."
Another option for PC-enabled customers is Virtual TEDS, a library housed at National Instruments' website that offers sensor data in the standard TEDS format for direct download to customers' data acquisition and signal conditioning systems.
The library of TEDS sensor data is the result of collaboration between National Instruments and sensor vendors including LEM, PCB Piezotronics, Weed Instrument, Honeywell Sensotec, Endevco, and Bruel & Kjaer.
To access the library, users enter the manufacturer model or serial number and can then receive the sensor's specific scaling and calibration information packaged in the IEEE1451.4 binary format.
Once users download the electronic data sheets, they can use the Virtual TEDS Editor, a standalone Windows application written in LabVIEW 7 Express to translate the file into script that the user can read and modify. With the Virtual TEDS Editor, users can view and edit sensor properties stored in the Virtual TEDS file, making large systems with hundreds of sensors easier to use and maintain.
"At the very least customers can take advantage of the 'paperless calibration' part of plug and play by accessing calibration data from their manufacturer's website," adds Armson.