Published

Sorting Out Wireless Standards for Smart Valves and Actuators

Back in the ’90s Profibus and Foundation fieldbus competed to replace 4-20 mA communication for field devices like valves and process transmitters with all-digital systems.
#standards #iiot #actuators

Share

Years later all three are still here, and now the battle is over the use of wireless networks in industrial settings. There’s WirelessHART, ISA100.11a, WiFi and the list goes on. It can be difficult to figure out which network is best suited for a particular application, and this isn’t helped by the fact that it’s difficult to figure out what standards govern them. This article will help to sort out the alphabet soup of different wireless standards.

 

The communication system in a plant or factory can generally be divided into several levels. At the top is the business information network, which carries information important to business functions like accounts receivable, plus production related matters such as Manufacturing Resource Planning and the like. Below that is the control-level network, which connects major components of the automation system like controllers, operator interfaces, I/O racks and so on. At the bottom level is the field device network, which connects sensors and actuators, including valve position indicators, back to the control system; that’s the level we’re concerned with in this article.

 

Wireless sensor networks and their topologies

Wireless field device networks can be traced back to the development of mesh networks. A mesh (see Figure 1 and other figures at the end of this article) consists of a self-organizing network of small, low-power nodes, each connected to its own sensor or other field device, and each able to both receive and transmit. The network also includes a gateway that provides connection to a larger system or wired network. The nodes are generally battery powered and have a communication range of only a few tens of feet, but each node acts as a repeater for the other nodes within range. If one node transmits a message its neighboring nodes will receive it and pass it along until it reaches the gateway (Figure 2); in this way a mesh can cover a geographic area considerably greater than the range of one node. If communication between two nodes in the mesh (or one of the nodes themselves) should fail the message will automatically be re-routed to reach its destination (Figure 3). This self-healing property makes mesh networks quite robust. The batteries in a mesh network can last months or even years, because the nodes spend most of their time in a powered-down condition (“sleeping”). At set intervals (which can be anywhere from less than a second to several days) they all wake up to communicate their messages, which tend to be only a few bytes long, then go back to sleep. WirelessHART and ISA100.11a networks use this arrangement.

An interesting characteristic of mesh networks is that they become more robust as more nodes are added, since more nodes create more alternate signal paths.

Not all low-power field device networks are organized as meshes; some use a star arrangement (Figure 4), in which the low-power nodes communicate only with a router (which is larger and generally runs on regular line power) and not directly with each other. The router communicates with the gateway. Multiple star networks can be combined (Figure 5), to form what’s called a hybrid mesh. This is the way ZigBee networks are arranged.

Some of the first mesh networks were intended for household and building automation use, such as turning lights and air conditioning units on and off remotely, but others were aimed at industrial uses, mostly for equipment monitoring. Perhaps the first industrial application was to equip pieces of rotating equipment around a plant with wireless vibration sensors that would periodically transmit data back to a central control system; this way a failing bearing could be spotted before it caused a shutdown. Temperature and pressure readings could be handled the same way. In recent years, however, applications have expanded, and wireless communication for valves has emerged as an important application.

Standards

Any network, whether wired or wireless, uses a series of protocols to standardize how messages are put together, transmitted, received, passed up the line to higher-level systems and so on. The general model used to represent the way the different types of protocols are arranged is the seven-layer Open System Interconnection Reference Model (OSI Model), developed by the International Organization for Standardization (ISO) (below).

OSI Seven-Layer Model for IEEE 802 networks

Number

Name

Function

7

Application

Network process to application

6

Presentation

Data representation and encryption

5

Session

Interhost communication

4

Transport

End-to-end connections and reliability

3

Network

Path determination and logical addressing

2

Data Link

Logical Link Control

Media Access Control

1

Physical

Media, signal and binary transmission

Numerous standards have been developed to suit almost every type of communication activity. Many of those dealing with local area networks (LANs) and metropolitan area networks (MANs) are handled by the IEEE (Institute of Electrical and Electronic Engineers) 802 working group, which is divided up into a number of working groups. For example IEEE 802.3 covers Ethernet networks, used in almost every office and many homes, while IEEE 802.11 covers wireless local area networks (WLANs), and IEEE 802.15 covers wireless personal area network (WPANs). Each of these working groups is further divided into more specialized working groups that publish the actual standards.

IEEE 802.15.4

With one exception all the networks to be discussed in this article use the IEEE 802.15.4 standard, which defines the two lowest protocol layers of low-speed, low-cost wireless personal area networks (WPANs): Physical (PHY, which covers things like radio frequencies and modulation methods) and Media Access Control (MAC, which provides reliable single-hop transmission). The IEEE 802.15.4 wireless systems discussed here shorten the rest of the stack as well, to just physical, MAC, network and application layers, as shown below.

IEEE 802.15.4 protocol stack

The different networks use different upper-level layers, and IEEE 802.15.4 gives multiple choices for each layer, so most of the different networks are not directly compatible with each other.

IEEE 802.15.4 gives a choice of four different physical layers using two different frequency ranges (10 channels in the 902-928 MHz band and 16 channels in the 2.4 GHz band) and three different modulation methods: binary phase-shift keying (BPSK), offset quadrature phase-shift keying (O-QPSK) and a combination of BPSK and amplitude shift keying (ASK). It also provides for two different spectrum-spreading technologies: direct sequence spread spectrum (DSSS) and parallel sequence spread spectrum (PSSS). It supports mesh, star and peer-to-peer topologies.

ISA 100 (previously known as SP100)

ISA (the International Society for Automation) formed the ISA100 committee in 2005 to establish standards for wireless field-level systems with the ultimate goal of creating “the next 4-20 mA standard.”

ISA100 ranks usage classes by criticality from 0 (emergency action — always critical and applied to safety systems) — to 5 (logging and downloading/uploading — no immediate operational consequence).

While the ISA100 committee is working in a number of areas, so far it has released just one standard, ISA100.11a, "Wireless Systems for Industrial Automation: Process Control and Related Applications." Officially adopted in September 2009, in ISA’s words, the standard “is intended to provide reliable and secure wireless operation for non-critical monitoring, alerting, supervisory control, open loop control, and closed loop control applications” where latencies of 100 ms can be tolerated.

ISA100.11a runs at 2.4 GHz, adds channel-to-channel frequency hopping to the regular DSSS, and uses a mesh topology and time division multiplexing (each node is assigned its own time slot and sleeps the rest of the time). It can be used for class 1 (non-critical) to class 5 applications. The first release of the standard concentrates on process industrial applications, but later versions will also support factory automation. Honeywell’s OneWireless system uses it; it also supports Profibus, Modbus and Foundation fieldbus.

Other ISA100 projects in the works include ISA100.15, for a wireless backhaul network (connecting the ISA100.11a networks to control systems); ISA100.14, for trustworthy wireless; ISA100.12, aimed at converging ISA100.11a and WirelessHART; and ISA100.21, for people and asset tracking and identification.

As an aid in getting companies on board and making sure everything works together, the ISA100 Wireless Compliance Institute was formed; it provides compliance testing programs, associated market awareness, and technical support to users and developers. Ten days after official issuance of the standard the ISA100 Wireless Compliance Institute deployed products from six companies in a field demonstration of a working system at an Arkema liquid organic peroxides plant in Crosby, TX. Included in the demonstration were sensors for temperature, pressure, contact closure, valve positioning, gas detection, corrosion detection and others.

Process control companies that are members of the ISA100 Wireless Compliance Institute include Honeywell and Yokogawa; the others are mostly users or universities. Other companies expressing interest, according to wireless equipment maker Nivis, include Banner Engineering, Cameron, ConocoPhillips, Dresser Masoneilan, Flowserve, Krohne, Magnetrol and NASA.

Advantages of ISA100.11A are the ability to work with multiple network protocols (HART, Profibus, Modbus, Foundation fieldbus, etc.), reliability and security. The main drawback is higher cost.

WirelessHART

WirelessHART uses a specification developed by the HART Communications Foundation and is intended to make it possible to extract information from the 85% of HART-equipped devices that the Foundation estimates users are not accessing, because of cost or perceived difficulty.

WirelessHART runs at 2.4 GHz, using O-QPSK modulation at a data rate of 250 kbps and, like ISA 100.11a, uses frequency hopping (channel-to-channel) to increase resistance to interference. It uses a mesh topology (every node a router) and time division multiplexing, with field devices that connect to process or plant equipment, gateways that provide communication between the field devices and host system applications, and network manager software.

The system supports adapters that allow existing HART field devices to be integrated into a WirelessHART network and short-range handhelds for direct communication with individual field devices.

WirelessHART is backward compatible to core HART technology such as the HART command structure and Device Description Language (DDL), and field devices include a maintenance port that allows connection of existing HART tools.

Emerson Process Management has mounted a major effort on WirelessHART, which is also endorsed by ABB, Elpro, Endress+Hauser, MacTek, Pepperl+Fuchs, Phoenix Contact, Siemens and Yokogawa. Products available so far include process transmitters, a wireless adaptor for existing HART devices, a wireless adaptor for control valves, and gateways to connect to industrial field buses and to Ethernet. Fig. 6 shows a control valve equipped with a THUM WirelessHART adapter.

WirelessHART’s advantages include low power consumption per node, reliable communications, resistance to interference and the ability to handle up to 1000 nodes. Its weaknesses include reduced throughput for bursty traffic, the existence of a single point of failure unless redundant gateways are used, and higher cost than other systems.

ZigBee

ZigBee, which is governed by the ZigBee Alliance, is perhaps the technology most closely identified with IEEE 802.15.4. It has been associated mostly with home and building automation, and ZigBee-equipped actuators are available to control flow through heating system radiators or to shut off gas valves in case of an alert from a gas detector, but ZigBee has also been getting attention for industrial applications. One early indication came in 2005, when, in a presentation at Sensors Expo 2005, Cirronet (since acquired by RF Monolithics) predicted that ZigBee would soon provide control functionality in systems where some latency can be tolerated (250 Kbps data rate < 10 ms per hop) and monitoring where only extremely low latencies can be tolerated. With the appropriate gateway, it continued, ZigBee would integrate easily into a Modbus network.

In the U.S. ZigBee can run at 915 MHz with BPSK modulation and data rates of 20 kbps to 40 kbps, or at 2.4 GHz with O-QPSK modulation and a data rate of 250 kbps. Range per node is 10 to 75 meters, and transmitting power ranges from 1 mW to 100 mW.

ZigBee devices include the ZigBee End Device (ZE or ZED), ZigBee Router (ZR) and the ZigBee Coordinator (ZC) that connects the network with the outside world. A ZigBee system can be configured as a star (with multiple ZEs connected to a ZC), mesh (with multiple ZRs — which may have associated ZEs — connected both to each other and to a ZC) or hybrid or tree (with multiple ZRs — with or without ZEs — connected to a ZC). Routers can be set up to stay awake at all times listening for messages from ZEs, other ZRs or the ZC (so-called non beacon-enabled) or to sleep most of the time, waking up periodically to signal the other nodes of their presence (beacon-enabled).

ZigBee’s strengths include low cost, low power consumption, flexible network architecture, support from many companies and ability to handle a large number of nodes. Weaknesses include low data rates and the existence of a single point of failure (the ZC).

Bluetooth

Bluetooth is unlike the other wireless networks discussed here. For one thing it is not based on IEEE 802.15.4, and uses frequency hopping spread spectrum rather than DSSS or PSSS, although it does operate in the same 2.4 GHz band. Modulation can be shaped binary FM modulation or PSK. . It uses a master-slave arrangement, rather than a mesh, with up to seven slaves in a star or star cluster (although more slaves can remain connected in a parked state), and frequency hopping rather than DSSS.

Three classes of Bluetooth devices are available, ranked by power output:

Three classes of Bluetooth devices are available, ranked by power output:

Class

Power

Level

Approx. Range

1

100 mW

100 m

2

2.5 mW

22 m

3

1 mW

6 m

Bluetooth finds its main use in such things as mobile phones, computer mice, keyboards and the like, but has found some use in industry. So far industrial Bluetooth devices have been introduced by Phoenix Contact, Wago, Scanning Devices, ABB and Schneider Electric — the latter two for electric power monitoring and control applications. In addition, both Rotork and Flowserve Limitorque have valve actuators that use Bluetooth for setup and configuration, while Emerson has a handheld field communicator that can be updated via Bluetooth.

Advantages of Bluetooth are low cost and resistance to broadband interference. Disadvantages are small network size and the small number of vendors supporting industrial applications.

Summary

Low-power wireless networks are becoming more and more accepted in industrial applications. The big challenge is in choosing among the available standards. The best choice is to examine both present and future needs of the application and match them up with what’s available.

 

Watch for the Winter 2010 issue of Valve Magazine, available in mid-January 2010, for another article on wireless technologies.

 

RELATED CONTENT