Fieldbus in the field
If you still connect devices and controllers with point-to-point wiring, it’s time to take a look at one of the fieldbus networking options available. An appropriate selection will save money now and later.
In factory automation there are lots of buses to choose from. Buses are only networking protocols, so don’t let what may at first seem to be a baffling array of options stop you from choosing one, or several. Don’t install point-to-point wiring just because that’s what you’ve always done. There are many benefits to using buses. And by benefits, we mean saving money.
Let’s take a look at some of the bus protocols with some guidance on making the right choices.
Industrial Ethernet is the top-of-mind networking protocol these days and rightly so, but there are places it does not fit well. For example, it makes no sense to connect a limit switch to an industrial Ethernet network. But a new automation project today should begin with industrial Ethernet and then add appropriate fieldbus and sensor/actuator networks at lower levels as needed.
What is industrial Ethernet?
Industrial Ethernet leverages commercial Ethernet and adds speed, determinism, and ruggedness. In non-hazardous areas, industrial Ethernet often replaces the older serial fieldbuses. If you think of a typical office environment, Email can be delivered in a few seconds or even minutes, but a command to turn off a factory conveyor better be delivered in milliseconds or else product may pile up on the floor. In addition to speed in the factory world, commands need to be delivered in a repeatable timeframe, which is determinism. An Ethernet switch from your local computer retailer may work electrically for industrial Ethernet, but it is unlikely to be rugged enough to survive physically in a hot, humid, oily environment on the plant floor.
The most used industrial Ethernet protocols today are Profinet, EtherNet/IP, EtherCAT, and Modbus TCP. A recent market survey showed that the installed base is roughly one-third each for Profinet, EtherNet/IP, and the others.
What is a fieldbus?
There are three general kinds of device-level networks, discrete and process fieldbuses, and sensor/actuator networks. The division is based on the types of devices they communicate with and the types of data they carry. All three types typically connect a group of devices to a single controller or higher network. All provide a digital, serial, two-way, multi-drop communication link among controllers and devices.
Discrete fieldbuses connect digital I/O to controllers. For factory (discrete manufacturing) applications in North America, Profibus DP and DeviceNet are the two most popular choices. Other choices include Interbus and CC-Link LT. Discrete fieldbuses communicate with devices such as:
• Motor controllers;
• Process instruments;
• Machine vision devices; and
• Other devices that generate complex data.
Such devices are normally externally powered, however DeviceNet can also carry power.
Process fieldbuses are designed to communicate with process instruments and actuators. The two most common are Profibus PA and Foundation fieldbus (FF). Process fieldbus deployments began in oil refineries and chemical plants where there are often huge populations of devices, but adoption has spread into many more areas such as water treatment and pharmaceutical manufacturing.
Process devices with fieldbus connectivity include:
• Pressure sensors;
• Level sensors;
• Control valves; and
• Temperature sensors.
Both process fieldbuses carry power to the devices and are designed to operate safely in hazardous environments. This is a major advantage where there is potential for fire or explosions.
Process and discrete fieldbuses are similar in that they can carry the same types of data. In fact, in some plants, a discrete fieldbus network may be used with process instrumentation, however this requires external powering.
A sensor/actuator bus is designed to work with simple devices that are either on or off. The most popular sensor/actuator bus is AS-interface (AS-i for Actuator/Sensor interface). It typically connects simple I/O devices over a two-wire cable that also provides device power. Typical AS-i devices include:
• Limit switches;
• Proximity switches;
• On/off valves; and
AS-i supports line, star, and tree wiring topologies. Because of its low communications overhead, AS-i is well suited to the low data requirements of this application.
A new technology has emerged called IO-Link. While not a bus, it superimposes digital communications over the discrete IO wiring. IO-Link is particularly suited for use with intelligent sensors and actuators. It, too, can be connected into a fieldbus or industrial Ethernet network through gateways or proxies.
Devices and controllers
The complexity of an individual automation device helps determine the appropriate bus to connect it to. Simple devices attach to a sensor/actuator bus; more complex devices attach to a fieldbus or industrial Ethernet network.
For example, if you were considering an application and trying to determine what kind of bus to use, you should begin by looking at the types of devices and data: a proximity sensor that has to determine if a piece is in the right spot on a conveyor may work with a sensor/actuator bus. If the sensor has to say how far away it is, you will need a discrete fieldbus. The same thinking applies to a pressure measuring device. A switch that triggers when the compressed air line falls below 80 psi is fine for AS-i. However if you have to know what the pressure is at all times, a process fieldbus is probably a better choice.
Any of the controller types popular in industrial automation are typically available with fieldbus and industrial Ethernet connectivity. A DCS (distributed control system) is typically used in process applications. PLCs (programmable logic controllers) started in factory automation with on/off outputs, but are increasingly being used in applications covering discrete and process functions. PACs (programmable automation controllers) combine PC-based and PLC functionality. Actuator/sensor buses most typically connect to a higher level bus through a gateway or proxy and not directly to the controller.
Choosing a bus
Once you determine that you are past the level where an industrial Ethernet network is the best choice, it’s time to start trying to select the appropriate bus protocol. If you’re looking at a relatively large deployment, the answer may be more than one.
The first place to begin is considering what devices will need to go on a given network. You will need to see what devices are available from various suppliers that communicate with that type of bus. For example, if you want to use a flowmeter on a FF segment, you have to make sure the flowmeter you want to use is available with that type of connection. Vendors don’t necessarily offer every device with every connection option. Some vendors have preferences for one fieldbus above others.
If you have a wide range of devices in mind, such as you might in an installation for a production unit in a chemical plant, you may have your process instruments on FF networks but the motor controllers driving your pumps on Profibus DP. Don’t assume everything has to go on one type of platform.
Similarly, on a discrete manufacturing line, encoders and machine vision devices may connect to DeviceNet, but limit switches may communicate via AS-i. Use the appropriate network for the application.
You can’t just hook up a wire and have devices communicate; they have to have the network’s protocol embedded in them. This means that if you want to connect a shaft encoder to Profibus DP or a pressure transmitter to FF, the device has to be configured for that purpose and offer that connectivity. Moreover, the protocols are different from one network to another. So you can’t connect a Profibus PA transmitter to a FF controller card – it won’t work, even though the two share the same bus physics (like message size and speed). This characteristic is a major problem in situations where users want to deploy a fieldbus in an existing installation with legacy devices and point-to-point wiring.
The beginning of this article promised benefits in the form of saving money. With something so complicated, how is that possible? At every phase of an automation project, using a bus, or buses, can help reduce costs.
During the design phase, bus configuration tools simplify engineering and documentation. Many of the configuration tools provide documentation as a matter of course. The reduction in wiring inherently simplifies drawings and bills of material. Installation is inherently faster due to this reduction, too.
Commissioning bus-based systems leverages generic and purpose-built tools to speed the startup. Diagnostic tools can pinpoint wiring errors and incorrect termination. For all the bus types, an advantage is centralized access. Configuration and testing can be done from the central control station, eliminating the need to traverse the plant floor with handheld configurators, signal injectors, and multi-meters.
During operation many of the benefits from commissioning apply. In addition, many devices provide asset management information. For example, a fieldbus-based valve actuator knows how many cycles it’s been through and that information can allow maintenance to respond when a critical value has been reached as opposed to acting on an inefficient time schedule. Some protocols can pinpoint failures very specifically. For example, the digital input on I/O device 3, module 4, channel 1 has a broken wire.
It’s not all sweetness and light, of course. There are some disadvantages in using these buses. It’s not hard to use them, but it isn’t what your staff is used to. Training is definitely recommended, and there are many opportunities. Web-based, free one-day training classes, certified training, and onsite installer/maintainer training are available from automation providers, network consortia, and third parties.
The reduction in wire, conduit, terminations, terminal strips, enclosures, and associated labor is easy to quantify, but may be the smallest part of the savings. Larger savings will accrue during commissioning where troubleshooting tools and techniques shorten the startup. The largest savings will be seen during operation and maintenance where improved diagnostics will shorten downtime. Unfortunately, this is the hardest to quantify.
There is no reason to be intimidated by the wide choice of buses available. Industrial Ethernet, fieldbuses, and sensor/actuator networks each have their place in the industrial automation architectural hierarchy. Advice and training are available from your preferred automation vendor and from the independent organizations that support the various buses. Training is available from multiple sources – some of it free, some of it web-based. Compare the choices in each category with a checklist of the features that you need, then start with a small project or retrofit to learn how the buses work. There is no substitute for doing the engineering and seeing the results.
Carl Henning is deputy director, PI North America. He has an electrical engineering degree and is a Senior Member of IEEE.