Smart I/O systems vs. fieldbus networks
Over the last few years, one-by-one, major process control system suppliers have been introducing new field I/O systems that offer some sort of simplified wiring scheme and easier configurability. (See article from January 2013 at the bottom.) These new systems are far less complex than traditional analog field wiring using junction boxes, marshalling cabinets, and techniques like half-knit and fan-out. A visit to a plant with this kind of wiring must include seeing waterfalls of cables and trays. (See article from March 2013 at the bottom.)
Promoters of fieldbus technology, primarily Foundation fieldbus and Profibus PA, look at these new configurable I/O systems and point out that the kind of convenience now being touted has been available for many years using fieldbus networking. As Carl Henning, deputy director of PI North America, asks, "It’s 2014—why would anyone architect a project with marshalling cabinets? In process applications, Profibus PA and Foundation fieldbus long ago eliminated the need for marshalling cabinets. These cabinets are expensive to engineer, build, wire, maintain, and troubleshoot. It’s better to skip cabinets and go straight from the controller to devices. A fieldbus system provides multi-drop connections that eliminate the need to wire devices to an intermediate point like a marshalling cabinet. This has obvious benefits in initial costs, but more importantly, the fieldbus provides diagnostic data that can be used to prevent or shorten downtime."
While Henning’s points are all valid, one might ask why there are not more fieldbus deployments and why companies still install systems with traditional wiring schemes. Plants that are committed to using fieldbus networks and have made the investment to install them and train users thoroughly, find them very reliable and remain generally loyal to them. There is a sizeable group of users concentrated in the oil, gas, and petrochemical industries who will consider nothing else. Over the last several years there have been a number of very large new greenfield fieldbus installations around the world to reinforce the point.
However, Klaus Erni, DeltaV and DeltaV SIS hardware product marketing/business development manager, Emerson Process Management, sees the fieldbus side of his business as essentially flat. "Big oil and gas companies that have selected fieldbus will stick with it for at least another decade," he says. "There are two types of customers now. One kind went for fieldbus a few years ago and they don’t want to switch direction. They say, ‘We standardized the whole fieldbus thing on the commissioning side. We have all the templates and testing done, so commissioning fieldbus for us is easier than HART.’ That’s fair, they want to continue using things they invented and procedures they have in place. The other kind of fieldbus user wants to pick and choose which approach depending on the application. Depending on functionality, they see situations where they need fieldbus and want to use it, but there are less critical situations where they say they are good with a HART device. They choose depending on the functionality and add those devices to their system, but that gets messy."
Joe Bastone, solution manager, Experion control and I/O for Honeywell Process Solutions, agrees but sees fewer of the second type precisely because of messy networks. "Companies decide, ‘We’re going to adopt fieldbus because of the diagnostic support it gives us,’" he says. "They aren’t going to decide on a onesy-twosy basis and say, ‘We’re going to use fieldbus here and we’re going to use HART over here.’ It’s more of a wholesale decision because there is other infrastructure that comes into play with asset management systems and the like."
Even so, Bastone says he has seen many plants with small isolated fieldbus deployments that were tried as an experiment but never used more widely. These can be problematic if technicians don’t work on them often enough to keep their skills sharp.
Diagnostics and functionality
The new smart I/O systems still use conventional analog and HART devices. In that context, a digital device is strictly on-off. The new systems do not provide any more diagnostic information than conventional HART-enabled I/O; however, they do take full advantage of what HART offers. Traditionally, the working assumption was that fieldbus networks were capable of providing more diagnostic information than HART. That is still true, but more current versions of HART have closed that gap somewhat. Devices using HART 6 and 7 have added new capabilities since version 5, and the new I/O systems can read it.
The nature of the diagnostic differences varies depending on the type of device, but there are some that are more or less universal.
Multiple variables—HART devices can provide variables in addition to the primary variable, but a fieldbus device can provide more information and do it more directly. It’s rare to find a plant that controls from a secondary HART variable, but it’s far more common with fieldbus.
Control in the field—Foundation fieldbus supports having a sensor and actuator on the same segment communicating in a way that allows the two to drive a PID loop without support of an external host.
Link active scheduling—Foundation fieldbus also allows a segment to continue functioning, even controlling a loop, when the host system has failed. As long as there is power, there can be enough intelligence resident in the devices to continue functioning.
Companies that have deployed fieldbus networks but find them unsuccessful often cite training issues as the main problem rather than problems with the technology. Fieldbus is not intuitive and technicians find that it requires some effort to grasp. Even companies that are enthusiastic about using fieldbus networks find the process of designing segments daunting. Fortunately, once that process is over, it rarely needs to be revisited.
Electronic I/O is growing
While Erni sees sales of fieldbus equipment solid but flat, Emerson’s electronic marshalling product line is growing steadily. The key factor of the equation is the flexibility offered by the new electronic systems, which effectively eliminates any need for designing I/O strategy. "Even fieldbus segments have to make some sense as to where devices are connected. You don’t mix devices on a segment, and data is bound to where that segment goes. The big advantage of universal I/O is that it avoids any need to design device strategy. Anything can go anywhere."
Bastone says that capability can make a huge difference in new applications. He adds, "The killer app of universal I/O is the ability to make late changes during a project and not suffer the huge penalties that come with realizations like, ‘Oh nuts, we forgot to design for this other compressor.’ Now we’re in a mad rush because we have to figure out where we’re going to land this I/O, how we’re going to get all the engineering drawings done, and trying to stay off the critical path for the project. That’s where the beauty of universal I/O and universal cabinets comes into play."
There’s no question that the influence of cloud technologies is beginning to change how we think about field devices. The ability to direct information from any kind of device to any point in the control architecture by typing a few keystrokes is pretty compelling to anybody who has had to troubleshoot a hardwired device or redirect its output by moving the relevant cable.
Fieldbus architecture is easier to work with than traditional wiring, and it represents the ultimate in diagnostic capabilities so it will still have loyal followers for years to come. Even so, the simplicity and convenience of newer I/O systems that move closer to a true instrumentation cloud are difficult to ignore. Either way, there is little reason to continue using old-fashioned hardwired I/O. As Henning reminded us, it is 2014 after all.
Peter Welander is a Control Engineering contributing content specialist. Reach him at email@example.com.
Read related articles more about field I/O strategies below.
For more information, visit: