CG0809_OTB
CG0809_OTB
CG0809_OTB
CG0809_OTB
CG0809_OTB

Will Wireless Replace Fieldbus?

Sept. 5, 2008
Hardwired Instruments Are Going to Be Around Until a Generation of Plant Operators Retires
By John Rezabek, Contributing Editor

If I’m running any wire for measurement and control/interlock signals, do the cost savings of wireless become less significant?  I heard a presentation from a wireless supplier a few weeks ago. One of the the attendees from an ammonia/urea facility asked,  “I want to hear about the wireless technology that’s going to replace fieldbus.” His facilities have an extensive fieldbus network, with scores of H1 (37.5 Kb) segments connecting virtually the whole facility. Installed in the early part of the decade, the selected host vendor’s implementation of fieldbus uses the network for I/O only, and to this day—nearly seven years later—the facility remains unable to use control in the field.

The same host’s Foundation fieldbus implementation requires users to pre-address devices (using a National Instruments configurator), not unlike Profibus PA.  While the facility was not planning to be the first large-scale industrial application of said supplier’s fieldbus implementation, the projects ahead of them in line were delayed or cancelled, and it became the unwitting vanguard. A number of bumpy years ensued. Life is OK now, but the memory of past pains remains pretty fresh.

The people with the longest memories and the longest careers in the local plant are the operators, first- and second-line supervisors, and their counterparts in maintenance. Many of these organizations are so conservative they put the kibosh on fieldbus, even when they’ve seen wireless in action in similar facilities or within their own companies!

Many of our operations “customers” are forward-thinking and willing to explore outside-the-box, but how do you suppose they’ll react when you tell them a critical measurement (or, even worse, a valve) is running on batteries and only waking up to communicate once a minute? I’m guessing any fast-talking at that point of realization is pretty useless.

We need to understand that we’re a link in the chain of delivering a service (measurements, information, control, automation, etc.) to our end users—mainly operations and the business itself. Operations takes a beating when the process goes down for no good reason. Appliances with the potential to put operations in that position, or have a legacy of doing so, get extra scrutiny. So until we retire a generation of plant operators, I’m betting that hardwired instruments are going to be around.

Even tank gauging, once seen as a “killer app” for wireless due to the long distances and vulnerable “aerial cable” required, may remain hardwired. Some hydrocarbon industry end users have done layer-of-protection analyses of their tank farms, and believe SIL-rated instrumented systems are required. Measurements communicated over radio waves may not measure up, even with redundancy.

If an ardent wireless true believer installs a system where only critical measurements and control valves have wires, he or she still ends up with a physical copper network connecting most areas of the physical plant. If this network is a fieldbus backbone, we need to evaluate whether the incremental cost to connect a given measurement to the wired infrastructure far exceeds the cost for wireless. When a fieldbus installation exploits IS or non-incendive wiring methods, the wire doesn’t even need to be in conduit. Hence the incremental cost of an added “wired” measurement is even less. SP100 has defined six classes of potential wireless applications, the lowest of which are “1— closed loop regulatory control” and “0—emergency action.”

Most of us will retire before operators are willing to run these classes of control without wires. Other classes are finding a number of applications. For those of us possessing or planning a fieldbus infrastructure for our facility, let’s hope that the incremental cost for wireless capability is based on scenarios like the ones described above, and based not the cost of running rigid galvanized steel clear from the control house.