Though not all process analyzers require sample systems and shelters, many of them do need some form of infrastructure in addition to the sensor itself. This infrastructure, which often includes the sample conditioning system, normally requires a number of instruments to control flow rate, pressure drop, and in many cases temperature, with some form of heat tracing and hence local controller(s).
Of course, having this equipment in one place also means there's likely to be reliable power to ensure that everything continues to operate, or as a minimum, fail safely in the event of power loss. Since, as we've indicated before, reliable power is one of the challenges faced by wireless systems, the logical conclusion is we should take advantage of having power, and “somewhere” (process analyzer shelter or enclosure) install equipment to place access points in the same proximity as our process analyzers.
Installing a Wi-Fi wireless infrastructure to connect process analyzer nodes to a central control room, maintenance shop or local network connection via copper or fiber (when installed with cybersecurity and VPN/VLAN setup to support connectivity of wireless sensor networks and maintenance systems) will increase information related to these systems, which when used appropriately, improves system reliability.
[pullquote]Though it's possible to use the same Wi-Fi network to “tunnel” process values, creating a separate maintenance network outside the process domain for the process analyzers as a minimum, and perhaps other devices as well, will ultimately reduce system complexity.
Many facilities have concerns about the potential vulnerability of Wi-Fi networks, and therefore connect all their wireless networks to a demilitarized zone (DMZ), keeping it outside both the business and control domains. It's also easier to connect from the DMZ to either the business or control networks with any required data transfers, for example, the status to the control system, while sending the more detailed diagnostic information to the maintenance and enterprise resource planning tools.
Because process analyzers require a minimum availability of 95% to be considered reliable, it’s good to actively monitor the health of the systems; for example, reagent levels for wet chemistry analyzers, dirty lenses, pressure or flow rate out of range, and other system or measurement-specific faults. All this information would typically reside above the control system layer in either the aforementioned maintenance or enterprise planning software suite.
There's also concern among some practitioners that capturing all the diagnostic information in the control system and transporting it across the networks may impact reliability of the control signals themselves because of bandwidth constraints. This, however, is somewhat of a red herring because Ethernet- and fieldbus-based systems have quality of service (QoS) capability to prioritize messages, so the control signals will always be delivered. QoS, however, is a subject for another column.
The maintenance software is also a logical place to incorporate quality control processes and charts to verify the integrity of analyzer signals. Basic statistical quality control (SQC) techniques are often used with process analyzers. This is because experience has shown the folly of changing the calibration every time the reference fluids show a difference, rather than verifying that the calibration has changed by more than could be expected from normal signal distribution.
If the calibration is adjusted every time, control actually deteriorates due to variation in the base line itself. Automated tools in the maintenance system can manage timing of the auto-calibration/reference check and recording, then as required, report the results or, if required, generate a work order for inspection to determine why the calibration requires adjustment.
Once installed, this same wireless infrastructure can be used for other maintenance and operation-support activities by providing access anywhere in the facility to suitably hardened computers and tablets, so technicians can see not only the equipment information in real time, but also associated work practices and reference materials such as user manuals.
Connecting process analyzer systems makes sense for more than making information available to increase reliability of the analyzers themselves. The resulting infrastructure for the analyzers also inherently has the required elements to provide an extended reliability network.