No one imagined how critical the medium pressure steam letdown control loop was until the safeties popped on night shift. The passive, spring-opposed vent valves protect everything from process vessels to your water heater at home. If you’re lifting them, it means the process has ventured outside the bounds below which the basic controls were meant to maintain it. The crew on shift was shocked to see that an 8-in. control valve through which many pounds-per-hour of steam had been flowing had gone to its no-power (fail closed) state. The process was stabilized once the valve was bypassed, and when the normal day shift maintenance folks arrived, they identified the cause: an accumulation of deposits from water vapor condensation in the positioner’s termination box had caused a short circuit.
Could this sort of upset-causing wiring defect be detected before it impacted the process? That’s one of the benefits we imagine are possible with “smart,” digitally integrated devices. But there are a number of potential issues that can evade even the most ardent observer of diagnostics and are best reserved for a process outage.
One of the most common is control-valve stroke testing. You may already have an extensive list of valves slated to be removed from the line and rebuilt during the shutdown. Potentially, some of these can be tested more thoroughly once the process is down. If the issues are minor or can be addressed “in place” with minimal repairs (replacing a terminal box, for example), maybe they can be cut from the always-overburdened shutdown scope of work. Conversely, valves that remain inline can be thoroughly checked to diagnose problems that could arise during startup and routine operations. In addition, this might be a time to improve the tuning of control valve positioners, which can be disruptive when the process is online. Modern valve positioners can have unique sets of tuning constants, so I’d recommend chatting with the gurus at the factory before getting too creative because destabilizing an adequately tuned positioner is worse than doing nothing.
It's also interesting to check whether instruments in the off-line process vessel or utility header come to rest in a reassuring state. For example, if you have a few dozen temperature sensors in a reactor, are they all reading ambient when the vessel is de-inventoried? Sensors that have gone adrift will stand out and can be replaced without any hardship on the part of the operations team. Do all the pressure transmitters on the steam header agree to your satisfaction? If your facility is in a cold climate, now might be the ideal time to refill impulse lines of instruments in steam or water service with a fresh antifreeze (e.g., glycol) solution of your choice.
Update firmware, verify redundancies
If you have components like managed switches, fieldbus I/O cards, or wireless gateways, a process outage is your best chance to ensure firmware is updated. If your gateways or cards aren’t redundant, connected measurements will go offline during an update. In the case of wireless, measurements will reappear only after the gateway has rebooted and the mesh network begins to rebuild.
Shutdowns can be your chance to run down other wiring issues like shields shorted to ground. Multiple trunk cables have been in the tray for years—even decades—and as solar UV radiation takes its toll, insulation degrades. Redundant or N+1 DC power supplies are designed to be hot swappable, but you may not know they have retained this capability until you take one or the other (of a redundant pair) offline. There’s never a better time to see if swapping a power module has any impact on devices or communications. If your network power supplies are long in the tooth, it may be the time for some preemptive replacements.
Processes meant to operate continuously may have you waiting years between shutdowns, so it’s good to have an evergreen list of issues that were addressed with first aid in the intervening years. Making the best use of these opportunities facilitates the controls specialist’s goal to minimize unforeseen process upsets and interruptions once the process is back online.
