The design of a process Safety Instrumented System (SIS) might contain one or more hard-wired switches in order to manually actuate shutdown devices. The switches might be wired as inputs to the SIS logic solver (relay, PLC, TMR, etc.) or in parallel with it. Frequently the switches utilize 24V DC normally closed contacts.

As the trend toward consolidation of control rooms progresses, the manual switches may have to be located relatively long distances (1 km or more) from the logic solver or shutdown devices.

What are the control community's thoughts on different ways to implement manual switches without long cable runs, while still maintaining the reliability needed for an SIS?

 

> The design of a process Safety Instrumented System (SIS) might contain one or more hard-wired switches in order to manually actuate shutdown devices. The switches might be wired as inputs to the SIS logic solver (relay, PLC, TMR, etc.) or in parallel with it. Frequently the switches utilize 24V DC normally closed contacts. As the trend toward consolidation of control rooms progresses, the manual switches may have to be located relatively long distances (1 km or more) from the logic solver or shutdown devices. What are the control community's thoughts on different ways to implement manual switches without long cable runs, while still maintaining the reliability needed for an SIS?> How about a bus work system. ASI networks have E-stop systems that work over the buswork. In most cases, the decentral equipement needs other signals also which can be sent over the buswork. The only limitation is the network max. length. This structure also simplifies cabling.

 

Responding to John Mangoff's and Zan Von Flue's queries:

As the inventor of the only truly "virtual" digital signal conditioner, i.e., detection and diagnosis of open-circuits, short-circuits, and ground-faults on switched devices, I of course, would prefer the hardware approach. This technology has even been used on a 5 km circuit. This approach has often been used on ship loading or unloading SIS installations. The advantage of diagnostics usually, but not always, eliminates circuit redundancy. I won't comment on sensor redundancy. That's another can of worms devoted to another List - The Physics of...
Voting Systems!

I have also used both hard-wired and fiber-optic links, with and w/o fault-detection.

Choosing which technology to use depends on many factors. Some ROE's (Rules of Experience) you can to explore or comment on, follow:

DC over AC is preferred, regardless of voltage level. I recommend low voltage (24 or less) rather than high voltage voltage (110-125 or higher). AC can, of course, be used. But, depending on circuit length, capacitive coupling could prevent the circuit from responding properly. Normally, this phenomenon doesn't exist in new installations. However, it first appears as an intermittent problem that results from system aging because of water ingress or insulation deterioration. Eventually the circuit will fail to operate at all. Unfortunately, the increasing deterioration is not readily detectable. I investigated one installation where it only occurred after a heavy rain!

While appearing obvious, bus systems should be used only after all the "What if?" scenarios have been evaluated! Remember, an open-circuit will cause the system to react adversely, one way, while a short-circuit, another. And of course, there is the most prevalent failure-mode, the ground-fault.

Regards,
Phil Corso, PE
(Boca Raton, FL)