A few weeks back, an interesting question came in from a customer regarding an article that I wrote about intrinsic safety. Here's a link to the article, "Intrinsic Safety 101 - Hazardous Locations."
The customer’s question had a bit of backstory:
“Lately, we've been engineering several midstream (pipeline) facilities, both natural gas and oil. The compressor station buildings range from single to 4-unit stations. These stations are all classified as Class I, Div. 2, derated from Class I, Div. 1 due to the installations having multiple air exchange ventilation systems and maintained fire and gas monitoring systems. All control panels within the compressor buildings (unit control panels, and station remote I/O panels) are being built with Class I, Div. 2 equipment, methods, and ratings.
The unit control panels (UCP) for the compressors contain a PLC and all I/O modules to interface to the Caterpillar engines and compressors. The panels were designed and built by a company that is well-experienced in compressor control. Several were built to interface to the on-skid vibration, temperature, pressure, and other instruments.
We’ve decided to take on the engineering/design of the UCPs for the oil pumping stations along the pipeline. Like the gas compressor buildings, the oil buildings will house from one to three 3500 HP pumps and are rated as Class I, Div. 2, having a ventilation system and fire and gas (F&G) system. The oil pump UCPs will have much less I/O than their gas UCP counterparts and will be monitoring the pump and motor bearing temperatures, vibration, and lube oil system pressures, temperatures, and levels.
Looking at the gas compressor UCPs (designed by others), I noticed that they had installed IS barriers within their panel on the discrete (24 VDC) inputs from the on compressor skid I/O signals. None of the analog instrumentation or discrete outputs (i.e., 24 VDC solenoids) were protected with the IS barriers. This was puzzling, so I began reading and studying like a madman, trying to figure out why they saw the need for the intrinsic safety barriers within a Class I, Div. 2 panel wired to Class I, Div. 2 instruments (most are actually rated at Class I, Div. 1), all installed in the same Class I, Div. 2 area.
I've studied ISA12 (the TRs and RPs), the ISA book on HazLoc Instrumentation, the NEC, and everything else HazLoc related that I can get my hands on. I'm struggling to find anything that states that the IS barriers are a requirement. Most lean toward the fact that the intrinsic safety barriers are a very good idea and enhance the safety of the systems that they are installed in.”
The customer’s question revolved around whether IS barriers were required for instrumentation that was rated for the Class I, Div. 2 area. If so, should all the field devices beyond the dry contacts utilize IS barriers?
Here is my answer for this type of question:
“You need only use the wiring method for the classified area as designated by the plant's manager. Since the area was reclassified as a Class I, Div.2 area, then you need only use Div. 2 wiring methods as long as all of the equipment used is also rated for Class I, Div. 2. This includes the relays. Odds are that the relays were not rated for the Class I, Div. 2 area, so using an intrinsic safety barrier was the easiest solution for the previous engineering firm.
With that being said, as long as all devices are rated for the Class I, Div. 2 area, then you can get away without using IS barriers. They are not required. This includes the dry contacts, but you must make sure they are rated for the Div. 2 area.
Now like you said, and if want to be extra safe and add that extra layer of protection, I would use IS barriers on everything! However, all field devices must be IS rated, which might be the case anyways, as you had mentioned that the area was initially rated a Class I, Div. 1 area. Finally, if you decide to go this route, you must use IS wiring methods, even though it is only a Div. 2 area. I would suggest that you consult the NEC for the proper installation of IS wiring, such as separation of IS from non-IS wiring, vapor seals, etc.”