Intrinsic Safety - Is My Client Wrong?

Hello all,

This is my first foray into intrinsic safety, but I feel like I designed a solid system. Inside the hazardous area, the only electrical devices are reed switches for pneumatic cylinders and RTDs for temperature monitoring. I have these interfaced with Pepperl-Fuchs Isolated Barriers and RTD Repeaters, respectively, in a control cabinet outside the hazardous area.

There are some circulating baths (120VAC) just outside the hazardous area, as well.

Whenever a hazardous condition is detected, I designed the system to go into an e-stopped condition, where I'm isolating all power leaving the panel. The circulating bath power is isolated as well as the SMC valve bank outside the hazardous area. NO POWER LEAVES THE CONTROL PANEL.

My client is telling me that this is not enough. They want ALL power inside the panel killed, as well, to include the PLC and HMI. When I explained to them that there would still be live 120VAC in the panel, they responded with, "yes, but there's no current flow."

I further explained to them that current flow is not the greatest hazard in an explosive environment but arcing, which would still be a potential threat unless they isolated all power coming into the ENTIRE ROOM. I also explained that they would essentially be unnecessarily "blind" by killing all their controls.

They remain unmoved. I'm going to give them what they want, but I just wanted to get some second opinions. Is this overkill on their part, or is my thinking correct?

Thanks for your honest opinions. I'm always open to learn.

you have the hazardous area.
you go through a wall or barrier.
within 18" of that wall on the safe side you have seal off fittings that are filled with cotton and epoxy (that's the seal).
you then go to the control panel with the safety barriers.
the other side of the safety barriers are the standard control voltages.

in regards to the pumps, I am making a guess on this one, but I would say that that is not allowed. electrical motors in the safe environment can short out and conduct in the fluid you are pumping. So you will have to get someone well qualified to sign off on this (a PE).

in regards to the rules, see nfpa 70 section 500-? hazardous locations.

the customer is wrong, you do not have to kill power to the control panel in the safe area. You can use ( I advise against it) a purge unit that will kill the power until the panel has been purged out.

the whole purpose of an intrinsic barrier is to reduce the voltage / current so that even if a short occurred in the open, there would not be enough energy in the spark to create an explosion.

this is what I had to go through in a class 1 div 1/2 and class 2 div 1/2 workplace.

my 2 cents,
james

Thanks for your reply, James!

in regards to the pumps, I am making a guess on this one, but I would say that that is not allowed. electrical motors in the safe environment can short out and conduct in the fluid you are pumping.

Click to expand...

They are circulating baths that are simply heating and circulating water into the hazardous space. They are not circulating the hazardous chemicals, themselves. The circulating baths are outside the hazardous area.

the customer is wrong, you do not have to kill power to the control panel in the safe area. You can use ( I advise against it) a purge unit that will kill the power until the panel has been purged out.

Click to expand...

My thoughts exactly. The truth is, I don't believe they have even given the space an official classification. Judging by the process, the hazardous area would be a Class I Div 1 area, and the area outside would be a Class I Div 2, AT WORST, and only because some ignitable gases might enter the space during an accidental rupture of the hazardous area.

If they're going to be consistent and demand that all power inside my control cabinet be killed, then they would also have to retrofit everything in the room to follow the same standard. For Pete's sake, there's a fluorescent light fixture hanging immediately outside of the hazardous area that is MUCH closer than my control panel.

There are outlets between the hazardous area and my panel with test instruments and computer plugged into them.

It's the failure to see their own inconsistency in standards that drives me nuts. Since their request doesn't create a hazard, only an inconvenience, I'm going to give them what they want, but it's not going to be free.

From what you have written I agree with your thinking, and the customers approach is not an approach I would have taken.
The best explosion protection is where you go as far as you can to minimize the risk of an explosive atmosphere to occur in the first place. That should always be the first approach, and after you have done that you can start assesing the remaining risks that cannot be circumvented.

If there is a risk that the cabinet could ignite gas, it is placed in the wrong spot (in a zone).
It is not only arcs that can ignite a gas, but also heat.

Friedrich said:

If there is a risk that the cabinet could ignite gas, it is placed in the wrong spot (in a zone). It is not only arcs that can ignite a gas, but also heat.

Click to expand...

Precisely! Thank you for your feedback, Friedrich!

Has a process hazard analysis been performed? This is the owner's responsibility, but it will dictate details of your design.

It doesn't sound like it if you don't have an official classification.

If you believe that the process area should be Class I, Group D, Division 1, then your panel location is BAD. If there is only a wall separating your panel from the process area and there is a door in that wall, you are in a Division 2 area at best.

Division 1 operating areas are rare these days. Exposure limits for the operators tend to preclude them.

If an intrinsically safe system is installed properly there is no need to shut down the control system. The intrinsic safety barriers limit available energy to below what can cause ignition. But this is only if they are installed in a safe zone and installed properly.

Something you said concerns me...
"Whenever a hazardous condition is detected, I designed the system to go into an e-stopped condition, where I'm isolating all power leaving the panel."

How are you detecting a hazardous condition? It may be appropriate to E-stop the control system based on an LEL sensor. But you should first (at a lower level) sound an alarm to notify the operators in the room so they can take appropriate action to deal with the situation. You may also take action to minimize the hazard by increasing the airflow in the HVAC system to evacuate the hazardous vapor.

Mellis, thank you for your reply. You asked:

How are you detecting a hazardous condition? It may be appropriate to E-stop the control system based on an LEL sensor. But you should first (at a lower level) sound an alarm to notify the operators in the room so they can take appropriate action to deal with the situation. You may also take action to minimize the hazard by increasing the airflow in the HVAC system to evacuate the hazardous vapor.

Click to expand...

Yes. I am detecting with a DetCon LEL Sensor. I have a warning threshold alarm configured as well as the critical threshold alarm. At the warning level, an alarm is generated with flashing lights and a horn to give the operators plenty of warning (assuming they are in the room to deal with it).

Concerning the design of the machine, think of something like a paint booth, but smaller. Everything is enclosed in an area with a exhaust hood. To my knowledge, they have probably performed a process hazard analysis, but whether the area has been officially classified as Class I Div 1, I couldn't say. The panel is outside of this enclosed space, on a wall, in a laboratory with lots of other non-classified equipment such as test instruments and computers.

In the event of a hazardous condition, my panel would be the only thing in the room required to die, according to what they're asking.

It sounds like they are wrong, but they're also writing the check. Since they appear to be going ovrr safe, there is nothing really wrong with it. Put in a shunt trip breaker, when estop is detected by your plc, energize shunt thus killing yourself. The operators will then have to toggle the breaker to power back up. If this is a common enough occurance and ut effects productivity, you will probably get to charge them again in a year or so to undo it.

g.mccormick said:

It sounds like they are wrong, but they're also writing the check. Since they appear to be going ovrr safe, there is nothing really wrong with it. Put in a shunt trip breaker, when estop is detected by your plc, energize shunt thus killing yourself. The operators will then have to toggle the breaker to power back up. If this is a common enough occurance and ut effects productivity, you will probably get to charge them again in a year or so to undo it.

Click to expand...

It depends on if the panel location is truly safe. If it is not, tripping a breaker automatically is a dangerous thing to do.

My final analysis before I head out to the plant to make the requested modifications:

In the very same environment where my panel lives, there are fluorescent light fixtures, plugged-in computers, test equipment, and other non-intrinsically-safe devices that are all consuming 120VAC. These devices will not be switched in the event of a hazardous condition.

To my knowledge, they haven't classified the area in which my panel resides. It's definitely not Class I Div 1 -- at BEST it could only be classified as Class I Div 2, and that's a stretch. If they have classified it as Class I Div 2, then not only am I in trouble, but so is the client. I have installed a panel that is not rated for a Class I Div 2 area, and I have exacerbated the situation by punching non-sealed penetrations into it.

They are in trouble because not only are they allowing sources of ignition (the computers, test instruments, etc.) into a classified area, but they are allowing voltages and currents into this classified area that are not intrinsically-safe.

Either way, they are writing the check, and the check just got a little bit bigger to pay for the change order.

Thanks for all your input, gents.

the customer MUST tell you the hazardous chemicals used !
if he doesn't, the material you use in the area might react with those chemicals and cause a bigger issue! I designed an acid tank system and specified stainless 316 (expensive), the head designer changed the order to stainless 308 (cheaper). this tank lasted 1 week before the acid ate holes in it.
the boss yelled at me until I showed him the prints that required 316. he then yelled at the head designer. that cost us about $5k to fix.

in regards, to the pumps, I understand that they are in the safe zone, the coolant is NOT. if a short develops in the motor, you can cause an electrical current in the coolant itself and that current will be in the hazard area. if you drop a metal tool on the pipe / tank for example, the spark will not be good.
I would see if you can isolate the motor from the pump as best as possible.

james

Hazardous area design requirements get pretty obscure at times.

A common occurrence in wastewater treatment plants is the call for all explosion proof light fixtures and panels in digester areas. These also have a boiler with a pilot light. Never did figure that one out.

Friedrich said:

The best explosion protection is where you go as far as you can to minimize the risk of an explosive atmosphere to occur in the first place.

Click to expand...

Not to be pedantic, but the best explosion protection is to not install ignition or heat sources in the hazardous area in the first place. This is something a lot of people miss ad jump into hazardous area stuff.



The only thing you did not mention is if your IS stuff goes in individual cables per circuit or multicore into a junction box. You would have to take into consideration the protection of the circuits, but if this is not the case your approach seems solid.



However, I would also look at how they defined their hazardous areas and not go by their word without justification alone. I find that a lot of people are overly optimistic (read, don't want to change a lot of stuff) when it comes to the risk assessment that leads into the selection of Hazardous area equipment.