Ventilation in natural gas compressor building

Status
Not open for further replies.
K

KentAT

Senior Member
Location
Northeastern PA
The company I work for has begun to reclassify the compressor buildings of our compressor stations from Class 1, Div 1 to Class 1, Div 2. This usually occurs when expansion needs arise.

I have always felt that classification of our areas fall under NEC, which points to NFPA 497, which recommends following the stricter guidlines of API RP500 due to the high-hazard criteria of our gas pressure which is >500 psig. Others do not always share my opinion.

The plan usually goes like this:
(1) Existing enclosed buildings are typically Class1, Div 1, inadequately ventilated.
(2) We have always had, and will continue to have fixed gas detection equipment which is installed for safety and other regulatory reasons that are exclusive of the API RP500 recommendations (but do meet or exceed them).
(3) Ventilation fans are typically installed to provide 6 air changes/hour when the fans are on. The fans are normally off. They are turned on when the gas detection system reaches a 20% LEL concentration. The powers that be consider this to meet the definition of adequately ventilated in 6.3.2.4.2 because this ventilation is providing for 6 changes/hour.

IMO it it is more important to be concerned with the gas reaching above 25% LEL as stated in the definition of "adequate ventilation" in 6.3.2.1. instead of just providing for the recommended # of air changes when a gas level is indicated. I think this is a gamble because you cannot predict the size or extent of the gas source (leak, etc) and you cannot know that the installed fans provide enough ventilation to prevent the accumulation above 25% until such time that an event occurs and the system continues to monitor it.

The "good engineering judgement" part comes in here because the policymakers know that our Emergency Shutdown Systems (ESD) will activate when the level reaches 40% LEL. At that time, the building and compressors are shutdown and all gas in the building is blowndown/released to atmosphere. I can see it their way if it were not for the fact that they and I all know that not all company compressor stations follow the strict company policies concerning bypassing of the ESD safety systems.

To me, the "good engineering judgement" would be to ensure full compliance with all company policies so this reclassification works, or don't do it.

FWIW I must admit that at first I was resistant to this change because I 'm one of the employees who might be inside this building when a gas leak (OK, "equipment failure") occurs and I want all the protection from accidental ignition I can get. But I do see the economic advantages and industry direction that points to these changes.

Ant thoughts?
 
If they are using API 500 as the basis, the ?powers that be? should read Sections 6.5.1 and 6.5.1a and its note more carefully.

If the gas sensors detect 20% LEL, it is likely it will be 40% within seconds ? even with good ventilation. The requirements of 6.5.2 are fairly onerous as are those of NEC Section 500.7(K).
 
I was thinking about this a little more last night. Have any records kept about how and when the automatic systems have operated?

My own personal opinion is that the system you described might at least be functionally acceptable from an API/NEC point of view if the facility is truly constructed to Division 2. Conversely, your environmental, safety and especially personnel health concerns are still quite valid with respect to potentially large releases in a confined space.
 
Last edited:
My concern always revolves around how many sensors there are and where are they located. For example, a 20% LEL at 30 ft above the floor will be much larger nearer the compressor or piping. Secondly, rate-of-rise should be given consideration when designing the logic, that is the only way to respond to the magnitude of the leak.

Turning-on the fans based on a 20% LEL is (IMO) sacrificing safety for economics. I would certainly think the "6 changes/hour" specification is for dispersion as well as evacuation.

In regard to your comment about bypassing safety systems, I believe that too often the design of the control-system itself invites abuse. Usually if there is a bypass incorporated in the design it is a "all or none". For example, to service/test a single component, the Tech must install a jumper, force an I/O, or activate a "system-wide bypass switch". We all know that jumpers and forces can remain hidden and be easily forgotten. IMO, safety systems should be designed for maintenance (with an audit-trail a mile wide) as well as functionality.

Another major concern is "how fast" can the compressors and piping be blown-down. Again, too often economics will be the determining factor for sizing the blowdown valve(s) and piping.
 
Have any records kept about how and when the automatic systems have operated?
Click to expand...

Yes, we have all records. Our systems are designed and have operated as designed to respond to 20% LEL, 40% LEL, flame, or heat.

Another disagreement is this:
Some of the decision makers feel that if the fans come on as designed at 20% LEL, and they are able to prevent the gas from reaching the 40% LEL blowdown setting which a big enough leak would do, then that's OK to remain in this mode as personnel figure out what's happening. (I didn't mention that 3/4 of the compressor stations are only manned during normal M-F 8-hr workdays and have a person on-call to respond to abnormal operations)

I, however, feel that (1) whether or not the fans are on, upon reaching the 20% LEL you now have knowledge that a gas leak is occurring in your Div 2 area and due to the pressure sources available (400-1200 psig, others at 50-100 psig) you best shutdown the place unless maintenance activities are known to be causing the elevated levels of a single sensor, such as opening up a compressor for normal maintenance work, since that is not a leak (2) Since no employee should be remaining or entering an area with a gas leak, why try to maintain the leak? They say for reliability, I say that shoudn't matter. (3) As stickelec points out, the reading is at the sensor(s), so of course there are areas much higher in gas concentration.

I feel that if the goal is to save money on installation costs, we better be prepared to act quicker and more prudent when gas is detected. As for the quality of following the wiring method rules in haz loc, we are very strict about them.
 
In the interest of full disclosure, I have always been philosophically opposed to gas detectors as a protection scheme for most classified locations.

When I was on CMP14, I accepted the three application schemes listed in 500.7(K) since, bluntly, they make very little difference. In two cases [500.7(K) (1) & (3)], the construction is still ultimately Division 2. In the other case, [500.7(K) (2)] since the “source” is Division 2, there is already a very low probability of flammable ingress. By the time all of the other requirements of 500.7(K) are met it is usually more economical to simply construct properly in the first place.

It should be noted there are NO other recognized applications for gas detectors than those listed in 500.7(K).

When I was on the CMP my “affirmative” comment noted that using gas detectors does not actually change the electrical area classification, it simply permits reduced installation requirements. One of my colleagues disagreed with that, but it doesn’t take too much analysis of the text to see my assertion is correct.

I also suggested the originators of the Proposal request that the Standards Council create a new committee project for gas detector applications.

I have attached a copy of the 2001 ROC Comment 14-51 CMP Action.

Recently, one of my fellow moderators in these forums (Don Ganiere) pointed out most flammable materials that require Class I classification are already toxic at levels below 20% LEL. So now, I am even more philosophically opposed to gas detectors as a protection scheme for most occupied classified locations.
 
KentAT said:
Yes, we have all records. Our systems are designed and have operated as designed to respond to 20% LEL, 40% LEL, flame, or heat.

Another disagreement is this:
Some of the decision makers feel that if the fans come on as designed at 20% LEL, and they are able to prevent the gas from reaching the 40% LEL blowdown setting which a big enough leak would do, then that's OK to remain in this mode as personnel figure out what's happening. (I didn't mention that 3/4 of the compressor stations are only manned during normal M-F 8-hr workdays and have a person on-call to respond to abnormal operations)

I, however, feel that (1) whether or not the fans are on, upon reaching the 20% LEL you now have knowledge that a gas leak is occurring in your Div 2 area and due to the pressure sources available (400-1200 psig, others at 50-100 psig) you best shutdown the place unless maintenance activities are known to be causing the elevated levels of a single sensor, such as opening up a compressor for normal maintenance work, since that is not a leak (2) Since no employee should be remaining or entering an area with a gas leak, why try to maintain the leak? They say for reliability, I say that shoudn't matter. (3) As stickelec points out, the reading is at the sensor(s), so of course there are areas much higher in gas concentration.

I feel that if the goal is to save money on installation costs, we better be prepared to act quicker and more prudent when gas is detected. As for the quality of following the wiring method rules in haz loc, we are very strict about them.
Click to expand...
My next curiosity, how many times did they operate at 20% LEL and not also operate at 40%.
 
First, thanks to both of you (and others as future posts might show) so much for the discussion. I don't think there is anywhere else on the internet nor industry publications where such an informative and frank exchange of worthy ideas and viewpoints can be found. As you can see by my posting history I just found this forum two months or so ago and have found the members, viewpoints, and explanations to be outstanding. (I only hope I can add an occasional comment now and then!)

My next curiosity, how many times did they operate at 20% LEL and not also operate at 40%.
Click to expand...

I don't know the number of events of the other stations, but my station has one medium sized building (60' x 180') of original construction and maintained as Division 1, and a newer building (60' x 50') originally constructed as Division 1 with original vent fans that has been since lowered to Division 2 as part of a recent controls upgrade so new equipment can be installed in the room as Div 2.

I think your question asks how many times the gas level had reached 20% without reaching 40%:

In our larger, unventilated building, we have the occasional (1 or 2x annually) 20% level reached such as when the building windows etc. are closed up over the weekend. Minor leaks on the compressors, fuel valve packings, etc. (engine/compressor vibrations take a toll on the packings) lead to an alarm level of 20% being reached due to the relatively stagnant air currents inside the buiding but never leading to 40% LEL. That's because we have quite a few sensors installed and the one over a leak picks up the small amount of gas. This scenario strengthens the decision makers' position.

In my smaller building (built Div 1, now Div 2 going forward), and a similar building of a station near mine, we have on average 1 high-pressure leak per year. The most recent ones were due to stainless steel tubing fitting leaks (high pressure gas). Again, normal vibrations were the cause of these. In all cases, the fans came on as designed. In all cases, the on-call personnel arrived at the buildings and either (1) manually tripped the safety systems to initiate a shutdown/blowdown before the automatic level of 40% was reached, or (2) 40% LEL was reached and the systems initiated the shutdown systems.

As far as using or not using gas detection systems for classification criteria, they are not primarily installed as a means to lower an area's classification. They are installed to protect employees and equipment and to comply with DOT regulations. Apparently, their existence is being used as a tool to save money in electrical materials.

My view is this. Look, you need a fuel, air, and a source of ignition to start a fire or explosion. We know we have a lot of fuel all the time. Why should we be considering playing the money game with the ignition part of that triangle? Build it to Div 1 and feel safe that in the event of the inevitable leak the ignition source due to the electrical system will have been minimized (the spark plugs on the gas-fired engines are another matter altogether - accepted as a risk that cannot be mitigated by the industry)
 
--- rbalex -

I could not open the attachment. The error message reads "Your user account may not have sufficient privileges to access this page. Are you trying to edit someone else's post, access administrative features or some other privileged system?"

How do I get to read the document?
 
I doubt seriously that designing to Div.1 requirements could be supported against a group of professionals equally intent on justifying a Div.2 classification. I being a "belt and suspenders" type guy, do agree that it would be advantageous, but I too am often overridden during the design-phase of a project in regard to (what I consider) safety related design.

Kent, in regard to sources of ignition, you mention that the engine ignitions are a serious risk regardless of the electrical wiring methods and controls. I agree, but another equally hi-risk source of ignition is the Turbocharger. Many years ago on two occasions I encountered fires (although small and isolated) in compressor buildings on 8GTLA White-Superior turbocharged engines. The fires were on the Wastegate Actuators which were on Instrument-Gas (BAD! BAD!) and located above the "cherry-red" piping. The seals on the Actuators failed, and the Instrument-Gas was ignited by the hot pipe. The point being, I think the presence of "an ignition source" should be considered as commonly present and every effort should be made to provide for reducing the potential fuel source.

Herein is were I would inject the need for: (1) Gas Detectors located not only in the upper parts of the building, but also above each compressor-cylinder. (2) Locating the unit suction & discharge valves immediately outside the building, not after the discharge-coolers which only increases the volumes required to blowdown. (3) Sizing the blowdown valve and piping sufficiently to provide for a reasonable "blowdown time" (ie, 20-30 seconds). (4) Design Exhaust Fans so that one fan could be down for maintenance and the other three still provide the six changes/hour. (5) When sizing the Exhaust Fans take into account the potential backpressue when they must discharge against a 20-30 mph wind. (6) Standards-for-Installation including specifications for materials and workmanship.

There are many other variables that I would consider worthy of serious consideration, some of which are (1) Is there a Standby Generator with an ATS present to keep the fans operating during an outage? (2) Is the ESD control system redundant, is it Relay Logic or PLC and if so, safety-rated, etc.? (3) Does the design provide for bypassing at the component-level, and with an audit-trail? (4) If a bypass is active, is an alarm present indicating the bypassed component? (5) Is the audible Alarm for the Gas Detection system unique?

Anyway... enough for now, thanks
 
Actually, I?m thinking more in terms of adequately ventilating the facility, rather than building it to Division 1. When you read all the side requirements of a recognized gas detection scheme [one compliant with Section 500.7(K)], a primary/backup fan system with simple air flow detection is usually more cost effective and reliable from an electrical area classification perspective.

Of course, gas detectors may still be required for other personnel protection reasons, but those are outside the scope of the NEC.
 
KentAT said:
The company I work for has begun to reclassify the compressor buildings of our compressor stations from Class 1, Div 1 to Class 1, Div 2. This usually occurs when expansion needs arise.

I have always felt that classification of our areas fall under NEC, which points to NFPA 497, which recommends following the stricter guidlines of API RP500 due to the high-hazard criteria of our gas pressure which is >500 psig. Others do not always share my opinion.

The plan usually goes like this:
(1) Existing enclosed buildings are typically Class1, Div 1, inadequately ventilated.
(2) We have always had, and will continue to have fixed gas detection equipment which is installed for safety and other regulatory reasons that are exclusive of the API RP500 recommendations (but do meet or exceed them).
(3) Ventilation fans are typically installed to provide 6 air changes/hour when the fans are on. The fans are normally off. They are turned on when the gas detection system reaches a 20% LEL concentration. The powers that be consider this to meet the definition of adequately ventilated in 6.3.2.4.2 because this ventilation is providing for 6 changes/hour.

IMO it it is more important to be concerned with the gas reaching above 25% LEL as stated in the definition of "adequate ventilation" in 6.3.2.1. instead of just providing for the recommended # of air changes when a gas level is indicated. I think this is a gamble because you cannot predict the size or extent of the gas source (leak, etc) and you cannot know that the installed fans provide enough ventilation to prevent the accumulation above 25% until such time that an event occurs and the system continues to monitor it.

The "good engineering judgement" part comes in here because the policymakers know that our Emergency Shutdown Systems (ESD) will activate when the level reaches 40% LEL. At that time, the building and compressors are shutdown and all gas in the building is blowndown/released to atmosphere. I can see it their way if it were not for the fact that they and I all know that not all company compressor stations follow the strict company policies concerning bypassing of the ESD safety systems.

To me, the "good engineering judgement" would be to ensure full compliance with all company policies so this reclassification works, or don't do it.

FWIW I must admit that at first I was resistant to this change because I 'm one of the employees who might be inside this building when a gas leak (OK, "equipment failure") occurs and I want all the protection from accidental ignition I can get. But I do see the economic advantages and industry direction that points to these changes.

Ant thoughts?
Click to expand...

I have some flea thoughts:D , don't know if it suffice.

All the other ant:smile: thoughts have properly addressed the issue, so I just add my own musings.

If you use gas monitors to assure that there is no hazardous accumulation of ignitables I would recommend to classify those instruments as SIS(ANSI/ISA S84.01 ? 1996, ?Application of Safety Instrumented Systems for the Process Industries,? (S84.01)).

I would use the gas monitor to remove all power to the system, not just to "shut down" the equipment.

Consider that when you reach an ignitable mixture and detect it with the instrumentation you are under conditions of Class I, Division 2, as it was proven to be non-hazardous prior, so your installation for Div. 2 should be perfectly safe from the ELECTRICAL standpoint. The premise of Div. 2 is taht you have infrequent presence of ignitable concentrations due to process failure, etc. and that the likeliness of a Div 2 equipment failure at the same time, under which condition it becomes an ignition source, is unlikely.

Classification per NFPA 70 and 497 does not take into consideration other ignition sources. Of course that does not mean that the Safety Department should not address the issue and require additional methods of control that are more stringent the Electrical Classification itself and may be implemented by components of the electrical system.

I would recommend the maintenace of continous 6x airexchange to comply with the Code requirement, not only when flammables are detected. If the LEL 20% is reached I would initiate an alarm and remove the power at 80%.
 
weressl said:
If the LEL 20% is reached I would initiate an alarm and remove the power at 80%.
[/FONT]
Click to expand...

Except for the Emergency DC Lighting (C1D1), right? Can't leave a guy on top of an engine...in the dark. :grin:
 
Status
Not open for further replies.
Top