Recently I was asked to help an organization with their Line Break/Process Opening Procedure and Permit. Anyone who has been through this process quickly realizes there is ZERO guidance from OSHA or any RAGAGEP as to what needs to be in the program and most notably the scope of the program. This organization has several hazardous materials that will fall within the scope of this program, but the “oh no” moment came when we were explaining the scope and impacts with regards to the refrigeration process with the refrigeration personnel. The “deviation” I am going to lay out in this article is common in a lot of refrigeration systems that utilize a “relief valve header system” in the design of their relief systems; although, this deviation applies to ALL processes that utilize a “relief valve header system”. Here’s is the potential issue…
Most safety and process safety professionals can agree with the application of a line break/ process opening program; noticed I said most! The principle behind a line break/process opening program (and permit) is that whenever we “open the process” that contains a hazardous form of energy (e.g. toxic, flammable, elevated temperature, pressurized, low/high pH, etc.) we need layers of protection for the risk associated with opening this normally closed process. It is pretty clear that when we open a fully closed system (i.e. a pipe) that the program/permit apply to this task; however, ask your personnel if they follow their “line break/ process opening” procedures and/or issue a permit (if one is part of the program) when they change out a relief valve. Sounds like a simple question, but…
I will bet you that almost ALL of these operators (and even some PSM managers) will say “no” and their reason being is that the process is already “open to the atmosphere”. I will be the first to admit that if EACH RV’s discharged DIRECTLY to the atmosphere via its OWN vent line then yes, we have an “open process” so breaking the connection on the discharge side of the RV to the vent line which is open to the atmosphere would NOT necessitate a line break permit for this opening/break. So, Is a line break permit necessary for changing a Relief Valve that discharges into an “RV Header/Manifold” system?
Basically a relief system that incorporates a “header” into its design creates havoc in traditional practices, such as line break. Consider this…
I have to change out an RV that discharges into a common header/manifold that other RVs discharge to. For the sake of this discussion lets just say 12 other RVs discharge into this header (by the way this number could be two (2) and the same issue is at hand!). In order for me to remove the RV from the ANSI/IIAR2 required RV “tree” I close the 3-way valve below the RV tree and I have to break the RV discharge connection to the header. Once I have broken this RV discharge connection to the header, I have now essentially created a monumental POTENTIAL hazard, as the header and my opening combine to create a “conduit” for any other RV discharging into the header to allow the NH3 to exit right at my location.
There are some that will claim they have (and I am being a smart ass here) “trained their ammonia vapors to ONLY discharge out the header discharge that is 15’ above the roofline as they designed it to happen” but basic laws of physics will tell us otherwise. In essence any premature failure of any of the other dozen RV’s tied into the header or an actual lifting of an RV due to an overpressure event, you will be inundated with NH3 vapors and in the case of an overpressure event, you will find yourself in an IDLH atmosphere rather quickly. Just ask yourself, how much ammonia will be flowing through my RV and into my header when the RV is sized for 65 pounds/min. Granted some vapors will exit the normal discharge location, but the entire header becomes pressurized in this overpressure event (and even in the premature lifting scenario) and the gas will find the least resistant path. The opening your working at, along with the header, have now acted as a conduit and delivered the ammonia right to you and out your “break/opening”!
NOTE: For those arguing that this is a one in a million scenario… ask yourself how many RVs have you changed out BEFORE the mandated “no more than 5-year period” schedule? RV’s have been known to lift prematurely and this is NOT all that uncommon; so before we go off and make this one of these “this will never happen”, know this… this very scenario has happened numerous times in the chemical processing industry and the ammonia refrigeration industry is NOT immune to it happening in their RV design basis! (See https://www.technicalsafetybc.ca/sites/default/files/bp_5615109.pdf for an incident where a RV lifted prematurely)
Here are a couple of items to consider in your risk discussions:
1) is there an alarm for EVERY RV in your process that will sound in such a manner before an RV lifts so that a worker who has the header open can exit the danger zone before the RV lifts, charges the header with NH3 vapors, and an exposure event occurs?
2) does your header have any type of indicator to alert this worker that ammonia has entered into the header and is the placement of this device(s) such that it will provide adequate and timely indication so the worker(s) who are at the “break/opening” can take necessary precautions to protect themselves. PLEASE NOTE that an indicator located at the normal discharge location (e.g. the end of the vent pipe to atmosphere) will provide ZERO level of protection so please do not consider this indicator as a protective measure.
3) when working at elevated heights or in situations with restricted egress, what level of PPE would be needed to provide ample protection for a worker to exit the ammonia release point safely should an exposure event occur?
I want to warn you that the RV that lifts and charges the header does not have to be in a separate room from where the work/break is taking place – in fact, it could be the RV right next to the one that is being changed! What would be the consequences if the RV right next to the RV that is being changed lifts prematurely or during an overpressure event that the worker is unaware of because there is NO ALARM before the set point of the RV lifting? Throw in the fact that the worker may be at an elevated location with limited means of egress to safety and now we have a meaningful discussion about the risk of changing an RV on a common header!
Now in the design where EACH RV has a DEDICATED VENT DISCHARGE, I am PROTECTED from an overpressure event or a premature lifting of another RV because the ammonia from EACH RV is directed to a safe location and NOT into a header that I am working on elsewhere in the process; however, when all the RV’s discharge into a header, this header acts as a conduit that allows the ammonia gas from any other RV on the header system to be carried directly to me and the “process opening” I have created right at my workspace. Just something to consider when discussing the application of your line break/process opening procedure applies to your RV header. Don’t be fooled into thinking that your header is “open” like a dedicated RV discharge vent, IT IS NOT THAT SIMPLE! So ask yourself this question… when personnel are changing out an RV that is tied into a header, what isolation do we have against ammonia from BOTH DIRECTIONS?
NOTE: I am in NO WAY advocating this, but many facilities have found that isolation valves in their relief systems can come in handy. Now I have written extensively about car seal programs and the RAGAGEPs that allow for isolation valves in relief systems, BUT IT REQUIRES A GREAT DEAL OF PROCESS DISCIPLINE to manage this type of design and it can prove to be MORE dangerous when not managed perfectly. But this is why facilities have implemented these types of car seal programs, as not having these isolation valves in an RV system essential means the ONLY way to safely remove an RV from a process is to REMOVE the potential for an overpressure event (e.g. shut down the process, drain the vessel/pipe, relieve ALL pressures, etc.) which is OK if the business actually does this when a RV needs to be changed. Unfortunately, the common practice is to fool ourselves into believing that since the header discharges to the atmosphere that somehow there is no risk associated with working directly at an opening we have created in this shared header/manifold which is connected to a LIVE PROCESS.
So again, I encourage you to sit down with the personnel who do the changing of RVs and walk through your design and the practice and determine if your process lends its self to this type of exposure scenario. Without isolation valves to protect them from what could come at them from the header, it should be obvious that this level of risk is unacceptable, especially when we consider most of the RVs are in elevated positions with very limited means of egress should something go wrong. Lastly what level of PPE would be needed if a worker was trapped in a cloud of the HHC for an extended amount of time due to their lack of ability to escape the release point? Could an APR’s Protection Factor be exceeded? Should an SCBA be required for these occasions if no isolation and limited means of egress from an elevated working position?
I will let each of you answer these questions for yourself, but I encourage you to have this discussion with your personnel as I would bet it may come as a surprise to some.
Here are some code references to consider:
NFPA 55, Annex A Explanatory Material
A.7.1.5.5.7.3 Where multiple valves or pressure relief devices are connected to a vent stack or common header, the potential exists for one valve to discharge into the header or stack while personnel are performing maintenance on another valve or pressure relief device. One solution is to install valves on the inlet and outlet of pressure relief devices for maintenance. Closing a full-area stop valve on the outlet of a pressure relief device that is being replaced prevents the backflow of gas from another device connected to the same stack or header. The ASME Boiler and Pressure Vessel Code requires that full-area stop valves be locked open and provided with manual vent valves for maintaining pressure during maintenance operations. Further, the ASME Code requires that if the full-area stop valve is closed, an operator must be present at all times to maintain the vessel pressure within acceptable limits and must lock the full-area stop valve in the open position before leaving the station. See the ASME Boiler and Pressure Vessel Code for complete details and requirements.
Requirement from National Board (NB) 23, Part 1 – Installation
4.5.6 INSTALLATION AND DISCHARGE PIPING REQUIREMENTS
a) The opening through all pipe and fittings between a pressure vessel and its pressure relief device shall have at least the area of the pressure relief device inlet. The characteristics of this upstream system shall be such that the pressure drop will not reduce the relieving capacity below that required or adversely affect the proper operation of the pressure relief device. When a discharge pipe is used, the size shall be such that any pressure that may exist or develop will not reduce the relieving capacity below that required or adversely affect the proper operation of the pressure relief device. It shall be as short and straight as possible and arranged to avoid undue stress on the pressure relief device.
b) A non-reclosing device installed between a pressure vessel and a pressure relief valve shall meet the requirements of 4.5.6 a).
c) The opening in the pressure vessel wall shall be designed to provide unobstructed flow between the vessel and its pressure relief device.
d) When two or more required pressure relief devices are placed on one connection, the inlet cross-sectional area of this connection shall be sized either to avoid restricting flow to the pressure relief devices or made at least equal to the combined inlet areas of the pressure relief devices connected to it. The flow characteristics of the upstream system shall satisfy the requirements of NBIC Part 1, 4.5.6 a).
e) There shall be no intervening stop valves between the vessel and its pressure relief device(s), or between the pressure relief device(s) and the point of discharge, except under the following conditions:
…
3) A full area stop valve should also be placed on the discharge side of a pressure relief device when its discharge is connected to a common header for pressure relief devices to prevent discharges from these other devices from flowing back to the first device during inspection and repair. This stop valve shall be arranged so that it can be locked or sealed open, and it shall not be closed except by an authorized person who shall remain stationed there during that period of operation while the valve remains closed. The valve shall be locked and sealed in the open position before the authorized person leaves the station. This valve shall only be used when a stop valve on the inlet side of the pressure relief device is first closed.
API 576, Inspection of Pressure-Relieving Devices
c. If a block valve is not installed on the downstream side of a relief device discharging into a common header, a blind or other suitable means should be provided to prevent discharge through the open outlet pipe in case one of the other relief devices opens.
d. A blind should be inserted between the pressure relief valve and any adjacent upstream block valve before a pressure relief valve is inspected or repaired in place. If the pressure relief valve discharges into a common discharge header, a blind should be installed between the pressure relief valve and the header.