Taking credit for a building in a WCS under RMP

I have touched on this topic in some of my other writings and have received many e-mails inquiring about my statements.  First let me state that this impacts ALL industries, but it appears it has struck a nerve in the NH3 refrigeration industry, so my response will be specific to this industry since 100% of the my e-mails came from clients and colleagues in this industry.

First thing that has to be understood is the principle behind why EPA made mention of this consideration.  Ammonia boils at -28°F.  If you have several hundred to a couple of thousand gallons of this chemical in a pressure vessel being stored as a liquid under pressure, when this vessel “catastrophically fails” and this liquid (BP= -28°F) goes to atmospheric pressure instantaneously, there will be a considerable pressure rise within the room where the incident occurred.  Understanding that in MOST cases, it does not take a huge pressure wave to do significant damage to structures.  In a 2011 post regarding vapor cloud explosions I posted some data that can be applied to this “catastrophic” vessel failure, as we are talking about “destructive pressure waves” in both cases.  You can see below just how little of a pressure rise can do to structures, especially windows and roll-up bay doors!

So as you can see it does not take much to blowout windows and common construction of walls.

Second and this is from EPA’s guidance to refrigeration facilities concerning their off-site consequence analysis:

Release Rate – Mitigated Releases: The rule allows you to consider passive mitigation in estimating the worst-case release rate. Figure E-1 displays the procedure to be followed to determine the release rate for the worst-case scenario. If the release takes place in a building, the building can be considered to provide passive mitigation, unless:

• • The building may fail as a result of the release. This is unlikely except in the case of a large vessel in a very small room. As a rough rule of thumb, if the room volume (V) divided by the quantity of ammonia (Q) in the vessel is less than 0.1 ft³/lb, you should look at the possibility that the release of ammonia will cause failures such as windows blowing out or doors blowing open.
  • The release takes place facing an opening in the building (door or window). In this case, you should assume that the door or window will be open, and the ammonia will be released through these openings.

If the building may fail as a result of the release, estimate the release rate as for an unmitigated release (Equation 1, QR = Q/10 lb/min). Similarly, if the release would take place facing doors or windows, the release rate is again the entire inventory uniformly distributed over 10 minutes (Equation 1).

If the above conditions do not apply, you can assume that rain-out of liquid droplets is facilitated by impingement on surfaces (in a compressor room, for example), and only a portion of the released material will become airborne. The remainder collects in relatively slowly evaporating pools and makes only a small contribution to the rate of release from the building. To estimate the mitigated release rate, assume the following:

The amount of material airborne in the building is four-tenths of the total inventory, or 0.4 Q.

-The airborne material includes 0.2 Q vapor and 0.2 Q liquid droplets.

-Exhibit E-1 provides factors for estimating the mitigated release rate from a building. To estimate the release rate using these factors, do the following:

-Estimate 1 as follows:

-Determine room volume, V, in ft³

– Calculate 1 from room volume divided by the quantity of ammonia initially released as vapor, or 1(ft³/lb) = V/(0.2 Q)

– Determine the active ventilation rate, N_v, in room volumes exchanged per hour (hr^-1), for the building.

– From Exhibit E-1, find the 10-minute building attenuation factor, FR₁₀, corresponding to your estimated 1 and the ventilation rate, N_v.

– Estimate the release rate in lbs/min from the building attenuation factor and the airborne quantity (0.4 Q) as follows, assuming the release takes place over 10 minutes:

QR_B = (FR₁₀ × 0.4Q)/10 (2)

Example 1

A high-pressure receiver containing 5,000 lb of ammonia is in a room of dimensions 20 feet x 50 feet x 30 feet = 30,000 ft³. Hence, 1 = 30,000/(5,000 x 0.2) = 30 ft³/lb. The nearest value of 1 on Exhibit E-1 is 1 = 25. The ventilation rate for the building is 5 hr^-1. For 1 = 25 and N_v = 5, FR₁₀ = 0.35, and the release rate to the atmosphere is QR_B = (0.35)(0.4)(5,000)/10 = 70 lb/min, using Equation 2 above.

Example 2

The 5,000 lb vessel in Example 1 is outside. The release rate is, therefore, 5,000/10 = 500 lb/min. It can be seen that the building provides extensive attenuation. However, to take advantage of this potential attenuation, you must be certain that the worst-case scenario cannot occur outside or adjacent to a door or window that may be open.

Thirdly, we have to consider if we have windows that can be blown out when the WCS occurs.  Not only windows, but large roll-up doors (metal and cloth) can even be a bigger issue.  If the room where your largest vessel has no windows or doors and is large enough and constructed such that is will be able to handle this increase pressure rise then more than likely your WCS in your RMP is just fine.  HOWEVER, if your room has windows (the more and larger windows makes this issue worse!), roll-up doors, etc. then it may be wise to review the “assumptions” that were made in the WCS.