Bob,

Your conclusion turns out to be right, even if your method is not quite.

Remember 9.3.6.5.1 - you are only required to use the largest collection of interconnected vessels for the calculation.  The total quantity of gas stored in the room is not relevant any longer.  

In your example, I assume you have three manifolds: a 12x12 N2, a 12x12 N2O and a 6x6 CO2.

The largest collection of interconnected vessels here is one side of one of these (the manifold controls can be taken to "interrupt" the interconnection for the purpose of this calculation).  Clearly, that is the Nitrous, at 12 x 567 = 6804.  6804/5=1360.  So I need a "motivator" (fan, blower or suchlike) with a capacity of 1360 cuft/min, but since the rule in 9.3.6.5.3.2 is that I never need more than 500 cuft/min, that is where I end up.

Yes, 2500 cuft does equal 500 cuft/min, but be careful how you get there!

If the limiting factor is the maximum flow of 500 CFM though an exhaust duct then the intent could be to install multiple exhausts which add up to the required exhaust flow.  I hope Keith Ferrari comments - I believe he spoke with someone who was knowledgeable regarding the intent of this ventilation code.

well omg / I fully understood the requirement for determining the volume of gas within the room ( based on the section # 9.3.7.5.1 that steve listed ):

But when it can to exhaust rate / volume, I was lost. So the answer can always be 500. I just need to pay attention to the make up air. Is that right?

Steve Bradshaw wrote:

Based on the words "common manifold" in this code, I think many AHJs would base the the exhaust flow on the whole manifold.

 

Reference:  NFPA 9.3.7.5.1* For the purposes of this section, the volume of fluid (gas and liquid) to be used in determining the ventilation requirements shall be the volume of the stored fluid when expanded to standard temperature and pressure (STP) of either the largest single vessel in the enclosed space or of the entire volume of the connected vessels that are on a common manifold in the enclosed space, whichever is larger.

Steve,

I think you are correct, AHJ's are rarely trained enough in medical gas to see the fine distinction between largest volume of connected vessels and "whole manifold", but that would be a misinterpretation.  Since the concern is how much gas could reasonably be expected to discharge at one time, the worst case as a practical matter would be something like a pigtail breaking and a single bank emptying.  Failure inside the common controls would of course be worse, as it could discharge both banks, but that is why we have all the relief valves piped to outside.