Comments for Review of NuMI RAW system 

                           Oct  24, 2001

                   from    Russ Rucinski
                       ------------------------

The piping diagram conforms with the ISA piping standard and is readily understood by 
those familiar with process piping.
 
The cooling water to the adsorber pile shielding has multiple parallel paths (and 
parallel/redundant paths) without an obvious means to equalize or set desired flow 
rates to the modules.  It understood that the amount of energy deposited will vary 
among the different modules.  I further understand that efforts were probably made 
to design in the correct flow impedances.  It is worth designing in a convenient 
means of adding valves should flow balancing be necessary.
 
As mentioned by a different reviewer in the meeting, if the purpose of the expansion 
tank (at H5 on drg. ME-406229, rev.B) in the adsorber raw system is to separate out 
evolved gas, the flow should be in one end of the tank and out the other.
 
Alternatively, an experience I have had with gas in water systems is that the air 
collects in the top of the full flow filters on the suction side of the pump.  That 
is a bleed off point that we commonly use.  You may consider moving the expansion
tank at H5 to a large port connection on your filter at E5.
 
The back pressure regulator on the expansion tanks should be replaced by a manual 
flow control valve.  The back pressure regulator and supply regulator will fight 
each other.
 
You probably should have plans to have a gas analyzer connection to the vent line 
of your expansion tank to determine the level of H2 in the nitrogen.
 
I agree with your decision not to rely on flow measurement as an alarm or interlock point.
 
Your expansion tanks seem to fall outside of the scope of ASME boiler and pressure 
vessel code, but to save yourself time and energy with the safety committees, you might 
consider putting a diverter valve and an additional relief valve on such.  It will allow 
you the means to pull off the relief and test it periodically.
 
I think the connections to the ejector pump may be incorrect.  The horn drain should go 
into the throat of the ejector pump.  The pump flow goes through the venturi.
 
Submitted by Russ Rucinski, NUMI reviewer