By placing the pump on the supply header you add to the static pressure and this reduces entrained air and noise, reduces the possibility of cavitation, and prevents air from being drawn in through the air vents. The expansion tank should be piped close to the suction side of the pump
The pump is located at the ground. The typical max. working pressure of the expansion tank is 150 psig considering the static pressure elevation which at 140 psig, is it safe to locate it on the ground with the pump.
Safe to locate on the ground provided you specify the pressure rating of the tank (it will be an ASME vessel at this pressure).
I totally agree with imok2 that the correct location for an expansion tank is at the suction of the pump... however in a tall building the tank will be much smaller is located at the top of the system. How much smaller (and if it is worth it) will be determined by sizing both scenarios.
Having said that, you need a place at the top of the system to locate the tank.
The best place for an expansion tank is where the system pressure is the lowest. This not only minimises the tank size, as suggested by Chris, you can avail full static head in the suction if yours is an atmospheric tank.
I suggest you to read the excellent article Understanding Expansion Tanks by Steven T. Taylor appeared in March 2003 edition of ASHRAE Journal.
Quark, isn't the pressure lowest at the pump suction in a closed loop piping system? With this pressure~pump relationship, may the design parameter of concern be where the pumps are located while maintaing a NPSH greater than the pump NPSHR?
The pump suction pressure equals the static head available at the pump suction - the frictional losses in the suction piping at the given flowrate. Now, start from the pump suction and trace back the suction line. At every position, the static head decreases as the elevational difference is reduced and even the frictional losses get reduced (as the pipe length is shortened).
So, at some point, the difference becomes zero (in the absence of an expansion tank and if you ignore temperature effect on the fluid) and this is your low pressure point. Most probably, this point will be where the suction pipe takes a downward turn towards the pump.
Generally, when we design a pump for closed system, the static head is considered to be zero and only dynamic head +pressure drop across equipment is calculated. This may indicate that, as the length of pipe increases, the pressure drop increases and so, the pressure will be the lowest at the pump suction. This is a wrong assumption as the pump pushes the fluid against a static head in the discharge leg and suction static head supplements the pump.
BTW, we need worry about NPSHa in a chilled water or cooling water circuit. This needs to be checked for a hot water circuit depending upon the temperature.
