Head Loss Through Parallel Objects in a System
Head Loss Through Parallel Objects in a System
(OP)
I read that pumps in parallel increase flowrate but not head. Does this mean other components in a parallel branch of a piping system will not add to head loss anymore than one, or am I drawing the wrong conclusion. I am designing a system where three heat exchangers are in a parallel configuration. It seems like this should be calculated like resistors in a parallel circuit. My question is "What is the proper way to calculate head loss through parallel components in a piping system?"





RE: Head Loss Through Parallel Objects in a System
Ultimately, you want to have a single system cure of DP vs W for the parrallel HX's, and teh intersection of this system head curve with the pump curves will yield the balanced system flow and DP.
For each HX and its associated pipes and valves, assume 3 different flows and calculate the corresponding 3 different pressure drops for each of the 2 HX's. Draw these 2 curves of Dp vs W.
Now, add these 2 curves together as follows: for 3 different pressure drops, find the flow across the 2 Hx's, for example , for DPa, DPb, DPc, find :W1,a; W1,b; W1,c; W2,a ;W2,b; W2,c; where W1 is the flow acorss HX1.
The system W vs DP curve is found by the new curve drawn from the following 3 points :
DPa , (W1,a+W2,a)
DPb, (W1,b + W2,b)
DPc, (W1,c+W2,c).
This new curve is drawn and superimposed over the DP vs. W curve for the parrallel pump combination, and theintersection of these 2 curves is the "balance point". Of course,t ehre are certain pump curves that will not be stable , so that issue should be addressed first.
RE: Head Loss Through Parallel Objects in a System
RE: Head Loss Through Parallel Objects in a System
RE: Head Loss Through Parallel Objects in a System
RE: Head Loss Through Parallel Objects in a System
If each had three different resistances , i.e. 10, 20 and 30kPa, the pump would need to generate pressure to overcome only the 30kPA circuit. The other two circuits would need regulating valves to prevent excess water flow and short circuiting. In essence the 10 circuit would need 20 added and the 20 circuit would need 20 added to get up to the 30 figure.(worst case figure)
Friar Tuck of Sherwood
RE: Head Loss Through Parallel Objects in a System
In general , heat exchangers will have a heat transfer performance that varies with flow. If the fluid that passes thru the HX has a specific volume that varies significantly with temperature, then the theraml-hdraulic characteristic of the HX and the system needs to be determined. This procedure is consisitent with the simple calculaitons I presented earlier, but requires that teh HX performance be determined with each individual flow assumed passing thru it. The performance check is to determine :
a) heat transfer for each flow
b) change in fluid specific voulme and fluid frictionla pressure drop thru HX and piping
c) change in overall pressure driop thru HX and piping including gravity head and acceleration
These secondary effects actually govern the balanced flow determination for some types of HX's , such as boiler evaporator circuits where 2-phase flow occurs, or if passing thru a supercritical circuits's "psuedo-critical "point.
For such a static pressure drop calculation, there may be flow instability, so-called "Ledinegg" staic stability criterion. So,not ony can the pumps and system have instability based on the pump's characteristic, but also be unstable from a thermal-hydraulic chracterisitic of the HX.
RE: Head Loss Through Parallel Objects in a System