How do I size a centrifugal pump for a cooling system when system curve has zero hea

[EleanorT]

I’m trying to size a centrifugal pump for a closed-loop cooling system, but I’m getting stuck on the system curve because my static head is essentially zero. My calculations for friction losses seem solid, yet the required pump head feels off when I look at the manufacturer’s charts.

[Evelyn_W]

I’ve run into that exact setup. In a closed loop the static head was basically zero, so the pump has to do all the lifting through friction. My hand calc matched friction losses but the chart didn’t line up at the expected flow. The culprit turned out to be the little stuff: valve losses and a bunch of short pieces of tubing that I had treated as negligible. When I included those, the system curve shifted and the pump head made a lot more sense.

[Scott.T]

One time I actually measured it. Put a flow meter in, cranked the pump, logged pressure drop across the loop, and plotted the real system curve. It climbed faster with flow than I expected, and the operating point landed near where the high head part of the chart sits, not where the simple friction calc suggested. Temperature and density changes mattered more than I thought.

[MiaJ]

I’m tempted to question whether the problem is the pump curve itself. Those charts assume clean conditions and a fixed density. In a cooling loop with heat exchanger, purging air and trapped pockets can make the effective head jump around. We saw a brief spike when air kept forming, then it settled after we bled the loop.

[Andrew.R]

I’m not fully sure what the real issue is, honestly. Could be the flow you’re targeting isn’t physically attainable with that loop, or the model is missing a piece. At one point I moved to a bigger pump just to have margin, then realized the curve still didn’t fit. Might be the wrong target flow or you’re chasing a point that doesn’t exist in your setup.