Sizing circulator

Terkuta

Hi!

   I am currently learning to do radiant heating systems. I am having trouble understanding how to calculate pump head for a circulator. 

   I am trying to size a circulator for a building that is 3 stories high and has radiant heat for each floor. 2 apartments per floor. It is a 2-pipe system with 2 inch copper pipes going through the middle of the house to the top floor. On every floor, there is a header branching out from the supply pipe, and 6 pex pipes, at 1/2 inch, going out to the floor, 3 per apartment and returning back into a header and into the return pipe. It is a zone controlled system with no additional circulation pumps added. As the apartment calls for heat, it opens the zone valve and lets hot water through that area and back into the return pipe. 

    So, how do I size a circulation pump for the main system ? Whenever you choose a pump on a manufacturer's website like Grundfos, they need two things, pump head and Gallons per minute. So if I have 18 pex lines, at 6 per floor, do I calculate the head for each pex line and add it up with the head of the copper main pipe? If the design flow is .75 g.p.m. per 1/2" pex, then it's 13.5 g.p.m. for all 18 zones. 

   The issue I have is calculating the head. How does it work? Because the head changes if half of the zones close. 

So I was wondering if there is someone who can explain that to me.

Jamie Hall

You really have to do the calculation in two pieces -- but that's not a big deal. Your maximum head loss in the system will be when there is maximum flow in the risers (part 1) -- so all zones open -- plus the head loss in the most restrictive (usually longest) loop (part 2) Add those two together and there you are.

The logic is this. The head loss in the risers is determined by the flow, and will be a maximum when the flow is a maximum. Any other condition will have less flow, and hence less head loss. The head loss through the radiant piping is determined by the flow in the pipes, and the head required will be that in the longest or most restrictive run. Other loops will have less head loss, and the flows will automatically balance so that the actual running head loss will be the same in all loops (less in the more restrictive runs, more in the less restrictive ones).

Now if you shut off all but one zone (doesn't matter which one is left running) the flow in the risers will decrease as will the head loss in the risers. The head loss through the loops will stay the same. The pump -- even a dumb pump -- will adjust to that all by itself (actually the flow through the pump won't drop by say two thirds -- you will actually be getting more flow through the zone which remains on than you did before -- but how that changes depends on the specific pump and is not, at least usually, a concern).

Ironman

Elevation does not factor in a closed loop. Like a Ferris wheel, what goes up is canceled by what comes down.

It’s the job of the fill valve to be set at a pressure high enough to lift the water to the top of the loop; it’s not the circulator’s function to do so.

Head is simply resistance to flow from the friction created by the water moving through the piping and its various components.

Terkuta

@Jamie Hall Thank you! I didn't the relationship between head and flow. The big thing that I got here is that head is not just a static number. You have to start with a given flow rate and the head is calculated from that, is that right? 

Jamie Hall

Terkuta said:

@Jamie Hall Thank you! I didn't the relationship between head and flow. The big thing that I got here is that head is not just a static number. You have to start with a given flow rate and the head is calculated from that, is that right? 

Got it. Furthermore -- the head produced by the pump (except :"delta P" controlled) also changes with flow. The more flow, the less head. The relationship depends on the pump model. At some combination of flow and head the pump is producing exactly what the system needs.