How to pressurize circulator pumps for an open radiant heat system

[rustylaurels]

We are installing a solar water heating system with a below-ground cistern to store heated water (an open system). We purchased Grundfos UPS15-58FC, 3-Speed Circulator pumps to send the water from the cistern to the panels, house, and barn and then realized that the pumps are supposed to be placed at least 3 feet below the lowest possible water point in order to work. Digging an outdoor location for the pumps seems like a bad option because of having to waterproof the space where they go and still keep it accessible for maintenance.

Now we're brainstorming options for what to do next and would like some advice. Should we:
- buy the expensive submersible pumps and put them in the cistern
- buy cheap submersible pumps for the cistern and use them in series with the indoor pumps (and if so, how can we calculate the amount of cheap pump we need to get the Grundfos ones working?)
- use two pressure tanks (about 20 gallons capacity) we scrapped from someone else's project to create a closed, pressurized system (and if so - how do you do this without the pressure tanks just running dry)
- some other approach?

Thanks!

[J.P.M.]

Maybe I've missed it or maybe I'm looking at a different Grundfos UPS15-58FC manual, but the one I'm looking through for that pump doesn't talk about the pump being located be located " 3 ft. below..." in order to work.

What I did find was an instruction to not locate the pump at the lowest point in the system "...where dirt and sediment may collect…". I also found a NPSH table.

Per your post, under "some other approach":

Pumps of this type and others may often/always need a certain amount of what's called "Net Positive Suction Head" (NPSH) at the pump inlet.

Reason: When running, the pump will create suction (lower pressure) at the inlet. Often that pressure will be less than what's called the saturation vapor pressure of the water or other fluid being pumped, particularly/usually if the water and it's warm/hot.

Long story short, when the pump is operating, the pressure at the pump inlet might be low enough to cause the water at the pump inlet and through the impeller to boil. Then, when the pump impeller increases the pressure of the water, the steam bubbles collapse (technically called cavitation). Not good and for lots of reasons, and something you want to avoid.

See the Grundfos operating manual for required NPSH as f(water temp.).

One way to get the required NPSH on a pump in an open (non pressurized) system if the static head (the vertical distance from the pump impeller centerline to the highest liquid elevation in the system) is insufficient to meet NPSH requirements is to use something called a stand pipe. If the pump needs, say 5 ft. of NPSH, a stand pipe that has at least a 5 ft. above the pump level may be sufficient.

Another way to perhaps reduce the NPSH requirements (although not recommended) is to use a larger than normal diameter inlet (suction) line to lower the imposed NPSH requirement ( that will lower the V^2/2g number - something called velocity head, and so the NPSH requirement, but that's usually of limited value and might not work at all because the pump inlet is a fixed diameter.

See plumbing how to books for details on the above.

[LucMan]

The first problem I see with that pump selection is that it is a cast iron pump used in an open system. A taco 0011 or equivalent would be a much better choice in solar systems as it is a high head pump. The total system head for the required flow rate would need to be calculated before final selection of a pump.
A stainless or bronze body pump is required in open systems.
Hopefully the cistern will be lined with plastic or rubber and insulated to prevent BTU loss.

[rustylaurels]

Hi, thanks for the responses.

@LucMan, yeah, a cast iron pump isn't ideal, but we went with it because it had the head requirements we needed to pump to the top of our panels. The cistern is lined with pond-liner and foam board insulation.

@j.p.m, the "Inlet Pressure Requirements" table at the top of page 2 of this Installation Guide (https://s3.amazonaws.com/s3.supplyho...FC-install.pdf) shows a 3 ft inlet pressure requirement or 1.3 PSI for a low-temp system. The text that caught my eye is underneath: "In a system open to the atmosphere, the required inlet pressure is the minimum distance the pump must be located below the lowest possible water level of the water source (tank, pool, etc.)." So I think you're right about this referring to the NPSH; thanks for giving that a term. We couldn't find any information online about how a stand pipe would help this - is there another term we should google?

Thank you both your input!

[J.P.M.]

Hi, thanks for the responses.

@LucMan, yeah, a cast iron pump isn't ideal, but we went with it because it had the head requirements we needed to pump to the top of our panels. The cistern is lined with pond-liner and foam board insulation.

@j.p.m, the "Inlet Pressure Requirements" table at the top of page 2 of this Installation Guide (https://s3.amazonaws.com/s3.supplyho...FC-install.pdf) shows a 3 ft inlet pressure requirement or 1.3 PSI for a low-temp system. The text that caught my eye is underneath: "In a system open to the atmosphere, the required inlet pressure is the minimum distance the pump must be located below the lowest possible water level of the water source (tank, pool, etc.)." So I think you're right about this referring to the NPSH; thanks for giving that a term. We couldn't find any information online about how a stand pipe would help this - is there another term we should google?

Thank you both your input!

See Wikipedia "Net positive suction head" for details about why pressure at the inlet to any plumbing system, particularly an open system is important.

Also note, and I don't know why, that the NPSH requirements for your pump is different for different Grundfos publications.

I should have been more descriptive about using a stand pipe as using one will require some parallel lines and a check valve or two to work properly. A bit complicated to get into here. Apologies.

What must happen for any pump application is the liquid pressure at the pump inlet must be less than the vapor pressure of the liquid being pumped. Imagine a centrifugal pump trying to lift water about 35 ft. from an open reservoir. As the pump sucks the liquid higher, the pressure at the top of the water column will decrease. At about 34 ft. of water column elevation above the reservoir surface or less (depending on the water temp.) the liquid at the top of the column will boil and the column height will drop.

Similarly, in real applications, the pump sucking action decreases the system pressure on the suction side of the pump. If that pressure decrease is large enough the liquid pressure at the impeller face(s) may drop below the "saturation" pressure required to boil the liquid. The common way to prevent that in open systems is to make sure the line pressure at the impeller inlet never gets sucked down to less than the saturation pressure of the liquid. One way to ensure adequate suction pressure (or "suction head") is to add additional head by decreasing the pump elevation relative to the tank (cistern) water level. One other way is to use a parallel stand pipe, but as I wrote above, it's a bit involved to write about. A third possible but no promises and probably impractical way, as I alluded to in my last post, is to use large inlet lines and keep them as short as possible. The impractical part of that is that the pump housing will negate most of what's called the "velocity head" advantage.

The best way to deal with your head requirement dilemma is to get a pump that meets your flow and head loss requirements to meet the duty while having low suction head requirements, and then keep its elevation as close to or below the bottom of the cistern as possible.

[Mike90250]

cast iron pumps in an open system, will fail from corrosion. The open system allows oxygen into the water, and causes the pump to rust. Because of the moving parts, the rust flakes are thrown off and fresh metal is exposed. In a short while, much less than the warranty period, the pump is dead and the damage is not covered under warranty.

[rustylaurels]

Thank you all for the helpful input. This has been a very tricky part of the project and there's a lot to learn about the requirements. That said, does this look like a workable pump (open system filled with water, height difference from the cistern to the top of the panels is 15+ feet): https://www.grainger.com/product/LIB...ersible-444A24. Here are the specs and installation guides: https://www.grainger.com/ec/pdf/444A24_1.pdf and https://www.grainger.com/ec/pdf/444A24_2.pdf.

Thanks again!

[LucMan]

I don't know how many panels that you are planning on installing, flat plate panels require 1-2 gpm, check the specs on your panels. The 1/2 hp pump you are looking at is way over sized, and use a substantial amount of power.
The delta through your panels should be 8-10 degrees. I have 72 sq ft of panels and circulate 2 gpm from my 0011 and the throttling valve is nearly closed.
I suggest you use 2 external pumps in series to achieve the flow rate and head requirements.

[NewBostonConst]

Once the system is purged the head pressure is only going to be the restrictions inside the plumbing. The height measurement at your flow doesn't mater. The pump only needs to be able to pump to your max height to get the air out and then the siphon effect takes over pump flow increases to the zero height flow numbers.

[J.P.M.]

I don't know how many panels that you are planning on installing, flat plate panels require 1-2 gpm, check the specs on your panels. The 1/2 hp pump you are looking at is way over sized, and use a substantial amount of power.
The delta through your panels should be 8-10 degrees. I have 72 sq ft of panels and circulate 2 gpm from my 0011 and the throttling valve is nearly closed.
I suggest you use 2 external pumps in series to achieve the flow rate and head requirements.

Yea, I'd agree with you that pump is probably way oversized, but I'd suggest we have no way of being sure of that with the information we have at this time about the system or project goals.

As you wrote, and among other things, I agree the OP needs to get familiar with calcing/estimating pressure drop through the system with some knowledge of what the system will look like and how much pressure drop all the various components will cause, and then size the pump to the application using the pump curves available for various pump candidates.

Seems to me this whole thing is starting to get more complicated than we've got information on the design to be able to deal with, starting with design goals and some information about duty.

As an example, putting a 1/2 horse pump that'll put out something like 30 or so GPM into what's probably a 3/4" line (from the other pump connection sizes) will produce a line velocity that may well cause problems from flow induced vibration or erosion fouling, not to mention pressure drop - to name a few. We don't even know the system duty, or the array size or layout and flow arrangement and other things.

Rustylaurels: IMO, and meant in a respectfully, helpful way, you're in over your head technically on this, and continuing on the way you're going without some help - and that means more than we can probably give in this forum format - I don't see this coming to a good end. Get some help, or get more education before you go further.

FWIW, a small journey sort of off topic: Seems to me you're in a place very similar to one I was in about 45 years ago both in terms of knowledge about, and immediate goals for a solar energy project not too different than the one you seem to be up to your best intentions in at this time.
At that time, I was seeking information about flow turbulence for a collector system I thought I was designing and sought help at the local university from some very educated engineer who suffered fools badly but had the professional charity to sit me down and insult me for about 20 minutes because, for starters, I didn't know what something called a Reynolds number was. I didn't care for his blunt honesty at first, but saw that he was absolutely right about my ignorance and the trouble it was causing. That was the cusp of a turning point for me. I returned to school part time, later full time, got another degree in mechanical engineering and started on a path that changed my profession to engineering. I also now know what a Reynolds number is. Point is/was, I didn't know what I didn't know and I was spinning my wheels with my ignorance leading to dead ends and creating more problems. I was clueless and didn't know it. You may be in a similar situation.