Boiler - Circ Performance and Question

[jebatty]

First, measured performance, then the question.

My Tarm heats 1000 gal pressurized storage through a 5 x 12 x 30 plate hx, 1" ports. The boiler loop is very short, Taco 009, assumed 8 gpm flow.

The tank loop is about 60 feet, 3/4" line, Grundfos UPS 15-58F, 3 speed, with check valve. I have a high temp water filter, 50 micron, on the supply side to the hx.

I just installed a variable area flow meter, which measures 2 - 20 gpm, on the supply side to the hx. Measured flow rates with the Grundfos: Low - 4 gpm, Med - 5 gpm, Hi - 5.5 gpm. Since btu varies with delta T, at an assumed delta T of 30F, btu transfer through the hx is: Low - 60,000 btu/hr, Med - 75,000 btu/hr, Hi - 82,500 btu/hr.

Question: the flow chart for the Grundfos shows on Low 4 gpm = 6' head, which also means that on Med flow should be 9 gpm and on Hi flow should be 12 gpm. Why are the measured flows on Med and Hi so much less than the flow chart indicates? Is it simply related to the flow resistance in a 3/4" line? Will a higher head capacity circulator (Taco 0011 or 0013, for example) materially improve flow performance? To increase flow will moving to a 1" or larger line materially improve flow performance? Is the water filter likely to be the limiting factor on flow?

I do have the ability to temporarily remove the water filter, so I probably can answer the last question myself, but other input will be helpful. Thanks.

 

[Donl]

I found the same think when I measured flows on my 3 speed Grundfos. I think the reason for the discrepancy between the actual flows and the Grandfos charts is that the head loss does not remain the same across the three speeds. As the flow rate increases so does the head loss. Does that sound right?

Don

 

[DaveBP]

As the flow rate increases so does the head loss.

Exactly. The friction (head) in the system varies with the square of the flow rate. Double the flow rate and the head increases 4 times. I'd guess that pipes alone have so little friction that the effect is not so noticeable but with all the little microturbulence in a filter and flat plate HX ( all those plates are corrugated in there) the friction is up where it takes a big toll.

the flow chart for the Grundfos shows on Low 4 gpm = 6’ head, which also means that on Med flow should be 9 gpm and on Hi flow should be 12 gpm.

Pump curves alone are deceiving that way. If the head remained the same at the different speeds that chart would be right.

Jim, did you think about piping a pressure gage across the filter unit to get a head-loss-when-new reading so you could watch the pressure increase as the filter gets crapped up with sediment? And what flow meter is that and where did you get it?

 

[steam man]

I have the same issue with my solar loop. I do have a wye strainer and possibly some other flow restrictions I can change. I suspect the wye strainer to be the biggest loss. Now that the system is clean I probably can remove the fine strainer.

Mike

 

[jebatty]

...did you think about piping a pressure gage across the filter unit to get a head-loss-when-new reading so you could watch the pressure increase as the filter gets crapped up with sediment? And what flow meter is that and where did you get it?

I don't have a pressure gauge for this purpose, but I can bypass the filter and see what the flow increase is, and thereby also compute the head added by the filter. My current filter is only about two weeks old and should be pretty close to minimal obstruction. The flowmeter also will tell me when the filter gets crapped up by showing reduction in flow, so it will be a good monitoring device.

Got the flowmeter off of e-bay, after doing quite a bit of research. I got the Blue-White F-410000LN-16, 1", 2-20 gpm. It's rated only to 150F, so I have it on the return line, and it's not usual that return gets this high on my system. Probably some wiggle room here too. Will have to pay attention though.

Thanks for the "square" info.

 

[jebatty]

...did you think about piping a pressure gage across the filter unit to get a head-loss-when-new reading so you could watch the pressure increase as the filter gets crapped up with sediment? And what flow meter is that and where did you get it?

I don't have a pressure gauge for this purpose, but I can bypass the filter and see what the flow increase is, and thereby also compute the head added by the filter. My current filter is only about two weeks old and should be pretty close to minimal obstruction. The flowmeter also will tell me when the filter gets crapped up by showing reduction in flow, so it will be a good monitoring device.

Got the flowmeter off of e-bay, after doing quite a bit of research. I got the Blue-White F-410000LN-16, 1", 2-20 gpm. It's rated only to 150F, so I have it on the return line, and it's not usual that return gets this high on my system. Probably some wiggle room here too. Will have to pay attention though.
http://www.blue-white.com/products/VariableArea/F-400/acrylictube.asp

Thanks for the "square" info.

 

[jebatty]

For those using or thinking about using a hot water filter to protect a plate hx on their system, using a string wound 50 micron filter, 3/4" plumbing, I can report on the impact of such a filter on system head. Today, using the flowmeter as the gauge, operated a Taco 009 on the system side of the plate hx, both with a 50 micron filter that has been in use for 17 days and without the filter. With the filter flow was approx 5.8 gpm and without the filter flow was slightly greater, but not enough for me to accurately judge the gpm increase.

Conclusion: such a filter has a minimal impact on system head. I will continue to leave the filter in place and determine how much, if any, greater time in use impacts gpm.

Filter I am using:
http://www.filtersfast.com/Pentek-PD-50-934-Sediment-Filter.asp

 

[steam man]

For those using or thinking about using a hot water filter to protect a plate hx on their system, using a string wound 50 micron filter, 3/4" plumbing, I can report on the impact of such a filter on system head. Today, using the flowmeter as the gauge, operated a Taco 009 on the system side of the plate hx, both with a 50 micron filter that has been in use for 17 days and without the filter. With the filter flow was approx 5.8 gpm and without the filter flow was slightly greater, but not enough for me to accurately judge the gpm increase.

Conclusion: such a filter has a minimal impact on system head. I will continue to leave the filter in place and determine how much, if any, greater time in use impacts gpm.

Filter I am using:
http://www.filtersfast.com/Pentek-PD-50-934-Sediment-Filter.asp

I looked up the specs on that filter. It is good for a 9gpm flow rate at less than 2psi drop which is 4.6 ft of head. Seems like an oversized filter for a 3/4" line which is good for minimizing flow loss. I would say it would depend on the pump quite a bit depending where you are on the specific pump curves, i.e whether it is steeper or flatter. I know I am at the point where it drops off quickly with a minor increase in head. I will post when I get the strainer out to see if it does just that.

Mike

 

[heaterman]

I found the same think when I measured flows on my 3 speed Grundfos. I think the reason for the discrepancy between the actual flows and the Grandfos charts is that the head loss does not remain the same across the three speeds. As the flow rate increases so does the head loss. Does that sound right?

Don

You nailed it Don. Head is not linear in a closed loop especially if the loop is undersized. The more you try to ram through it-the higher the head becomes-the less GPM you move per revolution of the impellor.

acceptable flow rates for copper tubing
(roughly speaking)

3/4" = 4GPM
1" = 8GPM
1.25" = 15GPM
1.5" = 22GPM
2" = 40GPM

If you need 8 GPM that 15-58 will easily do it with 1" tube. The water filter is questionable IMHO.