Selecting a solar water pump - new circulating pump technologies for solar hot water?
 

I'm just wondering if anyone has heard of or knows of any new technologies that are applicable to a solar hot water setup, mine to be more specific. The pump will be on a closed loop with polypropolene coolant circulating water from the panels to a heat exchanger in my hot water tank.

I don't think that there is a ton to be gained by having a variable speed pump, but I'm not 100% sure on that. Other than that, I know there has been a lot of development in circulating pumps the past few years, and I'd like to take advantage of anything that is out there.

Are you talking about a vacuum tube array?

I'm not sure why that would matter. Hotter temperatures? But, no, these are flat plate collectors.

Not much new in the pumping field that is better than a standard Grundfos pump unless you want to go DC. A variable speed control like the RESOL can give about 8% better output over a single speed system and can be used with any of the can type pumps.

All the systems that are not drainback will need propylene glycol as an antifreeze. Type of panel doesn't matter.

I really like the ElSid pumps, they are an induction drive, so the impeller is 100% sealed off from the electronics and they impeller is the ONLY moving part, I think they make both A/C pumps and D/C variable speed pumps.

El sid pumps are good but the system has to be designed to have low head pressures so depending on how your system is designed (length of piping runs, 90s and height of panels) it might be a good pump. An alternate to the el sid is the Thermo-Dynamics pump which is bigger and more robust.

Thermo Dynamics Ltd. - Solar Pumps

This pump will run off a 20w panel and needs no controller at all. You usually don't get a digital output for DC pumps so monitoring is harder. That may not matter tho....

Unfortunately, the el sid is way too small for my application. My panels are ~100 feet from the tank. I'm planning on running 1" PEX to my 200 sq/ft of panels. Combine this with the antifreeze which thickens up the water and increases head loss and this gives me a target flow rate of ~10 gpm @ 23.5 feet of head. Only the larger inline taco or grundfos pumps even come close to this.

I also emailed Gary at builditsolar to see if he knew about any new stuff I should look out for. He said there really wasn't anything. So, I'm looking at a Taco 0011 or Grundfos UP26-120U.

remind me of how many m2 of panel you are putting in? You only need .25-.6L/m2/minute so if you have 10 x 2m2 panels x .6L= 12L/m or 3.2gpm-MAX. Try the T-D pump with a 30w panel.

The other issue is max temp on the panels. Most PEX systems won't last very long unless you are keeping it really low temp (less than 180F) and I don't know the specs for the panels. If you are going to bury it you want to be sure.

200 square feet is 18.6 square meters.

To size my pump I've been following the pump sizing guide on builditsolar. He recommends .025-.075 gpm per square foot (which Heliodyne, a solar panel mfg also recommends). His testing shows siginificant gains to staying a little above .025.

Your recommendation of .6l/m/m2 max puts me at .015 gpm per square foot (unless my math is wrong which is quite possible).

What else would you recommend for piping/tubing?

I always use either copper with armafex insulation or the insulated flexi stainless steel. If you drop the flow rate, you can drop the tubing size from 1" to 3/4" without any drop in performance. You will lose a lot of electrical energy trying to run a 26-120 or a 0011 verses a DC pump or a variable speed AC of the correct size.

The more efficient the panel, the faster you will need to move the glycol to get the heat out. Your panels are probably not as efficient as a new german panel or the Heliodyne which is the same absorber as many of the german ones so i wouldn't be as concerned about maxing out on the flow rate. My info comes from Viessmann and from a number of textbooks I have from Europe and while the numbers are not that different, many manufacturers have a different idea about pumping. Some like higher temps and lower flows, others do not. We have 2 manufacturers here that do "microflow" systems which use 3/8" tubing and a high head pump (T-D is one of those) and T-D has a very high output for their packaged system.

I would use at most a UPS26-99 but the wattage makes it hard for variable speed controls which are often limited to 1A. Try the numbers with a UPS15-58 or a Wilo Star21. These are used in a lot of European pump stations. Gotta get to work, i'm late

Hi,
Just some thoughts on sizing the pump and supply and return line for this 200 sf collector.

Heliodyne recommends a flow rate between 0.025 gpm/sf of collector to 0.075 gpm/sf of collector.

From the point of view of thermal efficiency of the collector, within reason, more flow is always going to make for higher collector efficiency. More flow means the temperature rise over the collector absorber is going to be smaller, so the average absorber temperature is lower, so the heat loss out the glazing is smalerl, so the efficiency is better. From the collectors point of view, as far as I can see, there is no "ideal" flow -- the more flow the better the efficiency. Eventually you end up with a pump that requires a lot of power and flow velocities that are too high for copper, and I suspect that is what sets Heliodyne's upper limit.


Look at a Low and and High flow rate:
At the low end of this range -- say 0.03 gpm/sf or 6 gpm (for 200 sf)
At the high end of this range -- say 0.05 gpm/sf or 10 gpm:

Pressure Drops:
If the collectors are 100 ft from the house, that's 200 ft return, and maybe add an extra 50 ft for other loses (turns, valves, manifolds)... just to get a rough idea.

Pressure drop for 1 inch pex at 10 gpm is 32 ft of head, and at 6 gpm, 12.6 ft of head.
Pressure drop for 3/4 cpr at 10 gpm is 60 ft of head, and at 6 gpm, 23.6 ft of head.
Pressure drops from:
http://www.everhotinc.com/barrier-pex-tech-specs.pdf
Pressure Loss of Water Due to Friction in Copper Tubes

The 3/4 seems a bit small to me even at the lower flow? This is a very long pipe run.

Efficiency:
This page tries to give a quantitative answer to what happens to collector efficiency as you raise the flow rate. Its based on the logic listed above -- that is, more flow rate means lower collector temperature rise, which means lower heat loss and greater efficiency.
Determining Solar Water Heating Collector Flow Rate

The values in the table of collector efficiency vs flow rate are calculated using this collector efficiency calculator: Solar Collector Efficiency Calculator
The only thing that changes for each flow rate is that the average absorber temperature goes down a little as flow rate goes up -- this causes efficiency to increase a bit. The table assumes full sun, 40F ambient temp, and 100F collector input temp.

So, if you buy that argument,
Collector Efficiency at 0.03 gpm/sf is 49.4%
Collector Efficiency at 0.05 gpm/sf is 50.1%

Heat output:
Heat out for 200 sf at 0.03 gpm/sf = 29640 BTU/hr or 8687 watts
Heat out for 200 sf at 0.05 gpm/sf = 30060 BTU/hr or 8810 watts.

So, under full sun conditions, the higher flow rate produces about 120 watts more in heat output.

This 120 watts of increased heat out for the higher flow rate can be compared to the increase in pump power needed for the higher flow rate.
If they are close to a push, than I'd say go with the smaller pump.

Just looking at a couple of Taco pumps:
New Page 1
At the 0.03 gpm (6 gpm) and 12.6 ft of head
A Taco 008 is a good match and does the 6 gpm at 14 ft of head.
power consumption is 92 watts for the Taco 008.
http://www.taco-hvac.com/uploads/Fil...ry/100-1.9.pdf


At the 0.05 gpm/sf (10 gpm) and 32 ft of head,
The Taco 013 comes pretty close -- just a bit short on head at 10 gpm.
power consumption is 230 watts.
http://www.taco-hvac.com/uploads/Fil...ry/100-6.4.pdf

So, the pump power increase to go up to 0.05 gpm/sf is roughly 230-92 = 138 watts compared to the 120 watts for added heat output for the higher flow. So, it seems like the lower flow rate and smaller pump is the clear winner here?

This is the first time I've done one of these calcs where the higher flow rate did not win -- I guess because of the fairly large pressure drop along the long supply and return pipes.


I'd appreciate anyone's thoughts on these sizing methods.

Gary

Thank you very much for running through the calculations Gary. I think the only thing you forgot is that I'm running a closed loop system with probably a 50/50 mix of polypropolene, so that flow rate must increase to get the same heat out.

According to your site your recommend a 15% higher flow rate, and say your pressure head will increase by ~30%.

Wow Gary, great job!

Thanks for including all your references.

I know when I was trying to calculate why my loop field pump was using so much power, and what the fix was, I nearly went blind shifting from PDF to PDF, and trying to co-relate all the info.

There must be a spreadsheet or program that would inter-relate all those variables.

Best,

-AC

Gary, that is great. I just didn't have time to do the calculations.

What it does show is that the the flow rate is on the high side. Remember that the flow rate has a relationship to the amount of heat collected on the absorber so that is what your flow should depend on. As you don't know this, with this panel, I would start with a RESOL controller with variable speed where you can set the parameters. That way you will get the optimal flow rate.

Doax, you only need 40% propylene glycol not 50% so you pumping should be a bit easier.

Thanks for the help guys. I'll look into reducing the flow rate and using a smaller pump.

According to my NoBurst bucket, 50% polypropylene will give me 'burst protection' to -60F, and 'flow protection to -10F'. Its really not the clearest. Does that means its going to start turning to slush at -10F?

Hi,
Just wondering how you plant to insulate the 100 ft pipe?

Gary

Here is the Tyfocor propylene glycol report for Bosch (Buderus).

http://www.google.ca/url?sa=t&rct=j&...FqWeIw&cad=rja

It is a 2mb file and I didn't know how else to show it. I use this stuff almost exclusively and it is available from Viessmann or Buderus dealers as well as Calefi. Bar none, it is the best one out there.

If you have a way to make sure that the panels never get above 90C, I would say to use the PEX but there are dual PEX pipes available in a plastic case with urethane insulation meant for underground that you could get. It would ensure that you had minimal heat loss and keep it sealed. Remember that 100ft one way is a long distance.

viscosity for any oil gets higher when the temp gets colder but there are systems running in the Canadian arctic with 50% glycol. I have done a number of systems in Fort Francis, Ontario with 40% and AFAIK there has not been any issues.

How deep can you bury the piping and can you place a 2" sheet of SM over the tubing to allow the ground heat to come up closer to the tubing? It's just one way to help it out.

It will be surrounded by 3" of XPS similar to what you did with yours.



That can be done. I am building my own differential controller to control the system so I can do whatever I want.

I looked into commercial insulated options. IMO they don't offer nearly enough R value, and they're incredibly expensive. Most of the stuff I saw had only around R3 (where mine is R15), and it cost 3X more than what I am doing.

R3? Ive never seen one that low but if you are boxing it in that is fine. I would still put armaflex around it as well.

That sounds good -- hope you will take some pictures and post them.

Gary

There will definitely be pictures. I'm currently working on the trench and solar panel rack, and sharing my progress in this thread:

http://ecorenovator.org/forum/solar-...el-rack-7.html

I ran through the same calculations as Gary did a couple posts ago. I made a spreadsheet to compare the flow rates and different heads of pressure they create.

http://ecorenovator.org/forum/attach...1&d=1349466381

Looking through the available pumps I've come across a few that would work.

grundfos up15-100f 5.3gpm @ 1.1A (135W)
taco 009 gives me 6gpm @ 1.4A (161W)
grundfos ups26-99FC (speed #2) gives 7.5 gpm @ 1.5A (179W)

I'm still trying to dig through all the possibilities. However, I am leaning toward the UPS26-99FC at this point. Unless the power consumption ratings are way off, its very comparable pump wise to the UP26-64F, but the UP26-64F pulls 1.7A vs the 1.5A on the UPS26-99FC.

One thing I am thinking would probably be nice to have integrated into the pump is a check valve to prevent circulation when the pump is off. Its a relatively cheap option to add to the pump, and its one less thing I have to find fittings for and install.

The grundfos are all available with the check valve built in. If you get the Resol variable speed control, your pump will be running at much lower power 80% of the time and I really like the UPS26-99. it is a very versatile pump.

After a bit more thinking I think I'm going to go with the UP15-100F. I'll err on the cautious and cheaper side, especially since the gains for going to a larger pump really are quite minimal.

The one problem I'm having is finding the UP15-100F with the internal flow check valve. After looking at the grundfos spec sheet, I don't think I can actually get the pump with one, so I'll have to add that in as a separate item.

List price difference is about $120. I don't know what price you would be paying for it. A good inline check valve is probably $40+ but since you have such a long lineset that is not going up to the roof, you might be able to do away with a check valve altogether. You can use a heat trap to prevent any unwanted fluid movement.

As much as i like the 3 speed 26-99, it is too high an amperage for the variable speed controllers where the 15-100 will work with the RESOL controller so it looks like it is a good fit.

That is a good idea. Would you suggest a mechanical heat trap, or just the raised tubing type?

I would try one "U" on the supply going into the tank HX, 12" long should work, and one from the bottom of the HX as it is going to the pump. The second one may not be necessary but it is a cheaply made with a couple of 90s and could be added after if desired.

I ended up ordering the UP15-100F last night. The price was right around $140 plus about $7 for the flange kit. For more info on the installation see this thread:

http://ecorenovator.org/forum/solar-...anel-rack.html

where did you get the pump? The price is quite good.

I got the pump and plumbing items from pexuniverse.com.

It says Max. power input: 135 W , can you use half the amps for reduced consumption, if the pump is not required to lift the water 35 ft.
I'm curious what circulation rate gpm is used with a system like yours.
Seams like a powerful pump i guess you need the lift ?
With solar panels and a few batteries it could run it self.

A couple of pages ago, I suggested the Thermo-dynamics pump which is fed off a 20w PV panel. It is pricey but combines the variable speed with being totally off grid and a flow rate to match the hot water panels. He could have used 1/2" soft copper which is still more money than the pex but is easier to insulate and may have less heat loss than pex due to less surface area.

I have built many different types of systems from the micro-flow type which are popular here in Canada from T-D and another company (who shall stay nameless because I don't like them), to German system with high flow rates like Viessmann. According to the SRCC data, a T-D panel with a moderate selective absorber will be within 1-2% of a Sunearth panel with the same construction. The difference is the flow rate. The T-D is .4l/m/m2 and sunearth is .8l/m/m2 (These numbers are from memory so I may be off a bit). People concentrate on the panel and ignore the rest which is arguably more important.

The real difference is in how well you transfer that energy to the storage tank using a properly sized HX and matching the flow rates properly. It is almost like MPPT with PV, trying to find the best power output.

Since there are people doing Arduino and Raspberry projects on this forum, I think there should be solar controller project made for higher current pumps. There are controllers in Europe that take higher current but we tried to get on certified to a CSA standard a few years ago and gave up due to the pig headedness at CSA.

That said, I have been using a Taco 009 on drainback system with 4 panels with the RESOl controller (which is only supposed to take 1A) for 3 years now and the current draw is more than the Grundfos.

If you look at the chart, you'll notice if you don't have 35ft of lift you simply get a higher flow rate. The purpose of the calculations I've done is to try to find the pressure head that you have at the flow rate you want, and then find a pump that closely matches those two variables.

In the case of solar panels, greater flow means greater efficiency by lowering the solar panel operating temperature and lessening the heat loss to the atmosphere. However, at some point your pump starts using more power than you gain from the efficiency benefit you see from the greater flow rate. So, you have to play a balancing act when trying to size your pump.

The post that Gary made has links to his testing of systems that show what flow rates returned different efficiencies, so that is the data I used to size my pump. I'd be interested in seeing any additional information that Mikesolar can provide that shows that a pump powered via a 20W pv panel is going to provide enough flow to efficiently operate these panels.

At this point I'm just trying to get the panels up and running. Solar PV is great, but there are many other things I would like to do to my home conservation wise (insulation mainly) that would be money MUCH better spent. Perhaps some day!

What i meant was less volts not less amps, less volts would give you less amps and less flow, if you wanted a way to control the flow rate.
I must comment on the 20 watt pump, wow that is a tiny pump, a flow at 1/2 liter per sq m is required and the little pump does what 10 L per min?, how many sq ft room would the 20w pump be good for?, sorry my math is more then a little rusty so i'll leave it to those more adept at it..

Ok i just read some flow rates needed to move how many feet of water, i understand why you chose the 135w pump now ..

Attachment 2555

Here it is. I'm not against your choice of pump, just that this is also available.

As the speed goes up on a circulator, so to does the head pressure so it is somewhat self limiting but if it is really an undersized pump, that circ will just spin way too fast without doing much.

Sorry Mike, I don't see how that pump can get me up to the 5-6 gpm target I am looking to get. That flow chart only goes up to 3 gpm.

So, I've been brushing up on my hydronics lately because of the solar project and also installing hydronic floor heating in my office. I was reading John Siegenthaller's 'Modern Hydroinc Heating' and happened to be reading the chapter on pump selection. He mentioned that it is quite common to find that, going from a PSC (permanent split capacitor) motored pump to a ECM (electronically commutated motor) pump will double your efficiency! This isn't even taking into account the benefit of the variable speed advantage that the ECM motors provide.


My Grundfos UP15-100f uses a PSC type motor. With a current draw of 135W, I'd be very happy to see something in the 65-70W draw as a replacement. So, I shall be looking into what options are out there for ECM motor pumps.

In the book, he continues on to show that a 85W pump operated for 3500 hours per year with an electricity cost of $.14 per kWh (a bit high) will uses almost $1400 in electricity over a 20 year life. So, reducing that number in half gives quite a bit of room for cost savings in a pump even if the new pump costs substancially more.

I've started a thread about PSC vs ECM pumps here: http://ecorenovator.org/forum/applia...find-them.html

About the pump speed......if you have a 100mbtu boiler which needs 10gpm to work properly, would you use 10 gpm on a 50mbtu boiler? No, you would use a 5gpm flow rate to get the correct dT.

The same is true for a solar panel, if it is less efficient because the technology is older or it is a less efficient absorber, you choose a different flow rate. This is why i talked about the T-D solar pump.

I am putting panels on my wall which have a painted (Solkote II) surface which is not as efficient as a highly selective absorber and I am using the T-D pump at the lower flow rate. This will give the liquid a bit more time to take the heat out of the panel for a bit higher flow rate. I though you would benefit from this pumps as well given the panels you have.

How would you recommend figuring out the most optimal flow rate of any ol panel (be it new or old)? Could you make a setup where you put a thermal sensor before the panel and then after it, then adjust flow rate so that the temp delta was X degrees?