Heat transfer from roof loop to tank below expectations.

[LAWindsurfer]

The solar hot water system shown in the attached figure 2 has not been yielding the gas savings that I expected. The system consists of six, EC40 roof-mounted solar panels, two storage tanks plumbed in parallel with the output of those feeding a gas-fired water heater. The system uses heat exchangers that are built into the Rheem storage tanks. It is a one pump system with a flow back tank for freeze protection.

It seems to me that natural gas consumption for boosting water temperature to 120 oF could be reduced by improving the heat transfer between the circulating collector fluid and the water in the storage tanks. Early in the afternoon on a typical bright sunny day, the temperature of the circulating collector fluid is in the 170 oF to 190 oF range, while the temperature of the water in the storage tanks is typically about 110 oF . The temperature of the inlet water from the street is approximately 70 oF. If hot water consumption is low, the temperature of the water in the storage tanks will occasionally reach the 140 oF range after many hours. If a washing machine is being run, the solar system doesn't seem to be able to keep up with the demand even on a bright sunny day. More specifically, when the temperature at the bottom of the storage tanks is about 105 oF, the difference between the temperature of the collector fluid coming down from the roof and the temperature of the collector fluid being fed back to the roof is about 8 oF at most. As the temperature of the water at the bottom of the storage tank rises, the temperature drop of the collector fluid during its pass through the heat exchange decreases, and it most often is only about 5 oF. I expected a greater drop in the circulating collector fluid as it passes through the heat exchangers.

My rough computations indicate that with 400 BTU/sq.ft./hour of sunny southern California day sun energy and 240 sq.ft. of panel, approximately 96,000 btu / hr would be collected with an 100% efficient system. The storage tanks hold 180 gal. of water requiring 8.3 BTU/gal/oF for heat rise. Hence 1500 BTU is needed to raise tank temperature 1 oF. If system was 100% efficient, system could raise temperature of 180 gallons by 64 oF in one hour. My observation is that under the best of conditions, low water consumption on a bright sunny day, the tank water goes up about five degrees / hour at most. It is hard to get precise measure because I don't know how much water tenants in nearly empty building are consuming during the observation period.

QUESTION .. Does the performance of this system seem to be what would be expected? Would a second pump to circulate the water in the storage tanks significantly increase the efficiency of the heat exchanger even though it would result in the temperature of the exiting water being lower because the hottest water was no longer being allowed to rise to the top of the tank?

[MikeSolar]

hi there,

First, the inputs to the HX from the solar (hottest water) should be at the top port of the HX. The coldest water in the tank should go to the coil. Also, it appears that a cold water valve is open bypassing the solar tank. This must be closed. water must ONLY from through the solar tanks before getting to the back up tank.
Next, please post the height of the building, pipe diameter, controller type and pump specs. There are many things that can cause the issue because as an average family in your location, 2 - 3m2 panels (64 ft2) should do 75%+ of your hot water with a 300L (80-100USG tank)

[Naptown]

I hope that system is not plumbed in the way it is shown in the drawing.
Can you get photo's of the installation?

[russ]

Like Rich pointed out - if it is piped like the drawing it is a disaster.

[Art VanDelay]

Any gas usage data before installation?

Like Rich pointed out - if it is piped like the drawing it is a disaster.

Russ or Rich please explain further.

LAwindsurfer - It sounds like this system is providing hot water for an apartment building. If so, how many units, how many people per unit?

The high temperatures in the solar loop indicate that the heat exchangers may be undersized. It can be hard to balance the flow in a parallel system like this, so one of the tanks or heat exchangers could be starved for flow.

I have used the Rheem tanks and was disappointed in their heat exchange performance. These heat exchangers are simply 1/2" soft copper wrapped around a conventional steel tank. If that copper were submersed in the tank, they would work much better. But Rheem doesn't do that because corrosion is a problem. Most installers probably use two collectors per tank, and you have three. From my own testing, I believe this HX can only handle one 4x10 panel.
Rheem doesn't say. Nor do they provide ANY heat exchange data: http://www.rheem.com/documents/solaraide-he-rheem-solaraide-he-spec-sheet

"QUESTION .. Does the performance of this system seem to be what would be expected? Would a second pump to circulate the water in the storage tanks significantly increase the efficiency of the heat exchanger ..." NO
The problem with those HXs is a design problem with the amount of heat transfer area, and increased water circulation inside the tank cannot improve that.

But your solution isn't that difficult or expensive. Just add a pump: , and a brazed plate stainless steel heat exchanger: http://www.amazon.com/175-000-BTU-30..._sbs_indust_10

But once you start getting most of the available solar energy into the tank, you're going to find that you don't have enough storage unless the people use a lot of hot water during the day. Good design would have 360-500 gallons, but you have only 160 or 240 depending on which Rheems you have.

[russ]

No storage for an apartment building

With the 100 & 80 gallon tanks & heat exchangers piped in parallel as they are you have no control over the flow - it goes to the path of least resistance

What a 'flow back tank' is I have no idea

The hot water recirc pump is a horrible idea unless you have great insulation on all the piping - mine ha substantial heat loss so is disconnected

The T2 connection should go to the bottom of the tank - not top

For maximum water temp the T^& T+ connections should be reversed. As they are they would give maximum heat collection - take your choice

[LAWindsurfer]

Update .. to Heat Transfer ... below expectations

I'm preparing a better explanation of situation and will probably post later today with added data.

Last comment is correct. Installation is for an apartment house with 14 bathrooms and currently 19 people. Sketch was prepared to help service people know which valves to shut off if there was a leak and not necessarily for a design review - sorry. Sufficient attention was not paid to how roof circulation lines were drawn as they entered storage tanks; in fact return line is connected roughly mid tank where it is supposed to be connected and feed the to roof comes off the bottom of the tank. The "drain-back" tank is actually a large section of insulated pipe on the roof; it is sized such that if the pump is not running, all the water from the collectors puddles in that "drain back" tank.

The recirculation line is well insulated. I'm aware that if no water is being demanded, the gas water heater, and not the solar panels, keeps the water in the recirculation line warm. The recirculation line pump is controlled by a clock and thermostat. The pump runs continuously for about an hour in the prime morning an evening use periods. In non-prime hours, the pump only runs when the water in the return line, just before the water heater, drops below the set point - 105 oF. The thermostat controlling the pump has about a 1 oF hysteresis, so it is not necessary for the line to get really hot before the pump shuts off. Hence, during non prime hours, the water in the recirculation line is kept "warm" but not really hot.

More photos later.

Thanks for help and comments.

No storage for an apartment building

With the 100 & 80 gallon tanks & heat exchangers piped in parallel as they are you have no control over the flow - it goes to the path of least resistance

What a 'flow back tank' is I have no idea

The hot water recirc pump is a horrible idea unless you have great insulation on all the piping - mine ha substantial heat loss so is disconnected

The T2 connection should go to the bottom of the tank - not top

For maximum water temp the T^& T+ connections should be reversed. As they are they would give maximum heat collection - take your choice

[Art VanDelay]

Recirc, etc.

Russ,

Could you explain: "No storage for an apartment building" ?

LAW,

You have the optimum control strategy for your recirc loop. As Russ says, the only improvement would be to disable it completely, but that may invite a complaint from the apt. that showers first in the am.

[MikeSolar]

Russ or Rich please explain further.

LAwindsurfer - It sounds like this system is providing hot water for an apartment building. If so, how many units, how many people per unit?

The high temperatures in the solar loop indicate that the heat exchangers may be undersized. It can be hard to balance the flow in a parallel system like this, so one of the tanks or heat exchangers could be starved for flow.

I have used the Rheem tanks and was disappointed in their heat exchange performance. These heat exchangers are simply 1/2" soft copper wrapped around a conventional steel tank. If that copper were submersed in the tank, they would work much better. But Rheem doesn't do that because corrosion is a problem. Most installers probably use two collectors per tank, and you have three. From my own testing, I believe this HX can only handle one 4x10 panel.
Rheem doesn't say. Nor do they provide ANY heat exchange data: http://www.rheem.com/documents/solaraide-he-rheem-solaraide-he-spec-sheet

"QUESTION .. Does the performance of this system seem to be what would be expected? Would a second pump to circulate the water in the storage tanks significantly increase the efficiency of the heat exchanger ..." NO
The problem with those HXs is a design problem with the amount of heat transfer area, and increased water circulation inside the tank cannot improve that.

But your solution isn't that difficult or expensive. Just add a pump: , and a brazed plate stainless steel heat exchanger: http://www.amazon.com/175-000-BTU-30..._sbs_indust_10

But once you start getting most of the available solar energy into the tank, you're going to find that you don't have enough storage unless the people use a lot of hot water during the day. Good design would have 360-500 gallons, but you have only 160 or 240 depending on which Rheems you have.

The 006 is fine for the HX to tanks and I assume the pump to the panels is now a 009, high head and lower flow. He only needs 0.5 gpm/ panel so the 009 will give him the head needed and 3GPm. You are right though, the HX around the tanks are garbage and it should have an external HX. Also, the OP should have at least 2- 200GAL tanks and not mismatched ones.

[LAWindsurfer]

Heat Transfer from roof loop below expectations - additional detail 3

Sorry for the delay in updating my post but vacant apartment has taken priority.

The attached sketches provide additional detail as well as no longer implying inverted connections to the heat exchanger. One entitled "as built" shows current configuration. One entitled "add-on heat exchanger" shows proposed revised configuration. The photograph shows one portion of the temperature measurement instrumentation. The temperature of the collector loop is measured using "oven meat thermometers." I suspect they're not particularly accurate, but surprisingly they agree with one another within 1 oF, and it is temperature differential that provides an indication of the amount of heat being transferred from the collector loop to the storage tank. There thermostats were also inexpensive. The top left thermostat is showing time rather than temperature... it indicated about 110 oF.

A recirculation line is necessary for an acceptable hot water delay time as the distance between the water heater and the furthest apartment is more than 100 feet. The line is well insulated, but I suspect there is lots of heat leakage to the individual (inside the wall uninsulated ) risers. It can't be defeated and have happy tenants in front units.

There are 14 bathrooms in the building. There are 17 adults, one child and one infant. There are two, 30 gal/load, top loading washing machines, and about 3.5 loads (~100gals) are run each day on average. On Saturday, they're in nearly constant use between 10 AM and 3 PM. Assuming some pause for reloading between the 30 minute cycle times, thats 2x 8-loads or 16 loads and 480 gallons. Hot water usage is likely to be lower as not everyone does every wash with 100% Hot water. During the week, each machine is run for at least one cycle, frequently with the HOT setting for water wash temperature.

The gas fired water heater thermostat is set to ~135 oF which leads to a bathroom faucet temperature (as I recall) of about 125 oF at full flow. The 135 oF setting was established before the solar boost was installed. At the time, if it was kept much lower, I'd occasionally get complaints of inadequate hot water. I could be set the thermostat lower if the solar storage tanks still contained warm water in the morning, but that seems not to be the case. A thermostat that records max and min indicates that the output of the solar storage tank(s) drops to water-main temperature, ~ 65 oF, during a typical 24 hour period. I expect the storage tanks are at their lowest temperature early in the morning. I am well aware that it would be advantageous to have additional storage, but there was/is no more room in the laundry room or in the garage for additional tanks. The smaller 80 gallon tank was dictated by the smaller diameter required for there to be enough room to move washers and dryers out of the laundry room with having to bring in a plumber.

I am also aware that the unsymmetrical plumbing of the two units undoubtedly results in unbalanced heating of the stored water. However, I don't see that as an major issue as it appears that all of the preheated water is used up by morning. Furthermore, as the heat exchanger temperature differential will be less in the hotter tank than in the cooler tank, less heat will be delivered to the hotter tank and more to the cooler tank resulting in a somewhat self balancing situation. Again, as all the preheated water is being used up each day, I want to maximize the amount of captured heat, and it is not important to keep the tanks of equal temperature or to keep the hottest water at the top of the tank.

Could you please explain what is meant by, "The 006 is fine for the HX to tanks." What is the 006? Regarding the ability of the collector pump to overcome the required "head," ... I don't see the collector circulation pump being a problem as the water coming down from the roof is typically far hotter than the water in the storage tanks.

Art VanDelay said, "The problem with those (Rheem) HXs is a design problem with the amount of heat transfer area." Inadequate heat exchanger performance is what I suspected. Would the situation be improved by plumbing the two heat exchangers in series rather than in parallel? If I added a heat exchanger as suggested, would it be best to put it between the warm water take offs or the cool water feeds or some other configuration. (see second sketch) It would seem to me that the pump could be powered from the same controller as that used for the roof panel pumps, perhaps with an additional relay if the controller relay contacts are not sufficiently rated to run two pumps. How would I compute appropriate pump size and/or GPM flow for this second heat exchanger? I could install a three speed pump like the UPS-15 and just experiment to see which setting gives the best results.

Thanks again for your suggestions and help.




V0-SolarHotWater As Built.jpgV0-SolarHotWater Exchanger.jpgV0-SolarTempIndicate (12) (Large).JPG

[MikeSolar]

Sorry for the delay in updating my post but vacant apartment has taken priority.

The attached sketches provide additional detail as well as no longer implying inverted connections to the heat exchanger. One entitled "as built" shows current configuration. One entitled "add-on heat exchanger" shows proposed revised configuration. The photograph shows one portion of the temperature measurement instrumentation. The temperature of the collector loop is measured using "oven meat thermometers." I suspect they're not particularly accurate, but surprisingly they agree with one another within 1 oF, and it is temperature differential that provides an indication of the amount of heat being transferred from the collector loop to the storage tank. There thermostats were also inexpensive. The top left thermostat is showing time rather than temperature... it indicated about 110 oF.

A recirculation line is necessary for an acceptable hot water delay time as the distance between the water heater and the furthest apartment is more than 100 feet. The line is well insulated, but I suspect there is lots of heat leakage to the individual (inside the wall uninsulated ) risers. It can't be defeated and have happy tenants in front units.

There are 14 bathrooms in the building. There are 17 adults, one child and one infant. There are two, 30 gal/load, top loading washing machines, and about 3.5 loads (~100gals) are run each day on average. On Saturday, they're in nearly constant use between 10 AM and 3 PM. Assuming some pause for reloading between the 30 minute cycle times, thats 2x 8-loads or 16 loads and 480 gallons. Hot water usage is likely to be lower as not everyone does every wash with 100% Hot water. During the week, each machine is run for at least one cycle, frequently with the HOT setting for water wash temperature.

The gas fired water heater thermostat is set to ~135 oF which leads to a bathroom faucet temperature (as I recall) of about 125 oF at full flow. The 135 oF setting was established before the solar boost was installed. At the time, if it was kept much lower, I'd occasionally get complaints of inadequate hot water. I could be set the thermostat lower if the solar storage tanks still contained warm water in the morning, but that seems not to be the case. A thermostat that records max and min indicates that the output of the solar storage tank(s) drops to water-main temperature, ~ 65 oF, during a typical 24 hour period. I expect the storage tanks are at their lowest temperature early in the morning. I am well aware that it would be advantageous to have additional storage, but there was/is no more room in the laundry room or in the garage for additional tanks. The smaller 80 gallon tank was dictated by the smaller diameter required for there to be enough room to move washers and dryers out of the laundry room with having to bring in a plumber.

I am also aware that the unsymmetrical plumbing of the two units undoubtedly results in unbalanced heating of the stored water. However, I don't see that as an major issue as it appears that all of the preheated water is used up by morning. Furthermore, as the heat exchanger temperature differential will be less in the hotter tank than in the cooler tank, less heat will be delivered to the hotter tank and more to the cooler tank resulting in a somewhat self balancing situation. Again, as all the preheated water is being used up each day, I want to maximize the amount of captured heat, and it is not important to keep the tanks of equal temperature or to keep the hottest water at the top of the tank.

Could you please explain what is meant by, "The 006 is fine for the HX to tanks." What is the 006? Regarding the ability of the collector pump to overcome the required "head," ... I don't see the collector circulation pump being a problem as the water coming down from the roof is typically far hotter than the water in the storage tanks.

Art VanDelay said, "The problem with those (Rheem) HXs is a design problem with the amount of heat transfer area." Inadequate heat exchanger performance is what I suspected. Would the situation be improved by plumbing the two heat exchangers in series rather than in parallel? If I added a heat exchanger as suggested, would it be best to put it between the warm water take offs or the cool water feeds or some other configuration. (see second sketch) It would seem to me that the pump could be powered from the same controller as that used for the roof panel pumps, perhaps with an additional relay if the controller relay contacts are not sufficiently rated to run two pumps. How would I compute appropriate pump size and/or GPM flow for this second heat exchanger? I could install a three speed pump like the UPS-15 and just experiment to see which setting gives the best results.

Thanks again for your suggestions and help.




[ATTACH=CONFIG]2114[/ATTACH][ATTACH=CONFIG]2115[/ATTACH][ATTACH=CONFIG]2116[/ATTACH]

OK, It is not imperative but the anti-scald valve (mixing valve) that you have on the output of the storage tanks should actually be on the output of the gas tank, this way you can get a bit better storage capacity. It is actually in our code in Canada to have it set for 50C (120F), not sure about Cali.

Putting the new HX at the bottom of the tanks and piping in a reverse return will make the tanks more even. Reverse return mean that the piping from the solar panels enters on the left and exits on the right so that the pump sees both tank as being the same distance away and will then have more even flow. To do this on your drawing just take the tee that is on your return line from the 80 gal tank and put it on the 100 gal tank so the line comes from the left.

Art seems to like using Taco pumps so the 006 and 009 pumps are Tacos. As you have Grundfos, there is no need to change brands. If the current UPS-15-58 works well on the loops, keep it but remember that it will only work well as long no air gets into the system. This means no air vent on the roof. I have seen it many times but this is a cast iron body pump and sludge will form if air gets in.

The pump on the tank side needs to be stainless steel or bronze such as a UPS15-35 will do the trick. Take the drains out and put brass tees in place for the supply to the pump, then take the pump VERTICALLY after joining the tanks. It should pump into the HX from tanks and a strainer would be helpful. Then, up to the HOT out at the top of the tanks. The existing cold can stay for just cold water only.

The issue of balancing the tanks is actually important because it becomes the difference between using 40% of the resource and 50%. The same is true of the mixing valve. The extra volume in the gas tank can help.

[Art VanDelay]

LAW,

The 006 is the pump you could add to solve your heat exchanger problem. From my post above:

"But your solution isn't that difficult or expensive. Just add a pump: (above the link is still active for a Taco 006 pump) , and a brazed plate stainless steel heat exchanger: http://www.amazon.com/175-000-BTU-30..._sbs_indust_10 "

[LAWindsurfer]

Addition of supplemental heat exchanger and circulation pump.

OK, It is not imperative but the anti-scald valve (mixing valve) that you have on the output of the storage tanks should actually be on the output of the gas tank, this way you can get a bit better storage capacity. It is actually in our code in Canada to have it set for 50C (120F), not sure about Cali.

Putting the new HX at the bottom of the tanks and piping in a reverse return will make the tanks more even. Reverse return mean that the piping from the solar panels enters on the left and exits on the right so that the pump sees both tank as being the same distance away and will then have more even flow. To do this on your drawing just take the tee that is on your return line from the 80 gal tank and put it on the 100 gal tank so the line comes from the left.

Art seems to like using Taco pumps so the 006 and 009 pumps are Tacos. As you have Grundfos, there is no need to change brands. If the current UPS-15-58 works well on the loops, keep it but remember that it will only work well as long no air gets into the system. This means no air vent on the roof. I have seen it many times but this is a cast iron body pump and sludge will form if air gets in.

The pump on the tank side needs to be stainless steel or bronze such as a UPS15-35 will do the trick. Take the drains out and put brass tees in place for the supply to the pump, then take the pump VERTICALLY after joining the tanks. It should pump into the HX from tanks and a strainer would be helpful. Then, up to the HOT out at the top of the tanks. The existing cold can stay for just cold water only.

The issue of balancing the tanks is actually important because it becomes the difference between using 40% of the resource and 50%. The same is true of the mixing valve. The extra volume in the gas tank can help.

To MikeSolar: This message was prepared prior to Art VanDelay's 9:50 AM Post

I agree that placing anti-scald valve on output of the gas fired heater would increase storage capacity by allowing solar boost to feed water whose temperature exceeds 120 oF to the gas fed water heater. I need to check local code. As repositioning the anti-scald valve involves lots of plumbing, I'm probably going to wait on it until I see impact of circulation pump with supplemental heat exchanger.

I understand your suggestion on relocating T so extra length on input on 100 gal tank is offset by extra length on output of 80 gallon tank. I need to study the actual configuration of the plumbing to see how best to improve the balance. My original schematic diagram was a representation of the installation that was not intended to indicate actual pipe run lengths. In fact the connection diagram is far more complicated because of the physical constraints of the laundry room.

Thanks for alerting me to the potential of sludge / corrosion should air get into the circulation line. I need to check to see if there is an air vent.

You recommended pumping water from the drain through a heat exchanger and the back into the tanks via the hot water outlet. Does the attached sketch illustrate what you’re suggesting?

thanks again for the help.

[Art VanDelay]

LAW,

The drawing dated 6-26 looks pretty good.

As I said before, 180 gallons total solar storage is less than half of what you should have. Some excess solar can be transferred to the gas-fired tank.
But as you say, verify operation with the new brazed plate exchanger first.

You can install a 3/4" line from the aux tank to the cold line going to the solar tanks. In that line is a check valve and pump connected to a differential control. The pump kicks on when aux. tank temp drops below solar tank temp. (say delta T=5F ON, 3F OFF)"

It's like you are pretending that the solar tanks are solar collectors, and the aux tank is solar storage.

More discussion about a similar piping diagram is here

[MikeSolar]

To MikeSolar: This message was prepared prior to Art VanDelay's 9:50 AM Post

I agree that placing anti-scald valve on output of the gas fired heater would increase storage capacity by allowing solar boost to feed water whose temperature exceeds 120 oF to the gas fed water heater. I need to check local code. As repositioning the anti-scald valve involves lots of plumbing, I'm probably going to wait on it until I see impact of circulation pump with supplemental heat exchanger.

I understand your suggestion on relocating T so extra length on input on 100 gal tank is offset by extra length on output of 80 gallon tank. I need to study the actual configuration of the plumbing to see how best to improve the balance. My original schematic diagram was a representation of the installation that was not intended to indicate actual pipe run lengths. In fact the connection diagram is far more complicated because of the physical constraints of the laundry room.

Thanks for alerting me to the potential of sludge / corrosion should air get into the circulation line. I need to check to see if there is an air vent.

You recommended pumping water from the drain through a heat exchanger and the back into the tanks via the hot water outlet. Does the attached sketch illustrate what you’re suggesting?

thanks again for the help.

Yes, it looks good but you can completely disconnect the original HX in the tanks as the Brazetek HX will do the whole job. Actually, having the original HX in the loop may transfer some heat back to the panels at times so i would disconnect them. barring your ability to use a larger tank, this setup will give you the best efficiency. One other note: try to keep the potable water pump as slow enough to get about 30F dT under steady running conditions. You won't always be able to do it, of course, but it is something to shoot for.