Heat transfer from roof loop to tank below expectations.

[LucMan]

I would connect the HX out to the cold water inlet side of the tanks instead of the HW out. Pull the dip tubes and cut them so that they are about 12" from the bottom of the tanks and reinsert. Less turbulence and mixing of hot & cold water in the storage tank this way.

[MikeSolar]

I would connect the HX out to the cold water inlet side of the tanks instead of the HW out. Pull the dip tubes and cut them so that they are about 12" from the bottom of the tanks and reinsert. Less turbulence and mixing of hot & cold water in the storage tank this way.

Not the best way. If you have cold water coming in at the bottom and some of it is getting heated by the HX, you would not want to have that newly heated hot water mix with the cold near the bottom of the tank. Having it deposited at the top where it can properly stratify will allow the hottest water to be available to the gas tank. This means the gas tank will not see mixed water coming in and be fooled into turning on for short periods of time (which wastes gas).

We have designed many systems where there is no pump on the potable side and the heat will thermosyphon from the HX to the top of the tank but it takes careful HX design and pipework. This is why I said to keep the pump speed as slow as possible and also use the smallest pump possible to get a high dT across the HX.

[LAWindsurfer]

Optimizing flow rate on potable side of heat exchanger

Not the best way. If you have cold water coming in at the bottom and some of it is getting heated by the HX, you would not want to have that newly heated hot water mix with the cold near the bottom of the tank. Having it deposited at the top where it can properly stratify will allow the hottest water to be available to the gas tank. This means the gas tank will not see mixed water coming in and be fooled into turning on for short periods of time (which wastes gas).

We have designed many systems where there is no pump on the potable side and the heat will thermosyphon from the HX to the top of the tank but it takes careful HX design and pipework. This is why I said to keep the pump speed as slow as possible and also use the smallest pump possible to get a high dT across the HX.

If I understand the situation correctly, MikeSolar's suggestion to slow the potable side pump down to seek "a high dT across the Hx" may or may not improve things depending on the location of the heat exchanger. If the heat exchanger is located at or above the top of the storage tanks, one wants to have a high flow through it, so to bring through the cooler water. If the heat exchanger is located at or below the level of the storage tanks, where it is immersed in the cooler bottom water, you want to keep the circulation rate low to allow the water to stratify and preclude the mixing that will bring the warmer water through the heat exchanger. I'm not sure I can locate the heat exchanger at or below the bottom of the storage tanks in my situation because of physical space constraints. The most convenient place to locate the heat exchanger in my situation is against the ceiling above the storage tanks; if so done, you want to have good circulation. Locating the heat exchanger down low has the advantage that thermally induced circulation will minimize the electrical power and operating costs for the potable water circulation pump. I need to look at the economic tradeoffs; initial investment in building structure to raise the tanks up off the floor versus continuing operating costs for circulating the water.

[Naptown]

Not the best way. If you have cold water coming in at the bottom and some of it is getting heated by the HX, you would not want to have that newly heated hot water mix with the cold near the bottom of the tank. Having it deposited at the top where it can properly stratify will allow the hottest water to be available to the gas tank. This means the gas tank will not see mixed water coming in and be fooled into turning on for short periods of time (which wastes gas).

We have designed many systems where there is no pump on the potable side and the heat will thermosyphon from the HX to the top of the tank but it takes careful HX design and pipework. This is why I said to keep the pump speed as slow as possible and also use the smallest pump possible to get a high dT across the HX.

Most internal heat exchangers are simply a coated pipe within the tank.
Slowing down the flow too much will result in laminar flow in both the collectors and heat exchanger. You want turbulent flow through both for highest output.

[MikeSolar]

Most internal heat exchangers are simply a coated pipe within the tank.
Slowing down the flow too much will result in laminar flow in both the collectors and heat exchanger. You want turbulent flow through both for highest output.

If all you cared about was effective transfer of heat through the HX, I would agree with you but not if you want to promote stratification in a tank. You want to keep the mixing to a minimum, take cold from the bottom and put hot in the top without stirring it up too much. If you allow it to mix, the probability will be that the water pulled out of the tank, much of the time, will be below the setpoint to turn on the back up heater. If you let it move slowly through the domestic side of the HX, you may lose a bit of efficiency of the HX but you will increase the effectiveness of the system by keeping the backup off. In the end, that is what is most important. Slowing down the DHW flow does not influence the speed of the solar line. It is preset.

LA, although it is better to have the HX at the bottom of the tank, it will work quite well at the top. The only difference will be that the pumping power will be slightly higher to overcome the thermosyphoning. The other issue is that having it up there will promote thermosyphoning from a hot tank to the panels at night so I would put a 2ft U-bend on the line going out of the HX to the panels to stop it.

[Naptown]

If all you cared about was effective transfer of heat through the HX, I would agree with you but not if you want to promote stratification in a tank. You want to keep the mixing to a minimum, take cold from the bottom and put hot in the top without stirring it up too much. If you allow it to mix, the probability will be that the water pulled out of the tank, much of the time, will be below the setpoint to turn on the back up heater. If you let it move slowly through the domestic side of the HX, you may lose a bit of efficiency of the HX but you will increase the effectiveness of the system by keeping the backup off. In the end, that is what is most important.

LA, although it is better to have the HX at the bottom of the tank, it will work quite well at the top. The only difference will be that the pumping power will be slightly higher to overcome the thermosyphoning. The other issue is that having it up there will promote thermosyphoning from a hot tank to the panels at night so I would put a 2ft U-bend on the line going out of the HX to the panels to stop it.

I thought we were talking about the internal static heat exchangers not the second external.
In that case yes slow flow is better and a plate HX will have turbulent flow irregardless.
however I would not bring the return into the top. About 1/2 -2/3 down would be about right.

[MikeSolar]

I thought we were talking about the internal static heat exchangers not the second external.
In that case yes slow flow is better and a plate HX will have turbulent flow irregardless.
however I would not bring the return into the top. About 1/2 -2/3 down would be about right.

I agree on having a higher flow rate in internal HX. I do stick by the idea of putting the hot directly at the top for the reasons explained above. Plus, we measure the differential temp at the bottom of the tank so lets assume for the moment that you have cold water at that level, and the top of the tank is 80F. The pump will pump perhaps 120f water down to the tank and if it enters at the 1/2 way mark it will mix with the colder water and the 80F you had at the top could now be less (or it could be more but not 120F). If you draw water off for hand washing, you will get colder water than if you allowed the 120F water to enter at the top of the tank. The mixing would be far less. It works really well and I have done this on a lot of systems.

[Naptown]

I agree on having a higher flow rate in internal HX. I do stick by the idea of putting the hot directly at the top for the reasons explained above. Plus, we measure the differential temp at the bottom of the tank so lets assume for the moment that you have cold water at that level, and the top of the tank is 80F. The pump will pump perhaps 120f water down to the tank and if it enters at the 1/2 way mark it will mix with the colder water and the 80F you had at the top could now be less (or it could be more but not 120F). If you draw water off for hand washing, you will get colder water than if you allowed the 120F water to enter at the top of the tank. The mixing would be far less. It works really well and I have done this on a lot of systems.

OK than explain why every single tank manufacturer that uses an external heat exchanger directs the flow in this manner.
Doing this was a known quantity thirty years ago when I first got involved with this stuff and hasn't changed since.
Secondly I do not reccomend a second (or in this case third) heat exchanger at all. Waste of money IMHO.
I don't know why on earth they used 2 different size tanks unless there was a space issue. but flow should be able to be balanced through them. Otherwise put the two solar tanks in series with each other and the back up tank and run the loop from the collectors in series with the two solar storage tanks feeding the 100 gal first and proceeding to the 80 then the return to the collectors. Or better yet add a pump and split the tank heat exchangers so each operates on its own pump to the array and will be heated and controlled separately.

[MikeSolar]

OK than explain why every single tank manufacturer that uses an external heat exchanger directs the flow in this manner.
Doing this was a known quantity thirty years ago when I first got involved with this stuff and hasn't changed since.
Secondly I do not reccomend a second (or in this case third) heat exchanger at all. Waste of money IMHO.
I don't know why on earth they used 2 different size tanks unless there was a space issue. but flow should be able to be balanced through them. Otherwise put the two solar tanks in series with each other and the back up tank and run the loop from the collectors in series with the two solar storage tanks feeding the 100 gal first and proceeding to the 80 then the return to the collectors. Or better yet add a pump and split the tank heat exchangers so each operates on its own pump to the array and will be heated and controlled separately.

http://www.thermo-dynamics.com/solar_boiler.html

http://enerworks.com/products/#residential-product-lines

Both these companies use the method I outlined and the first one is a 30+ year old company with one of the highest output systems available.

Why do some companies have HX in/out at the top of the tank and some at the side? Company preference and we all know that most small companies just follow what their bigger competitor is doing, saves on research.

TBH, I don't know why they still do it this way but there is a lot of inertia in business and system changes cost money, especially things like UL, CSA or SRCC certs. The question is simple though, is it better to have the hottest water at the top of the tank fastest and for the longest time? If so, what is the best method?

[Naptown]

http://www.thermo-dynamics.com/solar_boiler.html

http://enerworks.com/products/#residential-product-lines

Both these companies use the method I outlined and the first one is a 30+ year old company with one of the highest output systems available.

Why do some companies have HX in/out at the top of the tank and some at the side? Company preference and we all know that most small companies just follow what their bigger competitor is doing, saves on research.

TBH, I don't know why they still do it this way but there is a lot of inertia in business and system changes cost money, especially things like UL, CSA or SRCC certs. The question is simple though, is it better to have the hottest water at the top of the tank fastest and for the longest time? If so, what is the best method?

Here is where the problem lies with returning to the top. (and the second returns about where I reccomend)
It's morning and the tank is stratified with 120 deg water up top and say 70 degrees at the bottom.
The controller senses a differential of 10d and starts up
So now you have 75-80 degree (probably closer to the 75d once the system stabilizes) water mixing with the 120d water at the top and cooling it down where if it were reintorduced lower you would actually be heating the lower portion without disturbing the top stratified layer. pulling from the bottom and returning to the top is never a good idea.

[inetdog]

Pulling from the bottom and returning to the top is never a good idea.

I would qualify that with "for solar water heating"

If the goal is to produce hot water to satisfy the consumer of hot water, absolutely right IMHO.

If the goal is to get the largest number of calories of heat into the tank and you don't care about stratification, then pulling the coldest water into the panel and returning the output to the hottest part of the tank will harvest the most total heat per hour from the panels. That sort of need might happen in a room heating application with a baseboard loop rather than a water heating application, and would give you the largest amount of stored heat by the time the sun goes down. (And so it might be specified in an a handbook oriented to HVAC rather than domestic water?)

On the other hand, it is probably just as likely that the writer did not think things through properly.

[MikeSolar]

Here is where the problem lies with returning to the top. (and the second returns about where I reccomend)
It's morning and the tank is stratified with 120 deg water up top and say 70 degrees at the bottom.
The controller senses a differential of 10d and starts up
So now you have 75-80 degree (probably closer to the 75d once the system stabilizes) water mixing with the 120d water at the top and cooling it down where if it were reintorduced lower you would actually be heating the lower portion without disturbing the top stratified layer. pulling from the bottom and returning to the top is never a good idea.

Agreed, assuming that you have a control that doesn't check the tank top temp. The Resol controls I use will not turn on the second pump until the liquid coming back is above tank top temp (if I set it up that way). I only use the 2 pumps system when I have a lot of collectors and I need a storage volume that doesn't allow for an internal HX. Damn those ASME rules, they are so archaic. Europe has a much better system, build it, test it and rate it for pressure, none of these CRN numbers. But that is a different discussion.

If I am using just a thermosyphon system with only the solar pump, there will be no charging of the tank until the temp at the exit of the HX is higher than tank top. That will happen quick enough if there is nothing taking heat out of the solar line and it needs no controls.

[MikeSolar]

I would qualify that with "for solar water heating"

If the goal is to produce hot water to satisfy the consumer of hot water, absolutely right IMHO.

If the goal is to get the largest number of calories of heat into the tank and you don't care about stratification, then pulling the coldest water into the panel and returning the output to the hottest part of the tank will harvest the most total heat per hour from the panels. That sort of need might happen in a room heating application with a baseboard loop rather than a water heating application, and would give you the largest amount of stored heat by the time the sun goes down. (And so it might be specified in an a handbook oriented to HVAC rather than domestic water?)

On the other hand, it is probably just as likely that the writer did not think things through properly.

I'm not sure I understand you but it looks like you agree with the way the two companies I referenced do it. the real benefit, aside from always having the coldest water going to the panels (it is not mixed at the bottom at all), is that, if you assume, that there is no back up tank and you want to do hand washing or take a shower, there is a higher degree of probability that (after 30 min in the morning from a totally cold tank) the thermosyphon system will give you enough hot water to do the job. Not returning it to the top means using a backup, whatever that is, to get the temp to 120F. It simply takes longer for mixed water to stratify. Sorry for the long sentence.

[inetdog]

It simply takes longer for mixed water to stratify. Sorry for the long sentence.

That sounds like you have the right idea.

Only one small problem:
Mixed water does not stratify. That would violate the infamous and repressive laws of thermodynamics. If you add heat to the top of the tank (via heat exchanger, electric element, gas flame or any method that does not require moving the water) then as you heat the top water only it will become stratified. And if the thermostat is at the top, the bottom will get heated only by conduction and it will take a loooong time to "mix" to a uniform temperature.

Hopefully that was just a poor choice of phrase on your part.

[MikeSolar]

That sounds like you have the right idea.

Only one small problem:
Mixed water does not stratify. That would violate the infamous and repressive laws of thermodynamics. If you add heat to the top of the tank (via heat exchanger, electric element, gas flame or any method that does not require moving the water) then as you heat the top water only it will become stratified. And if the thermostat is at the top, the bottom will get heated only by conduction and it will take a loooong time to "mix" to a uniform temperature.

Hopefully that was just a poor choice of phrase on your part.

Trying to type and watch spain vs portugal.........not very easy.

You are right, poor choice of words.