Run through both coils on tank?

[J.P.M.]

Continuing on with the back and forth exchange of information and speculation,

...building the system with some flexibility and doing some measurements would be interesting.

Which is pretty much precisely what I've done for the last 8 + years. Some of what I've spouted and bloviated in this thread is some of what I think I've learned/confirmed in the endeavor.

I offer it only as (educated I believe) opinion based on real, and I hope valid engineering based analysis and the experience gained.

I've not much more to off along this vein. Thank you for the lively (and civil) discourse.

[inetdog]

Continuing on with the back and forth exchange of information and speculation,

If your hot water usage is going to occur entirely at times when no solar input is available, then I agree that stratification is not at all a practical issue.
At the other extreme, if you will be using hot water at or near the rate at which the solar system is able to supply heat to replenish the hot water supply. then you would be in a constant state of high stratification and it would make more sense to be directly heating the incoming cold water to the extent possible. In that case a lower heating coil and a downward flow of heated working fluid is more likely to make sense.
A real installation, particularly DHW for a residence, is likely to be somewhere in between the two extremes and so one size fits all advice will not necessarily be right for a given user. To counter that, the user may not be able to quantify usage and do the calculations well enough to determine in advance what strategy to use, so building the system with some flexibility and doing some measurements would be interesting.

[J.P.M.]

Well, the energy use in the electric element in the smaller tank, neglecting heat loss from the tank between uses, is going to be directly proportional to the number of gallons used and directly proportional to the temperature difference between the water from the 80 gal tank and the setpoint of the 35 gal tank.

While we are on the subject, an observation or two:

Ideally a hot water supply tank which is heated directly be paid energy should be operated at the lowest temperature which will meet the needs of the user. If you only want to take 110 degree showers, there is no point in heating the water much above 110 degrees since you will only be adding cold water to it at the point of use. But that affects primarily the static heat losses in the tank and the pipes. If you have hotter water in the tank you will end up using less of it for the same point-of-use flow rate. The product of volume and T will remain almost constant.

If, on the other hand, you are harvesting solar energy to heat your water you probably want to be able to store as much energy as possible as high quality heat for use during the dark hours. So, depending on the size of your storage tank, you may be better off storing water at a higher temperature and accepting the increased static heat losses and the loss of collector efficiency.
If the storage temperature gets high enough, safety requires a tempering valve to limit the water temperature as delivered to point of use.

If, by your 1st paragraph, you're saying the electric energy use is proportional to (the product of the house draw) X( the temp. diff. (supply temp. - thermostat set point)) when the del. temp. from the small tank is less than the set point, I agree with that statement and apologize for my poorly written statement. But, strictly speaking, and without trying to be pedantic, the electrical energy used will be proportional to the # of gallons used only when no solar heat from the larger tank is added to the small tank. disregarding line and tank losses for a moment, if all the needed energy is supplied by solar as when the feed from the solar tank is at or above the demand temp. and/or the thermostat set point, no additional electric energy will be needed. That and other common scenarios don't sound directly proportional to anything.

I agree with your second paragraph and note that a larger storage volume for the solar tank will accomplish that lower temp. goal. As I wrote, the lower all the temperatures in a solar thermal system can be designed and still meet the duty, the better it will be for system efficiency and probably system longevity. That's one argument for larger thermal storage. Now, a 10,000 gal. tank would be counter productive, but within reason and common sense, bigger storage is better than smaller storage for applications such as this, provided increased tank standby losses don't get out of hand. Another trick is more tank insulation. FWIW, I've got 6" extra insulation on my tank sides and 2" foam top/bottom.

The piping losses from a overtemping a system (high tank temperatures by design) will affect the draw. Draw out of the small will be less with higher temps. as you note. However, the piping losses will be greater, meaning the end use temp. will be proportionately lower, meaning either (slightly) less point of use cold makeup to make flowrate, and/ then/or more tank draw to make up for the (probably) slightly lower shower head delivery temp. caused by the increased line losses, which is where the increasedheat load will show up in an energy balance. I also agree that in the end, the (volume) X (delta T) product will remain as a practical matter ALMOST, but not quite the same, the small difference showing up over time as more hot flow to make up for lower delivery temps. caused by what may be relatively large piping heat losses as noted.

I'm mostly in disagreement with your 3d paragraph. The goal for things like DHW, as I've leaned at least, is to have sufficient quantity of energy of sufficient quality to meet the design load in as safe, maintainable, well designed and cost effective way as possible. Higher quality energy (higher temp. water), while always possible is usually not the most cost effective or practical way to go. The idea behind good solar thermal design also seems to be one of having sufficient heat quantity to get through some (limited) sunless periods without overdesign to handle overly long and infrequent sunless periods.

Rather than go for high storage temperatures, a (perhaps) better first step might be to insulate the holy grail out of the storage. Or, back to the bigger is better with respect to storage, depending on cost/efficacy, a larger storage may be possible. Better yet, large and heavily insulated storage that will have a sufficiently sufficient load capacity and a sufficiently long thermal time constant. Higher temp. storage is always an alternative choice, but that would be my last one. My first would be low flow shower heads and other use reduction, followed by more insulation on a design with a large tank on a system with the collectors fit for purpose and correctly sized.

Respectfully,

[Graycenphil]

This forum is great; I can't say how much I appreciate the thoughts and discussion. Thank you.

I have my 35 gallon electric heater set at it's highest temperature, maybe 125 degrees. We don't take 125 degrees showers, of course, but do like the hot water when you need it. This has also allowed us to get by with just a 35 gallon tank for three people andwhen the power goes out, we have a longer supply of hot water.

I set the solar system at 165 degrees to start, but this is probably too optimistic; I'm not sure it will ever get close to that. I'm happy with whatever it delivers, but it would be fun if it got above the 35 gallon tank thermostat. I'm also thinking that in a power outage, we will have a much longer supply of hot (or warm) water now.

It has not been too sunny snce I installed the system, but the next few days look like full sun, and we will be away all day tomorrow (not tusing any hot water). I'm anxious to see what temperature I get.

[inetdog]

...

How much the electric element in the smaller tank will be fired up is mostly a function of the hot water use, not the temperature of the water delivered by the 80 gal. storage device.

Well, the energy use in the electric element in the smaller tank, neglecting heat loss from the tank between uses, is going to be directly proportional to the number of gallons used and directly proportional to the temperature difference between the water from the 80 gal tank and the setpoint of the 35 gal tank.

While we are on the subject, an observation or two:

Ideally a hot water supply tank which is heated directly be paid energy should be operated at the lowest temperature which will meet the needs of the user. If you only want to take 110 degree showers, there is no point in heating the water much above 110 degrees since you will only be adding cold water to it at the point of use. But that affects primarily the static heat losses in the tank and the pipes. If you have hotter water in the tank you will end up using less of it for the same point-of-use flow rate. The product of volume and T will remain almost constant.

If, on the other hand, you are harvesting solar energy to heat your water you probably want to be able to store as much energy as possible as high quality heat for use during the dark hours. So, depending on the size of your storage tank, you may be better off storing water at a higher temperature and accepting the increased static heat losses and the loss of collector efficiency.
If the storage temperature gets high enough, safety requires a tempering valve to limit the water temperature as delivered to point of use.

[J.P.M.]

Your array can absorb only a limited amount of energy lets say 20k btu per day, that is fixed. That energy is dumped into your tank, it doesn't matter if you have 1, 2 or 3 coils the amount is fixed. The total delta T of heating 40 gallons will be double of heating 80 gallons. Heating 40 gallons could bring your water temp up far enough to keep your electric resistance heater from coming on, not so if you are heating 80 gallons. If you only use 40 gallons a day the electric backup should not be energized as long as you have daily sunshine = 0 energy consumption.
1 BTU raises 1# of water 1 degree F
One gallon of water weighs 8.34# you can do the math.

That's not correct.

While the amount of energy delivered ("dumped") to a heat exchanger by a solar thermal collector of fixed area, type and orientation may be fixed for any particular solar input and local weather conditions (but not strictly true as described below), the amount of that heat that's delivered and subsequently gets transferred to the tank fluid once delivered to the heat transfer device(s) is most certainly related, and in a pretty much directly proportional way to, and limited or enhanced by, the size/amount number of coils available to transfer the heat. The amount of heat transferred will be == Q = U*A* LMTD as described earlier, where "A" is the coil area.

2 times the coil area, "A", will mean about 2 times the amount of heat transferred to the fluid, "Q", for any reasonable storage size.

It looks like there may be some confusion between quantity of heat (BTU) and quality of heat (temperature).

2 coils of the same size (that is, equal heat transfer area per coil) will transfer twice as much heat as one coil under the same temperature conditions. What will change in a 40 gal. to 80 gal. tank is how much the tank water temp. will increase. Neglecting tank thermal losses and line losses to/from the collector and the lower sp. gr. of hot water for the time being, if you add 20K BTU to a 40 gal tank, the tank fluid temp. will increase ~ 60 deg. F. (20,000)/(8.35)*(40). Aside from getting close to a temp. of something like 180 F. at the end of the day, or maybe 165 F. if tank and line losses are considered, the better thing to do is to use an 80 gal. tank with both coils and cut the tank fluid temp. increase to 25 - 30 deg. F., as well as cutting the standby and line losses with the lower tank temp., as well as keeping the keep the collector temp. lower and thereby increasing it's thermal efficiency and probably inducing less thermal stress on the collector, and thereby (perhaps) increasing it's probable service life maybe a bit.

Cutting the coil area in half with any reasonable storage volume size will reduce that tank fluid TEMPERATURE gain by about half by cutting the heat delivered to the storage device in half, regardless of the storage size. Additionally, since only half the collected heat will be deposited in the storage by a half size coil, the collector will run a lot hotter due to a higher collector inlet temp. (that is, higher coil outlet temp. on the collector loop side). That will compound the inefficiencies that a smaller coil area causes by lowering the collector thermal efficiency due to incresed thermal losses to the environment.

The secret to collecting a lot of heat very efficiently and cost effectively in solar energy applications is, in general, to keep the temperatures of all devices as low as possible while still meeting the duty, and do that while making heat transfer between collector and storage as easy as possible - which almost always means - in a practical sense, and to a point - large(r) and more efficient heat transfer surfaces.

I'd keep the 80 gal. tank and two coils plumbed in series. I'd also run the fluid in the collector/coil(s) loop at as high a flow rate as possible, but that's a separate subject.

How much the electric element in the smaller tank will be fired up is mostly a function of the hot water use, not the temperature of the water delivered by the 80 gal. storage device.

[inetdog]

Your array can absorb only a limited amount of energy lets say 20k btu per day, that is fixed.

What you are leaving out is that the net energy into the collector is the difference between the energy absorbed by the collector from sunlight and the amount of energy lost by conduction and re-radiation before it gets to the rest of the system. And there will be heat losses in the piping which are proportional to the pipe temperature.

These and other factors combined mean that your collector will absorb more net energy if it is operated at a lower temperature. That, in turn, means that you want to keep the temperature of the inlet water to the collector as low as possible.

[LucMan]

Thanks Lucman. The way I have it set up, the 80 gallon tank feeds my 35 gallon electric tank. We don't use a great deal of hot water. My hope is that I can get the 80 gallon tank above the temp of the 35 gallon thermostat, and keep it there so the electric never comes on. Whether I ever reach that goal, I'm using less electricity by starting with warmer water.

Your array can absorb only a limited amount of energy lets say 20k btu per day, that is fixed. That energy is dumped into your tank, it doesn't matter if you have 1, 2 or 3 coils the amount is fixed. The total delta T of heating 40 gallons will be double of heating 80 gallons. Heating 40 gallons could bring your water temp up far enough to keep your electric resistance heater from coming on, not so if you are heating 80 gallons. If you only use 40 gallons a day the electric backup should not be energized as long as you have daily sunshine = 0 energy consumption.
1 BTU raises 1# of water 1 degree F
One gallon of water weighs 8.34# you can do the math.

[J.P.M.]

I would guess that varies between a straight metal tank and a glass lined tank, which is a common option in DHW units. Even if the glass lining cracks and loses its effectiveness in protecting the tank wall from corrosion it will still be able to decrease the thermal shunt across the height of the tank.

I'd agree that it varies, and that lined tanks are common to the point of universal, but the heat transfer that might well disrupt stratification is not an all/nothing thing - probably somewhere in between.

The more recent data I could find on thermal conductivity of "glass" lining is similar to that for true epoxy linings and other such stuff I once designed with/around. That more recent data that gives a value for the thermal conductivity of the glass as about 2% +/- a bit of that of steel. However, that's not the whole story with respect to how well/poorly a barrier/lining conducts heat. A glass lining is much thinner than most pressure boundaries (steel walls) - something like ~ 0.3 mm or so, or something like 0.01" +/-some.

The rate of heat transfer through a homogeneous barrier whose thermal conductivity is isotropic is inversely proportional to it's thickness. I'd take an educated guess and say the OP's tank wall is about 0.13" thick. A back of the envelope on relative ability to conduct heat in a radial direction (outward) might be about 3 or 4 to 1, tank wall to glass lining, the tank wall being thicker but with much higher thermal conductivity vs. a much thinner glass lining that has much poorer thermal conductivity.

However, I'd also estimate (and here the heat transfer explanations/estimates start to get sticky and start to involve dimensionless groups like Biot numbers, Nusselt #'s and others, and boundary layer thickness estimates of the tank fluid under natural convection and other stuff) that the rate of heat conduction through the ling will still be of an order of magnitude or so more than the natural convection coefficient of the tank fluid at the wall, and that's what determines the relative importance of the tank wall's ability to disrupt tank fluid stratification.

Once the heat does get through to the steel portion of the tank, the greater tank wall thickness (the 0.13" or so) and the greater thermal conductivity of the steel will make it easier for conduction of heat from where the tank wall is warmer to where the tank is colder - in our example "down" the tank to the very same portions of the tank where the water is colder - and then, once it gets there, to transfer that heat in the same manner as when it was going "out", but this time in the opposite direction - going "into" the water, and thus causing a local heating of a portion of the fluid, changing that (local) fluid "chunk's" density relative to the surrounding fluid and moving that "chunk" in a direction that will tend to reduce stratification.

I'd be quite surprised if the cracks in linings that eventually cause most DHW tank failures contribute little, one way or the other to heat transfer or stratification.

I was not and am not trying to imply that this is a big contributing factor in killing tank stratification, but unless all the laws of Thermodynamics as they relate to entropy, and all the principles of heat transfer dealing with conduction and convection have been repealed or replaced, I believe the above is a reasonable, if greatly simplified, approximation of what will happen.

My purpose in citing tank/lining axial conduction was as an example of many of the (possibly) unknown or unconsidered mechanisms and /or conditions that might exist that would contribute to hasten the destruction of thermal stratification in situations similar to those the OP describes. It's also perhaps an example of the many contributory factors as to why thermal stratification is quite fragile and thus quite easy to disrupt or destroy, and not something to place a lot of reliance upon when looking for ways to increase thermal efficiency in things like solar thermal collector systems.

As usual, take what you want of the above. Scrap the rest.

[inetdog]

... Also, as I mentioned in another post, the thermal conductivity of the tank wall, which is usually about an order of magnitude greater than that of water will contribute to reducing or eliminating stratification. ....

I would guess that varies between a straight metal tank and a glass lined tank, which is a common option in DHW units. Even if the glass lining cracks and loses its effectiveness in protecting the tank wall from corrosion it will still be able to decrease the thermal shunt across the height of the tank.

[J.P.M.]

Thanks again for all the thought. Perhaps a dumb question: At night (when no hot water is being used or made) does the tank become more stratified, due to lack of movement? Or does it become less stratified, due to the hot water on top warming the cold water below?

Less dumb question: My tank has two temp sensor locations, upper and lower. The controller compares this to the temp of the solar array, and decides when to turn on the pump. Once I plumb in both coils, should I still use the lower sensor, as I do now with just the lower coil in operation.

First, and respecting LucMan's opinion while still having a largely opposite view on this particular issue, I'd keep both coil's connected. More surface area inside the tank will result in more heat transfer which is the goal.

Any additional heat in the working fluid will then transfer that additional heat to the tank via the lower coil. Without using the lower coil, that heat will be retained in the working fluid and only serve to lower the collector efficiency by raising the average temp. of the collectors. By leaving both coils connected, the worst that will happen (assuming the rest of the control system functions correctly) is more heat will wind up in the 80 gal. tank.

To your questions, which BTW are both far from dumb:

1.) When no heat is being added to or withdrawn from the tank, the water will cool by heat loss to the surroundings. The stratification will decrease a bit, somewhat in proportion to the overall temp. decrease of the tank To a very loose 1st approx. if the tank temp. drops, say by 10% of the difference between the tank water and the tank's surroundings, the difference in tank stratification top to bottom will probably decrease by about 10% or so. The stratification is mostly caused by specific gravity differences of the water at different temperatures. In a strict sense, the temp. gradient in a still tank is not linear, but follows about the same path as how the specific gravity of the fluid - in this case H2O - varies as f(temp.), and that is not quite linear in a strict sense. Also, as I mentioned in another post, the thermal conductivity of the tank wall, which is usually about an order of magnitude greater than that of water will contribute to reducing or eliminating stratification. Het will flow from the top fluid to the tank wall making that portion of the tank warmer than the lower portion of the tank. Since heat flows from warmer to colder regions, the lower portion of the tank will thus warm up and transfer its heat to the lower portion of the tank fluid, which will then rise and reduce the stratification.

In the Thermodynamic limit, if left long enough, the tank will be at a uniform temperature equal to the surroundings with no stratification.

2.) The second of your non dumb questions is one I might defer to LucMan as he likely has more practical and hands' on experience than I, which is close to invaluable. I'm quite sure he's seen more practical/actual residential situations than I have.

However, I'd think if the lower coil was the first one the fluid returning from the collector saw, which is how I'd plumb it for reasons given yesterday, and, provided there was not much actual stratification as my experience, measurement and education leads me to believe, I'd still use the lower sensor. It would seem to me that, again, a worst case would be that the tank sensor seeing a temp. that has the highest probability of being the lowest temp. would keep the pump running longer and removing more heat. As a practical matter, I'd SWAG it that the tank stratification under operating conditions will be small enough to make sensor temps. close to the same, maybe even within the controller deadband, but I'd also guess that a cold startup in the A.M. with a slightly stratified tank would benefit from using the lower sensor.

[Graycenphil]

Thanks Lucman. The way I have it set up, the 80 gallon tank feeds my 35 gallon electric tank. We don't use a great deal of hot water. My hope is that I can get the 80 gallon tank above the temp of the 35 gallon thermostat, and keep it there so the electric never comes on. Whether I ever reach that goal, I'm using less electricity by starting with warmer water.

[LucMan]

I suggest that you connect only to the top coil, an 8 tube array is under sized for an 80 gallon tank. A 30-40 gallon tank would be the correct match for the array you have installed. Using the top coil only will give you usable hot water at your faucets.

[Graycenphil]

Thanks again for all the thought. Perhaps a dumb question: At night (when no hot water is being used or made) does the tank become more stratified, due to lack of movement? Or does it become less stratified, due to the hot water on top warming the cold water below?

Less dumb question: My tank has two temp sensor locations, upper and lower. The controller compares this to the temp of the solar array, and decides when to turn on the pump. Once I plumb in both coils, should I still use the lower sensor, as I do now with just the lower coil in operation.

[J.P.M.]

Thank you for the detailed analysis.

I do have some doubts about your suggestion that the stratification temperatures in the typical DHW tank will be no more than on the order of 5 degrees.
At least during the periods when hot water is actually being consumed the difference can be much greater. How long that high differential will be maintained is another question though.
I am working on the assumption that there is a practical reason for having a top and bottom heater in an electrical DHW tank with interlocked thermostats that do not energize the bottom element unless the top of the tank (from which hot water is drawn for use) has already reached the cutoff temperature.
The design of the cold water inlet dip tube is intended to minimize active mixing of water in the tank as hot water is being drawn off.

Understood. You're welcome.

As you probably know, the stratification of water as f(temp.) is (or can easily be made to be) a complicated subject, quite application specific as to effects and degree, and probably not manageable in a forum type setting. Small example: it is affected by, among many other things, tank wall thermal conductivity.

On an experiential/experimental basis at my home - not as part of my prior engineering practice, I've got an 80 gal. ( ~ 5 ft. of liquid col. height) single electric (4.5 kW) element. I replaced a prior thermal collector about 8.5 yrs. ago, kept the relatively new tank (less than 1yr.), repiped and instrumented the entire (new) thermal collector loop w/ several thermometers, two rotometer type flow meters, and a bunch of pressure gauges and other thermometers. Long, boring story. The collector uses potable water directly from the tank without a HX. The system runs well with no problems.

FWIW, I believe I know more about heat exchangers than I know about solar energy. Knowing that, I do not believe the currently available technology, or quality for residential heat exchangers used in solar thermal energy applications is either cost effective, practical or serviceable. Most such products are junk and cause more problems due to fouling and performance penalties than they are worth. Most homeowners are clueless about them and service personnel are only marginally better, if at all.

Anyway, as you may guess, I have not stopped my old ways of piddling with stuff to see what practical knowledge I can squeeze from such efforts. I usually wind up confirming what the textbooks say, but there are always surprises 1X/awhile.

Two things I have among several toys on the system are thermowells w/thermocouples at the top/bottom of the tank, the bottom one in the outlet to the collector about 2 pipe dia. from the tank, the top one in one of 2 anode locations, the other anode being there of course, and checked annually.

Commonly, in the morning and with usually no hot H2O draw since maybe 10 P.M. the prior evening, the undisturbed tank temp. differential - top to bottom runs about 3 - 4 deg. F.

If hot water draw does occur, the cold water from mains replacement (usually ~ 60 - 70 F. depending mostly on season), does indeed tend to stay in the bottom of the tank as put there by the dip tube as you describe. There is usually not a lot of hot H2O draw around here as you might guess, so the level of cold water makeup in the tank tends to be below the bottom foot or so of tank level. Technically speaking, that results in a lot of stratification, but it is not as linear as naturally occurring stratification with a sharp "knee" at the cold/warm interface - and again, quite low in the tank and probably below any HX coils that might be in a similar tank. I sort of confirm that "sharpness" in the interface by being able to use the collector pump to recirculate tank H2O, bypassing the collector loop altogether. Then, at the tank return (at the top), another arrangement using a 3 way valve allows me to put the return back to the tank via direct injection at a variable rate of up to about 7.4 GPM through a 1/2" orifice pointing downward, or sideways horizontally and approx. at/along the tank wall to create a circular fluid motion in the tank, or some combination of the two. In so doing, and at a throttled rate and a horizontal return, the transition from the colder added water to the much warmer "upper" water is always quite sudden, as indicated by the 10 sec. or so it takes to go from say 80 deg. F. up to, say, 110+ deg. F., indicating to me that the cold water does indeed stay low and the demarcation cold/hot does tend to stay somewhat sharp for longer than one might think if stratification is left undisturbed.

Once collector circulation begins in the A.M., the tank stratification is quickly destroyed. That's due not only to the fragility of the stratification, but for other reasons as well. For my situation, that reason is mostly due to flow rate. I suppose this may spark some discussion, but I circulate through 2 ea. 4 X 8 collectors plumbed in series. I can circulate at up to about 4.3 GPM through the collectors, but usually run at about 1.7 GPM and keep the collector fluid Reynold's # in the transition zone for what I believe are several sound/practical engineering reasons, but that's sort of off topic. Details on request.