Domestic Hot Water Storage Tank Bank

mbshawaii

My friend's 12-story apartment building is suffering from poor hot water recirculation.  I went into their ground level hot water plant and found three (3) air-to-water heat pump water heaters.  They were piped to a bank of six (6) 120 gallon storage tanks that were connected in series.   It seems that if they were connected in parallel, the hot water recirc loop would only have to go through the equivalent of 1 storage tank, not all 6.  What is the reason for having them in series?  It also makes it tough to isolate a tank in case something happens.  Does series piping provide better stratification over all 6 tanks, which might make the heat pump COP's higher.  Thoughts?



Thank you very much!

Mark Eatherton

Follow your gut feeling....

I seriously doubt that the tank configuration has anything to do with the poor recirculation return your friend is experiencing. That probably has more to do with the way it is piped and controlled (direct return with poor balancing).



I assume the reason for all the storage is due to the limited hourly output capacity of the ASHP's.



The ONLY time it makes sense to have tanks in series, is in situations where you have numerous inputs, with differing temperatures of operation. For example, first stage is waste heat recovery (REAL low temp app) then solar DHW preheat (medium to high temp app) and then the final tank temperature (140 degrees F recommended which may be outside of the capacity of the ASHP's). The final temperature tank could be gas, electric or other fuel capable of reaching the needed higher temperature. Don't forget to provide an ASSE rated anti scald mixing valve to avoid scald potential.



One could pipe these tanks in a parallel reverse return, counter flow configuration, thereby minimizing the pressure drop associated with flow, and maximizing the energy input from the heat sources. You could pipe 3 pairs of tanks in series parallel if you wanted, but it would complicate the piping with very little benefit.



The lower the operating temperature, the higher the C.O.P. of the ASHP.



The reason for maintaing such a high final temperature is for bacteria control.



In large storage tanks, there is a condition called the "bushing effect", which creates significant resistance to flow on draw. Putting tanks in series compounds this pressure drop on draw. As you said, paralleling it would allow the system to "see" this pressure drop only once.



Have you checked with the manufacturer for their piping recommendations?



ME

mbshawaii

Series vs. Parallel analysis

It dawned at me that I need to look at the series piping

differently.  After sketching some things out, it started to make sense,

and I am suddenly in favor of the series storage tanks I've seen in the

field.  See attached PDF and comment.



 Sheet 1



Figure 1 shows a typical vertical storage tank

with a heat pump water heater and a typical hot water recirculation loop. 

This system demonstrates good thermal stratification, maximizing the delta T

between T1 and T6, and as such maximizing the COP of the heat pump.





 Figure 2 shows an identical tank which has

been divided into three equal volumes for modeling (still a single compartment

tank).  Temperatures T2 & T3 are equal, as are temperatures T4 &

T5.  Stratification and HP performance are identical as Figure 1.





 Figure 3 shows the same three volumes as Figure

2, now modeled as separate compartments.  Similar stratification and

performances must occur regardless of height differences between the locations

of T2 & T3, and T4 & T5.



Sheet 2



Figures 1, 2, 3 show the

typical hot water recirculation path for all three scenarios in ORANGE, when the heat pump (and integral

circulation pump + valve) is OFF.



Sheet 3



Figures 1, 2, 3 show the

typical heat pump circulation path for all three scenarios in GREEN, when the heat pump (and integral

circulation pump + valve) is ON.  Note the direction of flow through the

pipe connecting the tank has reversed.



Figure 3 also shows the hot

water recirculation path in ORANGE,

when the heat pump is ON.



Sheet 4



Figure 4 shows a typical

horizontal hot water storage tank.  Stratification between T1 and T6 is

poorer than that seen in Figure 1, especially during periods of high

demand when cold make up water is mixing at greater rates. 



Figure 5 shows a typical

cluster of storage tanks piped in parallel.  This configuration closely

mirrors the thermal effects of Figure 4.



 In conclusion, it appears that Figure 3 provides

greater thermal stratification, greater Delta T between T1 & T6, and better

HP performance than Figure 5.  In the examples I sketched, it is

assumed that the volumes in the series and parallel systems are equal.  As

such, a parallel system won’t actually deliver a greater supply of hot water

during a sudden spike in demand – they are the same.  Furthermore, the

materials and labor cost to pipe Figure 5 appears to be greater than Figure

3.  All three tanks In Figure 5 must be properly balanced to

ensure even temperature distributions across all three tanks. 

 Plumbing multiple smaller vertical tanks in series can produce the

benefits of a large single storage tank where space limitations, and / or costs

(anything over 120 gallons must be an ASME pressure vessel I think) prohibit it.



Thoughts? Comments!  Thank you for the input!!!

Mark Eatherton

My thoughts...

As long as you have recirculation return, I don't think you will experience any appreciable stratification.



As far as stratification is concerned, from the stand point of stored energy, it makes no difference. 1 btu can only still raise one pound of water one degree F, regardless of stratification. In other words, a stratified tank, if mixed up, still can only hold so many btu's.



If recirculation is stopped, the water will naturally tend to stratify (A.K.A. stacking), but I still don't see a whole lot of advantage to it, unless you or someone else can prove otherwise to me, and I always try and keep an open mind.



And having the tanks in series seems to me that when you do run out of hot water, you will be out of hot water for a long time. And then there is the issue of tank failure, which is inevitable. Are people willing to be without hot water for a long period of time?



Having the tanks in parallel allows the heat source to spread its load evenly among the tanks.



It just seems to me that if series were such a good idea, regardless of the heat source, that we'd see a LOT more of it going on, and I've never seen in other than in drawing form with numerous inputs, as I explained before.



I didi have a fellow instructor up at Red Rocks that wanted to test this very theory. Before we got the chance, he quit and now works for the State prison system, and is the official prison plumber.



Others thoughts?



ME

mbshawaii

response

Good points.  Stratification doesn't do much to overall energy storage, as you said 1 btu is still 1 btu.  What it does do though, is allow a greater average temperature to be supplied to the building, and allow a lower average temperature to feed the ASHP.  Vertical storage is typically preferred to horizontal storage for these reasons - similar in the solar water heating field.  In a horizontal tank, the stacking is less prevalent, mixing is greater, and the delta T between top and bottom is usually slightly less.   I'm just not convinced if tanks piped in series can truly model the same thermal effects of a single vertical tank.  It is also true that recirc will stir up the tank and lessen the stacking effect, but by how much, i'm not sure.   I'd like to do some testing.  I've been in 4 friend's apartment buildings recently out here in Hawaii and have found these series installations in every one.  I'm guessing a single contractor went around and did these retrofits 10 years back or so.  My inquiring mind just wants to know why.  Thanks for the comments.

Mark Eatherton

After thorough and extensive internet research....

I suspect that it was a local invention. I saw no evidence of any manufacturer who recommends a series tank installation. It just really doesn't make any sense to me to do that. Maybe someone will chime in with more information.



Hawaii eh... What a bummer. No heating systems to fix there, and energy that is so expensive that people WANT to do alternative energy and conservation. Sounds like a conservation persons paradise. To say nothing of the surfing possibilities :-)



ME

icesailor

Series Vs Parallel piping:

Piping multiple water heaters in series is a really dumb idea. All that happens is that once both or more water heaters are hot, the only one that does any work is the first one where the water going in is cold. The hot water in the first tank goes in the second tank and so on.

If I do two, I make sure that the piping between the two is exactly equal so that the balance pressure between the tanks is as equal as I can get it. If it is three or more I do it as a parallel reverse return. I've sort of found that even if one tank is cooler than another, the water temperature will more or less balance out with use. The hotter water being lighter, will leave the tanks first.

As far as the return, on two tanks, I've always piped the return into both tanks at the bottom. It will immediately stop stratification. And I personally think that stratification is a very bad thing.

That's what I think, that's what has worked for me.

But just because a lot of systems you see are piped one way doesn't mean it is right. It may be that someone did it wrong and everyone else copied it.

Years ago, I did a lot of houses for a designer that designed houses that weren't designed to have plumbing in them. No place was allowed to put electric water heaters so they had to go under the house in a 48" crawl space. Others were digging a hole in the dirt to put a 60" high electric 80 gallon water heater in place. I came along and put two 50 gallon short tanks. They then had 100 gallons stored and twice the recovery because one electric water heater will only recover 18 gallons per hour, no matter how many elements you have. But two recovers twice the amount (36 GPH) because you had two elements running. One in each tank. And if they were balanced with equal piping, they emptied at the same time. In other words, one didn't run out before the other. If they weren't, one ran out before the other.

Last year sometime, there was a discussion of this multiple water heater piping situation. I think that the conclusion was that reverse return parallel piping was the way to go.

mbshawaii

thanks for the comments

ME - yes i did endless searching online as well, and couldn't come up with anything substantial either - BUT that doesn't mean that it isn't a good idea - it needs to be properly analyzed before i will discard.   I may build a scaled mockup and log some temperature readings to get a final answer



In Hawaii, we pay for paradise.  Energy is very pricy.  Right now, residential electric rates on Oahu are around $0.37/kwh.  On the outer islands, its over $0.40/kwh.  Needless to say, its all about saving energy, and new approaches to common scenarios are going to continue to gain in popularity. 



Icesailor - I completely agree with you.  Piping water HEATER tanks in series is pointless, for the reasons you mentioned.  However, in Hawaii we see over 50% of the apartment buildings use air-source heat pumps for their central hot water plant because of the warm climate.  The tanks are purely for storage - no heating elements.  Heat pump efficiency (cost of running them) increases substantially when the incoming water is cooler.  Therefore, in our storage tanks out here, we love stratification.  We encourage it any way possible (same concept applies to the efficiency of solar hot water panels).  Tall vertical tanks are preferred for this reason, but they don't always fit, and are very pricy to chopper onto a 20 story roof.  These smaller series piped storage tanks coupled with the heat pumps may be the answer to commercial size central DHW plants in the islands.



I'm looking for something beyond a gut feeling, would like to have a thermodynamic reason for why it is a good or bad approach.    Mahalo all for the input!

Mark Eatherton

I look forward to your iterations...

I assume you will be using someones FEA program to generate these graphic representations. As I said, I always try and keep an open mind, but my minds eye keeps telling me its not a good idea from numerous stand points (pressure drop, service ability, thermal stream flows etc).



As you said, it might just be one companies interpretation and or idea.



Thanks for contributing, thanks for thinking outside of the pipe and watch out for corral reefs.



ME

mbshawaii

Update on series vs. parallel

Well, i never got around to building actual models, but i did some thermal simulations using software and got the results i was expecting. 



Working on another project, i stumbled across this literature today, which goes into length about optimizing air-to-water hot water plants using storage tanks plumbed in series.  Interestingly enough, the author - Colmac -  is the brand we are seeing out here in Hawaii.



Best reading....

Larry Weingarten

Interesting discussion

Hello:  This is an interesting discussion to me as I like one main benefit of series piping, which is it cannot get out of balance.  The problem of serviceability is dealt with by putting valves on both hot and cold tank lines and running a bypass between them with another valve.  This allows any tank to be taken off line without interfering with service to the building.  Flow loss is a concern.  I'd want to know the maximum draw and make sure the flow path through tanks is not overly restrictive.  Bigger tanks/connections or even more than one string of tanks could accomplish this.  One question that came to mind looking at the diagrams from Colmac (diagram 1) was the possibility of water making up it's own mind about whether to flow through heat pumps or tanks when water is drawn.  This could create temperature problems if the water chose badly.  As piped, it looks like a last in - first out arrangement, which helps to insure you get the hottest water from the tanks.  One other thing is the claim of COPs from 3 to 5.  This is substantially better than two, which is good for residential air to water heat pumps.  Much to consider!



Yours,  Larry

Mark Eatherton

Interesting concept for sure...

My brain is still telling my gut that parallel is still better, but they have both been known to be wrong before, and will probably be proven wrong again ;-)



Too bad I am not still up at RRCC. Might be able to actually do the installation both ways and log it to see what happens.



Thanks for the insight MBS.



ME

mbshawaii

It may just be a Hawaii thing

These particular colmac heat pumps appear to be best suited for environments such as that which we have here in Hawaii.  The COP of the heat pumps increases as enthalpy  of the air increases, and as the incoming water temperature drops.  Therefore, all effort is made to keep the bottom of the "cold tank" as cold as possible, to promote stratification and reduce mixing.  Colmac recommends using storage tanks with incoming water baffles or diffusers to reduce dilution, and feed the coldest water possible to their heat pumps. 



Larry I also was wondering about how the water flows through the system.  After some discussions I think I understand it now.  When on, the flow through the heat pumps (internal circ pumps) is fixed, lets say for the sake of discussion at 3gpm.  If valves are opened and hot water demand is at say 8gpm to the building, then 5gpm will draw off the hot tanks, and 3gpm will pass through the heaters to the building.  Cold water will replace the hot water in the cold tank at 5gpm.  Once demand falls back off, say to 2gpm, the flow through the heat pumps will continue at 3gpm and start to replenish the hot water tanks at 1gpm, feeding backwards through the hot water tank.  Its really a pretty slick system.  I did some field measurements of Colmacs installed out here and they are indeed getting COP's of 3.5 - 4.  If you take advantage of the free air cooling, you can get combined COP's that are very high. 



The only major disadvantage of the series system i still see is serviceability, as was mentioned.  You can't just valve off a tank to remove it.  Because the battle out here is always energy costs (remember we're paying $0.34/kwh, not counting demand charges, and Gas isn't much better), I think the benefits outweigh the serviceability issue.  I enjoy all of your inputs.  Mahalo!

Larry Weingarten

water flow

Hello:  I'm wondering if a positive displacement pump might not be the only way to really control flow through the heat pump units.  If there were a greater pressure differential caused by a water draw, wouldn't that pull more water through the pump than 3gpm?  About serviceability and valves, it really is not hard to run one main overhead, tie into it with hot and cold from the tank and put a valve between those connections to allow you to unhook one tank from the system, even while it's in operation.  I've done this with no problems.  I'm waiting for high COP units to come to residential!



Yours,  Larry