As promised in my main thread, here's a design detail question.

One thing I am weighing is the DSH design.  I was originally thinking a single tank supplemented by the GTHP, but now I'm not so sure.

As I understand it, the basic element here is that the DSH can get the tank water heated up to around 90F, and you use the electric elements in the water tank to get you up to full temp.  But for full efficiency it seems that you'd need some coordination between the tank's thermostat and the DSH.

For example, if the heat pump hasn't been running, then the water tank is going to heat the water to full temp using it's own heat elements.  When the heat pump kicks on, the water tank is already hot, so there's nothing for it to really do, right?  So it seems like you end up with more instances where you can't really take advantage of the DSH.

With a dual tank design, my understanding is that you have a feeder tank (an electric water tank that's essentially unplugged), which is where the DSH puts the waste heat.  So when the heat pump is running, you have a tank that could be as warm as 90F.

The feeder tank is where the primary tank draws its replacement water.  So while the electric tank uses heat elements to keep the water at full temp, the warmer replacement water means that the primary tank uses less energy to recover.  It would also seem like you'd slow the temperature decline if you are drawing a lot of water (like for a bath).

Am I missing something?

For the most part you are correct. A desuperheater will not keep a buffer tank @ 90*... It wiil be kept @120~130*.

Anyone NOT using two tanks is not getting the full bennifit of a desuperheater. With one tank, the elements will always try to keep the water at set point. (120~130*) The desuperheater wants to do the same, but if the tank is at set point the desuperheater has nothing to do. In a two tank set-up the desuperheater can do it's job whenever it's running. It has all day and all night to bring 40, 50 or 80 gallons up to temp.

The two tanks should be piped according to the attachment.

We draw water UP through the dip tube, into the desuperheater and re-inject into the bottom of the tank. This insures we do not draw sediments from the bottom of the tank into the desuperheater coax or the circulator. The 1/2" ball valves are then used to limit flow so a delta T of about 10* can be achieved. Please remember to remove the handles so they can't be moved after setting the delta T.
The 3/4" ball valve BELOW the 1/2" tee can be shut off and ALL the water can be forced through the desuperheater, flushing the lines in a snap!

Bergy

I've found that in single tank situ's the water can convect and carry heat from the water heater.
Piping diagram above is fine but the inspector in me requires a t&p between the throttling valves (source of heat + valves = code required relief valve), though most don't do it. Other piping methods work as well as this.
J

Bergy,

Could you please explain the reason for limiting the water flow to achieve a ∆T of about 10˚ between water temps in and out of the desuperheater? Thank you.

Regards, Masoud

Masoud;

Actually, a 5~15 Delta T is OK. We were taught to shoot for the ten degree mark. If the water is moving through the desuperheater coil too fast it can not pick up enough heat, making it less efficient .

Bergy

I never cease to be amazed by the persistence of the idea that overly rapid flow through a heat exchanger reduces the amount of heat exchanged. It is simply not true. Faster flow rates increase heat exchange though the relationship is not linear. Stated another way, the lower the delta-T across a DSH coil, the higher will be the total btuh extracted, but there is a point of diminishing return and other design constraints

There are at least two good reasons to limit flow through a DSH coil so that Delta-T is 5-15. One is pump power - the extra watts start to exceed the incremental extra btuh extracted. The other is stratification in the preheat tank - overly high flow into the tank will reduce or remove useful stratification - warmer water at top of tank for use.

The most efficient way to achieve this is to size the pump small enough so as to avoid having to throttle it with ball valves.

Other than that, the sketch and explanation in post #2 are first rate, especially the bit about routing flows to avoid sediment pickup.

Joe - Strictly speaking I follow your reasoning calling for a TP relief, but are there any circumstances where a DSH could produce temperatures high enough to set off a TP? I suppose any heating of totally isolated water would substantially raise pressure, but I would think the source compressor would trip out on pressure or overload long before boiling water in a DSH - comments?

Engineer;

Thanks for the info...I'll have to do some research to get my mind around it.

Bergy

"properly sized pumps" would be most effective as you describe however that would require specific pumps/job. CM for instance designs pumps that can reach 50' 1 way to a buffer tank where I think Bard is only 30'. If you happened to set the buffer tank exactly that distance away or leave extra ft of pipe coiled to create the distance, no throttling would be required.
To say a system doesn't need a t&p because of a high pressure limit on a compressor, suggests we don't need one on a water heater because it has a thermostat, or a boiler because it has an aqua stat. Code says...source of heat + isolation = relief valve. DSH piping requires closed cell foam insulation because some systems get hot enough to melt the cheap stuff.
Is it likely that someone will close the throttling valves at the same time contactor sticks? Of course not. Safety controls however are generally installed to protect us from the "unlikely events".
Now here's the real challange- if the code requires it, but your inspector doesn't (because he's uneducated) what do you do? Remember contractors, you are on the hook if something goes wrong even if it passed inspection, because you are required to install to code. The other concern is harm to components if water gets too hot, and that could occur below boiling.
The easier fix is to only install one throttling valve, then there is no isolation. Or if you like to isolate the DSH as I do you can use the T&P from your buffer tank providing it has no independant source of heat.
Just sayin' ;)

I'll concede the TP issue - I didn't know some DSHs become so hot.TP valves are cheap and easy to install.

Regarding pump size, I think the trick is to recognize the relatively low heat and water flows associated with DSH and size reasonably. The specs for WF call for 0.4 gpm per ton but don't say much about the need to hit that on the nose. So long as one winds up with flow somewhere in that range and delta T of 5-15 ( a few degrees either side won't matter much) and such a range implies allowable flows varying by a factor of 3.

Just about the smallest circ pump made suffices for most installations. What one doesn't want is a 3 ton HP with a 10 GPM circ pump throttled way down with ball valves.

Getting back out of the heavy-duty thermodynamics here...

I'm looking at doing a 40 gal buffer with a 80 or 120 gal primary.  I currently have a 50 gal propane tank which is sufficient most of the time.  The only time it runs out of steam is when we fill the two-person jacuzzi in the master bath (which isn't often).

First off, am I really going overboard with an 80 or 120 gal primary?  The slower recovery time of an electric (versus the propane) has me skittish on undersizing.  Am I worried too much about that?

Secondly, I have the option of a Vaughn hydrastone-lined tank or the Marathon polybutylene tank.  The former has 10-year warranty, the latter is lifetime and is supposedly more energy efficient.  The contractor says he's been doing the Vaughns for a long time and has never had one go bad.  The Marathons are much more expensive.

Seems like I could go lower-end on the buffer tank (since will basically be turned off most of the time), and maybe invest in the Marathon on the primary.

Thoughts?

David,

Why is there such a large disparity between the proposed buffer and primary tank sizes?

If we assume a desuperheater diverts 10% of the total heat pump's capacity to heating water, 45˚ entering water temp to pre-heat tank, and that desuperheater cuts off at 125˚water temp. A 40 gallon tank attached to a 5 ton (48 K BTUH heating) GSHP will be at its top capacity in less than 3 hours of heat pump run time (if my calcs are loosely correct), not enough heat storage for a cold long night in NH, if you will need 80 to 120 gal hot water next day.

I have kept my old propane hot water tank to use when electricity is out for more than a day, despite heat loss / gain through the flue.


Regards, Masoud

The feedback I've gotten (and I forget where at this point) was that sizing the buffer tank larger than 40 would be too much for the DSH to handle.  Is this a faulty assumption?  Would I be better off with 80/80, or perhaps flip them to 80 buffer and 40 electric?

Size the primary for household maximum use during peak hour. Ignore the DSH / buffer for this calculation - assume the primary has to tote the entire load since some days during mild weather your geo HP won't run at all. All mainstream storage heaters have a first hour rating. It is the total of storage + how much more water it can heat in one hour. Get this right and no one gets a cold shower. Err on the side of caution (one size up if incoming water is particularly chilly or there is a particular aversion to running out of hot water. Standby losses are ALWAYS cheaper than a divorce lawyer

Try to size the DSH buffer for as much of a day's total household use as possible. That means the DSH has all day to preheat water. Time of heavy DSH production varies by season: late night / early morning in winter; afternoon and evening all summer. Insulate yourself from the time shift between production and use by having the DSH work on a whole day's supply whenever it "wants" to.

DSH is pure gravy - undersizing the buffer could cause you to lose some of its potential.

Storing water loses heat. I would use an on demand tankless water heater in place of tank #2. And since the water is mostly preheated, it can be a small one.

David,

Engineer gives good advice. I have listened to him before, without adverse consequences.

If an 80 gal primary tank is a good fit for the household hot water use, another 80 gal tank for desuperheater might be a good choice also. Even though a 5 ton system's deuperheater can make more hot water than 80 gal, when it is running a lot, in a 24 hour period. The hot water generated by DSH in HEATING mode is cheap compared to electric resistance, but it is not free. The heat, if not used for hot water, would go to your home, thus shorter run times for the heat pump. In COOLING mode, DSH generated hot water is almost free, or less.

During last winter, Tranq. 27 3 ton DSH provided most of my household's hot water needs (for 2 people in Michigan) through a 50 gal buffer and a 50 gal energized electric tank. Often the pre-heat tank's water was hotter than the primary's, set at 120˚. In spring, I noticed the DSH's absence on my electric bills.


Regards, Masoud

......the opposite (oversizing causing lost potential) is seldom true, but usage is key and paying more for capacity seldom used is probably the quickest way to lose on a DSH.
If you have 2 people and a circ tub, I'd go with the 80 gallon (for the circ tub) and a 50 gallon buffer (as that will be closer to your needs).
If you  truly don't use the tub much you can dial down the 80 gal tank and mix less cold water and just crank it up on the rare occasion you wanna soak.
J

I think that an 80 gal primary is a given at this point.  The First Hour Rating on the 80 gal electric is similar to my current 50 gal propane, so worst case I'm looking at a lateral move.

On the buffer tank, given the price differential between 40 and 80 gal is only $150, I think it makes more sense to go 80 there as well.  The heat loss rating on the tanks I'm looking at is 1/2-degree per hour.  That should translate into the primary tank kicking on extremely rarely during seasons where the heat pump runs, and during those seasons I effectively have 160 gallons at my disposal, making the tub a non-issue. 

The Vaughn hydrastone looks like a very good unit.  The insulation of the Vaughn and the Marathon look nearly identical, which is the most important factor when the DSH is the primary heat source, right? 

The Marathon has a very high EF, and Vaughn (to my disappointment) does not publish the EF on their units.  Anyone know where to get that info?  Seems like with electric they can't be all that different, and with the DSH supplying the bulk of the energy, paying an extra $900 for two Marathons doesn't look like it will have a good payback.

Thoughts?

80 gallon preheat will work as well, just extraneous as you likely don't use that much/day. Not a fan of the ultra efficient water heaters.
j

Joe,

Is there software for optimizing buffer tank size?

David had mentioned KIDS at home in another thread last September. That's the reason I thought a 40 or 50 gal buffer may be too small.


David,

For comparing standby heat loss from everyday electric water heaters (different models within a brand), I used inches of insulation provided. I agree with your assessment about the importance of insulation to reduce standby heat loss. I'm not certain that DSH is going to be the primary heat source (providing most of hot water in a year) for your hot water needs. DSH works best during long GSHP runs.



Regards, Masoud