What is the location in the desuperheater circuit to place a check valve. I assume you would place it right after the circulation pump were there is a greater pressure differential when the pump is running. Correct?

Should this be a poppet type valve with a light spring? Any recommendations?

As you can see, I have a thermo siphoning problem.

http://welserver.com/WEL0058/

My DSH loop is to a Marathon 50 gallon tank on the other side of the utility room.  From the GHP, the 1/2" copper tubing runs straight up then along the bottom of the floor joist and down to the DHW tank.  DSH-Out is tee'd into bottom of tank and DSH-In is tee'd into cold water inlet.  Installed two ball shut off valves with drains to isolate and bleed air on the horizontal run (highest point). With ball valve closed, the DSH loop temperatures will equalize between DSH-In and DSH-Out.

I would like to do the easiest thing and fix the problem. That is why I was asking about a check valve.

I am hoping that some one who had experience this issue will respond. Without the WEL temp sensors, I probably would have not notice it. I am getting 4-6 hours period between the electric heating element has to kick in.

Bob G.

I'm not sure I understand your problem. What you are saying is when you stop the flow in the DSH loop, the in and out temps on the DSH connections equalize. Would you not expect that to happen? Maybe you could explain this a different way and I will try to help you.

If you valve out and disable the DSH, how often do the elements kick in? In other words, what portion of your standby losses are attributable to the loss via the DSH tubing?

If I'm not mistaken, when I looked at the WEL server previously, the temperature showed that the DSH In was several degrees hotter than the DSH Out. The Wel server is not showing this currently.

I'm not an expert, but here is a quick thought. The cold water pipe going into our water heater has a "T" in it for the DSH In. This pipe, on our Marathon, is warm to the touch due to heat from the tank (it is insulated with pipe insulation though). Just wondering where your DSH In sensor is located? Could it be sensing heat from the tank?

For the experts: For monitoring purposes, where is the best place to install a temperature sensor on the DSH In and Out lines? Do you also suggest putting sensors on the lines closer to the tank to see what the heat loss is in the DSH lines, or is this not typically an issue?

P.S.  Now the temperatures have changed.  Currently 93.6 DSH In and 86.2 DSH Out with the DSH Off and the upper heating element On.

Do you have a separate buffer tank installed?  Just curious, how are you reading the temperature at the upper and lower elements?

Thanks to every one for responding,

Geome 09/17/2009 11:48 AM
If I'm not mistaken, when I looked at the WEL server previously, the temperature showed that the DSH In was several degrees hotter than the DSH Out. The Wel server is not showing this currently. Now the temperatures have changed.  Currently 93.6 DSH In and 86.2 DSH Out with the DSH Off and the upper heating element On.

Strange... Since I hooked up DSH loop back in early spring, the temperature was always between 100 to 110 degrees when the system was off.  The DSH temperature is now 85.6/85.1 degrees.  This first time I have witness this below 100 degrees.  I have attached charts from the previous three days.  The last time the HP ran was Sept 15th late afternoon.

I'm not an expert, but here is a quick thought. The cold water pipe going into our water heater has a "T" in it for the DSH In. This pipe, on our Marathon, is warm to the touch due to heat from the tank (it is insulated with pipe insulation though). Just wondering where your DSH In sensor is located? Could it be sensing heat from the tank?

My sensors is located inside the heat pump cabinet attached to the DSH copper tubing which is insulated.

For the experts: For monitoring purposes, where is the best place to install a temperature sensor on the DSH In and Out lines? Do you also suggest putting sensors on the lines closer to the tank to see what the heat loss is in the DSH lines, or is this not typically an issue?

I originally had sensors inside the heat pump and at the DHW tank.  With the system running, there was no more a 1 degree temperature difference.  A few months ago I lost all my sensors after a nearby lighting strike so I did not replace DSH sensors at DHW tank.

Do you have a separate buffer tank installed?  Just curious, how are you reading the temperature at the upper and lower elements?

This is a problem to get a good reading with the plastic Marathon tank.   The Marathon tank's heating element temperature switches is mounted on an aluminum plate which is attached to the threaded portion of the heating element.  My sensors are located between the aluminum plate and the tank wall.   I could not get the temperature switches to reliably cut out  120 degrees for the upper and 110 for the lower element.  The best I could do is average plus/minus 10 degrees.

   
engineer 09/17/2009 7:33 AM 
If you valve out and disable the DSH, how often do the elements kick in? In other words, what portion of your standby losses are attributable to the loss via the DSH tubing?

I just shut one of the isolation ball valves and turn off the HP since we will not need air conditioning for the next few day.


geotek 09/16/2009 3:54 PM 
I'm not sure I understand your problem. What you are saying is when you stop the flow in the DSH loop, the in and out temps on the DSH connections equalize.

Yes the temperatures should equalize when the system is not running but this was not happening since I hooked up the DSH loop back in early spring.  However, it decided to stop thermal siphoning and equalize temperature. 


Thanks again,

Bob G.

Just a quick update after closing the ball valve on the DSH loop.

1.  The DSH temperature has dropped and equalize to room temperature.

2.  The DHW heating elements is cycling now about once every hour instead of about every 4 hours. 
 

I am going to take a shot at this.  The cycling of both heating elements was causing turbulence within DHW tank.  Not sure how this effect would cause thermosiphoning since the cold water dip tube goes down to about 12" from bottom of the tank.

Note1: On the Domestic Water chart, DSH-P is the upper heating element (primary) and DSH-S is the lower heating element (secondary).

Note2: The dishwasher was running at about 5 AM this morning   

Bob G.

Simple thermosiphon  (http://en.wikipedia.org/wiki/Thermosiphon)

Convective movement of the liquid starts when liquid in the loop is heated, causing it to expand and become less dense, and thus more buoyant than the cooler water in the bottom of the loop. Convection moves heated liquid upwards in the system as it is simultaneously replaced by cooler liquid returning by gravity. Ideally, the liquid flows easily because a good thermosiphon should have very little hydraulic resistance.

I am not aware of convection loss on occasions where buffer tanks are employed.
Contrary to popular opinion I also install process pipe opposite that described above and do not see a build up problem in DSH coils.
Flow controls are available for boilers where pumps instead of solenoids are used for zoning to stop convection.
I am not a fan of mechanical devices where not required....less to fail.
To sum, I supply buffer tank from the top and return to DSH on the bottom....no check valve....no convection.
Two DSH's on the same buffer indicate check valves but otherwise......
If I had to mechanically impede flow I think I would lean toward a zone valve or flow control, a check valve is generally for maintaining flow in one direction, not impeding arbitrary convection.
JMHO
Good Luck,
Joe

"To sum, I supply buffer tank from the top and return to DSH on the bottom....no check valve....no convection. "

I'm trying to picture this but can't - maybe I've run afoul of prepositions.

I supply DSH from buffer tank cold inlet, and pipe DSH return to the buffer tank boiler drain fitting, though reversing those two would probably work just as well, perhaps better, thermodynamically. My priorities are to allow natural convection to establish stratification in the buffer tank during times when no hot water is being drawn, and to not disturb that stratification when hot water is being drawn. The goal is for the main tank (or tankless...yuck) primary water heater to get the warmest possible water from the buffer tank for as long as possible during hot water draws.

We may be saying the same thing in different ways.

I would avoid at all costs drawing water from the boiler drain of any energized gas or electric water heater owing to the hard water mineral sediment likely to accumulate there.

I do not see a role for check or other actuated valves in a DSH system other than in the case of two+ DSHs feeding a single buffer tank. DSH circ pump head pressures run in the range of just a few, <10 feet of head, so any restriction will materially decrease flow.

I pump from DSH to buffer cold water inlet and return to DSH from boiler drain. I would not pipe DSH to any but a buffer tank. No gas or electric hooked-up. This works nicely for us with no check valves.
j

Thanks Joe and Engineer,

I agree that a buffer tank would be the best situation.  Unfortunately my utility room is also the laundry room.  There is not much space to work with.  Is it possible that the cold water dip tube is located close to the heating elements and when the elements turn on, water is heating up in the dip tube causing the warmer water to rise thus the thermo siphoning.  

Compare the 9/16th graph to the following current graph. 

The 9/21 graph shows what normal off should look like. There is no turbulence within DHW tank and the water is stratified .   No thermosiphoning in DSH loop (valve is closed).  Looks like upper part of the tank is maintaining a steady 115 degrees.   The lower part of the tank will fluctuate between 100-110 degrees as the lower heating element cycles on and off once an hour.

The 9/16 graph shows the upper part of the tank fluctuating between 115 to 135 degrees and the lower tank fluctuating between 100 to 115 degrees.  The DSH loop always above 100 degrees.


So does this means that all single DHW/DSH setups experience this thermosiphoning issue?

I am going to open the ball valve again to set if thermosiphon will resume.

Thanks,
 
Bob G.

Yes it is likely that convection is the problem. A buffer doesn't have to be in the same room. If you can't use one I guess I would try a hydronic zone valve with end switch to activate tank, though you could try flow control (which can be over come by pump but not convection).
J

we wont plumb a dsh to a single tank setup, if you cant employ a buffer tank, the simpliest solution is to disable the dsh

I wonder if adding a U-shaped section of pipe 2-3 feet down then up would defeat the convection?

Just a thought

Thanks Kurt,

Just to update this thread for future reference, this weekend I got around to installing a check valve vertical in the DSH-Out line on my GHP. This change has take care of the thermosiphoning between the DHW tank and the GHP when the GHP has not been running. The DSH pump draw was 0.2 amps before and after installing the check valve.

The remaining question will be whether DSH flow is impeded by the check valve - amps alone won't tell that story, and DSH pumps aren't specified to provide much head. Mind you, check valves are routine in multi-unit DSH installs - just need to be certain the check valve is proper for the application

Kurt,

I selected a check valve in which I can pop open the valve by blowing through it.   In hindsight, I could of used a 3/4" check flow on 1/2" CU pipe to lessen the restriction through the valve.

Thanks,

Bob G.

As long as it works...

Another great example how inefficient some installs/designs are (yes, it happened to me too) and many will never find out because they do not monitor their systems. We discuss issues which might or might not add 2% efficiency, but a single checkvalve can improve performance of a desuperheater/buffer tank by many folds. I got burned with one of those early R-410 units, where the DSH was stealing heat in A/C mode out of the tank. I also had no idea how much of an issue thermal siphoning was, until I installed my 1st Welserver. The final conditioning tank was loosing so much heat that the electric element was turning on every 2-4 hours (without any water consumption). After a checkvalve was in, it turns on every 15-20 hours to make up a 10 degree heatloss.