Register to reply

Design of hot water circulation system

by TSN79
Tags: circulation, design, water
Share this thread:
TSN79
#1
Jan31-07, 12:46 AM
P: 357
Here I've seen different solutions. Some people simply add a T to the hot water pipe from the heater and run the circulation pipe alongside the hot water pipe connecting the two near the end. They then connect a pump to this. I don't really see how this has any effect. Wouldn't the hot water simply circulate endlessly and still cool at the same rate, since it's not being re-heated in any way...?

A different way I've seen is pretty much the same as above, except that the T now connects the circ.pipe to the cold water pipe that supplies the water heater. To me this makes more sense because it forces the circulating water through the heater for re-heating. The backside of course is that if the circ.pump fails, you would get cold water in the circ.pipe, though it can easily be prevented using a one-way-valve.

Have I got it right...?
Phys.Org News Partner Engineering news on Phys.org
Greater safety and security at Europe's train stations
Fingerprints for freight items
On the way to a safe and secure smart home
Artman
#2
Feb1-07, 08:45 AM
P: 1,591
There is some heat gain in the DHW supply pipe (actually both the hot and the cold side) adjacent to the water heater. Before they used check valves on the cold water inlet to the heater, they put a dip in the piping called a "heat trap" to hinder heat transfer into the cold water side of the piping.

But you're right, most sources recommend running the return circulation pipe into the cold supply through the water heater. This also preheats the cold entering water (until the circulator shuts off), which is better for the water heater. This arrangement of piping, with the recirc piping mixing with the cold water supply, is also recommended by mixing valve manufacturers.

Yes, the recirc system should have both a balancing valve and a check valve (one direction valve).

(And if a check valve or backflow preventer is used on the cold water inlet, you need an expansion tank suitable for domestic water on the cold side of the water heater.)
TSN79
#3
Feb1-07, 06:33 PM
P: 357
Thx! Would my first described situation have any effect at all?

russ_watters
#4
Feb1-07, 07:37 PM
Mentor
P: 22,303
Design of hot water circulation system

No, the first system wouldn't have any point. The second is the typical way to do it.

Artman - what's the balancing valve for?
Artman
#5
Feb2-07, 09:05 AM
P: 1,591
The balancing valve allows you to reduce the flow to the minimum required to keep hot water available at the last fixture at a 20 deg F allowable drop. A properly sized circulator would eliminate the need for this in a single riser system. If there are multiple returns to a single pump, the balancing valves should be included and a means to check the returning temps on each return line. An aquastat on the line of a single system can take the place of a balancing valve, and will shut the circulator off when the return temp rises too high.
Artman
#6
Feb2-07, 09:19 AM
P: 1,591
Would my first described situation have any effect at all?
There would be some, but very little. The heat transfer surface would only be the water to water interface within the pipe at the 'T' and the heat transfered through the pipe adjacent to the 'T'. As the warm water is circulated away from the 'T' colder water takes its place and picks up whatever heat is available for transfer. Not much surface area for transfer.
Wray
#7
Feb24-07, 12:56 PM
P: 2
I just installed a hot water circulator. The return piping runs to a tee at the cold water input to the tank. I was surprised to find that the water to my dishwasher when it was running was 10 to 20 degrees cooler than before. I believe this is an unexpected interaction with the water heater.

I have a "Power Miser" water heater. I expect that it saves energy by using two effects: (1) the water in the tank stratifies and there is a region of hot water at the top enabling the OFF set point of the thermostat near the bottom of the tank to be set at a lower temperature and (2) the ON set point can be set low because the cold water entering will trip it ON.

The circulator disturbs these conditions. It continually "stirs" the tank and it heats the entering (re-circulated) water.

I am reluctant to just turn up the thermostat because I want to switch the circulator ON and OFF. I'm doing some tests. I didn't expect this complication.
Artman
#8
Feb26-07, 09:11 AM
P: 1,591
Quote Quote by Wray View Post
I just installed a hot water circulator. The return piping runs to a tee at the cold water input to the tank. I was surprised to find that the water to my dishwasher when it was running was 10 to 20 degrees cooler than before. I believe this is an unexpected interaction with the water heater.

I have a "Power Miser" water heater. I expect that it saves energy by using two effects: (1) the water in the tank stratifies and there is a region of hot water at the top enabling the OFF set point of the thermostat near the bottom of the tank to be set at a lower temperature and (2) the ON set point can be set low because the cold water entering will trip it ON.

The circulator disturbs these conditions. It continually "stirs" the tank and it heats the entering (re-circulated) water.

I am reluctant to just turn up the thermostat because I want to switch the circulator ON and OFF. I'm doing some tests. I didn't expect this complication.
Did you install a check valve on the Return Circulator to only allow flow toward the cold side and stop reverse flow from the cold side when a fixture demands hot water? Without this check valve the water will come from the path of least resistance, possibly the cold side, blending with some from the hot side.
Wray
#9
Feb26-07, 11:57 AM
P: 2
Artman,
Thanks for your thoughtful suggestion. Yes, I have a check valve. I followed the installation shown for the Laing SM-303 pump. (One reason I chose this pump is that their web site has a copy of the manual.)

I checked the operation of the check valve by measuring the water temperature at the kitchen sink with the circulator running. Then I turned OFF the pump and closed a valve at the heater in the return line preventing any flow in that line. The water temperature was the same as I measured before.

Your mention of the path of least resistance touches on another issue. The original primary hot water piping is 1/2-inch galvanized. The new pipe I installed is 1/2-inch copper. The galvanized pipe has 50 years of corrosion. I'm thinking now that the corrosion may be greatly accelerated by the circulator.

As another side issue, I have a plumbing story. Not being an experienced plumber, I asked at the plumbing store what was best for pipe joints, putty or Teflon tape. I was told that it doesn't make any difference, either is just for lubricating the threads, that the seal is metal-to-metal. In my experience, this is WRONG. I used plumbers' putty (out of a tube) because it was easy. The threaded joints were between brass valves and copper adapters. All of them leaked no matter how hard I tightened them. (In one case I cracked the brass housing by over tightening.) I later found that 6 wraps of Teflon tape (the thicker kind) worked. You need something substantial to seal imprecise pipe threads. I always wondered how pipe threads seal--they don't.
ricoricio
#10
May9-08, 07:18 AM
P: 2
Quote Quote by Wray View Post
Artman,
Thanks for your thoughtful suggestion. Yes, I have a check valve. I followed the installation shown for the Laing SM-303 pump. (One reason I chose this pump is that their web site has a copy of the manual.)

I checked the operation of the check valve by measuring the water temperature at the kitchen sink with the circulator running. Then I turned OFF the pump and closed a valve at the heater in the return line preventing any flow in that line. The water temperature was the same as I measured before.

Your mention of the path of least resistance touches on another issue. The original primary hot water piping is 1/2-inch galvanized. The new pipe I installed is 1/2-inch copper. The galvanized pipe has 50 years of corrosion. I'm thinking now that the corrosion may be greatly accelerated by the circulator.

As another side issue, I have a plumbing story. Not being an experienced plumber, I asked at the plumbing store what was best for pipe joints, putty or Teflon tape. I was told that it doesn't make any difference, either is just for lubricating the threads, that the seal is metal-to-metal. In my experience, this is WRONG. I used plumbers' putty (out of a tube) because it was easy. The threaded joints were between brass valves and copper adapters. All of them leaked no matter how hard I tightened them. (In one case I cracked the brass housing by over tightening.) I later found that 6 wraps of Teflon tape (the thicker kind) worked. You need something substantial to seal imprecise pipe threads. I always wondered how pipe threads seal--they don't.
RE: "best for pipe joints, putty or Teflon tape" I use both. So do; Journey-Men Pipe-Fitters, Certified Med-Gas Plumbers, and LP-Gas, Natural-Gas, Oil-Line Mechanics. About six wraps of tape, and an even coat of Oatey brush-on pipe joint compound will guarantee 100% success. Be neat, and Take your time. Try to avoid getting stuff inside the pipe, it clogs stuff, like valves, aerators, and doesn't taste to good either.
ricoricio
#11
May9-08, 07:19 AM
P: 2
RE: "best for pipe joints, putty or Teflon tape" I use both. So do; Journey-Men Pipe-Fitters, Certified Med-Gas Plumbers, and LP-Gas, Natural-Gas, Oil-Line Mechanics. About six wraps of tape, and an even coat of Oatey brush-on pipe joint compound will guarantee 100% success. Be neat, and Take your time. Try to avoid getting stuff inside the pipe, it clogs stuff, like valves, aerators, and doesn't taste to good either.
FredGarvin
#12
May9-08, 09:31 AM
Sci Advisor
FredGarvin's Avatar
P: 5,095
Quote Quote by Wray View Post
Artman,
As another side issue, I have a plumbing story. Not being an experienced plumber, I asked at the plumbing store what was best for pipe joints, putty or Teflon tape. I was told that it doesn't make any difference, either is just for lubricating the threads, that the seal is metal-to-metal. In my experience, this is WRONG. I used plumbers' putty (out of a tube) because it was easy. The threaded joints were between brass valves and copper adapters. All of them leaked no matter how hard I tightened them. (In one case I cracked the brass housing by over tightening.) I later found that 6 wraps of Teflon tape (the thicker kind) worked. You need something substantial to seal imprecise pipe threads. I always wondered how pipe threads seal--they don't.
Whoever told you that standard pipe threads (NPT) seal themselves is incorrect. The only thread form for pipes that will seal without the use of an applied sealant is the NPTF, Dryseal thread form. They are not manufactured to the same tolerances as the standard, off the shelf NPT thread. They have tighter tolerances and are machined with other modifications to the standard NPT thread form. Even with the Dryseal form, a lot of times a sealant is used to help minimize installation galling.
jid999
#13
Apr27-11, 02:16 PM
P: 1
Hi,
Please someone can give me any suggestion regarding..
As i'm doing hot water circulation pump calculation of five star hotel(G+7). So as per Design the hot supply pipe size is more than return pipe so when i'm calculating the head loss in return piping keeping the flow GPM same as supply so i'm getting a very high head in return piping as compared to supply pipe. so should i keep same GPM to return pipe as supply pipe or not if not then how much should i keep gpm in return pipe.


Register to reply

Related Discussions
Looking for a design company for the dishwasher water flow system General Engineering 3
Water distrubution system design Mechanical Engineering 2
The best engineering system design Engineering Systems & Design 1
Meridional Overturning Circulation vs. Thermohaline Circulation? Earth 3
Digital system design General Engineering 7