I also have a recirculating two-line 3/4" HW system for the third floor, 40 feet above the HW heater. We turned off the pump. HW circulates naturally, and pump is not needed.

Just so you are aware of what is happening, both a "gravity loop" (which you describe) or an "always on" recirc pump can increase the amount of energy dedicated to water heating by several hundred percent.

For example, when we had a thermostatically-controlled "always on" pump, our water heater ran 2.5 to 3 times more often per day than with no loop.

Once I installed motion sensors in the bathrooms, the water heater returned to the same usage as with no loop.

Your post is kind of off topic. I just mentioned the other system in passing.

There are many tradeoffs involved in deciding on such a system, and they have been well described in other threads on this forum.

However, most people who want the retrofit system buy a pump which may very well not be needed.

The valve I describe wastes minuscule amounts of HW, and people who could possibly benefit from might be unnecessarily alarmed by your post.

Hey Downtowner,

There have been a couple of posts about convection recirculating systems. Thanks for pointing out the valves and linking to them.

The post that shadow700 wrote was on target. Any circulating system, including convection systems, use additional energy-- some of which can be significant depending on the construction, pipe material,etc.

Any recirculating system should be fully insulated including the return lines - which is code. There is a question about how effective a convection system is if the return loop is insulated.

There have been other posts about using the cold water as the return line for a recirc system. One issue is that the cold water is now warm, somewhat negating the "water savings" argument for a recirc system. The other issue that has been raised is that there is often bacteria present in a hot water heater that may not be what you want to drink.

Finally, a convection system will not work with a tankless system.

None of this is alarmist - it's simply that there are other considerations that people should be aware of before they install any recirc system and particularly one that uses the cold water line as the return.

Convection will not operate nearly as well with insulated pipes.

The heat loss from the return side is what drives the convection process.

It also needs the same size return pipe as the supply pipe to operate usually.

Pumped systems on timers (or even motion sensor triggered) are much better.

At least then you only circulate heated water when it might actually be needed.

Brickeyee,

Thanks for the heads up about convection 2 pipe system. (I'm actually going to turn my pump back on 24/7. I don't mind the operating cost if it improves comfort.)

As has been said many times before, timers don't work for retired people, or people who work at home. They also add greater installation cost, and something else that will go wrong eventually. (I am semi-retired, and work at home.)

Motion sensors cannot give ultimate comfort, at least in theory. If you turn on the water as soon as you enter the bathroom, you've only speeded up the appearance of HW by the interval between the sensor seeing you and the time you turn on the water. That might make sense in a powder room perhaps, or a public restroom. In any case, and I admit it's unusual, you have to walk through the bathroom to get to my bedroom, and my bath is sort of a public room since it is built to be a spa (I've had business meetings there) and is as large as any of our bedrooms.

But all that is really off the main topic. The main point is that it is a no brainer, if you are using a retrofit system, to first install the valve under the furthest sink in each line and see if you are pleased with the results. If not, then hire a plumber and electrician to install the pump that goes near the HW heater. I have found this approach works to my satisfaction.

My first post was the diplomatic response, now you get to know how I really feel:

Good lord, it's fine to suck up a third of the earth's energy resources, fight wars over oil, drill in protected areas, change the earth's climate, cause oil spills that devastate marine areas and send tanker ships full of cash to countries that don't like us very much as long as your pampered butt doesn't suffer any minor discomforts and you can entertain in your spa.

Great.

For everyone else, this isn't a "no brainer". Putting a thermostatically controlled valve that requires an uninsulated return line to work or an "always on" pump, particularly without pipe insulation, is just stupid, wasteful, gluttonous and shameful.

The lighter density of hot water will cause some current. Some flow.
Some current is normal, when a return channel is provided. A full circuit.
Whether it is enough to overcome friction and go all the way to the tap while it is still deemed "warm", this is the question.
It's not easy to predict, even by someone who is on site and experienced.

Whether in insulated lines or not, hot water loses heat steadily as it moves towards the tap.
In a circuit based solely on a convection current, the height and the configuration of the riser (e.g. straight up, or long flat segment before rising, etc) are two big factors. Some circuits may be have greater height than others, e.g. if your basement is deep (high ceiling).

Have we all heard about heat-trapping loops in HW lines coming out of water heaters? I'll find an image and post it. These are designed to hold the normal HW convection movement back: extra pipe and elbows are added to make the water go down and back up (a loop).

Uninsulated lines are not a requirement. Heat is lost in both insulated and uninsulated, so the water's density is greater in the return portion than in the first portion of the circuit.

To get out of this discussion, either install an Instant IHW under your kitchen sink, or accept lukewarm water at the kitchen sink tap, or close this tab. : - )

HTH
-david

I don't mind the operating cost if it improves comfort.

Which is all well and good ... I am not going to fault anyone for evaluating the factors and making the selection that is best for them.

However, for readers that see your post, it is important to point out that it is not a one-time cost of $34. It is $34 + $x / mo ... where "x" is dependent on a number of factors. For my house, "x" is around $15.

As such, the "break even" point for my system compared to a simple gravity loop is somewhere between 12 and 18 months.

"Good lord, it's fine to suck up a third of the earth's energy resources ..."

Maybe we should just forgo heated water?

The US may use a very large fraction of the "earth's energy resources" but our production for the resources we use is one of the best in the world.

Divide energy use by dollars of GDP.

And of course letting water run until it warms up is not without cost either.

Hey Brickeye,

All you had to offer were false choices:

There's a pretty big continuum between a water heating system that wastes half of its input (standard tank heater with
EF of .6 - uninsulated thermosiphon recirc system) supplying a residential spa large enough to hold business meetings VS forging heated water.

I don't really care what our energy/GDP ratio is - what we are doing is unsustainable environmentally, politically, militarily and economically. Neither does it follow that increasing efficiency will decrease GDP.

Never suggested that running water until it warms up is without cost - again a false choice. Although I'm not a fan of recirculating systems, there are responsible ways of implementing them. It's just that a constantly on pump, particularly with uninsulated lines isn't one of them.