Hot Water Recirculating Without A Pump?

It could be just thermo siphon effect, but it is pretty miraculous if it is!

Our old hot water system was heated by that method and it needed a 1" vertical pipe virtually straight from boiler to hot water cylinder to make it work.

Maybe the pump is under the floor boards? Have you listened to the pipe work, generally you can hear a pump if you press the sharp end of a screwdriver against the pipes and the blunt end against your ear. Maybe you'd have to listen when the boiler first came on it the hypothetical pump is controlled by and equally hypothetical (differential?) temperature sensor.

Maybe it just works because you believe that it works !

Del

No Pump required. Cold water naturally will fall and hot water will rise.

Since the Supply side is larger, the extra mass of water will retain its temperature longer and the smaller return side will cool off faster. This ensures the water circulates naturally in the same direction all the time.

Interesting!

I wonder if I am making the problem too complicated:

If it was a non-continuous flow, couldn't there just be a temperature-sensitve value at the "far end" of the loop that, when the water in the pipe is cool enough, opens up, just like someone opening a faucet, but instead of water going down the drain, it goes back to the tank in the other side of the loop.

Makes me wonder why "most" installations of a recirculating loop use a pump at all! (Which makes me think I am wrong in some important way...)

Thanks
- Mark

It works best in an open loop. The convection currents take a while to get moving. With the open loop it simply becomes an extension of the hot water heater.

The disadvantage is the hot water heater operates more often because water outside of the tank is being cooled and returned to the tank. The excess heat also slightly warms the house and slightly reduces the heat requirement of the furnace in the winter but can increase cooling costs in the summer.

"Makes me wonder why "most" installations of a recirculating loop use a pump at all"

If the loop does not go back to the tank, it does not have a heat source to drive it.

If the loop is all the same size of piping it will not work. The downward movement of cold water will happen on both sides of the loop equally.

Then there is also the profit margin...............

Makes me wonder why "most" installations of a recirculating loop use a pump at all!

dT is the answer.

The natural circulation works only with a decent dT. So the return lines are not insulated at all. This invokes a serious loss in heat and added costs.

The first central heating systems all worked on this principle, very reliable and all losses went into the house. The design and setup was a bit difficult so the installers abandoned this method.

This is the first time that I hear the use of this system to be used for hot water distribution. I would not do this as you need to keep the water above 55°C to be sure not to have bacteria growth. (Legionella and all the other nasty buggars) The return point at the boiler should be >55°C.

If you like the luxury of instant hot water you should swap to a local electric heater or electric heat tracing on the distribution lines, this with decent thermal insulation.

"you need to keep the water above 55°C to be sure not to have bacteria growth."

That's where we have a problem in the UK. Too many people were scalded by hot water, so now the maximum setting of a hot tank thermostat is under 60⁰C, and there must be scald protection at the taps in addition!

Then the powers that be wonder why we are getting iller as a nation.

On the one hand there's that rule, and on the other there's the work place one:

the 'Elf & Safety lot have worked out about Legionella et al, and decreed that the water temperature much be over 42°C. At least, that's what our Facilities Manager was told on his last training course. This has resulted in him altering all the under sink mixers so that the hot tap is now too hot to wash your hands under! It's just about bearable for a quick post-toilet ablution, but if you're trying to get work dirt off, you've got to fill the sink to get water at a sensible temperature, and then do this several times to effectively remove all the cleaning agent.

And we're supposed to reduce water usage...

42°C is way to low to avoid bacteria growth.

What you need is a mixing tap: automatically the water that comes out is mixed with cold water so that the temperaturen is approx 35 to 38°C.

But I wash my hands in cold water, better for your skin.

Convection currents? Same as the flow in the beaker in the high school experiment.

Fore home use it consumes slightly more power because of the greater heat sink effect.

What is interesting is that this is old technology that has quickly been replaced by the cheap practices of plumbers over the years.

40 years ago , most water tanks were plumbed like this.

When the electric motor costs became cheaper and copper tubing got more expensive, it became easier or cheaper to put in a pump then running extra tubing.

Now they have even stopped using copper and switched to plastic.

What concerns me most is that the priority of most plumbing and electrical courses have completely shifted over to 'how to save money' instead of 'how to do it correctly or safely'.

Unfortunately, due to the price of copper nowadays, you have to think about price! If I was to rewire my house in normal copper cored cable, the difference in price for the cable today as opposed to 5 years ago is frightening! I'm a sparkie here in Spain and to do a house for a client will cost me quite a small fortune in outlay and to be honest, the clients would prefer to stay with there out dated, dangerous installations than pay the prices for new spec stuff! As for plumbers, I think plastic is the way to go! Progress!!!

Yes, copper pipe nicely done can look really good...I must admit tho' that plastic pipe and speedfit connectors is v useful at times (did most of my solar panel plumbing with it up to the the pump and where it joins the rest of the system...it looks a bit like the 'pipes' screensaver!)...

S'pose 'plumbing' is the wrong term now..it's been ages since we used 'Pb - lead'

maybe it should be coppering or plasticing?

Del

I can make plastic look pretty good too.

You wrote:-

What concerns me most is that the priority of most plumbing and electrical courses have completely shifted over to 'how to save money' instead of 'how to do it correctly or safely'.

How right you are, how many heating systems have been priced to the bone to attract a customer, but he cannot remove a radiator to allow new wallpaper, without draining the system completely for example.

The same goes for the computer world, the cheapest price for a system that will work for 6 months is sold. Not what the customer really needs....

There is a great generalized truth in that Andy. Current regs in the UK make it illegal for somebody to do much more than change a light bulb. Go near a socket outlet etc and in theory the courts could really stiff you. Astonishingly, there is no clear definition of who is 'qualified' to do electrical work. The legislation is so loose that either a cowboy could rig your house, or you could be imprisoned for doing a job yourself. If you want gas work done CORGI is mandatory - check how that works, people with membership are not qualified for all works. A few years back I had a guy give me a quote: I checked by phone and they freaked at the very thought that he was flashing a CORGI card and implying competence to do the job. He was registered etc, but in no way certified competent. Like a lot of stuff, just because someone flashes the card doesn't mean nothing.

If you want gas work done CORGI is mandatory....

I'm not even certain that this is correct except by default... e.g. Regs say things like 'suitably qualified', and 'competent' and Corgi is the only recognised testing/approving body...(in UK) It can be a bugger to even buy parts for a gas boiler now, I say I'm buying them for installation by a Corgi fitter .

Even so, you can do the work yourself and then have it tested/ checked by a CORGI (Fortunately there's a guy a few doors down, who will do a quick check over with his sniffer for £20)

It begs the question what does 'qualified' mean legally? Do 'Qualified' and 'Certified' mean the same legally....obviously one can be 'competent' without any qualification or certification at all. Similarly a piece of paper doesn't mean you are any good.

Personally I don't want any Corgis in the house...it's ok for HRH but I draw the line...

Del

You don't describe what type of water heater you have. Some kind of gas or oil burner ? As Del describes, it should be able to track noise of a pump. Depending on the system you maybe able to kill the electric power supply for a few seconds to help discern noise change ( Though be aware, a boiler may need re-igniting). Without a schematic it's hard to say, but if you've looked every place then maybe it is pure convection ( Is there a noticeable temperature difference with height in the building ?).

As described thus far, I'm not sure what kind of system you have. I'm currently involved in demolishing an old house and re-installing a new system. It's what would be considered 'standard' in UK, but USA practice may differ vastly. I have various contact in the industry and had considered various systems that would have baffled the average plumber - discrete systems that allow high/low pressure flow via feedback control, in-line-heating, etc. Many of these have involved pumps in 'unusual' places. A line pump is not that big, but it will have an electric feed - can you track it via this ?

I hate to quote CR4 guidelines, but can we have a bit more info ? My house is a nightmare, and it's 'ordinary'. Re-plumbing it next week is going to be a pain. I'm going for an easy option - rip the whole lot out and start over ! Luckily I have a second base, so can do this gradually ( I'm down to 1 cold water tap now, as I rip out the guts to help the professional do it quicker). I'm not telling this lot my woes as I do this, but mail me direct if I can help. When my folly is done I may even write it up as a follow up to my 'bath' thread. <groan> <sorry for the plug>

[Don't worry - I'll kick myself for that bad one.]

If the "loop" ties in to the CW supply side of the water heater, I could see it working, in theory, via the siphon action of the slightly denser water on the 'cool' side of the loop pulling on the "down" side. Without a check valve though there is no way to keep the cool water from backfeeding when you open the faucet, and if there is a check valve I would be amazed that the siphon would be able to overcome the valve's resistance. Is there maybe a special fitting in the line with a small orifice that would permit low volume flow back to the heater tank? That would possibly do the trick. You said that the loop pipe was smaller. How much smaller?

this 'loop' keeps hot water through the upper floors instead of waiting. The costs is the lost heat = tradeoff. Hotels pump a hot water loop as all rooms need hot water fast.

A lot of people now use instant heat gas/electric water heaters, saves thw heat lost by the tank all night, but instant heaters are lower efficiency in gas service

In relation to house heating :

Instantaneous heating is increasingly the norm now in UK. However, if people have a hot water tank located (typically) upstairs the heat is not 'lost' - it serves as an extra radiator since heat dissipates to surrounding rooms. Whether this is good or not depends upon geographic location ( ie seasonal temperature vs heating requirements).

In the UK 'combi' boilers are currently the main choice - gas provides heat for radiators by thermostat, and to taps on demand. The fuel costs of gas vs electric is not that different, but the capital/maintenance costs can differ significantly. So called 'condensing' boilers are slightly misleading in terms of efficiency - they recover heat from flue gas, but only at optimum when a higher temperature difference exists. In other words when the system fires up from cold, good efficiency is achieved but this decreases with house temperature.

Hi

I think you hit the nail on the head here... as I start to understand the problem better, the issue seems to be how the cooler water could make its way physically back into the pressurized hot water tank.

The larger diameter pipe is 7/8" and the smaller is 1/2". I don't see anything obviously included in any of the copper piping... no valves or seams anywhere.

Thanks
- Mark

I believe this process is called thermo-siphoning. I have seen this used on stationary diesel generators.

Hotels have actively pumped hot water loops to deal with the problem of the AM demand. In addition, multi floor hotels have loops for each floor with pressure regulators etc.

drill here

http://www.google.ca/search?hl=en&q=%22hot+water+system%22+%2Bhotel&btnG=Google+Search&meta=

Hi Milo

Thanks for the link. This looks to be the likely explanation... the more I look at the loop the more it looks like it goes straight up to the 2nd floor and then comes straight back down and then to the bottom of the tank. I may not understand how the returning water can overcome the pressure in the tank (change in pipe diameter is probably important), but clearly this technique is real.

Thanks
- Mark

Sorry Del, but what is driving the "natural circulation" isn't the temperature physically. Physically the pressure difference from cool to hot line is the driver. This pressure difference comes from the temperature as already was described by others: the hot water density is lower than the cool water density. For a given gravity height this makes the cool loop section more "heavy" than the hot side - and bringing the system into circulation ... if the built in flow resistances would allow it (because the driving pressure difference is rally a small one).

Having the driving mechanism clarified here some words to pipe sizes: the circulation in first view is independent from pipe size - according to rules for flow dynamics the gravity height in the closed circuit is important for the effect. However, the intensity of flow has to do with the pipe sizes (e.g. if cool loop is bigger than hot loop section it will flow faster (vice versa).
It is very difficult to get such circulation designed. Some decades ago where many heating circuits working with this principle (and even noise-less refridgerators) the circuits were simple and wide steel pipes were used. The wide pipes ensured a quite low flow velocity and therefore the circuits were of low flow resistance. At this point it should be clear those circuits are not suitable to run a multi storage heating (or hot water) circuits: parallel sections are able to make other sections running reverse or stopping.
With todays small pipe sizes (e.g. copper pipes) the flow resistances often too big, to allow a natural circulation. Another argument why natural circulation sometimes explizitely is blocked (e.g. check valve) is the loss of heat, because you need a temperature difference to create the required pressure difference. For a two story home, if carefully evaluated, you might be able to allow heat loss at locations where you can take advantage from (it not really is a loss - in cool seasons).

If you want to provide some control for the circuit be careful. You may use adjustable thermostatic valves, but these valves must be able to provide a wide diameter finally because nobody can tell you, at what additional resistance the circulation would stop.

be happy, if the circualtion heat loss not really is a loss for you: saving the common pump holds you more independent from availability of electrical energy - and is environmental-friendly.

Albert

No need to be sorry!

Our explanations are the same, it is semantics we differ on.

Weight of the water coulmn. Density. Pressure. Its all the same thing.

When I said there was no pressure differential. I was trying to explain that there isn't any actual differential pressure between the bottom of the hot riser and the bottom of the cold return, as they are probably very close to eachother (just the otherside of the heater. .... Half way up the two pipes things are different.

Pressure is possibly an unhelpfull concept as it could easilly be argued the wrong way...! E,G

For a given gravity height this makes the cool loop section more "heavy" than the hot side.

Well if it is heavier... the pressure will be higher... and we all know the flow will be from high pressure to low pressure..therefore the water will flow from cold side to hot side! QED

The trick of trying to explain anything is finding the terms which happen to 'work' for the person you are explaing to.

Thanks Del, you are right, there are different explanations possible for the same thing - and one may find this approach more comprehensive than that approach ...

... personally I found the smallest risk for going lost between the different aspects of an issue for me is: verifying the physical nature to watch thinking about technics.

Regarding the pressure assumption at main tank you are right. If we would assume a constant water temperature in all sections (no circulation) there will be a constant pressure (minus gravity distance for upstairs locations) all over the circuit - according the somewhat static supply pressure from the water grid. The natural cooling effect always is overlaid to this static pressure and in no way dependend on the static pressure (therefore the return loop does not need to overwhelm the tank pressure because we have a closed circuit where the static pressure is the basic pressure effective for all locations).

The matter changes as soon as water is flowing out of the circuit. But for those periods it doesnt matter if the natural circulation would stop. It is quite a question of the circuits design whether the outflow becomes fed from the hot section, the cool section or from both ... (another risk for using natural circulation getting the hot using water all times available at all outlets).

Albert

Yeh...

Once the water is flowing it all changes, the narrow bore of the down pipe presumably becomes V important as it pesents a higher resistance to flow than the big bore hot riser, this presumably then all works to set up some mometum in the system with water moving in the right direction.

It would be fun to actually know all the forces acting...but I'll bet it was worked out empirically in the first place....some guy noticed the effect and tinkered until it kept going on its own.

Cheers

Del

The big driving force is indeed the density difference of the hot water in the riser and cooler water in the return lines.

This is realised by the difference in insulation (typically the return lines are not insulated, the risers are) and a simple check valve. (swing type)

The boiler/heater needs to be part of the riser, this is the initial start force, and here the repairs fail: modern heaters have the in and out at the same height (mostly at the bottom) and the initial difference can not be made.

In heating systems you will typically find overdimensioned radiators: the cooler the water enters the return lines, the better this works. Initial priming can take a while as a serious part of the riser needs to be heated up first. The power available from this dT is also limited, so the velocity will also be limited.

A couple of people have mentioned check valves to prevent reverse flow. I'd bet a decent wad of cash on thermo-siphon not being strong enough to open a check valve.

For thermo-siphon to work the pipe work would need to be carefully aranges...maybe the small bore return helps prevent back flow as it will present a higher resistance to flow.

My gas boiler used to use thermo-siphon for the indirect hot water heating.

A big bore pipe came out of the top of the cast iron heat exchanger, straight up about 15' then into the heating coil of the hot water cylinder, the exit from the coil was also big bore and went straight down re-joining the heat exchanger at the bottom on the oposite side. There was a radiator coming off the side of this circuit as a sort of heat sink to prevent the water overheating.

Del

I had an insulated thermal loop system in my old house some years back, no solar heat at all, gas. It went from the top of the tank to the kitchen and upstairs bathroom and back down to an entry point 2/3rd of the way down the HW tank. Fresh CW entered at the bottom.Tank was well insulated, but even with no water being drawn the gas would come on every hour for a few minutes.

No waiting at all for hot water = great.

In these days instant heat was available in electrical units, but I never saw a gas one.

http://en.wikipedia.org/wiki/British_thermal_unit

Del, your comment to "listen" for motor noise is right on. Yes, thermo-syphon can move a "swing-check" but not a "weighted flow-check" that requires the ΔP of a booster pump.

A thermo-syphoning system generally can not provide adequate heating comfort. Without a velocity of at least .66 FPS, ( 1 gpm in a 3/4" pipe ), your system is considered "lazy" and will only heat the beginning of circuit if it long. Typical design criteria is 2 FPS.

Although, with enough vertical height, you could achieve surprising performance from a "gravity" system, it would still be "sluggish"

i am wanting to build a convection loop system. i have many questions. i have put the information about the system in an off topic post.

The top of the 5' run would need to be the highest part of the system, as the water will lose its heat the further it is from the tank - and the near horizontal part would be losing heat to the environment, so the cooled water needs to be encouraged to return via the smaller bore, to the bottom of the hot tank.

"was planning to use the same size pipe for my hot out and cold in (3/4"). would you advise increasing the dia. of the hot out?"

Since your return is going through the digester, (immersed in liquid) and the supply side is not, I don't think this would be necessary. Most liquids will absorb heat 7 to 10 times faster then air. (This is why you will get hypothermia faster if you are in the water.) I believe the increased temperature wicking would drive the system quite well.

Those who have been around for a while, may remember some of the early motor-cars.

These cooled successfully on the thermo-syphon system, thus not needing any water pump to circulate the water coolant.

I'm amazed that you guys haven't seen this before as gravity returns are commonly installed in the States. Convection/gravity is so powerful in moving water that I have a radiant slab floor in the basement (2200 SF with >1500 feet of PEX) that heats much of the time without the pump. This is even tho' the water heater is on the floor of the basement, the supply line runs to the ceiling of the basement (8') as does the return
and I have even installed a 30" heat trap and it still moves water on its own without benefit of the pump.

sghunter@wcnet.org