CGroves 17th July 2013, 06:46 PM I'm always interested in ways to save energy. We have a solar hot water heater but in winter I have to run the boost a few times a week. We have a wood stove (an Australia made Thermalux (Wise Living Products (http://www.metaldynamics.com.au/stovesandovens.htm))) which we use to cook on and heat the house with. Unfortunately at the time we bought it the advice we received was that a wet back couldn't be coupled with a main pressure hot water system.
Over the years I've had in mind to try some other alternatives to a wet back to boost the hot water. Given the way the tank, solar panels and stove are mounted a system using a thermo-siphon wasn't possible. A system where the boosted hot water was at risk of boiling didn't seem appealing because i'd basically have to have an open system coupled to the closed system through a heat exchanger needing extra header tanks etc. I also wanted a solution where I didn't have to remove the stove which would be a big PITA.
After looking on the net I came across this site: Skip's Woodburner -- Water Heater Pre-Heater (http://www.michigan-horse.org/preheater/)
I thought that could be a goer. I didn't need the water to be raised by much, just enough for a hot shower, ~5degC from the 37degC that the solar panels provided on a winter's day. The design would allow it to be placed on the side of the stove without the stove needing modification.
I picked up some scrap chequer plate and some angle and made up the frame:
277312 277313
277297277298
The bolt in the middle keeps the heat exchanger pressed onto the side of the stove. The stove is ~75mm away from a wall so the heat exchanger sits between the stove and the wall.
The heat exchanger was made from 1/4" copper tube. I had to hunt around for this as some of the prices were exorbitant. I ended up getting it from Uniflex. Approx 16 metres of copper went into the coil. Bending the coil was an exercise in patience. I ended up using copper motor winding wire to keep the coils apart.
277299277300
277314 277315
I then encased the coil in plaster of paris. I suggest that if you're going to do the same to practise with it. My first attempt was a miserable failure with it going off as I poured it. I had to break that lot out. The second attempt was better (marginally). It was after that I remember a friend saying not to stir cornice cement/plaster too much as it will make it go off quick....
277316
277301
After leaving it dry for a few days I gave it a coat of high temperature paint and gave it a go on the stove. With the stove roaring the maximum temperature I recorded on the heat exchanger was 80degC. Next I tried a water test. I connected the garden hose to it and the difference temperature difference between the incoming and outgoing sides was 5-6degC. Which I was quite happy with.
Now that the heat transfer part seemed to work now I had to consider getting that water into the hot water system. With the current system the water from the solar panels is transferred to the hot water tank via a 240V circulating pump. So I thought a similar setup could be used to re-circulate the water from the tank through the heat exchanger. I took off the heat exchanger to do some testing with a 12V recirculating pump. That was a dismal failure. It didn't have enough power to suck up water and push it through the coil. The "friction" from the 1/4" coil was too much. I tested the heat exchanger by attaching the garden hose which was connected to a water pump and I got a respectable flow of several litres a minute. So I thought a 240v recirculating "solar" pump was the go. I ordered one off ebay and repeated the test.
277302
277317
I was disappointed by the result. The flow rate is about 620ml a minute with the pump set at 100W. You can see from the picture that the test was based on sucking the water from the bucket however I tried it with the garden hose also connected to the pump and the pump didn't appear to make much difference to the flow rate.
So whilst the heat exchanger works to exchange the heat from the stove to the water I don't think with the flow rate using a pump is sufficient to warrant connecting it to my hot water tank. The tank is 315L, I typically run the stove for 5hr a day, say 4hours with the heat exchanger producing enough of a boost. So that would mean approx 150L would be recirculated over that time. 5degC boost was at the heat exchanger so taking into account loses in the pipes between the heat exchanger and tank its likely 3degC is a more realistic figure. It looks like that the tank would only be raised 1.5degC over that time.
Given that to install the new circulating circuit i'd have to drill a couple of 20mm holes through a double brick wall, drain the tank, install piping, pressure relief valves etc., it doesn't seem worth the effort.
Anyway I hope this helps if someone is contemplating doing something like this in the future.
Christian
Stustoys 17th July 2013, 08:00 PM Hi Christian,
Why plaster? I thought it was a pretty good insulator?
How do you feel about breaking into your heat exchanger with 1/2" pipe and setting up say 6 parallel loops?
This is going to run at tank pressure?
5C rise on 620ml in a minute with 100W. My maths says you'd be better off with a electric element on the tank.
Your pictures arent working for me :(
Stuart
CGroves 17th July 2013, 08:37 PM G'Day Stuart,
I updated the pictures.
I used plaster to increase the mass and to even out the heat between the coils. Others seemed to have used it for their systems. Being "pour-able" its easy to encase the coils. Only problem is if you want to remove the coils later on to make it parallel it makes it pretty hard :C.
My math comes to the same conclusion as yours unless there's a way to increase the flow rate and keep the temperature differential.
Regards, Christian
Wrongwayfirst 17th July 2013, 08:50 PM Howdy,
Connecting a Wood Stove (http://www.solazone.com.au/connecting-a-wood-stove.html)
Heat exchangers on a wood stove will heat water to boiling point and beyond(steam). Hot water expands when heated. If it were a closed mains pressure system you would need a pressure relief valve and about 20 mins after you started the fire water would start to run out the pressure relief valve and run and run and run.
the link above is just the first one from google. The systems available place a coil inside the mains hot water service that by convection heats the water.
payback period is the only concern, most of these types of systems are sold into areas without a reliable power supply or no power at all.
cheers
Stustoys 17th July 2013, 09:07 PM Hi Christian,
Thanks for the pictures.
While not fixing your current issue, one thing I've liked the look of*. Have you looked into waste water heat exchangers that are fitted to the outlet of the shower before sewer. They preheat both the cold water going to the shower and the cold water going to the hot water system. The ones I have seen claim some pretty good numbers for heat recovery.
I'll try and dig up some links if you like....... it was a while ago.
Stuart
*I can't fit one without a pit and a pump. Though the claimed savings are high enough to run a pump depending on one thing and other.
Ueee 17th July 2013, 09:19 PM This one Stu? WaterFilm (http://www.gfxtechnology.com/) or Certified and approved drain water heat recovery (DWHR) | RenewABILITY Energy (http://www.renewability.com/power_pipe/index.html)
Christian, I would have made that with at least 1/2" manifolds and then 1/4" risers. As you say the flow restriction on such a length of small pipe is high.
Cheers,
Ew
simonl 17th July 2013, 09:28 PM Hi Christian,
Congratulations for having a go. I'm really interested in this sort of stuff, although I myself are yet to try anything similar. One observation, your 5 deg C increase with 600ml of water is governed by the heat transfer of the exchanger. So, with your system as it is, if you were able to say double the flow and get 1.2L/min then you would enjoy a temp increase of only 2.5 deg C. I was under the impression that a small flow rate should not necessarily be an issue if the heat transfer is high and the temp change is accordingly high. 100 L at 50 deg C has the same amount of heat as 200L at 25 deg C. Using the specific heat of water, your flow rate of 620ml/min and the temp change of 5 deg C you can quite easily work out the rate of heat transfer in Kw for your system.
Specific heat water is approx. 4.2Kj/Kg/deg C
so a change of 5 deg C of 620ml is 4.2x5x0.62 = 13Kj heat. This happens over 1 minute (60 seconds) for rate of heat, or power transfer is
13000j/60 = 217W
So, based on your observations, your heater exchanger is extracting just over 200W of heating power from your stove. This is the limiting factor of the system, not the flow rate. You may find that you have used too much copper coil. I'm sure a bit of research would reveal the optimum heat transfer of a copper heat exchanger per unit length. you may well find that part of your heat exchanger are "giving up" some of the transferred heat.
You could experiment with shortening your copper tube, using a different heat transfer medium instead of gypsum and perhaps some insulation on the outside part of the exchanger (side not facing the stove)
Just spitballing. It's not all gunna be gold!
Cheers,
Simon
Stustoys 17th July 2013, 09:36 PM This one Stu?
Thanks Ewan, that be them.
Stuart
CGroves 17th July 2013, 09:43 PM Howdy,
Connecting a Wood Stove (http://www.solazone.com.au/connecting-a-wood-stove.html)
Heat exchangers on a wood stove will heat water to boiling point and beyond(steam). Hot water expands when heated. If it were a closed mains pressure system you would need a pressure relief valve and about 20 mins after you started the fire water would start to run out the pressure relief valve and run and run and run.
the link above is just the first one from google. The systems available place a coil inside the mains hot water service that by convection heats the water.
payback period is the only concern, most of these types of systems are sold into areas without a reliable power supply or no power at all.
cheers
Hello Wrongway,
I'm familiar with that page. The point of this design is that the water doesn't boil as it's on the outside of the stove. 20min and the stove itself hasn't even warmed up. The flue pipe and wet back type options run the risk of boiling the water. It is a good practise though to have a pressure relief valve in the circuit.
Christian
CGroves 17th July 2013, 10:05 PM G'Day Simon, Ewan and Stuart,
Yes some things take some experimentation. You don't know unless you try it out.
That 5degC figure is based on a flow rate of 2.5L/min. I got that with the garden hose connected. If I could heat the hot water system that much over a firing of the wood stove i'd be happy as I wouldn't have to turn on the electric element in the tank (btw before anyone mentions I know there is a regulation minimum tank temperature of 55deg to prevent nasties).
I thought that the length would have been OK. The design from the link I reproduced used 75ft (22.7m) of copper tube whereas I only used 16m. Teaches me to believe everything I read on the internet :-). A manifold design is probably the way to go to increase the flow rate.
Waste water heat exchangers look interesting to pre-heat the cold water to the shower but it would be a bit of a nightmare to plumb it up at my place. Seems good if you have a basement and your hot water is always "hot".
Christian
Stustoys 17th July 2013, 10:13 PM One observation, your 5 deg C increase with 600ml of water is governed by the heat transfer of the exchanger. So, with your system as it is, if you were able to say double the flow and get 1.2L/min then you would enjoy a temp increase of only 2.5 deg C.
Well not exactly. If you double the flow the temp rise wont be halved(though in this case it might get close to it). But you could keep the flow the same and use less power at the pump. So still 200W out but only 50W in.(in nice round numbers there is likely more to it, I'm pretty sure there is a velocity squared rule in there somewhere).
But yes, ditching the plaster on the heater side of the exchanger would be the first move I think.
Christian,
You gave 5C rise, but what are the temps? as your input temp goes up things are going to get worse than the bucket numbers.
Stuart
simonl 17th July 2013, 10:28 PM Well not exactly. If you double the flow the temp rise wont be halved(though in this case it might get close to it).
Hi Stuart,
Is that because the efficiency of heat transfer is expected to change with different flow rates?
Mathematically, for a given system, you can either have 620ml at 5 deg increase or 1240ml at 2.5 deg increase for the same amount of heat input & same conditions?
Simon
Stustoys 17th July 2013, 10:52 PM Is that because the efficiency of heat transfer is expected to change with different flow rates?
ummm that might depend on your definition of efficiency.
But, the greater the temp dif the more flow you get.
To put it another way, there comes a point where doubling the length of pipe will do nothing(well ok maybe "almost" nothing) to the temp output. So depending on one thing and another, cutting the pipe on half and doubling the flow might make no difference(well ok maybe "almost") to the output temp.
There are also the savings in pump power mentioned above.
Stuart
jhovel 17th July 2013, 11:11 PM I can't add anything to the maths, but we've had a wet-back stove heat exchanger into a mains pressure tank for 30 years (until we took the stove out a couple of years ago for 'domestic' reasons).
Our hot water service was an off the catalogue model (we had to order it to made brecause even then they were only made to order): it has a 6m of 3/4" copper coil in the top 1/3rd of the height of the tank. That coil is/was connected to the we back bottom to top and top to bottom respectively. The water in this cuicuit is moved by thermosyphon only and heats up the top 1/3er of the tank directly and preheats the bottom 2/3rds by internal thermosyhon. The hot water tank also has connections in the bottom 2/3rds for solar panels (but we never installes any). The stove circuit has its own header tank with float valve and overflow to leave it open for expansion.
The mains pressure tank has a pressure relief valve of course. When we had the stove going all day in winter, the pressure relief valve would occasionally discharge a little when the tank got close to boiling temp and actually boiled a few odd times in those 30 years, but not often. When no-one is home to use up any hot water, the stove wouldn't be going either....
Also, wet backs for stoves are available or can be made in different sizes to ensure that not too much heat is extracted and prevent the temperature in a given tank to get too high. Our tank is 400 litres.
I have no idea if you can still get those kinds of hot water services, but they were commonly used in split level or multi story rural houses to enable hot water outlets higher than the hot water service - as in our case.
Might provide some food for thought.
Michael G 18th July 2013, 08:09 AM When I worked at Beasley hot water we made all sorts of HW systems like the one that Joe described. Now they have been bought by Rinnai they may have been axed by the company. I remember one unit was a instantaneous gas heater screwed onto the side of a tank that could be connected to solar. Provided that there is a relief valve somewhere and a way of circulating the water the rest is just pipework.
Stuart, what you are describing with respect to heat exchangers is correct with parallel flow units but not necessarily true for the type of exchanger that Christian tried. Heat exchange relies on temperature difference (greater difference = greater transfer) If the side of the stove is assumed to be a constant temperature plate, the longer the tube the more opportunity for heat to transfer. When the water temperature approaches the plate temperature the transfer will taper off but if there was only a 5 degree gain I would suggest that point is a way off. (The relationship is actually logarithmic)
Christian, well done for giving it a go. Advice after the event doesn't help much but like most others I think the tube size is too small. On the Solar panels Beasley made the risers were 3/8" minimum (may have been 1/2") and the headers bigger again. Plaster is a pretty woeful conductor of heat. You may have been better off strapping blocks of Al to the pipes or even (as we did for the solar panels) soft soldering a sheet of copper onto the bent tube (incidentally I'm told although I've never tried it that using cold tea will slow down the setting of plaster*).
Michael
*I also briefly worked in a plasterboard factory. The setting speed of the plaster depends a lot on the way it has been ground. We had stuff ground in a ball mill that would wet set in less than 30 seconds. One of the uses was sealing leaks on the mixing equipment.
simonl 18th July 2013, 08:24 AM ummm that might depend on your definition of efficiency.
But, the greater the temp dif the more flow you get.
To put it another way, there comes a point where doubling the length of pipe will do nothing(well ok maybe "almost" nothing) to the temp output. So depending on one thing and another, cutting the pipe on half and doubling the flow might make no difference(well ok maybe "almost") to the output temp.
There are also the savings in pump power mentioned above.
Stuart
Hi Stuart. Think we are on the same page here. By efficiency I meant the amount of heat actually extracted by the heat exchanger divided by the total heat available in the system (ie total amount available to the heat exchanger) x 100
Flow rate, coil length and coil ID, the most you can hope to achieve is only what is available in the first place. But by tweaking these parameters I would expect they will either increase or decrease this efficiency. There would be people out there that have studied this subject and I would imagine there would be some basic principles (flow rate, optimal heat transfer per lineal metre of pipe, optimal surface area of heat exchanger, etc etc) that would need to be followed to achieve certain efficiencies.
Simon
CGroves 18th July 2013, 10:43 AM G'Day Michael,
I think that if I get around to making a mark 2 of the heat exchanger i'd go for 3/8" coil and use a copper sheet to increase the heat transfer as you suggested. I know now to play around with the flow before putting it all together :-). I also agree with your comment to Stuart about the heat transfer from the side of the stove. With the side of the stove anywhere between 65 and 80degC at the temperature I was wanting to heat at 37 - 42deg there was still plenty of temperature differential for the copper tube to pick up heat.
Christian
Stustoys 18th July 2013, 11:33 AM Stuart, what you are describing with respect to heat exchangers is correct with parallel flow units but not necessarily true for the type of exchanger that Christian tried. Heat exchange relies on temperature difference (greater difference = greater transfer) If the side of the stove is assumed to be a constant temperature plate, the longer the tube the more opportunity for heat to transfer. When the water temperature approaches the plate temperature the transfer will taper off but if there was only a 5 degree gain I would suggest that point is a way off. (The relationship is actually logarithmic)
But halving the length of tube gives you that greater difference(in fact by cutting the coil in half you would more than double the flow assuming the pump was up to it. Which at .6l a minute it would almost have to be). But yes as we've said the first problem would appear to be the plaster needs to go. But given Christians other numbers, even a 33C rise with the current flow rate will only warm his water tank 10C.
We are assuming of course that the water in the bucket wasn't 70C.
As long as the heat exchange is a fairly air tight fit to the back of the heater and the back was insulated, would radiation and convection do?
We had stuff ground in a ball mill that would wet set in less than 30 seconds. One of the uses was sealing leaks on the mixing equipment.
Sounds fun, how hot did it get?
Hi Stuart. Think we are on the same page here. By efficiency I meant the amount of heat actually extracted by the heat exchanger divided by the total heat available in the system (ie total amount available to the heat exchanger) x 100
In this case where the heat avalable is not fixed(we arent talking about a flame in an insulated box were some sort of counterflow heat exchanger could be used)
If all you are worried about is the heat exchanger efficiency, have as many parrallel loops
as you can and crank the pump pressure up to get minium temp rise. As flow goes up and temp rise comes down your heat exchanger efficiency goes up.......... but you have to run the stinking pump :~.
As you say, there would be a sweet spot for system efficincy.(or even building the dmn thing)
Zwitter 18th July 2013, 01:21 PM Great project.
I have 2 observations to make and add to the discussion.
Pump. These are designed for closed system and basically overcome the thermo-siphon so the hot water returns from the solar panels to the tank that is usually lower than panels. They are not very good lift height rated and low flow so maybe in a loop would perform better?
Coil. Being vertically mounted the hot water will want to stay inthe top half of the coil loops? Most solar panels and HX offer a bottom to top flow. As has been suggested larger tube top and bottom with many smaller vertical tubes may siphon better. And better flow with less resistance etc.
I did similar experiments about 30 years ago. Currently have solar on the roof and a month ago the safety valve failed and it rained hot water at 3am ! Fun.
My latest experiment is in monitoring the boost to see how much the solar saves.
James
Sent from my iPad using Tapatalk HD
eskimo 18th July 2013, 04:37 PM the coil should have been a serpentine arrangement for vertical install..pump will over come your circular exchange
Oval tubing in lieu round would give better heat transfer...annealed copper tubing could be run thru a roller skipping a small section where a bend is required
a bending spring or hand made wooden roller would give nice tight bends
heat conducting paste applied to pipe and plate will also give better heat transfer
If a copper plate had been used brown tip silver brazing tube to plate my have been the way to go
Ueee 19th July 2013, 12:09 AM Hi Christian,
A few years ago i um "theoretically" made a pair of SHW panels. They are 1800 x 1200, 3/4" manifolds with 1/2" risers every 150mm. Fins are .6mm ally, fully wrapped on the risers and siliconed in place. Theoretically of course. The panels double as a window awning on our NE wall, they are tilted to 60 deg to get the best out of the winter sun, and are double glazed with clear laserlite, insulated with high temp insulation at the back.
They reach well into the 80's regularly and i have seen them at over 100 in summer. They heat a SS tank (ex gas heated tank) which then feeds out boosted tank via temp valve.
Pump is a 240v 3 speed job, controller is home made using an Arduino and DS18B20 digital temp sensors. Theoretically.
If i did it again i would have done some things differently, but overall i am very happy with the system.
The idea came from here: Solar Water Heating Projects and Plans (http://www.builditsolar.com/Projects/WaterHeating/water_heating.htm#1KSolarWater)
Here are some hi res rendered images of the panels....
Cheers,
Ew
CGroves 19th July 2013, 11:20 AM Ewan, nice job on the solar panels. Its amazing what you can do with 3D rendering and photoshop to bring these "theoretical" projects to life ;-). BTW is that Felder green in the background?
On the todo list is a solar air heater but that will be some time off.
Eskimo I like the idea of flattening the tubing. I saw the oval section tubing on the link to the waste water heat exchanger that Stuart mentioned.
James, I would have thought the pump would have performed better. The lift height was rated to 5m maybe it does perform better when its in circuit but the performance wasn't convincing.
Again thanks for the suggestions. I think i'll get a coil of 3/8" and play around with length vs flow rates. In term of picking up heat I think a design with a squashed coil soldered to a sheet of copper would be the go.
Christian
| |