Replies

1. 
   CloudHidden | Nov 05, 2004 09:00pm | #1

   Just off the top of my head, I don't think you'd get sufficient flow. When my pumps aren't running, I still get some thermosiphoning from the basement heat hot water tank to the top floor, but the effects don't go more than a few feet beyond the manifold. Without sufficient flow, you won't get sufficient heat transfer.

   Open to being proven wrong, though. Someone like David Thomas could prove it one way or the other with real numbers.

   Edit: On second read, I noticed I was thinking RFH, while it's radiators that was being discussed. My bad.

   
   

   ---

   Edited 11/5/2004 9:43 pm ET by Cloud Hidden

1. 
   JRuss | Nov 06, 2004 12:42am | #2

   I agree. We've tried it a few times in the past at the owners insistence. It's just not a satisfactory result. A pump is an incidental cost item for assured performance.

   Never serious, but always right.

2. 
   royboy | Nov 06, 2004 02:44am | #3

   Bill - I'd weigh in with a seriously qualified yes. It is possible to move heat in a hydronic system through convective flow (aka thermosiphoning) but it takes very intentional design of the pipe runs to make this work. And, as has already been said, it may not deliver heat fast enough for you depending on how quickly you need the heat.

   Bottom line is that convective flow, driven by the density differences in the hot (supply) & cold (return) lines is a relatively weak force, and is easily overwhelmed by friction in pipe runs or in fittings. To make it work, best to think in terms of making the pathway as smooth and straight as possible. Larger diameter is better. Use sweep elbows rather than conventional ones if possible. If plumbing with copper, consider bending soft tubing rather tto minimize elbows. The greater the height differential the greater the driving force. Requires continuous slope upward from source to destination and then continuous downward on the return. Things like conventional check valves are too much resistance. (Zomeworks in Albuquerque used to make a bouyancy activated check valve to work in convective systems).

   My experience with convective systems started with solar water heating sytems (no pump needed - very elegant!), and I've done many water heating systems that convect from a heat exchanger in a wood stove to a storage tank (here convection is a much safer thing than a pump, cuz if the electricity goes off when you have a good fire going, it doesn't take long to to be making steam, which leads to all sorts of ugly & potentially dangerous things). And I did do one setup for a friend, years ago, where a heat exchanger on a wood stove in the basement convected to a copper header & riser layout that ran - properly sloped - in the joist cavities under his sunroom/kitchen nook floor. I was dubious that it would deliver enough heat to make a difference but he said it worked great.

   Upside as I see it: very elegant & satisfying solution to moving heat, simple, no parts to fail. Downside: requires very particluar piping design, won't move heat real fast, and not easy at all to calculate/predict performance.

   Writing this reminds me how much I love(d) doing convective hydronic systems ... and reminds me about the whole concept of intentionally designing for convective airflow for heating/cooling. Remember envelope houses? Thanks for the memory jog ...

   Roy

3. 
   SamT | Nov 06, 2004 03:07am | #4

   Just make sure that the runs always slop up from the inlet to the heater to the inlet to the radiator and slope down from the inlet of the radiator to the inlet to the heater.

   Internally and externally.

   I don't know how the minimum slope you will need. Obvioulsy, the greater the slope the better it will work.

   The longer the immediate verticle from the output of the heater the better. Do not insulate the return from the radiator and the longer the immediate vertical of the return from the radiator the better.

   Start with a minimum of 1% (1/8" per foot). If that doesn't work, rebuild it with greater slope.

   Then try again.

   Good luck and happy checking accounts.

   SamT

4. 
   moltenmetal | Nov 06, 2004 03:58am | #5

   Our old place has a free convective hydronic system like most old houses did in its day up here.  It had a coal "boiler" with radiators fed by a supply header of 1 1/2" pipe with a return header of 2" pipe.  The coal was replaced with a rental conversion burner sometime in the '50s I suspect.  The current system works with NO electricity.  The thermostat is driven by a thermopile run off the gas pilot light.

   Designed right, they work great- very few moving parts- no pump to fail, and you still have heat when you lose power.  The coal boiler has to be 70 years old and it's still going strong. But it's VERY inefficient.  The unit has to run at very high water temperatures to get enough density difference driving force to work, and the radiators need hot water to make them effective too.  That means significant heat wasted, going up the chimney.  But the heat produced is very comfortable.  Less bulk air movement means you feel warmer.  Response time is far slower than with forced air, but we can re-heat the house in a couple of hours after letting it cool down all night or when we're away.

   We'll be ripping out and replacing the old coal boiler with a modern pump-driven modulating condensing boiler when we do our addition.  These run at much lower water temperatures and are hence much more efficient than the hotter density-driven system is. 

   I haven't run the calcs, but f your hot water heater is gas-fired, I bet the cost of electricity to drive the pump will be paid for in gas savings due to increased thermal efficiency.  If it's an electric hot water heater, that goes DOUBLE- the pump will save far more electricity in increased efficiency than it consumes in power.   And it will allow you to move the hot water heater out of the crawlspace into the insulated envelope of the house, so that any losses of heat through the insulation still end up heating your house (as will most of the losses in your electric motor- they'll end up heating your house too, ultimately).

1. 
   drbill | Nov 08, 2004 11:33pm | #6

   Thanks for the advice about gravity flow, etc.