How to lower heating system temperature?

[Daox]

If you use modern switching power supplies the losses will be minimal. However, I don't think you'll see a significant voltage drop over that short distance either. Just looked up a site and using 20 gauge wire, your voltage drop @ 1A is only going to be .78V. With 18 gauge its .5V.

Voltage Drop Calculator Tool

I think the energy savings from using the fans will hugely overshadow this decision. I'd do whatever is easiest/cheapest.

[Xringer]

If you do the loss calculations, Voltage Drop Calculator

I don't think it's going to be much of a problem. Unless you use 26 AWG wire, or telephone wire..


If you do, tie the conductors together and make the equivalent of a larger diameter pair of wires.

[Piwoslaw]

I got a cheap 12V 1200mA wall-wart, hooked it up to turn on/off with the circulation pump, and drilled a lot of holes between floors and rooms to get wires out to the radiators. I used 0.75mm2 (AWG19-AWG18) wire and noticed a funny thing: The voltage at the power supply is 11.9V but at the other end (up to 12 meters away), the open circuit voltage is 12.08V

I hooked up the first fan a moment ago, more tomorrow, then I'll do some A-B-A testing of the temperatures. Hopefully I'll have my Arduino data logger working by then.

[Xringer]

"a cheap 12V 1.2mA wall-wart"?? I'll bet you meant 1.2 amps..

I don't think you have any voltage gain in your 12 meters of cable,
but a little drift (0.18v) in your power supply..

But, I have seen cases where my voltmeter gave me strange readings,
and traced the cause to RFI from a nearby radio transmitter.
RFI and emissions from long runs of wire can be a problem.
But twisting the pairs helps out a lot.. Twisted pair - Wikipedia, the free encyclopedia

[Piwoslaw]

OK, so I've had computer fans mounted on the 6 largest radiators in our house for the last few months, and they've been wired to go on and off with the boiler's circulation pump. The pump in turn has been wired to stay on after the thermostat turns the boiler off, until the water in the system cools off.

I finally did some A-B-A (actually, A-B-A-B-A) testing with the setup and here are the results:


The explanation:

• I logged 4 temperatures (from top: boiler supply and return, living room and outdoor temp), once per minute for 10 days (Feb 01 to Feb 10),
• The thermostat's on/off temps were set at 16.25°C/17.25°C,
• The first two days were with the fans and circulation pump option on, then two days off, then on again, etc.
• The gas usage was recorded every morning, between 6:30 and 8:30am,
• The Avg temp data and the HDD data (rel 15.5°C) are midnight to midnight,
• During the test it was very cold, but no clouds at all, so there was some solar gain during the few hours of daylight,
• Doing this test during the coldest days of the year increased heating needs, so any differences between A and B would be larger.

Now, the interpretation of the data (please feel free to comment or correct anything):

• With the fans off all the time and the circulation pump off when the boiler is off (days 3-4 and 7-8), the boiler would be on for very long (over 4 hours on a few occasions). The return temp by then would be high, not much lower than supply temp, so the burner's flame would be reduced to a minimum. In other words: not too efficient.
• On the other hand, with the fans on and circulation pump working after the burner shuts off, the boiler turns on more often, but for shorter lengths of time. Normally it is on for 1 hour, then the pump and fans run for another 40-50 minutes. Usually the t-stat would turn the boiler on again within 15-30 minutes. When it was warmer outside (just under freezing), then the boiler would be on for about 1h, then the fans+pump would go for another 50-60 minutes, and then the next on cycle can be 2-4 hours later.
• Once the circulation pump stops, the water in the plumbing in the basement will cool quickly to a low temperature (basement temperature is 2.5°C-4.5°C), but the water in the radiators up in the house stays at a higher temp. This doesn't show up in the data, since the return and supply temperature sensors are mounted on the boiler in the basement, not up in the house.
• Crunching some numbers gave this:
  

  Date	Gas (m3)	HDD	Gas/HDD
  Feb 01	16	30.7	0.521
  Feb 02	20	33.6	0.595
  Feb 03	20.6	34.2	0.602
  Feb 04	19.5	33.4	0.584
  Feb 05	17	32.4	0.525
  Feb 06	16.5	31.5	0.524
  Feb 07	18	26.2	0.687
  Feb 08	16.3	24.5	0.665
  Feb 09	15.5	26.5	0.585
  Feb 10	17.4	28.8	0.604

  
  Summing that up:
  

  Fans+circ. pump	Total Gas (m3)	Total HDD	Gas/HDD
  ON	102.4	183.5	0.558
  OFF	74.4	118.3	0.629

  
  In other words, not using the fans + circualtion pump increased gas usage by almost 13%.
• I don't know how much of this improvement is due to the fans only, and how much to the extended circulation pump operation, since I didn't test each separately.
• It is possible that the difference in favor of the fans+pump setup is a coincidence. On one hand I'd think that 13% is more than the testing "noise", but
  1. The HDD and gas usage intervals have an 8 hour shift (I didn't feel like going outside to the gas meter at midnight for two weeks). Maybe I could estimate better values for them?
  2. The house has a thermal lag of a few hours, or even days, when the outdoor temperature changes.
  3. I have no way of estimating when and how much gas was used for hot water.
  4. There should be more data, maybe a whole month's worth? The reason I stopped is that the weather was to change - ie go back to its usual semi-predictable norm with temperature swings every 2-4 days. This was not compatible with the test.

The bottom line is that having the circulation pump go for another 30-60 minutes after the thermostat turns the boiler off, and having fans blow air over the house's radiators, can reduce gas usage by 10% or more during the coldest days of the year. I think that is worth the extra electricity (115W) and slight increase in noise from the fans.