Improving the effectiveness of a single pipe system

[JohnHa]

I have a single pipe system in a large-ish 1930s house and, guess what: the radiators at the end of the run don't get as hot as those at the start of the run.

It has been so cold recently that I wondered if I could do something fairly simple to get them hotter. Has anyone tried any of these wacky ideas?

1. Can I install a "flow reversal" system so I pump forwards through the radiators for a time and then pump backwards through them? The flow reversal circuit would be after the pump and boiler so that the pump and boiler direction was always forward.

I can see an immediate potential problem or even hazard as the water in the boiler (Vaillant Greenstar 30CDi) would be stationary and/or slow moving during the reversal so would heat very rapidly. However, reversal could be done each time the room stat switched and CALLed. Would I disturb huge quantities of sludge? The Tee joints feeding the radiators are correctly swept at each end.

2. Would things be improved if I found a way to pump the circulating water from the pipe into the far radiator rather than relying on convection? It would only need a tiny pump. Or is the water too cold to make that worthwhile?

3. Is it simplest just to install thermostatic radiator valves on all the start of the run radiators so they stop accepting circulating water when they are hot?

A problem-ette is that TRVs for single pipe systems are quite expensive :-( I have a radiator with a Danfoss Randall Single Pipe Thermostatic Radiator Valve body (£29) and Danfoss Randall RA2910 Sensor (£20) - none of this "two TRVs for a fiver at B&Q" you double-pipers can get :-(

4. Anything else?

Increasing the size of the end of run radiators isn't really feasible as they are double panel radiators positioned under windows and the worst one stretches the whole width of the room. Add a fan to blow air across it so it convects more hot water up from the pipe?

I don't have easy access to under the floor - I was persuaded (much against my better judgement) to tile the floor :-(

I will be very interested in any thoughts anyone has.

Thanks.

 

[gpbeck]

Could you get access to plumb in a new return pipe half way round, reversing flows at the cold end, so effectively 2 separate one pipe systems with a new common return?

 

[Chuck]

If it hasn't been done recently, I'd start by giving the system a clean and flush to clear out the sediment and see if that improved circulation. Next step would be to make sure the circulation pump was working and correctly sized. Then rebalance the radiators by working along adjusting the lock-shields until you get the same temperature drop for each radiator.

Properly-installed one-pipe systems can work just fine but they do need a bit of TLC to get them balanced correctly.

 

[JohnHa]

Thank you for your replies.

Could you get access to plumb in a new return pipe half way round, reversing flows at the cold end, so effectively 2 separate one pipe systems with a new common return?

That's a good idea, but access is a bit tricky in my case as the single pipes are under the floor, one of which is tiled :-(

If it hasn't been done recently, I'd start by giving the system a clean and flush to clear out the sediment and see if that improved circulation. Next step would be to make sure the circulation pump was working and correctly sized. Then rebalance the radiators by working along adjusting the lock shields until you get the same temperature drop for each radiator.

Properly-installed one-pipe systems can work just fine but they do need a bit of TLC to get them balanced correctly.

It should be pretty clean but that is probably sensible.

Your suggestion to balance the rads is interesting as I had always assumed I needed to keep the balance valve fully open so as not to impede the convection and they are fully open. However, balancing is effectively the same as adding a TRV to that radiator so I will definitely try it.

One reason I have always been hesitant to add TRVs is they reduce the heat accepted by a rad so I reasoned the return temperature will be higher.

I measured some temperatures during one boiler cycle in the cold snap when it was ~0C outside. The boiler only runs the fan for two minutes after being called and then fires. The boiler limited at 75C and the return temperature tracked the boiler temperature about 12 degC below it. The Grunfoss UPS 15-60 circulating pump is set to 2 in a range 1, 2, 3.

When I set the boiler to Max, it stabilised with flow at 82C and return at 71C.

Does that suggest I need higher output radiators to get the return temperature lower?

 

[doitmyself]

2. Would things be improved if I found a way to pump the circulating water from the pipe into the far radiator rather than relying on convection?

If you have a pumped system, how is it relying on convection?

If the system is pumped, there will be a pressure difference across the radiator. So water will flow through the radiator not due to convection, i.e hot water rises, cold water falls, but because of the pressure difference - water will flow from the higher pressure to the lower. Incidentally this is how two pipe systems also work; flow through a radiator occurs beccause there is a pressure difference across it.

On a two pipe system there is the same drop, say 20C, across each radiator, while a one pipe system has 20C shared between all radiators. So the first radiator may have a drop from, say 80C to 78C, while the last may have a drop from 62C to 60C. The actual drop across each radiator will depend its size relative to the total output. The radiator ouputs will be completely different from that stated in manufacturer's literature - higher at the start and reducing as you proceed along the chain. So earlier radiators will appear to be undersized and the later ones oversized.

 

[JohnHa]

If you have a pumped system, how is it relying on convection?

Thank you!

I had always assumed, incorrectly it seems, that because the single flow pipe "short circuits" the radiator, the water enters the radiator by convection or, as Danfoss puts it below, by gravity. I guess there is a bit of both.

You are saying that the resistance to flow through the radiator is sufficiently low that the "short circuiting" pipe does not short circuit the radiator and hence water is forced into the radiator by the pressure difference across the "short circuiting" pipe.

The temperature difference between flow at 75C and return at 63C (measured with a very old clip-on thermometer so may be wrong) is only 12 degC for me rather than the 20 degC you mention. I have always been worried that anything I do to restrict flow into the radiators (TRVs or balancing) will end up with the return temperature rising.

This is what Danfoss says about its single pipe TRVs and it is a bit confusing.

"circulation relies on gravity" is as I had assumed.

"requires high flow rates at low pressure drops" is a bit ambiguous.

I wonder if they are referring to the pressure drop across the TRV due to the convected flow? Or, as you say, that the pressure drop across the TRV and rad needs to be low because the resistance of the "short circuiting" pipe is low?

rather than the 20C

 

[Jonnyh88]

John, did you ever figure out a way to improve the effectiveness of your radiators? @JohnHa

 

[JohnHa]

Sorry - just seen this.

I haven't tried anything so far.

I think chuck's comment #3 is the best. I think that, over the years, my system has probably become a bit unbalanced and I ought to rebalance it.

I had always shied away from thermostatic valves but they could be good as radiators would shut down when that room reached temperature. leaving "hot" water to get to the "cold" radiators. I know that I would need "single pipe thermostatic valves" (about £30 each) and not the cheap TS valves for two pipe systems.

 

[Agile]

If I understand your graph you have a flow temp of 75 C and a return of 65 C.

Not terribly ideal for a condensing boiler but only a loss of 10 C across all the radiators in the house. I do not think you you could improve that very much.

Although increasing the pump speed might increase the return temperature by a few degrees. But that is not going to make a vast difference. At the expense of lowering the boiler efficiency!

 

[Keith Cole]

If it hasn't been done recently, I'd start by giving the system a clean and flush to clear out the sediment and see if that improved circulation. Next step would be to make sure the circulation pump was working and correctly sized. Then rebalance the radiators by working along adjusting the lock-shields until you get the same temperature drop for each radiator.

Properly-installed one-pipe systems can work just fine but they do need a bit of TLC to get them balanced correctly.
[/QUOTE

Having adopted a 10 rad one pipe system at home myself 21 years ago, I can agree , correctly balanced , the one pipe system works fine but can take a couple of hours to balance......I can see you have a drop of approx 11° between flow and return at the boiler .....if you can replicate that drop at each radiator by adjusting/balancing the lock shields you will have optimised it as best you can .
I also have Danfoss TRV's .....I was forever freeing up stuck pins that slide in and out of the glands ......so I bought new white plastic trv heads, just to smarten things up and was able to buy new Danfoss glands off eBay which cured the sticking pin problem ......both of these items can be fitted without the need to drain down.
Good luck !