The story so far. The straight piping to a radiator (approx. 3m) has been re-routed by an inverted "U" up the wall, across the ceiling and down the wall to facilitate the fitting of a shower.
Since adding this copper (approx 11m extra out and 11m extra back) the radiator no longer gets warm (it used to work without a problem).
I have installed a bleed at the highest point of the inverted "U" and I an happy there isn't air in that part (although I cannot be sure of course).
The original plumbing (unchanged) feeds two radiators off this branch (the main circuit is under the upstairs floor, so all the downstairs radiators are fed by pipes coming down from the ceiling). The other radiator on this branch gets hot, so there is water and return in the circuit. I have turned this one off and it has no effect.
I have turned all radiators off (except for the one in question) and left the pump running for an extended period. No effect.
Suggestions? I'm running out of them!!
Is a more powerful pump required? or am I missing something more basic?
Not being party to the original discussion I am short on background detail but 22m of extra (presumably 15mm) pipework would add a lot of extra drag on the system. On the face of it with all other radiators off you should be getting some heat into the remaining radiator if there is not an actual blockage somewhere in that part of the circuit.
I don't think that a more powerful pump would help much in this situation as it stands at the moment. How much extra pressure is required even to put any heat into the problem radiator? And how will you manage to balance the system with such a huge imbalance between that radiator and the rest of the system.
If you have an old microbore system the extra drag of 22m of microbore pipe would probably be impossible to overcome. (Does anybody still fit microbore these days?)
If it is just pipe drag then replacing 15mm pipe with 22mm pipe on this leg should make the system less unbalanced but I am doubtful whether that would help much in this instance given the failure to get any heat into the radiator at the moment.
Thanks. The other item I should add is that if I remove the bleed valve from the (cold) radiator, there is reasonable pressure and the water coming out will eventually get warm, so it can push it out that far - it seems to be the return that's the problem. It is plumbed in 15mm copper.
No, I just put it in the return part of the "U" as by removing the bleed screw on the radiator I can get hot water all the way through the new plumbing to (and through) the radiator.
It's getting it back again (and completing the circuit!) that's giving me the grief!
Just to check - if you turn off the valve for the feed, then bleed the radiator, do you eventually get warm water back through the return? Also vice versa? The question was asked in the original thread but I couldn't find an answer.
Oh, and are the feed and return close together where they leave the original circuit?
I am trying to visualise this arrangement in my head, and I am wondering why hot water would want to follow this route instead of going the easier route straight past the entry to the feed. Water tends to follow the path of least resistance and that vertical loop with a radiator at the end must put up quite a fight.
My CH fettling is a while back so a couple of probably naive questions.
(1) Is your a single or double pipe system i.e does the flow and return attach to the same pipe run or does the flow come from one pipe and the return go to another pipe?
(2) In either case, it sounds (and I could be wrong) as if you have extended the branch off to the radiator instead of extending the main circuit past the new radiator location and having your new feed and return local to the radiator.
In my Visualisation of the Cosmic All I can see that with a single pipe system if a flow of hot water goes past a feed and return with a radiator above, a little of the hot water may rise up the feed, and a little cold water drop down the return pulled along by the flow of water in the main system. Convection does its wonderful stuff and soon you have the cold flowing down and the hot flowing up as the mass of cold water in the radiator flows out and pulls the hot water in. I think. Also, with a dual pipe system there should be a pressure difference between the flow side and the return side as the water is pumped into the flow side and sucked from the return side. I think.
However I have more difficulty visualising this happening if the radiator with the bulk of the water is seperated from the main circuit by a massive inverted U, and also can't see how upping the pump speed is going to make much difference.
There are loads of systems where pipes go up and down between floors but I think (and could well be wrong) that the main circuit(s) flow past all radiators and the flow and return from the main circuit to the radiator is relatively short and usually below the radiator.
So - if you have not extended the main circuit, but just the radiator feed, then this may be the source of your problem. I am assuming that this is so, because otherwise with all other radiators closed off there would be nowhere else for the feed and return water to go but around the extended loop. Assuming of course there is no way for the water to bypass the complete branch if it feels like it.
It could be friction in the pipes. I have some more modern figures somewhere but typically can't put my hands on anything but a relic from the dawn of central heating design which states that half inch pipe (almost indistinguishable from 15mm) can carry a maximum load of 15000 Btus/hour and should be restricted to a maximum length of 80 feet which is about 24 metres and you appear to have exceeded that by a comfortable margin if you add back that original 3m x 2 even without considering the rest of the leg back to the junction.
By way of comparison the figures for three quarter inch pipe are 38 and
110 so on that basis even 22mm pipe would be near its limit.
But I am curious - why is the diversion anywhere near 11m x 2? A typical room is about 2.5m from floor to ceiling so your diversion should have added about 5m of extra pipework if it is in a normal domestic situation.
I have now tracked down and read the original thread so I think I know what the set-up is now although I still have a bit of a problem with the extra distance being as much as 11m.
However have you done as one of the original contributors suggested. Bleeding the radiator with first one and then the other of the control valves closed? If you get a similar flow rate in both circumstances then it is quite likely that it is just the the extra length of the pipework that is to blame.
An outside possibility is that if you bent the pipework your self you might have flattened the pipes at that point thus restricting flow but the converse of that is that if you used rt angle elbows for the bends they have a resistance equivalent to some extra distance of pipe.
Another stray thought. Do you have a bypass circuit on the boiler controlled by a pressure operated valve? That could be robbing the victim radiator of flow if all the other pathways round the circuit are closed.
Dave R has already scouted the posibility that your problem radiator is on a branch and there remains an alternative route back to the boiler even with all the radiators shut off but I can't see that myself as the flow pipe should terminate at the last radiator on the branch.
That question was just addditional background :-) My main concern was to find out if the main circulation route for that downstairs branch went directly past the radiator itself in the new location, or if the radiator was just hung off a very long and convoluted side flow and return connection to the main downstairs branch which served the original radiator. To me it is a big ask to take a radiator away from a 3m link to the main flow and return of the downstairs branch circuit (presumably on the same general level as the radiator and slightly below) and add a (presumably)
2.5m up + 2.5m down + 3m across inverted U to give 11m each way to the nearest CH circuit.
From all I have read so far the downstairs branch circuit is working fine in that it is feeding the other radiator. The OP is not getting any pumped hot water through the U shaped extension run. Ergo this does not carry the pumped flow and return. It is relying on convection and a bit of suction from the water flow across the cold return to move the water to and from the flow and return pipework of the downstairs branch, just as you do for a radiator which is close to the flow and return path. This might work (just) if all the piping was at floor level but I can't see this working with the inverted U.
I suspect the OP has taken the apparently easy route of just extending the piping which went to the original radiator up and over the room.
Summary: if the flow pipe in the downstairs branch is not part of the extension as described (and I do not believe it is) then I can't see this working.
I would extend the full flow and return of the downstairs branch to this new radiator location, and then just connect the flow and return of the radiator via a short pipe run to the extended main branch circuit. With luck, this should not involve much extra work.
Replace the T connectors where the original radiator connected to the flow and return of the downstairs branch circuit with right angled bends and remove the short run of pipe between original branch flow and return (1 pipe). Add T pieces by the radiator to allow flow and return when the radiator is off (you could even re-use the pipe you took out from the main branch :-) )
Two pipe system is a bit more work, as you should really extend both flow and return of the branch circuit which requires an extra two pipes to go over the room. One bodge might be to extend either the flow or the return and treat this bit as a single pipe system.
and the dual loop which I think I had in my last house is merely described as "inherently balanced but rarely practicable ". However, I am trying to work out the difference between this and the 'trunk and branch' design shown in the Wiki
formatting link
would be good to know how the flow and return is handled - is this a single pipe system, a tree trunk design, or something else?
Looking at the diagram in the Wiki if the radiator is connected in this way (tree and branch) then with all the other radiators shut off there should be nowhere else for the water to go (unless there is a seperate return loop) and so the problem should be some kind of airlock - even with the long pipe run the system should be able to warm a single radiator, especially with the pump turned up high.
However if the system is a single pipe loop then the problem may be that the single pipe loop has not been extended to the new radiator location.
One way to check would be to feel the pipe runs between the radiators on this downstairs branch.
If, with the other radiator on the branch turned off and cold, the pipework in and out of the downstairs branch gets hot but the pipework to the problem radiator stays cold then hot water is flowing round the branch but managing to avoid the problem radiator. If all the pipework in the branch stays cold then there is a blockage of some sort in the new pipework.
is essentially a schematic. All the radiators are shown in parallel but not how it would be laid out in a house. In practice the flow pipe goes to each radiator on the loop in turn terminating at the last radiator and the return pipe starts with the outflow of the first radiator and continues back to the boiler picking up the outflow of the rest of the radiators on the way.
A single pipe system, as the name suggests has only a single pipe connecting all the radiators on the loop so the radiators are in series and to get the same heat output out of each one requires those further down the pipe to be larger than those preceding.
That figures but what the diagram does not show (not familiar with the Wiki but it should be there somewhere) is the bypass loop that is almost essential with any modern system where the boiler will literally boil if the pump shuts of at the same time as the zone valve. A crude bypass loop is an always on radiator across the boiler before the zone valve but that tends to waste heat so the usual practice is to include a pressure operated valve so that the bypass circuit only comes on when the pump is on overrun and the bypass circuit itself only needs to absorb sufficient of the heat remaining in the boiler to stop it firing its overheat stat.
At the present time my money would be on a combination of the high resistance of the radiator circuit in question and the presence of a pressure operated bypass circuit.
Thank you for all the advice, it seems like it just aint gonna work the way it is.
To clarify the situation (as I understand the plumbing!). We have been in the house for approximately 11 years. About 5 years ago the combi-boiler went kaput and after a few expensive quotes (as we had to resite the boiler as a condensing boiler could not be sited where the old boiler was). We took the decision to install a wood burning back boiler. This required some additional plumbing for the central heating and an indirect hot water cylinder. The main plumbing run for the central heating is under the landing upstairs and is in 22mm pipe. Where the old pump was in the combi has become plain pipe and the new boiler tee'd in part way along. This has worked fine feeding all the radiators.
All the downstairs radiators are fed one pipe off each 22mm and in general there are one or two radiators in parallel off that circuit.
The circuit in question is at the end of the 22mm run and cannot be extended as there radiators are in a single storey extension.
It used to work :-) The pipe run (feed and return) came down to a "T" then split off left and right to go to the kitchen and cloakroom respectively. Because I needed a clean wall, I decided to route the pipes up and across the room, I have measured the lengths and there is
2.3m up, 2.5m across and 2.3m down when compared to the previous installation, so an extra total of roughly 14.2m extra pipework, plus of course the vertical climb for both feed and return.
What is most telling I suspect is that if you shut of the radiator on the other side of that "T" it still won't get warm, so I assume like electricity, it's finding an easier (low resistance) route.
I need to look under the floorboards (and that wont happen for a week or more as I fly out to Pittsburgh tomorrow morning) and see if this is the very end of the 22mm copper (which I think it is) so it shouldn't have any choice but t go around.
My other thought was to put the pump where it used to be and see if that has any effect?
Is there an easer path effectively in parallel with the rerouted section as unless its the only way through you will never get it to get hot. Can you not get the pipe under the floor instead? Brian
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.