Visiting a friend today, I noticed that a lot of air was circulating
with the water through the radiators. I found a key and bled the
downstairs rads. Upstairs the rads were lukewarm, and one was full of
air. I then heard an overflow happening from the loft tanks.
I climbed into the loft to investigate, and found that the DHW header
was OK. However the C/H header was full of dirty water which was quite
hot and overflowing the tank. The ballcock was half submerged.
I immediately turned off the pump, and rang the plumber who had
installed a new pump in October. I assume that he checked the system
before he left the property. He cannot visit until next week sometime,
but said there must be a 'blockage' causing the overflow of C/H water
into the expansion tank. He told me to isolate all the upstairs rads,
and just run on the downstairs rads.
I have no idea how the upstairs & downstairs piping runs are
configured, but found the situation so scary that I have left the pump
off for now. This is a 1950s ex-council house with a back boiler gas
fired C/H system, possibly added after the original build.
Any ideas what might have happened here?
I might be missing something but it's quite normal for the CH water to flow
back into the expansion tank - that's why it's called and expansion tank ;->
Now, if there was lots of air trapped, then air expands more than water when
it gets hot, so assuming some of the air got hot, it may have caused the
expansion tank to actually overflow to the drain/outside overflow.
Under normal conditions, the water level will rise and fall within the tank.
When it rises, it may do so with normally hot (not boiling) water. The
ballcock should be set (arm bent usually) to fill the tank enough to cover
the outlet to the CH. When the system heats, the water level will rise
(typically a few litres) but will not overflow if the ballcock has been set
Did the plumber drain and refill the system when he changed the pump?
Usually this is not necessary if the pump isolator valves are working, but
either they weren't or he wanted to clean out and refill the system, perhaps
with fresh inhibitor. If so, then even if the system is bled, it may need
rebleeding later as dissolve air comes out of the new water.
On the face of it, it really doesn't sound that bad. If the water is rising
and falling in the tank, it is not about to blow up. If the overflow is
working then there shouldn't be any danger in running the system.
Can you continue to bleed the air out of all the rads and run normally?
I would bet that after the last of the air is gone that the tank will no
longer overflow, but that the water level will rise and fall withing the
confines of the tank. It is perfectly normal for the ball to be fairly
submerged as it should be set low anyway as I mentioned above.
On an aside, if the water is dirty looking, it sounds like there *may* be a
lack of inhibitor which should prevent iron oxide formation from the rads.
Having said that, CH water doesn't always look too clean even with
Has this problem only started since the pump was changed, or has it always
been like it?
*Why* was the pump changed?
Sounds to me like it's either pumping over (is water flowing out of the vent
pipe back into the tank?) or it's sucking air through the vent pipe. If it's
a new problem, either the pump is set at too high a speed, or it's fitted
the wrong way round - and pumping in the wrong direction.
Hi thanks for prompt answers here.
Lots of questions from Tim & Roger. And yes it is a lady!
1. Pump was changed recently because it was knackered and was making
knocking noise. (It had been doing so slightly for the past 2 heating
seasons, but my advice to get it changed was ignored!)
While I was away on holiday it got seriously loud, so a local plumber
was called in. As I was not around I have no idea if he totally
drained the system. He probably did because the old pump had corroded
isolating valves which have now been renewed.
He certainly did not add Fernox or any inhibitor.
2. The pump maybe is set at too high a speed. I can reduce that and
see what difference it makes.
3. It isn't fitted the wrong way round. It has been working OK for the
past month or so.
4. I cannot believe that the header tank (that is the expansion tank
-right?) should ever be warm. The one in my own house is always cold,
and at a steady level. The hot water must have come from the overhead
pipe which is the expansion line. And then the surplus overflowed via
the overflow pipe to outdoors.
5. Just before I left the house I also noticed another fault. The hall
(the only one) thermostat is NOT controlling the pump. It clicks at
the ambient temp, but the pump is not inhibited when the temp setting
I am now wondering if this C/H system is simply over-heating and is
relying on the programmable timer to not loose more water from the
I did successfully bleed one upstairs rad with the usual key. Masses
of air in there. But 2 others need a large screwdriver to effect the
bleeding. When did those radiators stop being used?
Could end up with insufficient flow throug hthe boiler.
The clue is in the name _expansion_ tank. As the water in the heating
system is heated, it expands back up into the _expansion_ tank, and
yes, that will be warm, or it wouldn't have expanded.
Pump is probably controlled by the boiler, and it will likely leave
it running for up to a couple of minutes after it stopped firing.
Overflowing the expansion tank could be because there's insufficient
clearance between the fill level and the overflow to allow the water
to expand without overflowing, or it could be because there's a leak
between the hot water system and the central heating system, which
is usually in the coil in the hot water cylinder. If the level doesn't
drop back to normal and stay there when the heating system is cold,
then it's the latter (holed coil in hot water cylinder).
[email address is not usable -- followup in the newsgroup]
On Sun, 6 Dec 2009 00:24:37 +0000 (UTC), email@example.com
(Andrew Gabriel) wrote:
I don't know about the maths and the volume of water in the system
that has to expand when it is hot - but there is a length of vertical
expansion pipework before the header tank. Surely the water expansion
will mostly take place in there?
About the idea of a hole in the coil. Would this not make the
domestic hot water as dirty as the radiator water?
No, because the main header tank is higher than the CH header usually -
so the water in the cylinder is at slightly higher pressure than that in
the coil. So fresh water enters the heating circuit, rather than the
other way round,
OK John, thanks for that info.
Going to revisit the problem this morning and look again with a 'cold'
start before I switch the pump on again. I'll take a infrared
thermometer along to see how the radiators heat up.
I'll see what the pump speed is currently set at, and I'll look into
the C/H expansion tank and see how that has settled down overnight.
I'll use a big screwdriver and bleed the two odd radiators upstairs.
I'm pretty sure they are full of air.
I'll set the hall thermostat at min, and leave it there initially. The
C/H should not start at that setting, I know. If it does then the
wiring to the new programmer need attention, I would think.
I'll report back how it all works out.....
There are several possibilities for this problem and it's easy to work out
which is the cause. I've suffered from all of them on my own system at one
time or another.
1) Holed coil in the immersion tank. With the heating and hot water switched
off this problem will still manifest as the water levels in both of the loft
tanks trying to equalise. The level in the large cold water supply tank is
usually higher than that in the small heating system supply tank and often
the CW overflow goes into the HS supply tank and thence to the outside via
that tank's overflow. If the HS tank is overflowing with no obvious inflow
from the ballcock or the CW overflow pipe then a holed coil is present. Only
clean water should enter the HS tank though.
2) Pumping over. If the boiler pump is at too high a setting or fitted the
wrong way round, or indeed if the pipework has been badly planned so the
vent pipes are on the pressure side of the pump, then the immersion tank or
boiler vent pipes can pump over into the HS supply tank. However this should
not cause the tank to overflow because no additional water should be
entering the system. The overpumped water just finds its way back into the
circuit again via the tank and keeps circulating. However the HS tank will
of course get hot.
3) Ballcock set too high. The HS tank needs some room for expansion as the
water in the system heats up. Usually you set it at about half the height of
the tank and the ballcock will end up somewhat submerged. If the level is
too high the tank starts to overflow as the water expands into it however it
should then stop overflowing as equilibrium is achieved and the water in the
tank should not get hot.
To those who suggested that expansion might cause the vent pipe to overflow
into the HS tank, that I'm afraid is just silly. The water in the vent pipe
and HS tank will always be at the same level because they're both open to
atmosphere i.e at the absolute most this will be the level of the HS tank
overflow pipe. Only if pumping over is occuring can water come out of the
So the most likely cause of both an overflow plus hot water in the HS tank
is pumping over which will obviously only manifest with the boiler running.
By seeing what happens with the boiler both off and on you should be able to
determine the cause.
FWIW when my own system started pumping over and it couldn't be cured by
reducing the pump setting I just capped the immersion tank vent pipe off
with an old radiator valve. Expansion, how my system was arranged at least,
could occur just as happily back up into the HS tank via its outlet pipe and
the vent pipe was actually redundant. I fact I can't envisage a system which
actually really needs a vent pipe although I'm sure that will lead to a
storm of protest.
Finally when my immersion tank sprang a hole in the coil and while I was
still in shock at the cost of a new one a lorry turned up with a brand new
tank on it which someone else had ordered and misdelivered it to my house
(or at least they left it with a neighbour because I was out). I never did
find out who it was really meant for but was grateful to the plumbing gods
for such a timely intervention.
Dave, many thanks for this very comprehensive reply, which I have
printed out. My thoughts now are that your Item 2 covers this problem.
#Whether the pump is in backwards we will establish later with the
plumber. Now, after reading all the replies when folks realised the it
was a gravity/pumped system, I'm not so sure again.
What I can do this morning is to restart the system, with the pump on
the lowest speed setting, and see how it all performs. I can check
what is going on in the header tank as well.
May I suggest that when you first turn the system on, you set it to HW-only
mode and note whether or not the pump starts. If it's a fully pumped system,
the pump *will* start on HW-only. If it's gravity HW and pumped CH, the pump
won't run until you also turn the CH on. That should help to determine what
sort of system it is.
The maths is easy - there are net expansion calculators on the web, or use
the expansion coefficient of water (this varies by temp, but it's good
enough to stab an average value in and see what the expansion is between say
15C and 80C. The harder bit is estimating the volume of the rads and pipe.
Nowhere near. That pipe will be "full" to the same level as water in the
tank. You'll get a good "couple" (+/-) of litres expansion out of a typical
system depending on volume of system.
On Sun, 06 Dec 2009 09:52:05 GMT, Harry Bloomfield wrote:
No, think about it. The water in the vent pipe is connected directly
to the feed pipe via the system. Water finds it's level, so the water
level in the vent pipe is the same as that in the tank.
When the system starts heating up the water expands and that increase
in volume is accomodated by water moving up the feed pipe into the
tank. Raising the level in the tank and thus in the vent pipe but
that is only a inch or so of pipe so not a significant volume
compared to raising the level in the tank by and inch or so.
Correct. At worst the tank should only get a little warm. Hot is bad
so is any flow from the vent pipe, even a regular drip or a short
dribble when the pump starts.
That would be hard as the pipe to the tank is full even when cold. To do
what you propose would require a magic ballcock that cut off when the water
was several feet down the pipe below the base of the tank!
No magic involved...
The boiler, rads, pipe and tank will of course all be full of cold
water initially. As the system heats up there will be point at which
the water in the pipe will hot below a certain level and cold above it.
As the water warms and expands, the interface between the two will
gradually rise up that pipe.
Thermo mixing will be a minimum in a narrow pipe, unless and until the
hot water expands up to the tank, where it should quickly mix with the
cold water in there and the tank will then cool. A warm header tank
means something is wrong.
OK - I see what you are saying - my bad. I though you were saying there
would be no appreciable increase in the water level in the tank. You were
actually saying the tank shouldn't get hot.
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It 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.