I guess that many of us have come across primary vented systems which
pump *over* - where there is a constant flow of water out of the vent
pipe back into the F&E tank.

But is the opposite also possible - where air is sucked in through the
vent pipe, and water forced back up the fill pipe into the F&E tank?

I have recently been tearing my hair out - what's left of it(!) - trying
to sort out a friend's heating and hot water system.

The system has a floor-standing conventional boiler (Potterton
Kingfisher Mf - c.2003) in the garage, with a vented primary circuit
which has a fill & expansion tank in the attic. The hot water system is
unvented, employing a Heatrae Sadia Megaflow indirect cylinder (c. 1995)
in the first floor airing cupboard.

The CH and HW primary circuits each have their own pump and are entirely
separate - only meeting at the boiler connections. [There is some relay
logic which runs the boiler when one or other or both circuits are
calling for heat]. The only motorised valve is part of the Megaflow
installation. This opens in response to the tank stat, and its auxiliary
contacts turn the pump on.

The basic system was designed and installed by my friend several decades
ok, but has had a new boiler and pumps, plus replacing the vented HW
cylinder by an unvented one in the meantime. The current configuration
has remained unchanged since about 2003 - and has worked ok until recently.

Problem is that my friend is now in a care home, and his wife sends for
me when things go wrong!

The initial problem reported to me was lots of gurgling and banging in
the pipes when the HW was being heated. When I investigated, I found
that the F&E tank in the attic was completely empty - the ball valve had
stuck shut, and all the water had evaporated. That would explain the
symptoms, I thought - with air entering the system because of the lack
of top-up water. So I freed up the ball valve, allowed the system to
re-fill and bled air out from all available places.

When I turned the system on, the boiler would run for a couple of
minutes and then stop. The HW pump was running, but there didn't seem to
be any flow - in that the flow pipe from the boiler wasn't getting hot.
At this point, the pump speed was on its minimum setting - which is how
I had found it. When I turned it up to max, the boiler fired for much
longer, and the pipes got hot. I could still hear air going through the
system though.

There is obviously still a problem. My feeling is that there is still
air in the system, which I haven't succeeded in removing. With the pump
on its lowest setting, there is no circulation - even though there used
to be - and the water obviously doesn't get hot. So presumably, there is
an air lock.

If I turn the pump speed up enough to get the water to flow, we're back
to the dreadful gurgling and banging - all the way up to the F&E tank.
On one occasion, the water in the F&E tank became very hot, and the
level rose to the point where it started to overflow. [I began to wonder
whether there was an internal leak in the coil in the HW cylinder, but
the level went down again fairly quickly once I had turned everything
off - so I'm pretty sure that's not the case].

My feeling is that, when the pump is run fast enough to make the water
circulate despite an air lock, air is actually being drawn into the
system through the vent pipe - what might be described as pumping
*under*, as per the title. Anyone come across this? In case it helps to
visualise the system, I've drawn a somewhat crude diagram of just the HW
circuit, which can be seen at
https://app.box.com/s/43v06520ifm7s3iw2usqj0vwp83o7nwz As can be seen,
the fill and vent pipes connect to that circuit in the vicinity of the
HW cylinder. Note: The pump is on the return side, close to the boiler -
pumping *towards* the boiler.

Any suggestions as to how I can get all the air out, and get the system
working as it did before the F&E tank dried up?

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

On 12/05/2018 21:37, Roger Mills wrote:
I guess that many of us have come across primary vented systems which
pump *over* - where there is a constant flow of water out of the vent
pipe back into the F&E tank.

But is the opposite also possible - where air is sucked in through the
vent pipe, and water forced back up the fill pipe into the F&E tank?

I have recently been tearing my hair out - what's left of it(!) - trying
to sort out a friend's heating and hot water system.

The system has a floor-standing conventional boiler (Potterton
Kingfisher Mf - c.2003) in the garage, with a vented primary circuit
which has a fill & expansion tank in the attic. The hot water system is
unvented, employing a Heatrae Sadia Megaflow indirect cylinder (c. 1995)
in the first floor airing cupboard.

The CH and HW primary circuits each have their own pump and are entirely
separate - only meeting at the boiler connections. [There is some relay
logic which runs the boiler when one or other or both circuits are
calling for heat]. The only motorised valve is part of the Megaflow
installation. This opens in response to the tank stat, and its auxiliary
contacts turn the pump on.

The basic system was designed and installed by my friend several decades
ok, but has had a new boiler and pumps, plus replacing the vented HW
cylinder by an unvented one in the meantime. The current configuration
has remained unchanged since about 2003 - and has worked ok until recently.

Problem is that my friend is now in a care home, and his wife sends for
me when things go wrong!

The initial problem reported to me was lots of gurgling and banging in
the pipes when the HW was being heated. When I investigated, I found
that the F&E tank in the attic was completely empty - the ball valve had
stuck shut, and all the water had evaporated. That would explain the
symptoms, I thought - with air entering the system because of the lack
of top-up water. So I freed up the ball valve, allowed the system to
re-fill and bled air out from all available places.

When I turned the system on, the boiler would run for a couple of
minutes and then stop. The HW pump was running, but there didn't seem to
be any flow - in that the flow pipe from the boiler wasn't getting hot.
At this point, the pump speed was on its minimum setting - which is how
I had found it. When I turned it up to max, the boiler fired for much
longer, and the pipes got hot. I could still hear air going through the
system though.

There is obviously still a problem. My feeling is that there is still
air in the system, which I haven't succeeded in removing. With the pump
on its lowest setting, there is no circulation - even though there used
to be - and the water obviously doesn't get hot. So presumably, there is
an air lock.

If I turn the pump speed up enough to get the water to flow, we're back
to the dreadful gurgling and banging - all the way up to the F&E tank.
On one occasion, the water in the F&E tank became very hot, and the
level rose to the point where it started to overflow. [I began to wonder
whether there was an internal leak in the coil in the HW cylinder, but
the level went down again fairly quickly once I had turned everything
off - so I'm pretty sure that's not the case].

My feeling is that, when the pump is run fast enough to make the water
circulate despite an air lock, air is actually being drawn into the
system through the vent pipe - what might be described as pumping
*under*, as per the title. Anyone come across this? In case it helps to
visualise the system, I've drawn a somewhat crude diagram of just the HW
circuit, which can be seen at
https://app.box.com/s/43v06520ifm7s3iw2usqj0vwp83o7nwz As can be seen,
the fill and vent pipes connect to that circuit in the vicinity of the
HW cylinder. Note: The pump is on the return side, close to the boiler -
pumping *towards* the boiler.

Any suggestions as to how I can get all the air out, and get the system
working as it did before the F&E tank dried up?

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?

Could the header tank drying out allowed some crud that was sat in / on
the water there to have been drawn into the system? Hence causing a
partial blockage - say at the zone valve in the HW primary... that could
cause unexpected circulation patterns.

--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

I once had a similar problem with constantly needing to bleed. I had exactly the same theory and turning down the pump fixed the problem. In my case building work had temporarily removed a number of rads causing the problem. You could seal something over the vent pipe (eg a bit of balloon) and see if it is sucking or pushing stuff out. Are you bleeding all rads with all pumps off? Also, assume the ball valve is working correctly now so fully replenishing the system as you remove the air?

If airlock is the problem you could try switching all rads except 1 on and run the pump then try bleeding it. Continuing one by one. Having said that, sounds like when you put pump on max it appears to be strong enough.

When the pump is on low, do all rads heat up and are there cold spots on them? Top cold implies air so bleed , bottom cold implies crud

On 12/05/2018 21:37, Roger Mills wrote:

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?

Are all the radiators getting hot? If not you may have a airlock in the
pipes supplying them which may be the source of the some air getting
back to the to the pump or boiler. If so run the pump at full speed and
just turn on one radiator at a time to try and force the water through
the problem pipe length.

Since the tank ran dry have the radiators been bled? While the tank was
dry air could have been sucked in and distributed to the radiators.

--
mailto : news {at} admac {dot} myzen {dot} co {dot} uk

In message , John
Rumm writes
On 12/05/2018 21:37, Roger Mills wrote:
I guess that many of us have come across primary vented systems which
pump *over* - where there is a constant flow of water out of the vent
pipe back into the F&E tank.

But is the opposite also possible - where air is sucked in through the
vent pipe, and water forced back up the fill pipe into the F&E tank?

I have recently been tearing my hair out - what's left of it(!) - trying
to sort out a friend's heating and hot water system.

The system has a floor-standing conventional boiler (Potterton
Kingfisher Mf - c.2003) in the garage, with a vented primary circuit
which has a fill & expansion tank in the attic. The hot water system is
unvented, employing a Heatrae Sadia Megaflow indirect cylinder (c. 1995)
in the first floor airing cupboard.

The CH and HW primary circuits each have their own pump and are entirely
separate - only meeting at the boiler connections. [There is some relay
logic which runs the boiler when one or other or both circuits are
calling for heat]. The only motorised valve is part of the Megaflow
installation. This opens in response to the tank stat, and its auxiliary
contacts turn the pump on.

The basic system was designed and installed by my friend several decades
ok, but has had a new boiler and pumps, plus replacing the vented HW
cylinder by an unvented one in the meantime. The current configuration
has remained unchanged since about 2003 - and has worked ok until recently.

Problem is that my friend is now in a care home, and his wife sends for
me when things go wrong!

The initial problem reported to me was lots of gurgling and banging in
the pipes when the HW was being heated. When I investigated, I found
that the F&E tank in the attic was completely empty - the ball valve had
stuck shut, and all the water had evaporated. That would explain the
symptoms, I thought - with air entering the system because of the lack
of top-up water. So I freed up the ball valve, allowed the system to
re-fill and bled air out from all available places.

When I turned the system on, the boiler would run for a couple of
minutes and then stop. The HW pump was running, but there didn't seem to
be any flow - in that the flow pipe from the boiler wasn't getting hot.
At this point, the pump speed was on its minimum setting - which is how
I had found it. When I turned it up to max, the boiler fired for much
longer, and the pipes got hot. I could still hear air going through the
system though.

There is obviously still a problem. My feeling is that there is still
air in the system, which I haven't succeeded in removing. With the pump
on its lowest setting, there is no circulation - even though there used
to be - and the water obviously doesn't get hot. So presumably, there is
an air lock.

If I turn the pump speed up enough to get the water to flow, we're back
to the dreadful gurgling and banging - all the way up to the F&E tank.
On one occasion, the water in the F&E tank became very hot, and the
level rose to the point where it started to overflow. [I began to wonder
whether there was an internal leak in the coil in the HW cylinder, but
the level went down again fairly quickly once I had turned everything
off - so I'm pretty sure that's not the case].

My feeling is that, when the pump is run fast enough to make the water
circulate despite an air lock, air is actually being drawn into the
system through the vent pipe - what might be described as pumping
*under*, as per the title. Anyone come across this? In case it helps to
visualise the system, I've drawn a somewhat crude diagram of just the HW
circuit, which can be seen at
https://app.box.com/s/43v06520ifm7s3iw2usqj0vwp83o7nwz As can be seen,
the fill and vent pipes connect to that circuit in the vicinity of the
HW cylinder. Note: The pump is on the return side, close to the boiler -
pumping *towards* the boiler.

Any suggestions as to how I can get all the air out, and get the system
working as it did before the F&E tank dried up?

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?

Could the header tank drying out allowed some crud that was sat in / on
the water there to have been drawn into the system? Hence causing a
partial blockage - say at the zone valve in the HW primary... that
could cause unexpected circulation patterns.

The above has just happened to me! Putting Fernox into the header after
an unplanned drain down, I was horrified to watch floating crud (some
sort of algal growth) disappearing down the make up pipe:-(

Difficult to judge any impact. Initially the boiler tripped a few times
and the circulation was very noisy. Lots of trapped air due to long
horizontal pipe runs which cleared over a couple of days. This is an old
boiler (25 years) so I am hopeful that the heat exchanger is not overly
susceptible to blockage. There is an air vent which I keep venting
manually.

The need for relay controls may imply that the valve does not have a
microswitch. I am currently trying something similar with a diverter
valve and plan to report back. One initial thought is that the relay
will be instant while the associated valve may take several seconds to
fully move. Possibly leading to pump flow going somewhere unexpected
particularly if the valve fails!
My circuit is pressurised.


--
Tim Lamb

On 13/05/2018 02:21, John Rumm wrote:
On 12/05/2018 21:37, Roger Mills wrote:


Am I missing anything obvious?

Could the header tank drying out allowed some crud that was sat in / on
the water there to have been drawn into the system? Hence causing a
partial blockage - say at the zone valve in the HW primary... that could
cause unexpected circulation patterns.


I suppose that's possible. But the water *is* circulating through the
coil and heating the HW. What would happen if it had gone up the vent
pipe, and blocked that?
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

On 13/05/2018 09:33, wrote:
I once had a similar problem with constantly needing to bleed. I had exactly the same theory and turning down the pump fixed the problem. In my case building work had temporarily removed a number of rads causing the problem. You could seal something over the vent pipe (eg a bit of balloon) and see if it is sucking or pushing stuff out. Are you bleeding all rads with all pumps off? Also, assume the ball valve is working correctly now so fully replenishing the system as you remove the air?

If airlock is the problem you could try switching all rads except 1 on and run the pump then try bleeding it. Continuing one by one. Having said that, sounds like when you put pump on max it appears to be strong enough.

When the pump is on low, do all rads heat up and are there cold spots on them? Top cold implies air so bleed , bottom cold implies crud


I'm satisfied that the ball valve is working ok. You may have missed the
point that there are two pumps - one for the HW and one for the CH. When
I first freed the ball valve and replenished the top-up water, I did run
both parts of the system, and the radiators did get hot. I bled them,
and very little air came out.

More recently, I've only run the HW side, so I wouldn't expect the
radiators to get hot. Bearing in mind the fact that the F&E tank is
connected to the HW circuit - either side of the coil - this is the main
candidate for air. When the HW pump is on low, the pipe going to the
cylinder coil doesn't get hot - indicating no circulation. When the pump
is turned up, its gets hot - but then the gurgling and banging starts!
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

On 13/05/2018 10:06, Tim Lamb wrote:


The need for relay controls may imply that the valve does not have a
microswitch. I am currently trying something similar with a diverter
valve and plan to report back. One initial thought is that the relay
will be instant while the associated valve may take several seconds to
fully move. Possibly leading to pump flow going somewhere unexpected
particularly if the valve fails!
My circuit is pressurised.



The relay is there because there are two pumps, one for each circuit -
and the boiler needs to run when either or both pumps are on.

The valve *does* have a micro-switch. The sequence is that the manual HW
switch (there is no timer) powers the tank stat which opens the valve if
the water temperature is below the stat setting. When the valve is open,
its micro-switch switches on the HW pump. *Then* the relay logic
switches on the boiler if it's not already running for the CH. So the
pump can't run with the valve shut.
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

On 13/05/2018 09:35, alan_m wrote:
On 12/05/2018 21:37, Roger Mills wrote:

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?

Are all the radiators getting hot? If not you may have a airlock in the
pipes supplying them which may be the source of the some air getting
back to the to the pump or boiler. If so run the pump at full speed and
just turn on one radiator at a time to try and force the water through
the problem pipe length.

Since the tank ran dry have the radiators been bled? While the tank was
dry air could have been sucked in and distributed to the radiators.


Don't forget that there are two circuits, each with its own pump.
Currently, I'm only running the HW circuit - so the rads are cold. But I
*did* run the CH and bled the rads (very little air came out) when I
first became involved.

I omitted to say that there is a towel rail in the bathroom which is on
the HW circuit, connected in parallel with the coil. *That* does collect
air and needs frequent bleeding.
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

On Sunday, 13 May 2018 14:47:27 UTC+1, Roger Mills wrote:
On 13/05/2018 09:35, alan_m wrote:
On 12/05/2018 21:37, Roger Mills wrote:

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?

Are all the radiators getting hot? If not you may have a airlock in the
pipes supplying them which may be the source of the some air getting
back to the to the pump or boiler. If so run the pump at full speed and
just turn on one radiator at a time to try and force the water through
the problem pipe length.

Since the tank ran dry have the radiators been bled? While the tank was
dry air could have been sucked in and distributed to the radiators.


Don't forget that there are two circuits, each with its own pump.
Currently, I'm only running the HW circuit - so the rads are cold. But I
*did* run the CH and bled the rads (very little air came out) when I
first became involved.

I omitted to say that there is a towel rail in the bathroom which is on
the HW circuit, connected in parallel with the coil. *That* does collect
air and needs frequent bleeding.
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

So just to be sure.... With the current setup, the heating is off and therefore the pump is off. When the hot water pump is off all is quiet and when the hot water pump is on you get all the noises?

You mentioned that the F&E tank is connected to the hot water circuit but assume it must also feed the heating one too? If not, do you have a separate F&E tank for the heating circuit? Finally, are the 2 circuits isolated from each other e.g. via a motorised valve? Wondering whether the water circuit could be pulling air from the heating circuit.

On 13/05/2018 15:43, wrote:


So just to be sure.... With the current setup, the heating is off and therefore the pump is off.

Correct

When the hot water pump is off all is quiet and when the hot water pump is on you get all the noises?

Correct


You mentioned that the F&E tank is connected to the hot water circuit but assume it must also feed the heating one too?

Yes, both circuits meet at the boiler.


If not, do you have a separate F&E tank for the heating circuit?

Not applicable - no.

Finally, are the 2 circuits isolated from each other e.g. via a motorised valve? Wondering whether the water circuit could be pulling air from the heating circuit.

There's only one motorised valve, and that came with the unvented
cylinder, and is in the HW circuit. Prior to that - when there was a
vented cylinder - there were no motorised valves at all. I can't see any
obvious reason why just one pump running could cause circulation in the
other circuit when its pump isn't running - BICBW!
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

Roger Mills formulated on Sunday :
The valve *does* have a micro-switch. The sequence is that the manual HW
switch (there is no timer) powers the tank stat which opens the valve if the
water temperature is below the stat setting. When the valve is open, its
micro-switch switches on the HW pump. *Then* the relay logic switches on the
boiler if it's not already running for the CH. So the pump can't run with the
valve shut.

Any chance that though the valve actuator might have moved to the valve
open position, that the valve might actually not have opened, or not
opened fully?

Obviously there is some obstruction preventing or severally limiting
the flow.

On 13/05/2018 17:58, Harry Bloomfield wrote:
Roger Mills formulated on Sunday :
The valve *does* have a micro-switch. The sequence is that the manual
HW switch (there is no timer) powers the tank stat which opens the
valve if the water temperature is below the stat setting. When the
valve is open, its micro-switch switches on the HW pump. *Then* the
relay logic switches on the boiler if it's not already running for the
CH. So the pump can't run with the valve shut.

Any chance that though the valve actuator might have moved to the valve
open position, that the valve might actually not have opened, or not
opened fully?

All things are possible(!) but I don't think that's likely. I've taken
the actuator off the valve a couple of times, and the valve moves quite
easily, and the mating parts are fairly substantial - so I don't think
the actuator can move far enough to operate its micro-switch *without*
opening the valve. Of course, the innards could still be blocked by
crud, I suppose.


Obviously there is some obstruction preventing or severally limiting the
flow.

I'm not totally convinced. True, there's no flow at low pump speeds -
but I put that down to trapped air somewhere. There seems to be quite a
lot of flow when the pump is given a bit more urge - but we then seem to
be drawing air in. I'm assuming that the horrible noises are caused by
air going through the boiler, allowing localised boiling to occur.

When airlocks occur in gravity (secondary) hot water systems, the cure
sometimes seems to be to force mains water back up the pipe from a hot
tap to the cylinder and header. Is there anything similar I can do with
a vented primary system?

Please keep the suggestions coming. Hopefully they will lead to a Eureka
moment, eventually!
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

[I'm purposely top posting to get to the meat of the issue quickly,
whilst leaving the original post below for anyone who wants to refer
back to it.]

I was round at my friend's today having another play with the system.
[Remember, this is a vented primary system with a single F&E tank in the
attic, and two separate pumps - one for HW and one for CH.]

I ran the HW and CH separately, and exactly the same thing happened in
both cases, namely:

With the pump running at low speed, the boiler cycled on its own stat
but the pipes didn't get hot. With the pump turned up, the pipes got hot
and the system appeared to work ok for 20 minutes or so such that, in
the case of the CH, the rads were getting reasonably hot.

But then, all of a sudden, all hell was let loose - with banging and
crashing noises coming from the pipework. When I looked in the attic,
the water in the F&E tank was very hot, the water level had risen
sufficiently to reach the overflow, and enough steam to power
Stephenson's Rocket was coming out of the vent pipe.

I no longer believe that air is being sucked in, but rather that the
boiler is kettling and generating steam which increases the system
volume and raises the F&E tank level. [The level soon goes down when
everything is turned off].

The fact that the pumps need turning up more than usual to get any flow,
and that the boiler is kettling, suggest to me that there is a partial
blockage and perhaps scaling up in the boiler's heat exchanger.

Does that sound like a reasonable explanation? If so, what's the best
way of curing it? I'm thinking along the lines of Sentinel X800 prior to
a manual flush and re-fill with fresh inhibitor. Any comments?

--
Cheers,
Roger




On 12/05/2018 21:37, Roger Mills wrote:
I guess that many of us have come across primary vented systems which
pump *over* - where there is a constant flow of water out of the vent
pipe back into the F&E tank.

But is the opposite also possible - where air is sucked in through the
vent pipe, and water forced back up the fill pipe into the F&E tank?

I have recently been tearing my hair out - what's left of it(!) - trying
to sort out a friend's heating and hot water system.

The system has a floor-standing conventional boiler (Potterton
Kingfisher Mf - c.2003) in the garage, with a vented primary circuit
which has a fill & expansion tank in the attic. The hot water system is
unvented, employing a Heatrae Sadia Megaflow indirect cylinder (c. 1995)
in the first floor airing cupboard.

The CH and HW primary circuits each have their own pump and are entirely
separate - only meeting at the boiler connections. [There is some relay
logic which runs the boiler when one or other or both circuits are
calling for heat]. The only motorised valve is part of the Megaflow
installation. This opens in response to the tank stat, and its auxiliary
contacts turn the pump on.

The basic system was designed and installed by my friend several decades
ok, but has had a new boiler and pumps, plus replacing the vented HW
cylinder by an unvented one in the meantime. The current configuration
has remained unchanged since about 2003 - and has worked ok until recently.

Problem is that my friend is now in a care home, and his wife sends for
me when things go wrong!

The initial problem reported to me was lots of gurgling and banging in
the pipes when the HW was being heated. When I investigated, I found
that the F&E tank in the attic was completely empty - the ball valve had
stuck shut, and all the water had evaporated. That would explain the
symptoms, I thought - with air entering the system because of the lack
of top-up water. So I freed up the ball valve, allowed the system to
re-fill and bled air out from all available places.

When I turned the system on, the boiler would run for a couple of
minutes and then stop. The HW pump was running, but there didn't seem to
be any flow - in that the flow pipe from the boiler wasn't getting hot.
At this point, the pump speed was on its minimum setting - which is how
I had found it. When I turned it up to max, the boiler fired for much
longer, and the pipes got hot. I could still hear air going through the
system though.

There is obviously still a problem. My feeling is that there is still
air in the system, which I haven't succeeded in removing. With the pump
on its lowest setting, there is no circulation - even though there used
to be - and the water obviously doesn't get hot. So presumably, there is
an air lock.

If I turn the pump speed up enough to get the water to flow, we're back
to the dreadful gurgling and banging - all the way up to the F&E tank.
On one occasion, the water in the F&E tank became very hot, and the
level rose to the point where it started to overflow. [I began to wonder
whether there was an internal leak in the coil in the HW cylinder, but
the level went down again fairly quickly once I had turned everything
off - so I'm pretty sure that's not the case].

My feeling is that, when the pump is run fast enough to make the water
circulate despite an air lock, air is actually being drawn into the
system through the vent pipe - what might be described as pumping
*under*, as per the title. Anyone come across this? In case it helps to
visualise the system, I've drawn a somewhat crude diagram of just the HW
circuit, which can be seen at
https://app.box.com/s/43v06520ifm7s3iw2usqj0vwp83o7nwz As can be seen,
the fill and vent pipes connect to that circuit in the vicinity of the
HW cylinder. Note: The pump is on the return side, close to the boiler -
pumping *towards* the boiler.

Any suggestions as to how I can get all the air out, and get the system
working as it did before the F&E tank dried up?

I'm almost tempted to suggest converting the primary circuit to
unvented, which should make it easier to eliminate the air. But it's
worked ok as a vented system for decades - so shouldn't be necessary.

Am I missing anything obvious?

--
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

Mmmm does sound a bit odd. It could be that I guess but generating all that steam through settling does sound a bit extreme. Wonder if the boiler is over heating the water somehow. I guess if there is a blockage maybe the water is moving too slow across the heat exchanger?