Repeated Air in Central Heating System - Help appreciated

[Vessy_1]

Hi, I'm hoping for some suggestions as to where I go next in my endless mission to stop my system taking on air.

The house has a gravity fed system, 'S' plan, Heat only boiler supplying CH & HW, with separate F&E and vent just before the pump in the typical H layout, pump & valves are all located in the upstairs Airing cupboard.

Boiler and pump have just been replaced as the old boiler was proving unreliable and underpowered for the house (4 Bed, 11 Rads) new Ideal Logic + 24kw installed with Grundfos UPS3 pump. This issue is not new, I have battled with air previous to the boiler and pump replacement and I was hoping this would be solved following the clean, flush and a fresh dose of inhibitor that was done during the work.

So, the problem, once I have all the air out of the system it will run perfect indefinitely, as long as the water is cold!! the air only starts to circulate once the return temperature is in the 50's, as the return temperature rises the air gets worse to the point that the pump will eventually cavitate and stop moving water around the system, this happens on CH, CH+HW and HW only, the air is however much worse on HW only as the return temp is a lot higher especially towards the end as the cylinder approaches set temp (60 Deg). Again on CH only it only slightly makes a small amount of air as the return is around 50 Deg.

I added some further x100 as I had 50% emptied a rad a few times (air gets trapped in the rad return line) and this has made no difference to the issue, I have tried the pump on all speed settings and this makes no difference, I tried the glass of water on the vent line and all seemed good (water dropped about an inch when the pump turned on and came back up within a few seconds) header tank level slightly drops on pump turn on so looks OK.

Checked all the accessible unions on the valves, Pump, HW Balance valve and looked over everything, no evidence of leaks. I even checked the return line at the boiler with the heating on and definitely had positive pressure as the filter could still be bled (Mag filter above boiler on return line) so no sucking in of air.

I am all out of ideas, other than corrosion somewhere in the return line, is that possible and will it make air so quickly? After I have bled the system the pump will cavitate after 5 mins once the return is into the 55-60 deg region.

Currently I can keep the system going but reducing the boiler output temp to around 62, this keeps the return on CH well below 50, I start to build air when HW only on demanded but not enough to cavitate the pump, only downside is the HW takes an age to recover and the boiler short cycles as its minimum output is too much for the small temperature difference.

Any ideas or suggestions would really help, as I mentioned above I can leave the system running for hours with the boiler temp set to minimum so cannot get my head round this being a leak.

 

[steadyon]

1. How did you bleed the air from the hot water cylinder?
2. Photographs if possible might help diagnosis, as follows:
2a. Vent over F&E showing height of inverted "J".
2b. Hot water cylinder showing inlet and outlet for coil, together with any bleed valve(s) fitted.
2c. Pump with indication of pumping direction.
2d. Motorised valves (least likely to be useful, so if difficult, don't bother).

 

[Vessy_1]

Hi Steadyon

I will get some pictures soon, I can provide some answers your questions,

1. With the boiler off I ran the pump with HW demand only, opened the balance valve fully, not much air was in there, after a min or so there was noise of water moving but no air noise, As I get this issue when on heating only (once the return temps are up) does this not rule out air in the cylinder?
2a I am getting no pump over into the tank, it is about 500mm above the top of the water level, cannot go higher due to low roof rafters, I do however get a very small dribble when the pump turn off.
2c Pump is fitted the correct way, arrow towards the two valves, away from the F&E pipes.

 

[Vessy_1]

Pictures As requested.

 

[Vessy_1]

Sorry, system is open vented, not gravity as I put in the first post!!

 

[steadyon]

1. Many people call their systems gravity when the are open vented. Not an issue.
2. There appears to be no means of bleeding air from the hot water cylinder coil. You could try running the system (HW and CH) up to a reasonable temperature, then turning the boiler off. Wait 15 minutes or so, then manually open the HW motorised valve, and slacken the compression nut nearest the pump until all air is out and water starts to appear. Then re-tighten the compression nut. Old towels advised.
3. The vent pipe should rise up then fall over the F&E. The volume in the rise of the inverted "J" should normally be enough to accommodate system water expansion. From memory it should be able to accommodate 4% of total system volume, and should rise a minimum 450 mm. Would you have scope to:
3a. Raise an inverted "U" between the rafters, dropping the existing vent pipe so that it only has a slight rise from the boiler end giving more room for a rise between the rafters?
3b. Rerun the vent up towards the apex of the roof to achieve a vertical height of 450mm, then drop back to over the F&E tank?

 

[John.g]

Even though you have a "accepted" system with boiler, then vent, then cold feed, then pump, (VCP), I strongly feel that the combined vent and cold feed like mw own almost 50 year old system is a far better arrangement as its in a "static" part of the system, see below. If your vent passes up close to the F&E tank or you can modify it to do so, then cut and blank the H at cold feed tee in and tee in the cold feed directly like mine then I would think that would solve your problem.

The UPS3 looks as if its on CP1 mode/setting, is that correct??

Edit: It is also acceptable (but NOT with solid fuel) to have no vent as long as there is no isolating valve on the cold feed, I personally don't like the idea and wouldn't dream of it except there is a a 3 bar PRV fitted at the boiler. I see there is a compression fitting on the end of your vent, have you been blanking this off?.

 

[John.g]

1. Many people call their systems gravity when the are open vented. Not an issue.
2. There appears to be no means of bleeding air from the hot water cylinder coil. You could try running the system (HW and CH) up to a reasonable temperature, then turning the boiler off. Wait 15 minutes or so, then manually open the HW motorised valve, and slacken the compression nut nearest the pump until all air is out and water starts to appear. Then re-tighten the compression nut. Old towels advised.
3. The vent pipe should rise up then fall over the F&E. The volume in the rise of the inverted "J" should normally be enough to accommodate system water expansion. From memory it should be able to accommodate 4% of total system volume, and should rise a minimum 450 mm. Would you have scope to:
3a. Raise an inverted "U" between the rafters, dropping the existing vent pipe so that it only has a slight rise from the boiler end giving more room for a rise between the rafters?
3b. Rerun the vent up towards the apex of the roof to achieve a vertical height of 450mm, then drop back to over the F&E tank?

I have often wondered about this 4%, I know that water expands by 4% at 100C, but stlll can't figure out its relevance as the cold feed takes the majority of the expansion volume and the level in both the vent and the F&E tank will only rise by a few mm.(u tube effect), Maybe the correct description is vent + cold feed and expansion.

 

[Vessy_1]

Even though you have a "accepted" system with boiler, then vent, then cold feed, then pump, (VCP), I strongly feel that the combined vent and cold feed like mw own almost 50 year old system is a far better arrangement as its in a "static" part of the system, see below. If your vent passes up close to the F&E tank or you can modify it to do so, then cut and blank the H at cold feed tee in and tee in the cold feed directly like mine then I would think that would solve your problem.

The UPS3 looks as if its on CP1 mode/setting, is that correct??

Edit: It is also acceptable (but NOT with solid fuel) to have no vent as long as there is no isolating valve on the cold feed, I personally don't like the idea and wouldn't dream of it except there is a a 3 bar PRV fitted at the boiler. I see there is a compression fitting on the end of your vent, have you been blanking this off?.

Hi John,

I’m up for doing anything proactive to fix this, but with no pumping over and no suction of air on pump start would this be of benefit? The air only builds in the system after everything is hot and the return line from the rads or cylinder is 55 deg +. This is why it only seems apparent when HW only is being demanded and the cylinder is almost to temp, it really has me baffled.

Regarding the pump setting, yes at that time it was on CP1, my logic for this was that once I had balanced all the rads for a 12 deg drop any upstairs rads shutting down on the TRV’s shouldn’t change the drops on the downstairs rads as the pump would reduce and maintain the same pressure in the system, similarly if HW and CH are both being demanded the pump will speed up to keep pressure so the rads should still have a 12 deg drop, is this setting not right?

I have today been running the system on fixed speed 3, although maybe slightly better it is still producing air when the return is hot.

Could I still have trapped air somewhere, maybe a capped off pipe somewhere that filled with air when the system was drained, then as the system temp increases the air expands and enters the flow/return pipes, feasible?

Just would have thought after 2 weeks the air would all be gone!

Do these UPS3 pumps cavitate easily?

 

[John.g]

There are no guarantees with any OV system, most problems seem to arise in systems plumbed like yours where the pump is mounted high up on the flow from the boiler and very near the motorized valves (all for convenience). Mine (oil fired boiler) has the pump located in the boiler heatpac downstairs, on the boiler return.
I can run it at any pump heand or temperature I like with absolutely no problems.

You say that a increase in return temp > than say 50/55C triggers the air ingress problem.
Is the boiler up to its target flow temperature before 50/55C is reached, if not then its the flow temperature (pump inlet) which is probably triggering this, difficult to see how the increase in return temp without a increase in flow temperature can trigger a problem on the flow side at a steady temperature, the flow temperature will increase by 5 to 7C each time the boiler cycles as the burner cuts out at target temperature+5/7C, again don't know if your problems start with steady target flow temp or only when the boiler starts cycling.
Also don't like the thought of that dribble from the vent, I would think that each time this happens that you are drawing in some air each time. Of course, if the problem arises under steady, continuous pump running then another reason for the problem.
Can you ensure that the circ pump keeps running during all boiler cycling as it should only stop when both CH and/or HW is satisfied.
Also check that the pump start/stop is from the motorized valve auxiliary contacts (end switches)

You might just try (again?) running o n PP2 mode as it should run at ~ 0.5M lower head and still give sufficient flow with maybe no dribble on pump stop.

Re UPS 3 cavitation, havn't seen the NPSHr (Net positive suction head required) for this model but I would think that ~ 1.5M (like the selectric) would be OK,, the NPSHa (Net positive suction head available) however is/should be way above this as :
NPSHa is the Atmospheric pressure (absolute) + the static head - the liquid vapour pressure - the suction losses.
vapour pressure at 70C=3.1M, 60C=1.9M, & 55C=1.2M. so assuming suction loss at 0.2M (between cold feed and pump suction) and static head at 2M (use your own numbers for static head and suction losses) then the
NPSHa at 70C = 10+2-3.1-0.2, 8.7M and 10+2-1.2-0.2, 10.6M at 55C, so unless some very big problem with the suction loss at the pump (or my calcs are crap) then can't see how flow temperature itself should cause any cavitation problems so possibly points to air ingress/release at higher temps, there was a long running thread on here abut someone with problems like yours but with the circ pump located low down on the system (flow) and again couldn't be run at anything > say 65C due to noise, the system was then sealed but problems still persisted, unfortunately, the OP never told us if he ever fixed his problem.

Edit: Just saw that you had/have been running on fixed speed3? hope not for long as this is a disaster IMO with your type of system, CP1 at 3M should be a very good setting.

 

[Vessy_1]

There are no guarantees with any OV system, most problems seem to arise in systems plumbed like yours where the pump is mounted high up on the flow from the boiler and very near the motorized valves (all for convenience). Mine (oil fired boiler) has the pump located in the boiler heatpac downstairs, on the boiler return.
I can run it at any pump heand or temperature I like with absolutely no problems.

You say that a increase in return temp > than say 50/55C triggers the air ingress problem.
Is the boiler up to its target flow temperature before 50/55C is reached, if not then its the flow temperature (pump inlet) which is probably triggering this, difficult to see how the increase in return temp without a increase in flow temperature can trigger a problem on the flow side at a steady temperature, the flow temperature will increase by 5 to 7C each time the boiler cycles as the burner cuts out at target temperature+5/7C, again don't know if your problems start with steady target flow temp or only when the boiler starts cycling.
Also don't like the thought of that dribble from the vent, I would think that each time this happens that you are drawing in some air each time. Of course, if the problem arises under steady, continuous pump running then another reason for the problem.
Can you ensure that the circ pump keeps running during all boiler cycling as it should only stop when both CH and/or HW is satisfied.
Also check that the pump start/stop is from the motorized valve auxiliary contacts (end switches)

You might just try (again?) running o n PP2 mode as it should run at ~ 0.5M lower head and still give sufficient flow with maybe no dribble on pump stop.

Re UPS 3 cavitation, havn't seen the NPSHr (Net positive suction head required) for this model but I would think that ~ 1.5M (like the selectric) would be OK,, the NPSHa (Net positive suction head available) however is/should be way above this as :
NPSHa is the Atmospheric pressure (absolute) + the static head - the liquid vapour pressure - the suction losses.
vapour pressure at 70C=3.1M, 60C=1.9M, & 55C=1.2M. so assuming suction loss at 0.2M (between cold feed and pump suction) and static head at 2M (use your own numbers for static head and suction losses) then the
NPSHa at 70C = 10+2-3.1-0.2, 8.7M and 10+2-1.2-0.2, 10.6M at 55C, so unless some very big problem with the suction loss at the pump (or my calcs are crap) then can't see how flow temperature itself should cause any cavitation problems so possibly points to air ingress/release at higher temps, there was a long running thread on here abut someone with problems like yours but with the circ pump located low down on the system (flow) and again couldn't be run at anything > say 65C due to noise, the system was then sealed but problems still persisted, unfortunately, the OP never told us if he ever fixed his problem.

Edit: Just saw that you had/have been running on fixed speed3? hope not for long as this is a disaster IMO with your type of system, CP1 at 3M should be a very good setting.

Morning,

I’m only running fixed 3 to see if it pushed / drew any air out of the system. PP2 mode from memory did not give me enough flow to the furthest rad in our living room when most of the trv’s had shut upstairs. I can get 12 deg drop across the furthest rad (with all trv’s removed) easily on constant speed 2 (haven’t tried constant 1) and cp1.

So, yes the boiler is already at 70 deg when this happens, I’ve been monitoring temps and it is purely driven by return temp, HW only with a cold cylinder the boiler is running with 70deg flow and has modulated down to minimum output, I adjusted the balance valve to give a 15 deg drop (measured at they cylinder) after a few mins the return temp starts to slowly rise as the water is heating, after roughly 20 mins the return temp is now into the 60’s and air is building, you can hear it going through the pump, not big glugs but lots of little bubbles. With further heating it gets worse and starts to whoosh round the system until causing the pump to cavitate.

Balance valve on the HW return is open 3/4 of a turn, this gives the 15 deg drop and doesn’t impact the CH system too bad when both are on.

Pump is controlled by valve aux contacts, same wire runs to the boiler for demand, yes the pump runs when the boiler cycles.

The boiler starts to short cycle right at the end of the HW heating, on min output the difference between flow and return is only 6-7 deg, it does all seem to happen at the same time but I think the cycling is normal?

 

[John.g]

Strange alright re return temp, it looks as if its releasing air on heating up as it "must" be air that's entering the pump suction, presume its not affecting something within the boiler Hx?.

When it starts cavitating, if you adjust the dT to bring the return back down to "normal" does the pump stop cavitating without you having to release any air?.

CC2 is still quite high IMO at 5.2M, suggest CC1 @ 4.2M which may run at ~ 3.8ish or CP1 which would be great if it satisfies your flow rates @ 3M. and may stop that dribbling.

Failing this suggest plugging the vent, there is already a compression fitting on it and lock open (tie wrap) any isolating valve(s) on the cold feed and see how it reacts.

 

[Vessy_1]

Hi John, firstly thank you for taking the time to reply.

So, todays fun and games, I tried the pump on multiple settings, with the loft hatch open and tank lid off I can confirm that I am getting some pump over, not a vast flow but more of a constant trickle, and the circumstances effect this. Let me try to explain what i've done today.

With all rads on (TRV's removed) the system is nicely balanced, results are as follows

CP1, 12 dT on each rad, 24 Deg dT at the boiler whilst the system was warming up (max output)

CP2, 10 dT on each rad 19 Deg dT at the boiler, again whilst warming up (max output)

Furthest rad has the lockshield fully open.

With CH and HW demanding 1 turn open on the cylinder balance valve starts the pump over. With a cold cylinder I need the balance valve open 3/4 of a turn to get a 15 deg dT so this stops the pump over.

Running on CP1 I get no pump over, even with the balance valve fully open, I'm pretty sure I ran CP2 with just the HW and again didn't get pump over (guess lower flow?)

So now the system is again running CP1, I started to get air when on HW only but it is more like aerated water, not big glugs and sometimes a whooshing noise, should the air not exit up the vent pipe? or is it usual for it to get dragged round the system.

I let the HW and CH run together but just had the nearest two rads fully open (both valves) I could hear little pin pricks of air entering the rads, the pump never cavitated and the air noise eventually almost disappeared.

Could this be purely oxygen rich water entering the system? Annoyingly the air in my system collects in the pipes that run to a vertical rad in a hallway, the pipes are tapped off the upstairs flow and return and run in the ceiling and are plastic, so probably a high point, getting the air out involves draining half the rad out, then open one valve so the inrush of water brings the air with it, not ideal but the only way. My thoughts are that this process is introducing more fresh water in every time, viscous circle!

 

[Vessy_1]

Need to add something to the post above, might have found something!

Whilst continuing to find the cause I have noticed that when on CH only I could hear a noise from the HW valve, not 100% sure if it is air being drawn in or water being pulled past, however if I fully close the cylinder balance valve the noise stops, start to open the balance valve and the noise starts again!

9 out of 10 times when the HW is demanded on a resting system I can hear a big glug of air as the pump turns on, could the HW valve be letting in air that is being trapped in the coil? I would have thought it would always be under a positive pressure but with it all being upstairs (close to the feed tank) and the pump circulating the rads could it be a negative pressure on the coil side of the HW valve?

System is running now CH only, balance valve is shut tight, boiler flow is 68, return is 52, each rad is showing around 12 dT. No air at all moving round the system.

I'm going to demand HW and open the balance valve once the Honeywell valve is open and see if I get any air, thus eliminating any negative pressure, will report back.

 

[John.g]

Don't think any motorized valve will/should pass when closed
The dimples look correct but check arrows on zone valves for correct orientation.

 

[Vessy_1]

Valve is fitted correctly, although looking at an internal image of one the ball blocks the inlet pipe when off, so if the shaft was letting in air it would be pulled in if there was a negative pressure.

Do you think it could be under negative pressure? Could this really be the issue?

Is it feasible to test this by running the CH, opening the balance valve and loosening the lower union on the valve to see if water comes out? (with HW valve off of course)

So, the system has been running now for over an hour, CH only with 68 flow and 54 return, no air in the system and quiet. I opened the balance valve 1 turn after giving it a HW demand and left this running for 10 mins, return temp came up to 63 as the cylinder is hot, again no air at all.

As I now have all the air out of the system and it is not making air I also had no spurts of water from the vent pipe on shutdown.

Has anyone ever had a valve let in air before from return suction?

 

[John.g]

It's certainly worth trying the above even though with system at rest one might have expected to see some sign of a leaking shaft seal but then again maybe not as head pressure so low.

 

[Vessy_1]

It's certainly worth trying the above even though with system at rest one might have expected to see some sign of a leaking shaft seal but then again maybe not as head pressure so low.

Yeah I would think the same too but I’ve seen other posts about rad valves passing air but not water, got to be worth a shot.

The head is just shy of 2m, so pretty low pressure.

Going to use the immersion for a few days with the balance valve hard closed and see how the CH performs

 

[John.g]

Very good, what is the Pump setting?.

 

[Vessy_1]

Very good, what is the Pump setting?.

CP1 presently, constant 1 seems similar but doesn’t perform as well when all trv’s are open, PP just didn’t deliver to the furthest rads when upstairs was all shutoff on trv’s. According to the book CP is meant for underfloor but it works for me!

 

[John.g]

CP1 on that pump is a excellent mode on that pump as it maintains a 3M head between 0 LPM and 48LPM, CC1 rises to 4.2M at 0 LPM and the head falls below 3M from a flow rate of 23LPM.

 

[Limecc]

I had this problem and it was caused by a leaking coil in the hot water cylinder leading to fresh water constantly entering the radiator circuit. Fresh water is aerated and was released when it heated up. Changing the cylinder and fitting a vortex type air separator (don't know the proper name) fixed the problem.

I think I diagnosed it by tying up the ball cock and emptying the cylinder via a hot tap. Water never actually stopped, now entering via the system circuit header tank.

 

[John.g]

A very good point. My F&E tank is higher than the CWST and I keep the mains supply to it (F&E) shut permanently via a 1/2 ins gate valve, I can't look into the tank but open the isol valve a few times/year so will hear any make up via the ball cock. Ideally the water level in the CWST should be higher than the F&E tank to avoid contamination if the coil fails, if so then both levels will equalize or the F&E might start overflowing so worth checking out in the OP's case but easy to overlook.

 

[Limecc]

Ideally the water level in the CWST should be higher than the F&E tank to avoid contamination if the coil fails

I'm in a bungalow so have them both at the highest possible point, but I think it's a moot point as there was contamination every time somebody ran a bath or shower. With all that air and no inhibitor I must have had some serious rad corrosion going on.

I fitted an Adey Magnaclean filter before the boiler at the coolest point and it's pulled out an incredible amount of crud. The back boiler no longer pops and bangs, I'd attributed that to limescale. Must get rid of the latter, will start a separate thread.

I placed the air separator/vent pipe just before the pump suction, the pump is fitted horizontally tilted about 30 deg. System is more or less self bleeding since.

 

[Vessy_1]

I had this problem and it was caused by a leaking coil in the hot water cylinder leading to fresh water constantly entering the radiator circuit. Fresh water is aerated and was released when it heated up. Changing the cylinder and fitting a vortex type air separator (don't know the proper name) fixed the problem.

I think I diagnosed it by tying up the ball cock and emptying the cylinder via a hot tap. Water never actually stopped, now entering via the system circuit header tank.

Hi, I’ve already considered this and ruled it out, the CWST level is probably 2ft higher than the F&E level, I have had no increase in height of the water level in the F&E. Heating been on all day today and I’ve had very minimal air, putting that down to the fresh water taken in yesterday to bleed the air out, time will tell!

 

[John.g]

I'm in a bungalow so have them both at the highest possible point, but I think it's a moot point as there was contamination every time somebody ran a bath or shower. With all that air and no inhibitor I must have had some serious rad corrosion going on.

I fitted an Adey Magnaclean filter before the boiler at the coolest point and it's pulled out an incredible amount of crud. The back boiler no longer pops and bangs, I'd attributed that to limescale. Must get rid of the latter, will start a separate thread.

I placed the air separator/vent pipe just before the pump suction, the pump is fitted horizontally tilted about 30 deg. System is more or less self bleeding since.

A lot of problem systems have the pump pumping downwards which doesn't help, also helps a lot if the pump head can be limited to 3M or less. My OV system with a combined cold feed and vent does have the pump pumping downwards but its installed downstairs and outside on the boiler return. I run the pump on a PP setting of ~ 4.3M which results in a effective head of ~ 3M and less as the TRVs throttle in.