Central Heating - Expert Assistance Required - Revisited

Hi all

Thanks to all respondents to original thread. Have shifted emphasis here so new thread in order.

My concern is pump over! As roughly outlined in my previous thread, the solenoid valve for heating is half way between ground and first floor, with upstairs feed to all rads in

15mm pipe from this point. The feed to the cylinder and vent is 22mm with the solenoid close to the cylinder on the first floor. On advice from previous thread, I tried to "force" flow to 2 upstairs rads by turning all downstairs rads off (this is a likely scenario if all trvs close doownstairs). I didn't get as far as checking these 2 rads because water started to pump over into the header tank immediately, even with the upstairs bathroom rad fully open (both lockshield and control valves).

My dilemma is that it appears pump over is inevitable with this piping. With downstairs rads off, the pump runs and is faced with 2 routes, either

22mm up passed cylinder and over to header tank, or 15mm with fittings and rad valve resistance. Surely, as installed, this arrangement will always pump over rather than feeding the upstairs rads?

I have two theories:

  1. The systems were never designed to have trvs, therefore some down stairs flow would always avoid the pump over.
  2. There is a problem with the heating solenoid valve (although it seems to work OK) and it is presenting some restriction. So add the further restriction of only 15mm pipe available and pump over occurs.

Anyone shed any further light on this please?

TIA

Phil

Reply to
TheScullster
Loading thread data ...

Yes, switch to sealed pressurised operation, if your boiler allows it. Much quicker and simpler than diagnosing a pump over problem.

Christian.

Reply to
Christian McArdle

Is it possible to extend the expansion pipe so it goes higher? Possible to site the entire header tank higher in the roof void?

Reply to
Dave Plowman (News)

What speed was the pump running at?

What is the height difference between the rads and the apex of the vent pipe? That (converted into an equivilent pressure) will tell you the maximum flow resistance that can be accomodated in the upstairs rads without pumping over.

Are you sure you don't have a restriction or blockage of some sort in the upstairs rad loop?

Do you have a bypass radiator or a bypass valve?

If it is all TRVs and there is no bypass, then there is always the possibility of pump over unless you also include a flow switch as an interlock.

Reply to
John Rumm

You can't get pump over in my system. The vent and filler pipe are combined. It takes a little longer to fill but thats about the only problem.

Reply to
dennis

In order to get pump-over, you need a difference in dynamic pressure between the vent pipe and fill pipe. This can happen if they are connected on opposite sides of the pump, or even on the same side - but at two distant points on a pipe carrying a lot of flow. If at all possible, they should both be connected on the suction side of the pump, and not more than 100mm apart. Mine actually have a *single* 22mm connection into the flow pipe - which then splits into a 22mm vent pipe and 15mm fill pipe.

Have a close look at how your pipes are connected, and see whether you can move one of them to prevent this problem. Take care to ensure that both pipes have a clear path to the boiler which cannot be interrupted by closing any motorised (or solenoid) valves.

Reply to
Set Square

Meant to include that - it's at speed 2 of 3.

Approx 3.5m above top of rads (not towel rail)

Only the pipe size of 15mm mains rather than 22mm I think. Original radiators heat up in just-about-acceptable time.

As stated, the bathroom rad valves were both fully open on the first floor.

Thanks for the help John

Phil

Reply to
TheScullster

The main issue with combined feed/expansion is that it requires the same safety features in the boiler as sealed pressurised operation.

Personally, if the boiler did have the safety features, then I'd go sealed instead of combined gravity feed, as the benefits are great.

Christian.

Reply to
Christian McArdle

So that is about 1/3rd of a bar then...

What happens if you close off the downstairs rads and set the pump to speed 1? Do the upstairs rads heat then or do you still get pump over?

It seems that with the downstairs rads closed, and a differential pressure such as you have, you ought to be able to get decent circulation through the upstairs rads. If even in that circumstance you don't get decent flow through them that suggests some form of blockage to me, rather than just a flow restriction.

I was asking for a different reason ;-) More a case of without a bypass of some form and TRVs everywhere there would come a point (when the house is up to temperature) that the only place the pump can send water is up the vent!

Reply to
John Rumm

Yes, Grundfos Alpha and sealed circuit. You know it makes sense.

Christian.

Reply to
Christian McArdle

Yes

Didn't try that. What would this show if it worked? The system is pretty slow to heat up as is without lowering the pump speed. I will try this tonight and report back.

Still not clear why the water would want to take the tortuous route through

15mm pipe/rads rather than the clear 22mm route via the vent pipe.

The heating solenoid valve did start making noises some time ago and has since gone quiet. But as this controls water flow to both upstairs and downstairs I'm not clear on whether this is likely to be the cause. Could this be causing a partial restriction which, when added to the 15mm pipe reduction, might cause the pump over?

Other than the bathroom, there is also a rad piped direct to the flow and return of the boiler before either solenoid valve. Opening this last night did not stop the pump over but may have reduced it. valves were both fully open on the first floor.

Do you not think that the bathroom rad should perform this function?

Thanks again for your time with this

Phil

Reply to
TheScullster

The pump speed makes little difference to this. All it needs to do is to provide sufficient flow to get the correct temperature differential between flow and return. Speeding up the flow beyond this only results in more noise.

Reply to
Dave Plowman (News)

Ought to be plenty... (the vented system I used to have ran fine on 2m of head)

It would show that the upstairs rad loop is not blocked. If however it still failed to heat (or did so very slowly, or still pumped over) then that would suggest something wrong in that pipe loop.

Because the 22mm path requires 1/3rd of a bar more pressure to get water to pump over. Even if the flow rate is not as fast through the smaller pipe, the pressure required to achieve *some* flow through it ought to be a good deal less.

Perhaps... Is the valve one that can be operated manually (perhaps by removing the solenoid)? If so does it move freely?

Yes it should, sorry I must have missed this detail before. A rad can be used as a bypass (should have lockshield valves at *both* ends in which case).

Reply to
John Rumm

Yup. That's about what I have from the top of the highest rad to the top of the water level in the header tank in a three story house.

Reply to
Dave Plowman (News)

Thanks for all the input!

Turned pump down to speed 1 last night and managed to close all rads except bathroom without pump over. As I started to close bathroom rad, pump over started - what you'd expect I suppose.

BUT

The two radiators piped in plastic numbers 6 and 7 on the sketch below are still not performing.

formatting link
cannot turn off the bathroom rad and leave these open cos pump over occurs. The towel rail has hot feed pipe right to valve, but no flow through it. The valves on this I am not familiar with. They have an almost ball shaped chrome body with white plastic control knobs and are "in-line" style. The lockshield has been fitted on the flow side, this has a free spinning knob! There is a chrome insert on the knob which prises out to reveal a slotted screw recessed. This appears to operate the valve as it comes to a stop when unscrewed, but I'm not convinced! The chrome insert on the control valve has a threaded screw behind it, which screws into the valve stem. Removing the chrome insert allows the valve knob to be removed revealing the valve stem. Don't know how to remove the free spinning knob on the lockshield end!

SO

There is less flow to the plastic piped rad and towel rail (with pump over if only these are open), and I cannot get flow through the towel rail. Are you familiar with these towel rails and associated valves? Both rad and towel rail bleed OK. Any suggestions on the way forward please?

Thanks again

Phil

Reply to
TheScullster

Yup you would...

Which suggests serious flow resistance through that loop.

If the pipe is hot you would expect some heat to get into the rad if only via convection... It sounds like it could be the valve not letting through here.

Have you tried giving it a good pull?

Not the one you descibe alas. Could they possibly be thermostatic? (and how cold is it in your neck of the woods?)

They would do that even with a blocked feed or return though - since water makeup could come from either side.

Perhaps if you drained the rad and turned off the feed side valve (if you can) and then see if it bleeds. Then open the feed and turn off the return side valve and see if it still bleeds, you may be able to confim that water can come through both valves.

none that don't involve making some changes to see what happens...

How difficult would it be for example to try an alternative pipe run with a reel of speedfit draped up the stairs or similar - perhaps tee into the flow before your zone valve, diconnect the feed on a rad and cap the original feed pipe/valve[1], and connect your new temporary feed direct to the rad.

[1] I use a stop end on a short section of pipe in a compression to 1/2" bsp femail connector - this can be screwed onto the rad valve when a rad is removed to make sure nothing is going to come out!
Reply to
John Rumm

undulate up and down at all? If so, there's a good chance there's an air lock in one of the pipes. To make sure that *both* pipes to each rad are clear of air, close one valve and open the other, and bleed several pints of water out through the vent (into a suitable receptacle, of course!). Then open the other valve and close the first one, and repeat.

Reply to
Set Square

This handwheel "appears" to be retained by a circlip near the adjusting screw. Did try pulling it but not too hard!

Excellent and seemingly obvious point that had escaped me!

This is possible as a test but permanent re-routing would be bad news as en suite floor is tiled!

Thanks John

I'll start by trying Set Square's suggestion of closing each valve in turn and bleeding.

Phil

Reply to
TheScullster

Yes I believe they do!

Thanks Set Square I'll give this a go!

Phil

Reply to
TheScullster

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.