converting gravity fed HW CH to something better

[nospam]

Hello,

We've got a gravity fed CH system. I think you call it plan C. I don't know everything because much of it is hidden under the floor or behind boxes but it seems as though we have 28mm pipe from the boiler to the cyclinder. There is only one pump (next to the boiler) and no valves. The pump is energised by the room thermostat for the CH. There is no cylinder stat. We can have HW only or HW and CH but not CH only. Does this sound about right?

A neighbour has just paid thousands to have his system converted so that HW and CH may be switched separately. Is it worth us doing the same (the converting, not the paying thousands!)

Is it just a matter of inserting a valve and a bypass to control whether the hot water goes to the cylinder only or carries on to the radiators? What else is involved?

Thanks.

[Ed Sirett]

This work would normally be done as part of a package of upgrading the heating system including a new boiler. There is no reason why you couldn't improve the existing arrangements. We have discussed heating over and over on this ng. The FAQs below, The wiki and the subject of C-plan should get you started.

[David Hansen]

On Mon, 17 Sep 2007 16:48:13 GMT someone who may be nospam wrote this:-

It sounds like the hot water is heated by gravity circulation of primary water. Fairly standard a long time ago, but increasingly uncommon.

It is worth converting to a fully pumped system. It sounds like your neighbour had a lot more done, or he was diddled.

No. The pipe that circulates primary water to the hot water cylinder almost certainly continues to form the expansion pipe. You can probably see this if you look at the cylinder connections and follow the pipes.

Installing a valve that will shut this pipe converts the heating system into one which has inadequate provision to cope with the expansion of the water as it heats up. The end result tends to be a loud bang and lots of very hot water being flung around.

[Roger Mills]

Whatever you've got, it ain't a C-Plan - because that would have a valve and a cylinder stat, and would permit CH without HW if desired.

How many pipes are connected to the boiler? My guess is that there are four - two for the gravity HW circuit and two more for the pumped CH circuit.

If that is the case, you can convert it to a C-Plan, which will give you independent control over the HW and CH - albeit not quite as efficiently as as a fully-pumped system, but nevertheess a lot better that what you currently have.

All you need is a 28mm 2-port valve of the right type (see the C-Plan diagram on the Honeywell site) and a cylinder stat. As others have indicated, you have to be careful where you insert the valve so that it never prevents the boiler from having a clear path to the expansion tank. If the pipe to the expansion tank tees off near the HW cylinder, the valve needs to be on the *cylinder* side of the tee. You'll probably also find that your programmer has a link fitted which forces the HW on whenever the CH is on in order to ensure that the boiler runs for the CH, 'cos the room stat only controls the pump. But if you re-wire as necessary to comply with the C-Plan diagram, you should be ok.

If you're doing it youself, it should cost you less than £100 for the bits.

[nospam]

Hello,

You are quite right that it isn't a C-plan. I thought the C-plamn was the simplest set-up but I re-read the Honeywell site and the C-plan does have a valve; so my set-up is not C-plan but something even more basic.

Yes there are.

The cyclinder is covered in foam insulation. Do I just carefully cut out a chunk to fasten a cylinder stat to? Then wire the cylinder stat to the controller.

What is the function of this? Does it switch off the boiler when the HW is at the right temperature?

I will have another look at the Honeywell site to see where the valve should go.

[snip]

Correct again.

[nospam]

I'm not sure exactly what he had done nor exactly how much he paid. I saw a new cylinder being carried in too. Perhaps he went fully pumped? What are the advantages of this: the HW gets hot quicker?

[nospam]

Thanks. I had already found and read the faq which pointed me to the Honeywell site. The faq is great but I needed to see the pictures at Honeywell. Where do I find the wiki?

[nospam]

Sorry to answer my own question but is it:

?

[Andy Hall]

Yes.

The thermostat closes the valve to stop circulation around the gravity loop and a switch on the valve switches off the boiler.

This is why you have to be careful where you install the valve - i.e. not on the path from boiler to vent pipe.

That's a stylised diagram so be careful. Trace the pipes to the roof tank and be sure that the valve is not between boiler and vent.

All of this set up only has the effect of saving some energy by preventing the boiler cycling when the water is already hot.

If you want to improve performance, and you are installing a new controller anyway, it's usually not much more work to convert to fully pumped using a three way valve.

Then you can have completely separate CH and HW control as well as the water heating more quickly.

[nospam]

Sorry to post so many questions!

I have had another look at the cylinder. Two 28mm pipes come into the airing cupboard. These are then reduced to 22mm pipe which connects to the side of the cylinder. The one pipe has a tee: one side of the tee connects to the cylinder, the other goes into the loft to vent in the tank.

So it looks as if I need a 22mm valve which blocks the flow into the cylinder but leaves the other side of the tee (boiler to vent) open at all times: is that right?

The two pipes are not marked. Is the one with the tee piece the flow into the cylinder and the lower one the flow out? Do you always put the valve on the pipe with the tee?

I have looked at the Honeywell diagram and read the wiki but I am still confused about one point. The diagram makes it look like a loop: boiler to cylinder to radiators to boiler. If a valve stops flow at the cylinder, won't it stop flow around the whole circuit? Doesn't there need to be a bypass so that when the cylinder is blocked water is diverted straight to the radiators?

Thanks.

[RobertL]

So it's not totally gravity fed, the CH is pumnped, only the HW is gravity fed.

My father's old system was completely gravity fed and the gas boiler did not use electricity (clockwork clock). It might not have been very efficient but it worked during power cuts!

Robert

[Andy Hall]

Yes that's it.

The thing is to avoid it being on the vent path and to make it only on the cylinder path. Other than that it doesn't really matter.

The boiler will be fired up if either the HW or the CH demand heat. Assuming that this is allowed by the timer, then the room thermostat indirectly controls the pump running and demand to the boiler. If you do the Cplan, the thermostat on the cylinder opens the valve and that fires up the boiler.

[nospam]

Thanks for your reply but I am confused about this: if the cylinder stat says that the water is hot but the room thermostat says the room is cold, won't this switch on the pump and close the valve? If so, doesn't the water that goes through the radiators also circulate through the cylinder to warm it? If the valve is closed, won't this water be unable to move, damaging the pump and leaving the radiators cold?

[Andy Hall]

yes.

No. It should go through radiators only.

If you are doing Cplan, the branch to the radiators needs to be before the cylinder and valve.

In effect this means that the CH circulation is based on whether or not the pump is running and the HW on whether or not the valve is open.

The valve and branch to the pump need to be positioned appropriately and the vent allowed either way.

That's why I said that it might be almost as easy to do a three way valve and pump all of it.

[nospam]

Thanks for your continued help. Is there a diagram of the C-plan other than on the Honeywell site because I think if I saw another schematic I would understand. It's just that the Honeywell diagram gives the impression that its one big loop, so I can't see how the rads can work when the HW valve is closed.

If I did this, what plan would this become? What advantages would there be: I think I read HW heats faster is this right? What more work would have to be done: just adding a pump to the HW pipes?

Thanks.

[Roger Mills]

Yes, you've got uncontrolled gravity hot water, and pumped CH. The HW heats whenever the boiler is on, and will eventually get up to a scalding temperature if you run the radiators hot.

Thought so!

Yes. It's probably held in place with a piece of stretchy curtain wire round the cylinder, which can go outside the foam - but the stat itself must make metal to metal contact with the cylinder.

In effect, though not directly. The cylinder stat opens the valve when the HW needs to be heated, and allows it to close (by its spring return) when the HW demand is satisfied. The actuator of the valve has a changeover switch - which is electrically isolated from the motor circuit - which changes position when the valve is fully open. This switch causes the boiler to run via the room stat and pump when the valve is closed (CH only) and causes just the boiler to run when the valve is open (HW only). If there is a simultaneous demand for HW and CH, the switch on the valve controls the boiler, and the room stat controls the pump. All very cunning - but it

*must* be wired exactly as per the Honeywell diagram if it is to work.

[nospam]

Sorry to answer myself but I have found this:

this diagram look right? If so, I finally understand it. According to this diagram the cylinder and the radiators are piped in parallel rather than in series. If this is the case I can understand how the cylinder can be closed off and the radiators can be heated at the same time.

But I am surprised because if the HW is on, what is to stop the heat "leaking" into the radiators? Is it just that it follows the path of least resistance (i.e. to the cylinder). Also how does the water get heated when the CH is on? I would have thought that all the hot water would get "sucked" away by the pump and very little would find its way to the cylinder.

Looking at these various diagrams they all seem to show the hot water from the pump going into the upper side connection of the cylinder and out of th elower side connection. I thought it would have been the other way around considering that heat rises, or is this to prevent shock?

Thanks.

[Roger Mills]

No! You've got two separate circuits. One goes from the boiler to the cylinder by gravity circulation, and back to the boiler. The other goes from the boiler and is pumped round the radiators and back to the boiler. Closing the valve stops the gravity circulation to the cylinder, but still allows the pumped CH circuit to operate.

In general, you'll want your HW at about 60 degrees and the flow to your radiators at 80 degrees. So the boiler has to operate at 80 degrees. As soon as the water in the cylinder reaches 60 degrees (or wahatever the cylinder stat is set to) the valve closes. The boiler can continue to heat the radiators, but the HW doesn't get any hotter - which is exactly what you want.

[nospam]

Thank you. The Honeywell diagram confused me because it is abit vague about what happens at the boiler.

You say I've got two circuits but I presume they must be connected because they both fill from the CH header tank don't they?

I know there are four pipes going to the boiler 2* 28mm and 2 * 22mm (HW and CH respectively) but I can't see what they do behind the boiler to work it out myself.

One suggestion was to put a pump on the HW too, but I just looked at this:

they have the pump positioned before the cylinder-radiator tee so that it pumps both. Is there are reason why two pumps are preferable to doing this?

Thanks again for your help (and patience)

[nospam]

I've done what i should have done before I posted: had a good look at the boiler: on one side there are two 28mm pipes and on the other two

22mm pipes. Just curious but are these heated separately by the boiler? If not, I'm wondering why there isn't just one inlet and one outlet to the boiler.

Thanks.