There have been a number of posting recently about gravity HW systems which I have read with interest, having put off the job of upgrading the current layout year after year. After doing a good search on google I think I have all the concepts, and think that the way to go is just to stick a valve on the cylinder.
Can anyone help me out, and let me know if there is a simpler/easier/better/more efficient etc etc option before I buy the bits.
I have stuck details on my system on a webpage, so you can see how it is all plumbed together, any thoughts, comments, recommendations would be gratefully recieved. I rather be 110% clear on what I am doing, and get it right first time, than fiddle and have no heating at all!
interlocks to prevent the boiler from cycling when neither CH nor HW require it.
I would thus be looking at either a C-Plan system or an S-Plan system. I would *not* go for a Y-Plan because it would greatly complicate the venting, and 3-port valves are best avoided anyway.
The choice really hinges on whether or not the HW performance is currently adequate. If it is, go for a C-Plan - it's a lot easier. You will need a
2-port valve in the HW circuit right next to the cylinder, *after* the vent pipe tee - so as to maintain an un-interrupted path from the boiler to the vent pipe. You will, of course also need a cylinder stat and a room stat. The radiator nearest to the room stat *shouldn't* have a TRV on it. I presume that you don't get any unwanted gravity circulation in the CH circuit when the pump is not running? If you *do*, you will need an anti-gravity valve - which only opens when the pump generates a bit of urge.
If the hot water takes too long to heat with gravity, you need a fully pumped system - so I would go for an S-Plan. Put a 2-port valve in the HW circuit next to the cyliner, as for C-Plan. Put another one in the CH circuit. The current pump location may be a convenient position. Move the pump to a position in the common bit of pipe before the circuits split. In order to prevent pumping over, move the fill pipe connection into the HW flow pipe - very close to where the vent pipe connects.
I would go for a programmable room stat for several reasons:
It gives you completely independent control over the timing of HW and CH - which most ordinary programmers don't. [Set your main programmer to HW timed and CH constant - so that the room stat conrols the CH]
It adapts to the characteristics of your system/house - and maintains the set temperature with less overshoot
With 'Optimum Start' it decides when to turn the heating on in order to achieve the set temperature by the required time
In the 'Off' position, it acts as an automatic frost stat - and still turns on the heating if needed to prevent freezing
I doubt whether you need a by-pass circuit - even if you go for an S-Plan system. These are needed when you have a boiler which only holds a small quantity of water and which will overheat due to the residual heat in the metal bits if the water isn't carried away quickly enough. Such boilers are only used in fully pumped setups - and have pump over-run stats which keep the pump running after the boiler stops firing until it has cooled down a bit. The bypass circuit provides somewhere for the water to go if all the valves are closed.
If a boiler works ok on gravity without the pump being on at all (when operating on HW only), it can pretty certainly take care of itself.
Not sure whether I have answered *all* of the questions. HTH, anyway!
Thanks for the comprehensive reply, you have given me a couple of things to think about.
Not exactly sure on how long it takes to heat HW from cold at the moment- my guess is that we have a a resonable amount of hot water within about 40 mins from cold - certainly within an hour. Up to now we have not had a problem with having hot water (if we have a bath we might stick on the immersion to "help" the heating) if someone wants a bath straight after. This makes me inclined to stick with c-plan as you suggest. See below re stats.
. In order to prevent pumping over, move the fill pipe connection into the HW
Am I right in thinking that if the fill pipe is moved to the HW flow pipe close to where the vent pipe connects that the system would then effectivly be a "combined vent and feed"? i.e both vent and feed sharing a pipe back to the boiler.
Was going to stick in a programmable room stat in any case - Not sure if a programmable cylinder stat (AFAIK only Danfoss make one - the WP75) would be of any use in saving gas? We have hot water all day (house always occupied), at the moment heating in the morning (no stat), topping up for an hour at about 2pm, and then again in the evening say 5pm - 10pm. by using a prog cylinder stat could use cooler water in day (only use it for washing hands etc) - and hotter in the evening for a bath?
Using a regular cylinder stat we could heat the water in the morning to the higher temp, similar to before, and then leave it to cool during the day as it used up, before reheating the water again to the evening tempreture? The only disadvantage with this is that if the water was used up in the day it would not be replaced to the evening - a prog cylinder stat would realise the water usage and reheat the replaced water to the lower temp.
A prog cylinder stat would realise that the temp of the water had fallen, and then switch on to "top up"
AFAIK the boiler does have a switch for "fully pumped" - therefore the valves could be closed and the pump run. Suggesting again to keep to C-Plan
I see your point. Looks like the boiler would be ok left on gravity (i.e no overrun) - even on a fully pumped syste,
Great. Thanks a lot for your help. I am pretty sure that C-Plan is the way to go, unless someone come up with something else.
Now I only need to decide between the Honeywell CM67RF or Danfoss TP7000 - and if to use a regular cylinder stat or programmable cylinder stat!!
Am temped to stick to danfoss with the TP7000 (optimum start/programmable etc) and a HW cylinder programmeable therm WP75.
The really important question is whether the boiler you have requires a fully open gravity circuit as a safety measure. If it does (as many ancient or solid fuel boilers so require) then your options are far more limited.
You're laughing then. Run in whatever way the plumbing works best. Ensure there is a direct unvalved path from boiler output to vent and a direct unvalved path from water inlet to boiler return. Then place pumps and valves where appropriate. You can share valves or use separate as you wish. You can run each of the 28mm or 22mm system as your desire. Pumps don't count as valves if in the water inlet path. A fixed (non automatic i.e. gate) bypass valve counts as a path for the water inlet. If the boiler has an additional manual reset overheat cutout, then the water inlet restrictions can be ignored. The water inlet could even be combined with the vent under these circumstances, which eliminates pumping over.
Sort of. It's probably best to keep them separate, but connect then both into the same main pipe fairly close together. Mine are connected in a few inches apart - the idea being that there should't be an appreciable pressure drop between the connection points. [This only applies to fully pumped anyway - and you can ignore it if you go C-Plan]
I can't see a lot of point in programmable cylinder stats. I suppose if you really want different temperatures at different times of day, and want more on/off cycles than are provided by the ordinary programmer, you could make a case for one. However, if your tank is well lagged, the heat losses should be very small anyway - and it does little harm keeping the water slightly hotter than you actually need for some of the time. Features like optimum start and the ability to adapt to the system are less likely to be useful on a water heating system.
Use whatever you're comfortable with. I've got a (non-RF) CM67 with optimum start (they don't all have it) - which I fitted a few months ago to replace the original conventional room stat. It seems to meet my needs fairly well. I've also got a conventional cylinder stat, which I don't intend to change. [FWIW, I've got a Y-plan system - which was in the house when I bought it - which works adequately but isn't what I would have chosen].
I still reckon gravity with control is the way to go. as we do not have a problem with HW recovery time and it is a lot simpler to fit.
separate as you wish. You can run each of the 28mm or 22mm system as your
(non automatic i.e. gate) bypass valve counts as a path for the water inlet
water inlet restrictions can be ignored. The water inlet could even be combined with the vent >under these circumstances, which eliminates pumping over.
The boiler is fitted with a manual reset overheat cutout, does this mean that the 2port could be fitted in the HW circuit close to the boiler? The vent would still have an open path - and the feed have the 2 port valve inline but would still be in the HW circuit and stop circulation.. This contradicts what set square recommened which was to fit a "2-port valve in the HW circuit right next to the cylinder, *after* the vent pipe tee - so as to maintain an un-interrupted path from the boiler to the vent pipe" whilst this has the benefit of leaving the feed and vent open at all time - the ability to stick the valve in the feed/HW circuit close to the boiler would mean that I could use a RF tank stat
formatting link
and save running a cable to the loft making fitting extremly simple.
Not much in it. Just put the pump right next to the zone valve and wired off it, set on lowest speed. Check for pumping over, both continuous and at pump start/stop and increase the length of the vent loop over if it does.
Now you're really laughing. This allows you to close off the old feed pipe and take the F&E cistern feed to the vent pipe with a T. Guaranteed no pumping over. Only safe with a manual reset overheat boiler. These don't require the quenching water flow for safety, so can live with having the pipework blocked by escaping steam until the thermostat blows.
Yes, with such a boiler, you can place the valve anywhere that doesn't interfere with the vent path.
Maybe an option to go PF then. Will have a climb up tomorrow and have a look how tight the pipework is to move the pump?
but....(brain ticking over)
1) is the bathroom radiator currently acting as a bypass?
2) if I place the 2port by the boiler would I cause any issues with not having a bypass in the system? would the pump not overun - and need a bypass? or is the system ok being fully pumped - and set to "gravity" mode?
3) If I go to Fully Pumped I assume a second 2port will be required after the pump for the CH circuit.
No. By-passes are only used in fully pumped systems where there may otherwise be no water path if all the valves are closed. That ain't what you've got!
I'm sure that Christian is technically right - but I would feel uneasy if I didn't have a clear path from the boiler to the open vent - even though the boiler has a manual cutout. Belt and braces, maybe?
What you probably *could* do is put the 2-port valve for the HW in the
*return* before the CH and HW circuits re-merge - and move the cold feed so that both vent and feed pipe are tee'd into the HW *flow* pipe. That way, you'd still have your open vent path, and the 2-port valve would be in a position to make the wiring easier - and you could have your RF cylinder stat. You could do all this regardless of whether you end up with a C-Plan (semi-gravity) or S-Plan (fully pumped).
3) If I go to Fully Pumped I assume a second 2port
If you go fully pumped (and assuming only *one* pump), the pump will no longer be in the CH circuit - you'll need to move it to the common bit, otherwise it won't pump the HW. I would put the CH 2-port valve where the pump *currently* is - because you'll have to modify that bit of pipework anyway.
I still don't think you need a by-pass circuit. The question is this: When the boiler burner turns off - either as a result of the boiler's own stat turning off, or the room stat turning the boiler off - does the water in the boiler continue to get hotter due to residual heat to the extent that it will overheat and trip its manual cutout *unless* water continues to circulate through it? The fact that your boiler is quite happy with a gravity system with no forced flow through it in HW-only mode indicates that the answer is "no". If the answer were "yes", the boiler would have to control the pump - by means of a pump over-run stat - and would have to keep the pump running long enough for the boiler to cool down. Such boilers will only work in a fully-pumped scenario - so yours isn't like that.
You may have misunderstood. There still must be a clear path from the boiler to the vent. The manual cutout removes the need for a clear independent path from the cistern feed outlet to the boiler. This path is then allowed to either be valved or shared with the vent path (but not both, obviously as the vent path would then become valved).
The solution I suggested actually involved an additional pump, rather than using the CH one. The CH circuit can stay just the same. The new pump just pumps the hot water circuit. The new valve just controls the hot water circuit.
You are correct in saying that you may now need an additional bypass, although the effect of not having one may not be predictable and it might not be required. With this type of system with two independent circuits, there should be a bypass on each circuit. The CH circuit probably has a bathroom radiator doing this task, whilst the hot water circuit should have an automatic bypass valve that is simply plumbed in parallel with the zone valve set to around 0.3 bar.
Not true, actually. The gravity circuit fails efficiency tests on several counts. Firstly, the long reheat time keeps the pipework and boiler hot for unnecessarily long periods, wasting heat. Secondly, the reduced heat requirement from the poor circulation coupled with a gravity boiler's typical lack of modulation ability leads the boiler to cycle excessively, which wastes energy.
Thanks for the clarification. In an earlier post, you appeared to be saying that if the boiler had an over-heat cutout which had to be reset manually, it was no longer necessary to have a clear path to the vent.
Using 2 pumps sounds a bit complicated to me. In a conventional S-Plan (with a single pump) each thermostat switches the appropriate zone valve - and the volt-free contacts on the zone valves switch the boiler and pump on whenever one or more circuits are calling for heat. In cases where the boiler has a pump over-run stat, the pump is controlled by the boiler rather than directly by the zone valves.
If you have two pumps, I assume that each stat would have to switch one pump and one zone valve - with the volt-free contacts switching just the boiler. This would mean: A) the pumps would start before the zone valves were open B) the pumps could not be controlled by the boiler, so there could be no pump over-run
My feeling is that the OP has sufficient scope to be able to convert either to a C-Plan or an S-Plan, without the added complication of two pumps.
This sounds like another of your unsubstantiated assertions - because you haven't explained WHY.
Is it because you haven't heard of C-Plan [See
formatting link
] and associate gravity circulation with the boiler being on all the time, whether required or not, and with no control over the HW temperature, so that it rises to near boiler temp?
If so, gravity systems don't *have* to be like that!
OK, I accept the modulation bit - but the comment was made in the context of someone working with his existing boiler - so changing to a modulating type didn't really apply. The other points are covered by my statement if you include the associated pipework with the boiler.
[I guess there is a danger that comments can be made about specific cases in such a way that they appear to be generalisations, and may not be valid as such. I accept that I am sometimes guilty of this.]
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.