Previously, the vent pipe and the feed/expansion pipe were effectively connected at two points on the gravity circuit with the cylinder. The flow rate and hence the pressure between the two will have been very low and not enough to cause pumping over from the vent pipe or sucking down through the feed pipe.
Now you are putting the resistance of part of the boiler heat exchanger, a run of pipe and possibly the pump between the two. There will be a much stronger flow because of the pump and there will be a pressure difference between the two places on the circuit where the vent and the feed pipe are connected. That may be enough to cause the aforesaidmentioned pumping over or sucking down. You don't want that because apart from the noise and the water vapour and heat loss in the loft, it will accelerate corrosion of the system because of introduced oxygen.
If the FE and the vent pipes are joined to a point on the circuit where they are close together, then the recommendation is for the two connecting points not to be more than 150mm apart. That way, you can't have a great enough pressure for anything bad to happen.
The alternative, is an air separator. Have a look at
formatting link
product code 11334 or 11335. To use one of these, you would insert it on the flow from the boiler before the pump and then connect both the FE pipe and the vent pipe to it. Effectively, you would remove the vent pipe from the old HW side of the boiler and move it to the CH side. The important point is that there is nothing between the boiler and the vent like any pumps or valves. This approach also has the advantage that after filling, air bubbles are expelled very quickly and easily and bleeding is less of an issue. This would leave you with a high point on the cylinder connection, probably. To that point, you would fit an air vent - either a manual one worked with a key or an automatic one.
It's probably best at this point to sketch out the whole thing and figure out whether you will be breaking the rules over where the pipes join. For example, it may well be the case that you will need to join everything to a point leading to the old CH side and do away with the HW side of the boiler altogether, simply plugging the taps.
Hang on a bit! For a fully pumped system with a single pump, you *can't* have two pairs of pipes connected to the boiler - you must have *only* one flow pipe as far as the pump, but you can have separate HW and CH returns - making a maximum of *three* connections.
It sounds as if you have separate upstairs and downstairs CH circuits - probably teeing off near the boiler connections? If that is the case, you must disconnect the CH flow pipe at the boiler, and blank off the boiler connection. You can re-use the pipe, but feed it from the *top* end, after the pump - so that water going to the downstairs rads goes up the 28mm pipe, through the pump, through the CH zone valve, and back *down* the 22mm pipe (i.e. the opposite way to before) to the downstairs rads.
As stated before, the rad returns - and that includes both upstairs and downstairs - must all join together before joining into the HW return or - if it's easier - they can go all the way to the boiler and not join into the HW return at all. From my current perception of your setup, that probably would be easier. Unless I'm mistaken, you already have CH returns from both floors going back to the boiler, and joining together just before the boiler connection. If that's the case, leave that bit as it is - apart from inserting a bit of pipe where the pump used to be.
On Sun, 27 Jan 2008 16:25:19 GMT someone who may be Fred wrote this:-
They certainly are. Otherwise how could the central heating circuit be filled up?
In that situation I would add a second pump for the hot water and have a two pump system. No motorised valves to go wrong and easy to adjust with the pump speeds. If the existing pump was in good condition I would move it to the hot water circuit and replace it with a smart pump on the heating.
With such a setup the position of the vent pipe and feed & expansion pipe might not be so critical. I might even try it and see if I could get away with it without pumping over.
I suppose it's possible that the boiler may require pump over-run if used in a fully pumped system, but not when in a gravity system. It obviously can't over-run the pump in a gravity system 'cos when only the HW is being heated, the pump ain't running at all!
As you say, a gravity HW system would act as a sort of by-pass (gravity flow only) to get rid of some of the boiler's heat. A manual by-pass circuit would also allow some gravity flow when both zone valves are closed - but there's no way that gravity flow would generate sufficient pressure to open an automatic by-pass. So unless the boiler explicitly requires a pump over-run in a fully-pumped situation and controls the pump - there's no point in having an automatic by-pass valve.
A pump - such as a Grunfoss Alpha - which varies its output depending on the resistance in the system. Typically used on CH systems which have TRVs on most radiators. When most of the TRVs are shut, so that the system flow resistance increases, an ordinary pump would build up a lot more pressure, whereas a 'smart' pump will back off. In simple terms, it's attempting to produce a more or less constant pressure regardless of flow - although it can't *quite* achieve that.
Sorry I was explaining what connections were on the boiler, used or not. You are quite right that I am using three of the four.
That's what I've done. So far so good.
I took out the pipes because they were touching each other with no room for lagging. I have put three back, with lagging, but I haven't re-laid the fourth. I was hoping not to, and share the HW return, but it looks as though I must?
'Fraid so! You *could* bring the downstairs return up to the airing cupboard and merge it with the upstairs return before connecting them into the HW return - but you'd *still* need the fourth pipe, and the downstairs water would have further to travel - so there's not much point. So take the upstairs return down, and merge it with the downstairs return before connecting the merged returns into the boiler.
Interesting. Does this mean that a bypass valve is not required in a system with such a pump?
Someone said something about when the boiler switches off, the hot water in the boiler will want to rise somewhere so would I need a bypass valve for that regardless of whether or not I have a smart pump? Surely any boiler that would be damaged if the water cannot rise would need an over-run anyway?
I've added the fourth pipe as per your other post. I'm unsure where I would fit a bypass to. Since the pump is in the airing cupboard, can I tee into the HW return (also in the airing cupboard) or must I run a fifth pipe downstairs?
It's a mistral oil fired boiler. Are they still made? The instructions left by the previous people look quite old and are nothing but an A3 page with specification on.
Does anyone know if Mistral exist, and if so how I can get in touch with them or whether the bypass is required with this boiler?
No really - the two are not directly related. A by-pass circuit doesn't protect the pump (most pumps will happily run stalled anyway) but protects the boiler by providing a flow path to enable residual heat to be dissipated.
A lot depends on the boiler design. Many modern boilers hold a relatively small amount of water. When an external control (room stat etc.) turns the burner off while the boiler is in full flight the residual heat in the metal parts continues to heat the water - to the extent that it could actually boil if there is no flow to carry the heat away. Such boilers will only work with fully pumped systems, and invariably directly control the pump - keeping it running for a while after the boiler stops firing. There needs to be an open flow path during this 'pump-over-run' phase - which is where a by-pass circuit comes in, to provide a path when all the zone valves are closed.
Older boilers are often physically larger, and hold more water - and are much less in danger of overheating when the flame is suddenly cut. The fact that yours has happily worked with an un-pumped HW circuit suggests that it's in this category. It *probably* doesn't need a by-pass but - as Andy H suggests - you really need to study the installation manual, or contact the manufacturer to determine whether there are any specific installation requirements, depending on whether it's being used in a gravity or fully pumped system. In a gravity system, there is always a flow path open - albeit convection flow only. With your re-arranged system there will be *no* flow path when both zone valves are closed. Only the boiler manufacturer will know whether this matters and, if so, what to do about it.
Thanks. On a slightly different note, if pumps will run stalled, what is the worry about having TRVs closed? I thought if all rads had trvs then you had to have a bypass to protect the pump?
You say this heat has to be pumped away; would the convection in a gravity HW system be insufficient to dissipate this?
I am in the process of trying to find my boiler manufacturer for their advice.
Sorry, I hit Send too soon in my last post, before I had replied to this paragraph.
I see no reason why the by-pass can't simply be in parallel with the HW zone valve and heating coil - i.e. connected into the flow pipe just after the pump but before the zone valves, and returning into the HW return. I don't think you need a fifth pipe back to the boiler.
The type of by-pass is important though. Unless there is some mechanism for keeping the pump running after the boiler demand has been cut, there's no point in having an automatic bypass because there will never be sufficient pressure to open it. So a manual by-pass may be your only option. The problem with manual by-passes is that they are always open, and partially short-circuit the HW and CH circuits even when not needed. At least there would always be a flow path (if the boiler needs one - which is by no means certain) and it would allow gravity circulation akin to what you previously had.
If the boiler can't directly control the pump (it will have specific pump connection terminals if it can) and if you want to play safe, you could do the following:
Fit an automatic by-pass valve in the position discussed above
Wire the pump to a bathroom fan timer[1] set to (say) 2 minutes. This would have the effect that, when the boiler and pump demand are cut, the timer would keep the pump running for a couple of minutes, the bypass (if correctly set) would open, and any residual heat would be carried away from the boiler.
No it doesn't. Where is is helpful, is if you have TRVs. As the valves close because the place is warm enough, the pump will begin to reduce output.
If you need a bypass valve, you need one.
It really depends on the boiler. The case that you are trying to cover is that the room or cylinder thermostat is satisfied and cuts off when the boiler is hot and burning. The burner is stopped instantly. However, without over-run and circulation to somewhere to dump the heat, the temperature of the water may rise to boiling point or close to it producing noise and perhaps steam. You don't really want that.
This is more likely to be an issue with boilers having a high thermal mass - e.g. cast iron because there is more retained heat. Newer boilers tend to have low thermal mass and low water content and it's less of an issue. For example, on mine, the pump is inside the boiler and the heat exchanger is stainless steel with a coiled tube inside. The bypass is a short length of thin tube inside the case of the boiler itself. Normally, the boiler is operating the motorised valves and is able to dump excess heat into the circuit immediately after closing down. However, if that isn't possible, there is enough by simply continuing to circulate the water inside the boiler.
That's why you really need to find out from the manufacturer what they require.
This is getting complicated. On the flow side, the bypass needs to come from a point after the pump and before the motorised valve(s). It should really be returned to a point after the CH and HW have been joined together. If that is effectively at the boiler because you are doing the final joining of these circuits are parts thereof at the boiler, then theoretically you need a pipe for it. However, the flow through the bypass will be throttled down anyway if you use a lockshield, so if there is any reverse circulation, it will not be much. There is an easier path through the main circuit. If you use an automatic bypass valve, it will only open for the couple of minutes after the boiler has been firing and the CH and HW thermostats are off. i.e. you won't really notice the effect - a couple of radiators might become briefly warm.
I suppose that you could suck it and see with a lockshield valve, but given the hassle of draining again if it doesn't work, I would put in an automatic bypass valve and be done with it.
We have a Mistral oil fired jobbie, floor standing about 4 x 2 x 3 feet containing a fecking great lump of cast iron holding 12 gallons of water... A bypass isn't required like it is for small high themal output, low volume boilers that literally melt if they don't have a bypass and pump over run.
Ours needs overun but that beacause it is a monster 38kW output with, under mosts circumstances, a much smaller load so it tends to trip it's overheat stat. If yours is better matched to the system I expect it'll be fine without a bypass. With the proviso that the control system won't have the pump running into a shutdown system.
I recently had a dig about on the web for Mistral. I'm not sure they are still trading as such, they last address/phone number I have are Telford based. Google is your friend.
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.