I'm in the throes of (slowly) fitting a new central heating system. The
house is a 2 bed terrace in North London. The boiler is a Keston Celsius
with the system kit and is located in the first floor bathroom. I've
removed the middle section of the chimmey breast to about head height to
create a cupboard and run the flue and intake up the chimmey. I've got
seven rads sized according to the Myson Heatloss calculator (although
I've reduced the size of the kitchen rad as the program can't take
account of all the cupboard volume). The two bedroom rads will have
thermostatic valves but the main living area (living room/dining room)
will be controlled by a room thermostat. I'm setting it up as an S-plan
with a Landis & Staefa RWB9 programmer set for 5+2 operation.
In any case I was looking around the Honeywell site as a result of
another thread (Central Heating - Controls) and I noticed this:-
"If a bypass circuit is fitted, an automatic bypass valve must be used.
(Slumber or Bypass radiators are only acceptable on solid fuel
on this page:- http://content.honeywell.com/uk/homes/Regulations.htm
I was intending to use the bathroom rad with a couple of lockshield
valves as the bypass but it now seems I'll have to scrub that. :-(
Is that correct?
Looking at the BES site (thanks Ed and why isn't it in the FAQ?), is the
Differential By-Pass Valve (12161) the approiate part?
Looking at the diagrams the bypass valve should be fitted from the flow
to the return between the boiler and the zone valves. But according to
the Keston installation instructions the flow/return temperature
differential should be 10 to 15 C. In the layout I was planning the zone
valves are only about a meter from the boiler. I can't see that short
run of pipework producing that much of a differential. Any thoughts?
Any other thoughts/observations would be welcome.
The idea is to set the bypass pressure on the automatic bypass such that
it will only open when all (or most of) the TRVs on the system rads have
closed and are up to temperature. When the bypass operates, it stops the
boiler seeing and 'open circuit' giving rise to pump and boiling problems.
When the boiler sees the return water rising in temperature, it will
modulate down and eventually turn off, which is what we want as the house
has reached the desired temperature. It will restart again when the temp
There is nothing to stop you having a towel rail on bypass and an
automatic bypass in parallel.
I experienced a problem (with a near identical system to yours) where the
bypass was close to the boiler and would tend to 'pop off' rather than
gently increase in flow as it was coming into operation. This resulted in a
nasty, but slow, pressure oscillation when the system was running on
bypass. To cut a long story short, I chose to limit the capacity of the
bypass by placing a 15mm lockshield valve in series with it. This was
enough to damp the oscillation.
Thanks for your reply , Fred.
I've now mailed and spoken to Keston and I'm still not clear about how
to deal with this issue. What is normal practice since Part L came in?
Are installers simply ignoring the boilers requirement for a temperature
drop across the by-pass circuit? Or are they creating an (artificially)
larger loop for the by-pass circuit? Or are they still using a rad as a
Thanks for your responses.
It is actually a lot easier than you think. Just think of the (automatic)
bypass not coming into operation at all when the heating is at full demand.
Its purpose is to provide an uninterrupted water path _only_ when the TRVs
on the rads are shutting down, which is a transient state just before the
boiler says 'right, the house is hot enough, I'm shutting down', so don't
worry about it. The absence of a temp drop across the bypass will tell the
boiler that it is time to shut down.
It is actually more complex than that as at first the bypass will come on a
little and the return flow will be a mixture of rad return water (at low temp)
and flow water ((hot) from the bypass). This will have the effect of
increasing the return temp a little, which will cause the boiler to throttle
back a bit. This is not a fault or a contravention of regs 'cos it is the normal
control loop coming into play and (eventually) shutting down the boiler
when the house is up to temp.
Hope that makes it a little clearer.
The main radiators in any house are in the hallways or reception rooms
(front lounge) so this radiator should, in actual fact, be your by-pass
loop. The proper thing to do with any heating system, is to place a
thermostat in the same room or area as the by-pass radiator. This way the
system is actually controlled by the loop and not from any radiator willy
nilly telling the system that it is to hot and should close down everything.
TRV's are there to tell the radiator that it is fitted to, that the area
that it is heating is up to temperature, that way each room, or valve
controlled zone, because that is what it literally becomes when you install
a TRV, will independently control its own temperature.
The main area of the house needs as much heat as it can to stop draughts and
colder air moving from room to room when doors or windows are opened, so the
radiator installed in the main area should be controlling the boiler and
other main parts of the system through an independent thermostat fitted at
the opposite end of the hallway from the radiator. That way the whole room
is up to the correct temperature setting and is now helping to keep the
whole house warm. No one actually sits in the hallway, but it is an
integral part to keeping the whole house heated.
I still think that underfloor heating is the best way to control the
temperature in a house. The heat is actually spread more evenly throughout
each room, and when a 'stat is installed in the main area of the house, then
you know that each area is at a comfortable and liveable heat.
So if you think of what the manufacturer is saying about a drop in
temperature over the by-pass loop, and think of the size of radiator you've
fitted in the main area of the house, then I think you'll have spotted where
the loop should be made. Along with a room 'stat if you want even better
I'm a bit confused by this.
If the bypass valve is normally shut your primary rad will be mostly out of
action until such time as all the other rooms are up to temp as specified by
the TRV's. Because the room that the primary rad is in is still cold the
boiler still has a heat demand which then flows through the bypass to heat
that area until the room stat says ok and shuts the boiler down?
What actually causes the bypass valve to function?
The Heating and Hot Water Guide to Part L states that:
"If the boiler manufacturers instructions advise that a bypass is to be
fitted, as a requirement ot the new boiler, then an automatic bypass valve
must be used in conjunction with any requirements for a minimum pipe length
specified in the manufacturers instructions."
Now if Keston specify an explicit temperature gradient for the loop then it
implies that whatever length of pipe for thet to happen is necessary. It
also doesn't specify that a rad cannot be used to achieve this.
However, if the bypass isn't normally an open loop now then there is a
potentially large functional loss to the system
I'm concerned by this as I am about to replace my boiler with a Celcius 25
and convert to a fully pumped system <probably W plan> with an open loop to
feed towel rads as a bypass.
It seems I may not effectively be able to do this any more and will be in
the same boat as the otehr Paul.
A by-pass loop is totally different from a by-pass valve. An unrestricted
flow of heated water from the boiler must be allowed to travel through at
least one radiator circuit to prevent the boiler from over heating in the
event that a fault occurs. The manufacturer states that a minimum
temperature loss must be lost in that circuit to make sure the boiler
doesn't have water it has just heated, flowing straight back into it. The
unrestricted loop is for your safety and in that case, it is in your own
interest to have the system fitted correctly.
But your interpretation bans the use of zone valves. When the room
thermostat is satisfied and the zone valve closes, where's your bypass loop
now? It needs some sort of automatic bypass loop to take the water as the
boiler pump overrun happens. Losing 10-15C over an automatic bypass loop
does seem somewhat inachievable, and probably illegal. Of course, the boiler
should have shut down as the zone valves started to move, so it shouldn't be
adding more heat to the circuit.
Part L does not allow any heating appliance to be uncontrolled. Every heat
outlet has to have either a TRV or thermally controlled zone valve in its
circuit, so can't be relied on as a bypass.
Thinking about it, it could be achieved by using the automatic bypass valve
to bypass a zone valve. You could use the hot water cylinder circuit (unless
it is unvented) or a heating circuit (assuming there is at least one
radiator physically locked on with no user controls).
Pain in the butt though....
If it's just 4 metres it's not so bad; I've just got visions of this
stuffing great pipe looping it's way round the house to burn off the
But if there is a stuffing big radiator within the loop, then heat is burned
off easily and then the pipework doesn't have to be all that long. All a
manufacturer asks is for you to make an unrestricted loop of some kind,
which will dissipate a required amount of heat so that if a fault occurs on
the boiler itself, the heated water is not circulating directly back into
the boiler to quickly and is loosing a required amount of heat to stop it
steam venting or splitting through the build up of pressure. I've actually
seen it done with a heated towel rail on the other side of a wall from the
boiler and no more than four feet away in the actual run of pipe, but it was
enough to satisfy the requirements for that system.
Because it asks for a loop, the loop itself doesn't have to be a loop of
pipework running all around the house. All it is, is an unrestricted
circuit which will dissipate the heat required and do it by what ever means
possible for the situation. A valve doesn't have to be used either, as a
normal open unrestricted loop will more than usually suffice for a domestic
How I was taught to do it was, place a circuit on the flow and return pipe,
but make it so that, that one particular circuit heated when a call for heat
request was given to the boiler even if the motorised valve or pump failed
to activate. The easiest way to do this, is to install a three port valve
that when de-energised, defaults to room heating closed and water heater
open. This makes the circulating coil for the hot water storage tank your
unrestricted loop. It is the perfect unrestricted loop between flow and
return with a perfect medium for dissipating heat from the boiler. Yes, you
got it. The "Water" in the storage tank.
Tell me how many times you've turned a conventional central heating system
on and not wanted it to heat the water you use at the sink or in the bath.
You've probably wanted hot water without heating the radiators in the house,
so you would fit a motorised valve to the radiator circuit to stop the flow
from circulating around the house, which is what it is supposed to do. But
I'd really like a show of hands on how many people call for room heating and
don't want the hot water tank heated to go with it.
The thermostat on the hot water storage tank is fixed in line with the three
port motorised valve and when it call for heat when only hot water is
required, then the valve is already in the default open position for that
circuit, so the valve doesn't have to move at all. When a call for heat is
made by both the room heating circuit and the hot water, then the valve
moves to mid position to give both circuits a flow of water. When the
thermostat on the tank reaches its setting, it then moves the valve to close
only the hot water side to stop the flow to the tank. This would only
happen when both the heating and hot water setting is selected on the
programmer or timer system by the connections to the switching of the valve
itself. When only the hot water is on, then the tank 'stat tells the boiler
to shut down.
We have a water flow underfloor heating system and it works perfectly by the
method I've just described above. It also exceeds the manufacturers
requirement of heat dissipation by a few degrees using the open loop through
the hot water storage tank, and our tank is only a couple of yards away from
the boiler. The main test on our system, consisted of the engineer
by-passing the thermostat from the boiler and allowing the system to run
itself silly for about half an hour with only the unrestricted hot water
loop in play, no heating circuit.
I'm glad to say that the water returning to the boiler never got hotter than
83 degrees centigrade, which I'm now reading from the pass certificate from
Scottish Gas, so it never allowed the boiler to boil already boiling water,
but we did have clouds of steam blowing out through the expansion pipe of
the hot water storage tank and the floor did heat slightly, and only
slightly, under the first half of the kitchen floor. The system was given a
good pass certificate and has been running merrily and economically for many
This wouldn't comply on an unvented cylinder, though. These have to be zoned
off and capable of shutting down their heating circuit independently. As the
bypass is particularly likely to be used when there is an overheat
situation, using a circuit that can be independently shut down by the
cylinder's safety systems would not be a good idea.
Well, I did mine before Part L came into effect, but I thought it
conformed to the draft. I use the rads in the room with the thermostat
as the bypass loop. They have lockshield valves only, and are controlled
by the thermostat which cuts off the boiler when there's no call for
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It 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.