Dave Baker wrote:

In the event of a control failure, it is a safety device; it releases
steam, prevents the system becoming over-pressurized and bursting. In
your system, such a failure would probably now cause the contents of
the heating system to be discharged into the loft and the boiler would
dry-fire to destruction.

Nonsense. The system isn't pressurised other than by the head of water from
the loft tank. The contents of the system can't be discharged into the loft
other than by the same pipe from the bottom of the loft tank which is
continuously filling it up anyway.

You and Drivel have misunderstood what I have written. It is not
nonsense, it is common sense.

If the boiler thermostat fails in the ON position, the temperature of
the water will quite likely exceed 100 degC. The water in the boiler
will boil, producing steam. This happens. Your heating system will THEN
be pressurized by the steam and the steam pressure will become greater
than the static pressure provided by the head of water in the F&E tank.


The water in the heating system will be pushed up the cold feed pipe
into the F&E tank. It will be pushed out rapidly, probably faster than
the F&E overflow pipe can remove it. There will be no water in the
boiler and it will continue firing.

This assumes that the cold feed pipe is clear and that there is not a
stopcock (acting as a non return valve) fitted to it. The cold feed on
my heating system was totally blocked with limescale when I bought the
house. This is a common fault. I had to cut out the tee and replace it.
There had been a longstanding leak from the heating system through the
coil in the indirect cylinder.

With a non-fubarred system, such a failure would cause steam to be
blown out of the open vent, but cold make-up water would enter the
system through the cold feed. This could continue indefinitely.

You can't get 2-way flow (steam up & water down) in one pipe, which is
probably what Faber & Kell were referring to. Their book (20 years
since I had a copy) deals with commercial installations on which such a
combined CF &OV would be inadvisable.