The message
from The Natural Philosopher contains these words:

snip

The more concentrated the heat and the higher temp the heat the more
there is a chance of excess loss over simple steady state analysis. what
about ultra hot feed pipes in the loft? sure they are insulated, but the
hotter they are, the more they lose..

The better your insulation the more likely you are to get little benefit
from time switching and actually stray into excess transient losses.

But the better the insulation the less the heating system has to do to
restore the status quo. Scalding radiators are not needed.

Scalding radiators happen when any system that is not modulated, goes
into heat mode.

It doesn;t matter whethert tu re one degree or 100 degrees below TRV
temp, thermostat temp. The rads and the boiler will be working full on.

What modulates the radiator surface temperature is in fact the time
they are on 'heat' because they have thermal mass.

Thats teh whole point.

I don't have any figures to hand but I would have thought that the
thermal mass of the radiator (as opposed to the water within it) would
be negligible. The temperature of the radiators does take time to get up
to full working temperature but not so long that the set temperature
isn't reached on a normal firing cycle. Given a 10 degree drop across
the boiler it is going to take several circuits of the heating loop to
get the boiler output up to the boiler set temperature unless the set
temperature is very low.

If you had e.g. a house made of polystrene with hot air heating,
ultimately probably te lowest theral mas possible, then I would almost
agree with your perspective.

But the more mass there is in the building, or the heating system, the
more a high heat regime will tend to overheat parts of it with respect
to other parts. And leave the possibility of higher heat loss from those
parts during the transient phase.

Surely a passive high mass acts as a damper smoothing out the overshoot.
I am not sure passive is the right word but I am trying to distinguish
between a house with plenty of solid masonry in its internal walls and a
concrete slab with underfloor heating below which is really a radiator
needing serious control gear to prevent a significant overshoot, but
even there you have a large time constant in the equation.


It may not be
entirely impossible to engineer a heating system in an extremely well
insulated house to use more energy if it is switched off for a lengthy
period but, other things being equal, the only way that is going to
happen is for a single firing of the boiler to produce a larger
temperature overshoot on the heating phase than the temperature drop on
the non heating phase. ISTM that that is much more likely in normal
operation than it is when the temperature drop is significant.

Of course deliberately buggering up the operation of a condensing boiler
would make the task easier but even the 10% loss of efficiency may not
be enough even if the set-up could be designed to give maximum
efficiency in normal operation and maximum inefficiency in sustained
firing.


Well the figures and experiences of people in high insulated houses tend
to bear out the proposition that it makes very little difference on a
modern house whether its run 24x7 or timed.

If you go back to my very simplified model there is no difference. Don't
forget that this exchange grew from Dennis' claim that turning the
heating off for 8 hours would save very nearly one third of daily fuel
use.

The worse the insulation and the lower the thermal mass, the more the
timed solution saves you. Simply because it both loses far more heat
(and gets icy cold when the heating is off) and because its not got much
to heat up mass wise, so it comes up to temp quickly.

As I see it the thermal mass isn't much of an issue being a function
both of the heat required to keep the house up to temperature when the
heating is on constantly and of the heat required to bring the house
back up to the set temperature when the heating is off for an extended
period. What is true is that a poorly insulated house will leak heat
much more quickly than a well insulated one and thus save more of the
heat that would have leaked out had the heating remained on (IYSWIM).

Houses that keep heat in them well, don't benefit much, and as I have
been trying to say, the heatloss from parts of the system that dont heat
the house as such at all, can be a lot higher when their average
temperature is now at 65C-70C for prolonged periods, I.e heating pipes
in the loft or walls. If you have to use high boiler temps to get the
house to heat up, you WILL be losing more heat during that phase.

A pipe, even in a couple of inches of foam, is nowhere near as well
insulated as the rest of the house should be.

Pipes per se give out very little heat compared with radiators even when
they are not lagged.

This is not a simple analysis.

Boilers modulate as well, because it is both better to run for longer
periods than to run at lower power levels. The reasons are simple: high
power means less efficiency as the condensers - if fitted - work less
well, and also the alternative to modulation - cycling - leads to start
up losses when the boiler fires.

I would expect only minor differences in boiler efficiency if the same
system was in use in both timed and constant mode which would be dwarfed
by the difference in heat losses.

And if you use the most efficient heater there is - a heat pump - the
efficiency severely degrades when called for high output temperatures.

I m not saying that its always better to run a system this way or that
way, just trying to point out that the simple analysis runs into trouble
the more modern the house is.

There is a law of diminishing returns. And sometimes the transient
behaviour of the system makes it a net loss.

But the more modern the house the more sophisticated the control system
has to be. I do not believe for one moment there is a single centrally
heated house in the country where running the heating 24 hours a day
results in less fuel usage than having the heating off for a solid 8
hours in 24 and very few where it would be possible to even engineer
such an outcome.

--
Roger Chapman