wrote:

OK. Lets see if we can find some things we can agree on. I dont know
what of the below we'll agree on, but whatever it is should make a
start.

I'll address what you wrote as a separate reply.


1. Fitting a dimmer normally results in a changed pattern of lighting
use (if it doesnt, there wasnt much point fitting it)

Agreed

At its simplest level it means the brightness will be turned down some
of the time....

2. This means that in real life we can name 2 situations, the Before
situation and the After situation. These are the lighting use patterns
for each user before and after fitting a dimmer.

Yup

3. In real life the fitting of a dimmer can not be assumed to result
in no other change than using it. Why? Some users fit a dimmer as they
previously switched some lights off to reduce brightness, and think a
dimmer would be easier, look nicer or offer a wider range of control.
Some users never dimmed before but will afterwards. (There are other
options as well, but another day).

Indeed true, but missing my point really. All I was saying is that
fitting a dimmer in the absence of any other changes - i.e. same bulb,
same on time, but now some of the on time will be at reduced brightness,
will not result in increased energy usage.

For many users of dimmers that I have observed this is a typical mode of
use. Not all certainly, but it would be misleading to imply that costs
will rise if you fit a dimmer and this is your usage profile.

For example, I have a remote control touch dimmer in my office - almost
all the time it is on full brightness. It allows me to turn the lights
on when it starts getting hard to see without interrupting what I am
doing, and it lets me dim the lights if I decide to take a break and
watch a movie on the computer.

4. Given the above, and a few more factors I'll skip for now, we could
not realistically conclude that if a dimmer is fitted, the change will
always be from a given wattage of undimmed lights to the same wattage
of lights now sometimes dimmed. Of course this happens in a percentage
of cases, but clearly not all.

Indeed. And my comments applied just to this one case.

5. If we're going to determine whether fitting a dimmer will reduce
one's energy consumption, we need to compare the real world lighting
use situations before and after, not just a single pair of possible
and sometimes so situations.

The figures we have so far illustrate:

Filament, Full Power, No Control
Filament, Full Power, Dimmer
Filament, Optimal Power, No Control
Filament, Switch Bank

What others would you like?

We could add "optimal power, dimmer" to those which would show a very
slightly lower cost than the optimal power no control figures.

The situation you frequently mention(i.e. fitting a bigger bulb then
dimming) is covered by the Full Power, Dimmer option. This shows a
slight energy saving over having the same bulb and just a switch, but it
is also contrasted as 50% more expensive than the switchbank solution.

6. Now, if we can agree that far, then we can look at the various
before and after options that often happen in the real world, and see
where each takes us.


Before 1: Single set of filament lighting of X watts on one switch. No
plug-in lights, no switchbank, no other means of reducing lighting
level ever used, and only filament lamps are ever used.

After 1a: Single circuit of filament lighting of X watts, sometimes
dimmed.
Result: Some energy saving.

Yup agreed (in our figures already)

After 1b: As 1a, but with the same brightness achieved by increasing
bulb wattage and reducing the dimmer setting a little. Why would
anyone do this? To enable occasional use of higher brightness, for
tasks such as:
- more thorough cleaning
- seeing clearly behind the computer, tv, or whatever else may get
things plugged in the back at times etc.
- better illumination for various other purposes that people sometimes
have
Result: more energy used

Agreed (in our figures already)

After 1c: User wants to save energy as well as reduce brightness, and
chose a dimmer without realising the benefits of other options. User
would have bought a couple of plugin lamps and used those for lower
lighting, but due to lack of knowledge went for the dimmer.

No necessarily lack of knowledge... they may just not like table lamps etc.

Result: Less energy use than previously, but more energy use than the
plugin option that sometimes would have been chosen if the knowledge
had been there.

Agreed - with the above proviso. You may accept the higher cost in
exchange for the simplicity and lack of clutter.

(personally I am not so keen on table lamps because inevitably you end
up looking at them which is not always comfortable - I can never deduce
the logic of people who stick a table lamp on or beside a TV for example
- you spend the whole trying to watch a program with a bright light in
your peripheral vision).

After 1d: A year later user decides to go over to CFLs, for any of a
few possible reasons. Dimmer prevents this, thus while it may or may
not have saved some energy initially, it now results in 3x as much
energy use as would occur with no dimmer.

Seems like a non issue to me. You just replace the switch.

However this point is included in the article already.

There is also another Before possibility, that the user used to leave
a kitchen light on to light lounge while in lounge, but now uses
dimmed lounge lighting.
Result: more energy consumption.

Depends on what the kitchen light is presumably - if the kitchen has
400W of halogen downlighters then it could equally save money.

There does not seem to be much point in providing figures for obscure
uses like this since there is an infinite number of them.

Another real world Before situation we have to consider, and its a
common one:

Before 2: Lighting on 2 or more switches, which are sometimes switched
so as to produce dimmer illumination. This is common in long rooms,
domestic corridors and many other rooms, where lighting is not
intentionally designed for dimming by switchbank, but partial lighting
area overlap means that this basic approach to reduced lighting level
is sometimes used.

In this situation, fitting and use of dimmers will in most cases
increase energy use. I assume I dont need to go over each possible
After situation to show this.

If there is little or no overlap in regions then they would typically
have to use both, so the dimmed solution will cost slight less. Only if
they switch from one light to two dimmed lights would it increase.

I think once one has made the point that running dimmed lights will cost
significantly more than switching some off, all these cases can be left
for the reader to work out for themselves.

Before 3: Plug-in lights as well as fitted lights, with the
combination used giving more than one brightness level.

The result of fitting a dimmer with this one will be the same as the
above 'Before 2', in the majority of cases more energy use.

Again this is just conjecture. The logic may equally be - "oh the top
light is brighter than we need with the table lamps on", fit a dimmer.
This will save a little, but less than sticking a smaller bulb in the
top lamp.

Before 4: a set of kitchen lighting was too bright for much of the
time, but all on one switch. So the lights were left with some on and
some individually switched off (or with dead bulbs in). One encounters
this where under cabinet worktop lighting has been used to provide
most of the light in a kitchen, or anywhere where the lighting
installer has gotten a bit too enthusiastic.

User fits a dimmer and now runs all the lights on but dimmed a bit.
Result: more energy use.

We already cover this situation in the figures. Full Power, Dimmed, Vs
Full Power Switchbank (in fact we need to include the "full power" text
in the table for this one).

Etc etc. The conclusion we can draw is that while dimmers sometimes
reduce energy use, they also sometimes result in more energy use.

Yup agree fully.

So why the difficulty in highlighting the cases where they will save
money? Since we also cover the cases where they won't.


All of these options occur in the real world. There is no way one
could realistically conclude that fitting a dimmer will consistently
save energy. The problem with some of what has been written before is

I never said it did. I seem to recall that the para you wanted to
"enhance" with talk of triacs and negligible efficiency etc was simply
one that said (paraphrasing), "if you use the lights on full power most
of the time, neither dimming or switching will save much, but you may
get a slight advantage on a dimmer with longer bulb life". (paraphrasing)

that it strongly appears to imply that a dimmer will save energy,
which is often not true.

Well the figures already show the dimmer costing 50% more than the other
filament options (with exception of running full power lights flat out
on a switch).

I can see one possible way forward with this. That is to remove all
mention of energy efficiency from most of the article, put it under
its own 'Energy Efficiency' heading, and mention all the possible
scenarios above, so the end user knows that in some cases they will
save energy, and in some use more.


Are we finding some areas of agreement?

I think we agree on pretty much all of it. I just had problem with the
words "Dimmers at full brightness have a slight energy efficiency
disadvantage since the triac voltage drop will run the lamps at very
slightly lower rms voltage and efficiency, but the amount of the effect
is trivial." - given this was in the section that discussed using lamps
on full brightness most of the time.

It is a very complicated way of not making any real sense. Using phrases
like "energy efficiency disadvantage" would to most people I expect be
interpreted costs more, rather than "costs a little bit less unless you
compensate for the reduction in bulb brightness by using a bigger lamp
in which case it will cost a fair bit more"

The other point I wanted to introduce somewhere was the potential
benefit of soft starting on halogens since this can represent a
worthwhile increase in lamp life (and hence life cycle energy costs)

--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/