Heat/Efficency of Speakers

What % of power sent to the speakers is turned to waste heat? That's the short version of my question.

I'm looking to minimize waste heat accross my stereo as my listening room is unforgiving come summer; no cooling and a computer system which cannot be relocated. I understand amplifier efficency & the classes as well as speaker efficency measured as W/db however the interplay eludes me.

Taking two hypothetical amplifiers: a Class A amplifier outputting 10W w/ 100W from the wall & a Class D outputting 200 w/ 220W draw I understand the D will be the cooler operator however this is where the discussion tends to end, D only wasting 20W vs the A amplifier's 90W. Considering appropriate speaker matches to each amp(as well as a standard HE speaker at say 95db/w), how do I determine the wattage converted sound and the watts spent as heat?

I'm asking because I was previously running a 10W tube amplifier in this room(4xel84 tubes) with 96db speakers. This was bearable in two hour doses this last summer. My friend assures me any Class D amplifier and many AB amps would have no such heating problems and says it's class not wattage that is my issue. Before I move to a different amplifier technology(and swap speakers, these voiced for SE tube partnering) I want to understand this issue fully. I'm unconcerned with power usage and only care about the heat.
Not a scientific response from me by any means but frankly I would not worry about how much thermal energy your speakers put out. Grab a sensitive measuring device and measure the air a few feet away from the speakers and right next to the speakers; I'd bet there is no meaningful variance in temperature (i.e. higher) right next to the speaker. You have to measure close in I would guess so you don't confuse the effect in that immediate area from your amp. If you are worried about speaker-produced heat, you should also be questioning how much heat your components and rack throw out as most isolation racks & platforms work on the principle of harnessing kinetic energy (vibration) and channeling it away as thermal...

The biggest concern is heat energy from tube and solid state gear, or hybrid (solid state with tube input and/or output stages, power supplies, etc...). In general class A and tubes of any kind are the worst with Class AB and ICE or Class D being the best in terms of producing little heat. The person who mentioned 'idling' characteristics is spot on...

Playing devil's advocate for a minute; if you have sound you like from gear and speakers that you like, instead of ditching the gear, taking a loss on the used market sale then buying-in again all from the beginning (speakers, amps, etc...), why not put a smaller some into a separate "split" heat-pump unit (so heat as well as A/C) for that room only. A split is a unit with a very quiet unit mounted in-wall with no duct work and the compressor/heat exchanger mounted outside on a concrete pad....just a suggestion.
Maybe I did not make my point clear.  Everything coming out of the wall ends up as heat.  Reduce current draw and you reduce heat.  Forget the speakers,  They are irrelevant.  They do not create energy, they only take the energy that is sent to them and convert it directly to heat or indirectly to heat by first creating sound waves, which then strike surfaces in rooms and is quickly converted to heat.  Energy enters the room as light, current, and body heat.  All of the energy in the room is conserved.  It does not disappear.  Reduce current draw and light entering the room and you have done what you can do.  As has been pointed out several times in this thread, a class D amp is about as good as you can do in reducing current draw.  Al, help me out here.  Am I missing something?  Isn't this just simple first and second law stuff?
I wish more manufacturers listed power draw, last night I found a few SET amp makers who listed this and I've been severely undervaluing it. 100W seems the minimum for any SE tube amp. Ok, I was assuming 10% efficency at worst hence my thought some class A was competitive. That helps put into perspective the other classes, I understand why some were frustrated with my arguments for low-power tubes.

I'd also like to clarify my dogmatic approach is about keeping the discussion as simple as possible, I won't be picking an amp based on its efficency but I will be counting out amplifiers which would prohibit summer listening moreso than my former SE Parallel EL84.

We have had a consultant look at the room before and were told it's not a candidate for AC. I do have an industrial ceiling fan down the hall but does neccitate listening breaks, it's loud. I ran smaller fans throughout the summer to blow equipment heat out the door and I couldn't tell you if it made a difference.

I believe I have the answers I set out to find, and am thankful for the posters above. Suppose next priority is finding amps which play best lower in their power band, and picking out new speakers without obsessing too much over db rating.
Hi Bill (Brownsfan),

I mostly agree with your points, as far as they go.  But regarding the relevancy of the speakers, I would add a couple of things.

First, for a given listening volume a given class AB or class D amp will draw less current from the wall if higher efficiency speakers are used than if lower efficiency speakers are used, resulting in less heat being introduced into the room.  Second, as Swampwalker noted greater speaker efficiency would allow the use of a less powerful amp, which within a given bias class will tend to correlate (at least loosely) with reduced power consumption.

Also, regarding "All of the energy in the room is conserved.  It does not disappear," some of it will in a sense disappear.  Although it is pretty much just an academic point some of it will disappear **from the room,** assuming the temperature of the room is higher than the temperature of the adjacent spaces (both indoor and outdoor), given that the thermal insulation between the room and those spaces will not be perfect.

Best regards,
-- Al
Redfuneral 12-12-2016
... my former SE Parallel EL84.
OK, then my previous supposition that the amp was biased in class AB was not correct, as a SE parallel amp will operate in class A.  So the 60 to 75 watt estimate I stated for the power consumption of the amp under typical listening conditions was probably somewhat low.

-- Al