Heated windows [longish]

Unknown

So, I did some testing of a window I purchased [not from the company ME is involved with, ONLY because of shipping across the country] It was pretty interesting, and I have some graphs and such, see below.



Please take these as they are intended, an interesting evaluation by a disinterested party. I am pretty confident of their accuracy, I am very confident that the relative accuracy is very good.



I have an evil plan for my own application that makes it pretty likely that I will use these, the worst case being that I will not always run them.



I think these definitely have a place especially in difficult to heat rooms



K





<a href="http://members.verizon.net/%7Evze4wpr3/68degree.htm">[url=http://members.verizon.net/~vze4wpr3/68degree.htm]http://members.verizon.net/~vze4wpr3/68degree.htm</a>



vertical scale is watts per square foot, horizontal is 'outside'

temperature deg F



This first graph is the most extensive tests I did, and I have great

confidence in the results since some of the test periods ran over 50

hours. I would have left a kink at the left of the graph for the two

warmest temperatures, as this was pretty evident while watching this

that it was a straight linear function at the colder temperatures, but

flattened out as it got warmer. The spreadsheet is a little limited. I would have liked to run some colder

temperatures, but the freezer was not able to pull it any colder. It is

of course possible that the higher power usage numbers are actually

error, but I find it more likely that they represent heat loss to the

room becoming more of a factor. In any event, I don't think that it is

likely that anyone would be running the windows at temperatures any

higher than the mid 40's so it is pretty moot.



<a href="http://members.verizon.net/%7Evze4wpr3/78degree.htm">[url=http://members.verizon.net/~vze4wpr3/78degree.htm]http://members.verizon.net/~vze4wpr3/78degree.htm</a>



again, watts vs outside temp



Running the window at 10 degrees above room temperature. While it is a

limited data set it was again some pretty long run times. My electronic

thermometer died during the warmest run, so I chose not to include those

numbers, but they were consistent.



<a href="http://members.verizon.net/%7Evze4wpr3/Extendeddata.htm">[url=http://members.verizon.net/~vze4wpr3/Extendeddata.htm]http://members.verizon.net/~vze4wpr3/Extendeddata.htm</a>



This is my feeble attempt to integrate these two data sets and answer

the question



 "I want to run the windows at this temperature, and it gets this cold

out, how much power will they use?"



It is simply DeltaT[outside]+{deltaT[inside]*2}



Where DeltaT[outside]= Temperature difference from the heated glass

surface   to the cold side and   deltaT[inside]= Temperature difference from the heated glass surface  to the room. I have no reason to think this is accurate other than it looks good, and seems make sense intuitively.

Heated window test result conclusions







With the heatloss design temperature in my area of around 70 degrees F,

100 square ft of windows with an R value of 3[such as the tested window

without power applied] we would have a heatloss of 2310 BTU's, or 677 watts [100/3*70] according to my graph, with the glass surface at 68 degrees F, the

windows would draw about 730 watts, or about 107 percent of the energy

required by the unheated window. In  Massachusetts, with electricity

over 17 cents per kWH, this is not very attractive. This lead me to do a

few 'what if' calculations.



If we were to use a triple pane window, with an insulation value of

about R4, the heatloss would be about 1750 BTU's or 513 watts. Next

comes another broad assumption, but one that I think not to far from

accurate: I am going to assume that the center pane in the triple pane

window is thermally the center, and thus at a Delta 70 the center pane

would sit at 35 degrees. I recently had a good friend run an FEA that

confirms that this is a valid assumption, within reason.  If we go back

to the graph it shows us that we should use about 400 watts, about 78

percent of the energy required by the unheated window.  It should be

noted that my setup could not exceed 9 watts per square foot, due to the

contacts being located on the wide side of a nearly 6 foot pane.



This number starts looking attractive. Although it would take heating

oil at $4.60 a gallon[at 85 percent efficiency] to be literally 'saving'

money, we are now in a place that we can discuss the true comfort of not

having cold windows.



Interestingly, going to quad pane does not get us closer. While it does

use less electricity, percentage wise it is not as good as the triple

pane. This is due to the flattening of the graph at the warmer outside

temperatures. It is my feeling that this flattening is due to some

heatloss to the room, which could be appearing as a  constant overlaying

on the  graph. Simply, the warmest spots of the window may be radiating,

for instance 1 watt per square foot to the room, no ,matter the outside

temperature, which is not noticeable when the heat loss to the outside

is 6 or 7 watts per square foot, but down below 4 watts is starts

becoming a significant source of consumption.This flattening could also

be error[1], due to the average temp of the window drifting higher at

the warmer outside temps, or the controller wasting energy trying to

keep the temperature too close to the setpoint.

Whatever the reason, most of the reasons I can think of end up being

heatloss to the room, which is generally a 'good thing' although it can

be expensive. This is also reinforced by the much higher electricity

consumption a the higher window temperature test.



With the tested window at 78 degrees F, a Delta T of 70 is going to use

about 1200 watts for that 100 square feet. In a well insulated and tight

house, one would think you would not run this temperature[or higher]

without warming the room enough for the window to cycle on and off



[1]



 I could also have incorrectly calculated the temperature on the warmer

end, since I had to use a thermostat with a deadband that the colder

temperatures[with the freezer running flat out] did not have. This last

cannot account for all of the difference, since it would require a shift

to the colder of greater than the possible error.

k_2

re

Thought there would be more interest, probably everyone fell asleep before reaching the end of the post.



I got some fancy color pics if you can see powerpoint files:





members.verizon.net/~vze4wpr3/Window-Report.ppt



A bunch of WAG's in this also, I don't even know what most of it means, but the one that shows the temps of the surfaces[page 10] is what I was after.



Lots of pretty colors tho............



Finally got my password reset, not until convincing Dan that it was my first day on the web...........

Mark Eatherton

I suspect that you and I are the only beings interested in this....

Thanks for the posting Keith.



We are still awaiting the ship to come into port, and in the mean time are continuing doing our own R&D individually.



One of your findings that echo's our findings is that the lower the glass surface temperature, the higher the operating efficiency, I am also finding that the glass really does not have to be that hot in order to significantly increase human comfort. In fact, anything over 85 degrees F gets UN comfortable over a period of time. This is what we call the thermally opaque mode (temps less than 85 but above 70). In other words, the glass surface temp is kept at a temperature that the human bodies skin can't feel it's presence.



Another interesting application is in the elimination of the production of condensation on windows. All we have to do is maintain a surface temperature of 70 degrees F, and we can guarantee NO condensation will form on the glass. This will significantly reduce the physical plant requirements in areas like hospitals and natatoriums where high humidity is fought by moving large volumes of air. The amount of energy consumed to maintain 70 degree glass surface temperatures is significantly less than the parasitic cost of operation of the blowers and fans associated with conventional means currently being employed.



Thanks for the info.



ME