Solar Inverter Sun Screen

[Anorak Bob]

A few years ago we had a solar energy system installed, 16 Schott panels and an SMA Sunny Boy inverter. The sensible location for the inverter would have been inside the laundry but it would have had to be positioned up near the 3.2 metre high ceiling. A bit awkward to read the display but in hindsight......... The installer recommended mounting the inverter on the external south east facing wall near the existing meter boxes. Not realising the effect sunlight had on the Sunny Boy's LCD screen, I went along with the suggestion. I should point out that it would have cost more for the laundry installation and as it was the 3KW setup was expensive.

Last week I thought I would have a look at the display and to my disappointment I could hardly read the thing. The screen had become cloudy and the plastic had started to craze. To halt the deterioration I had to devise a sun screen. The inverter generates heat so it couldn't be an enclosure. I had a sheet of 2mm aluminium measuring roughly 900 x 600 that I thought would do if I could mount it in front of the inverter and maintain space for ventilation.

The sheet needed to be readily removable. A version of a keyhole mount seemed easiest but cutting four 1" holes in the sheet with a reasonable level of accuracy presented some problems. I thought I could simply drill the holes. A trial run with some 3mm thick aluminium proved that I couldn't expect much in the way of accuracy. I tried a hole saw and the saw found a way of exploiting the drill's forty years of wear. I could have possibly filed the holes but they probably wouldn't have ended up round or accurately located.

So I ended up using the mill and some ex Greg Q's helix clamps and the ex GQ scope to centre the 2.5 mm holes drilled previously that align with the holes in the 40 x 6 flat serving as a mount for the round bars. All worked better than expected.

My wife has concerns that the aluminium sheet will act as a heat sink and aid in the cooking of the inverter. I am concerned too. I could affix a sheet of polystyrene to the rear face of the aluminium but how effective it would be I have no idea.

Any other ideas?

BT

[RayG]

Hi BT,

Nice job. I like the SMA inverters, good German quality design and well made, probably the pick of the inverters out there. You could probably get replacement lexan front panels as a spare part if you wanted.

One of our big selling designs is (was) a solar powered irrigation flow meter, these days it is made in Sydney, we made the first 3000. last I heard they had sold 40 million $ worth, ( wish I could get royalties on that. :) ) Anyway our solution to the UV degrading the lexan panel was to put a stainless steel hinged flap over the panel. Without that cover the lexan would start to deteriorate with a few short years.

If you are worried about heating, just get out the IR thermometer and measure the enclosure temperature with, and without the shield, I think you will find the shield will shade the inverter from direct sunlight and keep the enclosure much cooler. You have enough of a gap that the heating of the shield will create some convective air flow past the inverter that will help keep it cooler than shading alone would. ( very clever design ).

Ray

[C-47]

Bob,
I think you have done an excellent job on providing a thermal barrier and there's plenty of airflow space around the unit and heat rising would tend to create thermal flow anyway and also the Freo doctor would provide some afternoon cooling breeze. If you wanted to really go first class stick some sheep wool (insulation) on face of the plate, they don't overheat :D
Alan

[Oldneweng]

Why did the inverter have to go up near the ceiling? My brother had this problem with his installment. He was told that was how they did it from memory. He suggested that if they wanted to be paid for their work then they would put it where he wanted it, as the customer, and where it had been planned to be placed. They agreed to put it where he wanted.

We have a couple of PLC's at work with your problem. A stainless cover just bigger than the screen was made and hinged in position. It had the sides folded in to hold the face of the cover about 10mm away from the screen.

Dean

[Stustoys]

Mirror finish the Aluminum :D (at least on the "inside")
But as others have said I doubt you've made things worse.

Stuart

[Anorak Bob]

Why did the inverter have to go up near the ceiling? My brother had this problem with his installment. He was told that was how they did it from memory. He suggested that if they wanted to be paid for their work then they would put it where he wanted it, as the customer, and where it had been planned to be placed. They agreed to put it where he wanted.

We have a couple of PLC's at work with your problem. A stainless cover just bigger than the screen was made and hinged in position. It had the sides folded in to hold the face of the cover about 10mm away from the screen.

Dean

Dean,

There is a window at the end of the laundry, on one side there are 2400 high cupboards, and on the other side, where the inverter could have gone, there is a clothes dryer mounted above the washing machine. The dryer which is used maybe twice annually, would have needed to be externally ducted. Big deal.

Where it is mounted on the wall, I still have to stand on a milk crate to view the screen. :rolleyes:

All in all it doesn't sound like I'm going the bake the inverter. Thankfully.

And thank you Ray and Alan for your comments and approval.

Bob

[Anorak Bob]

Mirror finish the Aluminum :D (at least on the "inside")
But as others have said I doubt you've made things worse.

Stuart

Stu,

You replied while I was typing with one finger.

The aluminium is shiny on the inside. It wore a protective plastic sheet.

And you are right about parallax with our scopes. It was awkward positioning my eyeball yesterday while peering through the Isoma. The image would shift slightly while I struggled to get into a comfortable position. I couldn't.

BT

[Stustoys]

You replied while I was typing with one finger.

I'm back to six finger typing :) no more thumb typing for awhile.



The aluminium is shiny on the inside. It wore a protective plastic sheet.

Well the only way heat can get from the sun to the inverter from the Aluminium plate is Radiation right?(no conduction, no convection). Polished Aluminium radiates 2/10ths of stuff all... right?
Only problem I could see is if the inverter shed heat by radiation... you're now reflecting it. but I doubt it does??



And you are right about parallax with our scopes. It was awkward positioning my eyeball yesterday while peering through the Isoma. The image would shift slightly while I struggled to get into a comfortable position. I couldn't.
So the Isoma does it as well? thats a little sad. I should get a chance to play over the next couple of days. As I said, might be an easy fix, might not.

Stuart

[jhovel]

Hi Bob.
Be glad your inverter is outside. They make a very intolerable sound which changes with the brightness of the sun as clouds go past. I wanted mine (also SMA) installed inside and was strongly advised against it on the day. I'm glad I followed the advice!

[BobL]

Very neat job Bob.

Two other ways to drill neat holes in thin ally sheet are a Forstner bit or one of those sheet metal stepped drills.

The thing doing the damage to the plastic is almost certainly the UV so a sheet of just about anything transparent over the top would work and not need to be removable.
Something like a piece of polycabonate would do, It would yellow over time but it would remain transparent.

That aside LCDs can be affected by heat so the Ally is probably a better direct sun shield although I'd bet the heat from the unit itself is not inconsiderable, and the radiation off those bricks prolongs the heat exposure after the sun goes down.

BTW I would consider painting the ally sheet white to reduce reradiating to the other side.

[rodm]

Hi Bob,
Looking at the pictures the perspex looks to me like it has a protective plastic film on it. If so pull it off and things should be a lot clearer on the lcd. As you know I have plenty of acrylic so yell out if I can help.

My solar inverter bit the dust some time in the last two months and only discovered it on Monday. Just when our power usage is high as the air conditioner has been running quite a bit. Fortunately it is covered by warranty but it will take them 3 weeks to get here.

[Anorak Bob]

Hi Bob.
Be glad your inverter is outside. They make a very intolerable sound which changes with the brightness of the sun as clouds go past. I wanted mine (also SMA) installed inside and was strongly advised against it on the day. I'm glad I followed the advice!

Hello Joe,

Most people I know that have inverters have them located out of hearing range in an uninhabited space such as a carport or garage. My carport is on the other side of the house and if the laundry was inconvenient for the installers, the carport would have been more so. I didn't know about the noise, a yellowed LCD is a small price to pay for the lack of irritation some background noise would have certainly caused me. The older I get the more those noises wind me up.


Very neat job Bob.

Two other ways to drill neat holes in thin ally sheet are a Forstner bit or one of those sheet metal stepped drills.

The thing doing the damage to the plastic is almost certainly the UV so a sheet of just about anything transparent over the top would work and not need to be removable.
Something like a piece of polycabonate would do, It would yellow over time but it would remain transparent.

That aside LCDs can be affected by heat so the Ally is probably a better direct sun shield although I'd bet the heat from the unit itself is not inconsiderable, and the radiation off those bricks prolongs the heat exposure after the sun goes down.

BTW I would consider painting the ally sheet white to reduce reradiating to the other side.

Thanks Bob,

Had I not looked at the display last week I would have probably left it until there was nothing visible. It is a bit of a pain in the neck to get to if a milk crate or ladder isn't handy Realising I was running out of time I had to come up with the simplest means of keeping the sun off the inverter. Initially I carved up a large Nylex crate so that it would act as a cover. I'd fashioned a covered vent in the front ( the bottom of the crate ) and it wasn't until I stood on a ladder to work out how I would fix the crate back to the wall that I discovered how much heat the inverter generated. A cover would have cooked the thing.

The sheet of aluminium had been in the shed for 21 years so I was glad to have finally found a use for it. Surprisingly the plastic protection simply pealed off. The keyhole-ish mount was used rather than hinges or screws because it was the easiest thing to produce with the gear I had. I haven't machined much aluminium before, the stuff is like butter. I might acquire a taste for it.:)

I have some white acrylic and the sheet's non polished face is slightly oxidised. Should be a good tooth for some paint.

Bob.

[Anorak Bob]

Hi Bob,
Looking at the pictures the perspex looks to me like it has a protective plastic film on it. If so pull it off and things should be a lot clearer on the lcd. As you know I have plenty of acrylic so yell out if I can help.

My solar inverter bit the dust some time in the last two months and only discovered it on Monday. Just when our power usage is high as the air conditioner has been running quite a bit. Fortunately it is covered by warranty but it will take them 3 weeks to get here.

I had better have a close look in the morning young Rod.

The bloke I work for, the one who convinced me to drop the what now seems excessive amount for our installation told me a somewhat similar story the other day but his story had a happier ending. He had noticed a massive jump in his electricity account. He ventured down into his basement to find out his inverter was not working. A plumber had turned it off 3 months earlier.

Bob.

OK I checked the screen this morning. What looks like an additional layer in the photo isn't. The plastic is cracked around the edges. I sent SMA an email yesterday querying the availability of a replacement screen. No response so far.

[Neil317]

Polish the "inside" surface? Nup. Paint it flat black, stop it radiating the heat towards the inverter. Black bodies and all that.
Mirror polish the outside, but birds might attack it.
Our Mudlarks are really agro latelyhttp://d1r5wj36adg1sk.cloudfront.net/images/smilies/standard/biggrin.gif
Neil

[BobL]

[QUOTE=Neil317;1845254]Polish the "inside" surface? Nup. Paint it flat black, stop it radiating the heat towards the inverter. Black bodies and all that.
Mirror polish the outside,[ /QUOTE]

Sorry that's incorrect on both counts.

Black absorbs and radiates the most.
When bright and shiny, metal surface are the best reflectors of heat but within weeks of exposure to the atmosphere the oxide layer on the surface (it can be so thin that it's invisible) will be a poorer reflector than a white surface. Metal oxides are also very poor emitters in the infrared so any heat that is absorbed is retained.
This is why galvanised roofs get so hot and are a poor choice for roof colour.
Zincalume is supposed to be better but I noticed no difference when we had our roof re-clad.
Aluminium is better that galv or Znalum but a white painted surface will be better still
Gold is a very good reflector because it does not oxidise anywhere near as quickly.

[Stustoys]

[QUOTE=Neil317;1845254] BTW I would consider painting the ally sheet white to reduce reradiating to the other side..
My thinking was Inside surface. Wasn't thinking reflector, thinking low emitter. The invertor doesn't care how hot the aluminium gets if its not emitting IR. Now without looking it up :B I'm pretty sure paint is a pretty good emitter. So maybe paint the outside white and leave/polish the inside? Granted I could be over thinking things again.:U

Stuart

[Neil317]

Whoa! I was 100% wrong! http://d1r5wj36adg1sk.cloudfront.net/images/smilies/actions/doh.gif
Thanks for putting me on track Bob.
Neil

[RustyArc]

[QUOTE=BobL;1845263]
My thinking was Inside surface. Wasn't thinking reflector, thinking low emitter.

From what I understood of Bobl's post, the easiest solution is to simply plate both sides of the sheet with gold. Good IR reflection and low emissivity.

[RayG]

I wouldn't paint it. (although gold plating would look good :) ) The function it has to perform is to block UV, if it shades the inverter and keeps it cooler, that's just a bonus side effect.

In any case the emmissivity of Aluminium, ( whether oxidised or not is very low) , much lower than white paint... typically less than 0.2, whereas white paint is 0.8 or thereabouts.

Ray

[Ropetangler]

I wonder if painting it any colour would make much difference. If polished or painted white say, the aluminium plate absorbs less heat from the suns radiation, and conversely if painted matt black say it will absorb more of the suns radiant energy, but a hot plate will begin to heat the surrounding air, which will form a convection current just as a chimney does, and that rising air will also draw in extra air ( the Bernouli principal) which will tend to cool both the inverter and the sunshade.
One could improve the cooling effect by increasing the length and making it in the form of a hollow tube perhaps, and even just painting black the upper section above the inverter would improve the draft, without any re radiated heat going the inverters way. I'm not sure just when the Local Government Authority would expect you to submit a D.A. though:o, and I don't know the best way to keep the swallows out either.:U
Great work as usual Bob by the way, and plenty of interesting thoughts from all contributors too.

[Anorak Bob]

I received this response today -

Hi Robert

We can sell a new top lid which has the clear Perspex for $140.75+GST+shipping. I suppose
this is purely aesthetic. Alternatively, you can download Sunny Explorer and monitor your
system with a laptop or PC which has Bluetooth connectivity. The software is very easy to
use, and provides more information and beyond of what you see through the screen.

The link to Sunny Explorer is here:
www.sma-australia.com.au/products/monitoring-control/sunny-explorer.html (http://www.sma-australia.com.au/products/monitoring-control/sunny-explorer.html)

As much as we are able to sell you a new lid, Sunny Explorer offers the monitoring you
need for your inverter system if that's what you ultimately need. That said, if you wish
to purchase a new lid, call us on 02 9491 4200 to process an order and pay by credit card.

Kind Regards,

Cyril Laugeon
Technical Service Engineer
SMA Solar Technology


time for me to embrace Bluetooth http://d1r5wj36adg1sk.cloudfront.net/images/smilies/smile.gif

[Ropetangler]

The Bluetooth solution looks good Bob, but could you mill out the degraded perspex and Sikaflex in a small piece of plate glass in your present inverter lid, because a replacement seems pretty expensive for what you get. The free bluetooth app is nice, but it seems strange to me that it runs on a P3 windows machine running XP with sp2, but not any Linux or MacOS machines. One good thing I guess is that a P3 with XP sp2 would be pretty cheap at a salvage centre, if you were unable to source one from the roadside on a hard rubbish day:wink:

[Theberylbloke]

Hi Bob,

Nice job on the sunscreen.

LCD displays can also degrade from heat to the point where they go black and can't be read. I have observed this in the first generation of electronic electricity meters which are now about 20 years old. These are not exposed to UV being in switchboards. They also don't generate much heat, rather, it is the environment (welcome to the tropics) and manufacturing tolerances. Some go black faster than others even though the date of manufacture is similar.

Good to see SMA have spares available. About what you would expect from a premium brand. Schneider (spelling?) are also pretty good. The noisiest inverters around at the moment seem to be Solax which have a fan that runs up as the inverter powers up. Probably this is part of the power on diagnostics as it then switches off the fan before the inverter goes online. The Solax are setup for battery useage (I think, not too sure), so may have extra circuits to cool. The majority of inverters I have been around (likely a couple thousand by now) seem pretty quiet in use.

With regards to painting your sunshade? I don't see the point for a number of reasons. A roof exposed to the suns heat will reach an equilibrium temperature with it's environment. That temperature will be same no matter what the material the roof is made from nor the colour the roof is painted. Lighter colours will probably take longer to get to the same temperature as a dark colour, but it will be the same temperature none the less. A more emmisive roof will radiate heat back into the air making the air adjacent to the roof hot which then radiates/conducts the heat back into the roof. Your roof and mine is exposed to a certian level of insolation in terms of watts per square meter. Place any two objects in a furnace set to 100 degrees Centigrade and sooner or later they will both be at 100 degrees. It makes no difference what colour you paint the objects nor what they are made from.

To go even further out on a limb, to make a real difference to the emmisivity of an item the surface colour has to be better than painted on. I say that because if you look at the heatsinks used on electronic components, they are anodized aluminium and the black pigment has been absorbed into the anodized layer. Painting would be cheaper than anodizing and if it worked anywhere near as well would be the coating of choice. Aluminium is used for heatsinks because it's about the third best conductor of heat after silver and copper and is the cheapest of the three.

So, having played devil's advocate here, I'll just slip out to somewhere cool and shady hehe.

Cheers

The Beryl Bloke

[BobL]

With regards to painting your sunshade? I don't see the point for a number of reasons. A roof exposed to the suns heat will reach an equilibrium temperature with it's environment. That temperature will be same no matter what the material the roof is made from nor the colour the roof is painted. Lighter colours will probably take longer to get to the same temperature as a dark colour, but it will be the same temperature none the less. A more emmisive roof will radiate heat back into the air making the air adjacent to the roof hot which then radiates/conducts the heat back into the roof. Your roof and mine is exposed to a certian level of insolation in terms of watts per square meter. Place any two objects in a furnace set to 100 degrees Centigrade and sooner or later they will both be at 100 degrees. It makes no difference what colour you paint the objects nor what they are made from.

Without going into the science, clearly you have never set foot on a galvanised roof on even a warm day.
Even right now with the air temperature at a pleasant 28ºC with a decent breeze blowing, the galv roof on my back veranda is 42º.
When the air is still the temperature difference will be a lot more than 14ºC.

[RayG]

Without going into the science, clearly you have never set foot on a galvanised roof on even a warm day.
Even right now with the air temperature at a pleasant 28ºC with a decent breeze blowing, the galv roof on my back veranda is 42º.
When the air is still the temperature difference will be a lot more than 14ºC.

Hi Bob,

Just for fun, I'd like to hear about the science. My understanding is that it's all about the absorptivity to emissivity ratio,

So if we want an optimal solution we want the front to have low absorptivity, polished aluminium 0.09, white titanium oxide paint 0.17, so we leave the front unpainted.
and the back to have high emissivity, black paint = 1.0, white titanium oxide paint = 0.92, not a whole lot in it, but black is better.

So paint the back black, and leave the front alone. will lead to the coolest panel? does that sound right?

Ray

[Stustoys]

I'd like to hear the science as well.

Why do you want a high emissivity back Ray? Yes you get the coolest panel.....but thats not what we are after. The inverter doesn't care how hot the panel is, just how much IR it is receiving from it.(ok it probably doesn't even care about that :) )

My thoughts......


With regards to painting your sunshade? I don't see the point for a number of reasons. -snip- Centigrade and sooner or later they will both be at 100 degrees. It makes no difference what colour you paint the objects nor what they are made from.


The furnace analogy while correct doesn't apply to this case.
Leave the furnace door open, place your objects(say one black and one white) outside the furnace. Now measure the the temp of the back surface(in IR, not the actual temp of the surface). In this case you don't care if the white object is the same temp as the black object(which it wont be), you only care how my IR it is radiating.



To go even further out on a limb, to make a real difference to the emmisivity of an item the surface colour has to be better than painted on. I say that because if you look at the heatsinks used on electronic components, they are anodized aluminium and the black pigment has been absorbed into the anodized layer. Painting would be cheaper than anodizing and if it worked anywhere near as well would be the coating of choice.


Guess what...... they paint Aluminium heatsinks as well. Now painting some heatsinks would be a PITA so maybe thats when they anodize?, can't really say, but I can say that there are painted heatsinks. Further I think, with most heatsinks you'd be more interested in convection than radiation.

So question to BobL, are emissivity and convection related? or it emissivity strictly an IR thing? (I think the later but I'm not sure)

Stuart

[RayG]

I'd like to hear the science as well.

Why do you want a high emissivity back Ray? Yes you get the coolest panel.....but thats not what we are after. The inverter doesn't care how hot the panel is, just how much IR it is receiving from it.(ok it probably doesn't even care about that :) )



Good question, I was thinking cooler shield would be better... leaving aside convection cooling..

Which solution results in the coolest inverter,

A.. Where the back is high emissivity, and therefore the shield is cooler, so emitting less infrared (?)
or
B..Where the shield is hotter and has lower emissivity...

I guess it's a black and white question, not sure the answer will be quite so black and white... :D

Ray

[BobL]

Here's a little predictive test for you guys

I just put 4 pieces of metal in direct sunlight.
A) White paired steel both sides
B) White Painted steel front, black back side
C) Unpainted slightly oxidised Al both sides
D) Unpainted slightly oxidised Al front with a black back

Air temp is 28 degrees.
The Ally is slightly oxidised and the white paint has some scratches etc to simulate real world situation.
Black paint is pot belly black

What do you think the temperature differences will be between the air temp and the temp of the front of each plate.
Hint, see example involving my roof in a previous post

[RayG]

Here's a little predictive test for you guys

I just put 4 pieces of metal in direct sunlight.
A) White paired steel both sides
B) White Painted steel front, black back side
C) Unpainted slightly oxidised Al both sides
D) Unpainted slightly oxidised Al front with a black back

Air temp is 28 degrees.
The Ally is slightly oxidised and the white paint has some scratches etc to simulate real world situation.
Black paint is pot belly black

What do you think the temperature differences will be between the air temp and the temp of the front of each plate.
Hint, see example involving my roof in a previous post

In order of hottest to coolest C D A B, probably not a big difference between A and D of course if you used an IR thermometer, then you would have to make sure you correct for emissitivity on the Al..

[BobL]

In order of hottest to coolest C D A B, probably not a big difference between A and D of course if you used an IR thermometer, then you would have to make sure you correct for emissitivity on the Al..

My prediction was similar, C A,D B, and while there was bugger all in it, the measurements came out at A=D, C, B

All measurements were done with a thermocouple in the middle of the plate, the reading took about 5 seconds to reach a steady value but I left it there for 10 seconds just to make sure. Plates were left for 10 minutes to equilibrate in direct sunshine on top of a bare wood outdoor table.

Ray is right about not a lot of difference between A and D, in fact they were both the same at 13º above air temp
But they were also the coolest

Then came C at 14º above air temp, and then B at 15º

In other words not much difference between them.

Can I explain this, not really, the numbers are too close and there are too many variables.
We tend to focus on colours, but reflection and emission are also very sensitive to surface roughness and dirt, which are difficult to quantify and even harder to maintain in the real world.

Then there is "what else is around" that will reflect and radiate heat. e.g. painting the back black may absorb significant heat radiating away from a nearby brick wall.

Just moving D) [ally front/black back] from the top of the bare timber outdoor table ~600 mm above the (warm) patio brick paving rapidly raised the temperature difference from 13 to 16º.

I think it was Stu that said the temp of the plate was not that relevant and it should be the temp at some distance behind the plate that we should be looking at. If there was more than a couple of degrees difference in what we've seen so far I might pursue it but give that its so close I'm done.

[RayG]

Hi Bob,

Interesting result, I tried to find a source on-line for absorptivity/emissivity ratios, but the data is very scattered, lots of difference between polished aluminium and oxidized aluminium, and different paints and coatings vary widely. So controlling all the variables is going to be a bit problematic.

One area that might be worth looking at further is the super reflective roof paints that are around...

Ray

[Stustoys]

of course if you used an IR thermometer, then you would have to make sure you correct for emissitivity on the Al..
But, I would argue. I think.:rolleyes:...........correcting for emissitivity of the plate will give you the surface temp of the plate. By not correcting for emissitivity don't you get a relative reading of how much IR is being emitted?(or am I going a step to far?)

The only example I have at hand is some numbers I came up with while looking at ways to insulate the roll-a-doors on my garage.
Inside and outside surface of the door 65C.
Shade cloth a couple of inches in front of door, outside 51C, inside 52.5C.
Foil against the back of door, 77C outside, 36C inside.
Shade cloth and foil, 62C outside, 33.5C inside.
Now even though the foil is touching the door in places so I issue would be near as damn it the same temp as the door, standing inside the garage I can tell you I(or the inverter in this case) feels a lot cooler behind the foil.

Am I drawing the wrong conclusions?


Stuart

[BobL]

Hi Bob,

Interesting result, I tried to find a source on-line for absorptivity/emissivity ratios, but the data is very scattered, lots of difference between polished aluminium and oxidized aluminium, and different paints and coatings vary widely. So controlling all the variables is going to be a bit problematic.

One area that might be worth looking at further is the super reflective roof paints that are around...

Ray

Yes a lot of testing is done in the idealised state which changes dramatically one the real world kicks in.

One of my colleagues at work spent decades testing roofing paints and comparing them with Colorbond, galv and ZnAlume. He tested a lot of stuff and even built a weather chamber where he could simulate Summer heat 24/7. About the only thing that he came up with up with over the long term was that white painted surfaces maintained lower temperatures than either Zn and Zn Alum, (I don't think he tested Al), and the that the so called heat insulating paints were not as good as they claimed.