Black or grey roof with solar panels?

[hsnoil]

I'm redoing my roof prior to putting up solar and I need to choose a roof color.

It is well known black roofs absorb heat while lighter roofs reflect it. So from what I understand in hot climates closer to white is best, and cold climates black is best. But here in NY, we get both really cold weather and really hot weather.

So which one is the best option?

I am thinking since solar provides most of my electricity, I would have increased generation during summer anyways due to longer sun hours, so my solar panels would offset the increased AC consumption of a black roof. In comparison, if I go with a lighter roof, that means burning natural gas for heating during winter? So I am more leaning towards the black roof.

Though since a large part of my roof will be in solar panels anyways, I am not sure how much it will even matter one way or the other.

[hsnoil]

So far looking around on the topic, I found this that seems to show black roof for some reason results in higher efficiency?

https://www.reddit.com/r/engineering/comments/al8ec4/does_the_color_of_a_roof_affect_the_efficiency_of/

[J.P.M.]

If aesthetics or other issues are not a problem, go with white.
Reason: A white roof will tend to keep that roof deck and the ambient air at/around the roof cooler than a darker color roof. That means the array will run cooler and produce more energy.

In the winter, if your home is reasonably well insulated (as it ought to be before you even consider PV), snow will stay on the roof a relatively long time.
That means you'll gain little to nothing from a black roof when there is snow on it.

When there is no snow cover, particularly during summer, the white roof will tend to keep the panels much cooler than a black roof.

Example: Depending on the wind velocity and direction, the ambient air temp. around my roof top array runs anywhere from ~ 3C to maybe 8C above the ambient air temp. at ground level with a dark brown roof. That increased air temp. increases the cell temps. by about the same amount.

If I had a white roof, the rooftop amb. air temp. would be closer to and maybe even less than the ground level ambient air temp. making the array temp. lower by ~ the same amount.

As long as the roof is a diffuse reflector (that is, not specular), albedo from a white roof onto the array will do little to increase the cell temps. for a normally placed array that's parallel to the roof, except maybe a bit for the outer edges of panels located on the outside of an array.

If the panels are tilted to the roof, expect some panel heating from roof albedo but also some additional production from the same albedo, Quantitatively, those two effects tend to balance one another to some extent.

BTW, and a bit off topic, don't forget that a PV roof array will often/usually have reduced production because of snow cover.

[hsnoil]

But wouldn't the outer edges statement apply to both the black roof and white roof? Only difference would be that black would absorb the heat, while white would reflect it(some towards the panels). Most of the roof is in panels after all. And panels are slightly lifted off the roof to begin with. The only time I can see increased heat would be at night, but at that point panels won't be generating anything anyways. The reddit link shows an experiment where someone did a test and it showed solar panels on black surface generating more than on white surface. But there wasn't much in the reddit post with definite explanation other than guesses. And depending on time of day the snow is, it may actually help melt the snow. Cause if it builds up heat during the day and releases it, the snow is more likely to slide off, no? I read some mention of that looking around. Comparing black solar panels to brown roofs wouldn't really be a 1 to 1 comparison cause the solar panels themselves are reflective. They also generate far more heat than their surroundings.

[J.P.M.]

But wouldn't the outer edges statement apply to both the black roof and white roof? Only difference would be that black would absorb the heat, while white would reflect it(some towards the panels). Most of the roof is in panels after all. And panels are slightly lifted off the roof to begin with. The only time I can see increased heat would be at night, but at that point panels won't be generating anything anyways. The reddit link shows an experiment where someone did a test and it showed solar panels on black surface generating more than on white surface. But there wasn't much in the reddit post with definite explanation other than guesses. And depending on time of day the snow is, it may actually help melt the snow. Cause if it builds up heat during the day and releases it, the snow is more likely to slide off, no? I read some mention of that looking around. Comparing black solar panels to brown roofs wouldn't really be a 1 to 1 comparison cause the solar panels themselves are reflective. They also generate far more heat than their surroundings.

I'm having trouble posting just now. something about an invalid security certificate and this site may be being taken over/hacked.

MODS: Any idea what's going on ?

Hsnoil: I'll respond when I can get back on line.o

J.P.M.

[hsnoil]

I'm having trouble posting just now. something about an invalid security certificate and this site may be being taken over/hacked.

MODS: Any idea what's going on ?

Hsnoil: I'll respond when I can get back on line.o

J.P.M.

Their SSL certificate expired on Thu, 21 Jul 2022, hence the error. They either didn't manually renew it or the automatic script broke (so it is not a hack). And chrome is annoying as it keeps blocking everything. So I just use firefox to get around it.

[J.P.M.]

Their SSL certificate expired on Thu, 21 Jul 2022, hence the error. They either didn't manually renew it or the automatic script broke (so it is not a hack). And chrome is annoying as it keeps blocking everything. So I just use firefox to get around it.

OK. I'll give it another try if this post goes through.

The practical and also the technical explanation of why panels and arrays parallel to and slightly above (by 10 - 20 cm or so) light colored roofs will tend to run cooler (and so gather more energy on an annual basis) than if on a darker color roof has mostly to do with how much more heat the air flowing over the roof gains when the array is operating.

Do what's called a heat balance on the array (actually, to be correct and completely inclusive of the electrical energy produced, do an energy balance, but that'll complicate things more than necessary for this discussion),
The array will gain and lose energy by the 3 modes of heat transfer:

Conduction - through the panel frames to and from the supports.
Radiation - from the sun and from the surroundings via albedo going into the array, and to the surroundings via albedo (the to or from part depending on the relative temperatures of the array and its surroundings).
Convection - either forced (if the wind is blowing) or natural (via gravity) if the air is still.

For getting rid of excess heat in an array, for arrays parallel and close to a roof, the primary mode of heat transfer from an operating array will be via convective heat transfer. If there is a reasonable wind regime that has an average wind velocity of maybe 2 m/sec or so, convective heat transfer will almost always be the dominant and controlling mode of the three modes, sometimes by an order of magnitude. In still air under sunny skies, and depending on other environmental factors such as humidity and effective sky temp., radiation will contribute some but convection will still dominate by maybe a factor of two or more over thermal radiation. The conduction contribution through the frames will probably be no mor than a few % or so.

There's more to this post but not all of it went through.

Mods: Can this be fixed ?