Saturday, December 10, 2011

Summary/Conclusion - DIY Flat Plate vs. Concentrator Performance

Two types of solar thermal collectors are in common use. The parabolic trough concentrator (example mounted on the right side of the solar test jig in the picture), is used in large thermal plants in the desert but is not often home built like the "flat plate" type collector (example at the left) which has avid user/builder groups promoting and advancing it's design and use.

Both make great Do It Yourself (DIY) projects that can be adapted to local needs. Both work well. How does their performance compare? I decided to find out. I was surprised that my two fairly carefully researched and constructed models were fairly close in performance. But there were some big differences.

Both types of collectors can be built by anyone who is handy with common shop tools. Most parts and materials are from home building supply centers. I used easy-to-follow plans from the internet. I built the collectors myself in my garage near Toronto and then did a series of measurements (See INDEX link below) to assess the heat gathered by each under the same conditions. Both types worked well. Both were challenging DIY projects. A solar heat source, even if home built, is a significant investment of time and dollars. For longevity and best performance, a project of other than the smallest size should be carefully done and perhaps a few models built to check concepts and presumptions. These are two of my models.

I am not yet attempting to make electricity. For now, I want to efficiently capture the sun's free heat and channel it into a flowing liquid - water in this case.

So which performs better: the flat plate or the concentrator?

As a DIY project, the result may depend on your building skills and perseverance but possibly more specifically on whether you choose to rotate your flat plate toward the sun with a tracker. This is more complex but it may be worthwhile. This is the only case where my flat plate out-performed my concentrator, but it was close, within a degree. In the other tests, the concentrator did better.

To be sure, the concentrator can have a real advantage in situations where a higher temperature is required. In this respect, it is better to choose the system with the higher stagnation temperature. Here the parabola truly shines. Over 636°F / 336°C reached in with the concentrator vs 200°F / 93°C with the flat plate (no flow). But with the same flow rate, the amount of heat captured by each is comparable.

In terms of collecting the heat that falls on each under the same conditions my conclusions are:

Configuration and link to specific testAdvantage?
Uninsulated, stationary flat plateConcentrator
Uninsulated, both trackConcentrator
Insulated, stationary flat plateConcentrator
Insulated, both trackFlat Plate

In each case, the detailed test description and measurements are shown at the links.

Gary Reysa of in his similar comparative tests of alternate flat plate designs suggests a metric of "the higher temperature wins". The graphs in my tests clearly show which collector achieves the highest temperature in each of the tests. This is the basis for the table above. Other interesting stuff happens during the tests: one starts heating earlier in the day and one cools down faster. The graphs show these effects also and are discussed in some detail for each test.

As an alternative presentation, I calculated (integrated) the area between each temperature measurement and the ambient temperature and summed them (the number of degree-rise-minutes over the whole test). The collector with the larger number I called the winner at 100% and I divided the other number by it to calculate how it compared to the winner. See below for how that data looks. Where both collectors are insulated and both track the sun, over the day, the concentrator yields only 93.4% of the energy gained by the flat plate. Not many people track their solar thermal panels. Maybe they should? It does add complexity however.

Degree-rise-minutes/day over ambient normalized to 100%
Test methodConcentratorFlat Plate
Uninsulated, stationary flat plate100?*
Uninsulated, both track10078.9
Insulated, stationary flat plate10088.7
Insulated, both track93.4100

* I can't do the calculation for the uninsulated, flat plate stationary test since I did not record the ambient but the concentrator is certainly a 100 (the higher heat from the concentrator over the day can be seen from the graph).

There are other considerations than the raw performance which will influence your approach. These can be debated and some overcome, but not others (where I live):

Other comparative considerations
Material and labor cost,
Perceived difficulty to manufacture/maintain,
Features of the design: the concentrator does not require a rotating fluid joint and uses less material, specifically less metals,
Unreliable results in winter due to unclear sun, the reflector must be clear of snow and ice and the sun must be visible (no clouds), blockage of the mechanism with ice would be a problem, cleaning of the mirrors or the glazing of the flat plate, wind loading and etc.

These factors are not discussed here, only relative heating performance under a hopefully clear blue sky. Likely you are not in Canada and the cold weather concerns might not trouble you. You may have other challenges where you are?

Hopefully the results I have posted here will encourage the exchange of more concepts and experimental results. Your own situation, your solar resource, your skill level and the materials you have available will determine what is best for you. Hopefully I can show what was possible in my situation.

A major area for continued work for me will be improvement of the insulated concentrator collector (this was my first attempt) and the achievement of higher temperatures, including steam generation, something that is not possible with a flat plate collector.

Your comments and suggestions are welcome. Thank you for your interest.

George Plhak
george (at) ffwdm (dot) com

Thank you once again to Gary Reysa for the loan of his HOBO data logger and for his support and encouragement. See his writeup on these tests here.

If you would like to read the full report, you can begin here:
Index - Comparing concentrator to flat plate solar collector


Thomas said...

Interesting comparison, thanks !
Still, it would be nice to see how both perform under cloudy conditions. I feel like the concentrator would fail, while the flat plate would still take advantage of the diffused light that hit it.
Any clue ?
Cheers !

george plhak said...

If you go towards the end of Test 1, you will see an observation about the cloud performance between the two, a graph with red (heating) and blue (cooling) lines added. I wrote:

"It seems that the two systems heat at similar rates (the slopes of the red lines are parallel) but the flat plate cools more quickly that the concentrator (the blue lines for the flat plate slope downward more than for the concentrator."

Not definitive, but the difference in the graph is interesting. The concentrator does not "fail". Thanks.

george plhak said...

Here is a link for that test:

Thomas said...

Thanks very much ! I'll review the whole site then ! There probably more replies I'm wondering about.

Gray said...

Thanks for this. Was any comparison done with the concentrator in a horizontal position. If I'm not going to have a tracking system, it seems like the concentrator might work better as long as the angle was tweaked every so often. Thoughts?

Gray said...
This comment has been removed by a blog administrator.
George Plhak said...

Gary - I removed your duplicate post.

I did not test the collectors with them horizontal (parallel with the ground). Is that what you meant?

db said...

Hello George,

Again, thanks a lot for your wonderful site, i hope to start working a steam collector soon.
Did you ever see this it seems this can concentrate the sun without tracking? You think it is worth trying?


D. Bachmann

George Plhak said...

Here is the link D. Bachmann is refering to at "this" if it does not work for you:


Anonymous said...

I also had a look some time ago about compound parabolas, both 2D and 3D, I would say they are a bit hard to recreate to be perfect and work as intented, but worth it. You can forget about the tracking mecanism, but put more efforts into making the compound perfect.