panel output at various flow temperatures

[KLD]

Following on from the "Winter optimised solar to drive UFH" thread, where we also discussed how the panels behave for different flow temperatures, I took the recent stable weather to perform a three day experiment.

We have 40 tubes 47mm, roof mounted at 36° and oriented due South, loading a 216L thermal store. The previous week was rather bad (solar output-wise), and we didn't top-up by boiler. Thus, the store temperature had dropped below 35°C. The last three days then were very good, with blue skies and many hours sunshine. The store temperature climbed overall and is now (after our daily use) still above 55°C.

I measured the heat input into the store, and calculated the heating rate (kW) for various store (and therefore: flow) temperatures. See the attached plot. The four blue data points are actual measurements, whereas the red (top) point is taken from the SPF Testreport C782 (980W peak per 20tubes, assuming 20°C ambient temperature).

Also attached is one of the three days' plot of average store temperature and calculated heat input.

So, what are main factors governing the panel output ?

Discuss! 

Klaus

[SimonHobson]

That's quite interesting. I've been meaning to knock up a spreadsheet to allow option to be fed in and have the puter do the number crunching so you can ply "what if ..." games with things like "do I heat a store bottom up or top down" since there's is some discussion (such as this thread) as to which is best !

In there is a link to http://www.apricus.com/html/solar_collector_efficiency.htm where the factors (and formulae) are stated.

[EccentricAnomaly]

Many thanks for doing this. More thinking required than is suitable for a Sunday evening....

I'm a tad suspicious of that curve zooming off the top of the chart, though. Wouldn't one expect the curve to go the other way - i.e., to go flat as the flow temperature reaches ambient. Still, it's nice to see the power increase as the temperature drops.

[KLD]

EA

I'd suggest to take the solid line curve as nothing more than 'a guide to the eyes'. As long as we have no hypothesis on which to build a mathematical model, any line fitting to the data, and especially any extrapolation beyond the current data range, remains guesswork. Even the uppermost point is not real data, as I couldn't find what conditions SPF use for their testing.

Is there anybody here who had comparable data to fill in the gap towards lower flow temperatures?

Klaus

[ericw]

The output graph in the SPF report plots the output against (mean panel temperature - ambient)

If you are a real masochist the you might try playing with the programs in this old thread http://www.navitron.org.uk/forum/index.php/topic,5204.msg51097/topicseen.html#msg51097

The biggest factor controlling the output (apart from the sun obviously) works out to be the temperature delta between the system and ambient.

[KLD]

Eric said: "The biggest factor controlling the output (apart from the sun obviously) works out to be the temperature delta between the system and ambient."

Hmm, not sure that's the entire story. The heat pipes will be designed (by selecting the fill medium and pressure) to perform best in a certain temperature range. If I were to compare results from two identical panels in rather different environment, say the first at 5°C ambient and the second at 55°C, and I keep the system temperature during the testing near the ambient temp, and all other factors constant, then I'd expect a higher output from the colder system. That's not reflected in the simple formulae using (T_m - T_a).

Klaus

[KLD]

In other words, what I'm after is the temperature dependency of the conversion factor, eta₀.

Klaus

[desperate]

Hi all,

I don't know too much about all this, but how relevant is the ambient temperature, more to the point is the effective sky temperature, which on a cold frosty october day could be higher than on a hot humid july day. That would result in a higher temperature difference across the tubes.

Just bouncing an idea around really.

Desp

[EccentricAnomaly]

Klaus's graph with addition of curves of expected power output calculated from the SPF test figures assuming an ambient temperature of 20 °C. Seems to fit well with somebody's (Martin's, Ivan's?) recent comment of getting a maximum of 60% output from PV at this time of year.

I look forward to seeing some more test results in the bottom left corner of the graph.

[KLD]

EA mentioned the

"bottom left corner of the graph."

I can hear the perpetuum mobile mob raising their heads, wanting to extract the 100W from a panel under 0W/m² radiation 

Incidentally, I've never seen a much higher output from the panels under higher radiation (June data). Hence, I believe that radiation is not the limiting factor.

Does anybody have a textbook about heat pipes etc? I've googled for far too long already, yet wasn't lucky in finding any theoretical or experimental data about the temperature dependence of the heat pipe heat transport capacity. There are plenty descriptions of the limiting cases, where heat transport ceases due to too low or too high temperatures, and thus there must be some sort of bell shaped curve connecting the limiting points. Anyone any pointers?

Klaus

[skyewright]

I don't know too much about all this, but how relevant is the ambient temperature, more to the point is the effective sky temperature, which on a cold frosty october day could be higher than on a hot humid july day. That would result in a higher temperature difference across the tubes.

Is ambient maybe relevant because of the requirements of the "heat transfer compound"?

Reading download_73 (Solar Evacuated Tube Collectors.pdf) yesterday I see that Navitron use a material that "Under the vacuum conditions that exist in the heat pipe, and at low heat pipe temperatures (<30oC), this mixture will form a frozen “ball” located in the heat pipe tip."

So (if I've understood correctly, which may not be the case, and if so I'm happy to be corrected) at low ambient temps, the solar radiation needs to raise the temp of the heat pipe further before the "heat transfer compound" can start doing its stuff?

[billt]

Thanks to Simon Hobsons link I did a couple of graphs which show the theoretical efficiency and relative power of the Apricus panels, with different insolation levels and over a range of useful water temperatures. Of course there will be greater loss at high temperatures in a real system, as there is potential for a lot of loss in the pipes between collector and store.

To me, that demonstrates that temperature related efficiency reduction is not hugely important, compared to the huge loss of output that you get with a modest reduction in insolation (which was my point in the original thread).

To put insolation in perspective, 1000 Wpsm represents pretty clear sunlight, 700 Wpsm is hazy sunlight, 400 Wpsm is thin cloud. In other words insolaton levels that will produce useful output are all bright conditions. Complete cloud cover won't produce anything much. Also the eyes perception is logarithmic, so a perceived halving of intensity is actually a reduction in brightness by a factor of ten. So what is apparently a reasonably bright cloudy day probably won't actually have enough energy available to use.

[KLD]

The formula used to calculate these curves assumes only a temperature dependence with regards to losses. It's simply a 3 term development (constant, linear and quadratic terms) with experimentally determined weighting factors.

Maybe I'm missing something, but for heat pipes collectors, there should be an additional term describing the temp dependence of the heat pipe itself.


Klaus

[EccentricAnomaly]

Maybe I'm missing something, but for heat pipes collectors, there should be an additional term describing the temp dependence of the heat pipe itself.

I'm not sure it's another term as such. More that there will be an area down in the bottom left of the graph where the heat pipe stops working so the formula breaks down. That would presumably be dependent not so much on the deltaT but more on the actual temperatures of the heat pipe itself which is determined by insolation and ambient from a practical point of view. My guess would be that the deltaT and the manifold temperature is not so important; as long as the heat pipe is warm enough for the transfer compound to evaporate/sublime and it is warmer than the manifold then condensation at the top of the pipe will happen and heat will be transferred.

[ericw]

Anything in here of help?
http://www.enertron-inc.com/enertron-products/heat-pipe-selection.php