Saturday 21 March 2015

Solar Eclipse

Solar Eclipse

The solar eclipse as a spectacle wasn't on the scale of the 1999 eclipse because you needed to be more northern to get the fullest eclipse but it did give an opportunity to see what the eclipse looked like to my IR and Visible light meter as well as see it's effect upon a solar panel.

Two things about the eclipse had bugged me in the run up to it. Firstly the media were reporting that the loss of the sun during the eclipse would put strain on the power generation system and also scientists had been asking for "citizen scientists" to report on changes of clouds and wind. I didn't, couldn't and still can't see why it would have been a problem for the power generation in the country, the chances of a black out or power shortage seemed a farcical idea and more like an opportunity for a news headline and why would scientists need people to report on the weather change when they could monitor it themselves? Again, another PR opportunity to create "engagement" with the public I think.

Simple graph plots of the solar panel power output along with Infra-red, Ultra Violet and Visible light should be sufficient to answer both questions.

Click graphs for bigger image.

UV Index

Solar Panel Insolation in W/m2
IR and Visible Light

UV Index Rolling Average Over 1 hour


Obviously all graphs look incredibly similar but since we are monitoring the sun, which is being blocked up to 87% or so, that is what would be expected. The solar panel Y scale hasn't been calibrated and is probably around 18% too high but that is unimportant.

Potential power shortage?
To answer the question as to whether there was any likelihood of a power shortage caused by the dip in power from solar panels you need to consider that good solar panels are only around 15% efficient and since the sun's power hitting the panels is only going to be around 150 to 300 w/m2 at this time of year at the time when the eclipse happened a maximum loss of 15% of that equates to about 33 watts per hour. Across 2 hours that's 66 watts per meter. Considering it wasn't totally dark even less was actually lost. Also this time of day just doesn't generate much power anyway, compared to the mid day sun which can produce 150 watts per meter. Across 2 hours around mid day had the eclipse happened then, it would have lost 300 watts per square meter which would have been nearly 5 times worse. Very little power was lost due to the eclipse and when you consider that my panel had almost no cloud cover it wasn't much different to there being no eclipse and a fair amount of cloud. I conclude that the national grid had no issues what so ever compared to any normal March day during this time.

Scientists want to know weather conditions from citizen scientists during eclipse
Why? All 3 graphs, UV, IR / Visible and Solar insolation are "spiky" across most of the day. The spikes are cloud cover, pollution in the atmosphere and general humidity. They are there most of the time, even with little cloud because there is often a haze. The darker and thicker the clouds the deeper and bigger the spikes. Yesterday was a hazy but generally bright day at my location all day as the spikes show but during the eclipse the spikes disappear to a large extent. This matches what I felt, the wind seemed to drop, and it seemed to clear up the more the eclipse happened and then as the eclipse faded the wind rose a little and the clouds formed slightly more. The graphs certainly suggest the cloud cover thinned. If I can see this then why do the scientists want you to tell them what you witnessed? The calming of the weather often happens around dawn and dusk and can also be seen in the late afternoon on the graphs which also ties in with the sun having a regular effect on cloud and wind as it gets brighter in the morning and then dulls in the early evening. The sun being covered and reducing the amount of solar power hitting the ground will have also effected the temperature. Up to 300 watts per meter of solar power was reduced on the surface of the earth which is like having a 300 watt heater on every square meter turned down and then off.

It didn't go dark
A market stall owner had brought some torches just in case it went dark but as the graphs show, the solar panel output dropped to almost zero as expected at the peak of the eclipse but the other readings only dipped. A rough look at the graphs show that the light levels dropped about 84% (very roughly) which equates to the percentage of the eclipse cover, although there was also some extra scattered light getting through bouncing off of the atmosphere. Even with the sun being covered 84% or so it was still very light which goes to show how your eyes work. The pupils dilate as needed to let more of less light in. When it is very bright bright your eyes try to reduce the amount of light hitting them and when there is poor light levels your eyes open up more to let more light in and hence you din't notice the light levels drop by that much.   

A few minutes of eclipse?
The graphs also so, despite clouds and haziness,that far from the eclipse being a 10 minute wonder it happened between 8:40am to 10:30am at least and maybe longer.

Body clock
Another observation is that the birds quietened down as the eclipse happened and changed their behaviour. The previous eclipse I saw in 1999 demonstrated this to a much bigger degree, which shows that they, and us, are influenced by the rhythm of the sun and light levels.

Update
The UV sensor failed, caused by corrosion. The manufacturers have told me that the problem is caused by the lead free solder and other modern processes. Something to do with moisture, air and sulfur...it wouldn't have happened with non-eco type processes. I think it is caused by some copper pads not being covered fully by solder and the solder reacting / copper reacting with moisture. Either way it only lasted 6 months before shorting out. The white powder around the components is the problem. Replaced sensor and back up and running.


No comments:

Post a Comment