diy solar

diy solar

Solar Eclipse 2017

Opusdei

Solar Specialist
Joined
Sep 21, 2020
Messages
90
Solar Eclipse Watch in 2017

The great solar eclipse on Aug 21, 2017, has swept throughout all America. The shadow of the total solar eclipse cut a path of 70 miles wide. It was 1000 miles away from my home at Tucson, AZ. The total eclipse lasted a few minutes along the path of the totality, but a partial solar eclipse here in Arizona lasted nearly three hours long.

Over the summer in 2017, I worked with my daughters and her boy friend to develop a single axis solar tracking cart for the solar eclipse project. It took two months to assemble the solar cart during the steamy summer inside the RV garage. The idea was to use the solar cart to measure the shading effect during the solar eclipse.

solar cart 02.jpg

Figure 1. Team photo in front of the solar cart.

Solar cart consists of 4 major components, two 100 Watt mono-crystalline Si solar panels, 40 A charge controller, 170 Ah deep cycle AGM batteries, and 3 kW Xantrex freedom SW inverter. Framework of the solar cart is based on T-slotted aluminum frame 4545 system, which is easy to build and can be configured into the simple projects. Single axis solar tracking capability has been demonstrated by mounting the solar panels using variable angle pivots. Solar panels are wired paralleled with the 40 A charge controller to charge the batteries efficiently. Since the charge controller has a Bluetooth capability, we were able to measure the solar power generation data regularly. Two deep cycle AGM batteries are connected in parallel by AWG 10 wires. Emergency circuit braker and 15 A fuse between the charge controller and the batteries are installed on the solar cart system for safety. Xantrex Freedom SW inverter is tied to the batteries and there is another DC 300 A fuse installed between the batteries and the inverter.

control panel.JPG

Figure 2. Control panel with a charge controller and shunt power meter. Xantrex freedom SW inverter controler is at the center and circuit braker at the bottom inside the control panel.



At first, we thought about traveling with the solar cart to the nearest path of totality. We could not only see the total eclipse, but also monitor the complete solar eclipse by the solar cart. However, making a field trip to Oregon or Idaho would take more than 3 days. Rather we decided to measure the partial solar eclipse in Tucson, AZ. One big advantage of watching the partial solar eclipse at home is that we can enjoy the solar eclipse in the back yard. The weather was really good and sunny in Tucson, AZ, on the eclipse day.

One of the questions we had with regards to the solar eclipse is how much it will impact the solar power generation. About 60% partial shading was predicted at Tucson, AZ. Our hypothesis was that the solar power generation would be linearly proportional to the solar eclipse shading. Therefore we were trying to capture the images of the sun by the pinhole viewer during the solar eclipse. The pinhole viewer made out of a shoebox worked fine to project the image of the sun in normal day. However, we were not able to take a clear image of the eclipsed Sun because the smart-phone camera was not sensitive enough to capture the glimpse of the solar eclipse. Rather we decided to use NASA’s eye simulation program. NASA's eye simulation program is a very accurate web-based 3-D simulation, that can show the view of the solar eclipse from anywhere in US. Once we entered the GPS location of Tucson, 32.1839 N, 110.5713 W, NASA’s eye simulation showed what the eclipse should look like throughout the solar eclipse period.

1602309999741.gif

Figure 3. Screenshot of NASA’s eye simulation program. Red color circles and triangle were drawn to estimate the shaded area of the solar eclipses at Tucson location.

Since NASA’s eye simulation doesn't tell us about how much area of the sun becomes eclipsed by the moon, we developed an algorithm to estimate the area of solar eclipse. Assuming the apparent size of the Moon is same as the area of Sun, the area of the eclipse is equal to the twice of the difference between the circular sector and the triangle. Ratio of eclipses could be calculated by repeating this procedure. On eclipse day of 8/21/17, we measured the power generation by the solar cart and repeated the measurement once again on the other sunny day on 8/27/17, which can be compared with the eclipse data as a control. Solar cart power data was acquired every 5 minutes during the solar eclipse in order to study the variations and test if the eclipse area predicted by the NASA’s eye simulation would agree with the actual solar power generation.

1602309999774.gif

Figure 4. Comparison of the solar cart powers generated on eclipse day on 8/21/17 (red rectangle) and control on 8/27/17 (black line). Simulation data from the NASA’s eye (blue line) is fitted against the eclipse results.

1602309999806.gif

Figure 5. Solar power loss during the eclipse (red rectangle) and the estimates from the NASA’s eye simulation (blue line).

Figures above show the solar power profiles measured by the solar cart on 8/21/17 and 8/27/17, and the result of the NASA’s eye simulation. This figure shows how solar energy production goes down as the shadow of the Moon eclipses the Sun. It confirms the hypothesis that the amount of the solar power loss would be proportional to the shaded area during the solar eclipse. Simulation results matched very well with the actual solar power data generated by the solar cart. Especially the peak of the eclipse ratio data and its magnitude matched really well with the NASA’s eye prediction. Note that simulation results seem to underestimate the solar power loss in the solar eclipse results. Solar panel may have an intrinsic characteristic at different climate condition during the measurements at the eclipse and control. Indeed, Si panel may perform worse at low light intensity condition. As a result, we can conclude the solar power loss is linearly proportional to the eclipsed area of Sun during the solar eclipse.

With solar energy now having greater significance for national power generation, the solar eclipse in 2017 reminded us of the importance of solar energy in our lives. Last but not least, we became to appreciate the science by building the solar cart and really enjoyed watching the solar eclipse in 2017.

smoothie.JPG

Figure 6. Blueberry “solar” smoothies made by my daughters to celebrate the great solar eclipse in 2017.
 
Looks like that was a fun experiment, and great matchup to the NASA data!

Did your experiment account for the higher efficiency of the panels as they cooled? That might explain why the datapoints are closer at the start of the eclipse and farther away at the end.

For example the Pmax on my panels is -0.37 %/°C. So, if the panels are in direct sun they might be 140°F, but in reduced sunlight with a breeze might drop to 100°F (say 23°C) over the span of the eclipse (about a 8.5% wattage change).
 
Last edited:
Great observation. As shown in Fig 5, solar power loss during the solar eclipse was about 5-10% bigger, which can be gracefully explained by the negative temperature coefficient of the solar panel performance. Let me try find the temperature profile of the solar eclipse day at Tucson. Temperature could have been ~20 degree Celcius higher at noon, because we love the sun in Arizona ^^.
 
NOCT could be a contributing factor to the bigger loss in solar power generation at noon. The acture solar panel temperature could have been much higher than the ambient temperature at noon. I will try measure the NOCT directly from the solar panel next time. That will be another fun experiment!
 
Last edited:
Wow, awesome project. I fortunately live just a few miles from the totality path, and took a trip with my children to enjoy it. I’ve got lots of pictures of the event. I even had an astronomer parked by me with a huge telescope and I got some shots from that!
3 HOURS! Wow. Our path only lasted about 45 minutes, start to finish.
 
Wow, awesome project. I fortunately live just a few miles from the totality path, and took a trip with my children to enjoy it. I’ve got lots of pictures of the event. I even had an astronomer parked by me with a huge telescope and I got some shots from that!
3 HOURS! Wow. Our path only lasted about 45 minutes, start to finish.
Totality lasts just a couple minutes while partiality endures a couple hours. You need a special observation tool to watch the solar eclipse - solar eye^^
1602353700101.png
 
Totality lasts just a couple minutes while partiality endures a couple hours. You need a special observation tool to watch the solar eclipse - solar eye^^
View attachment 24815
Hey, where was that pic taken? All the Pharisees writing on the structure reminds me of when I was in Iran. I was with my father when he was there. We settled in Tucson after, and there was a lot of that around.
 
Hey, where was that pic taken? All the Pharisees writing on the structure reminds me of when I was in Iran. I was with my father when he was there. We settled in Tucson after, and there was a lot of that around.
One of my colleague
Hey, where was that pic taken? All the Pharisees writing on the structure reminds me of when I was in Iran. I was with my father when he was there. We settled in Tucson after, and there was a lot of that around.
I don't know the picture. I led a group of high school students and it was from the syllabus. Hey, glad to hear about coming from Tucson, one of best solar cities and a great place for stargazing.
 
I was at a conference in 2017 and one of the speakers was Michael Picker, Chair of the California Public Utilities Commission. He publicly stated that he was concerned about the impact the eclipse was going to have on the grid when the solar went offline. I guess he was assuming the whole state was going dark at the same time. I was surprised at his lack of understanding of physics.
 
Opusdei great thread and looks like the family had a fun project .
It brought back great memories of the eclipse. We were in the National Forest boondocking north of Dubois, WY (8,400' elevation) for it.
Start of eclipse was 10:18am
Start of total eclipse 11:37am
Maximum eclipse 11:38am
End of total eclipse 11:39am
End of partial eclipse 1:03pm
2hrs & 19 minutes.

This was the morning of the eclipse, no smoke from fires and no clouds.
20170819_094835_0cc176807d0defa74133bfdf8c9fd491c77093bc (1).jpg
I didn't record a whole lot that day for readings from the the displays.
Eclipes readings 082117.PNG
There was a decrease in production but still getting something. I guess I was enjoying the eclipse more than getting recordings.
This was looking up during the full eclipse (no special camera), I remember thinking how weird it was during the period of time of the full eclipse (nighttime feeling during the day) waiting to see zombies come walking out of the bushes. ;)
20170821_113635.jpg

Next total eclipse for North America will come on April 8, 2024. We're going try to be in an area of totality again, maybe I'll get better reading of the solar this time.:)
 
I was at a conference in 2017 and one of the speakers was Michael Picker, Chair of the California Public Utilities Commission. He publicly stated that he was concerned about the impact the eclipse was going to have on the grid when the solar went offline. I guess he was assuming the whole state was going dark at the same time. I was surprised at his lack of understanding of physics.
My original plan was to build dozens of the solar carts at different locations. I should have contacted the NASA for funding^^; The chair of CPUC didn't seem to understand the whole picture of the solar eclipse in 2017. Hopefully he learned from it.
 
Great picture of the solar eclipse in 2017! Totaliity started from Oregon in the morning and moved along to the east and ended in Florida in the late afternoon. As shown in the attached table, the current decreased nearly by 50% at 13:08. BTW your RV was equipped with 12V 500Ah LiFePO4 batteries for boondocking. I would like to hear more about your van life^^
 
Great picture of the solar eclipse in 2017! Totaliity started from Oregon in the morning and moved along to the east and ended in Florida in the late afternoon. As shown in the attached table, the current decreased nearly by 50% at 13:08. BTW your RV was equipped with 12V 500Ah LiFePO4 batteries for boondocking. I would like to hear more about your van life^^
Here is thread I started on another site back in 2016 till present if you like to look at it. Like most threads there is misc info on different things. Within the thread I have my system explained some, some year end #s and pics of places boondocked.
National Forest trip
Quick Summary
Day 1 installed 4/13/16, fulltime use.

Days boondocking living off of solar/ LFP
2016_ 200 of 261 days
2017_ 365
2018_ 365
2019_ 344 of 365 days
2020_ 216 as of Oct 12 out of 264 days.

As of 10/12/20
1,492 (boondocked) of 1,641 days.

The times I did have hookups the solar does all charging of the batteries. I turn off the charger on the inverter, but inverter is still on incase there is 120v power loss which has happened and didn't even know till I looked at the displays.
 
Last edited:
Here is thread I started on another site back in 2016 till present if you like to look at it. Like most threads there is misc info on different things. Within the thread I have my system explained some, some year end #s and pics of places boondocked.
National Forest trip
Quick Summary
Day 1 installed 4/13/16, fulltime use.

Days boondocking living off of solar/ LFP
2016_ 200 of 261 days
2017_ 365
2018_ 365
2019_ 344 of 365 days
2020_ 216 as of Oct 12 out of 264 days.

As of 10/12/20
1,492 (boondocked) of 1,641 days.

The times I did have hookups the solar does all charging of the batteries. I turn off the charger on the inverter, but inverter is still on incase there is 120v power loss which has happened and didn't even know till I looked at the displays.

Beautiful pictures! It took a while to scroll down through the eclipse day on 8/21/17 and I will try the rest of never ending stories later today. Campfire and coffee maker calls me out to the wild nature :)
 
Back
Top