Lightning off-grid charging Solar Array Build w/Agrivoltaics

[MurphysLaww]

First 7 days complete.

Compared the Offgrid system with the grid tied enphase system, and it looks positive. Somewhere around 65% of the house system output. I think I was fooling with angles the first two days. The last 5, I think are what is to be closer to what I'd expect, with maybe a bit more, once the auto-tracking is running.

This bodes well for March and the summer, where I could expect 20-22kW from the truck system. If we add the two panels we might to fight the mid-day shading, I wouldn't be surprised if we were between 25-30kw, on a sub-$5k system(after rebates). That may grow a bit though. The prices of the 200ah Powerqueens have dropped again, so we could pick up another 10kw of batteries to get us to 20kw for $2k. That would add another year to payoff, but would make the system a little more robust for the summer months.

Sponsored

 

[MurphysLaww]

Have some wind today. Keeping my eye on the racking. So far, so good.

sustained 18mph, and gusting to 30mph. Tomorrow is predicted to be a full sun day.

We did top the truck off with the chargepoint last night. This morning, the lightning, sitting at 91% doesn't appear to want to take a charge from the CSP. I'm wondering if this is a function of tapering charge and the low amperage we're using ? It doesn't appear to be complaining, it's just not pulling any load from the inverter, and showing "plugged in not charging". It is showing "Charge type as UKNOWN.

I'm trying to remember if I set something during setup to max charging at 80-90%. Communication with the CSP directly has always been a bit sketchy, and I don't want to break what I do have working currently. I guess I'll wait until it's run down below 80% again and see if that is the issue.

 

[MurphysLaww]

Some errands run and dropped it down below 80%, and it's back to charging perfectly.

Beautiful day, but here is the bane of this array's existence right now...

Ironically, a powerline shadow.

And this power pole shadow...

Those two shadows will cross the panels at around 1:15 pm and cut production by half for about an hour... As the powerlines are directly overhead and aligned with the panels, there isn't much I can do. Once the panels movements are controlled by software, I'll be able to move around it a little, but only so much, and I'll have to insert another sensor to pick up the shading. It's significant enough, though that the power cost to move the panels will likely be far outweighed by the increase in output.

 

[MurphysLaww]

Biggest charge day so far, 11.1 kWh added to the truck. The calculated miles... well, it's always a little scary the struggles that developers that get paid to do this have with consistent math.(and exception testing).

With the the LiFePO4's it's hard to manage the charging, as it was down to 50v while under load, and much lower than that, and the BMS in the batts starts to shut down pretty quickly, if I'm not monitoring it. When I shut down the charging from Fordpass, it bumps back up to 52.5v, but that is with 650watts of PV input, which drives the voltage up. Surprised the developers are not more on top of the AI on this, and modeling it for settings.

I may move the battery monitoring to my Victron smartshunt, but I wasn't overly impressed with their software voltage management either in the art studio array/bank.

 

[MurphysLaww]

The Lightning is at the dealer for some recall fixes. GF is borrowing her sister's Mach-E. Charge station pro doesn't care what it charges.

And started modifying the code to make the panels move with the sun. This is the system I've been using to move the 300watt panels on the G/F's Art studio shed, modified for two motor controller outs. That will change when the Progressive automation Hall sensor actuators come, and I'll likely control each actuator individually... I think. That will be a learning experience.

LDR light sensors and the shade block. Ignore the crap 3d print. The one on the shed is much nicer. I'll probably re-print this one.

12v dual 18650 batts I DIY'ed. This set was actually made to run the solenoid, and rain barrel pump for our smartgarden.

It's all sitting in one of the mobile chargers I've built to take out to the farm. I needed to get some 5v USB power for the arduino. That eventually will come from a step-down buck converter that will take a feed from a newpowa 100watt panel that will sit off the fence to the left of the combiner box.

Also added a 120v outlet to the charger subpanel. Eventually, it will get another 15amp circuit for the mini-split/Heatpump heater going into the garage probably in a few weeks. Lightning is gone, so I need to be able to use this gift from the cosmos somehow.

 

[MurphysLaww]

Well, on hold for creating the actual video content for the Charging system since Ford now has the truck and who knows when we'll be getting it back. We're a week in now, and apparently they are getting us a rental truck ? This is for a Truck that had no apparent issues.

I'm not sure if it's Ford, or if it's the Dealers that seem to be actively attempting to fail at this, or are trying to make it a hassle to own an EV. I'd lean towards the dealers. Too bad their profit motive seems to have a stranglehold on helping clean up the Earth.

Maybe they are going to take them all back and send them to the crusher again, like GM.

Their window, and really, our window, as a country is closing.

China's renewable advantage will be insurmountable before long.

 

[MurphysLaww]

Still no truck after three weeks in for the recall, so really no content. System is working well for running everything in the garage, though. Except for one of the LDR light sensors, which I need to replace. I've just been moving the panel's mid day with the arduino cloud app.

We're about to pull the trigger on ordering the Progressive Automations hall sensor Linear actuators, though.

I did rewire the dual 12v 18650 batteries, as the wire gauge was too small for the amps the actuators pull once it got cold out. 22 awg to 16 awg. The power draw ends up being close to 10 amps.

 

[invertedspear]

When I was in high school physics, one of our labs was building a toy solar car with minimal "intelligence" We had 2 solar panels the size of a stick of RAM, 2 small electric motors driving 2 wheels independently. The shape of the car was a bit like an old propeller driven airplain, with the solar panels making up the wings. The motors drove the opposite wheels. This had the effect of the car turning in the direction to collect the most son. If you can picture how brighter sun on the left would drive the right wheel faster to turn the car till the power to the wheels was matched. In this way it could even turn away from driving into shade, unless momentum carried it too deep into the shade anyway.

I bring all this up because I think you could control the angle of your panels in a simpler way than using arduinos. Using a tiny panel on each edge of your main panel at a slight angle and wiring each to their own small motor to position the large panel. If the small panels are getting the same amount of sun, the force of the two motors against each other should negate and your large panel should be optimally positioned. Do this on 2 axis and the main panels can follow the sun throughout the day and year with no coded movement, and thus no chance for error. Plus due to how partially cloudy days could mean more sunlight is actually coming from a reflection on another cloud than the sun itself, they would automatically adjust for optimal capture.

Anyway, thanks for reading my thoughts. Not sure if they spark any interest in exploring, but I love things like this and I don't have the time/equipment/money/expertise to explore them on my own.

 

[MurphysLaww]

When I was in high school physics, one of our labs was building a toy solar car with minimal "intelligence" We had 2 solar panels the size of a stick of RAM, 2 small electric motors driving 2 wheels independently. The shape of the car was a bit like an old propeller driven airplain, with the solar panels making up the wings. The motors drove the opposite wheels. This had the effect of the car turning in the direction to collect the most son. If you can picture how brighter sun on the left would drive the right wheel faster to turn the car till the power to the wheels was matched. In this way it could even turn away from driving into shade, unless momentum carried it too deep into the shade anyway.

I bring all this up because I think you could control the angle of your panels in a simpler way than using arduinos. Using a tiny panel on each edge of your main panel at a slight angle and wiring each to their own small motor to position the large panel. If the small panels are getting the same amount of sun, the force of the two motors against each other should negate and your large panel should be optimally positioned. Do this on 2 axis and the main panels can follow the sun throughout the day and year with no coded movement, and thus no chance for error. Plus due to how partially cloudy days could mean more sunlight is actually coming from a reflection on another cloud than the sun itself, they would automatically adjust for optimal capture.

Anyway, thanks for reading my thoughts. Not sure if they spark any interest in exploring, but I love things like this and I don't have the time/equipment/money/expertise to explore them on my own.

I've seen this before, and some other mechanical solutions. We're I just entirely going for as simple solution as possible, I had thought about this. Like the tracking on the top of our art shed roof, does not need to be too complicated, although I may have them lower based on weather reports for the day. (any day openweather reports potential wind over 40mph)

I don't have opposing motors, either. I'm just reversing voltage on the motor drivers on a single motor, in each actuator. I wonder if what you are describing is how the large solar farms move their panels.

For me, the Arduino is all about the IOT cloud access, and once the hall sensor actuators arrive, the ability to move the panels to a storm configuration based on anemometer sensing, and of course, just general geeking out.

 

[MurphysLaww]

Realized I hadn't uploaded pics of the closer to final configuration of the control box.

The upper half will hold the controllers for the rainbarrel fed, raised bed irrigation automated drip system.

For those interested, what will be in there is more or less here:

https://www.hackster.io/racingtogreen/arduino-cloud-controlled-smartgarden-995051

MPPT charge controller for the 100 watt panel that powers this on the left, Arduino RP2040 connect in the middle, and Drok motor driver for the actuators on the right.

 

[MurphysLaww]

We did order the progressive automations hall sensor actuators today. They should ship early next week.

sigh... hopefully the truck will eventually be back so there is a reason for the project to exist...

We will be using these actuators, and will be moving from an 18" actuator to 24", so we'll be getting more than the current 15 degrees each way off horizontal, so hopefully, even more production. I think the calculated optimal angle max for our area is somewhere around 25 degrees.

The increased likelihood for durability and fact that these are IP66 rated was almost enough to decide to use these, so the addition of Hall sensor positional control is really just an added bonus.

 

[MurphysLaww]

And the Truck is back.

First charging session of 2024.

Progressive automation actuators should be here tomorrow sometime.

Our next Build may be a battery/overland trailer build. We have a 400lb utility trailer that we'll likely lift, and we'll likely lift the Lightning 3" or so. We're going to make the battery system in the Garage mobile, upgrade it to 20kW, which should be another 400lbs, with the ability to be moved into the trailer, and we'll be working on 3 x 1kW fold out panels of flexible panels. It looks like we should be able to build those into about 50lb configurations, with at least one of them charging while traveling, then 3kW ground mounted when camping. So, that should be around a 1,000lb trailer/charging sytem+ overlanding gear.

Not sure how much super-interesting content there will be here, but we'll start the build videos for this system on youtube this weekend, likely.

First sun in awhile, and still putting out 500watts at 4:30pm...

Tracking is working perfectly with the new LDR sensors, fully waterproofed this time.

Pumping it in...

 

[MurphysLaww]

The Progressive Automations linear actuators came in late last night. The quality appears to be crazy good. I had never quite convinced myself that it was worth the additional cost, for the IP rating, but just looking at them, I'm pretty sure I don't doubt it now. These things are beefy, and waterproof, and have Hall sensors...

If I had any remaining worries about the stability of the system, I think they will be gone after installing these. It's worth another 6 months to payoff, as I'll know the system will survive to payoff. That said, the additional kW produced by the single axis tracking will pay for the actuators, and we're still on schedule for a 5 year payoff assuming we put 9k miles on the truck per year, which I'm pretty sure we'll surpass. This week was the first week the Truck was around, AND the auto tracking was working and we had our highest production day yesterday, by a full kW. It'll go up even more, when these are in, and we're getting closer to 25degrees of tilt both ways, from the current 15 degrees the 18" actuators are giving.

One of the EcoWorthy 12" actuators next to the 24" PA. The body is almost the same size as the fence posts.

 

[MurphysLaww]

Well, the system survived a sustained wind test. Pretty much 40mph all night with gusting to the mid-50's.

This with the skinny Ecoworthy 18" actuators that are soon to be replaced by the huge Progressive Automations 24"

Sat outside for 10-15 minutes in the 50mph winds, and the panels didn't move at all, tilted to the evening position towards the West/Southwest. That will likely be the storm position setting going forward.

 

[MurphysLaww]

Finished a little needed work to add the 100watt panel that charges the single axis controller to the grid and powers the actuators. It had just been balancing on the top of the fence, and the 12v battery it charges was noticeably low in the early mornings if it needed to be moved. Normally, it will move back to the morning position in early night, and then not move again, until like 12-1pm

Today, I'm not moving the panels as it's going to be another nearly fully clear day. We'll see how it compares to what we get moving them, and calculate how long it will take the tracking system to pay for itself at it's current 30 degree movement max. Then we'll try to get that movement with the new actuators up to 50 degrees total movement, and test again. Should be some mildly interesting data.

Charged the truck the highest amount yesterday at 15.6kW, but due to not charging the previous day at all. Had a good amount still in the batteries from the previous, not very sunny day.