Small system to track solar generation over winter

[ruceanthony2]

I live in northern MN. I would like to build a small solar system this fall so I can measure the amount of solar energy I can generate on my site. I want to do this before I build a large 10-15 KW off grid system so I can be sure it will be effective here and I can afford and place the panels. I'm looking for recommendations on a small charger/controller/inverter that has an interface that I can interface to a computer to log power generation when the sun shines. I plan to get a few inexpensive solar panels and a battery or two that will let me utilize the power for things like furnace, water pump etc.. I would then reuse these in the strategic system if I can prove the location will work.

Any ideas?
Bruce

 

[SamDeleted]

Hi Bruce, you could go for cheap/unbranded components and monitor with something like a victron smartshunt or bmv-712 .... they can link to a phone/PC and I think will provide you with everything you need monitoring wise


something also to consider, your solar panels won't produce power if your batteries are full , so to get a true reading of the solar production you will need to keep an eye on battery %soc(state of charge) and use some form of load to keep space in the battery

 

[OffGridForGood]

I live in northern MN. I would like to build a small solar system this fall so I can measure the amount of solar energy I can generate on my site. I want to do this before I build a large 10-15 KW off grid system so I can be sure it will be effective here and I can afford and place the panels. I'm looking for recommendations on a small charger/controller/inverter that has an interface that I can interface to a computer to log power generation when the sun shines. I plan to get a few inexpensive solar panels and a battery or two that will let me utilize the power for things like furnace, water pump etc.. I would then reuse these in the strategic system if I can prove the location will work.

Any ideas?
Bruce

I am north of your location and do well with a 9.6kW array and enough battery bank for two-three days of poor weather.
Faster and easier than building a test rig (nothing wrong with a test rig) use PVWatts free online calculator, input your location and size of array and direction/tilt. PV Watts predicted my solar within a few percent of actual.
If you have trees/shading issues, then this will cut into production considerably.
If you are looking to go completely off grid, I suggest you look at inverters with super low idle consumption, like Victon stuff and others, the Nov-Dec low light is a killer for my system for 80 days starting Nov 1st every year, just too many things all at once (fog, cloud, short daylight hours, snow all at once here), a good back up generator (dual fuel is my recommendation) and or have utility back up for low production time of year.
My advice from a similar region, either build the solar to tilt up in winter or build it at a set steep angle, to assist with snow removal. I use 30 & 60 degrees and it seems to work very well. I also recommend bifacial if you can place them up off the ground/mounting surface enough to allow light behind the panels. Been working well for me.
Post your progress and questions, we are all eager to help!

 

[efficientPV]

All you need to monitor is panel current into a fixed resistance which is lower than the calculated ideal resistance. Just record the voltage on that resistance. Use a Lascar EL-USB-3 Voltage Data Logger to record the data. A RC-5 would also be nice for temperature.

 

[Shimmy]

It is harder than people so far are making it out to be. You need an MPPT charge controller rather than the cheap PWM controllers that normally come with small panels. You also need one that is going to work with a single panel-- low enough minimum input voltage. On the regulated output side an adequately sized resistor with heat sink and a shunt for measurement will do the job. If you just want daily information then a fairly cheap/dumb shunt will do the job, but if time-of-day information matters to you (it does for me), then you are going to need your own DC logger or all the parts necessary for a working smart shunt.

There used to be pole mounted commercial loggers out there that would have small cells pointing in four directions at something like a 45 degree inclination that could give you information on total solar, direct/indirect solar, optimum orientation, and a lot more on a minute-by-minute basis. I have not seen them in a few years though.

 

[OffGridForGood]

I plugged in Duluth MN, 1kW PV due south, tilt at 47-degrees into PV Watts Calculator:
Total annual is estimated at 1255kWh/yr (this is per 1kW array size)
Best month August at 130kWh/mth, and Worst is November at 69kWh/mth - a ratio of best to worst of +/- 2:1
Tilting 30/60 Oct-Mar high angle and April-Sept low angle makes very little difference - except snow removal difference will be significant.

A typical 30kWh/day house would need an array of 8.8kW on average, but x2 for that Nov-Feb timeframe, 16-17kW array per 30kWh per day average consumption. If you run significant A/C loads, at least these arrive at the same time as the longer sunny part of the year, when PV production will be about double the winter values. Trees/ shade will reduce these PV numbers significantly if that is an issue. Gives you a starting point at least.

 

[OffGridForGood]

It is harder than people so far are making it out to be. You need an MPPT charge controller rather than the cheap PWM controllers that normally come with small panels. You also need one that is going to work with a single panel-- low enough minimum input voltage. On the regulated output side an adequately sized resistor with heat sink and a shunt for measurement will do the job. If you just want daily information then a fairly cheap/dumb shunt will do the job, but if time-of-day information matters to you (it does for me), then you are going to need your own DC logger or all the parts necessary for a working smart shunt.

There used to be pole mounted commercial loggers out there that would have small cells pointing in four directions at something like a 45 degree inclination that could give you information on total solar, direct/indirect solar, optimum orientation, and a lot more on a minute-by-minute basis. I have not seen them in a few years though.

I think the main draw-back of setting up a real world test rig and logging data is it will only tell you the data for the days it runs, it will not inform you of the long term averages. You will have no way to know if the data you collect is unusually high or low, or normal compared to the real long term average. It will take significant time to collect the data, and yet it's value will be unknown to you.

Example: suppose you set up the logger, and collect data from Sept to April 2023/24. You use this data as the 'worst case scenario' and design a complete system on it's results. Later you realize the 6-months of data collected was actually unusually high for your location but you had no way to know that at the time of the data collection and design. You end up with an undersized system. The value of the data would be improved if you had the time to log for say 10-years, and then base your design on averages, and standard deviations but who (other than a utility maybe) is going to log data for ten years prior to building?

 

[Shimmy]

I think the main draw-back of setting up a real world test rig and logging data is it will only tell you the data for the days it runs, it will not inform you of the long term averages. You will have no way to know if the data you collect is unusually high or low, or normal compared to the real long term average. It will take significant time to collect the data, and yet it's value will be unknown to you.

FWIW, you can correct with weather BIN data for a given location. It won't be perfect (nothing will), but it can help with some error bars at least.

As I thought about my original post, it wouldn't be hard to set up an ESP32 with four photocells to mimic the commercial logger I had seen before.

In a real pinch, you could also just put up a logging weather station with sun intensity, and then integrate for the information you want. It isn't perfect for low sun angles, but it should be representative of how much winter sun (W/m2) you get and its duration.

 

[OffGridForGood]

FWIW, you can correct with weather BIN data for a given location. It won't be perfect (nothing will), but it can help with some error bars at least.

As I thought about my original post, it wouldn't be hard to set up an ESP32 with four photocells to mimic the commercial logger I had seen before.

In a real pinch, you could also just put up a logging weather station with sun intensity, and then integrate for the information you want. It isn't perfect for low sun angles, but it should be representative of how much winter sun (W/m2) you get and its duration.

And it could be fun project!

 

[45North]

Expect snow clearing to be a regular maintenance task, and pitiful production due to very low sun angle for a month or two either side of winter solstice, Dec 22. Optimum winter panel slope is your latitude + 15 deg, so around 60 deg? Very steep which will also help the snow slide off.

 

[JRH]

I live in northern MN. I would like to build a small solar system this fall so I can measure the amount of solar energy I can generate on my site. I want to do this before I build a large 10-15 KW off grid system so I can be sure it will be effective here and I can afford and place the panels. I'm looking for recommendations on a small charger/controller/inverter that has an interface that I can interface to a computer to log power generation when the sun shines. I plan to get a few inexpensive solar panels and a battery or two that will let me utilize the power for things like furnace, water pump etc.. I would then reuse these in the strategic system if I can prove the location will work.

Any ideas?
Bruce

That’s a good idea to try…I hope it works out for ya…. More should do that…
J.

 

[OffGridForGood]

Expect snow clearing to be a regular maintenance task, and pitiful production due to very low sun angle for a month or two either side of winter solstice, Dec 22. Optimum winter panel slope is your latitude + 15 deg, so around 60 deg? Very steep which will also help the snow slide off.

Northern hemisphere: on the winter solstice, your lat plus 23.5 degrees will be the ideal angle for the sun at solar noon.

 

[rcrracer]

Personal Weather Station Dashboard | Weather Underground

Weather Underground provides local & long-range weather forecasts, weather reports, maps & tropical weather conditions for locations worldwide.

www.wunderground.com

Somewhat random, location, date, time frame. Solar radiation near the bottom.

 

[ruceanthony2]

Thanks for all the great input everyone. One of the things I want to study is how much the leafless(mostly) tree branches impact things. In the winter, my neighbors trees interfere with the sun somewhat in the one location I have to place the panels. It's something I can't change and before I buy a pile of panels I want to be certain it is not going to decrease output too much.

You've given me some good ideas to explore! I'll update as I procced.

 

[Mattb4]

Thanks for all the great input everyone. One of the things I want to study is how much the leafless(mostly) tree branches impact things. In the winter, my neighbors trees interfere with the sun somewhat in the one location I have to place the panels. It's something I can't change and before I buy a pile of panels I want to be certain it is not going to decrease output too much.

You've given me some good ideas to explore! I'll update as I procced.

Your prime time is aprox. 9 am to 5 pm or 10 am to 4 pm Winter. If your selected location has shading from trees during this time solar likely is not worthwhile. Small test system may not accurately model your potential since a large system could end up having panels shaded that a small system could avoid.

 

[Madcodger]

As others have noted, the big challenge is often to make sure you have a sufficient load on the system to simulate your actual needs (or a representative percentage that you can use for modeling). If you don't drain down the batteries sufficiently, or have enough load, then your system will just sit there, unable to do anything with the energy you are receiving. This is one of the things most people don't understand when they start down the solar path - that you need somewhere for the energy to go, before you can really understand how much of it you're able to capture. Also, as others have noted, the equipment you select can make a big difference. Inexpensive, PWM chargers and poor quality panels can mislead you. Whatever you decide, I really like the idea of splurging on a Victron SmartShunt to measure actual power in and out. I have mine connected to a Raspberry Pi running Victron's free Venus OS software (you'll need the right cable to connect the two). That will allow you to measure and monitor power production and use from anywhere if you can connect the RPi to wifi. Very handy.

Just so you know before you get started, this stuff can be addictive! I started to set up a somewhat small (inexpensive, I thought) system last year to power a small exercise studio and the lights in a wood-fired sauna at our current home, primarily to give me some experience with solar before we build a net-zero, possibly off-grid house next year at another nearby property we own. But then as I designed it, it just kept getting "better and better". I ended up using Victron equipment, which is fantastic, but holy cow it's expensive. Worth it, but not cheap. In the end, though, I'm very happy I did it and have found a new hobby that I really enjoy.

 

[efficientPV]

It confounds me why anyone here thinks you would need anything more than a simple data logger and a panel. A charge controller would just be insane.

 

[Symbioquine]

I think @efficientPV is right. If your goal is just to measure, then a small panel and a data logger of some sort is all you should need.

I think you just need to log the temperature of the panel and the voltage across a known resistance - ideally not the open circuit voltage. Since the characteristics of the panel are known in terms of voltage vs current at a given temperature and irradiance, you can calculate the irradiance for the area of your panel given the temperature and voltage over a known resistance.

That irradiance data will already take into account the angle and direction of the panel. So as long as your planned array is at the same angle and doesn't have any major variations in horizon/trees, you can figure out the expected yield using the irradiance data. Remember to adjust for;

* Different panel efficiency - if the test panel is a different kind than the planned array
* MPPT/conversion losses - to be safe, maybe assume you'll get 80%
* Storage/inverter losses - some battery technologies and inverters are worse than others

All that is fairly complex, but doing the math is cheap compared to buying and assembling unnecessary hardware. As others have noted it is also non-trivial to create a representative load for the scenario where you create a full test setup with SCC/battery/(inverter?/)etc.

 

[Mattb4]

I think @efficientPV is right. If your goal is just to measure, then a small panel and a data logger of some sort is all you should need.

....

His first post stated,

I plan to get a few inexpensive solar panels and a battery or two that will let me utilize the power for things like furnace, water pump etc.. I would then reuse these in the strategic system if I can prove the location will work.

Therefore his intentions are not just to measure but to have a usable (later reusable components) system.

Based on his post I would recommend something like the beginners system I made a Thread* about. At 1600w and 3kW 120vAC output he could actually power items such as a furnace blower, refrigerator, water pump (depending on type and HP) and small appliances.

*https://diysolarforum.com/threads/matts-starter-system-for-2500.67858/

 

[efficientPV]

"build a large 10-15 KW off grid system" Anything small would be useless in a large system. A data logger and temperature logger are really nice to have. The RC-5 temperature logger can be had for as little as $10 used. Notice the wire coming out. I remove the temp sensor an put it on a long lead so I can monitor temperature of tanks. Coin cells don't like sustained 130F temperatures.

Knowing the voltage on a fixed resistor gives the current. That can be multiplied by the expected MPPT of the panel with correction for temperature. Pretty easy for something like mathcad.