Hello everyone and happy holidays!
I’m interested in photovoltaic panels, it’s the future and all!
But with the subsidiaries and the general enshittification of search engines, all search results about photovoltaics leads to sites with wildly misleading information, IMO.
I don’t care about a 3kWc system with installation. What even is a kWc (I know what it is) and why is nobody explaining how much power the panels would typically yield instead? Per month? During the day?
I guess it is less selling if your installation is generating near nothing in December when you need it the most?
Okay sorry, rant off. My question is, where can I find reliable information about how much panels generate every month, during the day?
I know places have more or less sun, but that’s quite easy to figure out if you have the numbers for any place.
🌞
Edit: I don’t need a web calculator for how many panels I need. I’d like to know roughly how many watt a typical panel produces a specific day (or better hour) in the year.
Edit2: I am not looking for how to install or calculate a typical solar panel setup. I’m looking for the typical real world output of solar panels around the day and year.
Edit3: got my information, thanks oo1@lemmings.world ! You all can now continue explaining how many panels a home needs or what a kwh is, Merry Christmas to you all!
There are web calculators where you put in your latitude, angle of the panels and total kWp of your installation. It then spits out a kWh prediction for the year. Might still be shitty to find a good one tho. I can tell you that the system i installed at my parents house with 10 kWp has produced 8.4MWh of AC output this year. I live in southern Germany which is around 48° latitude and it was pretty gray and rainy this summer so could be much better.
This is daily total generation in kWh split up by how much went into the battery vs directly into live usage in the house vs exported to the grid.
This shows the sources of all the electricity that the house used over the year on any given day. Red being imported electricity.
This. There is too much local variation in sunlight angle and weather to give a straight answer. An easy method is to take the rated output and multiply by 0.2, but even that is merely a rough average over a year.
But there should be data on weather and climate variations. So theoretically you could include that data into the calculation. Theoretically. Who’s gonna do it?
There is. I do it, it’s my job as a solar engineer.
Basically, there are several leading softwares that solar engineers use to account for just about anything that happens in the real world.
I mainly use PVsol premium where I 3d model each site and the panel placements and electrical components and so on, then run a minute scale simulation based on the exact location weather data (using Metronorm 8.3)…
Almost no one outside my field understands what goes into my job. It doesn’t help that there’s a lot of untrained people pretending to do what I do…
I was having a smoke break with some colleagues once and was talking about how it should be possible to simulate solar panels in 3D, to account for occlusions by roofs, other buildings, trees etc. Didn’t know there was a dedicated software and job for that, that’s so cool!
That’s what the calculators are for. This has been done.
Yeah, solar panels put out power in proportion to the light that hits it and its efficiency. The latter is in the specs but the former requires knowing how it will be installed before you can determine expected output.
Some calculators can also consider weather predictions (cloudy days, etc)
Some calculator sites;
Thanks but link 2 and 3 doesnt work for non US/UK? Adress needed
Link 1 seems completely useless, like no I don’t want to know how many “panels” I need for an installation.
“Panel” in most cases means “400W nominal panel,” which may include higher-efficiency, same-size panels with 420W nominal, so you can just math whatever panel numbers they give you by 400 to get the answers you want. Like, if they tell you a “10 panel” system will generate 3240 W, you can figure that means 75-85% of nominal peak power. A lot of the calculators are meant to help sell installations, based on people’s current electric usage and constrained by their roof area. That makes ‘number of panels’ a very handy measure.
Yes but that is not the information I am looking for, I have edited the question forore clarity.
For every month? That would be helpful, just then need to convert the energy (kWh) to power (W) which is easy.
You can’t convert the kWh to W, that’s not how it works. The amount of sunlight is highly variable during the day and the way the sunslight gets converted into usable energy differs a lot depending on the installation.
Normally those calculators assume you can connect the solar installation to the grid and use the grid as a buffer. So when there is sun and you’re not using it, you deliver the power back to the grid. And when there is no sun, you get your power from the grid. When there is a little sun, or you’re using a lot, you use some energy from the solar panels and some from the grid.
It starts getting complicated depending on how your local grid works. Often supplying back energy to the grid means the power provider credits you a certain amount. Depending on your contract, this might be a day price or even hourly. Or it may be a fixed price. This often means you pay a lot more using power from the grid then you get back pushing power back into the grid. So you need to think more in terms of money and using the energy in a way that’s cheapest.
If you are thinking about an off grid installation, the amount of power is almost always dependent on your equipment and not so much the solar panels. For example you can charge up a large bank of batteries from the sun during say a week. Then when fully charged, you can draw huge amounts of power from them till the bank is empty. But depending on the batteries used and the inverter used to convert DC into AC, the amount of amps it can push can be limited.
When thinking of something simple like a use case where you directly use the DC from the solar panel, the panel specs always include the Wp value. You can use that to calculate the exact amount depending on your location, time of day and angle of the panel. Weather services these days also include a watt per square meter of solar energy for different locations, which is useful. And keep in mind it only works when it’s sunny, with clouds the output drops a lot.
I’ve seen huge swings in my pv installation year on year. So it isn’t a sure thing how much energy you get from the sun.
You all explain how a classic solar panel installation works which is, as stated, not the information I’m looking for.
To convert kWh to watts you divide by time. That is how it works. 1 watt for 3600 seconds is a kWh. 1 kWh collected during 3600 seconds averages out to 1 watt.
The k means 1000, 1 kW is 1000 Watts, 1 kWh is 1000 watt hours. Etc.
This is basic stuff people.
Edit: internet at its finest, downvots the correct information.
You aren’t posting correct information, you are making assumptions that are unwarranted and therefor claim stuff that simply isn’t true. Most things aren’t true or false in an absolute way when it comes to complex systems, they are only true under specific conditions or when taking into account certain assumptions. Often it’s easy to generalize and handwave away small details, because they don’t matter for the end result we are going for. But you then can’t say this is a fact and works in all cases, sometimes the details matter a lot.
When it comes to solar installations, those are very complex things. That’s why you only find the resources you found, they are meant to convey a specific understanding of a complex system as it applies to the general audience. For example how many panels should I get, or how much money will I save on my energy bill if I get solar. If you go into the nitty gritty of solar installations it gets complex super fast and small details can matter a whole lot.
For example I have a micro-inverter system where each inverter doesn’t take full advantage of the capabilities of my panels. However due to my location and the angle of my panels, the panels don’t reach that peak capability for almost all of the year. Maybe in perfect conditions on a couple of days a year in the middle of the day, I go over the peak. I could change the angle, but that means more ballast is required for the bigger wind profile, which means I need a structural calculation done which costs money. So I opted to go for cheaper inverters, saving something like $300 on the whole installation, but that means I miss out on about $100 over the lifetime of the system in generated energy. Since that’s a nett positive of $200, that’s what I went with. I could have optimized a bit more to make use of the low-end of the solar output better if I went with a single inverter system instead of the micro-inverters, but that means a large device in a place that was inconvenient and the pricing calculation was tight.
The reason you aren’t getting a straight answer is because it appears to be a XY problem situation (https://en.wikipedia.org/wiki/XY_problem). You are asking for something that doesn’t really make a lot of sense to ask. So it’s likely you are actually trying to figure out something else, have come to the conclusion this is the information you need and thus asked the question. However that conclusion probably wasn’t correct, as the question doesn’t make sense. So it would be really helpful if you ask the initial question you are trying to figure out and let people help with that instead.
When talking solar it would also really help if you state where about you are located on the planet and if you feel like sharing which country, as each country has very different ways of structuring the pricing of energy and that matters a lot when it comes to solar.
As posted below:
Just FYI I’mnot looking for a classic solar installation.
You all seems to base your answers on that, so obviously communication gets complicated.
I needed ballpark numbers for robotics and battery size calculations, napkin figures, to get somewhere to start off of.
I got my answer, as I edited in in the question.
Thanks BTW and sorry if you spent a lot of time trying to help me, I have definitely learned enough to set up solar panels at home on top of it all.
Alright, now we get to the real question, that’s something we can help with.
How big is this robot going to be? Because to tell you the truth: solar panels kind of suck. If the angle isn’t perfect and you don’t have a lot of square meter, the output is terrible. That means putting them on mobile things is kind of hard. There usually isn’t a lot of surface area available and almost always the angle to the sun is going to be terrible. And keep in mind they don’t work in the shade. So inside won’t work, in between buildings won’t work, under some trees won’t work, a bit of clouds and it won’t work etc. So you need a big ass battery onboard to buffer energy, so the robot can charge up in the sun and then run on battery power the rest of the time. Then we get to weight, those regular 400Wp panels you were calculating with are regular roof panels. Those have a aluminum frame, plastic back and glass front. They are around 120x180 cm and pretty heavy. They are also sturdy but can break easily since they are made of glass. For mobile applications you need to look at RV panels. Those are usually flexible, rugged and much lighter. The spec you are looking for is Watt peak. That gives you the amount of watt the panel produces in perfect conditions. You can then calculate the losses you get in your situation because it won’t be perfect. The batteries used in combination with solar are normally lithium iron phosphate. These can store huge amounts of energy and are very safe. Plus they last a long time and can do a lot of cycles. However they are very heavy and to get enough energy storage you need a big ass battery. They are also very expensive. For robots you are probably going to want lithium ion instead. These are cheaper, lighter and can still store a lot of energy. However they are very dangerous to work with and wear out faster. In a mobile application to prevent losses you also don’t want an inverter that converts stuff from DC coming from the panels to AC. It’s a lot better to stay at DC all the time, for example 24V. That way conversion losses are kept to a minimum. But you’d still want a good inverter with builtin battery charger, as the voltage coming out of solar panels fluctuates a lot. Just connecting them directly to batteries is a really bad idea. Solar panels also don’t like it if you draw as much current as possible, that ruins their efficiency. Because of how they work, there is a sweet spot in the voltage and current. Be sure to get an inverter that takes this into account (keyword mppt).
Take a look at something like this panel:
https://www.amazon.com/Flexible-Monocrystalline-Semi-Flexible-Trailer-Surfaces/dp/B0BQ1Y8JMH
If you get three of these you can get something going. If you do it right you can get about 100W of energy under perfect sunny conditions with 3 of these panels combined. That isn’t a lot of energy.
That’s why you don’t really see solar panels on anything mobile, it really kinda sucks. My advice for solar power robots: put a solar installation on the roof, put all of the energy generated over a year into the grid and simply charge the batteries for the robot from the grid. This way you can’t just charge when there is sun and if there is sun and you don’t need to charge the energy isn’t wasted. And being on the roof they are sure to be at a better angle and without any obstructions. They can also be as big as the roof, which helps a lot. This is what we do with electric cars and it seems to be the most efficient we can get.
Thanks, this information is really useful!
Okay so here I go, this is where I am at the moment, I have this old robot idea based on the hitcher robot (“HitchBot” for the curious) and as I’m somewhat aware of my shortcomings I’m trying for a very simple start robot and build on top of the experience from that one.
So the first robot would be, and don’t forget this is just napkin figures to get a first iteration, 40 cm long, 20cm wide (16x8"), and not too tall.
Inside would be, roughly, a raspberry, a camera, some 18650 batteries, two motors for moving (probably NEMA 8 stepper motors), electronics (controling the motors, charging the batteries) and a solar panel. I’ll have a usb charge module for convenience too and some way to measure the solar panel charge (so the robot can angle itself as good as possible). Maybe a servo for the camera angle, maybe one for the solar panel.
Connected over wifi for starters, 4G in some distant future.
The 18650 could theoretically hold like 10Wh each, and one could drive the PI for some 10-20 hours (or weeks sleeping) or the motors, hopefully an hour or two.
Theses calculations are wrong ofc but hopefully not too off.
The robot would roam around in the summer time, in France, so lots of potential sun. Battery power to the rescue when stuck in the shadows.
Just extrapolating the numbers you just gave (3 x those 100w panels could give 100w in ideal conditions) gives 8 watt for a 20x40cm panel in ideal conditions.
So the robot would have to charge for 2 hour for a 1h of operation if losses are not more than some 35%. Not very good but somewhat usable. If it can charge in ideal conditions which is of course not the case.
So maybe 1h of operation a day?
If so, then I’d be quite happy for a first iteration.
Thoughts?
Solar panels don’t generate constant power! That’s why it’s wrong because you’ll get wrong numbers. You can’t assume you’ll get peak output continously.
You *** must *** calculate incoming and absorbed light. The Watt output will vary continously as the sun moves and weather changes. If you have average kWh / day stats and a battery solution that can store a day of battery, THEN you can calculate average Watts if weather doesn’t change too much.
Just FYI I’mnot looking for a classic solar installation.
You all seems to base your answers on that, so obviously communication gets complicated.
I needed ballpark numbers for robotics and battery size calculations, napkin figures, to get somewhere to start off of.
I got my answer, as I edited in in the question.
Thanks BTW and sorry if you spent a lot of time trying to help me, I have definitely learned enough to set up solar panels at home on top of it all.
Removed by mod
Removed by mod
I edited my comment and added a screenshot from my grafana dashboard to show the trend over the year and some other numbers. Batteries are expensive but they are worth imo. ~75% of the electricity usage of this house with 6 people comes from its own solar production. There is however a cut off for how much battery capacity makes sense. To get the last 20% of self sufficiency you would need a disproportionally larger battery to make up for long periods of low sun. so 80% is as good as its gonna get while staying cost effective.
I did a calculation for a 20kWh battery for my installation, which would cost me about 18k to purchase and setup. I can do a lot of the work myself, so the equipment costs are the bulk of the price. Figuring a prediction in the price of energy, my usage pattern and a lifetime of 15 years for the batteries, it figured out to be not worth it.
In winter where I live I’m totally fucked, there is basically no sun at all and it’s cold AF so energy usage is highest. Any battery would be empty within days and not really able to charge it because the little bit of solar that might be available is used right away in heating. So it would be down to spring and fall where there is still some sun and also some request for energy.
For a house of 6 people we installed 5kWh which is basically perfect for them, but they are also very resourceful with their electricity usage, running washing, cleaning and cooking during the day as to not drain the battery at night.
But yeah if you dont get enough sun in your area its just a waste of money. My parents system will be payed off after 11 years if the sun stays at the level it has been at for the past 2 years.
Predicting how import vs export costs might develop is also a big factor
It’s very circumstantial, slope and orientation make a big difference. so it has to be a calculator where you put in specific information.
I think this does it: https://pvgis.com/
There’d nbe loads of others. I’d expect most reputable solar installers aimed at residential to have calculators, or recommend one as part of the planning process. Maybe with localised assumptions. There’s one from a govt funded body in my country that makes cost and price assumptions too tand gives a ‘return on investment’ guess.
Wow that’s kind of technical :-) but monthly irradiation is accessible!
Edit: according to those figures a 1m²@20% panel would yield 55w averaged out on 24h, in the summer, and 14w/m² in December.
That’s numbers I can work with, thanks!
The solar atlas is another good starting point for making these calculations. Just by looking at the map, you can easily see how location makes a big difference. Solar power in Spain will be pretty good, while in Germany it’s a bit meh. If you’re in Scotland or Norway, solar power will be even worse. Well, you can always compensate by buying more panels, but that’s not great either.
The actual output is also greatly influenced by quite a few variables, like angle, and efficiency of the panel. As the panel ages, the efficiency goes down. Also, higher temperatures decrease the efficiency of the panel, so the burning hot panels in Libya might not produce as much as you thought based on the solar irradiance map. In other words, it’s complicated.
Check out this: https://globalsolaratlas.info/map
Be sure to check out the details, it has the exact specs for your location.
That’s a neat tool. But it’s giving me a slightly confusing result. I have a solar installation and I’ve plugged in the details so far as I know them, just to see if I’m producing about what I “should” be. The peak production month is about right, but the minimal production month is only estimated to be like 25% less than that. My system has more like 50-60% drop, and some quick googling suggests that’s about normal.
Any thoughts on why this tool suggests a much smaller drop?
No idea. The ‘typical meterological year’ is likely smooth some extremes - that will likely have less variance than any actual year. Maybe the geographical resolution is poor leading to more averaging. But that sounds a bit large of a difference to be just that. Was the total annual production way out?
Check some other tools , maybe a local one - another one might have better data on some things. Some consensus of several estimates might be better than relying on one calculator only.
It’s not hard.
Take the rated wattage. Adjust for system wiring loss, shading, age, soiling, etc. most importantly adjust for location. Your angle of sun makes a difference.
Pvwatts is a good tool for this.
