diy solar

diy solar

Solar charge controller not reporting unused solar capacity

As a class of circuits, this is referred to as a shunt regulator. PWM on its own is not efficient. Energy must be stored during the off times in a capacitor bank to be efficient. Heating water is a convenient dump for energy which can be switched in and out very fast. Here is a half hour screen shot in scattered clouds of my array voltage GREEN and the current to the water heater element YELLOW. It remains relatively flat. Slight peaks up are when the heater element can't take all the current. The spike down is likely when the refrigerator turned The voltage is a little over power point because it is measured at my garage over 100 feet away. As this is 1:45 the east facing garage array isn't producing much power. I have to account for voltage drop of the wire. Still, this provides enough to heat the 40 gallon water tank for laundry. A second diverter provides the hot water for the house. This is set at a lower voltage over power point giving house hot water priority over garageDIVERSION_21-09-22.png.
 
The device your looking for is called a battery.

Just with my dry New England pragmaticism I can’t for the life of me figure out why one just wouldn’t buy a AIO with grid support and use 100% of whatever the panels would put out on any given day. Or am I missing something?
 
i don't think even that works. If there is no load, there will be no current, and voltage will be near voc, even in lower light. There simply isn't a way to know what is available, only what is being used.
You could use a pilot panel (small wattage) to determine sun intensity. That panel must be loaded with a resistor low enough so the panel is always in Isc. That makes the voltage to current linear. Then compare that number to what the CC is using
So basically if you’re smart and tinkery you could take a small panel of known parameters and monitor the variable output to input data to some processor that also monitors the main power system and use code to output comparative tables or more basically a flat delimited data set that you could export and open in excel or whatever.

That’s sounding like fun but I don’t understand the practical aspects. My goal with solar is to exceed my needs and store a day or two of power for use when the sun doesn’t shine. Nothing wrong with experimenting for fun though. I’ve done plenty of that with series, parallel, vertical, refrigerators on 200W…

From a practical perspective I’m not sure knowing what the potential you’re missing is. Knowing that you are not meeting a particular need and the time period and load of same is enough to extrapolate how much more you need. Daily conditions vary so much and panels are cheap enough just having extra panels doesn’t wind up at chapter eleven.
 
How about using the data from a nearby weather station from https://wunderground.com that tracks the solar radiation/solar irradiance in the form of W/m2. That would tell you what your potential is at any given moment and from there you can decide what to do with that information. Of course there could be a cloud nearby that is affecting your panel and not the stations panel but if you find a station close by, it might not make a big difference. Especially, since you haven’t really said what your going to do with that mysterious unused power that you don’t need when your batteries are full and how that will actually benefit anything.

Here is an example of a station that keeps track of the solar irradiance.

Or I suppose you could spend the money on equipment like they use to measure this irradiance but honestly, why bother when the data is already available at a local station near you for free?
 
How about using the data from a nearby weather station from https://wunderground.com that tracks the solar radiation/solar irradiance in the form of W/m2. That would tell you what your potential is at any given moment and from there you can decide what to do with that information. Of course there could be a cloud nearby that is affecting your panel and not the stations panel but if you find a station close by, it might not make a big difference. Especially, since you haven’t really said what your going to do with that mysterious unused power that you don’t need when your batteries are full and how that will actually benefit anything.

Here is an example of a station that keeps track of the solar irradiance.

Or I suppose you could spend the money on equipment like they use to measure this irradiance but honestly, why bother when the data is already available at a local station near you for free?
That would work if your solar panels were flat level on the ground.
 
That would work if your solar panels were flat level on the ground.
Well then, put the panels flat level on the ground and the problem will be solved. Obviously this unused power isn’t being used anyway so what’s the difference? It’s all speculative anyway at this point. I don’t get upset when the earth rotates and my panels are no longer in the sun. I don’t wish I could have my panels circumnavigate the earth just so I could capture all of their potential. Especially when I don’t need it or use it. This is why we size our systems to our own individual needs. I’m wondering how much measurable potential I’ve lost by reading and commenting to this thread? ?

 
Last edited:
An excerpt from the manual appears below, taken from the manual at this link. http://www.bogartengineering.com/wp-content/uploads/docs/QuickReferenceGuide.pdf

View attachment 81012
"Display yes ..."
This is because it isn't possible to know the available power, only that most of it is. But, say some clouds cover while nearly fully charged. You may have very little available, but only know that all of it is available, not the quantity.

If there is a large load on the panel, then it can be estimated by comparing generated V and I with the panels Voc an Isc. This is when it can display unused watts. But this only works until the battery is charged.
 
The topic of this thread is not what to do with extra power but how to measure it.
The only way to reliably measure it is to apply sufficient load for long enough for the MPPT to catch up and establish a true maximal power point.

There are other ways to estimate it, either via a secondary system with an "infinite" load (e.g. a grid tied system with no export limit), or with a predictive solar output service (more on that later).

I don't think there is any technical way for the solar controller to know how much power is available. There might be a way to calculate it if the solar controller knew the wattage of the panel, the voc, and isc.
as well as the sun ray's angle of incidence, the ambient temperature, clouds and other atmospheric conditions, the nature of any array shading at that moment, the age of the panels, how clean the panels are (did a bird shit on them today?), the wiring, the DC to AC ratio of the charge controller (many are "over" panelled), and with grid tied systems even the grid's voltage can result in PV output changes, etc etc all of which will impact the potential output of any given PV array at any moment in time.

This 'tester' would only run for seconds and therefore I don't think it needs a huge load like everyone would think, high power MOSFETs like identified in the below video are all that may be needed.
How long before an MPPT settles into the true maximal power point when load or insolation changes occur?

Just with my dry New England pragmaticism I can’t for the life of me figure out why one just wouldn’t buy a AIO with grid support and use 100% of whatever the panels would put out on any given day. Or am I missing something?
I guess not everyone is able to connect a system in such a way that export to the grid is permitted. For off-grid DIY, drawing power from the grid to support your own production is usually OK and done with minimal regulatory fuss, but exporting energy to the grid comes with a different range of regulatory requirements/restrictions.

How about using the data from a nearby weather station from https://wunderground.com that tracks the solar radiation/solar irradiance in the form of W/m2.
Much better is a service dedicated to predicting and assessing the theoretical insolation for a given array specification and detailed satellite imagery, for example Solcast:

These are not perfect tools but are used by grid scale solar farms where this stuff matters. They provide a free level of access for small chumps like us. The free access level doesn't permit for unique system tuning (e.g. to account for a local shading profile) but as a general indicator of what your system potentially could have produced during that interval as well as predicting what it could supply in the coming hours and days it's pretty darn clever and integrates many factors such as those I mentioned earlier.

I am assuming because this 'feature' is not available on any charge controllers I looked at that there is more involved than a software update to monitor the solar panel voltage/current and use a solar algorithm to do this? It seems like this is a VERY necessary item for self-consumption.
Having more solar PV capacity than is needed at times is a feature of off-grid systems, not a bug. Arrays are often sized to satisfy demand during periods of low insolation (e.g. Winter, rainy days) which means more often than not potential output on good solar days is just never used.

Finding ways to usefully access and use that unrealised capacity may or may not be worthwhile. That's really dependent on each individual use case.

My grid tied system is always operating at its maximal potential since it can export excess to the grid while my off-grid system rarely reaches its output capacity. But today is rainy and I need that off-grid PV capacity to keep the pump going and reduce the rate the battery discharges. It's pulling everything it can from the array, which at the moment is only about 10% of the array's nominal capacity.
 
Last edited:
Arrays are often sized to satisfy demand during periods of low insolation (e.g. Winter, rainy days) which means more often than not potential output on good solar days is just never used.
That’s how I see it.
I’m way overpowered for most summer days but a lot of winter days there’s nowhere near enough to reach capacity or fully charge, but I can run all day without supplemental power.

The extra panels and hardware are just the one-time cost of getting by.
 
Just with my dry New England pragmaticism I can’t for the life of me figure out why one just wouldn’t buy a AIO with grid support and use 100% of whatever the panels would put out on any given day. Or am I missing something?
I am from California and I can't figure it out either. I never have unused capacity beccause what ever I use goes to the grid. The grid is the best infinite load that I have found. I was delayed 14 months on getting a PTO and I don't know how my system got turned on but it was. I did not get any credit during that time but it offset some daytime loads and I charged my EVs as often as I could during the day.

I am still working on ways to self consume as much as I can but that is another issue.
 
Last edited:
I am from California and I can't figure it out either.
As i said earlier, the requirements and regulations for being able to export to the grid are highly variable. In my location there are export limits, and no DIY grid tied inverter will ever be approved for connection, that's for off-grid only. There are strict regulations on what the transmission company will approve and who is permitted to do the installation.

In some locations there is zero grid export permitted.
 
These are not perfect tools but are used by grid scale solar farms where this stuff matters.
Don't grid scale solar farms have an infinite load of the grid unless curtailed by a system operator? To what purpose would they use that data? I suspect their investors were given projections or forecasts. The typical financial report includes comparison to budget or forecast.
 
Don't grid scale solar farms have an infinite load of the grid unless curtailed by a system operator?
Curtailment occurs regularly.

To what purpose would they use that data?
If you are bidding to supply the wholesale electricity market, it kinda helps to know how much power you think you'll be able to supply. Else you end up with either a glut you need to sell off on the spot market for a loss (or curtail) or have to buy energy on the spot market also at a loss to make up for what you have been unable to supply because your supply projections were out.
 
I've wondered the same thing. Did I install too big of a solar panel array? For example, say I consume 2000 watts daily. But my array is a 10,000 watt array (capacity to produce that much). Sure, I could store a lot in a battery bank, but for argument sake or point of this question, lets assume my batteries are topped off at 100% SOC. The 10,000 watt array is not being used to its potential, right? Now, that's 10,000 watts based on the size of the panels (say they are 300watts each panel x number of panels = 10,000 watts, that's rated power). But, the sun , weather, clouds, my panels aren't able to make what the sticker on the back says they can make. So, I think I would need a sun irradiance sensor or meter installed somewhere out on my ground mount rack, near my panels, at top, middle. That sensor would report real-time what amount of sun is coming in, and based on size of panels what they would produce based on the amount of sun that's coming in at the moment (real time, updated every 1 second maybe). So, the sensor tells me what amount of power I SHOULD be able to make at the moment, which is not the amount of power I am using. The amount I am using is my load at the moment, it fluctuates of course, someone runs a light, a hairdryer, toaster, and the load changes all day long throughout the entire day. If I knew from the sun sensor mounted out at my solar panel array what my array COULD be making max at the moment, minus my current usage (load), then that difference would be what capability or power I have there that I could be using , but am not.

Does this make sense? I don't net meter (sell to grid). I am grid connected, but I like to use grid as last resort when my system cannot produce enough sun power to meet my demand. But it'd be nice to know, and report say over time, day, week, month, year of how much power I kind of let slip by me (without using it, or storing it in a battery). This data I think would be helpful. I would know if my array is over-sized, under-sized, and I would know if buying more batteries was a good idea or not, and I would also know how much extra unused capacity I have in my solar panel array to either buy another inverter (if the existing inverter is maxed out), or buy a water heater tank and pre-heat water so my tankless gas water heater doesn't have to work so hard (and use less gas in the process).

The only current (no pun intended) way for me to figure out if I have excess/unused power/capacity in my system is using math and some reports from PowerView which would tell me how much I'm using, how much load I am covering (or not), and knowing the size of my solar panel array and what it CAN make on a good day, and kind of make an inference as to what my system is doing and is capable of, but its not all that accurate because its theoretical based on wattage ratings of the panels, the estimate of sun hours for my zip code, and is not a number based on real-time data that a sensor would provide. The idea above with a sensor though would be almost real-time, almost down to the watt in showing me what I'm making, what I'm using, and what capacity is there based on my solar panel size (wattage) and the amount of sun is coming in. It takes a solar panel + sunshine to make power, and that known power is real time, its for the moment (the moment the sun is shining and no clouds, etc). Using a sensor to send that back to the inverter and the inverters firmware can take that number (the available capacity) and do the math to minus out what's being currently used, to show available array capacity would be I think a good number to know and be able to report. I don't think anything like this currently exists. I think its probably done during the design phase of taking the size of your solar panel array, and figuring your known loads that you want to cover, and building a system to meet the demand of known loads and once that's met, the people that sell and install equipment know they have hit the goal post and stop there. There's probably a ton of solar systems installed daily that are over-sized , yet they meet the designed load, and the owner is happy the system works as it should, never knowing that one or more parts of their system is slightly or maybe even grossly over-sized. If under-sized, they'd know it sooner or later, dead batteries, unpowered loads, equipment failing to start, etc. But over-sized, they may not live long enough to discover their system was and is over-sized. Please correct me I am wrong. I'm not an expert in solar, just a newbie with a lot of ideas and thoughts and not sure if some are valid or not but feel open forums like this are great for inducing thought and maybe even figuring a few things out.
 
So, I think I would need a sun irradiance sensor or meter installed somewhere out on my ground mount rack, near my panels, at top, middle.
Use the function of a voltage control board / relay to only send ‘bonus’ power out when yo battries are fully charged
 
My Sol-Ark has a feature called a "smart load" which does just that. It has parameters where I could send say for example 500 watts to a water heater any time my batteries are over x% SOC and there's excess PV coming in. When those conditions are met, it sends the power to the smart load. I haven't used the feature yet but want to try it once I can figure out if I have excess power that I'm not using. I could try it and if it works, it works, if not, then I know it doesn't work. Anyway, I think the feature works were I set a parameter say 1500 watts , if battery is at 80% or better. Then if the water heater element uses say 1000 watts, it only brings over the 1000 watts, the 1500 parameter setting is a max output if the conditions/parameters are met. Seems to be a pretty powerful feature but I haven't used it yet.
 
The only current (no pun intended) way for me to figure out if I have excess/unused power/capacity in my system is using math and some reports from PowerView which would tell me how much I'm using, how much load I am covering (or not), and knowing the size of my solar panel array and what it CAN make on a good day, and kind of make an inference as to what my system is doing and is capable of, but its not all that accurate because its theoretical based on wattage ratings of the panels, the estimate of sun hours for my zip code, and is not a number based on real-time data that a sensor would provide.
Solcast:

Real time estimated output and forward forecasts based on your exact PV array(s), location, and environmental/weather conditions at the time, and forecasted on a half hourly basis for the next 4-5 days, continually updated.

They have an API - I have mine integrated into Home Assistant.

Screen Shot 2022-04-03 at 1.44.19 pm.png

This was yesterday's actual (columns) and forecast (dotted line):

Screen Shot 2022-04-03 at 1.46.42 pm.png

Here's the expected daily output for next week:

Screen Shot 2022-04-03 at 1.50.47 pm.png

These forecasts are constantly updated, but you can see that at the moment. 5 April looks like a decent day but the rest of the week doesn't;t look so good for production.

And this is all free. You can of course subscribe for more features (e.g. fine tuning) and for more arrays but frankly, that's not bad.
 
Yep this is a huge step forward in knowing what your array could produce, in the future, and at the time, it appears the forecast and actual are darn close. I still think a sensor mounted somewhere on your array would be the absolute best real-time data, but we're getting there with what you just provided. Thanks. I don't use Home Assistant, but I think its a sister product for Solar Assistant, and has built-in solar features as well.

Back to my thoughts of a sun sensor, I don't think or know if something like this would work but this is what I kind of had in mind, a sensor that would be wired back into my Sol-Ark, but its firmware would need to expect this data and be able to handle it of course.

Obviously a hand-held unit wouldn't be mounted outside in the weather, they are just not made for the application I have in mind, but just thinking about what they measure and that a sensor does in fact exist on the market, means this could be done if someone could get the right meter, bring the data into a computer (pc, ipad, web, etc) and then have software to manage that incoming data appropriately.

But it kind of would get more complicated than that. Having a sensor outside on my mount is only one piece. Then either my inverter, or some device (processor) that has software that takes data from sensor, the number of solar panels I have in the array, their spec'd data (voltages, temperature, amps, etc etc), all entered into the software so it could make an intelligent calculation of what capability I have out there on the rack for any second of the day. The software I think would have to bring in data from the inverter to know how much power I'm using, to minus that from how much I could be making above-and-beyond what I'm already using. The software might even need to know how much battery bank you have and its contribution I don't know. It could get complicated. From there, reports over time day/week/month/year could be ran to show highs and lows, get a min/max on how much power I could have made over a given 6 months and that would be a pretty solid number to base decisions on, like do I need more panels, do I need more inverter, or maybe more battery, do I have excess power that I could send to a water heater (or not), and so on. As I mentioned, I think many people gently build a solar system over time, getting something in place, then running it to see how much they have in excess or are lacking in one area (maybe number of solar panels, maybe inverter, or battery) and make adjustments over time. That's more of a trail and error method. If one had a sun sensor with reports, you'd know a pretty big piece of the puzzle I think. It's probably one of those things that could be useful, but its not marketable and therefore nobody is interested in making it.

I think I'll stick with trial and error for now, and take a closer look at Solcast API. Kind of wonder if it would work with PowerView but I imagine the programmers would have to build something for it to work. I know what an API is, but I'm not a programmer. I took and passed C++ in college and that taught me that the side of my brain I rely on doesn't lend me to being a programmer. Coders use a different side of the brain than I do.
 
Back
Top