What to do with afternoon PV watts?

kaipo_boy

I've lived with my small battery tied system now for about a week, and noticed the charging algorythm goes like this: In the BULK stage the voltage steadily and slowly goes up over a couple hours in morning sun, current is more or less steady as the batteries are depleted from a long night... when the voltage hits the set point on my Kid, the cc switches over to the ABSORB stage and holds voltage steady at the setpoint, with current steadily decreasing until it hits FLOAT (well, sometimes; we've had lots of clouds lately).

So, in my area this transition to Absorb is roughly 1pm ish... on a mostly sunny/mixed cloudy day. Assuming there is enough sun to make the Absorb stage, I've been wondering; is there any way to utilize the excess current now that the cc has to shed it since the charge algorythm calls for only a decreasing amount of current and fixed voltage? There are still a couple hours of useful sun left at this point and the sytem might make 600w or 700w in bright sun in the afternoon.... but in latter stages of Absorb it looks like cc is only making use of a couple hundred watts. The rest appears to go to waste unless I'm home and can organize some opportunistic loads like doing laundry etc. Then if I have a load on, I notice the watts the cc is passing increases but only by about the amount I'm loading plus the Absorb requirement... as you'd expect. The cc is continuing its charge routine, and adding in as many watts as my load is calling for. This means potential PV production is being shed off and wasted as it can't be absorbed by the battery and can't be used if I'm not home to post a load. Does anyone have any ideas on how to use this trailing part of production efficiently (aside from opportunistic loads)?

CALLD

This dilemma is as old as the off-grid system itself, and does have some very good solutions depending on how far you're prepared to go to use those extra watts.

If you're happy to spend the money, get a PWM voltage dependant relay that can send any extra PV watts to a DC element in your hot water cylinder. It will cost you some money and some clever programming of the system.

Alternatively you can hook-up appliances that naturally use more power in the afternoon than the morning - fridges, AC systems tend to draw more power in the afternoon when it's warmer.

Mountain Don

I gave up worrying about the lost PV energy. I did think about using the excess PV to heat water. After researching that I came to the conclusion it was not worth my effort. YMMV. I get much better water heating by using the sun to directly heat water in its own system. So yes, many days I am sure there may be more energy "lost" than what is actually going into the batteries.

CALLD

Mountain Don wrote: »

I gave up worrying about the lost PV energy. I did think about using the excess PV to heat water. After researching that I came to the conclusion it was not worth my effort. YMMV. I get much better water heating by using the sun to directly heat water in its own system. So yes, many days I am sure there may be more energy "lost" than what is actually going into the batteries.

Yes, it makes sense not to worry about "lost production" when you have a small array compared to your battery bank as there won't be a significant amount going to waste. With your 624w array and 24v420Ah battery you are probably never going to see much more than 22amps or thereabouts going into them. That's not much more than 5% of the battery's 1hr capacity and as such you will never have much to spare during absorb. With the high price of batteries and low price of PV it's becoming much more common to have large solar arrays and smaller high-rate batteries like AGM's. It allows you to make the most of intermittent sunshine on cloudy days and of course helps a lot in the dead of winter. The trouble is on a sunny summer day you have a lot of spare power begging to be used...

The way I've dealt with this problem is by "overloading" my system with far more appliances it would be able to continuously power. Appliances that consume power during the day like fridges, freezers and cooling systems are very useful. You need an alternate power source for overnight obviously. I use a 2 pole changeover switch to switch between my inverter and the grid in the morning and the evening. So I'm effectively able to harness almost 90% of my 1490w solar array's power with only a 24v260AH AGM battery bank while cycling it no more than 30% of its capacity.

Why don't I go with grid tie If I have access to a power grid I hear you say? 1) because grid power isn't reliable in South Africa, 2) because I can't sell excess back, 3) because I like being independent of the grid.

If you have no alternate power supply to run overnight then unfortunately there isn't much you can do with those extra watts other than try to go sleep at sunset and wake at sunrise and try to use most of your power during the day...

kaipo_boy

Very sensible solutions, thank you!
Calld, my system is 3x 285w Trina panels and 4x ~208ah 6v batteries for 24v, 208ah battery + Midnite Kid + Cotek 1.5kw inverter but the system is only pieced together right now for a week or so, I haven't put down hard lines or anything, everything is only temporarily wired together to see how they will play with each other to determine if I have to make changes. I actually have 6 panels as I was going to make 2 of these small systems but aside from the PV purchase, have not yet started accumulating the other equipment for the 2nd system. I have grid access but like you, I hate being dependent on it.

I was thinking of adding one more panel to the existing 3 so that PV production gets boosted... but then noticed the production in the afternoon wasn't being utilized; so adding another panel would only aggravate this. I was originally running it with only 2 panels and it did fine in bright sun but as soon as the weather turned a little cloudy, reaching Absorb was not a possibility. I am unsure if I've reached a good equilibrium yet, I'm still monitoring the SOC of my batteries and like you, switching back to grid (manually, by moving the cords over) if the SOC goes too low in early morning. I am only loading the system with about 2kwh/day of loads (1 chest freezer, 1 fridge) and still playing a bit with settings on the Kid (playing with Absorb voltage and minutes settings) and guessing at SOC by watching the batt voltage... my refractometer should be here in a few days.

Mountain Don

CALLD wrote: »

when you have a small array compared to your battery bank as there won't be a significant amount going to waste. With your 624w array and 24v420Ah battery you are probably never going to see much more than 22amps or thereabouts going into them. That's not much more than 5% of the battery's 1hr capacity and as such you will never have much to spare during absorb.

For us this system has worked well. I do know by all rules of thumb we should have more PV. In reality we should just have less battery. However, the battery size was intentionally made larger than needed to cover the cold winter weather. The GC-2's were cheap enough at Sam's Club in 2009, and still going. The cabin sits unoccupied most of the winter and then gets used for a few days before sitting for a couple+ weeks. I did not want to be forced to run a generator unless weather was cloudy for a prolonged period as this is 100% off grid, so much so that winter access is by snowshoe. Also our NM location is blessed with mainly sunny weather with approx 320 days classified as sunny and even in winter we have more sun than many others have in summer. The elevation of 8800 feet gains a slight bit of amperage due to reduced atmospheric loss too. We are usually into float by 2 PM or so on an average day. When these batteries require replacing I may try a smaller bank. OTOH we are getting a good lifespan out of the present bank by not stressing them.

Photowhit

If you rarely use your freezer, there is a scenario of setting the freezer on a timer, so that it is only switched on for a half hour every 4 over night and through the morning and then run continuously during the afternoon. You could chart the internal temperatures, I can't recall the max temp to allow to prevent freezer burn, something like 10 degrees Fahrenheit. This maximizes the energy use of the freezer to the afternoon hours.

kaipo_boy

Photowit, that is a stellar idea! Do you know what type of timer a setup like that would require? I was thinking along the same lines, but in my mind I was trying to dump the battery altogether and run the freezer during the afternoon hours continuously and not at all during the night but was not sure if the temp swings would be too extreme. Couldn't figure a way to do that, though, need the battery in there to cushion the uneven voltage swings produced by uneven PV production.... but combined with a timer, that might be doable. My 7 cu ft chest freezer draws around 1.4kwh per day right now... about 3x the amt that my fridge uses. If I could limit its draw to the afternoon wasted watts, though, it might be a good way to use that power since its mostly wasted anyway right now. The only other answer was expensive, as in over a thousand $$ to get the SunDanzer DC freezer and eliminate the inverter from the equation entirely, along with conversion losses. A timer on THAT freezer to just use afternoon watts would be awesome.

Photowhit

I haven't checked recently, so I might be out of line, but I think 1.4KWHs is 2x what current 7cuft models use.

here's one from Amazon looks like a better choice for my A/C which has some odd wall wart thing.

Mine looks like this one, but can't believe I paid more than $10

solar_dave

kaipo_boy wrote: »

Photowit, that is a stellar idea! Do you know what type of timer a setup like that would require? I was thinking along the same lines, but in my mind I was trying to dump the battery altogether and run the freezer during the afternoon hours continuously and not at all during the night but was not sure if the temp swings would be too extreme. Couldn't figure a way to do that, though, need the battery in there to cushion the uneven voltage swings produced by uneven PV production.... but combined with a timer, that might be doable. My 7 cu ft chest freezer draws around 1.4kwh per day right now... about 3x the amt that my fridge uses. If I could limit its draw to the afternoon wasted watts, though, it might be a good way to use that power since its mostly wasted anyway right now. The only other answer was expensive, as in over a thousand $$ to get the SunDanzer DC freezer and eliminate the inverter from the equation entirely, along with conversion losses. A timer on THAT freezer to just use afternoon watts would be awesome.

I know a guy who setup a off grid system much like what you guys are talking about. he made a "bang-bang" controller to charge his Chevy Volt with excess power after his battery bank is charged. It even stops and starts the charging if a cloud passes in front of the panels.

http://gm-volt.com/forum/showthread.php?12082-Adaptive-Solar-Charging-hack

kaipo_boy

Photowhit wrote: »

I haven't checked recently, so I might be out of line, but I think 1.4KWHs is 2x what current 7cuft models use.

here's one from Amazon looks like a better choice for my A/C which has some odd wall wart thing.

Mine looks like this one, but can't believe I paid more than $10

Hi Photowit, yes, my freezer is an older one, so probably draws more than a current one. Still, my Cotek's fan comes on when my fridge cycles for maybe 5 minutes every hour; so I know that this new, current-model compressor draws a lot of amps... but stays on for only a short time. The freezer, on the other hand, never makes the Cotek's fan come on when it cycles, but it seems to be on for longer periods, noticeably longer than the fridge's compressor. The kill-a-watt says the fridge uses .500 kwh per day; the freezer, 1.33 kwh per day.
Thanks for the suggestion on the timer. It's a good gamble and I've ordered one... this is worth looking into. The alternative is spending over a thousand bucks for another freezer... which I'm also looking into, but not breathlessly.

Photowhit

kaipo_boy wrote: »

Hi Photowit, yes, my freezer is an older one, so probably draws more than a current one. Still, my Cotek's fan comes on when my fridge cycles for maybe 5 minutes every hour; so I know that this new, current-model compressor draws a lot of amps... but stays on for only a short time. The freezer, on the other hand, never makes the Cotek's fan come on when it cycles, but it seems to be on for longer periods, noticeably longer than the fridge's compressor. The kill-a-watt says the fridge uses .500 kwh per day; the freezer, 1.33 kwh per day.
Thanks for the suggestion on the timer. It's a good gamble and I've ordered one... this is worth looking into. The alternative is spending over a thousand bucks for another freezer... which I'm also looking into, but not breathlessly.

I can't find a number in this that I agree with!

Are you sure you haven't switched your figures?

It would be hard to find a modern refrigerator that only used .5 kwh a day, and freezers tend to run less than fridges, they are better insulated and are opened infrequently.

Here's the first freezer that came up on Google search a Non energy star compliant freezer that uses about .7 kwh a day costs $240 , I'm sure you can find an energy star one in the same range....

CALLD

I've found each of the 3 moderately sized fridges I have on my system draw about 1kwh in a 24hr period. But that is AVERAGED over a year. In the high temperatures and humidity of late summer it can be over 2kwh each with the compressors running nearly continuously. Humidity is the culprit, it makes humans feel hot and thirsty which results in fridges being opened a lot! When that humid air hits the inside of your fridge the resulting condensation releases a huge amount of latent heat that the compressor needs to work overtime to remove again. In the cool and crisp dead of winter they hardy draw much at all, less than 0.5kwh per day as the compressors are off most of the time. In winter I often wake up and notice none of the 3 fridges drawing anything until I open one...

kaipo_boy

Hi Photowit. Yes, I'm sure of the figures. My old side by side was drawing over 4kwh per day (bad door seal) so I just trashed it a couple months ago and picked up a new chest type freezer, 14.7 cu ft Frigidaire/GE/Whirlpool. It is a chest freezer with a Johnson Controls digital thermostat/relay to autonomously set any temp you want, I think most people use them to control oven temps for home brewing mash and such... but works very well in freezer to turn it into a fridge. As it is chest type, I don't scold the kids anymore if they open the door and stare into the contents while trying to piece together assorted neurons. But as the freezer comes with great thermal insulation already from the factory, it is plug and play when you get the digital thermostat (less than a hundred bucks) and the unit is very, very efficient. Been living with it a few months now and not looking back. Horizontal vs vertical storage continues to be a slight issue, baskets really help here. But compared to the savings in power, it is only a minor inconvenience. The other issue I've encountered is condensation, which requires you to manually sop it up with a towel about once a week from the bottom as there is no provision to drain it out (well, there is a drain in the bottom of the unit, but its blocked by ice from frozen condensation as I have the setpoint a bit on the low side to extend longevity for veggies). It helps if you purposely have the freezer/fridge on a slightly uneven floor so the condensation collects on one side, much easier to sop it up that way.

Photowhit

Did you try to reset the freezers thermostat? Might have saved you the money on the Johnson Controls unit....

Still can't believe it cost you $1000...

SolarPowered

Cloud cover afternoon days VS sunny after noon days.
Could be wrong about this assumption but it sounds like your problem on the float absorb is coming from polycrystalline panels.

Cloud cover on polys will yield a much lower VMP than the readings coming from the battery bank.
IMP always falls short.
Then there is a algorithm ramp up that polys tend to have issues with VS monos.

What is the name plate spec on your panels if you don't mind telling us?

Issue happens less with monos, but it still happens.

Also full battery depletion from a long night is not a good thing. CC have a hard time maintaining charges on fully depleted batteries anything under 11v for 12v batteries. In your case 24v but you may be falling short at under 22v.

I would suggest investing into another battery bank.

Better investment than buying a sundazer freezer/fridge combo.

kaipo_boy

Photowhit wrote: »

Did you try to reset the freezers thermostat? Might have saved you the money on the Johnson Controls unit....

Still can't believe it cost you $1000...

No, sorry, the freezer I was looking at is about $1100 and it's the SunDanzer 8 cu ft freezer... pricey but super efficient and runs on 12v-24v DC. That's the expensive unit I'm contemplating. My current chest freezer is a standard Magic Chef, maybe 12 years old. It has a thermostat like normal. My freezer-turned fridge is the 14.7 cu ft Frigidaire and I didn't look at the thermostat to see if I could adjust it. The digital Johnson allows too much flexibility so I wanted to use that one. Aside from the set point, you can also set a range, or what I call 'dither' so that cycle que is different going up than coming down, plus a couple other variables, it allows a lot of flexibility rather than just one number.

kaipo_boy

SolarPowered wrote: »

Cloud cover afternoon days VS sunny after noon days.
Could be wrong about this assumption but it sounds like your problem on the float absorb is coming from polycrystalline panels.

Cloud cover on polys will yield a much lower VMP than the readings coming from the battery bank.
IMP always falls short.
Then there is a algorithm ramp up that polys tend to have issues with VS monos.

What is the name plate spec on your panels if you don't mind telling us?

Issue happens less with monos, but it still happens.

Also full battery depletion from a long night is not a good thing. CC have a hard time maintaining charges on fully depleted batteries anything under 11v for 12v batteries. In your case 24v but you may be falling short at under 22v.

I would suggest investing into another battery bank.

Better investment than buying a sundazer freezer/fridge combo.

Hi SolarPowered, thanks for the input. The panels are Trina 285w models and I do believe they are poly, I wasn't aware there was a big performance difference... still learning, and thanks for the info. Also I mis-spoke (or being over dramatic), sorry, my 'full battery depletion' is to me 24.4v, or 50% SOC which I never let my bank fall below, I will simply disconnect loads if it gets that low and wait till morning to run some current into them if they go that low. The bank is about 2 weeks old.

SolarPowered

kaipo_boy wrote: »

Hi SolarPowered, thanks for the input. The panels are Trina 285w models and I do believe they are poly, I wasn't aware there was a big performance difference... still learning, and thanks for the info. Also I mis-spoke (or being over dramatic), sorry, my 'full battery depletion' is to me 24.4v, or 50% SOC which I never let my bank fall below, I will simply disconnect loads if it gets that low and wait till morning to run some current into them if they go that low. The bank is about 2 weeks old.

O.K so you will have to bare with me and this may take a few posts as this requires many detailed pictures using a micro inverter monitoring system and lots of math to show how this affects your CC for battery charging operation.

The first 2 pictures I am showing are that of a micro inverter 9.6 kWh system. This system use an LG 300N1C-B3 Mono Crystalline module so it is a good template to use to compare a hypothetical MPPT algorithm. Your Trina 285 module actually has a higher VOC/VMP, but is 1.44 IMP less than the LG300N1C-B3. VOC/VMP "REGARDLESS OF TEMPERATURE", and IMP of your poly is impacted by roughly 4% at lower light insolation levels (and cloud cover days due to the impurity of poly VS MONO). VMP shown in these graphs of the LG/ will be very comparative to that of the Trina. IMP will fall short mostly because the difference between a 300Wp panel and a 285Wp panel. This is the difference that makes the name plate 1.44IMP variance between the 2 panels, this has the major effect on algorithm ramp up, from level of irradiance. This is where your thought process has to understand the difference in I-V curve, we can see the reactions the of I-V curve (from cloud cover PTC), and how it affects MPPT algorithm and harvest other wise known as "MPPT CLIPPING".
When you see this MPPT algorithm graph it will show that at some points there aren't enough VMP, or IMP to drive your charge controller to charge the battery because the level of resistance for off grid battery systems, require 10% of applied power production to charge the battery bank. In most cases its not even VMP, but the IMP at some instances in the algorithm graph.

In the algorithm graph example to drive volts 38.3V, but is under .05 IMP which isn't enough in our hypothetical situation to drive the charge due to the resistance of the battery bank. Without the driving force needed for charging the battery bank, the controller will react as if that the battery bank is equal or greater in resistance, your CC will then be in float/maintained, unless the driving force and a CONSTANT ALGORITHM can drive the charge to the battery bank.


Picture 1 shows how the MPPT algorithm produces power with cloud cover. TOTAL PV watts for that day was 32.6kWh. It was a horrible PVwatts day on may 20th. Attachment not found.











The second picture is that of a perfect sunny day, recorded yesterday, the 9.6kWh system produced 66.9kWh more than double what was produced on the cloud cover day. Attachment not found.











So now lets explore the mppt monitoring diagnostics to one of our solar panels/micro inverter. Our panel for this hypothetical situation is an LG300N1C-B3 monocrystalline. We are using mono for reference because we understand that poly typically yields 4% less of VMP/IMP than mono crystalline, on cloud cover insolation/lower light levels. Attachment not found.
Attachment not found.
Attachment not found.

SolarPowered

So according to the MPPT algorithm graph.
There was an instance 20 minutes before 14:00 hours where VMP looked excellent at 38.3VMP, but IMP suffered drastically at .02 IMP, thats is not enough to drive a charge to the battery bank and it roughly occurred for 20 minutes.
Another instance where between 17:30hours and 17:50hours VMP averaged 32.6, but did fall from dips as low as 30VMP, IMP was between .07, and 1.3IMP.

We can also see in the algorithm graph that temperature doesn't always affect VMP. Recorded between 14:00hour the temperature was recorded at 84*F yet the panel reacted to produce 38.3V.
Also recorded between 17:30 and 17:50 it was cooler temperature condition at 77*F however VMP was recorded between 30~32.6VMP, most of this has to do with sun fall and less insolation greater than the temperature coefficient variable that would drop the panels VMP performance by .32%.

Understanding algorithm and I-V curve are very important and what we come to find out when we micro analyze our I-V curve panel specs VS just name plate, MPPT will react in very unorthodox conditions because of the constant variable of PTC, we also come to find out that not always do the variable percentages for coefficients in STC, do not at all react in the same manner to the PTC condition.

The MPPT charge controller requires a 4.5% conversion, and in order to get into bulk charge the wattage delivered to the CC prior to the MPPT conversion must equal a wattage that is greater than the resistance of the battery bank which is typically 10%. That is why the CC stay's in float/maintained on cloud cover.

So what did we learn?

If you want the system to charge on cloud cover days it will require a minimum 300watt mono crystalline panel, or a 310watt poly crystalline, that will allow the CC to deliver a charge that is greater than the resistance of the battery bank.

This is why I suggest you invest into another battery bank to prepare for winter insolation, or eventually, and slowly upgrade to a 300watt mono, or a 310watt poly.


OR

If you DON'T want to invest into new panels, If you DON'T want to invest into another battery bank.

You could reconfigure the system for 12V. I am all about 12V systems and HARVEST, HARVEST, HARVEST. This is why 12V systems outperform on harvesting and have left overs, because there is less VMP required for charging, there is enough remaining Volts to convert for amps to drive that power required for charge.

Then all you would need is a 12V inverter, and I guarantee it would get you out of float on those cloud cover days with your 285 trina.

Remember 12V is excellent for producing power up to 2kWh!

Kindest Regards