So, lets spec out and set up a system of LiFePo4 batteries and inverter

[karrak]

For instance, there is no penalty for making a big mistake and oversizing your battery needs (other than upfront in your wallet), when you could easily just live in say a 40-60 PSOC window. Heh, if you do that, you may never have to even think about balance at all, aside from an initial sanity check on the first charge just to get relatively close.

Lightbulb time! For those that want to spend gobs on balancers and such, just put that money into a larger battery instead, run psoc, and fuggeddaboutit.

I see a few problems with this:

The most obvious one being cost, my battery which is only a small one at 9.6kWh cost around AU$4,300. I utilise about 80% of the capacity of the battery and cycle from 10% up to 60% of the total capacity daily. If I only utilised 60% of the capacity I would still need the same total capacity so my battery would need to be 20% larger and cost an extra AU$860. My first battery monitor, a cheap little CellLog8 cost all up around AU$50 and my DIY battery monitor with which I can check the state of my battery where ever I am via the internet cost around AU$150.

The second major problem is how do you charge your battery up to only 80% SOC when you are charging from solar panels where the charge current can vary by a factor of ten or more and discharge down to 20% when you have varying loads?

Simon

[PNjunction]

The second major problem is how do you charge your battery up to only 80% SOC when you are charging from solar panels where the charge current can vary by a factor of ten or more and discharge down to 20% when you have varying loads?

Easy:

1) Do a sanity charge on each cell individually up to 3.6v and hold until C/20 / 0.05C current is reached.
2) Set your HVD to 13.8v / 3.45v per cell
3) Set your LVD to 12.8v / 3.2v per cell
4) Use an accurate quality device for hvd/lvd
5) Make sure your wiring infrastructure such as cell-links and other interconnects are tight and clean from oxidation.

Live long and prosper in a conservative psoc environment which will approximately stay out of the knees. One is relying on the manufacturer to be providing cells that are equally matched in both capacity and internal resistance. The closer the better.

Ideally run sub-c for discharge. Not suitable for EV. Void where prohibited. And of course, YMMV.

[karrak]

Easy:

1) Do a sanity charge on each cell individually up to 3.6v and hold until C/20 / 0.05C current is reached.
2) Set your HVD to 13.8v / 3.45v per cell
3) Set your LVD to 12.8v / 3.2v per cell
4) Use an accurate quality device for hvd/lvd
5) Make sure your wiring infrastructure such as cell-links and other interconnects are tight and clean from oxidation.

Live long and prosper in a conservative psoc environment which will approximately stay out of the knees. One is relying on the manufacturer to be providing cells that are equally matched in both capacity and internal resistance. The closer the better.

Ideally run sub-c for discharge. Not suitable for EV. Void where prohibited. And of course, YMMV.

You haven't answered the question.
Simon

[PNjunction]

Ok, I'll try - although I thought I answered earlier.

Cost - if you can't afford the up-front cost of LFP, then stick to lead-acid. It is proven, and works well with proper upfront engineering and maintenance. Perhaps one will merely have to replace batteries a little sooner. Will that replacement come out somewhat even compared to LFP in the long-run in both cost and lifecycles? Maybe so. Since my needs are small, I can easily afford the upfront cost of an LFP bank compared to replacing the equivalent with lead twice or more. With LFP, my biggest savings come from operational freedom and simplified (but not totally ignored!) maintenance.

Cost vs life cycles has been discussed to death since about 2007 or so. If one can handle it up front - fine. If not, Pb replaced twice or more works just great too. I am pleasantly satisfied with both.

Varying solar condition charging :

Earlier, (maybe not this thread, I'll have to go back through it) I mentioned that when using solar it would be VERY wise to be conservative and set your end point voltage very low, which most of us are doing now at say 14.0, or perhaps even lower like anywhere from 13.8 to 14.0. At these lowered voltages, absorb will be slower, (assuming that you have actually had a discharge the previous day), and will avoid the problem of reaching a full charge with .05C or less.

Charging at .05C or less, which can happen with poor weather conditions for solar, and especially when your voltage is set high, say to 3.6v per cell, means that IF YOU DO REACH THAT point with such low current, the voltage rise is actually being created by the chemical reaction of overcharge - where the rise in voltage is not coming from charging per se, but from electrolyte heating instead!

If you watch it close enough, and you have your voltage set high, but your current capability low, (like .05C or less), you'll see the absorb basically stall. Some time later, the electrolyte heats up, and a very fast voltage rise follows! This is different from the normal steep upwards charge slope you witness under normal conditions. Not that I'm not saying that just being able to supply more than .05C is a total solution either.

In the end, no matter what voltage you choose, if you let it absorb and try to "absorb to zero", rather than stopping at .05C, electrolyte heating is the eventual result, and a rise in voltage *due to that* is the problem. Charging was finished long ago. Now comes parasitic reaction time.

So, playing it conservative with a 13.8 / 14.0 max with solar is very smart!

[Sunking]

Give up PNJ, Karrak is not capable of understanding. You already know he is a liar from the other Forums you both belong too.

Karrak by his own admissions does not use Balance Boards, a BMS, or Balances his batteries. PN you know it is true as you have read it like I have on 3 other Forums. He does not understand the technology, only what a shill from manufactures tells him. One of the mods on this forum knows the truth like and does nothing but let it continue.

[Sunking]

So, playing it conservative with a 13.8 / 14.0 max with solar is very smart!

Not to mention requires no automation other than what your typical solar equipment already provides. Any BMS or Vampire boards are rendered useless.

Bottom line any commercial solution requires engineered automation. The public is too ignorant. Issue with it is the Automation is also the biggest threat to the batteries and sets up the conditions of a melt down. People are stuck in a box of One Must Fully Charge Batteries. Manufactures are in biz to sell you something and have liability issues. That is why your McDonalds coffee cups have warnings. The public is full of stupid people who do not know coffee is hot and will burn you if you spill it on yourself.

People who really understand the technology get much better performance by simply changing strategies and use Passive protection, In the process save a lot of money and stress like you and I PNJ.

[PNjunction]

Well, actually any low current at 3.5v CV or higher per cell, will eventually fully charge a cell, the only variable being the length of absorb.

By running conservatively, and with a daily discharge routine, one is unlikely to have enough solar charge time in the day for that absorb to try and drive down to zero. (unless they are running ridiculously small banks, or totally over-panelled).

Ideally, a smart controller would have to be set for the capacity of the bank to watch for .05C and voltage, perhaps a length of time to account for passing clouds etc, and just stop. Otherwise there IS the potential to try and drive the cells to zero/no absorb, say on consecutive days of no discharge, and on one of those days the transition from 100% SOC to electrolyte heating occurs.

I watched it happen with one of my 20ah gbs cells on purpose. Drove the absorb down to zero with less than .05C with cell sitting at 3.5v when it stalled. Amazed to watch the voltage later climb like a banshee after about an hour, but with NO current flowing!

Ordinarily, that is electrically impossible from a battery with no current, but from a chemical standpoint of electrolyte abuse, it was fascinating. Pulled the plug pronto when it wanted to leave the launchpad as it was no longer a battery but an electrolyte cooker.

But would this same effect be noticeable with a cell 10 times as large? Possibly not, or not after a much longer delay. However, the point is clear - letting the system absorb down past .05C is not healthy - with any upper cell voltage of 3.5v or larger under charge. My small cells let me witness the parasitic reaction quickly. On a large cell - maybe a meteoric rise in parasitic reaction voltage won't be noticed, but those reactions are happening anyway degrading the electrolyte.

So I don't really know how to *exactly* stop at 80%. I either just charge to a designated end voltage and do no absorb, or just pull it while in absorb, never letting it truly finish - with .05C being the trigger to stop absorb worst case.

What is clear is I need to save up for a 200ah bank to continue the testing.

[PNjunction]

And the wheel of discussion just turns over again regarding balance.

This is why I feel that if one needs exacting values of capacity measurement during use, along with an easier safety mechanism for overcharge and undercharge, bottom balance is technically the best way to go. Perfect for EV's.

BUT, like the VHS vs Beta tape comparisons (young people will need to look that up..), so-called top balance is convenient, and also relatively easy to implement, but you just want to be on top of things especially in a solar environment, balancing both upper voltage AND absorb current limits - trying to avoid the "absorb to zero" current problem.

I use top balance for convenience. I don't need exacting capacity usage measurements during use. For me, top balance is the 99% solution, just watch out for that 1% "gotcha". Whatever floats your boat. Next!!

[Sunking]

Ideally, a smart controller would have to be set for the capacity of the bank to watch for .05C and voltage, perhaps a length of time to account for passing clouds etc, and just stop. Otherwise there IS the potential to try and drive the cells to zero/no absorb, say on consecutive days of no discharge, and on one of those days the transition from 100% SOC to electrolyte heating occurs.

Very good and escapes most everyone here. There is no such product for Solar on the Market now is there? Only chargers that can do that require commercial AC Power, and all of them are made for EV's which is driving the Technology. Example an Elcon PFC 1500 is the largest 120 VAC battery charger you can buy. It will charge any battery type from 24 to 120 VDC. When you order it you specify the battery type voltage and any parameters you want. They can load up to 8 algorithms and interfaces to any BMS or Balance Boards using CAN BUS (Controller Area Network) used and developed for the Automotive Industry. CAN BUS is the communication Protocol used in EV so all parts can communicate. Things like the charger, BMS, Balance Board, Coulomb Counters, Motors, Controllers, GPS, Lighting, you name it.

As of today DIY Solar has not even made the first step to integrate all th epeices needed. Sure Outback has a comm bus but is proprietary and nothing for Lithium. On the Commercial side there are some Euro manufactures and Vapor Ware for USA like Tesla that have an integrated Lithium BMS, but not one bit of it you can access. Mainly because commercial systems use high voltage DC and by the very nature has to be made inaccessible to the user just like commercial EV's. All of it Proprietary and will not work with any other equipment because the Solar Industry is so far behind has not even developed a standard communications protocol.

Can a DIY solar person make something work. Sure, there are three-ways.

1. Know WTF you are doing and modify automotive equipment to integrate to available solar charge controllers and there are only a couple like Midnite and Outback that will allow you to remotely turn the Controller off. Very expensive and extremely complicated.

2. Use some hack equipment from Australia who claims they have a Solar BMS which is nothing more than cheap Vampire Boards and a LVD and HVC contactor they call BMS built in someones garage.

3. Do like PNJ and DIY EV guys like myself. Use a completely different strategy using Passive methods proven to work. Much safer than any manufactured solution using automation with longer cycle life and little to no chance of destroying your batteries. The only automation required is what is already provided in Charge Controllers and Inverters. Match everything up and no worries of ever over charging or the dreaded over discharging of the batteries. You just need to understand the technology and quit listening to manufactures and shills telling you that you must use a BMS of some sort. Nothing could be further from the truth.

There is a 4th way. Wait till Lithium is actually cost competitive with FLA and maybe 10 years from now the DIY Solar manufactures offer commercial solutions to support lithium batteries. Or just call Tesla and order the Vaporware.

[PNjunction]

Actually there is a 4th way. Some of the marine guys who absolutely, positively can't have any issues, and use a simplified passive charging environment without any balance boards etc, run VERY conservatively.

That is, they don't run at any voltages where absorb can actually take place. We're talking about 3.4 to maybe 3.45v max under charge. It takes 3.5v or more under charge for real absorb to take place, so they run much lower just in case.

Of course this now means that they are really riding in the psoc window, and by doing so, nearly have to double the amount of calculated capacity when sizing the bank.

Sound familiar? Like doing the 50% DOD calculations for Pb, and doubling the required ah size of the batttery, they may do a 60% calculation / de-rate of the specified manufacturer's ah rating if they also plan to go no lower than 80% DOD on the bottom side as well. Have wallet wide-open.

Still, if one desires to do this and top-balance at least once for a sanity check, that means running the cells up to 3.6v, and letting it aborb down to no lower than .05C. THEN you can really crank down the daily controller voltage to 3.4 - 3.45v, where no real absorb as we know it can take place.

[karrak]

In a properly designed Solar System with a 3-day reserve capacity forces to charge at least @ C/6 rate which is way faster than an EV will charge. C/2 is a very real possibility where Solar Insolation or Sun Hours drop to 3 hours or less. Example Seattle December/January drops to 1.2 Sun Hours. At that low of Solar Insolation you are looking at charging around C/2. Before anyone says "I do not use 3 day, I only use 2-day". That argument does not hold water because with less than 3 day autonomy means you have to charge at even higher rates approaching 1C or higher.

The terms Solar Insolation and Sun Hours are not interchangeable, Solar insolation is a measure of solar radiation energy received on a given surface area in a given time and Sun Hours is an indicator of cloudiness of a location https://en.wikipedia.org/wiki/Sunshine_duration.

Solar panels do not need direct sunlight to generate power, they will still generate power under cloudy skies. Using Sun Hours is not as accurate as using Solar Insolation figures and a little misleading. Using Solar Insolation figures and number of daylight hours is a better approach to work out the charge rates than Sun Hours.

Having a battery sized at three times ones daily usage means you have to generate on average C/3 Ah per day. For Seattle the shortest day is ~8.4 hours long. If we divide this by two to take into account that there is not much Solar Isolation in the early morning or late afternoon we get an average charge rate of ~0.08C (0.33C/4.2), nowhere near C/2.

We know that charging LFP batteries at higher rates decreases their lifespan so I think it is a sensible precaution when charging from solar power to limit the charge to a maximum of C/4 if possible.

Simon

[Mike90250]

.....

Solar panels do not need direct sunlight to generate power, they will still generate power under cloudy skies. Using Sun Hours is not as accurate as using Solar Insolation figures and a little misleading. Using Solar Insolation figures and number of daylight hours is a better approach to work out the charge rates than Sun Hours........

Uh, well, barely. If you think that clouds don't matter, my experience shows that when cloudy, panels generate 10% or even less, of nameplate. 2 cloudy days, and I'm running the genset. Not sure where you got the basis for this statement, because if you think, or cause others to think clouds don't matter, that is VERY WRONG.

[karrak]

Uh, well, barely. If you think that clouds don't matter, my experience shows that when cloudy, panels generate 10% or even less, of nameplate. 2 cloudy days, and I'm running the genset. Not sure where you got the basis for this statement, because if you think, or cause others to think clouds don't matter, that is VERY WRONG.

I did not say that clouds don't matter. Clouds do matter, they reduce the amount of power produced, the reduction is dependant on how thick the cloud is. If you can't generate enough power under cloudy conditions get more solar panels or turn the generator on.

Simon

[J.P.M.]

The terms Solar Insolation and Sun Hours are not interchangeable, Solar insolation is a measure of solar radiation energy received on a given surface area in a given time and Sun Hours is an indicator of cloudiness of a location https://en.wikipedia.org/wiki/Sunshine_duration.

Solar panels do not need direct sunlight to generate power, they will still generate power under cloudy skies. Using Sun Hours is not as accurate as using Solar Insolation figures and a little misleading. Using Solar Insolation figures and number of daylight hours is a better approach to work out the charge rates than Sun Hours.

They are indeed not interchangeable. Insolation is the time integrated sum of irradiance. It has a well defined meaning and use. The ill defined term "sun-hours" may or may not have some various definitions, a few of which may be loosely connected to an Irradiance quantity of 1,000 W/m^2.

If the term "sun-hours" is interchangeable with any other terms, those other term might be "useless", or "misleading" at least in the context of determining or describing the availability of the solar resource for power generation purposes.

I've seen definitions of "sun-hours" such as "duration of time when the sun casts a visible shadow". Or, when an instrument called a Campbell-Stokes sunshine recorder "operates" - basically, burns an image on to treated paper. Both those, and others are lacking in necessary precision to be of any use for a lot of testing or design work. I note the referenced article has neither a power or energy term in the "sun-hours" description.

If linked to some (natural sunlight) irradiance quantity, the term "sun-hours" seems to have definitional problems several reasons besides units, dealing with angles of incidence, distribution of beam, diffuse and reflected irradiance, energy dist. as f(wavelength) and other things.

Perhaps one semi useful, but unnecessary use for the term "Sun-hours" might be, rather than as an indicator of cloudiness as you write, as a very subjective measure of "sunniness". Howeve, Id suggest the term "sunny day" seems to cover that condition well.

For testing, design and performance evaluation, irradiance measurements are required. The time integrated irradiance measurement - Insolaton- is useful, and comes along with the irradiance measurements if the irradiance measurements are done in a time sequential manner.

The term "sun-hours" has little use due to it's ill-defined nature, and IMO anyway, only causes confusion.

As Mike writes, while some power can be generated under "cloudy" skies, the amount of power generated by a solar device is, to a first approx. anyway, somewhat directly proportional to the POA (Plane Of Array) irradiance.

Since cloudy skies don't allow as much GHI (Global Horizontal Irradiance), and since what does get through is usually and largely diffuse in nature, much less power generation is the result.

Since tabular (usually) estimated hourly values, and thus day long integrated sums of GHI are commonly available, and since software is available to translate those GHI values to estimated P.O.A. values for the work of a few mouse clicks, the quantity and availability of that more quantifiable data, and it ease of use, makes the whole idea of "sun-hours" quite as unnecessary as it is anachronistic.

Take what you want of the above. Scrap the rest.

[Sunking]

The terms Solar Insolation and Sun Hours are not interchangeable, Solar insolation is a measure of solar radiation energy received on a given surface area in a given time and Sun Hours is an indicator of cloudiness of a location https://en.wikipedia.org/wiki/Sunshine_duration.

That is some funny stuff, I don't care who you are. What does Sunshine Duration have to do with Solar Irradience? Nothing..........:ROFL: Two completely different things. Go back to school and pay attention.If you are going to use links to back up your ignorance at least use the right terms which actually have somethine o do with the subject.

There are 3 units of measure of Solar Irradadience:

World Meteorological Organization is the megajoule per square metre (MJ/m2) or joule per square millimetre (J/mm2).

Langley (1 thermochemical calorie per square centimeter or 41,840 J/m2) or irradiance per unit time.

The solar energy business uses watt-hour per square meter (Kwh/m2) aka Solar Insolation. All Solar Insolation Tables for Solar Energy use Kwh/m2. If expressed in WMO or Langley units must be converted with simple formula.

So Simon here is your education on a Kindergarten level even you can understand. PV Education Organazation and Measurement of Solar Insolation