connecting batteries in parallel or series, myth and theory

In alt.engineering.electrical N9WOS worth of daily driving. You can count the cycles. I had my controller's |> battery amps programmed to 400 amps max, a limit that I hit fairly often. |> No |> significant loss of capacity or range. | | Yes, but how fast do you charge them? Amperes over what time period? | I would say that it isn't close to the rate at which you charge them. | The key isn't how fast or how much you discharge them, it's how fast you | charge them.

Charging a battery is similar to electroplating. While they do try to find ways to speed up electroplating, doing it well is something done slowly. Slow charging is something I've always done, with a few exceptions in urgent cases.

No longer. Today the only solar work I do is helping people fix the problems caused by poor design.

The science is that cells all have different internal resistances.

As you wish, but, all the systems I designed using my training have worked to spec.

S*it, that was close to twenty years ago. Battery technology has not changed much since.

Don't know much about commercial installations. I only did standalone home power.

Now that sounds like an insult:-) Yes I sold batteries, as part of a home power system. The batteries were chosen for their suitability for the job.

Explain to me this... How do you go from sub 50% State Of Charge to 100% SOC in one solar day?

Explain how that is suppose to work in a real life system.

Same difference. Whether it's 1 day or 10, ignore the need to get batteries fully charged regularly and it won't matter whether they're

It can be, but it's more often just life. For example, a setup can be well designed for say, 2 hours use of TV per day. But if the user buys a larger model and/or watches it longer then he's likely to have a problem. The key IMO is a proper battery monitor. That provides the user with a routine method of knowing his battery's true SOC, which gives him a chance to adjust habits or hardware if needed. DJ, a solar dealer who used to post here, once wrote that he wouldn't sell a setup anymore unless it included a battery monitor. Either the customer would pay for it or he'd include it free if necessary. I'll add that I'd question the business sense of any dealer who installs a whole-house setup without a proper battery monitor. Besides doing the customer a disservice, he's inviting call-backs that will eat up his profit margin. In my own experience with strongly encouraging friends and acquaintances to buy and use battery monitors for their home power installations, I've yet to have anyone regret their purchase. Most rave that they should have had one from the beginning, and some blame their dealer for not including it. The necessary $50 shunts are installed in the popular Outback power panels, so many owners need add little more than the meter itself, which only costs about $150.

Batteries are rated to supply a specific number of amp-hrs. The amount will vary with the number and depth of discharge cycles, and can certainly be less than spec if subjected to poor maintenance or bad habits. But it can't be more unless you've discovered something that's escaped everyone else.

Of course, that's the prevailing wisdom, and this being Usenet, there are certainly alternative theeries. For example: "A lead acid battery stores a chemical reaction. Therefor every time you charge/ discharge the battery you use up some of the chemicals involved. Also the greater the depth of discharge the more chemical used = fewer cycles. When the chemical is gone the battery is dead."

Wayne

On Thu, 14 Aug 2008 12:30:21 -0500, "N9WOS" >

You've been making a habit of it, such as the other day for example when you wrote that narrow-minded baloney about off-grid AC lighting being "stupid".

So what? Home power batteries tend to be discharged a fraction of their capacity per day, and (ideally) recharged at a rate that will get them fully recharged most days. The charge rates tends to rise during the morning, and fall in the afternoon or anytime with concurrent loads. Batteries that are discharged over several days tend to take days to reach full charge again, and owners who suffer that scenario frequently, probably wish they had some overcharging to limit. In the case of grid-connected setups with batteries designed to be discharged in a shorter time, the charger is programmed to recharge at the ideal rate.

Nobody has said that exceeding recommended charge rates is good practice, and you're the only one who's claimed it's necessary. You're arguing with yourself.

So you've invented an unlikely scenario to support your case. I'm not impressed.

That's the first thing you've written that relates to a common problem with home power batteries.

So basically you're saying that if you start with a strawman-battery sized to prove that overcharging is a big issue in home power applications, and then enlarge that size to about what most people actually have, the problem goes away. IOW, it only exists in a tortuous Usenet post.

That's done automatically by modern multi-stage chargers. They supply full available current (which is highly unlikely to exceed the recommended rate in my experience) until bulk voltage is reached, tapered current as needed to maintain bulk voltage for a set time, then further reduced current to maintain a lower float voltage. PV is generally too expensive to size for overcharging, but generators are sometimes capable of it, so AC chargers are programmable and are normally set to the proper rate, then shut down after absorption. Finish charging is optionally delayed until PV is available again in order to prevent extended generator run times.

But this "damage" only exists in your strawman scenario. In normal home power applications, it's *low* charge rates that tend to be symptomatic of bad design or operation. The closest I've seen to an overcharging problem is where people who are chronically short on supply raise their bulk and float voltage settings to limit throttling. Which isn't too terrible unless there are days when the batteries are charged early, which wouldn't be often considering the shortage of supply that prompted the fiddling. My own setup suffers from that overcharging scenario but for a different reason - I have a single stage turbine charge controller in parallel with an MX60. The turbine charging voltage is set a little higher than the MXs bulk setting, which means that when the batteries are full but the wind is still blowing and loads are minimal, the batteries are being floated at above normal bulk voltage. That only happens occasionally and doesn't tend to last long, so it hasn't been a factor in my battery life which is 13 years and counting.

Wayne

On Thu, 14 Aug 2008 18:41:10 -0500, "N9WOS" >>them.

Why? It's *your* contention that SOC routinely goes from sub 50% to

Wayne

Hold it one second here...You are messing with my mind. "grid connected"? Why did you attach that into the sentence? What does grid connected system have the slightest to do with the post you were responding to?

Yes I didn't realize that you shifted the subject to grid connected system which had nothing to do with my post. My mind wasn't able to comprehend why anyone would use a design of a grid connected system to argue the point that zero day autonomy off grid systems (a system that can't ride through one day of overcast weather) are perfectly viable.

It sounded like you were saying

"Batteries can be well-maintained and last the maximum with zero days autonomy. It's done all the time"

Followed up by

"Nonsense. The #1 cause of premature failure of home-power batteries is chronic undercharging."

Which sounds contradictory. Yesh.....

My responses are going to be gibberish if the person I am debating with can't even stay on subject.

That is my excuse and I am sticking to it!!! :-)

All right. You stated "Whether it's 1 day or 10(days of autonomy)".

Lets look at one day of autonomy for the winter months. 5 hours sunshine. You have 19 hours between charging periods. To have exactly 1 day autonomy would mean that you have to ride through two dark periods and one cloudy charging period. Given an average power usage spread (lights at night and other stuff in the day) that will give you about 44% capacity usage between charging periods. So each charging day, you are starting from 56% SOC. (close to my original guess)

So I will refine that question.

How do you get from 56% SOC to 100% SOC in one charging period.

If it can't happen then it does mater if the system has 1 or 10 days of autonomy.

|>Sounds like a design problem. | | It can be, but it's more often just life. For example, a setup can be | well designed for say, 2 hours use of TV per day. But if the user buys | a larger model and/or watches it longer then he's likely to have a | problem. The key IMO is a proper battery monitor. That provides the | user with a routine method of knowing his battery's true SOC, which | gives him a chance to adjust habits or hardware if needed. DJ, a solar | dealer who used to post here, once wrote that he wouldn't sell a setup | anymore unless it included a battery monitor. Either the customer | would pay for it or he'd include it free if necessary. I'll add that | I'd question the business sense of any dealer who installs a | whole-house setup without a proper battery monitor. Besides doing the | customer a disservice, he's inviting call-backs that will eat up his | profit margin. In my own experience with strongly encouraging friends | and acquaintances to buy and use battery monitors for their home power | installations, I've yet to have anyone regret their purchase. Most | rave that they should have had one from the beginning, and some blame | their dealer for not including it. The necessary $50 shunts are | installed in the popular Outback power panels, so many owners need add | little more than the meter itself, which only costs about $150.

Based on the description of there being just one monitor wired in at the inverter, it would seem this monitor only checks the health of the whole string, not each individual battery/cell.

Exactly what parameter(s) does this monitor check?

|>| No. Batteries have a finite life-rating usually stated as a number of |>| cycles at a particular discharge level. |>

|>Which is only a simplistic view of the usage patterns of batteries. | | Batteries are rated to supply a specific number of amp-hrs. The amount | will vary with the number and depth of discharge cycles, and can | certainly be less than spec if subjected to poor maintenance or bad | habits. But it can't be more unless you've discovered something that's | escaped everyone else.

If the manufacturer rates the battery for its most ideal circumstance, then sure, you can't get the battery to do better than the rating. And most of the manufacturers are interested in diverting all sales to their products, so of course they will rate as high as they can. If there is some independent source of battery rating information that rates batteries based on how well they do in real life scenarios, then I'd be more inclined to use those kinds of ratings, which are likely to be less where the real life scenarios are not what is ideal for the battery.

| Of course, that's the prevailing wisdom, and this being Usenet, there | are certainly alternative theeries. For example: "A lead acid battery | stores a chemical reaction. Therefor every time you charge/ discharge | the battery you use up some of the chemicals involved. Also the | greater the depth of discharge the more chemical used = fewer cycles.

A theoretically perfect battery will perfectly reverse the discharged chemical reactions during the charging cycle. Even if real life gets very close to that one big issue is that the reaction doesn't reverse in the same way (for example the lead that redeposits on the plates does not form the plates in the same way as they were formed in manufacturing). And of course, as soon as any gassing occurs and the gas bubbles move someone else, that's rather hard to reverse without some means to move the gas back. The best we can do for loss of water is to add some pure water back.

In alt.engineering.electrical N9WOS >them. |>>| |>>| Baloney. Batteries can be well-maintained and last the maximum with |>>| zero days autonomy. It's done all the time with grid-connected setups |>>| that only have storage enough to last through short outages. |>>

|>>What about off-grid setups? |>

|> Same difference. Whether it's 1 day or 10, ignore the need to get |> batteries fully charged regularly and it won't matter whether they're |> 5 or 20 year batteries, whether they're wired in series or parallel, |> charged fast or slow, etc. | | Explain to me this... How do you go from sub 50% State Of Charge to 100% SOC | in one solar day?

With a very fast charge?

| Explain how that is suppose to work in a real life system.

If I'm aiming to charge at a C/20 rate, I wouldn't expect to get a full charge back within, say, 6 hours of good sunlight in the winter. So it would seem to be a good idea to rate the system so that if it takes 4 days to recharge from an XX level of discharge, you need to rate the system capacity and manage your usage so you would discharge it to no more than XX level of discharge during both the unlit times of those 4 days PLUS the number of days between those days of light (e.g. the cloudy/snowy days). So we need a new ratio: rate of usual discharge divided by rate of usual charge. And since we want to charge slowly while also avoiding a deep discharge, the rate of discharge needs to be quite small, such as C/100. That means a BIG system, possibly larger than practical to do with a single string of single monster cells.

I wonder where those replaced submarine batteries end up :-)

|> So you are a dealer of very large cells/batteries. | | No longer. Today the only solar work I do is helping people fix the | problems caused by poor design. |>

|> | I can only repeat what I learned getting my accreditation. That was |> | many years ago and I am sorry that I no longer have the notes to show |> | you. |> | |> | The science says that parallel strings do not charge and discharge |> | equally. Parallel strings should be avoided where ever possible and GC |> | batteries belong in a golf cart. |>

|> That's not science. Science would tell you why and how that happens. |> Further, science would tell you how they behave under various methods |> of connecting charging and discharging circuits. | | The science is that cells all have different internal resistances.

I guess you don't have an EE degree.

Not that I have one, either. But I respect the knowledge AND understanding real engineers need to have. The information I seek is what would be learned in an Electrical Engineering, Power Elective, curriculum (not all of it, of course ... just the parts I'm interested in right now).

|> You're taking the training you've been giving regarding certain finite |> methods to install battery systems, and calling it "science". Real |> science comes from research laboratories and investigative processes |> that explore all options and find out why things really happen and how |> control methods may or may not work. | | As you wish, but, all the systems I designed using my training have | worked to spec.

Are you talking about true design, or just mere deployment configuration?

Bob needs twice the capacity as Carl, so you install a system for Bob that has a string of cells twice as big as those you installed for Carl, that's NOT "design". That's "configuration". You select a design (one single string) and you select a capacity based on available models.

Configuration is when you select from existing known designs. Have you ever installed a system which was wired different than _any_ example you ever saw before?

|> | If you want to use them, fine. But in the long run you will say, "That |> | bastard was right". |> | |> | The tests were done at Royal Melbourne Institute of Technology in |> | Melbourne. The test was done on a rig of six parallel strings of six |> | two volt cells. They tried all the tricks like isolating the strings |> | with diodes and fancy interconnecting of cells. The results were |> | always much the same. Unbalanced charging across the array. |>

|> What kinds of chargers did they use? Did they have separate chargers for |> each string? | | S*it, that was close to twenty years ago. Battery technology has not | changed much since.

But you didn't want to answer my question. Don't worry, you are under no obligation to answer it. And, besides, this is Usenet. Most questions go unanswered, anyway.

Start small. If it works go big. Don't even consider trying to start small and expand the system. You can always use the "small" system for a dedicated application (like running a relatively low-drain system in the house - perhaps the computers and entertainment systems, or lighting circuits etc. Always good NOT to have "all your eggs in one basket" with an alternate energy system.

Why is anyone bothering to talk to you?

And I'll point out that the above is just the worst example, not the only one.

This paper may be what you are looking for.

My brain if frying today. The smoke is coming out of my ears right now. That is the paper that Ron linked to at the beginning of this thread. Just ignore the previous message please. :-<

Gee Floyd, looks like you have been taking lessons from Tweedledum. I never said;

Which either makes you a simpleton or a liar.