Battery equalise timer

[commanda]

Lets say you had a dump load controller. It had a switch to set the float voltage at one of 2 positions; float and boost (equalise). Lets also say it had a timer to reset it to float after a pre-determined time on boost charge.

What would be a good amount of time to set the timer for?

I'm presuming it would be a function of the ampere-hour capacity of the batteries, and possibly also the age or general condition (differences in hygrometer readings between cells maybe?)

What about a good rule-of-thumb based on amp-hour capacity. So many minutes (or hours) per so many amp-hours. Plus another factor based on the standard deviation of specific gravity.

I roughed out the circuit last night, using a 12 stage binary counter; I've currently got it clocked so it ranges from 20 seconds to 10 hours.

I'm using flooded lead-acid cells, but comments on other technologies welcome.

Amanda

[dinges]

Not really sure what you are asking; how long to do the equalization, or how long the interval between equalizations.

I do know that for my batteries (Sonnenschein gell-cells) that everything and more you ever wanted to know about them is in their datasheets. (with the remark that they SHOULD be equalized, despite the remark on otherpower.com and by readers here that gell-cells should NOT be equalized).

http://www.otherpower.com/images/scimages/3538/Sonnenschein_A400_t_e.pdf

I'd advice to search for the datasheets for your brand & type of battery. I tend to put more trust in a (reputable) manufacturer than, well, never mind.

If it isn't in the datasheets, maybe try contacting the manufacturer directly?

Peter.

[TomW]

Amanda;

I seem to recall my C-60 [before it died] would equalize for 5 hours [manually initiated] or until it reached a certain voltage according to the manual. Flooded lead Acid.

Cheers.

TomW

[WXYZCIENCE]

Amanda, I agree with Tom, my C-40 equalizes four hours into 6 x 6 volt flooded golfcart batteries. Joe

[SamoaPower]

I wish I had that simple rule-of-thumb answer to offer you Amanda but, as so often happens, the issue is complex enough to require a complex answer.

Since you mention load dump controller, I assume you have a RE charge source that is time-variable in output. If you were using a constant current mains source, the problem is simpler. The variable nature of wind and solar is what makes battery charging difficult and is why all commerical controllers (that I'm aware of) fall short of the mark in accurately controlling battery charge (not just equalization).

Usually we equalize as a prevenative maintenance measure or to correct an existing out-of-balance problem detected by SpG measurement. In the latter case, the best way uses a constant current source that charges until SpG's stabilize for a three hour period with battery temperature as the ultimate limit. I personally like to limit temperature to 40C given it's dire consequence on battery longevity. If the temperature tries to exceed the limit, current must be reduced.

Frankly, I don't think time is the right parameter to use to equalize although many use it. The Trace/Xantrex 'C' series controllers use time and a voltage cap. They limit time to two hours with a voltage cap of one volt above the set bulk charge point. This is probably safe in most cases with reasonably matched charge sources and bank capacity. However, consider the case of high output sources and a relatively small battery bank. If the batteries actually didn't need much equalization, the temperature limit could easily be reached in less than two hours even with the voltage cap. In the opposite case, low source current availability (cloudy day and/or low wind) and/or a large capacity bank, it's likely equalization won't be completed. I don't see how time can be used with a variable source current.

So, if we don't use time, what do we use? It would seem that Amp-hours charge (above 100% charge) related to bank ACTUAL capacity would be reasonable. The problem with this is threefold. Most users don't measure Ah into the battery. Most users don't know their ACTUAL bank capacity. Most users don't measure battery temperature.

My answer to this dilemma is to give the batteries a mini-equalization charge for each charge cycle by simply raising the bulk voltage set point of the controller. The normal set point for my flooded cells at 27C ambient is 14.2 volts. I set it to 14.8. The result of this is that with my four year old bank of T-105's, The SpG's still match within 5 points.

[Flux]

I have come to the same conclusion as SamoaPower. Solar may not be so bad as you have a reasonably constant charge, but with wind it is never the case unless you choose a day with lots of wind and things are furling.

Even detection of the point to start equalising is difficult, a battery may reach dump voltage long before it is charged at 20A and will continue to run below dump with lower currents.

All the battery data seems to be based on charge at a specified current and I have never understood what to do under wind charging conditions. The more advanced the battery technology the worse the problem.  For things like traction batteries it seems best to wait until they have been gassing well and then clamp them to a float voltage.

I have never found out what is the best way for AGM and I have never used gell as they seem even more fussy.

If you solve the detection level issue I still don't see how you time the charge with a variable current, which will often let the volts drop below detection level. I can only assume you use a charge integrator ( and this is virtually ampere hours), or time the intervals when the battery is at equalising volts.

It will be interesting to see if anyone has better answers to this, it has forced me to think about AH meters but even then life is not that simple but it does at least give a far better idea of state of charge, which volt readings are not very good at. As long as you can be sure there have been no deep discharges then for most batteries, setting the charge level to about 14.4v ( depending on temperature) seems to work tolerably well for AGM and for flooded then I would agree with Samoa that 14,8 is reasonable and if it continues for long periods it would need dropping to a float level at about 14.

Flux

[commanda]

Thanks for all the responses so far.

To summarise;

Trace seem to use 4 or 5 hours, so I'm in the ball-park here.

My system will be mixed solar & wind, rather than wind only, so some comments about wind only will be taken under advisement.

May need to consider monitoring battery temperature whilst performing an equalise charge.

I think my comment about standard deviation of Specific Gravity is probably a good indicator of when an equalisation charge is required. More research (or experimentation) is required to derive a workable formula.

Just so you all understand exactly how my dump load controller works. When in float mode, it will start switching in dump loads at some particular voltage. When switched to boost/equalise mode, it moves this voltage up (about 1 volt for a 24 volt system). So, if the charging current drops, it simply drops out dump loads trying to maintain the higher float voltage. Worst that can happen, with a fixed timer, is that we under-equalise; no harm done.

It still takes some intelligence on the users part to decide when is a good time to start an equalisation cycle.

Amanda

[SamoaPower]

My Trace/Xantrex 'C' Series manual states:

"It will automatically stop the equalization process after accumulating TWO hours of operation at a voltage above the bulk setting."

I also have a mixed system and notice as much variation in the solar output as the wind as the clouds float by.

"I think my comment about standard deviation of Specific Gravity is probably a good indicator of when an equalisation charge is required."

While that may be true, ideally, you would like the battery never to get to that point because it indicates some hard sulphate has already formed. Daily modest equalization causes no harm if temperature and water loss is monitored.

I think I may have some problem with your terminology. Float mode usually refers to a maintenance charge voltage to compensate for self discharge with minimal water loss. Around 26.4-27.0 volts for a flooded cell 24 volt system. Normal charging (bulk mode) is usually limited to 28.4-29.0 volts. Equalization is often done without a voltage cap to allow it to happen in the shortest time. but in any case should be allowed to go up to about 31.0 volts if equalization is actually required. I'm not sure how to relate all of this to how you describe your dump load controller.

I certainly agree with your comment about required user intelligence.

[Nando]

Amanda:

Equalization has a variable time current charging factor.

If one looks the charging current set at 2.4 volts/cell for new batteries the current will increase to certain level and stay for certain time, once all the cells have had deep charging the current starts to drop, more or less like a dome.

This is the moment to stops the equalization charge -- NOW -- for used batteries the equalization voltage is going to vary depending on the "age" of the battery and how much it has been "abused".

One needs to remember that most batteries are "murdered" by their owners because no good battery management.

The ideal charger needs to have some intelligence to see how the voltage and the current is varying to define the cut-off point.

The manufacturers imply or indicate some standard charging profiles for "STANDARD" battery usage, where the batteries are handled carefully, almost with "Silk Gloves".

Some aged batteries may require more frequent altered profiles for the bank to store enough energy for the following discharge cycle.

AGED batteries should have, periodically an equalization charge that may be different from normal.

RECENT CASE: A fellow has a 24 volts bank with Trojan L 105 = 1200 amp/hour capacity.

He has been charging the batteries following the manufacturers recommendation and for many months the hygro readings have been below the 1170 levels and some much lower.

He asked for help and during a period of several weeks, I showed him how to manage the charge to recuperate the batteries to a healthy hygro readings level, we succeed to save all the batteries except 2 that had a very low hygro readings in one cell each.

The BULK charge cycle Voltage was increased to 29.6 Volts and the time to 3 hours and the equalization raised to 2.58 V/cell for 4 hours and repeated about every 2 or 3 weeks, depending this on the hygro readings.

This way he has a re-vitalized battery bank, we are now trying to save the 2 batteries with bad cells.

A good charge controller should have the capability of reading the charge current and define the charging cycles times and as well proper charging sequences.

Many times is necessary to equalize the batteries for several cycles immediately after each discharge cycle to insure that the bank recuperates its capacity levels.

Does this help you ?.

LASTLY : Charging profiles are based in the volumetric capacity of a battery = amp-hour, that may vary from 10 to 30 % of amp-hour capacity .

Nando

[commanda]

Sorry about my terminology. I mostly use SLA's for UPS application.

Batteries usually have specified two charging algorithms;

one for standby use eg: UPS.

one for cycle use.

I did get the terminology mixed up. Of course I mean what you call normal or bulk charge, not float.

And thanks for the pointers on the voltages to set it all at.

I might also modify the timer into accumulate mode.

Amanda

[commanda]

Nando,

Thankyou for your comments. The negative slope detection method I am well aware of. Have designed Ni-Cad chargers which use this principle.

Also aware of dual-state charging techniques, like the UC3906. Use these in our in-house UPS for our beltweighers.

It becomes a little difficult when working on a live patient as it were; that is, where the battery bank is still running the household.

What I am trying to achieve here is not the perfect charger, nor even a close approximation of an ideal charger. The apparatus in question is a dump load controller.

The goal is a system which can safely be left to its own devices, without having to be nursed (or watched) constantly.

It is left to the users intelligence to monitor SpG and decide when to initiate an equalisation charge. Trying to automate this will spark a philosophical monologue of what I think of "idiotproof". Lets not go there.

Your experience with this fellow you are helping to salvage his battery bank, goes towards what I hinted at in relation to regular record keeping of SpG, and instigating an equalisation cycle based on average & standard deviation of those SpG readings. I think if one did this regularly, and also recorded the details of the equalisation cycle, and again SpG readings after equalisation,  one would soon develop a feel for what it takes to properly maintain any given battery set. This would automatically take care of changes as the battery bank ages.

Amanda

[ruddycrazy]

Hiya Guy's an Gal's,

                    With my house 24 volt array I have 12 - 2 volt 600 amp/hour Sonnenschein batteries, a 2400 watt selectronic sine wave inverter and a 80 amp battery charger ( Staff & stanbury). Every month I run the genset and don't even worry what the inverter is saying when the lcd on the battery charger is showing 49 amps going in the batteries the inverter goes into float mode but I leave the charger running until there is less than 2 amps showing on the battery lcd. Then I know the batteries are fully charged. As far as a stand alone system I bought for $25,000 before govt rebates I need to do this every month to ensure battery life. Now to other people in Oz DON'T USE THE SOLAR SHOP IN SA and if you need proof why I'll send you their warranty statement which basically leaves you in the lurch.

Cheers Bryan

[commanda]

Bryan,

The Selectronic inverter is also a charge controller; don't know how many people on here would know that. You're likely to get some strange replies to your post.

So you disconnect all house loads before running the genset?

What you're doing sounds fine. If I spent $25000 dollars on an RE system, I'd spend another $7.00 and buy a good hygrometer. I know it cost $7.00 because I bought one today.

As for warranties, they're about as good as insurance policies; not worth the paper they're written on. This is just my personal opinion, please not let us do the insurance debate again.

And if you really want the real story; I was 9 years old when our under-insured house burned down. I spent the rest of my childhood in poverty growing up in the slums; and no my parents couldn't afford to send me to uni.

Amanda

[wpowokal]

Amanda this may help.

http://www.homepower.com/magazine/downloads_batteries.cfm

I float my 24v system at up to 32V, temperature is critical but I can not ram enough down their throat to cause any probs in that area.

If I'm using the DG its around 40 amps plus any sola and wind so potentially say 80 amps max average for a 5500A/h bank aproximatly. I monitor SG of one particular battery daily, gives me a good indication of where i'm at.

If I was to set a time 5 hours would be reasonable IMHO so long as voltage and temperature were not exceeded.

Sandia National labs have some very good articals from their reserch,

http://www.sandia.gov/Renewable_Energy/renewable.htm

search around on their site, they recomend high end voltages with maybe disconnecting any voltage sensative equipment.

Realy it depends to some extent what source you are using to ram the last few amphours in. Re is "free" mains or a DG costs, and it is sugested posibly not cost effective.

I have a copy of a batery book taken from a Tasie Re companies site some time ago, however they seem to have removed it, very informative.

Get my 'e' mail from Bruce at Mt Tambarine and I will send a copy, I will not publish my adress as too many offers to buy Viagra arrive. I have ideas for an out of the box generator and would not mind bouncing some ideas off you since it involve a switch  mode power supply.

regards Allan on the other side

[oztules]

Hi Commanda

I'm not qualified to answer yet on the RE situation, but have had many hundreds of traction packs in my ownership for many many years.

In that time I learnt that American ferro resonant battery chargers in the hands of my cleaners, could destroy 4 or 6 t105's (24 or 36v) in less than three months. Italian non-ferro resonant chargers could kill off the batt packs in around the same time.

These were designed by engineers for a specific purpose, but not with our end users in mind. As a result, I have bought and had killed thousands of deep cycle batteries over the years by a combination of poorly directed end user design, and end users who couldn't read english, and didn't care if they did.... so my response is coloured by bitter experience.

A bulk charge at a reasonable current rate, and at 28.1v switching on a 14 bit counter at around 1 second per pulse. At the end of the count, a latching circuit pulled in thus stopping the charge.

This simple system was by far the best for trouble free operation.

So I ended up using this circuit (similar to your idea I think except 14 bit)

http://www.otherpower.com/images/scimages/5171/chgrschem.pdf

with this board

http://www.otherpower.com/images/scimages/5171/chgrpdb.pdf

Maybe you can get some ideas from this. The bulk charge stops at 28.1v and then charges at 2(power 14) timing pulses which  brings it in to around 4- 5hrs or so. It then shuts off (RC network on 4060 controls the length of pulse). All charging done above float voltage should be done at low current, otherwise a good maintenance regime is critical..... here you set dump voltage at 31v say for bulk and eq charge and then back to 28.2v float at the end of eq charge.

I can't take a pic as the camera is up the mountain at present (lost).

Well, I hope it helps in your deliberations a little, if you want a spare board to fiddle with, email me. If anyone else outside Aus wants the pcb file I will put it in my files as soon as I can get it out of my old computer as a simple pcb file (this machine has it stored as a dbb file.... too big.

...............oztules

[commanda]

So, you're basically confirming that what I'm doing is in the ballpark.

2 ^ 14 / 60 x 60 = 4.5511 Hrs.

I'm not actually building a charger as such. The dump load controller, when in normal charge mode, will bleed off current once the batteries reach a certain voltage.

When the user manually starts an equalisation cycle, the voltage at which it starts to bleed off current is increased. The timer will reset the dump voltage point once it has been at this higher voltage for a certain amount of time.

Perhaps a small failing with this, is we are not limiting the charge current when in equalisation mode. I don't see this as a major problem, so long as the system is sized appropriately, and the user uses some common sense.

Amanda

[commanda]

As a footnote, I should add that I'm adding a battery temperature meter with a comparator output which will terminate the equalisation charge should the batteries rise above a certain temperature.

Amanda

[wpowokal]

Amanda what is being proposed is like trying to determin how long a piece of string is.

However if I was equalising a 24v bank I would set a maximum voltage at 32V (note this may require disconnecting dc loads)

Gassing begins at around 28.5V, as to time it depends on the banks amphour rating, stay within 10%, battery temperature and which way the wind is blowing.

For my money I would equalise for 5 hours as indicated by others, frequency will depend on depth of regular discharge, age of batteries and which way the wind is blowing.

So for a set and forget systed, like I'm going to WA to see Allan for two weeks, set maximum before dump load to 28.5 volts for a 24v system.

Equalising frequency is also like how long is a piece of string, older bateries may/will require more frequent equalising, I like a two week interval, a month would be fine, some bold soles say 3 months. Realy do it at short intervals and if you are not adding large amounts of water then by my yardstick you are in the ball park. Realy there is no difinative answer, it is very dependent on each individual installation.

regards Allan down under