Lead acid batteries internal impedance vs technologies and capacies 1

[homoly]

Hello,

I would like to ask you if you have some experience with interval of spreading for internal impedance in the case of lead acid batteries manufactured with different technologies ( open, sealed, VRLA, gel .... ). The second question is if the internal resistance of the battery with higher capacity can be estimated with the data available for a battery with lower capacity, in the case both batteries from the same product family and manufacturer.

The point is I was trying to implement the calculator of the remaining capacity of the battery connected to the voltage inverter, and to data measured voltage by inverter and configuration done by an user ( capacity and technology ) for the proper calculation also the information about internal impedance would be necessary. So the question is if this can be given by approximation based on some statistical spread of impedances for each manufacturing technology with some reasonable precision, or it would be necessary for the user to give as a part of the configuration the exact impedance value from the battery datasheet.

Any hints are welcome.

Gabriel.

 

[IRstuff]

Wouldn't you think that the "cold cranking amps" rating of the battery would be a clue to its internal resistance?

TTFN

FAQ731-376

 

[homoly]

Well I am afraid it is not so easy as several lead acid battery technologies exists with different electrode construction and material doping and thus different internal impedance.

In general I think the most relevant parameters that the internal resistance depends on are actual charge, temperature and the aging. As far as the aging information is not provided in general by a manufacturer the simplistic model should cover Z( Q,T). I have made some research and I can to the point when the equation based on cranking amp was not sufficient ( probably you suggest Z = Ubat / I ).

The reason why I would like to get this info is, I would like to make a calculation of the remaining capacity for several lead acid battery types and to flash some alarm when the capacity reach f.e. 20% of the nominal. The practical usage would be to indicate to the user that the battery is deeper discharged and he should turn off all unnecessary consumers and left the remaining power f.e. only for radio or other more important devices.

Other usage should be to evaluate the area where in normal condition the impedance can be by curves ( one curve for equal temperature and each temperature would have its own curve ). The upper limit curve should be determined by aging factor ( or sulfation ) and the lower limit curve should be determined by a short circuit in one or more cells of the battery ( because very low impedance ).

This could be used to flash the alarm that battery is worn out, or detect a dead or sulfated battery.

 

[DanEE]

This is probably the best compilation of battery internal resistance (DC load measurements) you'll find online.

http://www.alber.com/BaselineRef.htm

Otherwise contact the engineering reps for the specific battery manufacturers you have in mind.. Our business has obtained that data for the GNB Absolyte batteries that we deal with extensively.

If you are talking impedances measured by AC signals passed through the battery, then I recommend you read some of Albers white papers on the fallacies in this method. No battery manufacturer we have talked to will accept this method on a warranty claim..

http://home.4x4wire.com/deddleman/

 

[homoly]

DanEE, Thank you for the suggestion, I know Alber materials and they are realy good .

I was doing some reasearch if the battery impedance evaluation could be done in some way with the software of the charger/inverter ( without needing to add additional hardware to charger or disconnect the battery and use the special measurement tool with DC current ).

The point is to indicate the remaining capacity and battery health status when charger/inverter is connected to the battery.

Unfortunatelly this is not the case when the charger/inverter can be connected by only one type of the battery but in general the device can be connected to SLA battery as well as to deep cycle marine batteries and this is the tricky part. In the ideal world we should supply the customer with a given types of battery that we know the parameters and in this case maybe also the effects of aging should be included in the software. Without knowing the battery type it would be very hard to evaluate the remaining capacity ( with reasonable precision ) only by reading out the battery voltage and discharge current and to evaluate the correctitive coefficients.

So this should be something like mathematical model of the batteries , but the question is for a given battery types what are the dependences of the internal impedance on the selected parameters and what is the spread on interval of "normal values". This might be an equivalent for the mathematical model of the motor in the case of motor drives as it was my intention.

Maybe the simplest solution for health checking in the case of closed systems ( stationary applications ) I was thinking about it to connect some external power resistor to the device and periodically discharge through this load resistor during some period of time and if the voltage will not fall bellow the preset the battery is OK. The backdraw of this solution is a need for additional external resistor and it is feasible to use only in UPS like aplications of the inverter/charger, not in the case when the inverter function is used mostly and the battery is charged only by an opportunity charging.

 

[ScottyUK]

Some UPS designs back off the input rectifier to a level high enough to carry the load but below the voltage of the battery during test. The battery supplies the load and there is no ripple from the rectifier. Systems I have seen using this method give a go / no-go on the battery but it may be capable of being adapted a little further.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[IRstuff]

I'm not sure what difference is where the resistor is. Either way, you'd have to radically reconfigure the charger to do a task it's not designed for. You're essentially asking for a circuit to sink current, when it was only designed to source current.

TTFN

FAQ731-376

 

[homoly]

IRstuff,

As I mentioned it before in text this is charger/inverter device ( you can imagine it like offline UPS ). The idea about external resistor is about loading the inverter section of this combi device and checking the voltage drop during fixed time interval. When the voltage would fall too low beyond the threshold voltage the battery should be considered as worn out.Lets say this testing would begin always from top charge stage of the battery charging and with fixed value of external load resistor the discharge condition will be every time more or less the same.

 

[DanEE]

The concept of power supply designs that could provide tightly regulated power and then sink current (in this case the supply's own stored energy) on command were fairly common on large scale computer power systems due to memory technology requirements back in the late sixies and early seventies.

For example, the bipolar technologies used by mainframes particularly in memory chips and the related logic families required multi-level voltages, one of which was essentially a low current draw biasing voltage and the much higher capacity primary operating voltage. There were very tight specs on the speed at which the primary operating voltage had to be removed if the bias voltage went out of tolerance. Otherwise you would have a very large, expensive system with a lot of destroyed logic.

Therefore the power systems included sequencing controls on start up and shutdown and, crowbar circuits in many cases to pull a supply down quickly on shutdown or a malfunction. These systems had large capacitor energy storage (to meet other requirements) and when crow-bar'd, considerable current flowed to get these systems down in the time required.

http://home.4x4wire.com/deddleman/

 

[IRstuff]

I think you need to read those white papers. The duration of the discharge has to be VERY short and VERY controlled. This requires special test equipment.

TTFN

FAQ731-376

 

[homoly]

Thank you for the entries! I will explain on concrete example why would I need the internal impedance spread depending on manufacturing technology ( VRLA, maintanence free, SLA, marine deep discharge different eletrode doping ).

Lets assume the capacity is given by a user in a configuration program, the battery voltage, discharge current and temperature is measured by a combi device.

Then for a given battery type of a given manufacturer I should be able to determine the state of charge ( SOC ) of the battery. Mathematically SOC = f ( capacity, temperature, discharge current, battery voltage ) and the function itself should be made based on a datasheet parameters and look up tables provided by a manufacturer.

But for a various types of lead acid batteries ( with various internal impedances ) the f function will produce an interval of probable SOC values instead of concreate values.

So the real question should be maybe if this interval is narrow enough to use only one generic function for all types of lead acid batteries or it would be necessary to evaluate more functions of each battery type.

The resistor discharge test and battery tester should in this case validate if this model is still working or the battery is worn out and it should be replaced.

 

[DanEE]

One thing I forgot to add that may be relevant is that several of Alber's white papers point out that the internal cell resistance does not vary significantly due to state of charge.. that is as long as you are staying within the proper SOC limits.

You hear a lot of people saying the voltage change from charge to discharge is due to the resistance changing based on SOC.

That is overlooking the obvious changes in electromotive force of the anode and cathode metals as they convert from pure lead, lead dioxide and lead sulfate as part of the charge/discharge process.

I'm not sure if you are trying to determine state of charge e.g. run time left in the battery, or overall battery condition (which internal resistance would be a good indicator).

I have found it difficult to find good publicly available scientific papers on internal battery chemistry, but this one (for AGMs) is one of the best.. and may have some useful information to you..

http://www.tms.org/pubs/journals/JOM/0101/Nelson-0101.html

The only comment I would make based on our field measurement data is that the issue discussed in this 2001 paper, of voltage spread due to state of gas recombination and uneven electrolyte fill has apparently been improved..

We see a number of batteries that maintain a very small voltage variance across the cells.

http://home.4x4wire.com/deddleman/

 

[homoly]

Thank you for the valuable post. Well, I was trying to implement both but run time left with higher priority. The best indicator of SOC is clearly the specific gravity of the electrolyte, but I was trying to find out the best method of evaluate the run time left ( SOC ) only from the parameters measurable directly with the combi device.

Similary I was trying to find out the method for dead battery detection based only on parameters measurable by a combi device.

The internal resistance spread ( as I thought ) should be important to know how to modify a Battery Voltage -> SOC % look up table for various battery types ( f.e. 10,5V -> X% SOC in the case of SLA battery but it can be Y% in the case of VRLA ). So the real question is if X and Y are close enough to neglect the difference and use only one battery voltage vs. SOC look up table for each lead acid battery technology.

Thanks for the link, I will study it for sure.