Calculating battery life 1

[kennyden]

Hi,

I'm trying to calculate the service life for my battery powered application.

Here's how I calculate it. Please le me know if this the right way to do this.

I'm using 3X1.5 Volts AA type alkaline batteries -put in series- (energizer titanium e2) rated at "3135mAh (down to 0.8 volts)". My application works at 4.5 volts down to 2.6 volts.

So to get the total energy from the battery I multiplied 4.5V by 3135mA by 3600 (sec) which gives about 50787 joules.

I then took the total battery energy and divided it by my application's average power consumption per day at 4.5 volts, which is about 16 joules. So 50787/16.06 = 8.6 years !!

Now I know that theorically, 8.6 years is real long for 3XAA type alkaline batteries. I would surely need to caculate the energy derating over the years.

Nevertheless, I need about 4 to 5 years of battery life. Assuming the application is used at 23 degC ambient, will this do it ?

May I use the manufacturer rated energy this way ? Is the 3135mAh a nominal value at 1.5 volts ? Has anyone been dealing with such a life span for alkaline batteries for low powered apps ?

BTW, the PEAK current drain from the batteries never really exceeds 65mA for no more than 20msecs (100 times a day). The application in standby consumes only 23uA.

Thanks for helping me. I'm no battery experts and my caculation results seem a bit unerealistic to me.

Thx

 

[IRstuff]

You should probably see if you can get the datasheet that has life in Whr. Duracell's Ultra AA is rated at about 4 Whr. Your numbers come out to about 14 Whr. Using Duracell's value and your daily energy value, the batteries will only last 2.5 yrs.

Not sure what power my watch draws, but the lithium batteries it has seem to last years.

TTFN

 

[kennyden]

First, thx for your quick reply.

Have you extracted this data from their specification curves or did you get this data elsewhere ?

Are you saying 4 watts per 1.5 volts battery ?

1.5 X 3135mA/h is about 4.7 watts for an hour (or possibly 4 watts nominal taking into account battery discharge), that is pretty close to your data !? But then again, that battery can't possibly be able to provide 3.135 amp for an hour !!

Again, thx for helping me see through this. I must admit, I'm a bit confused.

 

[IRstuff]

It's actually not that far off. The battery's life is HIGHLY dependent on the discharge rate. A standard AA is capable of dumping 10 Amps into a dead short, not for very long though... At the extremely low discharge rate that you're talking about, the battery should perform to its rating.

I'm not sure why you're surprised by the 4 Whr. Your calculated value of 50787 Joules, which is Wsec, when converted to Whr is over 14 Whr.


The datasheets are on the Duracell website, which I can't get to at the moment.

TTFN

 

[kennyden]

Yes, 50787 joules is what I can get from 3 times 1.5 volts AA batteries.

So basically what I'm trying to say is, the rating I got from 3 batteries @ 3135mAh times 1.5 volts is 14.1 Wh which gives a theorical life span of 8 years are so. You said that the ratings you had were in Wxh and are at 4 wattsxh. Then 3 times 4 wh is 12wh and so 43200joules or about 7.3 years. You mentionned 2.5 years so I figured you meant 4 watts for 3 batteries. Otherwise you would have ended up with the same results. Is that what you meant ?

If so, this again leads to the 7 or 8 years life span which I find surprising. Practically, I assume this value to be way too long for alkaline batteries or, this was my question : is it possible ? Or else, how much percentage of this value should I expect in real life ?

I know a lithium battery has a shelf life of 10 years. Assuming you draw micro amps from these batteries, this would make possible to make a low power apps (like a watch) live long.

To me, the rating "1.5 volts X mAh " is another way of saying "watts" X hour. Or is it (in terms of battery science of course) ? I've read that every manufacturers express their ratings in different ways, given different discharge rates, temperature, max current pulse etc...

And you are right, I should not be that surprised considering the internal resistance of the batteries being around 125 mohms (i thought that was higher, sorry), this makes it possible to get 10 amps from these cans (for a "short" while at least).

Well, I guess I'll need to read their specs over again or I'll try to contact them directly and see what they think of my calculations. I guess this battery life thing makes me a bit nervous since it is the whole key to success for this project. And no, I can't add more batteries or have C or D s instead ..

Thanks for your valuable time.

Regards

 

[IRstuff]

The battery ratings are only a guide, based on their nominal discharge rates, which is significantly higher than what you are trying to draw.

Their self-discharge rates need to be considered for this type of application.

TTFN

 

[kennyden]

Thanks, TTFN. I've asked some manufacturers for more specific data as neither of them consider "very low" current discharge.

Obviously, self discharge will become important in such low power apps, but how important will this be, that was my initial question really about. All self-discharge curves I'v seen were shelf life specifics.. no good for me.

I'll see what they have to say about it.

Have a good day.

 

[lcsjk]

Kennyden, you need one little bit more information to get your final answer. The battery capacity rating is called the C-Rate and is the current that the battery can sustain for 10 hours before dropping to the "fully delivered capacity" voltage. This means that your 3135mA-H battery can deliver 313.5 mA for 10 hours before its voltage drops too low, or 31.35 mA for 100 hours etc.
For currents which are much less than the C-Rate of 313.5mA the batterylife can be calculated based on the rated capacity of 3135mA-H.
In your case of 23uA and another 1.5uA based on averaging 65mA for 20msec(x100) during a 24 hour period to give a total of about 24.5uA, your current is very low and the battery discharges based on its full rated capacity. 3135mA-H divided by 24.5uA is 127960 hours, 5332 days, or 14.6 years. You can calculate it differently, but what it really amounts to is that the battery's internal resistance will probably be a bigger factor in run-down than your circuit, so look up the manufacturer's shelf life.