Battery performance tests performed in 1998 (long 8K)

["Joe Smalley" <joes@worldfront.com>]

I type this document tonight from memory with the spreadsheet in hand to
remind me of the actual numbers. The tests were run 15 months ago and the
test system has been disassembled and several key pieces of equipment
dispersed.

A personal note: I hate plain text because I can't paste the graphs in the
text where they belong. I understand the reason for plain text and agree
with it. I have a slow modem too.

Joe Smalley
Rural Kitsap County WA
Fiesta 48 volts
NEDRA 48 volt street conversion record holder
joes@worldfront.com


I was given two flooded 8D batteries fresh from the factory to test:
1) to determine if there was a performance advantage to the cranking battery
over the deep cycle battery
2) to determine the rate of degradation of the cranking battery to determine
when the deep cycle battery eventually outperforms the cranking battery.
3) to determine the optimum temperature for the battery at the start of a
race.

A system was set up with the following components:
A heater to heat the battery.
A insulated box to allow the heater to warm the battery significantly over
ambient temperature.
A charger to charge the battery to the same state of charge each cycle.
A discharger that had programmable and repeatable performance.
A computer to monitor the status of the battery and stop the charge when the
battery was fully charged. (The computer also stopped the discharge at a
programmable state of charge.)
A water cooled load.
A 500 amp circuit breaker.
A thermometer that reads remotely to 0.1 degree F.
A circulation fan to keep the air moving in the box when the heater was
turned off at the beginning of the test.
An Emeter to measure the voltage and amperage of the system.
A pair of contactors to connect either the load or the charger to the
battery.
A power supply to power the emeter and contactors.

The test was set up to run 12 cycles on each battery at various somewhat
random temperatures.
Each discharge was to be run at 250 amps(+/- 2%) until 10.5 volts then it
stays at 10.5 volts and the current drops off.
The first few tests indicated that the box needed an insulated floor since
the highest I  could get the interior of the box was 101 F.
After the floor was insulated the temperature was capable of going over
130F. This was to become the highest temperature I wanted to use on the
batteries. I did not want to kill them early.

The tests were basically run back to back with each cycle following this
sequence:
The heat is turned to the desired thermostat setting and the battery is left
to soak in the warm box with the lid on.
The heater energy consumption is noted and when the high setting starts to
cycle, low power is selected.
When the low setting starts to cycle it indicates the battery is almost up
to temperature.
The battery is left to soak for another hour.
The battery is put on discharge and the heater is turned off.
When the load drops to 100 amps, the load is disconnected and the air
temperature inside the box is read.
Since the battery mass is stabilizing the temperature inside the box and the
heater is off, the air temp is very close to the battery temperature at the
end of the test.
The battery is put on charge and the lid removed if the next test is at a
lower temperature or the lid is left on if the next test is at a higher
temperature.
The thermostat is set for the next test.
When the battery is charged, the lid is installed on the box and the cycle
repeated.

Analysis of the data indicated several things:
1) The temperature had a profound effect on the capacity of both batteries.
2) The temperature had a profound effect on the internal resistance of both
batteries.
3) The cranking battery was losing capacity during the tests while the deep
cycle battery was relatively stable.

The first thing I had to do was remove the first two effects in order to see
the third effect.

Current capability as a function of temperature:

Impedance of both batteries was about 5 milliohms at 50 F.
At 75 F, the cranking battery was 3.5 milliohms but the deep cycle had only
dropped to 4.2 milliohms.
At 100F, the cranking battery was still improving at 2.4 milliohms and the
deep cycle battery had plateaued at ~3.7 milliohms.
At 112 F the cranking battery plateaued at 1.8 milliohms and stayed stable
out to the maximum test temperature of 128F.

Conclusion: Neither battery needs to go over 110 F to maximize current
output. The deep cycle battery is within 10% of its hot impedance at 85 F
while the cranking battery is still double its hot impedance at 80 F.

Ampere hour capacity as a function of temperature:

The deep cycle battery had a nearly linear relationship between capacity and
temperature.
50 F gives 78 Ahr
70 F gives 84 Ahr
90 F gives 92 Ahr
110 F gives 102 Ahr
130 F gives 111 Ahr
I did not test the battery over 130F for fear of melting something.

The deep cycle battery exhibited a similar slope but was not stable over the
number of cycles and had plummeted to 66 amp hours on the last test. The
cycle life was a larger contributor to the demise of the battery than the
temperature after the fourth cycle.

Conclusion: the hotter the battery gets the more capacity is available. (No
data was taken above 130F)

Ampere hour capacity as a function of number of cycles:

Since we have determined the temperature characteristics of the battery in
the last step, we can calculate out the approximate contribution of the
temperature to the capacity of the cranking battery. By subtracting the
temperature coefficient of 0.4 Ahr per degree F out of the results, the
degradation of the cranking battery becomes apparent:

Conclusion: The cranking battery starts out with nearly 10 amp hours more
capacity than the deep cycle but loses three percent of its capacity each
cycle on the first six cycles but the loss is about 2% per cycle on cycles 7
through 12. By the 12th cycle the capacity is 65 amp hours while the deep
cycle battery is stable at close to 85 amp hours.

Impedance effects:

The lower impedance of the cranking battery provides a higher discharge volt
age and thus more watt hours output than the ampere hour capacity results
seem to imply. The temperature effects on the impedance blur the true trend
on this data. The order of the temperature settings during the discharge
cycles needs to be more regular than was programmed during these tests.

Using the 100F impedance of 1.8 and 3.6 milliohms for the cranking battery
and the deep cycle battery respectively, the 250 amp load produces a 0.45
and 0.9 volt sag respectively. If the loaded battery voltage was assumed to
be approximately 11 volts, the cranking battery will put out .45 more volts
producing 4 % more watts. This added voltage will make up for the capacity
lost during one more cycle of the battery.

Conclusion: The cranking battery will outperform the deep cycle battery in
capacity tests for the first three cycles at room temperature and for four
cycles at elevated temperatures. In quick tests (races), were the impedance
of the battery is the most important characteristic, the cranking battery
will outperform the deep cycle battery well past the twelfth cycle.