47
reported. However, the number of profiles performed is not necessarily the cycle life and should not
be reported as such. Detailed results of the reference tests are reported over life as described under
these specific tests, including the magnitude of adjustments made (if any) due to the measured
temperatures being above or below the nominal temperature. In addition, degradation of capacity,
pulse power capability, CD and CS Available Energy, and Cold Cranking Power capability as a
function of life (i.e., number of test profiles performed) should be reported graphically.
The value of cycle life to be reported for a device subjected to cycle life testing is defined as the
number of test profiles performed before end-of-life is reached. In general an end-of-life condition is
reached when the device is no longer able to meet the targets (regardless of when testing is actually
terminated). The ability to meet the targets is evaluated based on the periodic Reference Performance
Tests, particularly the HPPC Test results. When the power and energy performance of the device
(scaled using the Battery Size Factor) degrades to the point that there is no power and energy margin
(i.e., CS or CD Available Energy is less than the target value at the target power), the device has
reached end-of-life. In addition, the inability to meet any of the other technical targets (e.g., the cold
cranking power, efficiency or self-discharge target) also constitutes end-of-life. The basis for the
reported cycle life value (i.e., the limiting target condition) should also be reported.
39
If the cycle life
based on power and energy performance is very near the target, the end-of-life point may need to be
interpolated based on the change in HPPC performance from the previous reference test.
4.10 Calendar Life Test
The raw data from calendar life testing are the periodic reference performance parameter
measurements for all the batteries under test. The objective of this data analysis is to estimate battery
calendar life under actual usage in a specified customer environment. Typically, the environmental
specification will include a cumulative distribution of expected battery temperature over its 15-year
life in, for example, the 90
th
percentile climate among the target vehicle market regions. These
temperatures will vary, and will generally be substantially lower than the elevated temperatures used
for (accelerated) calendar life testing. Note that for most (> 90%) of its 15-year life, the battery will
typically be in a non-operating, vehicle-parked state.
Predicting battery life is a desired outcome of testing. There are various approaches to constructing a
battery life model. One is theoretical, using various physical and chemical processes that may occur
in the battery, which degrade its performance. A second is fitting a curve to the data. The following
discussion is limited to the latter approach and is meant to illustrate a general approach to construct a
reasonable, data-based model. For a more advanced treatment of life test results, refer to the
Technology Life Verification Test (TLVT) manual, Reference (4).
Curve fitting may be applied to resistance, power, energy, and capacity data and is transparent to
battery chemistry and technology. Curve fitting is an interpretative, deductive approach to
understanding the performance degradation process. Assuming that the battery test was performed
with a number of different temperatures over a number of reference performance tests, the most
39 Efficiency and Self-Discharge are not necessarily measured at regular intervals during life testing, so the point
during life cycling where such an end-of-life condition is reached cannot always be determined with high accuracy.
Typically the test results showing that the targets are not met would be reported, without attempting to interpolate an end-of-
life point using two test results widely separated in time.