Rechargeable batteries are used today in many applications ranging from powering personal cellphones, and up to backup power for telecom and industrial equipment.
Battery failures can pose safety concerns, and lead to failure of critical equipment.
Known examples are the Samsung Galaxy note 7 and Boeing 787 Dreamliner, however many similar incidents have occurred in various industries.
According to Failure Modes Database, the probability that a battery failure is due to short is 2.9%, and probability of failure due to battery expanded is 3.1%. These failure modes have the potential to cause an explosion.
Therefore, it is important to install and handle batteries according to the manufacturer recommendations.
Unfortunately, spurious failures may occur even if the batteries are properly installed. This raises the question:
How to calculate battery failure rate?
Calculating the battery failure rate is not as simple as calculating a resistor failure rate. While standard electronic equipment is assumed to have a constant failure rate, batteries are known to age. This means that the failure rate increases with time. A Weibull distribution is usually used for modelling battery ageing.
Consider a battery with Weibull shape parameter β=5, and MTBF=100,000 operation hours.
The failure rate follows the following curve:
How to reduce battery failure rate?
In order to keep the failure rate sufficiently low, periodic battery replacement is advised. For example, replace the battery every 20,000 hours. The following curve presents the failure rate with periodic battery replacement:
The effective constant failure rate that gives the same failure probability as the case above after 20,000 hours is 1.044e-8 (1/hour), and effective MTBF is 95,778,382 hours i.e. MTBF increased by a factor of 957.783!