VRLA battery lifetime fingerprints - Part 1

With over 20 years of continuous monitoring of batteries, archival of a trillion points of data, timelines, and trends of over 1.2 million battery units, we are finding some common aging history. As batteries age to the point of replacement, individually, and collectively in single and multiple string systems, we are finding individual and combined characteristics, collective and distinctive fingerprints indicating aging and end of life conditions. In this paper, this is shown and described both in descriptive and graphical presentation. These individual and collective traits cannot be simulated and seen conclusively in the laboratory setting. Accelerated life testing, due to the artificially shorter time required, reduced quantity of data, plus other restraints of simulation do not reflect real world data. These studied traits can only be reliably observed in real life usage conditions over months and years. This is achieved through continuous and frequent monitoring of key signature measurement parameters. Archival of data, trends, and events plus detailed analysis are essential to observe actual battery behavior over real lifetime periods. As a battery ages, there are changes over time in the internal ohmic value. As researched and stated in various IEEE standards, rises in Ohmic Value greater than 30 - 50% are significant and warrant investigation. While not definitive, this level of rise from a baseline of a 100% capacity battery is an indication that the battery has decreased below full capacity of when the battery was new and performing to manufacturer´s specifications. Further evidence showing correlation of ohmic values vs. capacity has been shown in extensive studies performed several years ago (2002) by EPRI (Electric Power Research Institute.) A great deal of controversy exists regarding the correlation of ohmic value vs. capacity shown and calculated during a battery discharge test. This controversy continues to exist regardless whether the battery us- r performs ongoing discharge tests after initial acceptance discharge tests or not. Most users elect to rely on ohmic values and trends plus other measurement parameters to determine RUL (Remaining Useful Life.) and replacement criteria. This is done without periodic discharge testing which can cause risk and additional costs. These aging changes are caused by time, temperature, electrical and chemical variances, mechanical and other anomalies, plus usage patterns, interfaced equipment, charge/discharge cycles, harmonics, and load levels. All aging factors will cause ohmic values to rise or change as the battery decreases in capacity. Some common aging factors causing Ohmic rise or change are listed in this paper. Continuous frequent monitoring and record keeping of battery ohmic values, plus other battery measurement parameters are essential to predicting end of life conditions. This is true for individual battery units, as well as the complete battery system. Measuring, trending, and archiving ohmic values over time, in addition to other measurement parameters that affect ohmic values allow for extensive predictive analysis. In addition, continuous monitoring, data analysis and archived trend analysis help maintain battery unit and system state of health (SOH), state of charge (SOC), and allow for the prediction of remaining useful life (RUL). Observation, collection of this data and the use of proven mathematical prognostic techniques and models can be combined for lifetime prediction and forecasts. This will be part 2 of this paper. Prognostic methods are varied and diverse. They include many modeling techniques. Examples include: Bayesian theories, Neural Networks, Moving Averages, Kalman Filters, and many other high level mathematical models. These methods allow for the prediction of future points in calculated curve series utilizing data from battery Ohmic values which are affected by other measurement parameters including temperatures, AC float and ripple v