BTECH Development of the Impedance Based Monitoring Concept
When BTECH's engineering team was still at Singer Corporation, years of research were done to determine the best method to predict battery health. At the time, many experiments were made to determine which properties to trend in order to obtain the most accurate picture of battery health and many different methods were looked at. The one that was found to tracked capacity with the most accuracy was the battery's impedance. By measuring the individual battery cell's impedance when new and using it as a baseline, an impedance rise was determined to best correlate to the capacity and thereby the health of the battery. A number of key decisions were made concerning the measurement method based on careful laboratory research:
1) Impedance the Leading Indicator of Failure – From the earliest of tests it was clear early on that an impedance rise is the leading indicator of failure for most battery types. In the real world, impedance has found to correlate with the capacity of the battery to a large degree, as any basic battery model includes the capacitance of the electrolyte as a component in the measurement.
2) Resistance Ruled Out – Any method measuring resistance using a DC discharge requires a deep discharge of the battery to obtain meaningful information. Because an on-line battery health monitoring system would be doing multiple reads more often, we needed to be sure that the method would not discharge the batteries – so any method measuring resistance was immediately ruled out. We could not see as much value in measurements not including the capacitive component of the battery – the ability of the battery to store and transfer energy – which resistance measurements do not include.
3) Test Must Not Place Load on Batteries – Impedance is a better bet because the alternating effect of the test current eliminates the need to do a deep discharge on the battery, making it possible for the monitor to operate in the voltage spread between float and the battery's natural open circuit voltage. Without bringing the battery under the open-circuit voltage during a measurement, we can state that the monitoring system does not discharge the battery.
Once impedance was chosen as the ideal parameter to track, other challenges needed to be overcome:
- Impedance Frequency – early on it was detected that using a straight 60 or 50 Hz line signal would produce unpredictable results, because of the problem of AC ripple inherent on stationary battery systems. This ripple current is constantly changing – within milliseconds – and is dependant on the UPS design, the state of health of the capacitors in the UPS and the actual load on the system at any given time. BTECH's experience is that this ripple is unpredictable, and any algorithm developed to take this into account will inevitably fall far short on most real-world systems. The end effect is inconsistent data. As a result, BTECH chose a frequency between the 3rd and 4th harmonic of 60 Hz as the frequency that provided the most reliable and consistent data as it virtually eliminates the effect of ripple on the data. This method is patented by BTECH.
- Impedance Signal Type – BTECH also found that using a sinusoidal signal that goes above and below the battery voltage will of course cause a current that will enter and leave the battery. Because the more complete model of a battery shows a non-linear charge vs. discharge current relationship, impedance measurements cannot be obtained accurately with this method. BTECH has chosen to use a pulsed DC signal that simulates AC without the charging-discharging cycles. Tests showed this method to provide the best and most consistent data. This method is patented by BTECH.
- Test Current Signal Strength – BTECH's experience has been to show how important it is to provide a test current signal strong enough to get data with the best possible signal to noise ratio. Our systems are configured with a load circuit customized to the battery size, providing a load signal between 5 and 20 Amps while assuring that no detrimental load is placed on the battery system.
- Advanced Digital Signal Processing – BTECH has also developed a proprietary algorithm that measures any background noise in the data. BTECH also removes this effect so that end results are to within 1% of accuracy.
- Actual Measurement Specifications – BTECH publishes real world measurement specifications showing a resolution of 0.1mOhm to 20 mOhms +/- 0.01mOhm. These are results obtained in the field monitoring batteries on real-world systems. Most competitors do not publish real world measurement specifications, and they can't even come close.
With over 6000 systems installed and a proven track record, there is no longer any doubt that our method works well in all real world applications. When you go with BTECH, you get the experience we've gained in over 23 years of building and researching battery monitoring systems in noisy, real-world applications; experience no other company has.