Battery Monitoring

     


The Benefits of Monitoring Batteries

Batteries are the most likely cause of a UPS failure. Batteries are the most likely cause of a generator start failure. We have been involved in automated battery monitoring since the early 1990s. It has been well proven that batteries that are not maintained with a thorough PM and replacement program will fail and will cause a loss of power to the critical load equipment. Annual, semi-annual or quarterly checks are good but that is not frequent enough to be certain the batteries will be good until the next visit. Batteries, particularly the sealed type which are the predominant battery today, most often fail in a 3 to 5 week period from good to failed status. Thus, inspections or testing need to be performed at least at monthly intervals if not more frequent. The only method of doing this is to automate the process. Benefits of automation are 1) the system does not forget to do the test, 2) the testing is done the same way at the same test points every time and 3) the system even monitors between specific testing.

Upon realization that automation is the best way to go, the question is how to test. There are two accepted methods. The older way is load testing wherein a significant load is placed on the battery and the voltage drop is compared to known good values. If the battery holds voltage up it is good at that time. For voltage monitoring systems a load is required, either put on by the test equipment or by actual outages. If the test equipment introduces the load, the test itself ages the battery and that is not the best thing to do. If actual site load is used, the testing may not be done often as there may not be frequent enough and long enough outages to get meaningful data. Thus the monitoring system may be present but incapable of running as needed. If scheduled test outages are used, we are back to degrading the battery which is not good.

The second method is battery impedance. In this system an AC signal at high frequency is passed through the batteries during the test and the battery impedance is measured. This test does not degrade the batteries and can be done at any time. Battery impedance is a documented predictor of battery health and is the best early warning characteristic available to measure. The automated battery impedance test technology was invented by BTECH and is patented by BTECH. We have utilized BTECH equipment for years at critical sites and it has prevented many a failure.

Deciding whether or not to invest in the BTECH system is a decision to be based upon economics. The cost of the system needs to be compared to the cost of downtime. When downtime prevention warrants a UPS it probably also warrants automated battery monitoring.

A word of caution to those who believe the built-in battery monitoring in a UPS means their batteries are already monitored. UPS based built-in monitors are looking at overall DC voltage. The UPS cannot determine individual battery health as a bad battery is masked by the overall system. It is just not possible. If the overall battery plant is drastically failing a UPS based system may find out but it is already too late to solve without an expensive total battery changeout. Monitoring individual batteries as the BTECH does allows you to correct the bad battery before it causes the whole battery plant to fail.

Below are samples of battery analysis via the BTECH system. A history of every battery in a plant will be recorded and analyzed automatically through the automated system software.  

Batteries often go from good to bad over a short few week period. This shows the monitoring software picking out a battery, #15 in this case, weakening and failing over this period.
 
During an extended outage (>15 seconds) individual battery voltages are recorded and displayed to show relative strength of batteries in a battery plant. This shows a battery, # 14, which has dropped to under 9 volts during this total 44 second outage. This proves the battery needs replacement. Failure to do so will weaken the string and cause a load loss outage during a future outage.
 
The BTECH software for battery #14 above had shown this battery to be bad months earlier and the battery should have been changed but was not. Voltage of the battery had initially dipped when the impedance showed it was bad but a few weeks later the voltage came back but obviously the battery had not. This shows why impedance monitoring is a better method than voltage monitoring.
Battery #14 Impedance Graph
Battery #14 Voltage Graph
 


   

Site Update 03/26/08
  904-880-0118 * www.wilsonengineeered.com