A battery is an electric storage device that converts chemical energy to electrical energy. Batteries are comprised of one or more cells or jars, wired in series or in parallel. Stationary batteries are typically designed to deliver power for a specific load requirement over a specified period of time.
• Flooded or Wet Cells The cell plates (commonly a lead alloy) are suspended in a bath of liquid electrolyte (typically sulfuric acid).
• Gel Cells The liquid electrolyte is replaced with a thick gel electrolyte.
• AGM (Absorbed Glass Mat) Cells The space between plates is filled with a fiberglass, mat-like material that holds liquid electrolyte. In the early days of telephony, wet cell batteries were used exclusively. Typically two-volt cells or 12-volt mono-blocs, these jars were strung in series to form a battery. This (wet cell) type of battery is still used today and has a long life expectancy (20 years or more) with proper maintenance. Disadvantages include the size of the cells (very large and heavy), the safety challenges they present and the equipment required to alleviate hazards, such as special flooring, bags of inert material placed around the battery banks and special maintenance techniques like specific gravity testing and adding distilled water, as necessary.
Today many CATV systems, wireless carriers and major telecom providers use either gel cell or AGM batteries. Advantages are that they are much smaller in size, cost less, have reduced hazard considerations and require no specific gravity testing or watering maintenance. The main disadvantage of gel cell or AGM batteries is that they do not last as long as wet cell batteries, and therefore, require careful periodic testing and monitoring.
A cell is a collection of plates connected to a single positive and single negative terminal, immersed in electrolyte, and contained in a single jar. A single lead acid cell typically has an open circuit voltage of 2.1 Vdc.
A jar is comprised of one or more cells, connected in series, and packaged inside a single container (typically polypropylene).
As batteries age, they lose their ability to deliver power. According to the “IEEE Std 450 2002” document, when a battery has lost 20% of its rated capacity, it is no longer viable and should be replaced. Battery capacity is typically measured in Amp-hours which is a battery’s rated ability to deliver a specific amount of power, using a given load, for a specific period of time. Figure one The figure above shows an example of a very simple lead acid battery model, which can be thought of as a set of series resistances in parallel with a capacitor. R1 represents resistance of metals and electrolyte, R2 represents the “Charge transfer resistance” (the ability of cell to accept a charge, which reduces as the acid or electrolyte becomes saturated near the electrode/electrolyte interface region) and C represents the total battery capacitance. In a lead acid battery R2 represents only about 40% of the DC resistance. Therefore, most of the battery’s impedance will be determined by metallic resistance, and capacitive reactance variations, though charge transfer resistance will also play a part.
Batteries are designed life expectancies in specified number of years. The actual life span of a lead acid battery is greatly dependent upon the type of battery, its operating environment (including battery temperature, charging voltage, the number of discharge/charge cycles) and mechanical performance (deterioration of internal connections and proper valve operation). In our discussions with operators in the telecom, wireless and CATV industries, the general consensus is that these batteries do not always meet the designed life expectancy. Common reasons for failure to meet design life expectancy are:
• Improper float charging (which can cause excessive gassing and eventual dry out)
• Excessive ambient temperature (IEEE 450 2002 estimates 50% loss-of-life expectancy for an eight-degree C temperature rise above 25 degrees C.)
• Deterioration of straps, grids and post connections due to corrosion, sulphation or mechanical failure
• Thermal runaway
Today’s lead acid batteries need to be “trickle” charged continuously at a very specific voltage to maintain their proper state of charge. It is very important to verify that all cells or jars in a battery are being charged properly at the proper voltage (The proper charge voltage window for a cell is only several hundredths of a volt) to achieve the expected life span.
Overcharging or undercharging will shorten the battery’s useful life. A battery’s ability to deliver power decreases with age. Testing batteries at regular intervals of every three months or less, with Ohmic measurement techniques have proven useful, as the battery’s capacity to deliver power tracks generally with its internal resistance. Manual PM techniques (hand held meter) are partially effective, at best, because of inconsistencies created by using different meters, inconsistent probe placement, and how forcefully probes are placed.
Temperature is very important, since the temperature a battery is subject to can severely impact its life expectancy. (A lead acid battery stored at 93 degrees F has about half the life expectancy of a battery stored at 77 degrees F.) To anticipate how long a battery might be expected to last, knowing the temperature of its operating environment is critical. Also, thermal runaway is a dangerous failure mode and can be detected by monitoring charge current and battery temperature. A visual inspection is also necessary periodically to make sure that the battery has no physical problems, such as cracks in the case or bulging.
Typically, a technician is dispatched to the site and manually takes, at a minimum, voltage, conductance/admittance and temperature measurements for each cell or jar. The technician determines whether any of the batteries have failed or are close to failing, as well as performs a visual inspection. The technician’s results are then recorded and later transferred to an excel spreadsheet or similar program for further analysis.
Operators who perform manual preventative maintenance would like to adhere to recommended practices listed in the IEEE 450 for flooded cells and IEEE 1188 documents. For VRLA cells, however, some of the tests which are now performed quarterly are, in fact, recommended to be performed monthly by the IEEE. Intervals of “quarterly” testing can actually be closer to five months in length (if testing is done early in one quarter and late in the next). According to actual field experience and discussion with a number of operators, batteries sometimes fail between these manual testing periods. This testing methodology is, at best, only partially effective, especially when budget considerations also affect the regularity of preventative maintenance testing. After these manual tests are performed, the information is gathered and manually transferred to log files, later saved in spreadsheets or other software programs. The problems associated with manual testing are:
• It is primitive in nature. Ohmic measurement results vary by technician, meter type, placement of test probes, how forcefully test probes are placed on battery post terminals, etc., all of which make it impossible to trace a meaningful trend line.
• It does not provide real-time battery status.
• Its information is not readily available to all appropriate personnel.
• It provides little or no compiled, historical information.
• Does not provide enough statistical data to show accurate trends.
• Operators will likely miss impending failures.
Impedance testing is a method that forces a small AC test signal (usually under an amp) into the battery’s terminal posts and measures the small terminal post AC voltage component caused by that current. Impedance testing is also governed by ohms law, except that the voltages and currents are AC voltages and AC currents. Using a very small AC test results in very small AC signal voltages, but small AC signal voltages are easily separated from the cell’s large DC component by use of AC coupled voltage amplifiers. An example of an AC impedance test setup is shown below: AC Impedance Test In this example, if we assume the same cell properties referenced earlier (ignoring the capacitive component), then the internal resistance of the cell is R = R1+R2 = .001 ohm. If the test signal generator develops a current Iac = 1 amp through the battery, then the terminal post voltage (Vcell) will also have an AC component (Vac) superimposed upon it.
According to Ohm’s Law: E = I x R so, Vac = Iac x R Vac = 1.0 x .001 = .001 VAC This resultant AC voltage on the battery terminals is clearly very small and would be very difficult for a normal measurement instrument to resolve in the presence of the much larger Vcell voltage of 2.2V. We can solve that problem, however, by using an AC-coupled voltage amplifier, which does not respond to the very large Vcell DC voltage. If we assume that the AC amplifier has a voltage gain of 1000, then the signal at the amplifier output would be 1VAC, which is easily measured. A simple arithmetic manipulation can be used to relate the measured AC voltage at the amplifier output to the internal resistance of the battery.
A battery is a DC power source made up of one or more electro-chemically based voltage generating “cells”. Each cell has internal resistances that limit the amount of current the cell can supply to the load. In addition to a cell’s pure DC properties, it also exhibits characteristics of a large capacitor. The effect of this capacitance on battery testing methods will be discussed later. Figure one If the equivalent cell circuits above were analyzed by drawing a short-circuit DC current from its terminals, the capacitive component would not come into play because the currents are pure DC, and the cell would deliver a maximum current determined by Ohm’s law:
I = E/R where I is the maximum current the cell can supply, E = Vcell and is a voltage determined by the internal electro-chemical construction of the cell, and R=R1+R2 and is the combination of all resistances inside the cell. As an example, an ideal lead-acid cell with a chemistry-determined cell voltage (Vcell) of 2.2V and an internal resistance (R1 + R2) of 0.001 ohms would deliver a maximum instantaneous short-circuit current (Iss) of: Iss = Vcell/(R1+R2) Iss = 2.2/.001 = 2200 amps The terms “maximum” and “instantaneous” are used here because: from the moment the load current begins to flow, the cell begins to lose charge, causing the internal resistances to increase, which causes the voltage at the terminals to drop, causing the current to decrease. A traditional test to determine the “amp-hour” capacity of a cell is called “resistive load testing” and consists of connecting a moderate to heavy resistive load of known value to the cell and determining how long it takes for the cell’s terminal voltage to drop by a specified amount. This type testing, though informative and effective, is also time-consuming and destructive. If the cell is called upon to deliver power shortly after this type of discharge test is performed, it might not have sufficient charge remaining to deliver the expected run-time. Because of the limitations and destructive nature of resistive DC load testing, an advanced and minimally intrusive alternative called “impedance testing” has been developed and proven effective.
Load tests are very good for determining battery capacity, but are only done every 2-3 years because they are intrusive and expensive. It is also problematic in that:
• This type of testing is time consuming and labor intensive, which means expensive.
• The test provides a snap shot in time, however in between load tests (two or more years) the exact battery health is unknown.
• The battery plant must be taken offline and connected to a precise resistive load bank, and the battery is discharged over a number of hours to cut off voltage, which takes the battery backup out of service for most of the day
• The test is intrusive. Lead acid batteries have a finite number of cycles (charge and discharge) usually numbering in the hundreds. The discharge and recharge as is necessary in load testing will slightly reduce the battery’s usable life.
• Once testing is complete, the battery plant must recharge. To fully recharge will take many hours during which back up is not available.
It does not cause errors if the testing is designed correctly. The above AC measurement example would be a precise indicator of the battery’s internal cell resistance except for the matter of the cell capacitance and its affect on the AC measurements. The cell capacitance is a function of the amp-hour capacity of the cell and is usually estimated to be about 1.5 Farad for each 100 amp-hours of cell capacity. At high AC test frequencies, the AC reactance of this capacitance could be small enough so that it would ‘mask’ the value of the R2 component of the battery resistance, by effectively shorting it out. For this reason, impedance measurements are commonly performed at very low frequencies, where the capacitive reactance is not significant compared to the resistance that it shunts (R2). Frequencies well below 50 Hz are typically used.
Yes, an ideal battery would have infinite ability to deliver power. However, a real battery’s ability to deliver power is limited by non-ideal internal elements (i.e., metal straps, plate and post resistances, electrolytes, the resistance between the electrolyte and more). AC impedance/conductance testing is considered non-intrusive and can uncover battery irregularities. Though impedance/conductance testing has some inherent measurement inaccuracies (including manufacturer’s stated reference value, and meter measurement repeatability) this method has been proven to be a reliable barometer. Any accuracy errors can be overcome easily when trended over many data points and time. However, these inaccuracies can be misleading if only a few measurements per year are performed, like quarterly manual testing permits. As batteries age, their internal resistance will increase and further limit the battery’s ability to deliver power. In any case, the value of the cell’s internal capacitance is also correlated to amp-hour capacity, so tracking the combined cell’s relative impedance (which is the combined effect of resistance and capacitance) over time is very useful in determining changes in the cell’s state of health, even if the absolute impedance measurements are not 100% reflective of actual battery capacity.
Remote monitoring systems typically have some type of sensor that connects to each cell or battery and take accurate, consistent and continuous measurements for voltage, battery post temperature, admittance and charge current. This information is relayed to a controller, which typically provides standards-based interface to the outside world via an IP network. Many modern controllers use SNMP. Standards-based software programs interrogate the controller, and warehouse, analyze and provide executive-level reporting, as well as trending data and graphs to anyone with proper access to the software.
• Greatly reduced maintenance costs and better battery information
• Real-time visibility of DC power plant and UPS batteries
• Consistent and continuous measurement data, which produces meaningful and accurate trend line information
• Ability to look at every battery in the entire enterprise and its history from a single location
• Executive-level reporting compared to manual excel spread sheets
• Much more effective means of managing resources in crisis situations
• Ability to predict battery failures, as opposed to reacting to discovery of a dead battery
• Automated alarm notification
• Standards-based hardware and software
Technology advances have dramatically reduced costs for implementing remote battery monitoring. In addition to reducing manual maintenance visits and extended battery life, the ROI has shown to be well under two years. Operators will also be provided with much more reliable data, asset tracking management, complete battery history and the ability track warranty issues, and remote monitoring will virtually eliminate the need for collating and analyzing manually gathered data. Monitoring systems will identify bad batteries well in advance of failure, allowing operators to schedule replacement instead of reacting to a crises event. Identifying and removing a bad battery in the string will allow operators to significantly extend the life of other batteries in the string, instead of unnecessarily replacing all of the batteries in a string vs. just one.
Some can. Intelligently balancing the charge voltage so that each two-volt cell in a string is within 10mv or each 12-volt bloc is within 60mv has been proven to extend the life of batteries. Balancing charge prevents overcharging good cells when a weak cell is in the string and provides extra charge to weak cells in the sting so that they get the charge they need. For example, when a cell or jar in a string is replaced with a new cell from the warehouse, that new cell could have been sitting in the warehouse for months. If that cell was not being constantly charged, which many are not, then it has possibly lost a significant amount charge. When that weak cell is placed in a string of fully-charged cells, the fully charged cells can prevent the weak cell from getting the proper charge. By individually balancing each cell, weak cells will get the charge they require and all cells in the string will be equally charged.