Practical Battery Testing

 

Practical battery testing

 

The Battery testing matrix below may help guide even the most skilled battery testing technician and will help simplify the recommended practices. The following is a description of some of the tests or maintenance parameters.

 

Capacity test

Capacity test is the only way to get an accurate value on the actual capacity of the battery. While used regularly it can be used for tracking the battery’s health and actual capacity and estimating remaining life of the battery. When the battery is new its capacity might be slightly lower than specified. This is normal.

There are rated capacity values available from the manufacturer. All batteries have tables telling the discharge current for a specified time and down to a specific end of discharge voltage. Table below is an example from a battery manufacturer.

End Volt. /Cell	Model	8 h Ah Ratings	Nominal rates at 25º C (77º F) Amperes (includes connector voltage drop
			1h	2 h	3h	4h	5h	6h	8h	10h
1.75	DCU/DU-9	100	52	34	26	21	18	15	12	10
	DCU/DU-11	120	66	41	30	25	21	18	15	13
	DCU/DU-13	150	78	50	38	31	27	23	19	16

Common test times are 5 or 8 hours and common end of discharge voltage for a lead acid cell is 1.75 or 1.80 V.

During the test it is measured how much capacity (current x time expressed in Ah) the battery can deliver before the terminal voltage drops to the end of discharge voltage x number of cells. The current shall be maintained at a constant value. It is recommended to select a test time that is approximately the same as the battery’s duty cycle. Common test times are 5 or 8 hours and common end of discharge voltage for a lead acid cell is 1.75 or 1.80 V. It is recommended to use the same testing time during the battery’s lifetime. This will improve accuracy when trending how battery’s capacity changes.

If the battery reaches the end of discharge voltage at the same time as the specified test time the batteries actual capacity is 100% of the rated capacity. If it reaches the end of discharge at 80% (8 h) or before of the specified 10 h it is shall be replaced. See figure 3.

Procedure for capacity test of vented lead acid battery

 

1.      Verify that the battery has had an equalizing charge if specified by the manufacturer

2.      Check all battery connections and ensure all resistance readings are correct

3.      Record specific gravity of every cell

4.      Record the float voltage of every cell

5.      Record the temperature of every sixth cell in order to get an average temperature

6.      Record the battery terminal float voltage

7.      Disconnect the charger from the battery

8.      Start the discharge. The discharge current should be corrected for the temperature obtained at point 5 (not if capacity is corrected afterwards) and maintained during the entire test.

9.      Record the voltage of every cell and the battery terminal voltage in the beginning of the discharge test

10.  Record the voltage of every cell and the battery terminal voltage one or several times at specified intervals when the test is running

11.  Maintain the discharge until the battery terminal voltage has decreased to the specified end of discharge voltage (for instance 1.75 x number of cells)

12.  Record the voltage of every cell and the battery terminal voltage at the end of the test. The cell voltages at the end of the test have special importance since weak cells are indicated here.

13.  Calculate the actual battery capacity

Figure 3 if the battery reaches the end of discharge at 80%

(8 h) or before of the specified 10 h it is shall be replaced.

Figure 4 Replacement of battery is recommended when the capacity is 80% of rated.

It is important to measure the individual cell voltages. This has to be made a couple of times during the test. Most important is to measure the cells at the end of the discharge test in order to find the weak cells.

It is also very important that the time OR the current during a discharge test is adjusted for the temperature of the battery. A cold battery will give less Ah than a warm. Temperature correction factors and methods are described in the IEEE standards.

Manufacturers can also specify their batteries at constant power discharge. This is used where the load has voltage regulators. Then the current will increase when the voltage drops. Procedure for testing these batteries is the same but the load equipment must be able to discharge with a constant power.

Batteries can also be tested at a shorter time than their duty cycle, for instance at 1 hour. Then the current rate has to be increased. Advantage is that less capacity is drained from the battery (valid for lead-acid) and it requires less time to recharge it. Also less man-hour is needed for the test. Contact your battery manufacturer for more information. At higher rates it is more important to supervise the battery’s temperature.

Between load tests, impedance measurement is an excellent tool for assessing the condition of batteries. Furthermore, it is recommended that an impedance test be performed just prior to any load test to improve the correlation between capacity and impedance.

 

Impedance test

Impedance, an internal ohmic test, is resistance in AC terms. With regard to DC battery systems, impedance indicates the condition of batteries. Since it tests the condition of the entire electrical path of a battery from terminal plate to terminal plate, impedance can find weaknesses in cells and intercell connectors easily and reliably.

Basically, impedance test is determined by applying an AC current signal, measuring the AC voltage drop across the cell or intercell connector and calculating the impedance using Ohm’s Law. In practice, not only is the AC voltage drop measured but so is the AC current. The AC current is measured because of other AC currents in a battery that are additive (subtractive). Other AC currents are present from the charger system. The test is performed by applying an AC test signal to the terminal plates. Then measure both the total AC current in the string and the voltage drop of each unit in the string by measuring each cell and intercell connector consecutively until the entire string is measured. The impedance is calculated, displayed and stored. As the cells age, the internal impedance increases as depicted in figure 2. By measuring impedance, the condition of each cell in the string can be measured and trended to determine when to replace a cell or the string which helps in planning for budgetary needs.

The impedance test is a true four-wire, Kelvin-type measurement that provides excellent reliability and highly reproducible data on which to base sound decisions with regard to battery maintenance and replacement. Impedance is able to find weak cells so that proactive maintenance can be performed. After all, the battery is a cost but it is supporting a critical load or revenue stream. If a single cell goes open then the entire string goes off line and the load is no longer supported. Therefore, it is important to find the weak cells before they cause a major failure.

The graph in figure 5 shows the effect of decreasing capacity on impedance. There is a strong correlation between impedance and capacity so that weak cells are ably and reliably found in sufficient time to take remedial action. The graph shows the reorganized impedance data in ascending order with each cell’s corresponding load test end voltage. (Impedance in milliohms coincidentally is the same scale as the voltage, 0 to 2.5). This view, that is ascending impedance/ descending voltage, groups the weak cells on the right side of the graph to find them easily.

 

Impedance theory

A battery is not simply resistive. There is also a capacitive term. After all, a battery is a capacitor, a storage device, and resistors cannot store electricity. Figure 6 shows an electrical circuit, known as the Randles Equivalent Circuit that depicts a battery in simple terms. There are those who would have people believe that the capacitive term is not necessary and that the resistance is the only part that needs measuring.

Impedance measures both the DC resistance (the real component in impedance) and the reactance (the imaginary components in impedance). Only by measuring both can the capacitive term start to be understood. The other argument used against impedance is that frequency is a variable in the reactance part of the impedance equation. That is true except that since Megger uses a fixed frequency, namely 50 or 60 Hz depending upon geography, it is always the same. This variable, 2πω, now becomes a constant and, therefore, frequency does not affect the final result in any way. The only parts that affect the final result are the parts that vary within the battery, namely resistance and capacitance, which paint the whole capacity/condition picture.

 

Figure 5 Ascending impedance with corresponding end voltage

In the diagram shown in figure 6, Rm is the metallic resistance, Re is the electrolyte resistance, Rct is the charge transfer resistance, WI is the Warburg impedance and Cdl is the capacitance of the double layer. Rm includes all of the metallic components one post to the other post, i.e., post, top lead and grids and to a certain degree, the paste. Re is the resistance of the electrolyte which doesn’t vary that much on a bulk basis. But at the microscopic level in the pores of the paste, it can be significant. Rct is the resistance of the exchange of ions from the acid to the paste. If the paste is sulphated, then Rct increases or if that portion of the paste is not mechanically (electrically) attached to the grid so that electrons cannot flow out of the cell. Warburg impedance is essentially insignificant and is a function of the specific gravity. Cdl is what probably makes the most important contribution to battery capacity. By only measuring DC resistance, capacitance, an important part of the cell, is ignored. Impedance measures both DC resistance and capacitance.

A battery is complex and has more than one electrochemical process occurring at any given time, e.g., ion diffusion, charge transfer, etc. The capacity decreases during a discharge due to the conversion of active material and depletion of the acid. Also, as the plate’s sulphate, the resistance of the charge transfer increases since the sulphate is less conductive than the active material. (See discussion about the differences between the thicknesses of the plates in long duration versus short-duration batteries.)

 

Intercell connection resistance

 

Intercell connection resistance is the other half of the battery. A battery is comprised of cells connected in a series path. If any one component fails the entire series connection fails. Many times batteries fail, not because of weak cells, but due to weak intercell connections, especially on lead posts which can cold-flow. Generally, hardware should be tightened to the low end of the torque scale that is recommended by the battery manufacturer. But torque wrenches are a mechanical means to verify low electrical resistance. It is far better to actually perform an electrical test using an appropriate instrument. It is a low electrical resistance that is desired. This test should be performed before the battery is commissioned. Proper intercell connections are necessary to ensure that discharge rates can be met. The instrument of choice is a DLRO® or a MOM which can easily verify that all connections have been made properly. It can even find minor errors before the battery is commissioned, preventing possible causes of failure or damage to supported equipment.

Testing intercell connection resistance performs two functions:

·         Validates intercell connection resistance

·         Finds possible gross errors with top lead internal to the cell

 

By following IEEE Recommended Practices, intercell connection resistance can be validated. Those recommended practices specify that the variation of intercell connection resistance be less than ten percent. This translates into 7 micro-ohms on a 70-micro-ohm intercell connection resistance. This method can even find a washer stuck between the post and the intercell connector whereas torqueing will not. They also specify that ten percent of the intercell connectors be measured quarterly and all intercell connectors annually.

In multiple post batteries, it is possible to find those rare gross errors in a cell’s top lead. (See multiple post battery diagram in figure 1). On multiple-post cells, measure straight across both connections, then measure diagonally to check for balance in the cell and connections. Measuring only straight across does not adequately test for either intercell connection resistance or for gross top lead defects. This is due to the parallel circuits for the current.

The graph in figure 7 shows the data obtained from an actual 24-cell telephone (CO) battery the peak at connector #12 (cell 12 to 13) is an intertier cable connection. Connector #3 was out of specification and it was determined that one of the two bolts was not properly torqued. It was retorqued and retested. It came within ten percent of the string average after retorquing.

The negative plates (odd-numbered plates #1 through 15) are all connected through negative top lead which is connected to both negative posts. Positive plates (even-numbered) are connected to each other through positive top lead which is connected to both positive posts. There are two intercell connectors between neg post 1 and pos post 1 and between neg post 2 and pos post 2.

 

Figure 6 Randles equivalent circuit

Figure 7 Intercell connection resistance bar graph

The higher the current draw the more critical is the proper sizing of current-carrying components both internal to the cell and external. UPS batteries are usually designed for a high rate discharge lasting typically only 15-20 minutes. However, a telecommunications CO battery may have only a 500 Amp draw but can discharge for up to eight hours. So either combination can have disastrous effects due to improperly sized and maintained cells and intercell connectors.

 

Testing and electrical paths

In order to properly test a multiple post cell, one must understand its construction. Based on the diagram in figure 1, it can be seen that there are two parallel paths for the test current to travel. If the test leads are placed on neg post 1 and pos post 1, the two parallel paths are (1) directly from neg post 1 to pos post 1 through their intercell connectors and (2) neg post 1 down to the top lead, up to neg post 2 and across the intercell connectors to pos post 2 down to the pos top lead and back up to pos post 1.The two paths are parallel circuits and hence indistinguishable. If one bolt is loose, there isn’t any way to determine that since the test current will follow the path of least resistance. The better method to measure intercell connection resistance is to measure diagonally from neg post 1 to pos post 2 and again from neg post 2 to pos post 1. Compare the two readings for highest confidence. Admittedly, diagonal measurements are still parallel but the comparison becomes more interesting due to the increased influence of top lead and loose hardware. Diagonal measurements do not allow for a direct connection from post to post. In the case of six-post cells, measure diagonally across the farthest posts in both directions.

 

Voltage

Float voltage has traditionally been the mainstay of any testing procedure. What is voltage? Voltage is the difference, electrically speaking, between the lead and the lead oxide on the plates or between the nickel and the cadmium. The charger is the item that keeps them charged. The sum of all of the cell voltages must be equal to the charger setting (except for cable losses.) This implies then that voltage merely indicates the state-of-charge (SOC) of the cells. There is no indication of a cell’s state-of-health (SOH). A normal cell voltage doesn’t indicate anything except that the cell is fully charged. An abnormal cell voltage, however, does tell you something about the condition of the cell. A low cell voltage can indicate a shorted cell but only when the voltage finally drops to about 2.03. If a cell is low then other cells must be higher in voltage due to the charger setting. Remember that the sum of all cell voltages must equal the charger setting. Those cells that are higher are counteracting the low cell and generally speaking the higher cells are in better condition because they can tolerate the higher voltage. But those cells are being overcharged which over-heats them and accelerates grid corrosion and water losses.

Let’s say for the moment that the low voltage cell is not yet at 2.03, it is at 2.13 V. At 2.13 V it is not low enough to flag a concern but it is degrading. It may or may not be able to support the load when an outage occurs. Impedance is able to find that weak cell sooner than voltage. In this case, impedance will decrease since it is an impending short circuit.

A similar example can be found in VRLA when it comes to dry-out or loss-of-compression. Voltage will not find this condition until it is far later in the battery’s life, until it is too late. Impedance finds this condition much earlier so that remedial action can be performed.

So don’t confuse fully charged with full capacity.

As said above, cell voltage divergence can be caused by a number of factors and one way to solve this problem could be to make an equalization charge. In an equalization charge procedure, the entire battery is charged at a higher (than normal) voltage for several hours to balance the voltage in all the cells. The procedure can lead to heating and possibly water loss. It is recommended to follow the manufacturer’s procedure to avoid damaging the battery.

 

Specific gravity

Specific gravity is the measure of the sulphate in the acid of a lead-acid battery. It is also the measure of the potassium hydroxide electrolyte in nickel-cadmium battery but since the potassium hydroxide electrolyte isn’t used in the chemical reaction, it is not necessary to measure it periodically.

Specific gravity traditionally has not provided much value in determining impending battery failure. In fact, it changes very little after the initial 3 to 6 months of a battery’s life.

This initial change is due to the completion of the formation process, which converts inactive paste material into active material by reacting with the sulphuric acid. A low specific gravity may mean that the charger voltage is set too low causing plate sulphation to occur.

In a lead-acid battery the sulphate is a closed system in that the sulphate must be either on the plates or in the acid. If the battery is fully charged then the sulphate must be in the acid. If the battery is discharged, the sulphate is on the plates. The end result is that specific gravity is a mirror image of voltage and thus state-of-charge. Specific gravity readings should be taken when things are amiss in the battery to obtain as much information about the battery as possible.

Different battery applications and geographies have varying specific gravities to accommodate rates, temperature, etc. Following is a table that describes some applications and their specific gravities.

 

 

Float current

Another leg of the Ohm’s Law triangle is current. The charger voltage is used to keep a battery charged but voltage is really the vehicle to get current into the battery (or out of it during discharge). It is current that converts the lead sulphate back to active material on the grids.

There are two types of DC current on a battery: recharge current which is the current applied to recharge a battery after a discharge and float current which is the current used to maintain a battery in a fully charged state. If there is a difference between the charger setting and the battery’s voltage, that difference will cause a current to flow. When the battery is fully charged [1], the only current flowing is the float current which counteracts the self-discharge of the battery (<1% per week). Since the voltage differential between the charger and the battery is small, the float current is small. When there is a large voltage difference such as after a discharge the current is high and is limited by the charger until the voltage difference becomes less. When the current is on the plateau in the graph below, this is called current limit. When the voltage differential becomes less, the charge current is reduced as depicted on the downward sloping charge current line on the graph shown in figure 8. The charge voltage is the voltage of the battery, not the charger setting which is why it is increasing.

Float current will vary with battery size. The larger the battery is, the more float current it will take to keep it fully charged. Float current can increase for a couple of reasons: ground faults on floating battery systems and internal battery faults. Ground faults are discussed later. As a battery’s internal impedance increases, it takes more current to pass through that higher impedance. The increase in float current can be an indicator of battery faults. In lieu of measuring float current, many of the same conditions are found with impedance.

In VRLA batteries, float current [2, 3] seems to be an indicator of battery problems, namely thermal runaway. Thermal runaway is the result of a battery problem, not the cause. Some of the causes that can lead to thermal runaway are shorted cells, ground faults, dry-out, excessive charging and insufficient heat removal. This process takes anywhere from two weeks to four months to occur once the float current starts its increase. By measuring float current, it may be possible to avoid catastrophic failure of the battery and damage to connect and nearby equipment. Impedance will find many of these same errors.

Figure 8 Constant-voltage Constant-current charge characteristics

 

 

 

 

Ripple current

Batteries, as DC devices, prefer to have only DC imposed on them. The charger’s job is to convert AC into DC but no charger is 100% efficient. Frequently, filters are added to chargers to remove the AC current from the DC output. The AC current on the DC is called ripple current. Battery manufacturers have stated that more than about 5 A rms of ripple for every 100 Ah of battery capacity can lead to premature failure due to internal heating. Ripple voltage is not a concern since it is the heating effect of the ripple current that damages batteries. The 5% ripple current figure is a rough estimate and depends also on the ambient temperature. Ripple current can increase slowly as the electronic components in the charger age. Also if a diode goes bad, the ripple current can increase more dramatically leading to heating and premature death without anyone knowing it. Although impedance is not a measure of ripple current, ripple current is measured because of the way Megger designs its impedance instruments.

There is anecdotal evidence [4] that low frequency ripple (<10Hz) may charge and discharge a battery on a micro scale. More research is necessary to prove this hypothesis. Excessive cycling can lead to premature death of a battery regardless of the reasons for the cycling, be they outages, testing or maybe micro-cycling. One thing is true: the lower the AC is on the battery system, the less the damage is that can occur. VRLA batteries seem to be more sensitive to ripple current than their flooded counterparts. It is then advisable to filter their chargers for ripple current/voltage.

 

Temperature

It is well known that low temperatures slow up the internal chemical reactions in any battery; the degrees of reduced performance vary according to the technology. For example, at temperatures around freezing, a VRLA may need capacity compensation of 20%. The lead-calcium cell using 1.215 specific gravity acid will require a doubling of capacity, while the Ni-Cd will need about 18% increased capacity.

At the other end of the temperature range, high temperature is the killer of all batteries. There will be no surprise to find out that this impact varies from one technology to another. Lead acid at 95˚F will experience a 50% shortened life, while Ni-Cd will have a 16-18% shortening of life.

By applying what Arrhenius learned about chemical reactions, for every 18º F (10º C) increase in battery temperature, battery life is halved, battery life can start to be managed. The increased temperature causes faster positive grid corrosion as well as other failure modes. By holding a lead-acid battery at a temperature of 95º F (35º C) instead of the designed 77º F (25º C), a 20-year battery will last only ten years, a ten-year battery only five years and so on. Increase the temperature by another 18º F to 113º F (45º C), a 20-year battery will last only five years!

A battery is rarely held at a certain temperature for its entire life. A more realistic scenario is for a battery to heat during the day and cool down at night with higher average temperatures in the summer and lower average temperatures in winter. It is unfortunate but cooling the battery off to below 77º F (25º C) will not gain back the life that was lost. Once the positive grid corrodes, it cannot be converted back again. Furthermore, positive grid corrosion occurs at all temperatures, it is merely a matter of speed of the corrosion rate. The end result is to control, as best as possible (back to cost versus risk), the temperature of the batteries in the network.

IEEE 450, Annex H offers a method for calculating the impact of high temperatures on a lead acid battery.

Data analysis

The essence of any testing methodology is how to interpret the data to make some sense of it all. The same is true of battery testing. If the data are to be hand-written and filed or if a printout from an instrument is reviewed then filed, then there is no useful analysis except if there is an emergency at that very moment. The real value in battery testing lies in the trending of data to determine if problems are imminent or a little farther out. Trending of battery data, especially impedance and capacity, is an excellent tool for budgetary planning. By watching the batteries degrade over time, a decision can be made as to when to replace a battery. With trending, emergency replacements decrease dramatically.

The first time a battery’s impedance is tested can cause concern because there is no baseline. In these cases, it is good to compare each cell against every other cell in the string. Weak cells stand out. It is these cells which require further investigation. The table below provides a guideline depending upon the length of time batteries have been tested.