Although VRLA batteries have dominated the Industrial Battery Market since they became commercially available in the mid 1980’s the lead-acid Planté battery still has an important part to play. It has been argued that the Planté battery is the most reliable emergency power supply available. Whether this is true or not is a matter of opinion. What is true, is that there are many Planté batteries still in service that were installed over 20 years ago and some are still in reliable operation over 30 years of age. Planté batteries are still manufactured by a small number of companies and remain popular in many industries where reliability is paramount.

Because the voltage of VRLA batteries is not as stable as Planté batteries and specific gravity measurements cannot be taken, instruments that measure the ohmic resistance of batteries has gained popularity in trying to determine their condition. Ohmic measuring instruments were first introduced for use on VRLA monoblocs of 6V and 12V up to about 400Ah and not for 2V 1000Ah Planté batteries where the accuracy is debatable. The accuracy of these instruments remains the subject of much discussion even for VRLA batteries. In some respects it is unfortunate that these ohmic measuring instruments are sometimes used to determine the state of health of Planté batteries because they were not designed to work on large capacity cells. There is little doubt that some readers of this article will disagree arguing that they have an instrument that will work on any battery including vented NiCd. This argument is not the subject of this article but is intended to give a more practical understanding of measurements that can be readily taken with a high degree of reliability on Planté cells.

Specific gravity values published by manufacturers are not definitive and are difficult to relate to actual in service values. This is because a lead-acid cells an electro-chemical device, which is subject to manufacturing process variability and consequently the final product also has some variability. Additionally, the specific gravity does not change instantly with state of charge and therefore the values obtained need to be carefully analyzed to ensure they are meaningful.

Lead-acid batteries do not follow conventional electrical principals, nor do they follow conventional chemical reactions and as such they need to be considered as a variable product. In conclusion, while float voltage, ohmic measurements and specific gravity values are not finite; we can use these values to good effect in establishing if a battery is in good condition or needs remedial action.

This article considers records that have been, or will be taken on batteries that have been on float charge for some time, i.e. as they will have been in normal service.

For Planté batteries, manufacturers publish typical expected voltage variations for the average float voltage. Most manufacturers quote +/- 0.025Vpc. This value is typically derived from laboratory evaluation where batteries have been on a stable charging system, with no load attached and at a constant temperature. In practice the +/- of 0.025Vpc is difficult to achieve for the following reasons: -

A) Chargers vary in design and age and ability to control the voltage at different loads. The voltage may be very stable if there is no standing load but at a high load the variability may be considerable which will affect the stability of the cell voltages. With variable loads, the cell variation is exacerbated.

B) The operating temperature may vary and this will change the charge acceptance characteristics of the battery and the variability in float voltage will be affected. Variations in cell temperature across the battery can easily be 5°C and this will affect the individual cell float voltage.

C) Charger ripple current will affect the cell voltages.

Looking at the practical values of float voltage, a variability of +/- 0.05Vpc is generally acceptable. For an average float voltage of 2.25Vpc this results in a maximum of 2.30Vpc and minimum of 2.20Vpc. Anything outside these limits must be considered suspicious but it does not indicate a failed cell. If a cell voltage is at any time lower than the average by more than 0.05Vpc, i.e. below 2.20Vpc in this example, it should be investigated. The greater the deviation is and the more serious the consequences will be if no action is taken. Any individual cell with a voltage below 2.15Vpc must be investigated without delay. Low voltages are more of a concern than high voltages and some relaxation of the maximum voltage may be tolerated up to an absolute maximum of +0.08Vpc above the average. In this example anything between 2.20Vpc and 2.33Vpc (+0.08Vpc to –0.05Vpc) may be considered acceptable but a note on the record file should be made highlighting any cells that are at or close to the limit. This way, the next time the service is carried out it can be immediately seen that some cells were on the limit. If any cells are at the limit, or above, it would be prudent to measure again after approximately 6 months to verify the first set of readings. If, at the second set of readings, the voltage is seen to move further out of the range +0.08Vpc to –0.05Vpc, investigations must be carried out to establish the cause. If the readings were within +0.08Vpc to –0.05Vpc after 6 months, it would be reasonable to revisit at the annual service, i.e. after a second 6 months. If the voltages continue to be within the limits, it would be reasonable to continue with normal service checks and no action need be taken. However, if the cell voltage variability has increased, it would be prudent to investigate without delay.

Just because the voltage of any individual cell is outside the recommended limit does not mean that the cell needs replacing. In the first instance, investigations need to be made to establish why the voltage is outside the specification and if no good reason can be found and otherwise the cell appears to be healthy, a “boost charge” maybe considered.

A cell will only need to be replaced after detailed evaluation. Planté batteries are very reliable and rarely fail without warning. However, it is important to keep records for comparison over the lifetime of the product.

Below are examples for guidance that relate to the tabulated results further down the page.

Tabulated results headed Battery A) All cells within voltage limits, no action. Tabulated results headed Battery B) Repeat the measurements and if confirmed, investigate further because several cells are well outside the voltage limits. Tabulated results headed Battery C) On 25th June 2014 three cells high voltage and on 10th December 2014 (approx. 6 months later) all cells within voltage limits, no action required. Tabulated results headed Battery D) On 25th June 2015 one low voltage cell and on 20th December 2015 three cells out of voltage limit and one of the three has deteriorated. Investigate further.

Internal resistance values that are published by manufacturers are generally produced from D.C. discharge tests carried out on one or two cells from the same “family” and the results are extrapolated to produce values for each individual cell type. These tests are usually carried out in accordance with BS 6290 Part 1 where two discharges are carried out, and the results extrapolated to obtain a nominal value. The results give a D.C. resistance value in ohms. Hand held or fixed battery monitors usually give results of impedance or conductance and not a D.C. resistance. Impedance and conductance results do not easily correlate with the D.C. resistance value published by the manufacturer. Variations between ohmic measuring instruments occur because of the different methods used by the different instruments used, including applied frequency and amplitude to obtain the value.

The use of ohmic measuring instruments may be used to good effect for establishing the condition of Valve Regulated Lead-Acid (VRLA) cells where a visual inspection and specific gravity measurements cannot be made. The end of life of VRLA cells is generally due to positive grid corrosion and growth often accompanied with drying out. Planté cells have a different end of life failure mode that does not follow that of VRLA cells. Ohmic measurements may not be 100% reliable but can be used to compare different readings for cells in the same battery where an out of line and poor ohmic value may indicate a low performing cell. If a poor ohmic value is observed, even if the voltage is acceptable, investigations should be carried out to establish why. In the first instance, the measurements should be repeated to verify the results. If they are confirmed and all other parameters including float voltage and a visual inspection show no abnormalities further investigations should be made. A possible problem could be a high resistance inter-cell connection joint, which may not show up with float voltage measurements. It may also mean that the cell is in a poor state of health and needs to be replaced.

When considering % deviation from the base value, we need to consider where this base value came from. Remembering that the product is an electro-chemical device and subject to variability, we can expect this to be different not only from cell to cell but also from manufactured batch to batch. It is not unusual for the ohmic value to deviate by +/-7% from batch to batch. With this in mind a +/- deviation of 90% to 107% can be expected under ideal conditions. Many users of VRLA cells would consider a low conductance as 80% from the base number to be acceptable. Some consider 60% as the “danger point” for replacing cells. However, if one cell is showing a 20% deviation from the group average, this is sufficient to consider further investigations. Also, it should be understood that the actual method used to obtain a % from the base value necessitates that each reading is measured in exactly the same way with the probe attached to the same part of the cell pillar for all readings. For example, on the terminal pillar including the inter-cell connector. With this in mind, the 20% deviation, from the mean value is a reasonable starting point to consider further evaluation of the cell.

As in the case of voltage measurements, it is best to “trend” the results rather than consider a “snap shot” as a definitive answer to the battery condition. With VRLA cells the performance of the product deteriorates with time (age) and the product is normally replaced when this has fallen to 80% of the nominal. However, Planté cell performance is stable over a very long period and these cells are considered at the end of life if the performance is anything less than 100%. Ohmic measuring instruments cannot determine the difference between 99% and 100%; but neither can float voltage or specific gravity measurements. As with out of line float voltage, a “boost charge” may resolve the problem.

The following examples may be used as guidance.

Measurements of the cell specific gravity (s.g.) may be made with several different instruments but the most popular methods are by glass float hydrometer, or digital hydrometer. Glass float hydrometers are inexpensive but the results are open to interpretation by the operator and the results need to be temperature corrected, written into a log sheet for future evaluation and the instrument is easily broken. Digital hydrometers are more accurate and not, in the main part, influenced by the operator error and they are more robust. Some instruments have the ability to store the record value, which can then be down loaded for future evaluation, and the results are usually temperature compensated automatically.

There are several issues that must be considered when evaluation s.g. readings. They are subject to the following: -

A) Temperature variation. Providing the electrolyte temperature is within +/-10°C of the standard temperature, no adjustment to the reading is normally required; the results will be sufficiently accurate. This only applies to actual field results where an approximation is acceptable and not laboratory evaluation. If there is a need to temperature compensate the results, the correction formula is: - ((Reference temperature – Actual temperature) x 0.0007) + reference s.g.

Example:

i) Reference temperature: 20°C, ii) Actual temperature: 10°C, iii) Reference s.g.: 1.205 iv) Formula: ((20°C − 10°C) x 0.0007) + 1.205 = 1.212

This may be converted to give the corrected s.g. as: -

Corrected s.g. = Measured s.g. −((Reference Temperature – Actual Temperature) x 0.0007) Example for measure s.g. = 1.150. Corrected s.g. = 1.150 − ((20°C − 10°C) x 0.0007) = 1.143

B) We also need to know the following

i) The fully charges s.g. of the cell being evaluated. ii) The fully discharged s.g. of the cell being evaluated. iii) The fully charged s.g. at maximum electrolyte level of the cell being evaluated. iv) The fully discharged s.g. at minimum level of the cell being evaluated.

We need the above to calculate the state of charge at any electrolyte level.

In practical terms, the only time that s.g. measurements are truly meaningful is to establish the fully charged condition with the electrolyte at the maximum level such as during commission charging. Even then, it is more appropriate to consider the s.g. change over time rather than an absolute value. During the commissioning charge the s.g. will rise steadily until it is fully charged when the s.g. will remain at a constant value.

Although out of the scope of this article, it is worth noting that many batteries are incorrectly commission charged and the following give a typical reason for this. To complete a full and true commissioning charge, the Planté battery must be charged at a constant current of between 7% C10A to 10% C10A until the specific gravity plateau is reached. The voltage during the commissioning charge must not be limited and in practical terms, it will rise to a value in the order of 2.80Vpc. Frequently, a lower charging voltage is used and when the charger voltage limit is reached the current will fall away rapidly. When using a voltage of about 2.45Vpc to 2.55Vpc the battery will only be approximately 90% charged when this voltage limit is reached. The current will fall away rapidly and from this point, the state of charge will rise very slowly. Because the s.g. follows the state of charge, this will only rise very slowly and the cell may be interpreted as being fully charged when in fact it is not.

Some Planté battery manufacturers state that the battery can be commission charged wising a modified constant potential where the maximum voltage is in the region of 2.40Vpc to 2.50Vpc. While this is possible, it will take longer to complete the charge. This will be in the order of several days not a few hours that can be acceptable using constant current. Additionally, it is generally accepted that even after several days at 2.40Vpc to 2.50Vpc the battery will not be truly 100% charged. In practical terms this is not significant because all currently available Planté batteries are under stated. It is not unusual for a Planté battery to exceed 4h when tested at the 3h rate. It follows that even if the battery is only 90% charged, it is likely to pass any on site discharge capacity test. For guidance on commission charging at these lower “constant voltage” levels, always consult the battery manufacturer.

Below are examples of the specific gravity measured for the batteries in the examples used above for voltage and ohmic values.

Battery A) readings are all within a good range.

Battery B) readings show that although 9 cells are below the voltage limit, and 3 cells below the conductance limit, no cells are lower than the typical expected specific gravity. This could be for several reasons. One reason may be due to the battery being recently part discharged. This can disrupt the float voltage and ohmic value but because the true state of charge has not changed substantially, the s.g. value has not reduced. The correct procedure is to investigate. BATTERY C) 25th June 2014 records, show that no s.g. values are outside that expected. Three cells have high float voltages but as discussed above, this is not significant without further information. The ohmic conductance of cell No 23 is lower than expected but the voltage is acceptable and the s.g. is within expectations. It is reasonable to take no action but to check in 6 months time.

BATTERY C) 10th December 2014 records show some variation from 6 months earlier. However, the voltages and s.g. values are acceptable and it is only the ohmic conductance value of cell No 4 & No 6 that are marginally out of the expected limit. We know that ohmic conductance values can be misleading and because the volts and s.g. are acceptable, it is reasonable to take no action other than to ensure the battery is checked after a further 6 months.

BATTERY D) 25TH JUNE 2015 records show that although the s.g. of cell No is acceptable, it is lower than the typical. Together with the low float voltage and low ohmic conductance is sufficient to warrant further investigations. NOTE: no action was taken after this service visit.

BATTERY D) 20TH JUNE 2015 records show further voltage, conductance and s.g. deterioration of cell No 15. Cell No 26 also has low values of voltage, ohmic conductance and s.g. and further investigations need to be carried out to establish the reason.

Float voltage, ohmic measurements and specific gravity readings can all assist in establishing if a Planté cell is in good condition or not, particularly if trending is used. In addition, a visual inspection of the cell will further help to confirm the state of health. We must not forget that we have the opportunity to look inside Planté cells. Very often, the appearance can give more information than any instrument readings. However, if the person looking at the cell has little or no knowledge of what to look for then we have to rely on instrument reading. Instrument readings alone must not be taken in isolation; it is better to trend the results.

The use of instrument readings only will give us a reasonable assessment and confidence level but it will not be definitive. If the product is in question, some form of additional checks must be made and it may be appropriate to charge the complete battery at an elevated voltage or use a single cell charger to charge any cells that are suspect. Ultimately, it may be decided to replace the product.

6) GUIDANCE NOTES

A) Voltage Measurements

Float voltage measurements are the most accurate to use in detecting a low performance cell. In general terms, the float voltage will fall before a low ohmic value occurs and before a low s.g. is seen. However, a low voltage does not always confirm a low performing cell. Trending of voltages is the most accurate way to confirm the battery is in good condition. A cell float voltage more than +/- 0.10V from the average must be investigated without delay.

1. B) Ohmic Measurements

An ohmic variation of more than +/- 10% deviation from skewed the average, i.e. 90% to 110% should be considered for investigation. This average should exclude any cells that are outside the mathematical average by +/- 20%. Any cell having a value that is in the negative direction by more than 20% of the skewed average should be looked at more closely. By adopting this principal, exceptionally low or high cell readings are excluded from the average. For example, if the mathematical average is 1550mohs then the limit is 1395mohs to 1705mohs. If we have 5 cells below 1395mohs and 2 cells over 1705mohs these should be excluded from the mathematical average and a skewed average calculated made which may be lower than the true average. For example, the skewed average may be 1750mhos and any cells below 1575mohs should be investigated.

As with voltage, trending of the results is the best way to determine the condition. Manufacturers information must never be used as a pass / fail measure and should only be used as a guide. Instruments are often sensitive to probe position and the same position must be used for each cell measured and for subsequent readings.

C) Specific Gravity

Measurements of specific gravity can be misleading because of temperature differences, electrolyte level and cell-to-cell variability. They are good in confirming a bad cell but not cells which may be marginally low. They are also good for confirming a fully charged condition during commission charging. A value that is more than 0.030 s.g. lower than the average must be considered for further evaluation and in the first instance the float voltage should be measured and compared with other cells as described above in section 4) above.

D) Additional Comments for Guidance

Comparing measurements with each other is the best way to determine if a cell is suspect. A falling float voltage is typical of something going wrong. An ohmic value going in the negative direction can also indicate an anomaly taking place. If an s.g. value is more than 0.030 below that expected, it should be investigated. Ultimately, additional investigations need to be carried out to establish the condition of any suspect cell. Although the only way to establish if a cell or battery is in good condition and fit for the intended purpose is to carry out an actual discharge test, by sensible and intelligent use of instruments, a good confidence level can be established.

It is important not to blindly measure and record the results, instead think about what has been measured and if in doubt measure again to confirm the result.

Occasionally we are asked very interesting questions. Recently we were asked how much heat an industrial standby battery generates. It is fair to say, it depends on whom you ask. Different battery manufacturers have different answers to this question and the different method of calculation gives significantly different answers.

The heat emitted or generated is sometimes referred to as “heat loss”.

The author of this article makes no recommendations to the methods given below. The article is produced to show that there is conflict between the various methods used.

In general terms, the question is being asked to calculate ventilation requirements and this article explores different methods and demonstrates the variability of the results.

Heat is generated on recharge, float charge and discharge. The heat generated on charge is finite, i.e. once the battery is fully charged no more heat is generated but at this point the battery enters the float charge phase and as long as the battery is on charge, heat is being generated. Heat generated on discharge is also finite because once the battery is fully discharged no more heat will be generated. Therefore, we have three conditions to consider:

1) heat on recharge.

2) heat on float charge.

3) heat on discharge.

We all know that lead-acid batteries are heavy and have a large thermal mass. Because of this, during recharging, float charge and discharge, the heat generated within the cells will not dissipate to the surrounding atmosphere immediately and there is a difference of opinion on how quickly this will be. Part of the differing opinion is the result of different sizes and shapes of the cells or monoblocs making up the battery, and whether they are VRLA AGM, VRLA GEL or vented types.

In basic terms, heat is watts and watts can be calculated from V x I (volts x amperes) or we can use I2R (amperes x amperes x resistance). This principal these formulas may used to calculate the heat generated.

In this article the following battery system is used in the examples. The examples consider the following: -

a) A battery supplying 300kW for 15m at 20°C to not less than 408V (1.70Vpc average).

b) The battery consists of 3 parallel strings, each comprising 40 x 12V monoblocs; i.e. 240 cells.

c) Float voltage 2.27Vpc = 545V.

d) The nominal capacity of each string is 110Ah i.e. 330Ah total battery capacity.

e) Each monobloc internal resistance is 3.8mOhms. This value is from the battery manufacturer’s information. Therefore the battery resistance is 3.8mΩ x 40 blocs / 3 strings = 50.7mΩ total resistance.

f) Fully charged float current 1mA per Ah = 330mA. The value of 1mA per Ah is the I float. (note below) value from BS EN 50272.

g) The recharge parameters are 10% current (33A) and 2.27Vpc (544.8V) constant voltage.

(Note) - The fully charged float current may be obtained from the battery manufacturer. However, within BS EN 50272 (Safety Requirements for Secondary Batteries and Battery Installations) the typical value can be found in Table 1. The table gives the values for current when charging with IU or U chargers. While these values are used to calculate the gas emission on charge, they may also be used to estimate the current when fully charged. In practical terms, the values are worst-case scenarios with a safety margin built in.

For vented lead-acid batteries, VRLA lead acid batteries, and for NiCd batteries, the value is given as 1mA per Ah for float voltage conditions. We should consider the Ah as the nominal at the 10h rate for lead acid product and 5h rate for NiCd product.

First, we need to define “recharge” and in this context, we refer to the current / time required to return the capacity removed for the previous discharge. We are only considering the time to fully charged.

The amount of heat generated does not change appreciably even though the recharging parameters may be different. For example, the charger current i.e. 5% or 10% or 15% C10 amperes or using true float voltage (e.g. 2.27Vpc) or elevated voltage (e.g. 2.40Vpc), do not significantly change the heat generated or heat loss from the battery. However, the heat generated will be substantially different depending on the depth of previous discharge. For industrial standby batteries and in this article, we consider a constant voltage / limited current recharge characteristic; otherwise known as an IU or modified constant potential method such as 2.27vpc or 2.40Vpc, or similar, with a current limit.

It is worth noting at this stage that some battery manufacturers consider the heat generated on recharge may be calculated using the same method as if the battery was on float charge. This method is used in 1.1) below. This view is taken because any heat generated on recharge will not be released immediately because of the thermal mass of the battery.

The heat calculations are complicated when we take into consideration the specific heat characteristics of the battery and at least one battery manufacturer has produced results based on actual battery type and configuration. This does not help in establishing the heat generated for every battery configuration and we need something much simpler to use in an everyday situation. After all, we are looking at a typical value that may be used for room cooling purposes not a finite “laboratory evaluation”. In practical terms a good approximation is sufficiently accurate.

It follows that if the heat generated on recharge varies with the previous discharge then all other parameters are broadly irrelevant. We can then estimate the heat generated on recharge as a function of the previous discharge. To make the calculation a little more accurate we should estimate the time to fully charged based on the IU characteristics and previous depth of discharge. Most manufacturers have tables or even a software based method to determine the time to different states of charge including the time to fully charged. However, generally speaking it can be said that the time to fully charged will be many hours but the time to 80% charged will depend on the IU characteristic. During recharging, most of the heat will be generated as losses up to the battery reaching 80% charged which will be the “constant current” part of the recharge. During the constant current phase i.e. up to 80% charged, the heat may be estimated using the I2R principal. From 80% to 100%, the float current may be used to calculate the heat. Some battery manufacturers consider the current from 80% charged to 100% charged as twice the theoretical float current. In context with the actual heat, this may be considered as a reasonable method. This method is used in 1.2) below.

1.1) Considering the heat to be the same as if the battery is on float charge we have: -

V x I = W, or the alternatively method of I2R = W.

1.1.1) V x I = watts.

The only issue is deciding what voltage and what current to use.

For voltage, it is reasonable to consider the voltage as the actual float voltage across the battery terminals.

For current, it is reasonable to use the I float value as defined in BS EN 50272.

Calculate for 1 bloc: -

2.27Vpc x 6 cells x 110mA = 1,498.2mW

Therefore, for 40 x 3 blocs = 1,498.2 x 40 x 3 = 179,784mW = 179.784W.

This heat will be for the recharge time of 76h. Therefore the heat may be expressed as 180W x 76h = 13,680Wh but over 76h = 180W.

1.1.2) I2R = watts

We can use the same current as above, i.e. I float and for voltage R we can use the resistance of the bloc i.e. 3.8 mΩ. Calculate for 1 bloc: -

110mA x 110mA x 3.8mΩ. = 0.04598mW

Therefore, for 40 x 3 blocs = 5.5176mW.

This heat will be for the recharge time of 76h. Therefore the heat may be expressed as 5.5176mW x 76h = 0.42Wh but over the 76h recharge time = 5.5mW.

1.2) Heat to 80% Charged plus Heat from 80% to 100% charged

1.2.1) Heat to 80% Charged

Considering the battery system described above, we know that the recharge current will be 33A up to 80% charged and from 80% we will use the 2 x float current which is if we use the 2 x float current method, a current of 330 x 2 = 660mA. We need to establish the state of charge after the discharge. Assume the worst case of maximum current for 15m: -

The maximum current = 300kW x 1000 / 408V = 735A

The capacity removed = (735A x 15m) / 60 = 184Ah or 146Ah charged (330Ah – 184Ah).

This 184Ah represents 56% discharged or 44% charged.

We know that the recharge current of 33A (11A per string) will flow until the battery is 80% charged. The 80% charged condition is = 330Ah x 0.8 = 264Ah.

Time from 146Ah in the battery at the end of the previous discharge to 264Ah in the battery = 118Ah / 33A = 3.6h.

We can now estimate the heat from the commencement of recharging up to 80% charged as below.

Using I2R per bloc: -

11A x 11A x 3.8mΩ = 495.8mW.

Therefore for 40 x 3 blocs = 59,496mW

This current will flow for 3.6h, which can be expressed as 214Wh.

NOTE: The internal resistance of industrial batteries does not change significantly from 100% charged to 10% charged. Therefore, the I2R principal is valid.

1.2.2) Heat from 80% to 100% charged

We need to establish the time from 80% charged to fully charged and the battery manufacturer should supply this information. However, a reasonable assumption for the purposes of estimating the heat would be 72h. It is generally accepted that a fully discharged battery can be recharged using float current and between 5% and 15% recharge current within 72h. If we assume the full 72h, we are considering a worst-case scenario.

The heat per bloc can now be estimated to be: -

110mA x 110mA x 3.8mΩ. = 0.04598mW

Therefore, for 40 x 3 blocs = 5.5176mW.

This heat will be for the recharge time of 72h. Therefore the heat may be expressed as 5.5176mW x 72h = 0.40Wh and if we double this we get 0.79Wh.

Adding 1.2.1) with 1.2.2) we get 214Wh + 0.79Wh = 215Wh. This is over the full recharge time which equates to 215Wh / 76h = 2.83W

Most battery manufacturers consider the heat on float charge as a simple volts x current. V x I = W, i.e. volts x current = watts. Alternatively, the I2R principal may be used.

For current, we can contact the battery manufacturer or we could refer to International Standards such as BS EN 50272.

We can now make a calculation. Below is the calculation for the same battery considered in above, i.e. a battery comprising 40 x 12V monoblocs of 330Ah. Two alternative calculations can be made. In 2.1) we use the V X I method and in 2.2) we use the I2R method.

2.1) Considering the V x I method: -

Considering for 1 bloc: 2.27Vpc x 6 cells x 1mA per Ah x 110Ah = 1.496W.

Therefore, for the complete battery of 40 blocs and 3 strings: -

1.496W x 40 x 3 = 180W.

This heat will be generated for as long as the battery is on float charge.

2.2) Considering the I2R method: -

Consider for one bloc: 110mA x 110mA x 3.8mΩ = 0.04598mW

Therefore, for 40 x 3 blocs = 5.5176mW or 0.005W.

This heat will be generated for as long as the battery is on float charge.

Interestingly, many battery manufacturers do not quote a value for the heat generated on discharge because lead acid batteries are considered as endothermic. However, manufacturers generally accept that the internal components and external connections all have a resistance and will generate heat when a current is flowing.

Again, a simple mathematical calculation may be used and most battery manufacturers accept I2R as a reasonable approximation to the heat loss on discharge. We need to know the discharge current and internal resistance of the battery system.

Using the same 40 x 12V battery discharged at 300kW for 15m we first need to modify the 300kW to a current that can be used in the calculation. The “safe option” is to consider the end of discharge voltage and then calculate the maximum current. The end of discharge voltage has been given as 408V (see above). Therefore, the maximum current is 300kW x 1000 / 408V = 735A.

The heat loss is calculated as: -

735A x 735A x 50.7mΩ = 27.4kW.

This may be expressed as Wh, i.e. 27.4kW x 0.25h = 6.85kWh

Because the battery has a thermal mass it may be many hours before this heat is transmitted to the surrounding air. The battery in this article would weigh approximately 4800kg. Some manufacturers consider the heat being dissipated to the room will be spread over 10 x the discharge time. In this example, this will be 2.5h. This would calculate out to be 2.74kW for 10h.

It is worth looking at the total battery dimensions and weight to give an appreciation of the heat loss when compared with the physical parameters of the battery. If the heat were generated within 1m3 it would be considerable. However, if the heat were within a 10m3 volume the impact would be minimal. The following parameters are real for the battery of 3 x 40 x 110Ah x 12V blocs give this perspective.

While the dimensions and weights given below are real, we have to remember that the stand is the open type with a large free volume around the monoblocs. The total volume considering the open area within the cells and between rows and tiers is calculated as: -

3.7 x 0.8 x 1.3 = 3. 8m3

Stand type: 2 row x 3 tier open steel type.

Length: 3.7m

Depth: 0.8m

Overal Height: 1.3m

Volume: 3.8m3

Weight: 4000kg

It is difficult to substantiate the results of the heat when the battery is on recharge or float charge because batteries do not follow standard electrical characteristic and therefore the results must be questionable. We know that ohms law when applied to batteries does not work. This is largely because of the BACK EMF characteristics of batteries, which makes the V x I calculations questionable. Therefore, any mathematical results that rely on this principal must be suspect. Accordingly the V x I calculations must be suspect. To understand this more fully, we can calculate the theoretical float current using the I = V / R method. In our examples, we know that the applied float voltage is 2.27Vpc i.e. 13.62V for the 12V, 6 cell bloc and we know the resistance is 3.8mΩ. Applying ohms law, the float current should be I = V / R = 13.62V / 3.8mΩ = 3584A. Clearly, this is not correct.

If the V x I calculations are not reliable then we must also question the I2R results. What we do know, is that the current is real value and the internal resistance is also real. Therefore, the results must be more accurate, we hope!

The I2R results are more real because we know what the current is and we know the internal resistance of the product. The I2R results for recharging are very small and in all practical terms, the heat may be ignored. This is only 5.5mWh in the example.

Again, if the I2R results are more real and the V x I method is unreliable then the 0.005W heat on float charge may again be considered irrelevant.

The only method that seems to be used for heat on discharge is I2R and, as expected, the heat on discharge is considerably higher than that on recharge or float charge. What we have to remember is that the heat will not be given up immediately and some estimation of the time this is given up has to be made. Without doubt this will be hours rather than minutes but this is a matter of opinion without consulting a heating engineer.

For recharging and float charging the heat is very small, particularly when we consider the mass of the battery. This is fortunate because although there are different methods in use, the results are insignificant when put into the context of removing heat from a battery room.

For the heat generated on discharge, the situation is different because most battery manufacturers accept the I2R method as the most accurate. In addition, we can more readily accept the results because there is no BACK EMF on discharge. In this example, the heat generated can be expressed as 27.4kWh but when considering the mass of the battery we must consider this heat to be given up over a longer time than the actual discharge period of 15m. Not all manufacturers consider a time of 10 x the discharge time, but it is clear that the heat will not be given up instantly.

This article discusses test discharging of VRLA batteries. The principal is almost identical to the procedure that should be followed when testing vented Lead-Acid batteries but some important changes are necessary to obtain the correct results.

The article covers the broad spectrum of testing and some manufactures and users may prefer to test batteries in a slightly different way. Never the less, this article follows the UK accepted practice as given in BS 6290 Part 4 1997. It also discusses the International Standard IEC 60896 -

The article does not discuss “fault finding” which can be found in another article.

1.1) BS 6290 Part 4 principally considers the 3h to n x 1.80Vpc at 20°C rate as the preferred discharge rate when capacity testing. IEC 60896 suggests 3h to n x 1.70Vpc at 20°C or 25°C. Within the BS additional discharge rates are listed for discharge tests between 1h and 10h. Also listed are 4 discharge rates of 1m, 5m, 15m, and 30m. There is a recommended final voltage for all discharge rates.

1.2) Frequently, the user will want to test at a discharge rate other than those recommended in the BS or IEC and this can normally be accommodated. For example, the 10m is frequently specified for UPS batteries to a final voltage of 1.70Vpc. Any test can be carried out but it is important to agree these before commencement and agree the pass / fail points.

2.1) Before any test is carried out, the battery must be commission charged in accordance with the manufacturers procedures. For VRLA AGM batteries, this can normally be achieved by applying a voltage to the battery terminals averaging either typically 2.40Vpc for 48 hours or at a voltage of typically 2.27Vpc for 6 days. Other voltages and times may be used but the manufacturer must approve any method. For VRLA GEL product, the voltage may be increased to typically 2.45Vpc for the 48h charge and reduced to typically 2.25Vpc for the 6-day charge.

2.2) AGM 2.40Vpc and 2.45Vpc GEL for 48 Hours Method.

For all commissioning charge procedures, the battery must be located in a well ventilated environment at an average voltage of 2.40Vpc / 2.45Vpc with a minimum current equivalent to 5% of the A·h capacity expressed as C3 A and a maximum current of 10% C3 A·h capacity expressed as A. The battery and battery room must be at a temperature of between +10°C and +35°C. If the battery and room are outside the range 15ºC and 25ºC the charging voltage should be adjusted to compensate. For the correct applied voltage the manufacturer must be consulted.

2.3) AGM 2.27Vpc and 2.25Vpc GEL for 6 Days Method.

Charge the battery in a well-ventilated environment at a voltage recommended by the manufacturer. This voltage is normally the voltage at which the battery will be charged at when in normal service. As with the 2.40Vpc / 2.45Vpc for 48h method, the current available must be not less than 5% of the C3 A. However, the maximum current is often not limited and frequently, 100% C3 A is acceptable. The battery and battery room must be at a temperature of between +15°C and +25°C. If outside this range, the manufacturer must be consulted to establish the correct charging voltage.

2.4) At the end of the 48 hours, 6 days or other approved method, measure and record the typical cell or monoblocs temperature, ambient temperature, charge current total battery voltage and individual unit voltages.

2.5) If any individual unit voltage differs from the average significantly, the battery manufacturer should be consulted before the test carrying out the discharge test. For guidance, the following values may apply as limits.

For 2V cells: maximum difference from the average: 0.10V For 4V units: maximum difference from the average: 0.14V For 6V units: maximum difference from the average: 0.22V For 12V units: maximum difference from the average: 0.44V For 16V units: maximum difference from the average: 0.55V

2.6) No matter which charging method is used, the manufacturer must approve it and a record of the charge must be made.

3.1) Typically and to comply with the requirements if BS 6290 Part 4, the battery must be held on open circuit for not less than 1h and not more than 24 hours before commencing the discharge test. However, in a real live situation when the mains power fails, the battery will go from float to discharge immediately. Therefore, many users stipulate that the test shall be performed “direct from float”. It is rare that this procedure is a handicap to the battery and in most cases; it is to the batteries advantage.

3.2) Before starting the test, record the surface temperature of several units. This temperature shall be used in the final analysis for the temperature corrected discharge time. This is the only temperature that is used to “temperature compensate” the discharge time. For vented products the temperature is recorded at intervals and the average is used to compensate the discharge time. Also measure and record the ambient temperature near the battery. The unit and ambient temperature must be within ±5ºC. If they are not, do not start the test and contact the Company for advice.

3.2.1 The battery may be discharged at any agreed discharge rate including constant current, constant power or constant resistance. In all cases, the parameter should be maintained within ±5%, for the whole of the discharge.

3.2.2 During the discharge test, measure and record at regular intervals, the parameters below. Typically, at least 10 measurements should be taken at regular intervals.

3.2.3 The individual unit voltages, including one inter unit connector. 3.2.4 The overall battery voltage, at the battery terminals. 3.2.5 The discharge current or power if the discharge is at constant power.

3.3 When the overall battery voltage, when measured at the battery terminals, falls to the minimum accepted, record the time, and all individual unit voltages. Stop the discharge test.

3.4 The battery should be recharged as soon as possible. For high current discharged such as 5-minute constant power, the battery may be elevated above the ambient by several degrees. In spite of this, the battery may be recharged immediately. If the recharge cannot be started immediately, the manufacturer usually accepts a stand period of up to 12 hours. However, this should be confirmed.

3.5 If during the discharge test, the voltage of any individual 2V cell falls to below 0.5V, it is recommended that the test be stopped. Similarly for individual voltage of 6V units falling below 3V, 12V units falling below 6V, and 16V units falling below 8V, stop the test.

4.1 The test shall be deemed passed providing the discharge time corrected for temperature meets the requirement

4.2 The battery manufacturer should be consulted for the actual temperature correction values for the type of battery and discharge that it has been tested at.

4.3 The British Standard allows for individual cell or monobloc voltages to be below the average but the battery, as a complete system must meet the requirements. For example, BS 6290 Part 4 states that as a complete battery the terminal voltage shall be not less than n x 1.80Vpc at the end of a standard 3h test and all cells shall be not less than 1.75Vpc after 2h 30m. It will be apparent that if an individual cell has marginally exceed the 1.75Vpc at 2h 30m then there is still a likelihood that the battery as a complete system will fail or the individual cell voltage will be lower than the suggested 0.5V (see the last paragraph in section 3.4 above. This principal may be extended to cover different discharge rates. It is suggested that a minimum voltage of 0.05Vpc below the minimum average accepted as a pass at the end of discharge time at 83% of the discharge time. For example, if the discharge test has to run for 10m with a minimum voltage of 1.65Vpc, then every cell shall have a voltage of not less than 1.60Vpc after 8m 20s.

4.4 The graph below illustrates a battery tested at the 10m rate where the total battery voltage exceeded the requirement but one cell collapsed after about 4m. This battery test should be regarded as failed. In these circumstances, the faulty cell should be replaced. Although the customer may insist on a complete battery re-test it is often acceptable to test a single cell separately and insert it into the battery system. This avoids a complete battery re-test. This battery system consisted of 333 single cells of 950Ah each. In a subsequent teardown analysis of the battery the fault was discovered as an internal short caused by a bent end negative plate shorting to an adjacent positive plate

The document is intended to give the reader a better understanding of the difference between the major classifications of BS 6290 Part 4 (Lead-acid stationary cells and batteries – Part 4 Specification for classifying valve regulated types) and IEC 60896 – 22 (Stationary lead-acid batteries – Part 22: Valve regulated types – Requirements).

This document is not intended to comment on the specification classifications or make any judgement to the validity of tests or claims by manufacturers. The article is intended to draw attention to the different classifications and how the British Standard varies with the latest International Standard.

When VRLA AGM types first became popular in the UK in the early 1980’s there was no standard covering these types. The first British Standard for VRLA AGM product was introduced in 1987. This was a fair attempt in making a comparison between VRLA AGM cells and High Performance Planté cells which dominated the UK stationary battery market at that time. However, it soon became evident that the specification did not cover many aspects that users were interested in and in some cases it was misleading. Over the following 10 years a revised Specification was drawn up and this was issued in 1997.

The 1997 version changed several requirements and introduced different classifications rather than one classification which covered many factors. It also clarified many issues including gas emission when in the overcharge float condition. Some changes to the specification were made because the 1987 version was difficult or in some cases impossible to meet without incurring considerable manufacturing cost. Some products under development at that time were intended to comply but would not because of idiosyncrasies within the specification wording. The 1997 version was a vast improvement. Significant introductions and clarifications to the Standard were; a) Safety Class, b) Performance Class, and c) Durability Class. The three classifications are discussed below.

a) Safety Class.

This is a reference to the flammability of containers, lids and covers and 3 classes are quoted i.e. FV0, FV1, and FV2. The highest level being FV0. To all intents and purposes FV0 means that the plastic will self extinguish. FV1 and FV2 are not as good as FV0 with respect to flame retardant level and may continue to burn even when the ignition source has been removed. There are 3 levels i.e. 1, 2, or 3. Product compliant with FV0 is referred to as level 1, FV1 to level 2 and FV2 to level 3.

b) Performance Class.

This is the compliance that the product has with respect to the published data. The manufacturer must state the compliance at seven different discharge rates as specified in the Standard. This section has 4 levels where the highest compliance is 1. In all cases, the product does not have to give 100% performance on the first discharge after a fully commissioning charge. In reality, this is very acceptable because although some cells may fall short of 100% compliance, in the vast majority of cases, the complete battery would easily exceed the claimed performance

c) Durability Class.

This part of the Standard refers to the life at elevated temperature and is more commonly referred to a “design Life”. However, it is not simply an accelerated test because several additional parameters are evaluated. This class has 5 levels where level 1 is the highest compliance and in terms of “design life” can be considered as 20 years. The “lowest” class is level 5 which equates to a design life of 4 years. In reality, we now know that the real life and design life are not the same and it is the view of the author of this article to consider a real life of 80% of the design life.

There are other tests within the Specification that the manufacturer must give answer to for the product tested. One of these is High Current Endurance and product is classified as “H” or “L”. Basically, the “H” indicates that the product is suitable for high current discharge applications such as switch tripping and closing or engine starting. The “L” classification signifies that the product is only suitable for low current applications such as discharge currents lower than the published 30 minute rate. The overall classification for a product having a safety class 1 with high current capability, performance class 2 and durability class 3, will have an overall classification of 1H23. A classification of 1L23 should not be considered as inferior. There are good commercial reasons for a product designed for low current applications. The product should not be considered as inferior. Similarly, product having a classification of 1L34 may still meet the specification required.

What we have to remember is that a product which is fully compliant with BS 6290 Part 4 1997 may have an overall classification of 3L45. This product has the lowest classification for flammability, it is only suitable for low current applications, the performance class could mean that when the product is assembled as a battery it may fail to meet the published performance and the design life is only rated as 4 years. However, it is a fully compliant product. Time moved on and in 2004, the IEC specification 60896 – 22 was published.

IEC Standards are, as far as possible, an agreement for technical matters on an international basis. They are not “European Mainland” specifications and the UK organisations have a large input to the content. IEC 60896 looked very closely at BS 6290 Part 4 and has adopted many points plus additional requirements to make it more comprehensive. It does not have the same classification spectrum and as one example it refers to the Eurobat Guide for the design life which is not within the IEC 60896 Standard itself.

Making the comparison to the BS 6290 Part 4 specification, the following is applicable.

a) Safety Class.

The Eurobat document considers 4 “effective” classes whereas the BS has 3. The Eurobat refers to “standard” flammability for Vertical Flame of FV1, FV2 plus a lower specification of Horizontal Burn (HB). In addition to FV1, FV2 and HB the Eurobat Guide refers to a “High Premium flammability rating FV0”.

In conclusion: -

BS6290 Part 4: FV0, FV1 and FV2 Eurobat Guide: FV0, FV1, FV2 and HB.

b) Performance Class:

Tests carried out on a sample of 6 units as despatched from the factory must have at least 95% compliance with the rated performance. In effect, this means that every battery may fail to meet the rated performance when new. This needs to be put into perspective with the real world. We know that all lead-acid batteries will lose performance as they age. For VRLA product is it normally accepted that the end of life is when the product will not meet 80% of the rated value. Accordingly, batteries need to be oversized to allow for this end of life characteristic. It follows that if the battery can achieve 95% or more on first discharge then the installation is safe. Interestingly, it is reasonable to assume that all VRLA cell performance will increase by typically 5% during the first few years of float operation.

In conclusion: -

BS6290 Part 4: A complex list of different compliance levels at different rates. Eurobat Guide: Not less than 95% compliance at the test rates.

Note to avoid confusion: Planté batteries do not need to be oversized because they are deemed to have reached the end of life when they will not meet 100% of the published performance. They still have a real life of 25 years on float charge systems at 20°C.

c) Durability Class.

In the Eurobat Guide this is referred to as Design Life but the classifications are similar as below. Note: The BS quotes days on test before failure. The author has reclassified these to the nearest whole year to reflect years “design life”. For example, Durability Class 1 is >648 days which is corrected by multiplying by 11.36. This results in a calculated life of 7361 days i.e. 20.2 years.

In conclusion: -

BS6290 Part 4: >4 years, >8 years, >12 years, >16 years and >20 years Eurobat Guide: 3 – 5 years, 6 – 9 years, 10-12 years and 12 years and longer.

It is fair to say that neither “standard” is easy to understand. As such, the required battery specification from the user has to be clear and this must be documented to avoid confusion. It is entirely possible for the users required specification to call for every cell or monobloc to have not less than the published data on the first discharge. Under these circumstances it would be reasonable for the manufacturer to accept this after a commissioning charge, but this requirement may result in a premium price.

Many large users have their own requirements which may be a mixture of different standards. Reputable manufacturers will have no issues with this as a buying specification.

It would be irresponsible or foolhardy to simply request a battery that complies with IEC 60896 or to BS 6290 Part 4 without at qualifying the parameters as discussed above as a minimum. It has to be remembered that there are several classifications within the standards and these must be specified by the user.

Let the buyer beware and specify the requirements in detail.

Finding the correct battery size for your requirement can prove difficult, with terminology appearing somewhat confusing. Most lead acid batteries are manufactured to 2V, 6V or 12V, however the 'Ah' rating can be greatly varied. So what does ‘Ah' mean and how does this affect your choice? What other details should be taken into consideration?

‘Ah’ is an abbreviation of ‘Ampere Hour’, though sometimes you may also see ‘Amp Hour’ used, and is a standard unit of electric charge, equal to the charge transferred by a constant current of one ampere flowing for one hour. Looking at lead acid batteries the voltage and Ah is generally described on the battery itself, for example the Yuasa NP7-12 is described as 12V 7Ah. In this instance the 7Ah is given as the 20 hour rate, this tells us that at a constant discharge over 20 hours the battery will produce 0.35 amp per hour (20 hours x 0.35 = 7Ah).

An important factor to consider when comparing Ah rates between batteries can be the ambient temperature in which the Ah rating is based on, for example, a 12V 18Ah battery using the parameters of a 20 hour rate at 20 degrees celsius may also be described as 12V 20Ah if the parameters of 20 hour rate at an increased temperature of 25 degrees celsius is applied. The higher rating is produces as the second rating was calculated in a warmer ambient environment, batteries will produce more power in an environment with higher ambient temperature though this can be at the cost of service life. You could be forgiven if in replacing an 18Ah battery with a 20Ah battery in thinking that the new battery has a higher power rating, however it may just be the case that the rating has merely been calculated using differing parameters to the original.

Similarly lead acid batteries can be described at differing hourly rates, the most common for VRLA lead acid probably being 10 and 20 hour rates. This also needs to be a consideration when comparing battery types of similar power descriptions, as often the longer the hourly rate the more efficient a battery will operate, so a 20 hour discharge will most likely produce a higher Ah rating than a 10 hour discharge period.

As always the team at Blue Box Batteries are on hand to help with any questions or concerns you may have. Should you be in any doubt as to the correct battery rating for your application we will always work to provide you with the assistance you need.

The open circuit voltage of lead acid cells can give us useful information providing it is used correctly.

This article shows the relationship between the open circuit voltage and state of charge of lead-acid cells and how to interpret the values. The values are always considered as that of a single cell and monobloc voltages must be averaged according to the number of cells. This method was developed primarily for VRLA cells where the specific gravity cannot be measured. Measurement of the specific gravity will give a good indication of the state of charge. For measurement of either specific gravity or open circuit voltage, the fully charged and fully discharged specific gravity must be known. Used correctly, the method described in this article will give reasonable results to the state of charge of cells on open circuit.

Some specific rules apply when using this method to determine the state of charge as follows: -

a) The cell or monobloc must have been on open circuit for at least 12 hours and preferably more than 24 hours.

b) The cell or monobloc temperature should be between +10°C and +30°C for best accuracy but any cell or monobloc temperature between 0°C and +40°C will give reasonable results.

c) The measuring instrument must be accurate to within ±1% and capable or recording voltages from 0.5V to 2.2V to at least 2 decimal places.

The open circuit voltage of vented cells such as Planté, tubular and flat plate pasted types can be used in a similar way to the method used for VRLA AGM and VRLA GEL cells. However, when the electrolyte specific gravity can be measured, this will give a more accurate result. Never the less, open circuit voltage measurement can provide sufficient accuracy in most cases and the ease of measurement means the process is quick. A graph below gives the typical parameters. Similar graphs for VRLA AGM an VRLA GEL products may also be found below.

It can be seen that the relationship between the open circuit voltage and state of charge follows a straight line. This, in the most port is correct within the limits required for lead-acid batteries. The mathematical equation is: -

Open Circuit Voltage = Specific Gravity + 0.84.

Example: OCV = 1.15s.g. + 0.84 constant = 1.99V. Note: this is close to 2V which is why the measuring instrument needs to be reasonably accurate.

If we know the fully charged and fully discharged specific gravity, we can draw the straight line graph.

Typically for VRLA AGM cells, the fully charged specific gravity is 1.310s.g. at 20°C and the fully discharged value is 1.10s.g. These reflect an open circuit voltage of typically 2.15V when fully charge and 1.94V when fully discharged. It should be noted that even when the cell is fully discharged the voltage will be close to 2V again illustrating that the measurement accuracy is important. When it comes to monobloc voltages, the total voltage needs to be divided by the number of cells to find an average.

We can make use of this characteristic in several ways but the most practical is to establish the state of charge of cells that have been in storage for some time. Typically, for VRLA AGM cells, manufacturers will recommend a freshening charge after 6 months in storage. This will be shorter for higher temperatures and longer for lower temperatures. For example, 6 months at 20°C will typically become 3 months when the storage temperature is 30°C. Although the storage time may be extended if the temperature is lower, this is not recommended beyond the manufacturer’s recommendations because irreversible sulphate may result and special charging techniques may be required to fully recover the product. This technique of measuring the open circuit voltage can be used to identify low voltage cells prior to installation. For monoblocs this is more difficult but can be used to identify low cells within a monobloc that may be the result of a soft short resulting from a manufacturing fault.

For example if a battery contains 55 VRLA AGM cells it would be expected that the voltages will all be reasonably the same such as within 0.02V of each other; e.g. 2.10Vpc to 2.12Vpc. If any cell is outside the 0.02V limit, this should be questioned. Where soft shorts are concerned, the voltage differential is likely to be more like 0.05V for cells that have been on open circuit for some time.

Although the principal can be used for multi cell monoblocs, it is not as easy to identify rouge product. Even soft shorts which may result in one cell having a voltage 0.05V lower than the average are not very easy to identify. For example, if we consider a 12V 6 cell monobloc having a voltage of 12.70V this could be made up of 6 cells each 2.117V or 5 cells each 2.12V and one cell of 2.10V (total 2.12V x 5 = 10.6V + 2.10V = 12.7V). Under normal circumstances this would not be rejected and the rouge bloc may be missed.

For monoblocs, the voltage deviation from bloc to be needs to be wider otherwise good product may be rejected. For 6V 3 cell blocs a limit of 0.1V of each other is reasonable to question the unit. For 12V 6 cell blocs a variation of 0.15V is reasonable. These limits may seem close but in the context of monoblocs, all the cells would have been stored in exactly the same conditions and this cannot be said for individual cells. In fact, individual cells may have come from different batches and it is possible that a group of cells may be different from another group of cells. The technique must be used with some knowledge and expertise and if in doubt, it would always be wise to contact the supplier or manufacturer.

Looking in isolation at bloc voltages may not show up any rouge units but if a battery comprised 33 blocs and 32 are 12.70V and one bloc is 12.60V, a question has to be raised. The installation process could be completed but a note should be made and if the bloc voltage does not recover on float charge after 6 months and perhaps deteriorates or the differential between other blocs increases, then there is a reasonable case for exchanging the rouge bloc.

A final comment about open circuit voltages is to mention that for NiCd cells you can not apply this principal because the s.g. of the electrolyte does not change with state of charge. It shall also be noted that the NiCd open circuit voltage is not stable and typically varies between 1.2V and 1.3V.