A Guide To VRLA Battery Formation Techniques

A Guide To VRLA Battery Formation Techniques
www.firing-circuits.com
A Guide To VRLA Battery Formation Techniques

By Mike Weighall and Bob Nelson
Mike Weighall is an independent consultant with 36 years experience in the battery industry. He obtained his Chemistry degree from the University of Manchester Institute of Science and Technology. He has spent most of his working career associated with the battery industry, in a range of technical and managerial roles with major UK employers, including Lucas, Crompton, Cookson, and ENTEK International. In recent years he has played an important role in the ALABC (Advanced Lead Acid Battery Consortium) as Chairman of the European Technical Committee, member of the Research Management Team, and currently Chairman of the Project Advisory Team on Separators. He has previously written the Battery Test Guide for Digatron/ Firing Circuits. He has presented nine papers at International Battery Industry Conferences, four of which were published in the Journal of Power Sources.
M.J. Weighall MJW Associates 12 Low Stobhill Morpeth Northumberland NE61 2SG Tel: +44 1670 512262 Fax: +44 870 056 0376 Mobile: +44 7977 459819 Email: mjweighall@battery1.demon.co.uk
Bob Nelson is an independent consultant with over 23 years experience in the VRLA battery Industry. He obtained his Chemistry degree at Northwestern University in 1963 and his PhD in Analytical Chemistry/Electrochemistry at the University of Kansas in 1966. After spending 11 years in teaching and research at the university level, he joined Gates Energy Products where he worked for 13 years in various positions dealing with the development and manufacture of both spiral-wound and flat-plate VRLA products. He has also worked with other specialty VRLA products during work tenures with Portable Energy Products and Bolder Technologies. In between, he spent three years with ILZRO, where he was responsible for organizing and managing the Advanced Lead Acid Battery Consortium. In addition to the publication of some 39 refereed papers, two book chapters and 22 invited presentations at national and international conferences during his academic career, he has published 42 papers and given 41 presentations on VRLA battery technology over the past 23 years.
Dr. Bob Nelson Recombination Technologies 909 Santa Fe Drive Denver Colorado CO 80204 Tel: +1 303 573 7402 Fax: +1 303 573 7403 Email: nelson909santafe@aol.com
2001 Firing Circuits, Inc.
Printed In U.S.A., All Rights Reserved
Table Of Contents
Paragraph Page Paragraph Page
1. Introduction ....................................................................... 1 2. Plate Formation vs. Jar Formation .................................. 2 3. The VRLA Formation Process Jar Formation ........................................................................ 3 3.1 The Filling Process ..................................................... 3 3.1.1 Acid Density for Filling ..................................... 4 3.2 Fill-to-Form Processing .............................................. 4 3.3 Formation ................................................................... 4 3.3.1 Battery Preparation for Formation: Open Formation ..................................... 4 3.3.2 Fill and Spill Formation ................................. 5 3.3.3 Saturation/ Electrolysis Formation ................... 5 3.4 Formation Time .......................................................... 5 3.5 Completion of Formation ............................................ 6 3.6 Formation Algorithms ................................................. 6 3.7 Initiation of Current Flow ............................................. 7 3.8 Constant Voltage Charging ........................................ 8 3.9 Constant Current Charging ........................................ 8 3.10 Taper Current Charging ........................................... 9 3.11 Pulse Current Charging .......................................... 10 3.12 Rests and Discharges ............................................ 11 3.13 Sample Formation Algorithms & Profiles ................ 12 3.13.1 A Simple Algorithm ...................................... 12 3.13.2 More Typical Charge/ Rest/ Charge Algorithms ........................................ 12 3.14 Development of a Suitable Formation Algorithm ...................................................... 14 4. Temperature Limits for VRLA Jar Formation ................. 15 5. VRLA Battery Manufacture using Plate Formation ....... 16 6. Technical and Theoretical Background .......................... 17 6.1 The Formation Process Explained ........................... 17 6.2 Formation Processes and Ah Input .......................... 18 6.3 Key Differences Between Flooded and VRLA Batteries .......................................... 19 7. Jar Formation Additional Information ....................... 20 7.1 Battery Preparation for Formation Sealed Formation ........................................................... 20 7.1.1 Plate Curing and Carbonation ....................... 20 7.2 Acid Filling ............................................................... 20 7.3 Control of Formation Temperature ............................ 23 7.4 Completion of Formation .......................................... 25 7.5 Alternative Jar Formation Options ............................ 25
8. Battery Design ................................................................ 26 8.1 Plate Height/ Plate Spacing Ratio ............................ 26 8.2 Battery Case Draft ................................................... 26 8.3 Active Material Additives ......................................... 26 8.4 Electrolyte Additives ................................................ 27 9. Separator Optimization .................................................. 27 9.1 Volume Porosity ....................................................... 29 9.2 Saturation Level ....................................................... 29 9.3 Caliper ..................................................................... 29 9.4 Compression ........................................................... 29 9.5 Grammage ............................................................... 30 9.6 Surface Area ............................................................ 30 10. Separator Designs to Improve Wet Formation ............ 31 11. VRLA Gel Batteries ....................................................... 33 12. Formation Equipment and Layout ............................... 34 12.1 Battery Connections .............................................. 34 12.2 Formation Bay or Circuit Configurations ................ 35 12.3 Critical Maintenance of Formation Equipment ....... 36 12.4 Power Quality and Equipment Costs ...................... 36 13. Battery Monitoring During Formation ......................... 37 13.1 Electrical Monitoring .............................................. 37 13.2 Temperature Monitoring ......................................... 39 13.3 Gas Monitoring ...................................................... 39 14. Post-Formation Handling and In-Line Product Testing ...................................................... 40 14.1 Visual Standards .................................................... 40 14.2 In-Line Product Testing .......................................... 41 14.2.1 Open-Circuit Voltage Measurement ............. 41 14.2.2 AC Impedance Measurements .................... 42 14.2.3 High-Rate Discharge Measurements .......... 43 15. Troubleshooting: Problems and Solutions .................. 44 16. References .................................................................... 48
Appendix 1:
Glossary of terms and abbreviations ..................................... 49
Listing of Figures
Figure Page Figure Page
Figure 1. Examples Of Techniques For The Initiation Of Formation Charging ............................ 7 Figure 2. Examples Of Single-And Multi-Step Current-Limited Constant-Voltage Formation Profiles .................................................. 8 Figure 3. Examples Of Stepped Constant-Current And Conventional CC Formation Profiles Compared To An Ideal Formation Curve. ............... 9 Figure 4. Examples Of The Progressive Influence Of Temperature Monitoring On The Efficiency Of The Formation Process. .................. 10 Figure 5. Taper Current Charging. ...................................... 11 Figure 6. Examples Of Pulsed Charging Algorithms. .......................................................... 11 Figure 7. Typical Constant-Current Formation Profiles For A 12V/20Ah VRLA Battery. ................ 12 Figure 8. Typical Constant-Voltage And Taper-Current Formation Profiles For A 12V/20Ah Battery ............................................. 13 Figure 9. Typical Constant-Current Formation Profiles With Rests Or A Discharge For A 12V/20Ah VRLA Battery. ............................. 14 Figure 10. The Filling Process Within A Vacuum And Non-Vacuum Fill. .......................................... 21 Figure 11. Conceptual View Of the Filling Process For A VLRA Cell. .................................... 21 Figure 12. Action On The Leading Edge Of The Liquid In A VRLA Cell Filling Process. ................. 22
Figure 13. 2.5 Ah And 25Ah Spiral-Wound Single-Cell Internal Temperatures During Different Fill-To-Form Conditions. ............. 24 Figure 14. 6V/100 Ah Prismatic Battery Temperature Data During Fill-To-Form Time With Different Conditions. ............................ 25 Figure 15. Solubility Of Lead Sulfate In Sulfuric Acid At 25C. .......................................... 27 Figure 16. Mean Pore Size Vs. Kr Bet Surface Area. ....................................................... 30 Figure 17. Impact Of Surface Area (m2/g) On Water Wicking Height While Under 20% Compression, After 24 Hours. ..................... 31 Figure 18. Effect Of Fiber Mix And Segregation On Vertical Wicking Speed. ................................. 31 Figure 19. Upward And Downward Wicking Height For Oriented And Non-Oriented Fibers. ........................................... 32 Figure 20. Battery Connections For Series Strings, Series-Parallel Arrays And Series-Parallel Matrixing. ..................................... 35 Figure 21. AC Ripple Voltage And Current Representation And Its Effect On Cell Temperature And Cycle Lifetime. .................. 37 Figure 22. Typical Self-Discharge Curves For VRLA Batteries. ............................................. 41 Figure 23. High-Rate Discharge Voltage/Time Curves For Acceptable And Unacceptable Battery Performance On A 5-Second Test. ............................................ 43
Listing of Tables
Table
Table 1.
Page
Typical ampere-hour inputs in relation to wet paste weight and dry cured paste weight. ....................................................... 19
Typical AC Impedance Values For A Variety Of Thin-Plate VRLA Single Cells And Batteries Fully Charged At 25C. .................. 48 Sample OCV Chart Used In Manufacturing To Sort Cells Or Batteries After Formation Or Recharge. ............................. 49
Table 2.
Table 3.
1. Introduction
The purpose of this brochure is to guide the battery manufacturer in the formation of VRLA (Valve Regulated Lead Acid) batteries. The information is nominally confined to small VRLA batteries with capacities in the range 1.2 Ah to 100 Ah. Because jar formation of VRLA batteries is far more difficult than plate formation, this aspect of VRLA battery formation will comprise the largest section of the brochure.
The brochure is divided into several sections:

s
Sections 2 through 5 deal with practical issues related to VRLA battery formation, and deal mainly with jar formation. Section 6 deals with the technical and theoretical background. Section 7 gives additional information about jar formation. Sections 8 through 10 give battery and separator design guidance. Section 11 is a brief overview of VRLA gel batteries. Sections 12 through 14 deal with formation equipment, battery monitoring and product testing. Section 15 deals with troubleshooting formation problems.
s s s s s
Practical issues related to VRLA battery and jar formation
2. Plate formation vs. jar formation for VRLA batteries

The first decision the VRLA battery manufacturer has to take is whether to use a plate formation or jar formation process, and this section highlights some of the issues that need to be taken into account before making this decision. Plate formation may result in fewer manufacturing and technical problems in respect of battery design, process control, quality, performance and life. The merits of plate formation are particularly apparent for larger, higher capacity batteries where a long cycle life and/or calendar life is required. However, particularly for the smaller batteries being discussed in this brochure, many battery manufacturers are choosing jar formation. This may be for reasons of cost and convenience, but may also be because the battery design does not lend itself well to plate formation. This may apply for example with thin plate cylindrical or prismatic battery designs. The decision as to which process to use will be dictated by the detailed battery design and manufacturing constraints as described in more detail later. However, there are other manufacturing and cost issues to be considered. The total cost of the plate formation/dry charge process may be higher than jar formation when one takes into account the following factors:
s
The decision as to whether to plate form or jar form will be based on a number of factors which the battery manufacturer needs to take into account, and will be discussed in more detail later. Some general guidance is given below:
Plate formation should be used in the following formation circumstances: circumstances: s Plates for tall batteries s Plates for large, high capacity batteries s Plates for very long life batteries s Battery with high L/d ratio (>100) (see section 10.1) Jar formation consider in the following circumstances: formation circumstances: s Cylindrical battery design s Thin plate prismatic battery design s Battery with low L/d ratio (<100) (see section 10.1) s Large separator fringe area s High separator grammage (>=2g/Ah) s High surface area separator Other issues, which also need to be considered for jar considered formation: formation: s Whether single cells or monoblocs, and how many cells in the monoblocs (e.g. 3 or 6 cells). This will have an impact on the efficiency of cooling and temperature variations between cells. s If plates have been cured to produce high levels of tetrabasic lead sulfate (4BS) the plates may be more difficult to form, and require a higher charge input during formation than for tribasic lead sulfate (3BS) cured plates. s The inclusion of red lead in the positive paste mix will assist jar formation and enable lower Ah input and shorter formation times. It will also improve the initial electrical performance.
In principle all VRLA batteries could use plates prepared using plate formation/dry charge: but not all VRLA batteries can be successfully jar-formed. The information above is given for guidance only, and the suitability of a particular battery for jar formation should be established by careful experimentation. Jar formation of VRLA batteries is actually quite a complex process and will now be dealt with in detail.
The cost associated with the neutralizing and cleaning or disposing of the plate wash water. This water must be neutralized and cleaned of heavy metals before it can be recycled or discharged into a public sewer system. The capital and operating cost of the dry charge operation (e.g. inert gas drying). Post assembly charge and discharge cycles to recover the capacity loss that is inherent in the dry-charge process. Plate lug cleaning before final assembly.
3. The VRLA Formation Process Jar Formation

3.1 The filling process
The formation process for valve-regulated lead-acid (VRLA) cells and batteries really begins with the filling process. Several approaches can be taken to filling, including:
s s s
reproducible in terms of fill weight and the final saturation level is typically ~95% (i.e., the plate stack doesnt saturate). Push fill is a specialized technique for spiral-woundtype products where electrolyte is forced up through the wound element, either from the bottom or using a probe in the wound-element mandrel space. This is faster than the gravity-fill techniques (around 30-60 seconds) and, thus, requires more care in thermal management. Soft-vacuum filling involves drawing a moderate vacuum level and allowing the element to suck in electrolyte at its own rate. As this approach doesnt usually result in uniform electrolyte distribution there is often a pushpull (pressure-vacuum) finishing step to physically move electrolyte around to help diffusion. The filling rate is moderate (30-60 seconds) so thermal management is mandatory, along the lines of that given above for the gravity-fill approach. Hard-vacuum filling is a very rapid technique (on the order of 1-10 seconds for sizes 1.2-25Ah) and is, thus, attractive for high-volume manufacturing. However, it also requires extreme care both during filling and for processes prior to filling. In addition to speed, hardvacuum filling can result in uniform electrolyte distribution due to the almost total absence of air displacement. However, the absence of air also means that the paste is very reactive and the rapid introduction of electrolyte results in very high heat generation over a short period of time. Thus, thermal management is critical with this type of filling and it is impractical over a size of ~50Ah due to the inability to dissipate the heat rapidly generated, even with the chilling steps mentioned above. Poor thermal management can result in staining of the AGM separator by dislodged paste and/or plate deformation and case bulging due to heat and, possibly, steam generated from the filling reaction. In this type of filling, the formation of hydration shorts (lead sulfate in the separator) is also possible due to the high temperatures and low acidity conditions that can be generated. Plate carbonation during processing is also a problem because the rapid introduction of electrolyte can result in a burst of liberated carbon dioxide, which can help to defeat the vacuum created and result in low fill weights. Further liberation of CO2 can cause regurgitation of electrolyte in extreme cases. Separator damage can also result from the hydraulic action of the electrolyte if it is added too quickly, thus promoting plate-to-plate shorting due to the removal of overlapping separator between adjacent plates.
Gravity top fill, single or multi-step Gravity bottom-up fill Push fill where electrolyte is pumped into the cell or battery, usually from the bottom up (usable only with spiral-wound products) Soft-vacuum fill (>~20mm Hg), single or multistep, possibly with a push-pull step to distribute electrolyte more evenly Hard-vacuum fill (<~10mm Hg)
The first, a gravity top fill, is the simplest approach that can be used for any cell or battery (hereafter referred to, collectively, as batteries) and just involves pouring electrolyte into the headspace at a rate that the battery can accommodate. It can be done slowly with a single addition or in several measured amounts. This is a relatively slow process but it has advantages in that heat is generated slowly, there is not likely to be damage to the AGM separator and there is only a limited effect from carbon dioxide released from carbonated paste surfaces. There is the possibility of incomplete wetting due to trapped gas pockets. Heat generation in larger batteries can be counteracted by chilling the electrolyte (typically to 0 to 10oC) and/or the unfilled elements and, if necessary, putting the filled battery into a chilled water bath. The measures used depend in large part upon the size of the battery. For small products (1.2-10Ah), simple bath cooling after fill is sufficient (and this may not even be necessary for very small products). For larger sizes (10-100Ah), chilled electrolyte and bath cooling may be mandatory. Fill times are of the order of 10-40 minutes. Gravity bottom-up, or dunk, filling simply involves dipping a cell or battery into a bath of electrolyte (the case having a hole or holes in the bottom to allow ingress of acid) until wicking has resulted in complete filling of the separator and plate pores. This is also a slow process (several minutes), and has the advantages and drawbacks listed above for gravity top fill. An added disadvantage is that the filling hole has to be sealed before the battery goes into formation. In fact, simply letting the battery take as much acid as it wants is very
3.1.1 Acid Density for Filling This will depend on the battery application, the desired final density, and the amount of sulphation achieved during paste mixing. A typical filling acid density would be 1.26 with a final density of 1.28. Finished acid densities are normally in the range 1.28 to 1.32, depending on the application. Standby batteries tend to have lower densities (1.28) while high-rate batteries (aircraft, engine start) may have a higher density (~ 1.30). For deep cycle batteries the specified acid density may be in the range 1.28 1.32. With open formation the acid s.g. can be checked and adjustments made if necessary. For sealed formation, calculations need to be accurate as it is not possible to correct the acid s.g. after formation. Some VRLA battery manufacturers monitor the fill/ formation weight loss and then add an equivalent amount of water or dilute electrolyte before sealing the battery.
3.3 Formation
There are a number of factors to be considered in matching the correct formation algorithm to a given product, among these being:
s s s s
Product sealed or open? Temperature control and the use of air or water Formation time Desired level of completion of formation (i.e., % PbO2) Formation algorithm used (CC, CV, taper, pulse, rests, discharges?) Battery connection series-strings only, seriesparallel strings or series-parallel matrix? Monitoring parameters during formation Critical maintenance of formation equipment Post-formation handling and product testing
s s s
3.2 Fill-to-Form Processing

The time gap between electrolyte filling and initiation of the formation process is more important than may be realized. Even though VRLA batteries are electrolytedeficient and there is more than twice the amount of pastes as there is acid (on an ampere-hour basis), a significant amount of acid remains unreacted if batteries are put into formation immediately after filling. The longer the delay between fill and form the more lead sulfate is formed. This facilitates the formation process, but it also increases the resistance of the unformed plates (particularly the positive), as lead sulfate is an insulator. This is usually overcome by using 10% or more of red lead, Pb3O4, in the positive paste. Longer stand times after filling can also aggravate the conditions that can initiate hydration shorts by allowing lead sulfate to slowly dissolve and diffuse into the separator. With a good filling process this is not a problem, as even a mildly acidic condition will suppress lead sulfate solubility, particularly if sodium sulfate is used as an additive in the fill electrolyte. However, in cases where fill conditions resulting in areas of the plate stack where hot electrolyte depleted of acid can exist, batteries are put onto formation as quickly as possible after fill. When this is done, there is the danger of the battery overheating, as the formation process generates heat, particularly early when high plate resistances result in high I2 R heating, and additional heat is still being created by the ongoing filling reaction (which is very exothermic). In order to allow the filling reaction to go to completion and allow the battery to cool adequately, a fill-to-form time of between 2 and 4 hours is recommended.
4 www.firing-circuits.com
Battery Preparation Formation: 3.3.1 Battery Preparation for Formation: Formation Open Formation For jar formation the easiest technique is open formation, which usually means a condition where the vent valve has not been put in place. (The alternative, less frequently used technique of sealed formation, is discussed in section 7.1). Open formation may also indicate a formation where the battery headspace is open to the air. In either case, batteries are usually flooded, or close to it, and have the capability of removal or addition of acid during processing. Open formations are useful in that plate processing is not as critical (in terms of carbonation), heat dissipation due to gassing is greater by about an order of magnitude than in sealed formations (because the battery is formed in the flooded state) and adjustments in saturation levels are possible at any time. There are several approaches to open formation for VRLA batteries, among them being:
s
Saturated or near-saturated condition with provision for excess electrolyte handling; So-called fill-and-spill formation, where batteries are formed saturated and then the electrolyte level is adjusted at the end of formation; a variant of this is two-step formation, where the battery is first formed with dilute electrolyte which is replaced after formation with a higher specific gravity acid closer to the desired operational level; Saturated or near-saturated formation open, followed by saturation and electrolysis to achieve a target saturation level, usually ~95%.
Formation 3.3.2 Fill and Spill Formation This formation approach involves formation in a flooded state, followed by simple pouring off of excess electrolyte. This results in a near-saturated condition following formation (trapped gas in the plate pores ensures that some electrolyte is absorbed and, thus, there is a small amount of void space in the formed battery), which may result in higher-than-usual overcharge gassing and weight losses early in life and, possibly, acid leakage during heavy overcharge. This approach has been promoted by H&V for use with their Hovosorb II organic fiber/glass separator [1] (see also section 10) and is particularly well suited to manufacturing processes with high manual labor inputs and those where precise control over finished battery quality is not required. The saturation levels in the cells are not precisely known and significant cell-to-cell variations could exist. Heavy hydrogen gassing during formation must be accounted for, but the high levels of gas generation help with heat dissipation. In principle, a two-step formation could be used for VRLA products as is done for flooded lead-acid batteries. Here, the battery is filled with a relatively dilute electrolyte for the initial formation process, after which the forming acid is dumped and the battery is refilled with an electrolyte close to the desired final specific gravity after a finishing charge. The major problem for VRLA batteries is that the AGM separator (unlike flooded lead-acid separators) holds most of its electrolyte and any manipulation of the formed battery is likely to result in separator damage. Saturation/Electrolysis Formation 3.3.3 Saturation/Electrolysis Formation In order to have an accurately known saturation level in the region of 95% after formation, a method has been developed where a standard open formation is carried out, followed by over-saturation and pouring off of excess electrolyte (much like fill and spill above). The fully-saturated, formed battery (still open) is then subjected to a period of electrolysis at a known current level to drive off an accurately-known amount of water, thus getting the battery to the desired saturation percentage, after which completion of battery assembly can be carried out.
3.4 Formation Time

Very early work done by Ritchie at Eagle-Picher showed that flooded lead-acid batteries could be formed in ~2 weeks to a very high PbO2 level with minimal weight loss and consumption of just over the theoretical amperehour input of 241 Ah/kg of PbO. This is not feasible in large-scale manufacturing, but it does set a baseline against which more practical formation algorithms and times can be measured. Suggestions for typical amperehour inputs in relation to wet paste weights are given in section 6.2. In practice, small VRLA batteries are formed within 24-48 hours. The actual formation time will be dependent on a number of factors, but a general rule of thumb is that cell/battery capacities of ~20Ah or less require a roughly 24-hour formation, while larger sizes up to 100Ah can normally be formed in ~48 hours. Smaller batteries are easier to form because they are more compact and voltage drops across the plates are less. They also tend to have better heat-dissipation properties and can be formed at higher currents. Formation time is not the only criterion. Heat buildup is an issue that will tend toward longer formation times. High PbO2 levels move the time in the same direction, as does the requirement for lower formation weight losses. The positive plate (PbO2 level) is the key indicator of the completeness of formation (see also section 7.4). In fact, the negative active material, or NAM, forms relatively easily and it is rare that formation of the NAM is the limiting factor. If manufacturing throughput were not an issue, all formations would be done over several days, as most benefits flow from long formation times. However, formation throughput is of paramount importance and so the goal is almost always to achieve complete formation in a minimum amount of time. In order to accomplish this in large-scale manufacturing, it is important to have a deeper understanding of the chemistry of the formation process, and more detail is given in section 6.1. The chemistry involved in formation is fairly complex, consisting not only of the basic chemistry of conversion of lead sulfate to sponge lead and lead sulfate, coupled with the overcharge processes involving the decomposition of water, but also more subtle issues such as electrolyte diffusion and gas bubble formation.
3.5 Completion of formation

One or more of the following parameters can be measured to determine whether formation is complete:
s
3.6 Formation Algorithms

The best algorithms to use in forming VRLA batteries depend upon a number of factors, ranging from capital investment to desired product quality and the intended application. The battery manufacturer will need to carry out tests to establish the best algorithm for his specific manufacturing process and battery application, but the following guidelines may be useful. The optimum algorithm is likely to include a number of steps:
s
Battery, cell and individual electrode potentials become high and constant. Electrode potentials are determined using either cadmium wire or mercury/mercurous sulfate reference electrodes on test batteries. Top-of-charge voltages (TOCV) become constant, but will vary for individual VRLA batteries, depending upon the amount of oxygen recombination occurring at the end of formation; the more oxygen reduction taking place the lower the TOCV. This will be influenced by whether sealed or open formation is used. In fact, if a given formation system monitors TOCV values these can be used for matching small batteries into larger units, as this is a critical performance parameter. At the same time, electrolyte specific gravity becomes constant at some high level (relative to the starting density) due to conversion of sulfate in the unformed plates to sulfuric acid. Both plates gas uniformly and strongly. Temperature rises steeply toward the end of formation at a given applied current, reaching values as high as 65-75oC if the finishing current is not reduced. Internal examination of a cell or battery would show that the plates are uniformly colored and are without white spots (unformed lead sulfate); the negatives are soft with a gray metallic sheen and the positives are hard and are dark brown to black in color.
Low initial current to minimize temperature rise at the start of formation. There may be a continuation of the heat production from the oxide/acid filling reaction. There may also be a variable fill-to form hold time due to the time lag in filling a formation circuit queue. The low initial current will compensate for possible high battery resistances. The low current charge should be continued until the battery temperatures have fallen below 50C. One or more steps at a higher current during the main part of the formation process while battery resistance is at its lowest and heat generation is at a minimum. One or more steps at a lower current as the gassing phase is reached towards the end of formation. The formation process may also include rest period(s) and/ or discharge step(s).
s s
The chosen charging approach will usually depend on the desired amount of capital investment and/ or experience from making flooded lead acid products. The possible choices are:
s
In practice, most of these parameters cannot be measured routinely, particularly in a large matrix of batteries in a formation bay. With a given VRLA product, the normal approach is to carry out test formation algorithms on a few batteries, which are monitored and then autopsied in order to determine if they meet the above criteria. Following the establishment of a production-worthy formation algorithm, all batteries are then manufactured using it, with some form of periodic sampling and testing to ensure the desired level of quality and uniformity.
Constant-voltage (CV) Constant-current (CC) Taper-current (TC) Pulsed-current (PC)
s s s
When making the choice of appropriate charging equipment, it should be noted that the charging equipment from Digatron/ Firing Circuits Inc., offers computer control with optional battery monitoring to provide optimum control of charging parameters.
More information about each of these processes is given on the next page.
Additional information is given in section 7.4
3.7 Initiation of Formation Current Flow

Before formation current flow is initiated, most manufacturing operating procedures include so-called continuity checks, where individual battery strings are checked with an ohmmeter to ensure that the resistance of the string, while very high, is not infinite (indicating a battery with an open connection (poor COS or squeeze weld) or a defective formation connector lead-to-terminal contact). If an abnormally high or infinite resistance reading is taken, the formation room personnel must identify the source of the problem; otherwise, a complete string of batteries will not be formed and, in a seriesparallel array, the other strings will receive too-high current levels. Initiating current flow can be difficult if the plates are heavily sulfated and/or the fill-to-form time has been long, thus depleting most of the sulfuric acid in the electrolyte and raising the liquid resistivity. The use of red lead in the positive paste, carbon in the negative paste, as well as sodium sulfate in the fill electrolyte will help to minimize the high initial resistance. If a high-inrush current level is applied when the initial battery resistance is high, the voltage will be driven to high values and most or all of the current will go into heat generation and gassing. After a period of time these processes will diminish and conversion of lead sulfate to the active materials will take over and the formation will proceed in a normal fashion. However, a short initiation charge period of 0.5 to one hour at low currents can be applied in order to generate some acid and improve the conductivity in the plates, or the current can be ramped up slowly over an hour or so, as shown in Figure 1 before the main formation current is applied. Figure 1, Initial resistance to proper current flow can be detected either by an immediate rise in charge voltage to very high levels (or to the voltage limit if constant-voltage charging is used) or by a relatively sharp temperature increase. Unless the battery plates are very heavily sulfated, after a short period the voltage and temperature will drop to normal levels, which are temperatures below ~50oC and voltages of ~1.8-1.9 volts per cell. There will then be gradual rises in both temperature and voltage, but because almost all of the formation current is going into lead sulfate conversion these increases will be very gradual.
Figure 1
Examples Of Techniques For The Initiation Of Formation Charging
A. Low-current initiation
Current, amperes
Formation time, hours
B. Ramp-current initiation
Current, amperes
C. Abrupt or high-current initiation

Current, amperes
3.8 Constant-Voltage Charging

Current-limited constant-voltage (CV) charging is commonly used for cyclic charging of VRLA batteries, but its utility in formation is more limited, largely due to cost and its effect on product uniformity. Precise voltagecontrol limits are expensive in terms of formation electronics and in forming VRLA batteries they are not necessary. The primary drawback is that, as in CV recharge, the current taper toward the end of formation results in relatively long charge times. In order to minimize this and speed up formation times, multi-step CV algorithms can be used, by programming for currentlimit reductions when the voltage limit is reached. This then becomes a stepped constant-current formation, but with a voltage limit (usually ~16V for a 12V battery) to minimize gassing and grid corrosion. Typical examples of single-step CV and stepped-CV/CC algorithms are shown in Figure 2. The last step usually allows for a Figure current taper when the voltage limit is reached, the duration depending upon the desired formation time. There are a number of advantages and drawbacks to CV charging. On the plus side, overcharge is minimized due to the current taper during the finish of formation and so the charge efficiency is relatively high and concerns about gas monitoring and ventilation are less important. Balanced against this (in addition to cost) are a number of drawbacks:
s
Figure 2
Examples Of Single- And Multi-Step Current-Limited Constant-Voltage Formation Profiles
A. Single-Step Current-Limited Constant-Voltage Formation Profile
Voltage/ Current
Formation time
B. Multi-Step Current-Limited Constant-Voltage Formation Profile
Voltage/ Current
Formation time
= Formation Charge Current = Formation Charge Voltage
With a significant time in the current-taper mode, the total Ah input must be integrated electronically (rather than simply timed as with CC formation). In single-step CV, the long charge tail lengthens formation time significantly. Because voltage is applied to long strings or series-parallel arrays as a multiple of a given volts-per-cell, actual charge voltages for each cell can be highly variable. More seriously, paralleled strings can draw different currents based upon their cumulative DC resistances; this can have the effect of routing high currents through individual strings early in formation, which can result in large imbalances of total ampere-hours passed through different strings. In the extreme, this can result in strings with low initial DC resistance going into thermal runaway, particularly for large batteries with poor heat-dissipation capabilities. If strict voltage control is desired, temperaturecompensated charging must be used, which further increases cost. Temperature issues can be minimized by using lower charge voltages, but this will increase formation times significantly.
3.9 Constant-Current Charging

In CC charging, voltage control is not required (although there is always a voltage limit such as 2.80 volts per cell) and this reduces the cost of the charging equipment. Using single-step or two-step formations at high currents can also reduce formation times, but this results in lower charge efficiencies and large amounts of overcharge and gassing. The most common approach is to use a single-step CC algorithm, possibly with one or more rest periods (see below) or discharges. This is not efficient, since at low currents overcharge is minimized but total formation time is long whereas with high currents the forming time is shortened but the overall charge efficiency is reduced. More innovative, multi-step algorithms are now in use where relatively high currents are used early in formation and lower finishing currents are then applied, either as a two-step or multi-step algorithm. Dramatic gains in charging efficiency can be realized, as shown conceptually in Figure 3 [2]. In some Figure cases this is done as a fixed, programmed algorithm with defined current levels for pre-set time intervals. Other approaches involve monitoring of battery
parameters in order to apply optimal current levels for as long as possible. One example of this is shown in Figure [3], Figure 4 [3] where battery temperature is used as the control variable. As can be seen, this allows for an initial high CC level, followed by step-downs to lower currents based upon battery temperatures. The major advantages of CC charging are that it is easily programmable, it is relatively rapid and the total ampere-hour input can be determined easily. In addition, the current level is controlled, so even in series-parallel arrays battery damage due to high charge currents as noted above for CV formation is largely avoided. However, several drawbacks also apply:
s
Figure 3
Examples Of Stepped Constant-Current And Conventional CC Formation Profiles Compared To An Ideal Formation Curve.
Note The Disparity In Overcharge Amounts
Stepped Constant-Current
I
Ideal current curve Several-step current curve
Single-step CC formation is either very lengthy (low current) or very overcharge-intensive (high current). Heavy overcharge results in high heat production, grid corrosion and gassing. Voltage regulation on charge is not possible, except for the high upper limits used (2.8-3.0 V/ cell or more)
Current curve of several-step formation
Conventional CC Formation
Ideal current curve Several-step current curve
On balance, this is the simplest approach to formation and is the most commonly used, particularly in multistep algorithms. As programmable controllers for formation systems are now commonly available and inexpensive these approaches, though seemingly complex as shown in Figure 4 are very straightforward. Figure 4, They are also a more tolerant approach when poorly regulated input (i.e., dirty) power and/or inexpensive formation electronics are used.
Current curve of conventional formation
3.10 Taper-Current Charging

Taper-current (TC) charging for formation combines some of the best aspects of CV and CC approaches and is probably the least expensive of the three. As it is not a common approach, Figure 5 shows a typical circuit Figure for TC charging, along with a typical charging curve. This is really the simplest of circuits. A power supply is wired in series with a load resistor and a battery string or strings to be formed. If desired, some form of sensing of battery parameters (voltage, temperature, etc.) can be included to provide feedback control. When formation is initiated, current flows according to the rating of the load resistor and the voltage difference between the power supply setting (typically a high value of 2.6-2.8 V/cell) and the battery array (which will initially have a very low voltage). At the beginning of formation, the voltage difference is great, on the order of ~1V/cell, and the inrush current is relatively high, as in CV charging. As the cumulative battery array voltage climbs the formation current decreases because of the decreasing voltage difference with the power-supply setting. Unlike CC charging, when the battery array voltages climb into the gassing region the charge current is tapering off. However, the current does not taper off as sharply as with CV charging because of the higher voltage setting. Moreover, if the TOCV is low for the formation array due to an unusually high oxygen-recombination current draw the current at the end of formation will increase due to the widening gap between the power supply voltage, which is fixed, and the decreasing end-of-charge voltage. Initial and final currents are roughly set by choices of power-supply settings and resistor values. These yield an approximate formation voltage/time profile, but the exact shape of the curve will vary considerably depending upon the charge-acceptance properties of the battery array being formed. This can be viewed as both a strength and a weakness of this
www.firing-circuits.com 9
Figure 4
Examples Of The Progressive Influence Of Temperature Monitoring On The Efficiency Of The Formation Process.
A. Normal formation. The batteries are placed in a non-flowing water pool.
approach. It is well suited to VRLA formations, as many products do not require precise voltage control. In summary, TC charging has the following advantages:
s
Capital input for formation charging equipment is low if inexpensive power supplies are used; however, poorly regulated supplies can not only transmit, but also at times amplify, line fluctuations to the charging system. It allows for a relatively fast formation with only moderate overcharge amounts by allowing high inrush currents (relative to CC) as well as high finishing currents (relative to CV). To some extent, the formation profile is variable according to the charge-acceptance characteristics of the battery array being formed.
s
Temperature degree C Voltage V Current A
Time
There are also a number of drawbacks, as follows:

s
B. Formation in flowing coolant, with maximum initial current. Limit set on voltage but not on temperature.
Because the current tapers, Ah input must be determined by electronic integration. The amount of overcharge is high relative to CV or multi-step CC charging. Voltage and current are not controlled, so the uniformity of formation profiles for different battery lots is not high; total Ah input may vary significantly with product uniformity going into formation. The use of unregulated power supplies can result in shortened lifetimes in service.
s
Temperature degree C Voltage V Current A
Time
3.11 Pulsed-Current Charging

Pulsed-charge algorithms can be applied to the formation of VRLA batteries just as it is used in charge/ discharge service; typical algorithms are shown in Figure 6. Figure 6 As can be seen, profiles analogous to CV or TC charging as well as pulsed CC can be used. Note that in the off periods, rests or partial or complete discharges can be applied. The discharges are thought to be beneficial in eliminating surface charges from the plates, which can result in lower gassing levels; it has not been unequivocally established if this is, indeed the case. A good deal of work has been done on using pulsed methods, but it remains unclear whether product quality gains can be realized with this approach. There are clear advantages in enhanced heat dissipation while allowing the use of relatively high currents (even late in formation) and in reduced gassing due to the reductions in coulombic input per pulse as the gassing region is approached late in formation. While most battery companies have investigated this for the above reasons, it is not commonly used.
C. Formation with initial maximum current limit, followed by temperature and voltage limits.
Temperature degree C
Voltage V
Current A
Time
= Current = Voltage = Battery Temperature = Coolant Temperature
10
Figure 5
Taper Current Charging.
Typical TC Circuit
Switch
Power Supply
V
Sense
RTaper
Charge current ( I ) =
Power Supply Voltage - Cell Voltage (V) Load Resistor ( RTaper )
Typical TC Current/Voltage/Ah Curves
I
Charge Current Charge Voltage
E Ah
Ampere-Hour Input
plate pores. The charge efficiency of the positive electrode is relatively low even when completely formed, but in formation itself it is so poor that gassing of oxygen can begin after only a few hours, or even less. Later, the negative plate will also begin gassing and in both cases this hampers proper conversion of unreacted lead oxides deep in the plates to lead sulfate and then subsequent reaction of the sulfate to the active materials. The first step requires acid generated at the plate surfaces early in formation to penetrate into the plate interiors and the second reaction requires water to produce sponge lead and PbO2. When either or both of the plates goes into gassing, this will force liquids out of the plate pores and into the glass-mat separator; eventually, with heavy gassing much of the electrolyte will be forced into the head space or even out of the battery as regurgitated acid or acid spray. These conditions are easily avoided by inserting one or more rest periods or discharges into the formation algorithm. In both cases, when the charge voltage is removed gassing ceases and time is allowed for water and acid to diffuse into the plate interiors. This allows for reaction of acid with any PbO remaining after filling and fill-to-form. When formation is reinitiated more lead sulfate has been generated and water is present as a part of the filling reaction. When formation is continuous gassing seriously impedes these processes. Thus, use of significant off times can actually result in faster, more complete formation processes. Rests or discharges can be put in at fixed points in formation or they can be initiated when a trigger voltage is reached. These considerations probably apply more to thicker-plate
Time
3.12 Rests and Discharges

One of the major electrochemical problems in using any of the above approaches in a continuous way (with the exception of pulsed charging) is that gas generation can severely impede the efficiency of the formation process by retarding the diffusion of acid and water within the
Figure 6
Examples Of Pulsed Charging Algorithms.
In all cases, the coulombic input decreases as the top-of-charge is approached.
Current
0
s s s
Time
Current
Time
Constant Period, Decreasing Amplitude

Pulse/Rest Pulse/Discharge Pulse/Discharge/Rest
Constant Amplitude, Decreasing Period
11
products (2.0 mm or more) than those with thin plates (where diffusion paths are shorter and plate wetting is more efficient due to the higher surface areas). Which approach is better? Discharges are clearly more complex in terms of capital equipment and they will lengthen formation time relative to rests due to the requirement for replacing charge taken out during the discharge. Discharges are thought to be beneficial because, in principle, they should increase the porosities of the plates and further aid acid and water penetration, as well as improve post-formation discharge capacities. Little documentation is available comparing the effects of rest periods and discharges, so the technologist is left to weigh the possible benefits given above against the significantly higher costs of discharge equipment. Both are clearly beneficial in reducing formation weight losses and in improving finished-product quality. For VRLA batteries requiring high post-formation PbO2 levels (90% or greater) and long shelf lives (low residual PbO levels in the formed positive plates) the use of one or the other is almost mandatory.
Figure 7
Typical Constant-Current Formation Profiles For A 12V/20Ah VRLA Battery.
A. Voltage, Temperature and Gassing Curves for a One-Step CC Formation
Voltage, Temperature, Gassing Rate
0 4 8 12 16 20 24 Formation Time, Hours 28 32 36 0 2 4 8 12 16 20 24 Formation Time, Hours 28 32 36
= Formation Current = Voltage = Temperature = Gassing Rate
Formation Current, Amperes
B. Same Curves for a High-Inrush Two-Step CC Formation Algorithm

8
3.13 Sample Formation Algorithms & Profiles

3.13.1 A Simple Algorithm Sample formation profiles will now be considered that might be recommended for a typical 12V/20Ah VRLA product. The simplest approach would be a singlestage CC formation over, for example, 36 hours with a total Ah input of four times the rated capacity, or 80Ah. Over a 36-hour period this would be a CC level of ~2.2A, as shown in Figure 7a. This approach results in Figure relatively high temperatures toward the end of formation and large overcharge amounts and gassing levels, but it will form the battery successfully. The pore structure may not be optimal due to the low initial current and so a modification of this would be to use a two-step CC algorithm with, say, 2 hours at 8A (16Ah) followed by 34 hours at 1.88A. (Figure 7b). For a CV (Figure formation, somewhat more time may be required or a high inrush current may be needed, accepting a somewhat lower charge input at 36 hours, as shown in Figure 8a. In order to increase the charge input toward Figure the end of formation a taper-charge algorithm may be used, as shown in Figure 8b. This has a high inrush Figure current as with CV but the current only tapers to ~30% of its initial value. This results in a higher Ah input, but also higher temperatures and more gassing (weight loss) in the final 12 hours or so.
More Typical 3.13.2 More Typical Charge/Rest/Charge Algorithms An intermediate level of complexity can be applied without going to the type of feedback approach used in Figure Figure 3 (which is certainly acceptable). In this typical case, two rest periods are introduced into the 36-hour formation with the focus on CC charging. The rest periods can be shifted toward the end of formation as there is, initially, a great deal of lead sulfate from the filling process and it will take some time to consume this material. (Given the current level imposed and the amount of lead sulfate expected to be present, a rough time period can be calculated to where significant
12
Voltage, Temperature, Gassing Rate
Formation Current, Amperes
Figure 8
Typical Constant-Voltage And Taper-Current Formation Profiles for a 12V/20Ah VRLA Battery
A. Voltage, Temperature and Gassing Curves for a One-Step CV Formation
Voltage, Temperature, Gassing Rate Formation Current, Amperes
during the rest of the formation. After this, a fixed CC level can be used in combination with the two rest periods, as shown in Figure 9a. The rest periods are Figure beneficial not only in providing time for electrolyte penetration but also for keeping the temperature down compared to a continuous one- or two-step CC algorithm. Because the time spent in overcharge and resultant gassing is lower overall with rest periods (even though the charging current is higher to compensate for the 4-hour off time), weight losses are also reduced somewhat. If a discharge were to be used instead of the two rest periods, it would be most beneficial to have it near the end of the formation, as shown in Figure 9b. As can be Figure seen, only a partial discharge is carried out; a complete discharge would obviously be more effective in promoting pore formation and electrolyte penetration, but it would also require substantially more time for the full discharge and subsequent recharge. If this were done within the 36-hour schedule time it would require much higher charge current levels but it could be done. However, as noted, there is no clear evidence indicating that a discharge is more effective than rest periods. One advantage for a discharge is that it could be used as a matching tool for building battery modules into high-voltage packages, using discharge capacity and top-of-recharge data recorded during formation. This would, of course, require that all batteries be monitored and that the data be collected and processed. The major cost, however, would be for equipment to carry out the discharges; in addition, if the formation time were extended this would reduce the battery throughput level somewhat and would require more formation stations to process the same number of batteries. However, finished battery quality and uniformity would be improved significantly. These are just a few examples of formation algorithms that might be employed for the processing of VRLA batteries. The great flexibility in choosing an appropriate algorithm also introduces an equivalent amount of uncertainty. It is recommended that for a given VRLA product significant R&D work be put into the definition of suitable formation conditions. While each company has its own approaches, the following is a recommended procedure that should work for the majority of companies.
12 16 20 24 Formation Time, Hours
28
32..........48
B. Same Curves for a One-Step Taper-Current Formation Algorithm

Voltage, Temperature, Gassing Rate Formation Current, Amperes
0 2
28
32
36
= Current = Voltage = Temperature = Gassing Rate
gassing will begin; alternatively, voltage can be monitored, as shown here, and trigger levels used to start the two rest periods). Four hours total have been allocated for rest periods; this could be put into one or two rests; but it makes more sense to use two. More rest periods and longer total rest times may also be suitable for some VRLA thick-plate products. In order to have a relatively fine pore structure in the positive active material, a short, high-inrush current period has been used to provide smaller, more numerous PbO2 seed crystals upon which to build
13
Figure 9
Typical Constant-Current Formation Profiles With Rests Or A Discharge For A 12V/20Ah VRLA Battery.
A. Current and Voltage Curves for a CC/Rest Formation Algorithm
8 Formation Current, Amperes Formation Voltage, Volts
Weigh the batteries prior to formation, but at the end of the fill-to-form period. Dont assume that the electrolyte fill weight can be added to the prefill battery weight in order to get the pre-formation weight; all batteries, especially large ones that are processed open to the atmosphere (i.e., without the top and vent valve in place), will lose weight between filling and formation due to evaporation and, in some cases, acid spraying or regurgitation. The primary loss is from evaporation, which can be several percent of the total fill weight. To as great a degree as possible, batteries should be configured as they will be in a formation bay in manufacturing. Formation of single units or a few in series when they will be in large series-parallel arrays in production will not give an accurate idea of the effectiveness of the formation process. In addition, thermal conditions should be close to those that will be seen by the batteries in manufacturing. Initial studies can be done with forming of small numbers of batteries, but it should not be assumed that product quality would be the same as in full-scale manufacturing. Wire up the test batteries so that the following parameters can be monitored: voltage, time, current (also with integration if CV or TC charging is used, but also for CC to ensure that the correct Ah input is applied) and internal pressure (if batteries are formed sealed). Reference-electrode measurements should also be taken and at some point several batteries should be formed with gas collection and analysis being done. In order to get an idea of the capabilities of the battery design for formation, an initial run should be done using a very simple one-step CC, CV and/or TC charge, just to see how the battery reacts to these baseline conditions. Then, several preferred algorithms should be applied, covering a range of times and currents, using rest periods and, possibly, discharges (even if this is not to be done in manufacturing due to cost).
2 4
12
32
36
B. Same Curves for a CC/Discharge/Recharge Formation Algorithm

8 Formation Current, Amperes Formation Voltage, Volts Charge
2 0
Discharge
-8 0 2 4 8 12 16 20 24 28 Formation Time, Hours 32 36
3.14 Development of a Suitable Formation Algorithm

It is assumed that a 6V/100Ah VRLA battery is being developed for Telecom use and it is necessary to find out how to form the product most effectively. The optimized formation algorithm will depend largely upon the desired formation time, the design of the battery and the user requirements. Without going into such details, the following steps can be used to define a suitable algorithm.
s
Take at least 12 filled modules and monitor temperature between fill and formation; note the battery temperatures at the initiation of formation.
14
After formation, batteries should be weighed and carefully inspected for cosmetic and product defects (acid spray or leakage at lid/box seals or terminal posts, label damage, etc.). Teardowns should be carried out to look at the plates in detail (visual inspection for white sulfate, color, hardness (PAM) or softness (NAM), distortion, massive corrosion or growth). The separator should be inspected for holes/tears, damage from filling, staining by expander or paste and the possible presence of lead sulfate (hydration shorts). The latter can be determined using a sodium iodide solution sprayed on the separator; insoluble lead dioxide shows up as a bright yellow precipitate. Electrolyte should also be squeezed out of the separator at several points to determine specific gravity levels. Negative plates should be dried and prepared for SEM, BET and porosimetry analysis; other tests may also be carried out. Positive active material should be treated similarly. In addition, several positive plates should be stripped of active material and the grids should be inspected and weighed for general or localized corrosion during formation. Wet-chemical analysis of the NAM (free lead, sulfate) and PAM (PbO2, sulfate, unformed PbO) should also be done. XRD should be applied, if available, to define the amounts of alpha- and beta-PbO2 generated at different locations on the positive plate surfaces. Taking all of the data above, several iterations of formation algorithms should be applied to ensure that the most effective algorithm has been developed. As a final step, a pre-production run should be carried out under actual manufacturing conditions to ensure that the development work done on a limited number of batteries (particularly the thermal conditions and the series-parallel configurations) is relevant to full-scale production. In addition to the above analytical work, full electrochemical characterization of the formed batteries should be done to ensure that nominal quality levels and the desired uniformity have been achieved using the selected formation algorithm. Self-discharge (shelf life) measurements should also be done to ensure that the degree of formation of the positive plate and remaining unformed oxide amounts are acceptable.
4. Temperature Limits For VRLA Jar Formation

For jar formation of conventional flooded batteries, a maximum formation temperature of up to 65C may be permitted with no apparent harmful effect on the battery performance: this is certainly the case for SLI battery designs. Industrial battery designs may have significantly longer formation times and lower recommended maximum formation temperatures (e.g. 50C). The temperature during all stages of the filling and formation process is much more critical for VRLA jar formation. The control of temperature is necessary from the initiation of formation until its completion. Sometimes it involves active control and at other times it dictates passive processing conditions. The latter is true going into formation, where the battery has been filled with electrolyte and allowed to stand for some time before being placed in the formation environment. With VRLA batteries, high formation temperatures may result in the formation of lead dendrites and/ or hydration shorts. Therefore the maximum formation temperature should be kept below 40C: and normally this will require water cooling or forced air-cooling. The formation regime may also include brief rest periods. Some VRLA battery manufacturers may specify a maximum temperature of 50C or even 60C, but there are risks associated with this approach. In comparing formation at 60C with formation at 40C, it has been found that the PbO2 content is higher at 60C, and the a/b PbO2 ratio is lower. However, the higher temperature has an adverse effect on the negative plates, resulting in a decrease in battery capacity at high discharge rates. The surface area of the negative plates is decreased if formation is carried out at high temperature, possibly because of deterioration of the negative plate expander [4]. It is important to note that if the measured temperature at the top of the cell is 60C, the maximum internal temperature inside the cell may be significantly higher, 70C or even as high as 80C. This has implications in respect of the stability of the negative plate expander, and it has been found that the surface area of the negative plates is significantly reduced. Localized overheating may also result in grid corrosion and/ or increased risk of lead dendrite formation.
15
In practice, sufficient time must be allowed after filling and before the start of formation to allow the heat generated during the filling process to have passed its peak. The thermal management during the filling process should not be too efficient or the exothermic acid-oxide reaction may shut down if the battery is too cold, and start up again generating excessive heat when the formation process is started. The degree of cooling (or even heating) during formation needs to be related to a number of factors including:
s s s s s
Product size Temperature at start of formation Cooling technique Plant temperature Sealed or open formation
A critical issue with plate formation is to ensure that the plates do not bow during formation. This is a particular concern with positive plates if they have been over pasted on one side. A high current density should also be avoided. If the plates are bowed, the plate group pressure will be non-uniform when the plate group is assembled which will cause other problems including a reduction in battery life. Because VRLA batteries need to be assembled with a controlled plate group pressure to ensure long battery life, bowed plates are a far more serious problem with VRLA batteries than with conventional flooded batteries. Partly because of this issue concerning bowed plates, the maximum temperature for plate formation should be kept below 40C. A higher maximum temperature is unlikely to have an adverse effect on plate performance, but there is a greater risk that the plates will bow during the formation process. A high current density should also be avoided.
Additional information is given in section 7.3.
5. VRLA battery manufacture using PLATE FORMATION

In plate formation the pasted and cured positive and negative plates are placed in bottomless slotted crates with relatively wide spacing so that no separator is needed. The positive and negative plates are placed alternately so that the lugs of all the positive plates are on one side of the crate, and all the negative plate lugs are on the other side of the crate. Tacked or tackless formation can be used. With tacked formation lead bars are tacked onto the lugs joining all the positive plates together, and all the negative plates together, to form a 2v cell. In tackless or burnless formation, the plate lugs make contact with lead bars wedged between the walls of the crate and the plate lugs: a special clamp ensures close contact between the plate lugs and the lead bar. Or the plate lugs make contact with lead bars at the bottom of the crate. The crates are immersed in dilute sulfuric acid (e.g. 1.100 s.g.) and a formation charge passed through the crates: a stepped current may be used to maximize formation efficiency. Because the sulfuric acid is present in excess, there is rarely any problem with excessive formation temperatures. After formation, the plates are washed and dried to produce dry charged plates. (Oxygen needs to be excluded during the drying of the negative plates). This stage of the process is exactly the same for VRLA batteries as it is for conventional flooded batteries. The formation regime, total Ah input, and dry charge process can be exactly the same as for conventional flooded batteries.
16
Technical And Theoretical Background
6. Technical And Theoretical Background

6.1 The formation process explained
The chemistry that takes place during formation has a lot to do with the performance and lifetime of the VRLA battery in service. It also has a large impact on how batteries can be processed, particularly in the times required for proper formation. The chemistry taking place during formation can be characterized on a simple level as follows. [Theoretical treatments can be found in textbooks written by Hans Bode (Lead-Acid Batteries) and Dietrich Berndt (Maintenance-Free Batteries, 2nd Edition) ]. The purpose of the formation process is to convert the pasted plates to lead dioxide at the positive and lead at the negative. After pasting, both positive and negative plates have essentially the same composition, except that the negative plates contain additional non-leady additives (negative plate expanders and floc). The chemical composition of both the positive and negative plates after pasting and curing is essentially lead monoxide (PbO), monobasic lead sulfate (PbO.PbSO4), and tribasic lead sulfate (3PbO.PbSO4) [TRB]. The positive plate may also contain some tetrabasic lead sulfate (4PbO.PbSO4) [TTB]. The relative proportions of TRB and TTB are important for formation because TRB pastes form much more easily than TTB pastes. TTB is formed during the plate curing process at temperatures of ~ 70C or above. TTB can also be unintentionally created during the filling process if the internal battery temperature is at or above 70C for an appreciable amount of time. When the unformed plate is immersed in lead sulfate, the acid reacts with the lead monoxide and basic lead sulphates as shown below: PbO + H2SO4 PbO.PbSO4 + H2SO4 3PbO.PbSO4 + 3H2SO4 4PbO.PbSO4 + 4H2SO4 PbSO4 + H2O 2PbSO4 + H2O 4PbSO4 + 3H2O 5PbSO4 + 4H2O
positive paste and/or 2-5% of graphitic carbon in the negative paste provides some conductivity to aid with current flow. Sodium sulfate in the electrolyte (typically 1.5% by weight of electrolyte) is also useful in improving conductivity. Some manufacturing processes also involve having a small amount of sodium sulfate in the pastes themselves. All of these measures are more important in thicker-plate VRLA designs (plate thicknesses and plate spacings >~2.0 mm), as nominal resistance values are higher in such products. Even with the above materials being present, there is a strong initial resistance to current flow. Because of this, formation algorithms often start with a short, low-current step; immediate use of higher currents can result in heavy gassing from water decomposition due to the high voltage needed to overcome the high battery resistance. When the electrical current is switched on, the electrochemical reactions which take place convert the lead oxide and lead sulfate to lead dioxide (PbO2) at the anode (positive plate) and to lead (Pb) at the cathode (negative plate). In some products, formation is from the grid strands out toward the plate surfaces and in others it is the opposite. However, in both cases the lead sulfate-active material conversion also results in the production of sulfuric acid from the water present in the plate pores and separator. This creates a strongly conducting environment in the plate stack so that higher currents can be applied at low voltages. 2PbSO4 + 2H2O 2PbO Pb + PbO2 + 2H2SO4 Pb +PbO2
Sulfuric acid is regenerated during the electrochemical reactions, and because some sulfate was present in the basic lead sulphates of the cured plates, the final acid density at the end of formation is higher than that of the filling acid. For a significant period of time, the conversion of lead sulfate to active materials proceeds with very high efficiency, close to 100% in some cases where the rate of water diffusion into the plate pores is high enough to keep up with the current flow. This is a critical stage in formation, as the basic pore structure of the plates is established as the chemistry proceeds. High-inrush currents (following a short, low-current step to promote current flow) are felt to be useful because they create more seed crystals in the positive paste and a higherporosity lead dioxide structure is created; seed crystal formation is also aided by the use of red lead as
A significant amount of heat is generated in these reactions, and the higher the initial acid density, the greater the heat that is generated. Sulfuric acid is consumed in the reactions, so there is also a reduction in acid density. The pastes are largely converted to neutral lead sulfate on their surfaces. These conditions increase the resistance of the unformed battery to current flow. The use of 10% or more of red lead in the
17
described above (the red lead formula is Pb3O4 and each molecule contains one molecule of PbO2). The conversion efficiency is greater at the negative plate than at the positive plate. Negative plates form relatively easily and it is almost always the positive electrode whose formation efficiency is poorer. Thus, at some point relatively early in formation the positive charge efficiency will decrease and gassing will begin. If individual plate potentials are monitored during formation, it will be observed that there are two clearly defined potentials for the negative plate. At the higher potential (+0.1 to +0.2v with respect to a cadmium reference electrode), lead oxide and lead sulfate are being reduced to lead. As the conversion process nears completion, there is a rapid change of potential to about 0.2/-0.3v, at which potential the evolution of hydrogen gas by electrolysis of water becomes the predominant reaction. However, for the positive plate, the difference between the potential at which the lead dioxide is being formed and the potential at which competing gassing reactions occur is less clearly defined. The plate potential is always very close to that at which the electrolysis of water can occur. Therefore the formation process is inherently less efficient than at the negative plate. The positive plate potential (with respect to a cadmium reference electrode) at the end of formation is about 2.35-2.4v. Because the plate interiors are only partially wetted by electrolyte during the filling and fill-to-form processes, the complete conversion of pastes to active materials via lead sulfate formation depends upon continuous penetration of the acid generated by electrolysis into the plate interiors. When heavy gassing in the late stages of formation physically expels acid from the plate pores this continued diffusion is limited. For this reason, rests or discharges are desirable in VRLA formations, particularly those carried out in an acid-starved condition. The overcharge processes result in the loss of oxygen and hydrogen gases in a 1:2 ratio equivalent to a given amount of water. This concentrates the initial electrolyte strength and reduces the liquid volume. A moderate water loss is unavoidable, but if this loss is substantial additions of water or electrolyte following formation must be made.
6.2 Formation processes and Ah input

(Additional information about lead acid battery formation of conventional flooded batteries is given in the Digatron/ Firing Circuits brochure Lead Acid Battery Formation Techniques by Dr. Reiner Kiessling). For conventional flooded batteries, the choice of formation processes is as follows: 1. Plate formation. See Section 5 for details. 2. Two shot jar formation (also known as box formation). The battery containing dry unformed plates is assembled into the final container. The battery is filled with dilute sulfuric acid (e.g. 1.120 s.g.) and subjected to a formation charge: this can be a stepped charge to control the maximum temperature (smooth out temperature peaks) and maximize formation efficiency. After formation is complete, the acid is tipped out, and replaced with a higher density acid. Because some of the low density acid is retained in the plates and separators, the density of the acid for refilling may need to be as high as 1.350, to achieve a final density of 1.280. A short equalizing charge for 2 hours at a low current is desirable in order to mix the acid before the battery is finished and dispatched. Because the acid is more restricted in jar formation than in plate formation, additional care needs to be taken to avoid excessive formation temperatures. 3. One shot jar formation. This is now generally preferred over two shot jar formation. The battery is filled with acid of density such that the density after formation is the correct density for battery dispatch without the need for further acid adjustment. The required filling acid density will depend on the battery design (e.g. interplate spacing, active material to acid ratio etc.): for example typically 1.230 to achieve a final acid density of 1.280. Extra care needs to be taken with one-shot formation compared with two-shot formation. The higher density filling acid results in a high battery temperature during the first hour after filling, because of the vigor of the exothermic reaction between the acid and the active materials. The formation regime needs to be designed carefully: possibly including one or more rest periods so that the peak temperature does not exceed 65C. Also, a higher ampere hour input may be required than for two-shot formation because the efficiency of conversion of the positive active material to lead dioxide is poorer the higher the density of the formation acid.
18
Typical ampere-hour inputs in relation to wet paste weight and dry cured paste weight are given in Table 1 below. These figures should be used for guidance only as they will be influenced by a number of factors in relation to the plate and battery design and formation regime.
Table 1
Theoretical Plate formation (1.100 s.g.) 2-shot jar formation (1.13 s.g.) 1-shot jar formation (1.23 s.g.) VRLA jar formation Ah/kg wet paste weight 200 275 300 350 350+ Ah/kg dry cured weight 226 310 340 400 400+
The valve-regulated battery is designed so that any gases generated during charge are recombined within the battery. Each cell contains a self-sealing valve that vents gases to atmosphere if the pressure within the cell rises above a preset limit. So the cell/ battery is not hermetically sealed but is valve regulated. The separator (normally of microfine glass) completely fills the space between the electrodes. The sulfuric acid is contained within the pores of the plates and the separators and there is no free acid. The separator is not quite fully saturated with acid (e.g. 95% saturation), so that any oxygen gas generated from the positive plate is able to pass through unfilled pores in the separator and recombine with the active lead surface of the negative plate. As a result, the VRLA battery is unspillable, maintenance free throughout its design life, and can be operated in any orientation. In the jar formation of VRLA batteries, the cell or battery is not normally sealed until the formation process is complete and the separator saturation level is deemed to be correct. Therefore in open formation (section 3.3) the recombination process is not an issue. However, because the separator completely fills the space between the plates, the cell design and the properties of the separator have a critical influence on the acid filling and formation process. Since there is also less acid in a VRLA cell than in a conventional flooded cell, thermal effects are also more important. (Sealed formation is dealt with in section 7.1). Cell/ battery reproducibility & variability is also much more of an issue with VRLA batteries than conventional flooded batteries [5]. To make cells as uniform as possible in the manufacturing process, electrolyte amounts are accurately metered into the cell/ battery elements in the filling operation. However, normal tolerances in upstream processes & materials e.g. grid casting, pasting, separator materials may result in plate groups with variable amounts of void space within fixed case dimensions. As a result, the filled and formed cells may have slightly different void volumes. In subsequent duty, the cells may behave slightly differently during the charging process due to the small differences in the available void volume, and cell-to-cell differences in the compressed separator structure.
For guidance, the Ah input in the above table for VRLA jar formation is approximately equivalent to 4x the 5 hour rate capacity, dependent on the battery design and active material density. It is possible that certain VRLA battery designs may require higher Ah input than given in the table above, and extended formation times. However, any increase in Ah input should be minimized by experimentation to establish the optimum formation regime (e.g. by including brief rest periods or even a brief discharge partway through formation). If choosing jar formation, the battery manufacturer will normally choose 1-shot rather than 2-shot formation. This is in spite of the fact that VRLA batteries tend to be specified with a higher final acid s.g. which therefore requires a higher filling acid s.g, resulting in greater heat generation during filling. However, on balance there is no particular benefit in using a 2-shot jar formation process, because of the difficulty of controlling acid volumes and final acid density. (Section 3.13 has already given sample formation algorithms and profiles for VRLA jar formation).
6.3 Key Differences Between Flooded and VRLA Batteries

The comments below relate to VRLA batteries containing special separators generically known as Recombinant Battery Separator Mats (RBSM). These are typically glass separators also referred to as Absorptive Glass Mat (AGM) or Microfine Glass (MFG). However, other separator types may also be used, for example containing a blend of glass and polymeric fibers. A discussion of gel VRLA batteries will follow later (section 11).
19
Additional information about Jar formation
7. Jar Formation Additional Information

7.1 Battery Preparation for Formation Sealed Formation
The first part of this brochure (section 3) dealt with open formation, which is the more popular and widely used technique. It is also possible to have the battery acid-starved and fully sealed during formation. This will depend upon the battery design and size and the type of thermal management contemplated. For efficiency of manufacture and to minimize handling it is preferable to form a product in the starved state fully sealed. Thus, following formation the battery can be checked for quality, be electrically tested and, if passed, be shipped to the customer with minimal handling. However, there are stringent requirements on the design and processing of a battery subjected to sealed formation. In order to be able to form a battery sealed, the processing must be such that little or no carbonation of the plates has occurred (formation of lead carbonate by reaction of the lead oxide in the paste with CO2 from the atmosphere); carbon dioxide liberated as the plates form can cause expulsion of acid through the vent valve in the form of acid spray. The battery must also be filled to a starved condition with no more than ~95% saturation. Forming sealed, weight losses are minimal (~5% of the fill weight) so that there is only a small increase in specific gravity relative to the nominal level. Because of the low gassing levels, heat dissipation from gassing is low, so for large batteries more care must be given to keeping the battery in the prescribed temperature range. If this is not done, acid spray may again result due to elevated temperatures in the battery during the gassing phase of formation. In practice, there are very few VRLA products that are formed sealed, due to the above issues. While it is desirable and can be done, the cost penalties must be weighed against the ease of processing during and after formation. As noted above, it is relatively difficult to design and process VRLA products so that they can be formed sealed. Without adding electrolyte after formation, it is difficult to achieve a roughly 95% finished saturation level without having acid leakage and spray during formation. The starting saturation needs to be at 9798%, so unless the battery has a large headspace there is likely to be physical loss of acid. This may also be the case if a low surface area or a hybrid glass/organic fiber separator is used in the battery design. Such separators do not hold electrolyte as well as a high-surface-area
AGM material and thus the headspace is easily flooded. In order to handle the acid that is forced into the head space during the gassing phase of formation, some manufacturers use devices fitted into the filling port to take up the expelled electrolyte; when the formation current is reduced or terminated the acid can then flow back into the battery and be retained. This approach also minimizes gassing water loss because when the overflow device has taken up acid the plate stack and separator have enough void space to allow for a significant amount of oxygen recombination. Because an accurately measured amount of electrolyte can be added at filling, the final saturation level can also be known precisely from the materials amounts and the formation weight loss. The external device to accommodate free acid during formation returns the acid to the battery when gassing is completed; the reabsorbed electrolyte amount then results in an accurately known saturation level. Following formation the vent valve is put in place and the assembly of the battery is completed. 7.1.1 Plate Curing and Carbonation A detailed discussion of plate curing is outside the scope of this document. However, if sealed formation is proposed, certain precautions need to be taken to minimize carbonation. This is also important if high vacuum filling is used. The reaction of CO2 with the plate pastes is actually more likely to occur not during the curing process itself but after drying, when the plates are taken out of the curing ovens. To minimize carbonation, plates should be cooled down in a dry environment when removed from the ovens (~ 20C and ~ 10% RH). Plates should also be used as soon as possible after drying. The use of a desiccation system on the drying oven may also help.
7.2 Acid filling

The glass mat separator has a critical role in electrolyte filling. Any change in the physical properties of this material can drastically change the quality of the filled and formed cell or battery. The separator structure, degree of compression and fiber composition have a significant influence on how well an unfilled element will accept electrolyte. While high levels of compression are desirable for extended life, this may make the filling and formation process more difficult. When the separator is compressed, the pore size is reduced, and the space available for electrolyte between the plates is also reduced. This will make the filling process more difficult. (Section 8 gives more detail concerning separator optimization).
Figure 10
The Filling Process Within A Vacuum And Non-Vacuum Fill.
Pre-Fill Stage Vacuum Process Non-Vacuum Process
gets hotter, due to the exothermic reaction with the plate pastes. In the extreme case this heat build up can generate steam, which will impede the further ingress of acid/water, and if severe enough may also cause buckling of the plate stack [6]. (Figure 12 Figure Figur 12) Another danger is the formation of hydration shorts/ dendrites. As the acid reacts with the plate pastes, the sulfuric acid electrolyte becomes progressively more dilute. Lead sulfate is relatively soluble in the hot electrolyte with a pH close to that of water, and soluble lead sulfate will diffuse into the separator. This will hasten the formation of lead dendrites and/ or hydration shorts. A short circuit may develop and be detected during formation, or more subtly the battery will fail prematurely in service due to the formation of lead dendrites. Sodium sulfate is a useful additive to help to prevent dendrite formation: however during the filling and formation process the common ion effect may not be strong enough to prevent dendrite formation if the electrolyte turns to hot water during the latter stages of the filling process. If the filling process is poor, individual cells may also have dry areas after filling in which little or no liquid is present. These dry areas will slowly become wetted during and after formation, but massive grid corrosion may result due to the high temperatures and alkaline conditions prior to and during formation.
Plate pores, separator are evacuated of all air bubbles.
Plate pores, separator are filled with air.
Electrolyte Added
Separator is open, all plate pores are accessible, rapid reaction ensues.
Ingress of acid is impeded by air bubbles, slower rate of acid with pastes.
Figure Figure 10 is a schematic view of what happens in vacuum and non-vacuum filling processes (section 3.1). A high-vacuum fill allows faster acid ingress, therefore shorter filling time and higher productivity. However, because it removes air from most of the plate pores, it greatly increases the reactivity and thus the rate of heat generation [6]. The battery design and manufacturing process need to be able to cope with this rapid surge of heat. When electrolyte is added to the cell, the ideal situation is that all areas are wetted as much as possible by the same amount of acid so that there is perfectly uniform distribution of electrolyte throughout the plate stack when the filling process is completed. This ideal situation is difficult or impossible to achieve in practice, as there is a dynamic competition between the separator and the plate surfaces for the electrolyte [6] (as shown in figure 11 As the electrolyte penetrates into figure 11). the plate stack, it is held up by the separator (the capillary forces tend to hold the electrolyte fairly strongly), and at the same time the electrolyte is depleted by the exothermic reaction of the sulfuric acid with the plate pastes. As the liquid front penetrates deeper into the stack it becomes more dilute and also
Figure 11
Conceptual View Of the Filling Process For A VLRA Cell.
Fill Electrolyte
positive plate sponge
negative plate sponge

plate height, I
21
Dry area?
Separator
interplate spacing, d
When I/d is high, proper filling is difficult or impossible
Figure 12
Action On The Leading Edge Of The Liquid In A VRLA Cell Filling Process.
Fill Electrolyte
PbSO dissolves in hot water, releases soluble PB(II) into separator
Pb(II) Pb(II)
hydration shorts, dendrites
liberated CO2 from carbonated pastes exerts back-pressure on electrolyte expander is leached out by hot water
pressure, heat from steam may buckle plates, soften and bulge plastic case
Hot Water Steam paste particles are dislodged by steam, stain separator
Separator Positive Pasted Plate Negative Pasted Plate
Heat dissipation can be aided by the following options:

s
A cell design with a high surface area to volume ratio, which allows a longer time for acid ingress, and thus a longer time for heat dissipation. Chilling of the unformed element and/or electrolyte prior to filling. Chilling the electrolyte will be more effective than chilling the unfilled elements. Chilling of the filled element, using water rather than air cooling due to its greater heat capacity.
Use multiple fill ports and channels in the battery case to guide acid and remove the y factor. Use push/pull massage after electrolyte introduction. Use a lower surface area AGM: but this may have an adverse effect on performance. Reduce compression to create a higher mean pore size: but depending on the battery application this may have an adverse effect on battery life. Use more separator and more electrolyte so that the heat-sink properties are improved.
Dendrite growth and grid corrosion result from poor acid ingress and distribution. The following factors are important in promoting faster, more effective acid ingress:
s s
Minimize heat build-up and steam formation. Avoid significant carbonation of unfilled elements by taking special care in the drying and cooling of the unformed battery plates. Use a fluffier more open separator with a lower grammage for a given caliper. However, this may have inferior compression characteristics, with implications for battery life. Use a high or rough vacuum fill.
Unfortunately, design or materials changes that improve battery performance and/or life also tend to make proper filling more difficult. This includes e.g. high surface area glass fibers, high levels of compression, and thin plate designs. Thin plate designs make possible the creation of small, powerful batteries. But they also mean higher surface areas, smaller interplate spacing and generally greater l/d ratios (section 10.1). These impact negatively on the filling process, therefore extra care needs to be taken in filling and formation, to avoid tipping over the knife-edge into battery problems.
7.3 Control of Formation Temperature

Thermal management of the battery must be tailored to its size, the type of filling process and the formation conditions used. Obviously, there is no set formula for all VRLA products, so handling of batteries prior to and early in formation must be developed by the manufacturer. In general, it can be said that very small VRLA batteries (25Ah or less) are much more tolerant of temperature and cells more so than 6 or 12V batteries; for these products minimal thermal management is required (but the risk of over-cooling is high). For larger products the thermal requirements become more stringent and in the 60-100Ah range, particularly in multi-cell batteries, great care must be taken that the batteries are adequately cooled, but are not cooled too much. The VRLA battery manufacturer could use the following checklist in order to determine the necessary temperature control to apply:
s
poor that the degree of formation and initial discharge capacities are reduced. It is an unfortunate fact that negative plates are formed best at temperatures of 40C or less; positives form better at higher temperatures. When positives are formed below 40C there is an increase in the alpha-PbO2 content, a decrease in total PbO2 and a fine pore structure susceptible to clogging by lead sulfate during discharge can result. Because of all of these factors, a good formation system for small VRLA products should include some form of forced, circulated air that can be either heated or cooled, depending upon the ambient plant conditions, the product size and shape and the amount of heat production at each stage of formation. Larger batteries (25-100Ah), particularly those with unfavorable envelopes for heat dissipation (a cubic structure is the worst case), almost always require cooling during formation. This is necessary to avoid high damaging temperatures and also to have uniformity of temperature among the cells in a 6- or 12V battery. In the best case (small size, favorable envelope, long low-current formation) passive cooling (radiation, convection) can be employed, but there is likely to be a large swing in product quality between summer and winter conditions unless the plant has an outstanding air-conditioning system. It is more likely that active cooling will have to be employed and this can range from forced ambient air to forced chilled air to stationary water bath to chilled, circulated water bath. The use of water for cooling compared to air is significantly more efficient due to the higher heat capacity (four times that of air) and thermal conductivity (roughly 15 times greater) of water. Overall, heat transfer away from batteries during formation is roughly nine times more efficient for water over air. Clearly, water is preferable to air for a number of reasons but it usually involves a higher capital input and more maintenance. In some cases it may not be necessary, but perhaps the best argument for its use is that with the efficient cooling of circulated water shorter formation times can be used for some products, thus maximizing the capital input for formation equipment. Different products will each require a specific type of temperature control.
The size of the product (i.e., single cell or 6/12V battery) The battery envelope (i.e., the surface-area-tovolume ratio for heat dissipation). How has the battery been handled prior to the start of formation? Battery temperature at the start of formation How will the batteries be cooled or heated during formation (ambient, forced air, water, circulated water)? What is the plant temperature (air conditioning, summer or winter)? How much heat will be generated in the formation process and what is the duration? Is the battery formed sealed or open, starved or flooded (i.e., what are the contributions of the oxygen cycle to heat generation and of gassing to heat dissipation)?
s s
The smaller the battery and the higher the surfacearea-to-volume ratio the more easily heat is dissipated. In the extreme, heat dissipation may be so good that small VRLA products actually require heating during formation! This is particularly true in a plant environment with poor temperature control i.e., hot in the summer and cold in the winter. Small VRLA batteries require a minimum temperature for efficient formation and a maximum above which damage can occur. At the low end, formation of the PAM can be so
23
Figure 13
2.5 Ah And 25Ah Spiral-Wound Single-Cell Internal Temperatures During Different Fill-To-Form Conditions.
140 120 100
A D cell (2.5 Ah), 20 minutes in 10C cool and wash, balance in 23 air. B BC cell (25 Ah), 30 minutes in 10C cool and wash, balance in 23 air. C BC cell, air cooling (23C) only. D BC cell, electrolyte at 10C, 30 minutes in 10C cool and wash, balance in 23C air.
Temperature, C
80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
B A
C D
Time After Fill, Minutes
Several examples of the relative effects of air and water cooling are shown in Figures 13 and 14 for Figures several different battery sizes. These figures compare cell and battery temperatures with different combinations of chilled and room-temperature electrolyte and the use of air and water during the fillto-form period. As can be seen, there are significant differences in heat dissipation using air and water and this also applies during formation [7]. Thermal management during the formation process itself is also very important. As noted earlier, at the beginning of formation the battery resistance is quite high and if high inrush currents are employed ohmic heat generation can be substantial. The formation reaction itself is highly exothermic, as is the continuing oxide/acid neutralization reaction that takes place as acid generated by electrolysis penetrates deep into the plates as formation proceeds. The heat of reaction for the formation process is ~394 kJ/mole and the heat of neutralization is ~161 kJ/mole both of which are substantial. Later in formation the overcharge processes of water decomposition and hydrogen
evolution begin and this also contributes to a higher temperature level. If the battery is formed in the starved state there is a further contribution to heat generation by the oxygen-recombination reaction involving oxygen reduction at the negative electrode. Finally, if the battery is formed sealed heat dissipation due to gassing is minimized, whereas maximal heat dissipation occurs if the battery is formed in a flooded state and/or is open to the atmosphere. To some extent the formation algorithm can be used to control the temperature during the formation process, and smooth out temperature peaks. For example, longer formation times at lower currents not only minimize ohmic heating but they also provide longer times for heat dissipation. This is usually in conflict with manufacturing pressures, which favor the shortest possible formation times. The use of rests or discharges will provide time for heat dissipation, but they also lengthen formation. However, there are other technical advantages to these steps as discussed in section 3.13.
24
Figure 14
6V/100 Ah Prismatic Battery Temperature Data (Middle Of Center Cell) During Fill-To-Form Time With Different Conditions.
110 100 90
A Room temperature electrolyte. Cooling water at 8C. B -20C electrolyte. Cooling water at 10.5C. C -20C electrolyte. Cooling in air at 21C.
Temperature, C
80 70 60
B C
47C at 360 Min.
50 40 20 40 60 80 100 120 140 160 180 200
Time After Fill, Minutes
7.4 Completion of formation

In theory, formation is complete when there is conversion to 100% lead at the negative and 100% lead dioxide at the positive, but this is not possible in practice. The required degree of conversion will also depend to some extent on the battery application. Close to 100% conversion may be possible for the negative plates, but is more likely to be in the range 90-95% conversion to lead dioxide for the positive plates. A lower PbO2 percentage may be acceptable for SLI batteries because it is assumed that the battery will be used fairly quickly and even though it is not completely formed the battery will be able to start the vehicle. In addition, the shallow discharge service will slowly help to complete formation; because the battery is usually charged by the alternator at a relatively low voltage and will typically only be at 6070% state of charge in use. On the other hand, many industrial batteries are required to deliver nearly 100% of rated capacity when they are put into service by the end user; in addition, they may be held on the shelf for long periods of time before they are commissioned (thus requiring very low levels of unformed oxide which equates to a low self-discharge rate). Because of these considerations, completeness of formation must be relative to the product and its intended use.
7.5 Alternative Jar Formation Options

The battery manufacturer may wish to consider some rather more unusual assembly/formation options to overcome some of the disadvantages of jar formation already mentioned. One option might be to place the assembled plate group in a container which is larger than the final container so that the plate group is under little or no compression during formation. After formation, surplus acid is drained off, and the plate group is placed in the final container to give the required degree of compression/plate group pressure. This approach would eliminate the problems already mentioned, which can occur when formation is attempted with a high plate group pressure. In particular it would eliminate the problem of changes in plate group pressure during the formation process (arising from changes in the volume of the battery plates during formation).
25
Battery and separator design guidance
8. Battery Design
The battery design has a critical influence on the filling and formation of VRLA batteries. Some of the potential problems with VRLA battery filling and formation can be minimized or eliminated by careful battery design. Unfortunately, some of the design strategies to improve filling and formation may have an adverse effect on battery performance and life, so some compromise may be necessary. The battery design parameters which may influence VRLA cell/battery filling and formation include:
s
8.1 Plate height/plate spacing ratio (L/d)

The ratio of plate height to plate spacing (L/d) can be used as a rough measure of the difficulties encountered in filling. For a L/d ratio equal to or less than 50, easy filling results. If the ratio is between 50 and 100, care should be taken to avoid potential problems. Filling becomes more difficult when the ratio is between 100 and 200, and is almost impossible at ratios above 200 [8]. It can be deduced from this that the worst case is a tall battery with a close plate spacing: the best case is a short, narrow battery with a wide plate spacing.
Battery height: tall batteries are harder to fill than short ones. Battery width: short, wide batteries are more difficult to fill if filled from a single filling port. Plate thickness and interplate spacing Plate height and plate height/interplate spacing ratio Filling port position Battery case draft Active material additives (expander/reinforcing fibers) Gravity: liquid will only wick so high before being defeated by gravity: not a factor in gravity filling (top to bottom). Separator properties: - Volume porosity and pore structure (mean pore size). Finer pores wick more slowly but to greater final heights - Saturation - Compression: results in a finer pore structure with high tortuosity, therefore slower wicking. For example, 15% compression will double wicking time to a given height. - Caliper (thickness at defined pressure) - Grammage (g/m2) - Surface area/ fiber diameter. Finer fibers (higher surface area) result in finer pores, hence slower wicking. - Wettability - Fiber structure (coarse/fine fibers; inclusion of synthetic fibers). Organic fibers inhibit wicking (not wetted by sulfuric acid), and also promote faster drainage. - Fringe area of separator (area of separator not covered by plates)
8.2 Battery case draft

Battery case draft can result in a 10% compression change from the top to the bottom of the plate. For example, with a target compression of 25%, the compression may actually vary from 20% to 30% between the top and bottom of the plate. The effect of this on the performance and life of the battery may be highly significant. This effect should not be ignored in the formation process as well. The separator at the bottom of the cell will be subject to a higher compression resulting in a smaller pore structure which will influence the speed at which the acid fills the separator during the acid filling process: there will be a slower acid drip speed as the acid approaches the bottom of the plate. Because smaller pores have a greater force to pull liquid, this may also increase stratification. [9], [10]. There is a slower acid drip speed as the acid approaches the bottom of the plate.
s s
s s s
8.3 Active material additives

Additives in the active material can also affect the filling operation. The expander or reinforcing fibers in the paste may interact and result in excessive gassing during acid addition. This will result in a longer fill time or even in an unacceptable product. Care needs to be taken when any new material is used since the VRLA battery should be considered as a system and all the ingredients interact.
It can be seen that the separator properties are critical, and these are discussed in more detail in Section 9. The other critical battery design parameters are discussed as follows.
Figure 15
Solubility Of Lead Sulfate In Sulfuric Acid At 25C.
Solubility of Lead Sulfate (mg/liter)
7 6 5 4 3 2 1 0 1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4
9. Separator Optimisation
The separator properties have a critical impact on acid filling and jar formation [6]. Any change in the physical properties of this material can drastically change the quality of the filled and formed cell or battery. The type of separator used is dictated more by the intended battery application, but its properties can also partially determine the filling and formation conditions used. During the filling process the acid wicking rate is important. The acid wicking rate is primarily a function of the mean pore size of the glass-mat separator; this, in turn, is largely a function of the fiber mix (represented by the fiber specific surface area as measured by BET), the density of the glass mat and the compression level in the unfilled plate stack. In practice, wicking is only directly applicable for top-down gravity and bottomup filling methods where wicking is the primary mode of fluid transport. For soft- and hard-vacuum filling techniques separator properties also have a role, but the vacuum level and filling speeds are additional control elements, in addition to the electrolyte temperature and its resultant viscosity. A further variable is the use of 1020% organic fibers mixed with glass, as in the H&V IIP15 material (see also section 10). The organic fraction confers greater tensile strength and it also facilitates filling due to the hydrophobic nature of the organic fibers. Since the organic fibers are not wetted by sulfuric acid, the electrolyte is not held as strongly as by glass fibers. This clearly facilitates filling, but it can result in flooding of the negative-plate pores with acid and electrolyte regurgitation and spray may be significant during formation. The actual separator compression in the plate group will influence the ease of acid filling and jar formation as well as impacting on the performance of the battery. High separator compression has been shown to be beneficial in extending the life of VRLA batteries by inhibiting positive plate expansion, but unfortunately the process of filling the battery with acid becomes more difficult. When the separator is compressed it reduces the pore size significantly and also reduces the space available for electrolyte between the plates. This adversely affects the wicking properties of the electrolyte. However, smaller pores and higher compressions may mitigate variations in saturation and acid strength in the vertical plane (stratification). It is also important to optimize the ratio of plate and separator pore volumes to ensure sufficient electrolyte.
Density of Sulfuric Acid (mg/liter)
8.4 Electrolyte additives

Most VRLA battery manufacturers use sodium sulfate as an additive to the electrolyte. It is added to the electrolyte in powder form, at about 1% by weight. Sodium sulfate acts by the common ion effect to prevent the harmful depletion of sulfate ion which is a danger in the discharge of acid-starved batteries. The addition of sodium sulfate provides an inventory of sulfate ions that are available without increasing grid corrosion [11]. The solubility of lead sulfate increases significantly as the concentration of the sulfuric acid electrolyte decreases, as shown in figure 15. figure 15 The solubility increases more than fourfold as the sulfuric acid density decreases from 1.300 to 1.100 kg/l. Under certain conditions of over-discharge, the amount of dissolved lead sulfate may be such that on recharging the reduced lead will form metallic bridges between the plates. The addition of sodium sulfate will reduce this risk. Alternative electrolyte additives may be used, which have a different mode of operation. This class of additives is known as dendrite prevention additives (DPA) [13]. These operate by actively seeking out and deactivating the lead dendrite growths. They are polar organic compounds that are believed to deactivate a growing lead growth by coating its tip with a layer of oriented molecules. Once the lead growth is deactivated, these molecules are available to move onto the next site.
27
The easiest filling is achieved by using a combination of glass and organic fibers with a low specific surface area (~0.8-1.4 m2/g) in a low-density material (i.e., high percent porosity of ~95% or more) that has a relatively low compression level (25-30 kPa dry or less) in the assembled, unfilled plate stacks. This gives an open structure that is not completely wetted by the electrolyte and, thus, has the best chance for uniform fluid distribution. This type of separator would be best suited to gravity filling. However, this type of separator and cell construction also is most susceptible to electrolyte drainage and stratification, particularly in deep-cycling applications. On the other hand, the best type of separator to minimize drainage and stratification is a high surface-area glass (~2.0-2.6m2/g), high density (90-92% porosity) all-glass with high compression. This will give excellent deep-cycling results but it is extremely difficult to fill, particularly in large batteries. One might think that a high-vacuum fill would be best for this type of separator but, in fact, this would only be true in relatively small VRLA batteries (~25Ah or less) due to the large amounts of heat generated in short times in high-vacuum fills. If the battery configuration cannot dissipate the large burst of heat generated by the filling process there can be permanent damage in the form of plate buckling, separator staining by paste and/or expander, bulging of the case and destruction of terminal seals; internal cell temperatures in excess of 110oC can be achieved for relatively long periods of time. Gravity fills with this type of separator system will take much longer times (possibly up to 30-40 minutes), but thermal issues will be minimal. In the formation process itself the separator can have an influence on gas management and electrolyte distribution. The importance of this influence will depend in part on whether open (section 3.3.1) or sealed (section 7.1) formation is used. For open formations, the type of separator used is not critical, as provisions are available for gas escape and fluid management, whether by using a completely open top or by having hollow tubes attached to the cell fill ports. For sealed formations, however, the separator type and amount, the saturation level, the formation algorithm and the design headspace are all important. Forming sealed under pressure can obviously lead to acid regurgitation and spray, but it also puts pressure on all of the seal areas and can damage the functioning of some vent valve designs by allowing electrolyte to leak into the seal area between the valve and the fill stem, possibly leading to
valve sticking (where the valve wont open and release gas at the design opening pressure). If a battery is designed for sealed formation, the separator chosen should have a relatively high surface area (~1.6m2/g or more), be all glass and should have at least moderate compression in the finished plate stack. This is to ensure that the separator reservoir holds its electrolyte as tightly as possible without having unacceptably poor filling characteristics. Moreover, the amount of separator per ampere-hour of capacity should be relatively great (1.4-2.0g/Ah) and the saturation level after filling should be ~95% or slightly less; it is virtually impossible to form a VRLA battery sealed in the fully saturated or flooded state. These parameters will allow use of a sealed formation even with fairly aggressive algorithms (high currents, short times). However, care should be taken to minimize heavy gassing by using multiple rest periods when gassing potentials are reached. These will also allow more electrolyte to soak into the plates and more void space will be created in the separator, enhancing electrolyte retention. The rest of this section will give further background information concerning the influence of the separator choice and separator properties on the filling and formation process. Typical design parameters are given below, and are then discussed in more detail: Volume porosity: porosity: 92%
95% Saturation: Compression: Compression: 30% Acid utilization: 8.8 9.5 ml/ Ah Separator Related to interplate spacing and caliper: required degree of compression Separator grammage: Separator surface area: surface area: > 2g/Ah preferred > 2m2/g preferred to minimize stratification, but filling will be more difficult. 4+ times rated capacity. It is necessary to be careful about too much overcharge during formation: this may damage the positive plates and increase the overall acid density.
formation Jar formation Ah input:
28
9.1 Volume Porosity

This is a very important figure as it will determine how much acid the separator will hold and is therefore a critical parameter in the battery design and the determination of the battery capacity. For 100% glass separators, the figure for the volume porosity in the nominally uncompressed state is typically in the range 92-95% (as measured at 10 kPa). Compression of the separator will reduce this by a few percent, so that in the compressed state it will typically be 90-92%. There has been some experimentation with hybrid separators containing a percentage of polymeric fibers, or non-glass separators such as the Daramic AJS separator. These separators may have a lower volume porosity and will therefore hold less acid. This needs to be taken into account in the battery design.
9.3 Separator Caliper

There is a standard BCI method for measurement of separator caliper (thickness). Because of the fluffy nature of the RBSM, the caliper is measured at a controlled pressure of 10 kPa. Although this may be referred to as the uncompressed thickness, it is important to note that this is referenced to a controlled pressure of 10 kPa. Specifications of e.g. 20% compression or 30% compression are referenced to this thickness as measured at 10 kPa. The required separator caliper needs to be calculated in relation to the interplate spacing and the specified % compression. For example, a battery with an interplate spacing of 0.11cm and a design separator compression of 30% will require separator material with a caliper of 0.16cm. When compressed by 30%, this separator material will have a caliper of 0.11cm. Recent research has highlighted the importance of plate group pressure rather than % compression as the key to extending battery life (see comments below).
9.2 Saturation Level

The cell/ battery is not normally sealed until the formation process is complete and the separator saturation level is deemed to be correct (normally 95%) (except for sealed formation, section 7.1). If in doubt, the battery manufacturer should err on the side of overrather than under- saturation of the separator. If the separator is over-saturated, and the cell is then sealed, the recombination process will be less efficient initially, some water (as hydrogen and oxygen gases) will be lost from the system, and the efficiency of the recombination process will increase, preventing further water loss. However, if the separator is under-saturated in relation to the design value, the cell may contain insufficient acid to meet its design capacity. An accurate calculation of the amount of acid absorbed by the plates and separator is needed when precision acid filling is used [12]. The separator needs to be 9095% saturated, which corresponds to the separator in the area between the plates holding about 6g of acid for every 1g of separator. The fringe area of the separator also needs to be considered. Larger fringe areas allow for additional acid and thermal capacity in the battery: this will help in the thermal management of the cell. There may also be some contribution to the low rate capacity of the cell. The acid gravity is normally between 1.290 and 1.320 and for guidance a figure of from 9 to 11 ml of acid per Ah as measured at the 20hour discharge rate should be used in the battery design calculation. Fully discharged, the acid gravity can be around 1.08.
9.4 Separator Compression

ALABC research work has shown that high compression battery designs can extend battery life by maintaining a high pressure against the positive plates and eliminating or minimizing the phenomenon known as premature capacity loss. In fact, it may be more relevant to refer to plate group pressure rather than % compression. Some recent separator designs are less compressible but may be able to maintain a higher pressure against the positive plates than conventional glass separators [13]. Unfortunately, a high plate group pressure/ high compression design may also be more difficult to fill. A higher compression will generally result in lower fill rates. When the separator is compressed it reduces the pore size significantly and also reduces the space available for electrolyte between the plates. This adversely affects the wicking properties of the electrolyte. Another issue that may need to be considered is that of changes in plate group pressure during formation. There are changes in the volume of both positive and negative active materials during the formation process as the lead oxides are converted to lead at the negative plates and lead dioxide at the positive plates. This may have some effect on the separator compression and applied plate group pressure. This needs to be taken into account in the design of the battery and the specification of the separator and the initial compression level.
29
Mean Pore Size (microns)
However, recent ALABC research has shown that under some circumstances the separator compression may drop significantly after formation [14]. The exact causes of this are not yet certain, but may be related to a relatively low initial plate group pressure. There may be a critical compression that holds the fibers in place (for a particular separator) and if there is significant gassing at the end of the charging process there may be a loss of integrity in the fiber mat. While this problem is not yet fully understood, the following design issues need to be considered to minimize the risk of loss of plate group pressure during jar formation:
s
Figure 16
Mean Pore Size Vs. Kr BET Surface Area.
20 18 16 14 12 10 8 6 4 2 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Assemble cells with the maximum practicable plate group pressure (> 40 kPa) Maximize available acid volume and increase separator grammage to >= 2g/Ah. Increase the fine fiber content of the separator. Use a formation algorithm that minimizes the gassing at the end of charge.
BET Surface Area, m2/g
s s
9.5 Separator Grammage

The separator grammage is the weight of the separator per unit area. The amount of acid held by the separator is very important, therefore separator grammage as well as thickness needs to be considered carefully in the battery design. The aim should be to maximize available acid volume, which will improve the heat capacity of the battery and enable thermal management to be improved. The amount (g/ Ah) of separator used will also have an impact on battery processing and performance. A higher amount of separator (around 2g/ Ah or greater) will have a beneficial impact on practical levels of compression, gas recombination and acid stratification. This also implies a greater plate spacing and larger electrolyte reservoir. This will make it more practical to carry out jar formation on the completed cell/ battery [15].
9.6 Separator Surface Area

The surface area of the glass mat is very important because it has a great deal to do with wicking during fill and fluid movement in fill/formation. There is a reasonably well-defined relationship between surface figure area and pore size, as shown in figure 16. This curve was constructed from data on various separator samples from a wide range of manufacturers [16].
The separator surface area for a glass separator is related to the ratio of coarse/ fine fibers. A lower surface area (higher proportion of coarse fibers) separator has advantages in the filling process, but may have other disadvantages depending on the battery application. A higher surface area correlates to a smaller pore structure and results in a lower wicking rate, but a greater ultimate wicking height [17]. (Figure 17) The smaller pore (Figure 17). structure will also help to decrease stratification within the cell. The pore structure of the separator provides for a highly tortuous path which helps to prevent dendrite growth and minimizes the size of any dendrite if formed. However, this also creates a tortuous path for acid and air movement. This increases the filling time for each cell. A battery designed for deep cycling should use a high surface area separator, but extra care will need to be taken during the filling process. The higher surface area separator will require additional time to add the acid, since the acid wicks more slowly through finer pores. Also, the way the acid is added to the battery is critical. If the acid is added too rapidly from the top, the air within the plates and the separator may not have enough time to escape, and dry spots may result. If the filling process allows the acid to wick up the separator, entrapped air can escape since it does not have to diffuse through the electrolyte. It is also necessary to allow sufficient time to allow for complete filling of the pores of the separator. With a high surface area separator, an advantage of the longer time for acid ingress could be a longer time for heat dissipation. The filling procedure is critical to providing a quality VRLA battery.
30
Figure 17
Impact Of Surface Area (m2/g) On Water Wicking Height While Under 20% Compression, After 24 Hours.
8.00
10. Separator designs to improve wet formation

For many years H & V have marketed a hybrid glass/ organic separator Hovosorb II, covered by US patent 4,908,828 [18], [19]. This separator contains a synthetic fiber with reinforcing glass strands, the balance being microglass. The synthetic fibers are hydrophobic, and these hydrophobic sites within the separator matrix offer controlled wetting properties of the separator and modification of the recombination process. It is claimed that this assists the filling and formation process, and allows a fill and spill formation technique to be used. The unformed cells can be flooded with electrolyte prior to formation, and drained of excess free electrolyte after formation. It is claimed that the use of Hovosorb II together with a fill and spill formation system may result in a more uniform cell-to-cell electrolyte concentration than is obtained with volumetrically filled, in-container formed cells. With Hovosorb II, because of the hydrophobic sites within the separator, recombination can occur even when the separator is fully saturated. This is because some pores within the separator remain unfilled, even in the presence of excess free electrolyte, so that oxygen is provided with a path for transfer to the negative electrode surface. Oxygen recombination may even be enhanced at those areas on the negative electrode surface that are in direct contact with the non-wettable material contained in the separator. The separator also has improved compression properties compared with 100% glass separators. Hovosorb II-P-15 is a refinement of the original Hovosorb II, and has improved puncture resistance.
150
Wet/Dry Weight of Separator
7.00 6.00 5.00 4.00 3.00 2.00 1.00 0 20 40 60 80 100 120 140
Height (cm)
= 0.8 = 1.3 = 2.2 = Hybrid @ 1.0
Figure 18
Effect Of Fiber Mix And Segregation On Vertical Wicking Speed.
1000 900 800 700 600 500 400 300 200 100 0 0
Slower Wicking
Time / s
Faster Wicking
25 50 75 100 125
(Height)2 / cm 2
= 100% Fines = 20% Fines = 0% Fines = AMER-GLASS = AGM(B) = AGM(A)
Note: Strips of AGM not compressed. All AGM samples are single layered except AMER-GLASS, a multi-layered AGM.
An alternative approach is to use separators that consist of two or more layers of different fibers: this may be helpful in the filling process since layers of coarse fibers are soaked more quickly. Battery filling is made easier with an oriented separator that has separate layers of coarse and fine fibers: the fine fibers against the positive plate, and the coarse fibers against the negative plate. This has a very fast wicking characteristic both upward and downward [20] [8]. In figure 18 the influence of fiber figure mix and segregation on the vertical wicking speed is shown. Figure 19 shows the upward and downward Figure wicking height for oriented and non-oriented fibers. During the filling process, the fine fiber component absorbs acid quickly, but when the battery is filled from the top, the looser, coarser fiber structure permits an
31
Figure 19
Upward And Downward Wicking Height For Oriented And Non-Oriented Fibers.
UpWard Wicking Height (cm)
80 70 60 50 40 30 20 10 0
50 100 150 200 250 300 350 400 450 500 Compressed @ 10 kPa; band 5cm
= Non Oriented AGM (mixed fibers) = Oriented AGM (segregated fibers)
Wicking Time (Min.) UpWard Wicking Height (cm)

40 35 30 25 20 15 10 5 0
5 10 15 20 25 30 35 40 Compressed @ 10 kPa; band 5cm
Wicking Time (Min.)
easier access to the electrolyte which then permeates instantaneously to the fine fiber side. When the process is in reverse and acid is spilled out of the battery, the forces binding the electrolyte to the coarser fiber structure are weaker, so that electrolyte will be preferentially lost from this part of the AGM. The desired partial saturation of the separator is thus quickly reached. This multi-layered AGM such as that manufactured by Amer-Sil, has faster wicking properties which may be of great value in the gray zone of filling where the ratio of plate height to plate spacing is between 50 and 200.
Another possible option is to include a thin microporous sheet as part of the separator system: this may help to eliminate the problem of lead dendrite formation. This might also help to control the diffusion of oxygen from the positive to the negative plate. An example of such a microporous separator is the DuragardTM separator recently announced by ENTEK International at 7ELBC in Dublin. Amer-Sil has also developed a composite separator which includes a microporous sheet between two layers of glass. Results with this separator system have been reported in the ALABC Research Programme.
32
VRLA gel batteries overview
11. VRLA Gel Batteries

Gel is an older technology than RBSM technology for valve regulated lead acid batteries: gel batteries have been around for many years but have not been widely used except for special applications. The gel system has an inferior performance at high discharge rates, therefore it is not suitable for applications such as SLI requiring a high rate discharge capability. The gel technology may be more appropriate for tall cells because the gel system does not suffer from stratification problems. The gel is produced by adding finely divided (fumed) silica to sulfuric acid: the concentration of silica in the gel is about 6%. Fumed silica is a high purity silica manufactured by high temperature hydrolysis of chlorosilanes in a hydrogen/ oxygen flame. It is typified by a small particle size and a very high surface area, in the range 200 400 m2/g. When the fumed silica is mixed with sulfuric acid, a viscous solution is formed, which develops into a thixotropic gel on standing. It has been found that the inclusion of phosphoric acid is also beneficial, this gives the battery a much longer cycle life, greater than the addition of phosphoric acid to a flooded battery design. The optimum H3PO4 concentration is 17-30g/L. A typical process for preparation of the gel is given below: 6% by weight of fumed silica is added to 1.280 S.G. sulfuric acid with continuous stirring. 20g/ L of 85% phosphoric acid is added and stirring continued. Stirring is continued while pouring the gel into the battery: the gel will set as soon as the stirring ceases. The battery should be filled under vacuum and stirring should not be stopped until the gel is in the battery. Because the gel itself is unable to prevent the penetration of lead dendrites that can cause short circuits between the plates, a conventional separator is required, which has ribs on both sides of the separator. Stratification is negligible, because the gel is more strongly fixed in the plates and separators. It is not possible to carry out jar formation of gel batteries, the battery needs to be assembled with dry charged plates before adding the gel. Sufficient time needs to be allowed for the gel to set up, and the cells should then be given an equalizing charge. During this equalizing charge free electrolyte may percolate from
the gel as the gel cracks and shrinks. The free acid can be removed from the top of the cell before sealing the cell and this will ensure that the cell is in a recombinant state prior to going into service. The electrolyte removed from the cell can be measured to provide an accurate volume of the electrolyte present in the cell. The total amount of electrolyte in the gel cell may be slightly less (~ 80-90%) than that added to a comparable AGM cell. The electrolyte is immobilized in a 3-dimensional structure set up by the very fine particles of silica [20]. In an aqueous medium, these particles are fixed by the presence in their midst of negatively charged sulfate ions. The gel formation happens at a molecular level and allows the electrolyte to be present in the cell in an altered state that is neither liquid nor solid. It does not have the required mechanical strength properties to separate the plates, therefore a mechanical separator is needed. The separator in the gel cell is purely a spacer, similar to the separator in a conventional flooded cell. Gas transfer occurs via void spaces which develop within the gel structure as it dries out slightly and opens up under the pressure of the oxygen gas bubbles. The density of the gel will depend on the silica content of the gel. The higher the gel density (higher silica content) the harder it will be for crevice creation, and the greater the need for a very open separator. There may be different consistencies of the gel dependent on the silica content of the gel: a soft gel has a lower silica content than a hard gel. The soft gel will have a relatively weak structure and only very low shear forces are required to break it. The consistency of the gel will influence how many cycles of discharge/ charge are needed to achieve recombination. Initially when the silica gel is formed there is total water saturation. The gel structure is completely filled with the electrolyte, which is also present in the active material and in the gel/ electrode surface interfaces. For the first few cycles, the gelled VRLA cell functions similarly to a flooded lead/ acid cell and water loss occurs, particularly at the end of the charging periods. As water is lost in the initial charging cycles, there is a slight dry-out of the gel structure, which creates micro-channels in the gel
33
VRLA gel batteries overview / Formation equipment, battery monitoring and product testing
through which gas can pass. The softer the initial gel, the sooner the recombination process starts, because the process of gel cracking can start more easily. The formation of the gel is not influenced by the type of separator used or its design. However, the separator does have an important influence on the filling process and the ease of oxygen transport during the recombination process. The separator for gel batteries is a conventional separator as may be used in flooded lead acid batteries but with some important design differences. There are deep ribs on the negative side of the separator as well as on the positive side of the separator. The ribs are normally vertical and relatively widely spaced to permit easy filling with the gelled electrolyte. The interplate spacing is also very important: with a thin plate battery design having a close plate spacing, filling with the gelled electrolyte may be difficult if not impossible. For gelled VRLA cells, the separator porosity needs to be as high as possible. This is because the separator needs to retain as much of the electrolyte as possible in its structure and to minimize the barrier to oxygen transport through the separator. The pore size of the separator is also important in the context of oxygen transport through the separator. The optimum mean pore size needs to be in the range 1 - 10. Separators with a mean pore size < 1 (=micron) severely inhibit oxygen transport through the separator so that the only route for oxygen transport is around the edges or over the top of the cell group. The maximum rate of the internal oxygen cycle is lower in gel cells than in cells with AGM separators. The maximum rate is typically 10 A/100 Ah in AGM batteries, and 1.5 A/ 100 Ah in gel batteries [21].
12. Formation Equipment And Layout

Apart from the choice of electronic equipment (rectifiers, power resistors, power supplies, etc.), there are some critical practical issues in choosing battery connections and configuring formation bays. These can have as much to do with resulting product quality and uniformity as the choice of an effective formation algorithm.
12.1 Battery Connections

This topic is often neglected, but in practice can have a lot to do with product uniformity and scrap levels. There are several common methods of hooking units together in a formation bay, depending primarily upon the hardware used, the degree of automation, the numbers of units handled and the types of battery terminals. A few common connection methods are as follows:
s
Wires and alligator clips (usually for larger batteries) Manual plug-in of flat tabs to forming strips (usually for small single cells/batteries) Loading of trays that may contain 20-50 small batteries and insertion into formation bays using snap-on contacts or pressure springs.
Whatever the connection method, a few common areas of concern apply:

s
Are the battery terminals and formation contacts clean? If a pressure contact is used, are any components fatigued so that spring pressure is not adequate? Are the contact surface areas substantial so that localized resistive heating and/or oxidation doesnt take place? Are the battery terminals and formation contacts made from low-resistance materials that will not produce significant voltage drops? Are these materials easily corroded or oxidized?
These issues may seem trivial, but many formation connectors have minimal contact areas with battery terminals and the contacts and terminals are made from materials that are easily corroded by battery acid (steel, tin, copper) and/or are air oxidized at the high temperatures generated at the contact points (copper, copper-bronze). The use of more rugged materials such as nickel plating will minimize the above problems, but such materials may have significantly higher resistivities. High scrap levels or high formation recharge category can be created by the improper choice of connector
Formation equipment, battery monitoring and product testing
designs and/or materials and, more commonly, by infrequent inspection and cleaning of formation equipment contacts.
Figure 20
Battery Connections For Series Strings, Series-Parallel Arrays And Series-Parallel Matrixing.
A. Simple Series-String Battery Connections
The cleaning of battery connections is also a key health and safety issue. Hydrogen and oxygen gases are generated during the formation process: a potentially explosive mixture of gases is formed, which only needs a spark from a corroded connector to cause an explosion or even a fire. This has the potential to cause injury or death to personnel in the area: quite apart from the almost certain damage to batteries and/ or equipment.
Additional information about gas monitoring is given in section 13.3.
B. Series-Parallel Array Connections
12.2 Formation Bay or Circuit Configurations

There are three approaches to how batteries are configured in formation circuits:
s s s
C. Series-Parallel Matrix Connections
Series strings Series-parallel arrays Series-parallel matrices
By far the most common approach is to have a series string of batteries charged by a single rectifier or power supply. This is an effective approach and can result in good finished battery uniformity if voltage drops due to high contact resistances are not an issue. The major drawback is that one open contact (due to one open internal battery connection or to poor or broken formation bay mechanical contacts) can result in the loss of a whole string if not detected. Series-parallel arrays are useful in that more batteries can be formed from a single rectifier or power supply by dividing the available current (with all strings having the same total voltage), but there is a danger of overformation or even thermal runaway if one string, due to even a slightly lower overall resistance, draws a disproportionate amount of the total formation current available at the expense of the other strings. If this is even a subtle effect there may be a significant variation in product quality string-to-string. This can be avoided in the setup of the formation circuits by using steering diodes or other electronic measures to ensure that roughly equal currents flow through each of the paralleled strings. If an open condition occurs in one of the strings, however, it is lost as would be the case in a simple series-string arrangement; in addition, the total current available will now be distributed among the remaining strings according to their individual series
resistances. Thus, current would flow through the remaining strings, but at significantly higher levels than planned. If undetected by personnel or monitoring equipment, this would result in over-formation, and possible destruction, of the remaining batteries. The most effective approach, but also the most complicated from a mechanical standpoint, is to use series-parallel matrix connections where two or more series strings are put in parallel, but there are also crossconnections at some interval so that a matrix is created. This approach, along with the two previous constructions, is shown in Figure 20. The major Figure advantages of matrixing are that current is distributed more uniformly and in the event of an open battery or contact current continues to flow around the defective position. Batteries immediately adjacent to the defective position are affected to some extent (i.e., they experience somewhat higher formation current levels), but those further removed are not noticeably influenced when cross-connections are made between all units. The important point is that current continues to flow through all of the strings in a roughly equal fashion. Complete strings are not lost as in series-string and series-parallel designs and whole strings are not heavily overcharged in the event of an open battery. Matrixing is
www.firing-circuits.com 35
usually done for single cells or small batteries where large numbers of units are formed in individual bays. With these products, trays or carousels are used in which connections are made automatically so that cross-connections can be done between each unit. This approach gives the most uniform current distribution and ensures that the effects of defective units are minimal, but there is an added cost.
Cooling and heating equipment should also be maintained effectively. If forced-air heating or cooling is used, CFM and temperature measurements should be taken frequently in each of the formation bays to ensure that airflows are correct. If water baths are used for cooling, water should be checked frequently for temperature, flow rate (if applicable) and pH to ensure that acid buildup is not severe. All removable trays, racks, tables, etc. should be washed and cleaned of acid frequently in order to extend life and to minimize the occurrence of cosmetic rejects in batteries formed at later times (primarily terminal and label damage).
12.3 Critical Maintenance of Formation Equipment

To obtain optimal results from formation equipment frequent maintenance is required. Some of this involves electrical equipment such as rectifiers and DC power supplies, which must be calibrated on a regular schedule (this is essential for ISO 9000 or QS 9000 certification). The quality of the incoming AC line power should also be analyzed periodically. Computer control and monitoring equipment must similarly be maintained and calibrated, but perhaps the most critical aspect of formation-room equipment is the actual hardware that accommodates the batteries and which may be exposed to the high temperatures and acid fumes and spray that go along with such close contact. One of the most important parts of maintaining a formation room is to carry out regularly scheduled abrasive or chemical cleaning of hardware used to connect batteries in formation bays. Oxidized or corroded contacts can contribute large voltage drops to battery strings and this can cause undercharging or no charging at all in extreme cases. This can be detected if current/voltage monitoring is done on all battery strings, but if this is not done some batteries will be taken off formation with low voltages or in a dead state (i.e., zero voltages) and it may be attributed to the batteries and not the connectors. Because of this, connector hardware must be cleaned regularly, often enough that oxide and corrosion buildup cannot accumulate to levels that affect contact resistances. If an automated monitoring system is not in place, poor contacts can be detected manually by taking voltage and current readings on individual batteries or strings.
12.4 Power Quality and Equipment Costs

An important practical issue to consider is one of power quality and equipment cost and how these will impact upon a specific VRLA product that is being formed. Power quality varies greatly throughout the world and in some areas it is very poor. High levels of AC ripple and harmonics can and do feed directly into formation charging equipment; if filtering electronics is not included in the formation rectifier or power supply, large amounts of AC ripple will then be superimposed upon the nominally DC charge profile. In formed batteries out in field service it has been shown that AC ripple can result in severely-shortened lifetimes by generating heating that results in accelerated PAM softening and grid corrosion, as shown in Figure 21 [22]. Due to their low impedances, Figure VRLA batteries are affected to a greater extent by AC ripple than comparable flooded lead-acid products because low ripple currents translate into higher ripple voltages with the low impedances attributable to VRLA products. AC ripple is a form of mini-cycling that can wear out a battery prematurely in service, but it can also have negative effects in formation, depending upon its amplitude, symmetry and frequency [23]. In addition to temperature effects and enhanced grid corrosion, ripple can reduce oxygen over voltage values thereby increasing gassing during formation. Ripple is also
36
Figure 21
AC Ripple Voltage And Current Representation (Upper Figure) And Its Effect On Cell Temperature And Cycle Lifetime.
Volts 7.0 VCM 6.5 Test Battery is 8 Ah 12V SLA Type Amps 12
13. Battery Monitoring During Formation

Formation system monitoring is essential for both quality and safety reasons, as well as to keep scrap levels minimal. Formation scrap is the most expensive because units that are scrapped here have the maximum input of materials and labor of any stage of manufacturing other than finishing. Therefore, it is imperative that some form of monitoring equipment be used, even if this is only a formation-room worker with a multimeter.
ICM
13.1 Electrical Monitoring

The outputs of the charging system must be monitored for compliance with values for voltage and current as set out in the manufacturing documentation. In addition, measurements may be taken to ensure good power quality in terms of low levels of AC ripple and harmonics reaching the batteries being formed. Continuous recording of these data by Quality or Manufacturing personnel, with SPC charting posted in the formation area, is highly recommended. For the batteries themselves, the following parameters should be monitored and analyzed (monitoring alone is not enough: the data must be interpreted and appropriate action taken if necessary):
s
6 0 -6 -12 0 10 20 30 40 Time (mSEC) 50 60 70
Correlation of Battery Voltage and Charger Ripple Current

Ripple current (effective val.) Net Current 0.1C 0.2C 0.5C 0.8C Value 0.1 100 0.2 97 0.5 93 1.2 77 4.0 61
Item Temperature Rise (C) Proportion of life (%)
1.0C 6.0 50
Initial currents (CV or TC charge) or voltages (CC charge) Top-of-charge (TOC) voltages (peak voltage toward the end of a CC charge step, usually the last one at the end of formation) Finishing currents (CV or TC charge) Integrated ampere-hour input Presence of any timing kickouts and whether or not they were properly re-initiated Overall voltage-time or current-time formation profiles (sampling basis only) Discharge capacities if the formation algorithm includes a discharge
not always symmetrically-imposed on the DC signal and so there can be a net increase or decrease in the total charge applied (i.e., greater or lesser Ah formation inputs, respectively) depending upon whether it is skewed to the charge or discharge side. Given all of the above, it is highly recommended that formation power supplies be provided with the proper filtering equipment (capacitors and chokes) so that minimal AC ripple is fed to the forming VRLA batteries. It should be noted that the charging equipment available from Digatron/ Firing Circuits provides inductive filtering to minimize ripple current.
s s s
37
Initial currents or voltages are important because they indicate whether or not the initial current flow is at the prescribed level and that most or all of it is going into the formation process (conversion of lead sulfate to the active materials) and not into gassing because of high resistances. Low currents on CV or TC charge or abnormally high voltages on CC charge indicate that some or all of the unformed units are heavily sulfated and/or there may be connector/terminal problems (high resistances). Even though current is flowing the distribution between strings may not be uniform and if timed formation algorithms are used without current integration this will result in low Ah inputs to the batteries. Electrical energy that goes into resistive heating or gassing will, obviously, not count toward forming the active materials but it will be included in the total formation Ah input. Careful voltage monitoring during CC formation algorithms is useful as a feedback tool for triggering rest or discharge steps and for defining product quality; as with the other parameters discussed, TOC windows are defined for manufactured products and these values are used for accept/reject purposes. When voltages are low (when they should be high), this indicates that the batteries are being under formed; if voltages are in the gassing region (above ~2.35 V/cell) for appreciable periods of time weight losses and, possibly, grid corrosion will be high, battery internal temperatures will be elevated and dangerous over-formation of the positive plate is possible. If TOC voltages are unusually high (as they would be for a flooded lead-acid product), it indicates complete saturation of the plate stacks, which may not be desired for some products formed sealed with ~95% saturation levels. (This depends on whether the batteries are formed open or sealed, see sections 3.3.1 and 7.1). While it is often not possible to monitor every battery, compliance of the overall voltagetime or current-time profiles with manufacturing standard curves should be ensured. Non-compliance can indicate problems with either the formation equipment or the filled batteries going into formation or both. It is also important to monitor finishing currents during CV or TC charging, as they can indicate rectifier or power supply charging voltages being set too high or one or more strings in a series-parallel array being open. High finishing currents can also indicate that batteries are too hot or they have low saturation levels and, thus, high levels of oxygen recombination. In the extreme, high finishing currents can be a precursor to a part of the formation bay going into thermal runaway.
For CV and TC charging, integration of ampere-hour inputs is necessary to ensure that the proper amount of formation current is being applied; for CC charging, monitoring of the applied current level and the step times is sufficient to determine accurately the Ah inputs. A serious practical problem in many formation systems is proper re-initiation of the formation profile following an unscheduled power interruption. Ideally, the electrical system should pick up exactly where it left off but even when this is done the forming batteries will have experienced an unscheduled rest of some duration (or even two or more). If interruptions occur early in formation the effect is probably minimal but during overcharge the liquid diffusion and gassing processes will be significantly affected. If it occurs within the first hour or so, the formation schedule can simply be reinitiated. If it occurs in the last hour or so, early termination will probably not affect product quality severely. The bigger problems are if and when they occur in the middle of the formation, if they are extensive and if there are multiple outages. For manually controlled formation this can create confusion and errors. For computer-controlled systems proper programming can account for outages, but in any event the overall manufacturing schedule will be disrupted. However, if there is no compensation for outages the uniformity of product quality will be greatly affected, as each group of formed batteries may have widely different Ah inputs. If one or more discharges are part of the formation algorithm their duration and Ah output can be used as an indicator of the quality of the formed batteries; beyond this, the values can be used in matching VRLA batteries into larger arrays. In matching of modules to make higher-voltage (series) and ampere-hour (parallel) batteries it is a reliable rule-of-thumb that matching of like-capacity modules is of paramount importance more so than the nominal capacity. Thus, high- and lowcapacity modules must be matched with comparable partners rather than having them mixed. A battery comprised of all low-capacity modules will give better initial performance than one comprised of a mixture of high- and low-capacity units. Obviously, the best performance is obtained when all high-capacity modules are matched.
13.2 Temperature Monitoring

Temperature monitoring of the formation room environment, some of the formation equipment and the batteries themselves is important. VRLA batteries are very sensitive to the formation temperature, so if a formation room is being used without active heating or cooling (forced air, water) it is imperative that temperature be monitored at a number of independent positions throughout the room to ensure that a nominal formation temperature exists and that it is relatively uniform within a range of, at most, several degrees Centigrade. If the formation-room temperature is too high the batteries will be over-formed or, in the extreme, can go into thermal runaway. This will result in high gassing levels and weight losses, possibly high levels of grid corrosion and poor positive-plate quality. If the ambient temperature is too low under-formation can result, with low initial discharge capacities (particularly the positive plate) and high levels of unformed oxides (resulting in poor shelf life). In some VRLA products, if the initial formation is done at too-low temperatures it is difficult or impossible to effectively recharge the batteries following formation; the reason for this is not known. Both the nominal temperature for the formation room and consistency throughout the year are important, particularly if the manufacturing plant is not heated and air-conditioned. The worst condition is to simply form batteries on tables in a plant without temperature regulation of any type; this will result in variable product quality throughout the year and it is likely that batteries produced during either the summer or winter (or both) will be of inferior quality. It should be noted that critical formation equipment such as rectifiers, power supplies, power resistors, monitoring equipment and control computers must be located in a completely separate room from the formation room itself. The high temperatures and acidic atmosphere in the formation room can damage delicate electronic equipment and shorten their lifetimes considerably. As noted earlier (section 7.3) as a general rule small VRLA batteries (~25Ah or less) may form better if heated and larger batteries require cooling for optimal formation. In either case, appropriate regulation of the temperature immediately around the batteries (as opposed to the formation room ambient temperature) will result in superior and consistent product quality. Thus, temperature monitoring of forced air or still or circulated water used to heat or cool batteries, respectively, is also recommended.
If monitoring of air or water is done as described above then monitoring of battery surface temperatures is not necessary, as the temperature differences will be quite small. However, in cases where no active cooling or heating is applied (e.g., bench formation with no treatment) it is a good idea to monitor battery skin temperatures (selected batteries on a sampling basis, particularly center batteries in large groupings) in order to correlate this information with the formation algorithm timing and to ensure that batteries do not overheat or form too cold.
13.3 Gas Monitoring

In the later stages of formation, the forming batteries give off various gases, particularly when they are formed open. Early in formation, the positive plate goes into overcharge, generating significant amounts of oxygen gas. Vented oxygen poses no particular problems in formation rooms. Later, when negative plates go into overcharge hydrogen gas is given off. This is also not a problem unless it reaches a level of at least 4% by partial pressure; at this point and above it forms an explosive mixture with air in the presence of a spark source (which are usually abundantly present in formation environments). Other gases that are routinely generated during formation are carbon monoxide and carbon dioxide. In cases where battery internal temperatures reach levels of 160-190oC (this normally only occurs in a thermal-runaway situation), hydrogen sulfide can be generated on negative plates during heavy overcharge. Hydrogen sulfide is extremely toxic to humans, even at very low concentrations. Moreover, it attacks any copper or copper-coated electrical components and forms an insulating surface coating of copper sulfide, rendering the devices useless. While it is unusual in practice to have significant amounts of hydrogen sulfide generated, it is recommended that several types of gas monitors be installed in manufacturing formation rooms. At the very least, hydrogen monitoring should be employed for safety purposes. Monitors should be placed at locations in the formation room where hydrogen gas can be generated in large amounts or where it may accumulate. Clearly, this also calls for adequate air movement and ventilation to ensure that hydrogen gas buildup does not occur in pockets that may explode. Carbon dioxide and carbon monoxide are not likely to be generated in large quantities but CO monitors are common and inexpensive and it takes little investment to install several of these in a formation room.
39
Perhaps the most serious concern from a safety and equipment standpoint is the potential buildup of hydrogen sulfide, which is generated when VRLA batteries are seriously overheated during overcharge. Incidents are documented where telecom and UPS installations have been destroyed by VRLA batteries that generated large quantities of hydrogen sulfide on float when the systems went into thermal runaway [24]. Human toxicity is likewise severe, so it is highly recommended that hydrogen sulfide monitoring be a part of any VRLA manufacturing formation room. In most manufacturing operations, batteries in formation will occasionally explode, or pop, due to hydrogen being present internally during overcharge and a spark being created from a poor COS or squeeze weld contact. Such a defective connection will conduct current but at some point an arcing condition may occur. If hydrogen gas within the battery ignites there may be enough force to rupture the plastic case, creating plastic shrapnel near the battery. Clearly, this is a hazard for employees and by itself makes the wearing of safety glasses or, better, face shields mandatory. The frequency of this popping can be used as a quality indicator for the upstream cast-on-strap and/or throughthe-wall squeeze welding operations; if they occur frequently the welding equipment and procedures should be carefully inspected.
recharged). Multicell batteries which clearly have one or more dead cells (e.g., a 12V battery with an OCV of ~10V or 8V) should also be identified and scrapped. OCV monitoring directly after formation may not be feasible for single cells or small batteries, but for larger monoblocks it is useful in removing scrap batteries and thus minimizing future wasteful handling and storage.
s
For batteries formed open, completion of assembly is done at this stage. This may involve simply mounting one or more vent valves and caps and/or it may require heat sealing, gluing or ultrasonic welding of an outer cover. In any case, areas of the battery involving these operations must be dry and completely free of acid. Batteries are then put into a stable environment, possibly temperature-controlled, for several days until their electrical characteristics have stabilized and they can be sorted for future disposition.
14.1 Visual Standards

Many VRLA companies have a formal set of visual standards for manufacturing personnel to use in evaluating batteries for shipment following formation. Sample criteria that may be used for sorting are as follows:
s
14. Post-Formation Handling and In-Line Product Testing

When formation is completed, batteries are not necessarily ready for shipment immediately, although some manufacturers with very strong confidence in product quality (no doubt backed up by extensive sampling of formed batteries over a long period of time) do ship directly after formation. For most manufacturers, however, batteries from formation are subjected to the following general processes:
s
Category Category 1. Batteries with only minor defects that can be cleaned up and are acceptable after such treatment. Category Category 2. Batteries with cosmetic defects that will not affect performance and are acceptable electrically can be used in closed-case applications. Category Category 3. Batteries with major cosmetic damage but acceptable electrical performance can be reworked, if feasible. Category Category 4. Batteries with major cosmetic damage that may impact performance and/or lifetime must be scrapped.
Batteries are cleaned of any excess acid and inspected for physical case damage and possible corrosive attack on terminal posts or labels. Any batteries that clearly have leaks in seal areas (lid/ box, terminal posts) must be scrapped. Batteries immediately after formation will have very high open-circuit voltages (OCVs) due to trapped gas and excess surface charge on the plates. However, OCV readings directly after formation will identify dead batteries (to be scrapped) and those with low, but significant, voltages (to be
Each product must be evaluated and a series of visual standards covering all of the problems seen in postformation batteries must be developed to put batteries in one of the above four categories. Examples of common problems included in the above categories are: case/ label acid damage, poor seals/acid leakage, bulged or cracked cases, corroded terminals, improperly applied labels and deep scratches or dents.
40
14.2 In-Line Product Testing

Following cleaning and inspection after formation and storage for several days, batteries are then put through a series of tests in order to determine product quality and segregate them into several categories for further handling. The primary tool used is OCV measurement, but in some processes AC impedance and high-rate discharge performance are also measured. It must be stressed that all of these parameters change with time off formation, so in assessing product quality charts with listings of daily parameter levels should be employed (see below). Open-Circuit Voltage Measurement 14.2.1 Open-Circuit Voltage Measurement OCV values tend to drop on a daily basis by a few millivolts per cell depending upon the product type, the finished electrolyte strength and the effectiveness of formation. In addition, batteries with higher voltages after formation show a slower decrease in OCV values than those with lower (but acceptable) post-formation voltages. All batteries have a relatively steep drop in OCV over the first couple of weeks and this then flattens out to a lower rate. In addition, the rate of decrease in OCV is a function of temperature, so it is imperative that storage and measurement be done at a controlled temperature level, within a window of 3-4oC at most. Typical self-discharge curves for VRLA single cells are shown in Figure 22; multiples apply for multi-cell Figure batteries. OCV measurements are usually not simply a pass/fail affair, as batteries are often sorted into a number of categories, among these being (in order of increasing OCV):
s
Figure 22
Typical Self-Discharge Curves For VRLA Batteries.
A. Self-Discharge Curves for One- and Two-year Shelf Lives
2.2 2.15
Open-Circuit Voltage, V
2.1 2.05 2 1.95 1.9 1.85 1.8 0 5 10 15 Time on Stand, Months 20 25
B. Self-Discharge Curves At Different Storage Temperatures

100 90 80 70 60 50 40 30 20 10 0
% Nominal Capacity Available
0C 10C 20C 40C 10 20 30 40 50 60 70 Months of Storage from Full Charge 80
Dead, low voltage (e.g., <2.000V) or reversepolarity (all are reject category) Recharge category Low acceptable or high acceptable (to be subjected to further test and, if acceptable, shipping to customers) High-voltage (reject category)
s s
short circuit, or low fill weight; in any event, it is considered too dangerous and wasteful to attempt to recover these batteries and they are also scrapped. Recharge category batteries have low OCV values, but they are considered to be possibly recoverable as acceptable product. They are incompletely formed for some reason and often have relatively high internal impedances due to sulphation. Because of this they are often recharged in groups using high-voltage (3.5-4.0 V/ cell or more) constant-current charging for some period of time, after which they are subjected to the same inline test criteria as for the original formation. Some manufacturers will use their formation equipment for recharging, using a short, modified algorithm. In some manufacturing processes a second recharge treatment is allowed, but it should be recognized that with each
Dead batteries have either an internal open connection (COS, intercell weld) or they were not properly connected to the formation system during charge. Reverse-polarity batteries were formed in reverse and, thus, positives were formed as negatives and viceversa; both are scrap. Low-voltage batteries also may have had a poor connection in formation or they may have a near-fatal internal problem such as an incipient
41
follow-up charge treatment it becomes much more likely that these batteries contain some type of defect that may show up early in field service. Recharging, particularly multiple times, has the distinct drawback that, if effective, it can mask these subtle defects (such as small leaks, hydration shorts, etc) so that defective product is shipped. Manufacturers must be very careful in setting recharge procedures and standards for acceptance, as too-generous guidelines can and will result in higher levels of customer dissatisfaction and returns. The vast majority of batteries should normally be in the acceptable categories. If a companys batteries go into general use there may be only a single voltage window for product to be shipped immediately. However, if some of a companys markets involve the application of highvoltage systems or have high-performance requirements there may be two (or more) acceptable categories, one (A-group) being superior in terms of OCV to the other (B-group). In general, higher post-formation OCVs (A category as opposed to B) indicate a better-formed product and, thus, one with longer shelf life and superior electrical performance. However, some batteries can be and are over-formed (high-voltage, or HV, rejects), resulting in too-high weight losses and possible internal damage (enhanced grid corrosion, low saturation levels, overformed PAM, too-high electrolyte specific gravity). Clearly, if the formation process involves adjustment of electrolyte volumes after formation this type of defect is not as likely to be detected. For batteries formed in a sealed state this is particularly pertinent. In addition to possible internal damage, voltages may be so high for these batteries that they will tend to be undercharged in applications where they are mixed with normal OCV batteries in high-voltage strings. This will be particularly true in some float applications; in others, their high voltages may cause undercharging of the normal batteries and, in extreme cases, partial discharge. An example of a chart used in sorting batteries after formation is given in Table 3; this same Table can be used for sorting batteries after recharge or a short boost charge of any type. Obviously, any charging process will be the zero point for the chart, with days off form really reading as days off charge, whatever the type and duration of charge.
Measurements 14.2.2 AC Impedance Measurements In addition to OCV, the measurement of AC impedances of batteries can be a valuable sorting tool. Here, it is usually the case that a pass/fail value is used for each product, as batteries of different designs, voltages and Ah capacities will have different nominal impedance values. In batteries, single-cell impedances are additive depending upon the number of series-connected cells. Larger batteries (in terms of Ah capacity) and those with thinner plates have relatively lower impedances, as this parameter is a function of plate surface area and plate spacing. Typical impedance readings for a number of different VRLA single-cell and 12V products are given in Table 2. These are all thin-plate products (thicknesses of ~1.2mm or less) and so the AC impedance values will be lower than for comparable (in terms of voltage and Ah ratings) thicker-plate batteries. AC impedance is also a function of the state-of-health for a given VRLA product type and is an indicator of the quality of the materials and processes in use. The higher the impedance reading the poorer the battery quality (relative to the nominal value for that specific product). Thus, the nominal acceptable impedance reading is an allowable maximum value, with batteries above that value being sorted out for recharging (moderately high) or as scrap (very high). It should be noted that it is possible for VRLA batteries to have acceptable OCV values but unacceptably high impedance readings, as these two parameters have little in common electrically. There is no acceptable minimum value for impedance because it can never be too low. In practice, the lower the impedance the better the battery quality. Thus, average impedance readings for production batches of batteries are an indicator of the effectiveness of the manufacturing process and should be used by Quality and Manufacturing personnel as such. AC impedance values are usually taken at 1 kHz with a Hewlett-Packard Model 4328A meter or an equivalent instrument. As this involves a four-point measurement using delicate probes it is necessary to use great care and reproducibility in taking measurements on batteries, as simply varying the location of the probes on the battery terminals or applied measuring probe pressure can result in significant variances in measured values. It should be noted that AC impedance is a no-load test, so it says nothing about the capability of a battery to sustain certain discharge currents and/or the integrity of the internal battery connections. For this, an additional test is needed.
42
Measurements 14.2.3 High-Rate Discharge Measurements In addition to OCV and (possibly) AC impedance, some manufacturers also carry out a short-duration high-rate discharge test, usually at about the 10C discharge rate at room temperature. As this is not a complete discharge, variations in ambient conditions are not as critical as they would be for determining full capacities. The discharge time is for ~5-10 seconds, so a recharge is not necessary. This is a useful test, as it not only defines finished-product electrical quality but it also acts as a check on the integrity of the internal battery mechanical connections. It is often set up as only a pass/fail test, using a minimum voltage threshold at 5 or 10 seconds; any batteries whose voltages are at or above this voltage are accepted; those slightly below it may be put into an acceptable but inferior category for use in certain noncritical low-rate discharge applications; alternately, they may be sent back for recharge. In practice, it can also be used as a test to sort nominally acceptable batteries (i.e., those whose voltages are above the threshold value) into normal and superior categories, particularly for use in high-rate discharge applications such as engine start or UPS. This type of testing is somewhat subjective, as it depends upon the shapes of the discharge and voltage rebound curves, as shown in Figure 23. With Figure programmable testers, collection of data and analysis, even on such a detailed level, is not only possible but also commonplace. Figure Figure 23a shows a typical discharge and rebound curve for a strong battery, one with outstanding pore structures in the plates and good diffusion kinetics. On discharge, the coup de fouet (initial, instantaneous voltage drop) is followed by strong voltage recovery and an increasing voltage at the 5-second point, indicative of good electrolyte diffusion into the plate pores to support the high discharge current. Following termination of the discharge the voltage recovery is sharp, again suggesting effective diffusion kinetics. Figure 23b is an Figure example of the discharge and voltage recovery curve for an inferior battery. During the discharge the voltage is not strong and may even begin to drop off toward the end of the 5-second duration. The recovery is gradual, indicating restricted diffusion of acid from the separator reservoir into the plate pores. Such a battery will not be a strong performer in high-rate applications.
Figure 23
High-Rate Discharge Voltage/Time Curves For Acceptable And Unacceptable Battery Performance On A 5-Second Test.
A. Discharge/Rebound Curve Typical for a "Strong" Battery
Discharge Initiated
Test Voltage
Discharge Kickout
Note the strong voltage during discharge and the sharp voltage rebound after kickout.
Discharge/Rebound time, Seconds
B. Discharge/Rebound Curve Typical for a "Weak" Battery

Discharge Initiated
Test Voltage
Discharge Kickout
Note the weak voltage during discharge and the sluggish voltage rebound after kickout.
Discharge/Rebound time, Seconds
While post-formation testing and evaluation are not technically a part of the formation process, it is hoped that the foregoing will illustrate the necessity and power of this stage of manufacturing in validating the effectiveness of formation or catching and correcting its shortcomings. In some cases, it can also provide valuable information on processes earlier in the manufacturing stream (plate quality, COS integrity, heatseal effectiveness, etc.) This is an area that is often overlooked or given short shrift, but it is one that will have a major impact on the quality and uniformity of product reaching the end user.
43
Troubleshooting formation problems
15. Jar Formation Toubleshooting: Problems And Solutions

15.1 Filling and Fill-to-Form
1. Batteries overheat following acid filling
a. Chill unfilled batteries to 30oC (Caution: this process is energy intensive and the low temperature may cause damage to seal areas, and the battery case may be susceptible to cracking during handling) b. Chill electrolyte to 10oC c. Immerse filled batteries in chilled water bath d. Slow down filling speed
4. Electrical continuity check shows an open reading reading on a string or circuit circuit
a. Carefully inspect connectors for corrosion or oxidation b. Inspect connectors for broken wires c. Inspect wires coming from rectifiers/power supplies d. Measure individual batteries for open resistance readings (with and without connectors attached)
15.3 Formation Process

1. Batteries are too hot during formation are formation
a. Increase airflow or water circulation rate, decrease cooling water temperature b. Reduce ambient temperature conditions c. Reduce current levels, extend formation time d. Put in rest periods to dissipate heat e. Provide more space between batteries f. Lengthen fill-to-form time, reduce initial battery temperature g. Shorten time for high initial current charge, if used h. Change plate materials, use red lead in positive paste, carbon in negative paste i. Use sodium sulfate in electrolyte j. Shorten fill-to-form time to reduce sulfation level, lower battery resistance k. Some parallel strings receive too much current l. Use pulsed-current algorithm
2. Separator above plates is damaged during acid addition

a. Slow down fill speed b. Use acid-diffuser nozzle design
3. Lead sulfate hydration shorts form in the shorts form glass-mat separator glass-mat
a. Use sodium sulfate in electrolyte b. Use slower filling to yield more uniform acid distribution c. Keep temperature down during the fill-to-form time. d. As a last resort, put batteries onto formation immediately after fill
4. Conversion of PbSO4 to PbO2 is poor

a. Increase fill-to-form time
2. Batteries are too cold during formation are formation

a. Increase airflow rate, air temperature b. Increase cooling water temperature, slow circulation rate c. Increase ambient temperature d. Increase charge current, shorten formation time e. Some parallel strings receive too little current
5. Batteries regurgitate acid after filling regurgitate

a. Use cooling bath after filling b. Check dried plates for excessive carbonation
15.2 Pre-Formation Conditions

1. Batteries are too hot are
a. Allow longer fill-to-form time b. Use chilled water cooling between fill and formation c. Use a low-current initial charge for 1-2 hours d. Reduce cooling efficiency during fill, fill-to-form; extend filling and fill-to-form times
3. Formation ampere-hour input is too low Formation ampere-hour

a. Check for power outages during the formation period (did the formation programmer compensate?) b. Check rectifier/power supply setting - too low? c. Ambient temperature too low for CV or TC formation algorithms d. Cooling air or water too cold e. Check formation programmer for the correct formation algorithm f. Batteries heavily sulfated, current-time profile is low (CV or TC) g. Batteries under filled, resistances are high, currenttime profile is low h. Connector hardware is corroded, oxidized i. Battery terminals are corroded, oxidized
2. Batteries are regurgitating acid are regurgitating

a Form open instead of sealed b. Use external acid tower fitted to the fill port to take up expelled acid c. Modify cure/dry conditions to minimize plate carbonation d. Process batteries rapidly after cure and dry, minimize time on floor before fill and formation
3. Loading of formation bays takes too much time, formation fill-to-form time is high and variable ill-to-form
a. Use multiple acid fillers to speed up loading of bays b. Use smaller numbers of batteries on the formation circuits
44
4. Formation ampere-hour input is too high Formation ampere-hour

a. Check controller for correct algorithm current/voltage settings, timing b. Too-high rectifier/power supply setting c. Ambient temperature too high for CV, TC formation algorithms d. Cooling/heating air/water too warm e. Interrupted formation reinitiates from the beginning
9. Toxic, explosive gases are generated in large Toxic, are amounts, alarms sound alarms
a. Too much overcharge ampere-hours on the negative, high hydrogen levels b. Ventilation equipment inoperative, malfunctioning or under-designed c. Some formation bays have batteries in thermal runaway, hydrogen sulfide is being generated evacuate and shut down affected bays!
5. End-of-charge currents are too high in CV or currents are TC formation formation

a. Batteries are too hot due to poor temperature regulation b. Rectifier/power supply voltage too high c. Low fill weights in one or more cells and/or high formation weight losses results in high O2recombination level d. Impurities lower oxygen over potential on positive, increase recombination current e. Non-uniform distribution of current in a series-parallel string formation bay
10. Batteries explode during formation, may cause formation, plastic shrapnel plastic
a. Poor internal battery connections cause sparks during overcharge, H2 ignites, and cases are distorted or ruptured
11. Batteries explode during formation, catch fire formation, fire

a. Use flame-retardant case/lid materials b. Equip formation room with sprinklers, smoke detectors
6. End-of-charge voltages are too high in CC formation are formation

a. Batteries are flooded, no oxygen recombination taking place b. Too much electrolyte in negative-plate pores, acts like flooded battery c. End-of-charge current higher than the programmed level
15.4 Post-Formation Handling and Visual Inspection

1. Acid and/or water damage to batteries
a. Apply measures as in Section 16.3, Section 5 to avoid future acid damage b. Clean and dry battery thoroughly, examine terminals, labels for damage and scrap if severe
7. End-of-charge voltages (EOCV) are too low in CC are formation formation

a. Low fill weights in some cells and/or high formation weight losses create too much void space, therefore too much oxygen recombination, lowers EOCV b. Too little electrolyte in negative-plate pores, too much recombination c. High impurity levels reduce hydrogen over potential at the negative and/or the oxygen over potential at the positive d. Batteries are too hot, O2-reduction current increases, lowers EOCV e. End-of-charge CC current is lower than the programmed level
2. Acid damage to formation hardware formation hardwar dware

a. Clean thoroughly before re-use b. Replace steel, copper connectors if damage is severe
3. Oxidation damage to formation hardware formation hardwar dware

a. Clean thoroughly before re-use b. Replace if damage is severe
4. Exposed steel parts are rusted parts are

a. Clean with abrasive if minor
5. Terminal plating is corroded or burned away by acid Ter erminal corroded burned or shorting, respectively shorting, respectively
a. Scrap battery if severe; if minor use in closed-case applications
8. Batteries regurgitating acid, acid spray occurs regurgitating

a. Batteries too hot going into formation b. Batteries are too hot for the reasons given in Section 1 above c. Heavy, deep carbonation continues to generate CO2 from plates d. Voltages and/or currents are too high, gassing is excessive e. Attach acid tower to fill port to catch regurgitated acid f. Use all-glass separator, higher surface area separator, higher stack compression to hold electrolyte more tightly g. Battery headspace is inadequate h. Batteries are overfilled and/or void space is too low (high paste/grid weights) i. Equip formation room with sulfuric acid mist monitoring, alarms
6. Terminal posts are burnished or partially melted due Ter erminal are burnished partially to dead short short
a. Scrap battery
7. Minor scratches and/or dents to battery case battery

a. Rework and use as normal b. Train formation room personnel in better handling techniques
8. Major dent or cracked/chipped case that may indicate internal damage internal
a. Scrap battery
45
9. Acid leakage evident from lid/box seal, vent- valve from area or terminal post seal area. area terminal area.
a. Lid/box seal area scrap battery or rework for closed-case use b. Terminal post seal area see below c. Vent-valve area clean and dry
3. Recharge-category batteries (voltage is >LV but Recharge-category >LV below acceptable)

a. Slightly low acid fill weight b. Slightly inferior connector contact during formation c. Formed adjacent to dead battery in series-parallel matrix formation (less-than-normal Ah input) d. One or more cells may have an incipient hard short or hydration shorts in the separator e. Missing or damaged/leaking vent valve f. Large air leak due to cracked case or seriously defective seal g. Incomplete formation (high PbO level in PAM), poor acid penetration/no rests h. If all batteries in a bay are recharge category; check rectifier/power supply setting (may be low) i. If all batteries in one string of a series-parallel array are recharge category-low string current draw, other string(s) may be HV j. Ambient temperature low for CV or TC formation algorithm (all batteries affected) k. Formation kicked out prematurely, not reinitiated (all batteries affected) l. Battery terminals corroded or oxidized
10. Top and/or sides of battery are bulged visually Top battery are or to touch
a. If minor, hold for evaluation; if major, scrap battery
11. Dirt, oil and/or grease contamination Dirt, grease

a. Clean thoroughly and use
12. Acid around terminal post(s) (may be simple acid around terminal contamination during formation or a defective post seal) formation
a. Clean away acid and put battery on a 2- or 3-hour overcharge b. If no leakage, accept as good product; if leakage occurs during overcharge, scrap battery
15.5 Post-Formation Electrical Evaluation OCV, Impedance, High-Rate Discharge.

1. Battery is dead zero voltage Battery zero
a. Internal open connection in battery b. No fill acid c. No connector contact on formation d. Battery was short-circuited during handling (terminal damage will identify this) e. Battery is part of a whole string that received no formation current f. For a-d, batteries must be scrapped; for e batteries may be put back through formation although they may not form properly and it is safer to scrap them also.
4. High-voltage (HV) batteries, >~2.2 V/cell

a. String draws too-high current in series- parallel array; all batteries in string are HV b. All batteries in a bay are HV; check for high rectifier/ power supply setting c. Formation kicked out; reinitiates from the beginning (too-high Ah input, high weight losses for all batteries in a bay) d. Ambient temperature high for CV or TC formation algorithm, high Ah input (all batteries affected)
2. Low-voltage (LV) battery (<2.000 V/cell) (LV) battery

a. One or more cells shorted out b. Low acid fill weight c. Poor connector contact on formation d. Battery was partially short-circuited during handling (terminal damage) e. Battery is part of a string that received very low formation current f. Battery is poorly formed, went onto formation heavily sulfated g. Early formation kickout malfunction not reinitiated (all batteries affected) h. The usual procedure is to scrap any LV battery
5. High impedance readings readings

a. Fill weights too low b. Batteries over-formed, high weight losses (probably HV category also) c. Incomplete formation (dead, LV or recharge category also) d. Poor plate processing, usually at cure/dry e. Meter probes too high on terminals f. Meter probes corroded or oxidized g. Impedance meter out of calibration h. Battery terminals corroded or oxidized
46
6. High-rate discharge (HRD) voltage reading too low reading

a. Battery over-formed or incompletely formed, high impedance creates large voltage drop on HRD (impedance reject) b. Poor battery internal connections (affects HRD but not impedance measurements- impedance is a no-load test) c. HRD tester probes and/or battery terminals corroded or oxidized d. Poor connection of tester probes to battery terminals e. Discharge current set too high or HRD tester out of calibration f. Poor electrolyte distribution between separator and plates (too little in plates) g. Low active-material porosities, restricted electrolyte diffusion h. Plate stack compression too low, results in poor separator/plate contact i. Grid-PAM passivation layer present
3. Excessive grid corrosion after formation corrosion formation

a. Poor filling process created watery or dry areas, rampant alkaline corrosion occurred in these areas (the grid or areas of it may be completely corroded after formation) b. Dry areas in separator occur during filling due to tight fiber structure, trapped air (very localized), again alkaline corrosion occurs c. Low negative-plate potential pushes positive potential into high-corrosion region in CV formation (occurs over entire plates) d. Formation Ah input and/or temperature are too high e. Impurities present (chloride, organic compounds that break down to form acetic acid) catalyze extensive grid corrosion
4. High levels of unformed oxide, particularly in the unformed particularly positive plate
a. Low Ah input, batteries under formed b. High percentage of formation charge goes into heat production, gassing, grid corrosion c. Heavy gassing doesnt allow acid, water penetration into plate interiors during formation (use rest periods) d. Dense lead sulfate layer on plate surfaces (too-fine oxide, heavy calendering during pasting, too long fillto-form time) inhibits fluid penetration e. High orthorhombic PbO level in the positive paste creates an impervious structure for liquid penetration
15.6. Battery Tear-Down and Analysis

1. Negative plate shows areas of white lead sulfate areas
a. Insufficient formation Ah input b. Fill-to-form time too long, too much lead sulfate formed (high resistance) c. Long fill-to-form time, no carbon black in expander d. High impurity level, lowers hydrogen over potential, negative cant be formed e. Too much oxygen recombination due to low fill weight and/or high formation weight loss (too much void space in the separator) f. p:n value too low (<~0.8), negative- plate sulfate cant all form out g. Solutions: use high-purity materials, increase carbon black amount, use p:n ratio >/=1.0 h. CV formation voltage and/or temperature too low
5. High alpha:beta-PbO2 ratio in PAM PAM

a. Increases with >temperature b. Decreases with >acid density c. Decreases with >current density d. Decreases with increasing amount of positive-paste sulfate content e. Generally, lack of acid penetration increases alphaPbO2 percentage f. High alpha-PbO2 levels form in areas of the plate stack where the fill electrolyte is dilute; most prevalent in thin-plate products, pastes with little or no sulfation (high density)
2. Positive plate shows areas of white lead sulfate areas

a. Insufficient formation Ah input b. Fill-to-form time too long, as in 1b c. High impurity level, lowers oxygen over potential, positive material cant be formed d. High Ah input into grid corrosion, doesnt go into lead sulfate conversion e. High negative-plate voltage (flooded?) doesnt allow sufficient positive-plate polarization to complete CV formation f. CV formation voltage and/or temperature too low
47
References
16. References
[1] G. Zguris, Batteries International, April, 1991, pp. 80-81 [2] H. Chen, et al. Journal of Power Sources 59 (1996), pp. 59-62. [3] C. Ressel, Batteries International, April, 1990, pp. 24-25 [4] Exide Technologies, First Semi-Annual Report, March 2001. [5] R.F.Nelson, The Battery Man, May 1998. [6] R.F.Nelson, Batteries International, 36, July 1998, 95-103. [7] K. Matthes, B. Papp and R.F. Nelson, Power Sources 12, T. Keily and B. Baxter, eds., 1988, p.1. [8] A.Ferreira, J.Power Sources 78 (1999), 41-45 [9] The Battery Man, January 1995. [10] J.Power Sources 67 (1997). [11] A.Ferreira, Batteries International, 46, Jan 2001, 43-50 [12] J.Power Sources 73 (1998) 60. [13] M.J.Weighall, ALABC Project No. R/S-001, Final Report, October 2000. [14] CSIRO, ALABC Project S1.1, Progress Report 1, JulyDecember 2000. [15] R.F.Nelson, Batteries International, 43, April 2000, 51-60 [16] R.F.Nelson, Batteries International, 34, Jan 1998, 87-93 [17] G.Zguris, 16th Annual Battery Conference on Applications and Advances (2001), 163-168. [18] Hovosorb Technical Manual, April 1993. [19] M.J.Weighall, ALABC Project No. R/S-001, Final Report, October 2000, 44. [20] A.Ferreira, Batteries International, 36, July 1998, 83-90. [21] Dr D.Berndt, Oxygen Cycle Meeting, 7ELBC, Dublin, Ireland, 19th September 2000 [22] P.R. Stevenson and O. Enoki, Proceedings of the 5th ERA Seminar, London, U.K., April, 1988 Paper 3.2 [23] R.F. Nelson and M. A. Kepros, Proceedings of the 14th Long Beach Battery Conference, IEEE, 1999, pp. 281287 [24] R.S. Robinson and J.M. Tarascon, J. Power Sources 48 (1994), 277-84
Table 2. Typical AC Impedance Values For A Variety Of Thin-Plate VRLA Single Cells And Batteries Fully Charged At 25C.
2V/2.5Ah Cell 6V/2.5Ah Battery 2V/5.0Ah Cell 6V/5.0Ah Battery 2V/25Ah Cell 12V/25Ah Battery (Hawker SBS) 12V/25Ah Battery (Hawker Genesis) 12V/35Ah Battery (Hawker SBS) 12V/36Ah Battery (Hawker Genesis) 6V/100Ah Battery 5.0 milliohms 15.0 milliohms 3.5 milliohms 10.0 milliohms 1.5 milliohms 7.0 milliohms 5.0 milliohms 5.5 milliohms 4.5 milliohms 1.8 milliohms
By comparison, thicker-plate VRLA products have the following published impedance values:
12V/2Ah Yuasa Battery 2V/2.6Ah Portable Energy Products Cell 2V/4.3Ah Portable Energy Products Cell 80 milliohms 30 milliohms 24 milliohms
48
References
Table 3. Sample OCV Chart Used in Manufacturing to Sort Cells or Batteries After Formation or Recharge
Days Off Form 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 HV >2.199 >2.196 >2.194 >2.192 >2.190 >2.188 >2.186 >2.185 >2.184 >2.183 >2.182 >2.181 >2.180 >2.179 >2.178 >2.177 >2.176 >2.175 >2.174 >2.173 >2.172 >2.172 >2.171 >2.170 >2.169 >2.168 >2.168 >2.167 >2.166 >2.165 >2.164 >2.164 >2.163 >2.163 >2.162 >2.162 >2.161 >2.160 >2.160 >2.159 >2.159 >2.158 >2.157 >2.157 >2.156 >2.156 >2.155 >2.155 >2.154 >2.154 >2.154 >2.153 >2.153 >2.153 >2.153 >2.152 >2.152 A 2.199 - 2.170 2.196 - 2.167 2.194 - 2.165 2.192 - 2.163 2.190 - 2.161 2.188 - 2.159 2.186 - 2.157 2.185 - 2.156 2.184 - 2.155 2.183 - 2.154 2.182 - 2.153 2.181 - 2.152 2.180 - 2.151 2.179 - 2.150 2.178 - 2.149 2.177 - 2.148 2.176 - 2.147 2.175 - 2.146 2.174 - 2.145 2.173 - 2.144 2.172 - 2.143 2.172 - 2.143 2.171 - 2.142 2.170 - 2.141 2.169 - 2.140 2.168 - 2.139 2.168 - 2.139 2.167 - 2.138 2.166 - 2.137 2.165 - 2.136 2.164 - 2.135 2.164 - 2.135 2.163 - 2.134 2.163 - 2.134 2.162 - 2.133 2.162 - 2.133 2.161 - 2.132 2.160 - 2.131 2.160 - 2.131 2.159 - 2.130 2.159 - 2.130 2.158 - 2.129 2.157 - 2.128 2.157 - 2.128 2.156 - 2.127 2.156 - 2.127 2.155 - 2.126 2.155 - 2.126 2.154 - 2.125 2.154 - 2.125 2.154 - 2.125 2.153 - 2.124 2.153 - 2.124 2.153 - 2.124 2.153 - 2.124 2.152 - 2.123 2.152 - 2.123 Categories B 2.169 - 2.140 2.166 - 2.137 2.164 - 2.135 2.162 - 2.133 2.160 - 2.131 2.158 - 2.129 2.156 - 2.127 2.155 - 2.126 2.154 - 2.125 2.153 - 2.124 2.152 - 2.123 2.151 - 2.123 2.150 - 2.123 2.149 - 2.123 2.148 - 2.123 2.147 - 2.123 2.146 - 2.123 2.145 - 2.123 2.144 - 2.123 2.143 - 2.123 2.142 - 2.123 2.142 - 2.123 2.141 - 2.123 2.140 - 2.123 2.139 - 2.123 2.138 - 2.123 2.138 - 2.123 2.137 - 2.123 2.136 - 2.123 2.135 - 2.123 2.134 - 2.123 2.134 - 2.123 2.133 - 2.123 2.133 - 2.123 2.132 - 2.123 2.132 - 2.123 2.131 - 2.123 2.130 - 2.123 2.130 - 2.123 2.129 - 2.123 2.129 - 2.123 2.128 - 2.123 2.127 - 2.123 2.127 - 2.123 2.126 - 2.123 2.126 - 2.123 2.125 - 2.123 2.125 - 2.123 2.124 - 2.123 2.124 - 2.123 2.124 - 2.123 2.123 - 2.123 2.123 - 2.123 2.123 - 2.123 2.123 - 2.123 2.123 - 2.123 2.123 - 2.123 R <2.140 <2.137 <2.135 <2.133 <2.131 <2.129 <2.127 <2.126 <2.125 <2.124 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 <2.123 LV-Dead <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 <1.999 www.firing-circuits.com 49
Appendix
APPENDIX 1 Glossary Of Terms And Abbreviations

VRLA SLI UPS V I R Ah OC OCV TOCV EOCV LV HV CC CV TC PC DC AC HRD Pb PbO PbO2 Pb3O4 H3PO4 CO CO2 3BS/ TRB 4BS/ TTB NAM PAM SEM BET XRD S.G./ s.g. H&V AGM RBSM MFG L/d ratio Kpa CFM COS Valve Regulated Lead Acid Starting, Lighting, Ignition Uninterruptible Power Supply Voltage Current Resistance Ampere Hour Open Circuit Open Circuit Voltage Top Of Charge Voltage End Of Charge Voltage Low Voltage High Voltage Constant Current Constant Voltage Taper Current Pulse Current Direct Current Alternating Current High Rate Discharge Lead Lead Monoxide Lead Dioxide Red Lead Phosphoric Acid Carbon Monoxide Carbon Dioxide Tribasic Lead Sulfate Tetrabasic Lead Sulfate Negative Active Material Positive Active Material Scanning Electron Microscope Technique for surface area determination X-Ray Diffraction Specific Gravity Hollingsworth and Vose Absorptive Glass Mat Recombinant Battery Separator Mat Microfine Glass Plate height/ Plate spacing ratio Kilo Pascals Cubic Feet per Minute Cast on Strap
50
= Factory Locations = Representatives

Providing World-Wide Support for our Customers
Digatron/Firing Circuits, Inc. 230 Long Hill Cross Road Shelton, CT 06484, U.S.A Tel: 1 203 446 8000 Fax: 1 203 446 8015 E-Mail: info@firing-circuits.com www.firing-circuits.com An ISO 9001 company Digatron Hong Kong Co. Flat D, 3/F Cheerrick Bldg. 65 Sau Fu Street Yuen Long, Hong Kong Tel: ++852 2478 1011 Fax: ++852 2474 5060 E-Mail: digatron@hongkong.com Digatron (Qingdao) Electronic Co., Ltd. 8 Keyuanjing 3 Road, Hi-tech Zone 266101 Qingdao, China Tel: ++86 532 8870 5292 Fax: ++86 532 8870 5259 E-Mail: info@digatron.com.cn Digatron Industrie-Elektronik GmbH Tempelhofer Strasse 12-14 52068 Aachen, Germany Tel: ++49 241 168090 Fax: ++49 241 166465 E-Mail: info@digatron.de www.digatron.de

A Guide To VRLA Battery Formation Techniques

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Guide To VRLA Battery Formation Techniques

Uploaded by

Copyright:

Available Formats

A Guide To VRLA Battery Formation Techniques

A Guide To VRLA Battery Formation Techniques

2001 Firing Circuits, Inc.

Printed In U.S.A., All Rights Reserved

The brochure is divided into several sections:

Practical issues related to VRLA battery and jar formation

2. Plate formation vs. jar formation for VRLA batteries

Practical issues related to VRLA battery and jar formation

3. The VRLA Formation Process Jar Formation

Practical issues related to VRLA battery and jar formation

3.2 Fill-to-Form Processing

Practical issues related to VRLA battery and jar formation

3.4 Formation Time

Practical issues related to VRLA battery and jar formation

3.5 Completion of formation

3.6 Formation Algorithms

Constant-voltage (CV) Constant-current (CC) Taper-current (TC) Pulsed-current (PC)

Additional information is given in section 7.4

Practical issues related to VRLA battery and jar formation

3.7 Initiation of Formation Current Flow

Formation time, hours

Formation time, hours

C. Abrupt or high-current initiation

Formation time, hours

Practical issues related to VRLA battery and jar formation

3.8 Constant-Voltage Charging

B. Multi-Step Current-Limited Constant-Voltage Formation Profile

3.9 Constant-Current Charging

Practical issues related to VRLA battery and jar formation

Ideal current curve Several-step current curve

Current curve of conventional formation

3.10 Taper-Current Charging

Practical issues related to VRLA battery and jar formation

There are also a number of drawbacks, as follows:

3.11 Pulsed-Current Charging

= Current = Voltage = Battery Temperature = Coolant Temperature

Practical issues related to VRLA battery and jar formation

Power Supply Voltage - Cell Voltage (V) Load Resistor ( RTaper )

Typical TC Current/Voltage/Ah Curves

3.12 Rests and Discharges

Constant Period, Decreasing Amplitude

Constant Amplitude, Decreasing Period

Practical issues related to VRLA battery and jar formation

Formation Current, Amperes

B. Same Curves for a High-Inrush Two-Step CC Formation Algorithm

3.13 Sample Formation Algorithms & Profiles

Voltage, Temperature, Gassing Rate

Formation Current, Amperes

Practical issues related to VRLA battery and jar formation

12 16 20 24 Formation Time, Hours

B. Same Curves for a One-Step Taper-Current Formation Algorithm

12 16 20 24 Formation Time, Hours

= Current = Voltage = Temperature = Gassing Rate

Practical issues related to VRLA battery and jar formation

16 20 24 28 Formation Time, Hours

B. Same Curves for a CC/Discharge/Recharge Formation Algorithm

-8 0 2 4 8 12 16 20 24 28 Formation Time, Hours 32 36

3.14 Development of a Suitable Formation Algorithm

Practical issues related to VRLA battery and jar formation

4. Temperature Limits For VRLA Jar Formation

Practical issues related to VRLA battery and jar formation

Additional information is given in section 7.3.

5. VRLA battery manufacture using PLATE FORMATION