AU5485099A

AU5485099A - Separator for lead-acid cells or batteries

Info

Publication number: AU5485099A
Application number: AU54850/99A
Authority: AU
Inventors: Michel M. Lawrence
Original assignee: GNB Technologies Inc
Current assignee: GNB Inc
Priority date: 1998-08-18
Filing date: 1999-08-17
Publication date: 2000-03-14
Also published as: EP1031170A1; WO2000011746A8; WO2000011746A1; JP2002523880A; CA2306691A1; KR20010031203A

Description

WO 00/11746 PCT/US99/18499 SEPARATOR FOR LEAD-ACID CELLS OR BATTERIES FIELD OF THE INVENTION 5 This invention relates to lead-acid cells and batteries and, more particularly, to separators used in making such cells and batteries. BACKGROUND OF THE INVENTION A wide variety of applications, often termed "industrial battery" 10 applications, utilize conventional, flooded electrolyte lead-acid cells and batteries, or sealed lead-acid cells and batteries, often termed VRLA cells and batteries ("valve-regulated lead-acid"). In stationary battery applications, the lead-acid cells and batteries provide stand-by power in the event of a power failure. For this type of application, such cells and batteries are maintained at a full state-of-charge 15 and in a ready-to-use condition, typically by floating at a constant preset voltage. Stationary batteries are used for stand-by or operational power in a wide variety of applications, including, by way of illustration, telecommunications, utilities, for emergency lighting in commercial buildings, as stand-by power for cable television systems, and in interruptible power supplies for computer back-up 20 power and the like. Other applications in which lead-acid cells and batteries may be used involve a. variety of motive power applications in which an array of cells or batteries provides the motive power for vehicles ranging from Class 1 to Class 3 trucks, various automated guided vehicles, mining vehicles and also railroad 25 locomotives. The performance requirements for motive-power vehicles are quite different from the performance requirement for stationary power sources. In stationary power applications, the depth of discharge in service is relatively shallow, and the number of discharges is smaller, as most batteries are in float service. In direct contrast, motive power applications require a relatively deep 30 depth of discharge to be achieved on a continuous cycling basis over a period of time. Indeed, a common requirement for Class 1-3 trucks is that, in an 8-hour WO 00/11746 PCT/US99/18499 2 shift, the cell or battery assembly must be capable of delivering an 80% depth of discharge and that performance is required for 300 cycles per year with a useful service life under those conditions of 4 or 5 years. The widely varying requirements for these many applications have 5 presented substantial problems and an extremely challenging environment for manufacturers of lead-acid cells and batteries. This environment has resulted in, to a large extent, custom designs which satisfy particular applications. As a consequence, lead-acid cell/battery manufacturers have had to develop families of cells and batteries in an attempt to satisfy the diverse electrical 10 performance criteria. Such criteria vary widely, often requiring large cells connected in parallel, series, or both, to provide a satisfactory power/energy source. The space requirements often are also quite constricted, with closely defined dimensional requirements. Many types of steel trays and the like are used 15 to house the cells required. To achieve the family of cells and batteries requires grids of various sizes so that the capacity and other electrical performance requirements for an individual cell for a particular application can be satisfied. One approach utilized for VRLA cells has been to provide a series of grids having essentially constant 20 width while varying the height of an individual grid and the number of plates used in a particular cell to achieve a variety of capacity of other electrical performance requirements. While an effective solution, this approach does create challenges that have to be addressed, as will be discussed hereinafter. The internal configuration of as such VRLA cells can vary widely. In 25 general, such cells are disclosed in U.S. 3,362,861 to McClelland et al. As is thus known, such cells utilize highly absorbent separators; and the necessary electrolyte is absorbed in the separators and plates. Such cells are normally sealed from the atmosphere by a valve designed to regulate the internal pressure within the cell so as to provide what is termed an effective "oxygen recombination cycle" (hence, 30 the use of the terms "sealed" and "valve-regulated" as well as "recombinant").

WO 00/11746 PCT/US99/18499 3 Recombinant battery separator materials (sometimes termed "RBSMs") have traditionally comprised a highly absorbent glass fiber mat. Separators of this type have adequate absorbency to hold the amount of electrolyte desired and possess some vacancy of pores to allow the oxygen recombination cycle to 5 proceed. A wide variety of suitable glass fiber mats are commercially available and are in use in VRLA cells and batteries. Despite the widespread use of such glass fiber mats, substantial efforts have been made to develop other recombinant battery separator materials, perceived to satisfy varying objectives. U.S. 4,908,282 to Badger summarizes 10 many different prior art attempts to provide satisfactory separator materials for recombinant cells and batteries. Yet, Badger states that there has not previously been a suggestion of a separator which, when saturated with the electrolyte, leaves a residuum of unfilled voids through which a gas can transfer from one plate to another because the separator is not capable of holding an amount of electrolyte 15 which is sufficient to fill all the voids (col. 2, 11. 20-26). More particularly, Badger discloses a separator having, in general, two types of fibers. A first set of fibers imparts to the separator an absorbency greater than 90% relative to the electrolyte and a second set of fibers that have a different absorbency which is less than 80% relative to the electrolyte. The first and second 20 fibers are disclosed as being present in proportions such that the absorbency of the overall separator is from 75-95%. Specifically, a separator is disclosed which is made of a mixture of two different grades of glass fibers, one grade of chopped glass strand and a certain grade of polyethylene fibers. Another prior art attempt to provide a RBSM is U.S. 4,216,280 to Kono et 25 al. The '280 patent discloses separators which comprise glass fibers entangled in the shape of a sheet without the use of a binder and have a first and second portion of glass fibers. The first portion comprises glass fibers having a fiber diameter smaller than one micron; and a second portion uses glass fibers having a fiber diameter larger than 5 microns, as well as an average fiber length of at least 5 30 millimeters. Such separators are stated to have high electrolyte retention, good mechanical strength, and good shape recovery.

WO 00/11746 PCT/US99/18499 4 Yet another prior art attempt to provide RBSMs is U.S. 4,367,271 to Hasegawa et al. By way of background, the '271 patent thus states that one prior proposal comprises a glass fiber mixed with a synthetic resin serving as a binding agent, while another type proposed involves mixing a glass fiber with a synthetic 5 resin monofilament fiber. Hasegawa et al. state that such prior approaches are inadequate because these approaches suffer a remarkable decrease in liquid absorption and that the improvement in the mechanical strength is small. The '271 patent is said to provide a separator which is high in liquid absorption, high in strength, and is easy to handle. Such separator materials, according to the '271 10 patent, are produced by a process which uses glass fiber substantially 1 m 2 /g or more in specific area, mixed with about 10% or less, by weight, of fibril-formed synthetic fibers which have 350 cc or less in "freeness." With all of this prior effort in this field, there still exists a need for such separator materials which can overcome the significant problems that arise during 15 the service life of VRLA cells and batteries, particularly when the required service life is relatively long. First of all, as has been previously noted, one approach in this field provides a family of cells and batteries which utilizes grids of a constant width while the height of the grid is varied as well as a number of plates to provide the desired capacity and other electrical performance requirements. In cells of this 20 type, and, indeed, in many cells that have a relatively large height, cell "dry out" can become a prevalent problem, particularly, in industrial cells designed for a relatively long service life, e.g., 10 or 20 years or so. Such "dry out" can result from electrolyte loss, uneven cell saturation, RBSM pull away or a combination. Among the reasons for occurrence of "dry out" may include lack of appropriate 25 resilience of the RBSM and/or formation of saturation differential from cell top to cell bottom. In addition, electrolyte stratification can cause a variety of problems such as sulfation of the negative plate and uneven usage of the plate from top to bottom. Another problem, which perhaps may be related to cell dry out, is the loss 30 of appropriate cell compression. More particularly, as is known, providing satisfactory electrical performance in VRLA cells requires intimate contact WO 00/11746 PCTIUS99/18499 5 between the cell plates and the separators. Such contact is typically provided by compressing the separators by as much as 20% or more (based on their uncompressed thickness) in the cell so as to attempt to maintain satisfactory contact throughout cell life. 5 Thus, as far as can be perceived, and despite the many prior attempts to provide improved RBSMs, there appears to be a fundamental lack of understanding as to how to coordinate the many and diverse requirements for such materials so as to provide a satisfactory solution. This seems particularly true for RBSMs where the anticipated service life is expected to be 10 years or more. 10 Accordingly, there exists a clear need for both an understanding as to how to assess the relative efficiency of RBSM candidates, as well as providing such materials capable of imparting to VRLA cells and batteries improved performance over the anticipated service life. It is accordingly a principal object of the present invention to provide 15 sealed lead-acid cells and batteries utilizing separators capable of enhancing the electrical performance over the service life of such cells and batteries. A further and more specific object provides separators for large VRLA cells and batteries capable of achieving enhanced resistance to electrolyte stratification over the service life. 20 Yet another specific object of this invention lies in the provision of such sealed cells and batteries having separators with a preselected separator springiness so as to provide improved performance over the service life of such cells and batteries. Another object of the present invention is to provide separators for such 25 sealed lead-acid cells and batteries exhibiting enhanced mechanical strength so as to facilitate cell and battery assembly. A still further object lies in the provision of test methods useful for selecting suitable separate materials for VRLA cells and batteries. Other objects and advantages of the present invention will become apparent 30 as the following description proceeds. While the present invention will be described primarily with respect to use in sealed lead-acid cells, it should be WO 00/11746 PCT/US99/18499 6 appreciated that the present invention can be advantageously used in any other application where separators of the type disclosed may find utility. SUMMARY OF THE INVENTION 5 The present invention is, in general, predicated on the discovery that enhanced performance over the service life of sealed lead-acid cells and batteries can be achieved by utilizing separators having preselected properties and compositions. More particularly, it has been found that the coordination of particular properties, in a particular manner, will allow separator materials that 10 will achieve VRLA cells and batteries having improved performance over their expected service life. It has thus been found that superior separator materials are achieved when such materials combine preselected porosity and pore size characteristics and basis weights with preselected springiness and saturation and stratification characteristics. 15 BRIEF DESCRIPTION OF THE DRAWING FIGURE 1 is an isometric view of a lead-acid cell of the present invention, the cell jar being partially broken away to show the internal components. 20 DETAILED DESCRIPTION OF TE INVENTION FIGURE 1 illustrates an exemplary lead-acid cell in accordance with the present invention. The cell 10 has a container or jar 12 containing a plurality of positive and negative plates 14 and 16, respectively. As illustrated, the cell contains plural positive and negative plates. Of course, the cell can utilize the 25 necessary number of plates to provide the capacity and other electrical performance characteristics desired for the particular application. In accordance with the present invention, the plates 14, 16 are separated by absorbent separators 18 having the characteristics preselected as will be discussed hereinafter. In the preferred embodiment, the separator extends slightly past the 30 electrode to prevent an inadvertent short circuit of the cell. In addition, the WO 00/11746 PCT/US99/18499 7 separator may be folded around and between the plates by employing a U-fold 20, as illustrated in FIGURE 1. The plates 14, 16 preferably fit snugly within the container 12, that is, the electrodes and separators should stay in the assembled condition when the 5 container is inverted. Indeed, as is known, the cell configuration should ensure that the plates and separators maintain adequate compression and good contact so as to enhance the electrical performance of the cell. Preferably, as illustrated in FIGURE 1, the separators and plates are compressed so as to be in intimate contact with one another. Suitably, adequate compression can be achieved by 10 ensuring that the thickness of the separator in the cell is compressed typically by at least about 20% to 25% or so of the uncompressed thickness. The plates are connected to one another by conductive straps 22 and to external terminals, 24, 26, by conventional means. The thickness of the plates will vary depending upon the application to 15 which the cell is intended. An illustration of a useful range is from about 0.050 inch to about 0.300 inch, or even more. Preferably, the container is normally sealed from the atmosphere in use to provide an efficient oxygen recombination cycle as is known. The container utilized should be able to withstand the pressure of the gases released during 20 charging of the cell. Pressures inside the container may reach levels as high as, for example, 0.5-10 psig. Release venting is provided by a low pressure, self resealing relief valve, such as for example, a valve 28. An example of such valve is illustrated in U.S. 4,401,730. An electrolyte is also included within the container 12. Preferably, the 25 electrolyte is absorbed within the separator in the positive and negative active material. The electrolyte typically is sulfuric acid having a specific gravity in the range of, as an example, about 1.270 to about 1.340 or even more, as is considered appropriate for a particular application. The size of the plates can vary depending upon the necessary electrical 30 performance requirements for the cell. For example, conventional sizes of WO 00/11746 PCT/US99/18499 8 positive plates are rated ranging from 58 ampere-hours (AH) to 98 AH, 108 AH, 188 AH, and even greater. As illustrative examples, the height of the cell can exceed 28 inches. Indeed, the cells for many applications, require separators having thicknesses of 5 from about 0.04 to about 0.135 inches and up to 28 inches in height, or perhaps even more. As may be appreciated, consistent with utilizing separators having the preselected characteristics of the present invention, the internal configuration of the sealed lead-acid cells and batteries can be varied as desired. A wide variety of 10 types of sealed lead-acid cells and batteries are known and may be used. What has been found pursuant to the present invention is that certain characteristics and properties of the separators need to be preselected to provide enhanced performance during the desired service life. By the terms "preselecting" and "preselection," it is meant that the designated properties of the separator are 15 determined prior to assembly of the sealed cell or battery and that the separator material utilized in the cells is included with knowledge of such properties. Such knowledge can be obtained through testing by either the separator manufacturer or user. Also, as may be appreciated, there is no need to re-test for the desired preselected properties, once reproducibility in the process parameters has been 20 determined. One aspect of the present invention provides a test technique that adequately evaluates the springiness characteristics of RBSMs so that the preselected springiness characteristics will allow enhanced performance over the service life. More particularly, a load cell test has been developed which measures 25 the decay in the springiness characteristic over time in a continuous fashion so as to predict the ability of the RBSM to maintain compression within a cell over the service life. The load cell test utilized in the present invention measures the amount of pressure exerted by the separator sample (having adequate electrolyte absorbed 30 within the test sample to achieve 95% saturation and an initial compression of WO 00/11746 PCT/US99/18499 9 20%) against a load cell. A graph of the pressure versus time can be generated. Over time, the pressure exerted will level off. According to the present invention, useful separator materials should have a load cell pressure at two months of at least about 6.0, preferably at least about 5 6.5 or 7.0, and, even more preferably, at least about 8.0. The test sample is prepared by cutting a 3"x3" square. The actual thickness of the test sample is determined at any three points using a TMI gauge and the BCI test procedure (BCI Spec. No. 3-006 through 9) for determining the thickness of a recombinant mat. The sample is then weighed to the nearest milligram (Wsep). 10 The amount of 1.31 sulfuric acid to achieve 95% saturation is determined as follows: (1) Helium pycinometry is used to determine the true density of the test sample, Dhp; (2) The basis weight, bw, is determined by using the BCI 15 procedure for recombinant mats (Method A); (3) The density, Da, of the material in the compressed state (i.e., in the battery application), is determined by dividing the basis weight by the sample thickness at 20% compression; (4) The value for 100% separation at 20% compression (K) is 20 then derived as follows: K = Dacid 1 grams acid . aid a Dhp grams separation wherein Dacid is the specific gravity of the acid and (5) The amount of acid to achieve 95% saturation,

W

1 , is then determined as follows: 25 Wi = Wsep x K x 0.95 The amount of acid, WI, is then added to the separator sample, and the sample is enclosed in a 0.003" polyethylene film or bag, heat sealing the edges together so the sample has no excess air and little void space. Any polyethylene film or bag can be used although it is preferable to utilize a film or bag having WO 00/11746 PCT/US99/18499 10 sufficiently low water vapor transmission properties so that the loss in electrolyte over two months' testing is no more than 0.20% of the total electrolyte weight. To test the sample prepared, the sample is centered onto an aluminum base plate of a fixture. The top to the fixture (a 6"x6"x0.5" polycarbonate slab) is 5 placed on the test sample. Then, the load cell (LCAA-100 with an LBC-014 button, Omega) is put on top of the top plate so that the button is balanced on the top plate. The base of the load cell is then tightened down until the top plate of the fixture touches a spacer machined for the sample at 80% of the sample thickness and having 0.004" for the thickness of the polyethylene. 10 The data collected is a continuous plot of the pressure sensed by the load cell button over time. While a test time of two months is the time of choice for sample evaluation, lesser times can be used, if considered appropriate. Pursuant to another aspect of the present invention, the properties of the separators utilized should be preselected to have Saturation and Stratification 15 characteristics preselected to have a variance of no more than 10% for at least one, and preferably both of these characteristics, preferably a differential of no more than about 5%, and, even more preferably, no more than about 2%. Maintaining these characteristics is necessary for any VRLA cell or battery designed to function in service in an upright condition and is particularly important for taller 20 cells, i.e., cells requiring separators having a height in excess of 10 inches or so. To test the Saturation and Stratification characteristics, a 3"xl 1" rectangle is cut. The thickness of the chosen sample piece and the weight of the sample to the nearest milligram are determined, as was the case regarding the load cell test. Similarly, as was likewise done in the load cell test, the grams of 1.31 sulfuric acid 25 necessary to be added to achieve 95% saturation was determined utilizing the protocol as described for the load test. Then, after the addition of the requisite amount of electrolyte to the sample, the sample was again enclosed in a 0.003" polyethylene film with the edges heat sealed as previously described in the load cell test. The thus-enclosed sample is 30 then sandwiched into the Saturation and Stratification test fixture. This test fixture comprises front and back 6"x24"x0.5" polycarbonate slabs. One of the slabs has WO 00/11746 PCT/US99/18499 11 3.5" sliver/grooves laser-cut therein every inch. The slabs have equidistant 3/8" holes for the hold down mechanism, viz., nuts and bolts. The separator is aligned to the bottom slit of the slitted slab. Spacers are added to the bolts, machined specifically for the thickness of the sample as previously described. The other slab 5 is then brought on top of the separator sample using the bolts for guides. The top of the fixture is then tightened to the spacers using the nuts. The set up is allowed to stand for two months. After the stand time, the separator sample is cut into 1"x3" pieces using the slits of the fixture as a guide. Each piece, on the plastic, is sequentially numbered. Each piece is then weighed 10 with the numbered plastic sheath intact. The specific gravity is then checked for each sample using a Reichert-Jung Refractometer, and the samples are placed in their corresponding numbered beakers (viz., 250 ml beakers). Based upon this data, the acid gravity versus the height of the cell can be determined, and this determines the stratification. 15 To determine the respective Saturation, each plastic piece is dry wiped and then weighed. The wet weight of the sample, W,, can be determined by subtracting the weight of the plastic from the weight of the 1"x3" separator/ sheath pieces. The sample pieces are then rinsed with deionized water using a simple soaking method. The pH of the baths are then monitored and changed 20 approximately 4-5 times until the pH stays above 4.5 (or whatever is the pH of the deionized water used). After soaking, the samples are placed into an oven, removed from the beaker and placed in a marked aluminum weighing dish. The oven is maintained at a 120 0 C for 24 hours with no convection. The samples are then removed from the oven and placed into a desiccator until cool. The samples 25 are then weighed to the nearest milligram, providing the dry sample weight, Wd. If any flaking of the sample occurs, this should be added to the dry weight. The weight of the acid, Wa, in each sample is determined by subtracting the dry weight of the separator sample, Wd, from the wet weight of the separator sample, Ww. 30 Then, the amount of acid, A, in grams is calculated based upon how much the separator sample would hold if 100% saturated. This calculation utilizes K, WO 00/11746 PCT/US99/18499 12 determined as previously described, and then multiplied by the dry weight of the separator sample, Wd: A=KxWd. This calculation is carried out for each separator piece. 5 The saturation level of each piece of separator is then calculated by dividing the actual acid held in the sample, Wa, by the 100% saturated weight, A, and multiplying by 100: Percent Saturation = Wa/A x 100 From this data, the percent saturation versus the height of the cell can be 10 determined. While the separator samples described in the test have a sample height of 11 inches, it should be appreciated that the height of the sample can be increased to the height required for a particular VRLA cell or battery. In that event, the Saturation and Stratification characteristics still should, preferably, be within the ranges previously described. 15 Generally, there is not a difference in specific gravity from top to bottom of material in this testing procedure as described. This same procedure is used on separator material from tested cells, for verification, and changes in saturation as well as specific gravity are observed. One preselected set of separator characteristics is the pore size 20 characteristics. More particularly, in accordance with the present invention, the mean pore size varies from about 1.0 to 2.2 microns. Further, the pore size at the maximum distribution preferably varies from about 1.0 to 2.5 microns. The mean pore size and the pore size at maximum distribution are determined using an Automated Capillary Flow Porometer (Model CFP 1100 AEX from Porous 25 Materials Inc.). Suitable parameters are as follows: Pulse width = 1 sec., max. pressure = 10 PSI and max flow = 20000 cc/m. Pore Size Distribution is calculated from the pressure - pore size relationship by commercially available PMI software which uses the relationship of gas flow rates, both wet and dry. Further, it has been found preferable to utilize separators wherein the maximum 30 pore size is below 15 microns, more preferably less than about 14 microns, and even more preferably, less than about 10 microns or so WO 00/11746 PCT/US99/18499 13 Pore size distribution can be used to predict the ability of the RBSM to resist the development of saturation differentials. Mean pore sizes greater than 2.8 microns have a propensity to cause saturation differentials to form, even in cells where the separators are only 5 or 6 inches in height. It has been thus been found 5 that separators having such preferable mean and maximum pore size distribution characteristics provide enhanced performance during service. Such enhanced performances are particularly evident with cells having a relatively high aspect and intended to be used in service in an upright position. However, such separators also may be used when the cells, in service, are intended to be used in a 10 pancake fashion, viz., positioned in service on their side. Thus, it is also believed that these characteristics assist in providing a satisfactory oxygen recombination cycle. For most applications, it will be preferred that the normalized basis weight (where the RBSM contains polymer fibers, as will be discussed hereinafter) should 15 be at least preferably above 150, between about 170 to about 200 grams/meters 2 /millimeter. Such basis weights can be determined by the BCI technique previously referenced. As may be appreciated, the higher the normalized basis weight, the relatively more material that will be required. Therefore, the expense of the 20 separator is increased. However, from the standpoint of achieving the more preferred load cell characteristics, it has been found that the use of higher basis weight materials tend to enhance the load cell characteristics in comparison to lower normalized basis weight materials. As to the porosity of the separator according to the present invention, it is 25 necessary that the porosity be adequate to provide the appropriate size of the electrolyte reservoir. Thus, the porosity becomes a design issue; and it is preferred, accordingly, to utilize separators having a porosity of at least about 90 or 91%, up to about 93% or so. From the standpoint of providing separators having the desired 30 combination of preselected properties as discussed herein, it has been found desirable to provide a glass fiber mat which includes, based upon the overall WO 00/11746 PCT/US99/18499 14 weight of the separator, from about 5% up to perhaps 30% or so of plastic fibers, preferably no more than about 12%. Such separators can be made by known and conventional processes, simply incorporating into the slurry from which the separator is made, the relative amount and type of the selected fibers. Indeed, it is 5 believed that incorporating the selected level of plastic fibers tends to provide separators having more consistent and reproducible properties in comparison to fibers containing only glass. The plastic fibers utilized in the separators of the present invention are preferably polyester fibers, more preferably, polyethylene terephthalate fibers. 10 Suitable amounts of such polyester fibers range from about 5 to about 12%, based upon the total weight of the separator. As alternatives to the preferred polyester fibers, anyother acid-stable polymers could be utilized. Suitable fibers include polyolefins such as polyethylene and polypropylene. The addition of such polymeric fibers increases the resilience of the RBSM as determined by improved 15 load properties. However, if the polymer fiber content becomes too high (e.g., 50% or more), the load properties diminish. RBSM materials fabricated with a glass/polymer matrix have superior load/springy properties. These materials, however, are more hydrophobic; and the hydrophobicity directly correlates with the percent polymer content. Because of 20 their hydrophobic nature, these materials experience difficulty in maintaining satisfactory equal saturation levels from the top of the cell to the bottom. Controlling the mean pore size to below 1.5 microns can help, but does not solve the saturation problems. Such polymer fibers contained within this RBSM matrix can be treated by any known means, if desired, to convert the surface of the 25 polyester fibers to polar aprotic-type groups, such as SO 2 , SiO 2 and other known polar aprotic groups. Such known surface treatments involve transesterification, chemical vapor deposition, grafting and/or chemical laser treatments. Any means capable of modifying the fiber surface to enhance the wettability to the sulfuric acid electrolyte can be utilized. Increasing the level of fine fibers in the 30 glass/polymer material also aids in resistance of the material to saturation and stratification problems.

WO 00/11746 PCT/US99/18499 15 As may be appreciated, whether such surface treatment is necessary, depends upon the service life requirements of the cell and the total plastic fiber content. Thus, a certain degree of hydrophobic properties is desirable for the separator to aid in the initial recombination cycle when the cell is saturated to the 5 level in excess of about 90% or so, as would occur after assembly and during formation. This is particularly important in float applications. Separators utilizing only about 5% polyester or so may require no treatment at all. Pancake cell designs should not require the plastic fibers to be surface treated at all, regardless of the plastic fiber content. 10 As is known, many of the currently used RBSMs comprise a combination of coarse and fine glass fiber blends to provide a surface area for such RBSMs of about 1.2 m 2 /gm or so. It may be suitable for some applications to increase the proportion of the fine fibers up to 100%. Such increased fine fiber levels will assist in achieving the necessary preselected pore size requirements to maintain 15 even saturation from cell top to cell bottom and retard stratification. This is important for electric vehicle (EV) or cycling applications. The mean pore size for 100% fine fiber RBSM is typically less than 1.5 microns . These materials generally have acceptable springback properties; however, the cost of these materials can be as much as 3-4 times that of standard grade RBSM. 20 In accordance with another aspect of the present invention, it has been found that the performance of the separators of the present invention can be enhanced by adding appropriate levels of silica. Such additions can be made either during or after formation of the separator. Preferably, the level of SiO2 added ranges from about 5 to about 30, based upon weight of RBSM. It is 25 believed that the addition of the SiO 2 may provide a "cross-linking" matrix or "glueing" action between the glass and plastic fibers. The SiO 2 ions may also provide SiO 2 to preferentially dissolve to the glass fibers over time and under various temperature conditions, which preserves the strength and integrity of the glass fibers. However, whatever the mechanism, it has been found that the 30 inclusion of such silica not only enhances acid wettability, but also serves to maintain the springiness of the separator in service. Even further, it has been WO 00/11746 PCT/US99/18499 16 found that the inclusion of the silica enhances the ability of such separators to maintain saturation and prevent stratification of the electrolyte from the top of the cell to the cell bottom. Fillers other than silica can be used, if desired, particularly if such fillers enhance wettability and/or maintaining saturation and/or preventing 5 stratification and/or springiness of the resulting separator. Any such fillers should not, of course, be unduly deleterious to the desired performance of the VRLA cells. Further, to improve the integrity and strength of the separators, coupling agents can be utilized to bond the fibers in the separators into a cohesive matrix. 10 Suitable coupling agents should be stable in the cell environment and not be detrimental to the cell performance. Still further, and particularly for longer service life applications, the stability and integrity of the RBSMs in the cell environment becomes an important issue so as to avoid, or at least minimize, long performance declines due to 15 separator deterioration. In such cell environments over an extended service life, glass fibers are superior to polymer fibers as far as stability and integrity are concerned. Accordingly, pursuant to one aspect of the present invention, a multiple layer, dual functional, separator is used. To this end, a glass separator layer is 20 positioned adjacent the-positive electrodes and a composite glass/polymer separator layer having the desired preselected properties is positioned adjacent the glass fiber layer to achieve the desired porosity, springiness and saturation and stratification characteristics, and protect the polymer/additives from oxidative degradation. 25 The respective thicknesses of each layer can be varied as desired to satisfy the necessary characteristics considered appropriate. In this regard, by way of example, it will generally be desirable for the composite separator layer to comprise the majority of the thickness, up to perhaps 70% or so of the total thickness. 30 Indeed, while such dual functional separators can provide desirable benefits, the separator used in the VRLA cells can comprise multiple layers of the WO 00/11746 PCT/US99/18499 17 glass/polymer separators having the preselected properties, if desired. Also, the spatial configurations other than that previously described can be utilized all through the interface adjacent the separator and the positive electrode should be the region where separator stability will be most severely tested. 5 Thus it will be seen that a novel and improved separator has been provided which attains the aforementioned objects. Indeed, the present invention provides criteria and test regimes that allow, it is believe for the first time, predictable separator performance. Various additional modifications of the embodiments specifically illustrated and described herein will be apparent to those skilled in the 10 art, particularly in light of the teachings of this invention. The invention should not be construed as limited to the specific form shown and described, but instead is set forth in the following claims.

Claims

1. A valve-regulated lead-acid cell or battery having at least one cell having positive and negative plates and an absorbent separator between said positive and negative plates, said separator being preselected with a load cell pressure at two months of at least about 6.0.

2. The cell or battery of claim 1 wherein the load cell pressure is at least about 6.5.

3. The cell or battery of claim 2 wherein the load cell pressure is at least about 7.0.

4. The cell or battery of claim 3 wherein the load cell pressure is at least about 8.0.

5. A valve-regulated lead-acid cell or battery having at least one cell having positive and negative plates and an absorbent separator between said positive and negative plates, said separator being preselected to have saturation and stratification characteristics having a variance of no more than about 10% for at least one of the saturation and stratification characteristics.

6. The cell or battery of claim 5 wherein both of the saturation and stratification characteristics have a variance of no more than about 10%.

7. The cell or battery of claim 5 wherein the variance is no more than about 5%.

8. The cell or battery of claim 7 wherein both of the saturation and stratification characteristics have a variance of no more than about 5%. WO 00/11746 PCT/US99/18499 19

9. The cell or battery of claim 5 wherein the variance is no more than about 2%.

10. The cell or battery of claim 9 wherein both of the saturation and stratification characteristics have a variance of no more than about 2%.

11. A valve-regulated lead-acid cell or battery having at least one cell having positive and negative plates and an absorbent separator between said positive and negative plates, said separator being preselected to have a maximum pore size below about 15 microns, a mean pore size less than about 2.8 and a porosity of at least about 90%.

12. The cell or battery of claim 11 wherein the maximum pore size is less than about 14 microns.

13. The cell or battery of claim 12 wherein the maximum pore size is less than about 10 microns.

14. A method for predetermining the suitability of an absorbent separator material for a valve-regulated lead-acid cell or battery which comprises selecting a proposed absorbent material, testing said proposed absorbent material by at least one of the following test protocols: (1) load cell pressure, (2) saturation and stratification characteristics, and (3) maximum and minimum pore sizes to determine whether said proposed absorbent material meets certain preselected criteria, and then not using said proposed absorbent material in the cell or battery which does not meet said preselected criteria.

15. The method of claim 14 wherein the test protocol used is the load cell test and the preselected criteria is a load cell pressure at 2 months of at least about 6.0. WO 00/11746 PCT/US99/18499 20

16. The method of claim 14 wherein the test protocol used are the saturation and stratification characteristics and the preselected criteria is that at least one of these characteristics have a variance of no more than about 10%.

17. The method of claim 14 wherein the test protocol used are the maximum and mean pore sizes and the preselected criteria are a maximum pore size below about 15 microns and a mean pore size less than about 2.8.