WO2013181121A1

WO2013181121A1 - Equalization of string battery configuration

Info

Publication number: WO2013181121A1
Application number: PCT/US2013/042803
Authority: WO
Inventors: Enders Dickinson; Jack Shindle; Philippe Westreich; Vladimir Vichnyakov
Original assignee: Axion Power International, Inc.
Priority date: 2012-05-30
Filing date: 2013-05-28
Publication date: 2013-12-05

Abstract

A method of equalization of a string of batteries includes charging at least one string of battery modules until a first single module reaches a preset maximum voltage; reducing or pulsing a charge current to the at least one string; and repeating the charging step. The at least one string of battery modules includes at least two series connected modules that display concave down, increasing-shaped charge or voltage profiles.

Description

EQUALIZATION OF STRING BATTERY CONFIGURATION

This PCT international application claims priority of U.S. Serial No. 61/653,073 filed in the U.S. Patent and Trademark Office on 30 May 2012.

I. TECHNICAL FIELD

The present invention is directed to methods for equalizing at least one string

configuration of battery modules and to battery systems so equalized.

II. BACKGROUND OF INVENTION

Batteries that are required to deliver high energy and power output are typically provided by connecting several multi-cell or mono-block modules in a series string configuration, in which individual batteries, or modules (a term used when discussing strings of batteries), are connected end-to-end (e.g., positive of module 1 connected to negative of module 2 connected to positive of module 3, etc.). This results in a large "battery" with a voltage equivalent to the sum of the module voltages. Likewise, each individual module is composed of a series connection of identical cells composed of alternating positive and negative plates (also with separator and an electrolyte) required for the respective battery charge/discharge reactions. A schematic of this battery arrangement is illustrated in FIG. 1 .

FIG. 1 shows a battery comprising a series connection of 3 multi-cell modules. The nominal cell voltage (fundamental unit) is 2V; the nominal voltage of each multi-cell module (6-cells) is 12V; and the overall nominal voltage of the whole battery (entire series string) is 36V. These numbers are representative of common battery types (lead- carbon and lead-acid); while other battery chemistries may have different fundamental cell, multi-cell, and overall string voltages. While series string configurations are the most common method of providing high energy and power output, it is challenging to optimally manage both module and individual cell variation in order to preserve consistent battery discharge and charge performance and life. Typically, the only known voltage is that of the battery as a whole, a measurement that simply provides the sum of all module or individual cell voltages.

When operating the battery using string metrics, the true state of each module may be receiving too much or too little charge/discharge. This may be the situation with a string comprising modules with a high performance spread due to manufacturing variation, or can be a situation that develops over time as the modules (and/or cells within the modules) show a gradual spread in performance due to various levels of at least one of durability, wear, aging, or the like (despite low manufacturing variation).

Because such performance spread limits the charge/discharge output and life of a battery, there exist a variety of methods to minimize and/or correct for this module-to- module variation (in essence cell-to-cell variation - although individual cell metrics are not always accessible as the positive and negative contacts of each cell are often enclosed in a single multi-cell product). The goal of these methods, often referred to as "string equalization" methods, is to minimize the extent and/or impact of module variation in batteries composed of series strings, thus extending their useful charge/discharge output and life. Some of these methods and their limitations are briefly described below. In a Dissipative Equalization method, proper electronic hardware can be added to each module that allows optimized distribution of energy throughout the modules/cells of a series string. This method then singles out the highest charged module/cell (i.e., the "strongest") indicated by higher voltage, and removes excess energy through a bypass resistor until the voltage or charge matches the voltage or charge of "weaker"

modules/cells. Another variation of the Dissipative Equalization method halts charging when the strongest module/cell is fully charged, and then discharges the stronger modules/cells into a load until they reach the same charge level as the weaker modules/cells. Finally, a third variation maintains charging until ALL modules/cells are fully charged, but with a voltage limit applied to individual modules/cells to bypass charging when the limit is reached.

This Dissipative Equalization method levels DOWNWARDS (i.e., highest module/cell discharged to eventually matched lower module/cells) and requires low magnitude bypass currents that significantly extend the overall equalization charge time. More importantly, the overall battery performance is determined by the weakest cell and is inefficient due to wasted energy involved in the bypass actions.

A Shunting Equalization method levels UPWARDS by charging all modules/cells until a single module/cell has reached its rated voltage, at which point the full charging current is bypassed until all weaker modules/cells reach rated voltage. While this method is rapid and maximizes battery charge/discharge output, it requires expensive high current switches, high power dissipating resistors, and complex control logic, which is typically beyond the scope of the application.

A Charge Limiting Equalization is a basic way to lessen the impact of module/cell variation by simply removing the charge current when a first cell reaches the fully charged state during charge and the fully discharge state during discharge.

Unfortunately, this approach terminates the operations prior to all modules/cells being adequately charged/discharged, thereby significantly underutilizing the energy of the battery and reducing overall cycle life due to undercharging.

The methods described above are considered "monitored management" methods and require the accurate determination of state-of-charge (SOC) of the modules/cells in a battery string. While a straightforward voltage measurement is often used as a metric of SOC, this approach can be prone to error and lead to an increase in module/cell variation. While this approach may cut-off charge to high voltage modules/cells, these cells may display higher voltages not because of a higher SOC, but instead due to higher internal impedances (often the result of manufacturing variations). This is an additional unknown that can lead to significant errors in equalization efforts and will place additional stresses on these modules/cells with repeated cycling, thereby shortening life. As a result, more complex methods related to precise voltage

measurements and coulomb counting with consideration of temperature, historical usage, and age are required. Often, these complex methods are incorrectly

implemented and therefore do not provide sufficient improvements in actual

applications.

An alternative to these "monitored management" methods is the addition of chemical additives that "absorb" excess charging energy when the module/cell is above a voltage set point. The addition of such a "redox shuttle" is common in Li-ion batteries and minimizes the need for electronic module/cell balancing. Furthermore, lead-acid batteries charged above about 2.3 volts per cell undergo gassing (electrolysis of water), which can eventually balance the modules/cells. However, both of these methods require long charge times to implement the effect across all the modules/cells of a string (thereby reducing application use of the battery) and reduce the cycle life due to aggressive water loss and resulting dry out.

All of the methods described above provide suboptimal results including, but not limited to, insufficient variation reduction, limited battery utilization, reduced life, excessive offline maintenance time, or any combination of these.

The methods developed by Applicant are simplified "monitored management" methods that rely on the specific shape of the module/cell charge or voltage profile. The feature of this equalization method is referred to as a "concave down, increasing" (CDI) shaped charge or voltage profile and the equalization methods are termed CDI Equalization.

III. SUMMARY OF INVENTION According to a method of the present invention, equalization of a string of batteries is characterized by charging at least one string of battery modules until a first single module reaches a preset maximum voltage; reducing or pulsing a charge current to the at least one string; and repeating the charging step. The at least one string of battery modules comprises at least two series connected modules that display concave down, increasing-shaped charge or voltage profiles.

According to another method of the present invention, equalization of a string of batteries is characterized by charging at least one string of battery modules until a first single module reaches a preset maximum voltage; reducing a charge current to the at least one string; and repeating the charging step. The at least one string of battery modules comprises at least two series connected modules that display concave down, increasing-shaped charge or voltage profiles.

According to yet another method of the present invention, equalization of a string of batteries is characterized by charging at least one string of battery modules until a first single module reaches a preset maximum voltage; removing a charge current to the at least one string; and repeating the charging step. The at least one string of battery modules comprises at least two series connected modules that display concave down, increasing-shaped charge or voltage profiles.

According to still another method of the present invention, equalization of a string of batteries is characterized by charging at least one string of battery modules until the at least one string reaches a preset maximum voltage; reducing or pulsing a charge current to the at least one string; and repeating the charging step. The at least one string of battery modules comprises at least two series connected modules that display concave down, increasing-shaped charge or voltage profiles. An advantage of the present invention is that the methods provide an efficient, rapid, low cost, and effective means of battery equalization. Another advantage of the present invention is that the methods maintain low module-to- module (and cell-to-cell) variations of voltage and SOC, thereby resulting in improved battery charge/discharge performance and life.

Still another advantage of the present invention is that the methods may result in significantly less gassing during repetitive cycling, thereby extending life and minimizing the module-to-module voltage/SOC variation. In the following description, reference is made to the accompanying drawings, which are shown by way of illustration to specific embodiments in which the invention may be practiced. The following illustrated embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It is to be understood that other embodiments may be utilized and that structural changes based on presently known structural and/or functional equivalents may be made without departing from the scope of the invention.

As used herein "substantially", "relatively", "generally", "about", and "approximately" are relative modifiers intended to indicate permissible variation from the characteristic so modified. They are not intended to be limited to the absolute value or characteristic which it modifies but rather approaching or approximating such a physical or functional characteristic.

In the detailed description, references to "one embodiment", "an embodiment", or "in embodiments" mean that the feature being referred to is included in at least one embodiment of the invention. Moreover, separate references to "one embodiment", "an embodiment", or "in embodiments" do not necessarily refer to the same embodiment; however, neither are such embodiments mutually exclusive, unless so stated. Thus, the invention can include any variety of combinations and/or integrations of the

embodiments described herein. IV. BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows a battery comprising a series connection of 3 multi-cell modules. FIG. 2 shows charge or voltage profile curves for lead-carbon batteries and lead-acid batteries.

FIG. 3 shows an example of CDI Equalization according to the present invention for three series connected lead-carbon modules starting at 3 different voltages/SOCs.

FIG. 4 shows an example of CDI Equalization according to the present invention for three series connected lead-acid modules set at 3 different voltages/SOCs.

FIG. 5 shows sample charge or voltage profiles for Li-ion batteries (top) and Nickel- metal Hydride batteries (bottom), indicating a degree of CDI-shape.

FIG. 6 shows repetitive cycling with CDI Equalization implemented of lead-acid (left) and lead-carbon (PbC^® right) 3 module strings with modules preset to a range of voltages/SOCs.

FIG. 7 shows a comparison of 120 day cycling (industrial duty cycle) of 24-module lead- acid and lead carbon (PbC^®) strings (left and right, respectively).

V. DETAILED DESCRIPTION OF INVENTION

The equalization methods of the present invention are directed to batteries that display a charge profile or voltage profile with a shape described by a concave down, increasing (CDI) function of voltage versus charge time. For any battery technology that displays the CDI voltage profile, the methods according to the present invention provide an efficient, rapid, low cost, and effective means of battery equalization. Batteries displaying some degree of this characteristic charge or voltage profile, from high to almost none, include, but are not limited to, asymmetric lead-carbon batteries (e.g., batteries with a positive electrode comprising lead and/or lead dioxide and a negative electrode comprising activated carbon, also known commercially as PbC^® batteries available from Applicant), Ni-Metal-Hydride (NiMH) batteries, Li-ion batteries, and lead-acid batteries.

FIG. 2 shows voltage profiles for equivalent lead-carbon (PbC^®) and lead-acid batteries. The lead-carbon battery voltage profile is clearly concave down, increasing (CDI). The voltage profile for the lead-acid battery is nominally linear with a slight degree of concave up, increasing shape at the end of the charge. The dotted lines included as visual guide to overall general shape of the voltage profile curves.

These two batteries represent differences in applying the CDI Equalization methods according to the present invention, with lead-carbon batteries showing significant string performance improvement with CDI Equalization, compared to lead-acid batteries showing no performance improvement under the same conditions.

According to a first embodiment of an CDI equalization method of the present invention, at least one string configuration of battery modules is charged until a first single module reaches a preset or predetermined maximum voltage, for example about 13.2V to 13.8V (e.g., 13.5V) for a 12V battery. Then, the charge current is reduced, for example by half, and the charge step is repeated. This method maintains low module-to-module (and cell-to-cell) variations of voltage and SOC, resulting in significantly improved battery charge/discharge performance and life.

According to a second embodiment of the present invention, a CDI equalization method may comprise charging at least one string until a preset or predetermined maximum string voltage is reached. The charge current is then reduced and the charge step is repeated. This embodiment avoids the need for individual voltage sensing of each module and still provides the conditions necessary for module/cell equalization (variation reduction), but may not be as efficient as using a single module/cell voltage maximum.

In a third embodiment of the present invention, a CDI equalization method comprises removing the charge current for a period of time (i.e., pulsing) instead of reducing the current before the charge step is repeated.

According to the CDI Equalization methods of the present invention, the "reduced- charge" or "off-charge" periods provide time for the module/cell voltages to decrease and thus for "weaker" (lower voltage/SOC) modules/cells to "catch-up" to the "stronger" (higher voltage/SOC) ones. Thus, these CDI equalization methods maximize

convergence of all modules/cells around the preset or chosen charge voltage set point. The CDI Equalization methods significantly reduce module-to-module voltage across a string and maximize the string's performance and life.

The more pronounced a CDI-shape of the voltage profile for a particular battery architecture or technology, the more effective the "catch-up" periods may be. The magnitude of current reduction and duration of reduction steps may be optimized to a particular battery technology, design, configuration, application, or the like.

According to fourth embodiment of the present invention, a CDI equalization method comprises not repeating the charge step with a lower current, or successive pauses, but simply charging to a module/string maximum during cycling. Due to the CDI shape of the voltage profile, any charging will, to some degree, serve to align module

voltages/SOCs. However, current reductions and/or pauses (pulsing) significantly improve the process efficiency (increase) and duration (decrease).

FIG. 3 shows an example of CDI Equalization for three series connected lead-carbon modules having at three different initial voltages/SOCs (X1 , X2, and X3). The string is charged until a first module (X3) reaches a preset maximum voltage. The charge current is reduced for a time, and then the charging step is repeated. The CDI Equalization allows lower voltage/SOC modules (X1 , X2) to "catch-up" to those with higher voltages/SOCs (X3)

In contrast, without a CDI-shaped voltage profile, the CDI equalization method of the present invention has little measureable effect, as shown for lead-acid batteries in FIG. 4. FIG. 4 shows an example of CDI Equalization for three series connected lead-acid modules set at three different voltages/SOCs (X1 , X2, and X3). The string is charged until a first module (X3) reaches a preset maximum voltage. The charge current is reduced for a time, and then the charging step is repeated. The concave up, increasing shape of the lead-acid voltage profile negates the leveling effect of repeated charge times with lower currents, that is, low voltage/SOC modules (X1 , X2) cannot "catch-up".

A battery lacking a CDI-shaped voltage profile not only means CDI Equalization is substantially ineffective, but also forces the use of one or more suboptimal equalization techniques discussed above in the Background (e.g., gassing voltages must be entered to properly align modules/cells). In comparison, batteries with voltage profiles displaying the CDI-shaped voltage profile do not require aggressive equalization techniques and can be charged without necessarily bringing the entire battery (module string) to high voltages (e.g., even partial charging forces modules closer in

voltage/SOC - as evident by examining the outcome of using just half of the T-i charge period in FIG. 3). In addition, the CDI equalization methods of the present invention may result in significantly less gassing during repetitive cycling, thereby extending life and minimizing the module-to-module voltage/SOC variation. FIG. 5 shows sample voltage profiles for Li-ion batteries (top) and NiMH batteries (bottom) indicating some degree of a CDI-shape. Given the less pronounced CDI features of these voltage profiles compared to that of lead-carbon batteries, it is expected that while CDI Equalization may have some effect on Li-ion and NiMH strings, it may not be as substantial compared to lead-carbon battery strings. The reference for the voltage profiles are from T. Cleveland et al., "Developing Affordable Mixed-Signal Power Systems for Battery Charger Applications", Microchip Technology, Inc. (URL: http://www.microchip.com/stellent/groups/designcenter_sg/documents/rn

cation /en027883.pdf).

Applicant has demonstrated the effectiveness of CDI Equalization by comparing both small (3-module) strings of lead-carbon (PbC^®) modules versus lead-acid modules, as well as larger (24-module) strings of each.

In a smaller string (3 modules as shown in FIG. 1 ), CDI equalization clearly moved all lead-carbon modules closer to a chosen charge voltage target, with the weaker lead- carbon modules "catching-up" to the stronger lead-carbon modules. However, the same conditions imposed on a 3-module lead-acid string showed no improvement in voltage spread across the modules, as shown in FIG. 6.

In particular, FIG. 6 shows CDI Equalization according to the first embodiment of a valve-regulated lead-acid 3-module string (left) and of a 3-module lead-carbon (PbC^® right) string. The modules were preset to a range of voltages/SOCs. As shown, the lead-acid modules remain separated in voltage/SOC, while the lead-carbon modules rapidly converge to the voltage/SOC maximum. This variation reduction is a result of the CDI-shape reacting as illustrated in FIG. 3 to the CDI Equalization charge currents.

A larger string configuration is illustrated in FIG. 7, which shows a comparison of 120 day cycling (industrial duty cycle) of 24-module lead-acid and lead-carbon strings (left and right, respectively). As shown, the lead-acid string required two equalizations EQ (long time, low current charges) and 9 total battery replacements BR (greater than 1 /3 of the total modules in the string). The lead-carbon string required no equalizations or battery replacements and maintained a module-to-module variation up to 10x lower than that of the lead-acid string.

Furthermore, due to the normalization of voltage/SOC difference with CDI Equalization on lead-carbon strings, if replacement batteries are at any time required, the initial voltage/SOC of the battery does not necessarily require exact preconditioning, and/or full string equalization following the replacement, as is the case with lead-acid modules used for string replacements. The resulting extremely low module-to-module variation (M2M Variation) across the lead-carbon string results in higher string utilization, significantly less application down time (less frequent replacements/ no equalizations), and longer life.

VI. INDUSTRIAL APPLICABILITY

The present invention is directed to methods for equalizing at least one string

configuration of battery modules, thereby maintaining low module-to-module (and cell- to-cell) variations of voltage and state-of-charge and improving battery charge/discharge performance and life.

Although specific embodiments of the invention have been described herein, it is understood by those skilled in the art that many other modifications and embodiments of the invention will come to mind to which the invention pertains, having benefit of the teaching presented in the foregoing description and associated drawings.

It is therefore understood that the invention is not limited to the specific embodiments disclosed herein, and that many modifications and other embodiments of the invention are intended to be included within the scope of the invention. Moreover, although specific terms are employed herein, they are used only in generic and descriptive sense, and not for the purposes of limiting the description invention.

Claims

WHAT IS CLAIMED IS:

1 . A method of equalization of a string of batteries, characterized by:

charging at least one string of battery modules until a first single module reaches a preset maximum voltage;

reducing or pulsing a charge current to the at least one string; and

repeating the charging step,

wherein the at least one string of battery modules comprises at least two series connected modules that display concave down, increasing-shaped charge or voltage profiles.

2. A method according to Claim 1 , characterized in that the battery modules comprise at least one of lead-carbon batteries, NiMH batteries, or Li-ion batteries.

3. A method according to Claim 1 , characterized in that the battery modules comprise lead-carbon batteries.

4. A method according to Claim 1 , characterized in that the battery modules comprise NiMH batteries.

5. A method according to Claim 1 , characterized in that the battery modules comprise Li-ion batteries.

6. A method according to any one of Claims 1 -5, characterized by reducing a charge current to the at least one string prior to repeating the charging step.

7. A method according to any one of Claims 1 -5, characterized by removing a charge current to the at least one string prior to repeating the charging step.

8. A method according to any one of Claims 1 -5, characterized by charging at least one string of battery modules until the at least one string reaches a preset maximum string voltage.

9. A method according to any one of Claims 1 -5, characterized by the repeat charging step being at a lower current than the initial charging step.

10. A method according to any one of Claims 1 -5, characterized by:

charging the at least one string until a single module or battery cell reaches an upper voltage limit; and

reducing the current, thereby allowing charging to continue and providing a weaker module or battery cell an opportunity to reach a higher voltage/SOC.

1 1 . A method according to any one of Claims 1 -5, further characterized by repeatedly cycling of the at least one string to maximize convergence of all modules or battery cells around the present maximum voltage.

12. A method according to any one of Claims 1 -5 characterized by not repeating the charge step.

13. A battery treated by the method any one of Claims 1 -5.

14. An energy device treated by the method of any one of Claims 1 -5.