It •:
EPA-460/3-74-025
DECEMBER 1972
NICKEL-ZINC BATTERIES
FOR HYBRID
VEHICLE OPERATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Alternative Automotive Power Systems Division
Ann Arbor, Michigan 48105
-------
EP A-460 /3-7 4-025
NICKEL-ZINC BATTERIES
FOR HYBRID
VEHICLE OPERATION
by
Martin J. Sulkes
u. S. Electronics Technology and Devices Laboratory
u. S. Army Electronics Command
Fort Monmouth, New Jersey
Interagency Agreement
EPA Project Officer: Dr. Jalal T. Salihi
Prepared for
u . S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
. Office of Mobile Source Air Pollution Control
Alternative Automotive Power Systems Development Division
Ann Arbor, Michigan 48105
December 1972
-------
This report is issued by the Environmental Protection Agency to
report technical data of interest to a limited number of readers.
Copies are available free of charge to Federal employees, current
contractors and grantees, and nonprofit organizations - as
supplies permit - from the Air Pollution Technical Information
Center, Environmental Protection Agency, Research Triangle Park,
North Carolina 27711; or, for a fee, from the National Technical
Information Service 5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the U. S. Environmental Protection
Agency by the U. S. Army Electronics Command in fulfillment of
an interagency agreement and has been reviewed and approved for
publication by the Environmental Protection Agency. Approval
does not signify that the contents necessarily reflect the views
and policies of the agency. The material presented in this report
may be based on an extrapolation of the "State-of-the-art." Each
assumption must be carefully analyzed by the reader to assure that
it is acceptable for his purpose. Results and conclusions should
be viewed correspondingly. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
Publication No. EPA-460/3-74-025
ii
......
-------
Abstract
Nickel-zinc batteries have been evaluated against the requirements of
a hybrid electric vehicle regime. This regime consists of many short dura-
tion pulses at high rates of discharge with rapid recharge. High rate de-
signed conventional Ni-Zn batteries of a nominal 24 Ah capacity yielded up
to 26,000 Ah total output (equivalent to 150,000 cycles) before failureS
this compares to a desimioutput of 13,000 Ah or 400,000 cycles. The
principal failure mechanisms vere shorting and loss of voltage due to high .
carbonate concentrations caused by degradation of the cellulosic separator.
Zinc shape change (10-20j) was surprisingly low in view of the large number
of cycles. The nickel electrodes were essentially unaffected by the cycling.
Rapid and complete recharge was accomplished with a modified constant
potential of 1.83V/cell. The use of this voltage resulted in low heat
generation, approximately 7-8 BtU/CeU/h and a low percentage of overcharge.
Overall energy efficiency on the hybrid vehicle test regime is as high as
8~. Two types of inorganic separator material were evaluated, and were
found to be unsuitable. A full size battery, to power a hybrid electric
vehicle, would consist of 120 cells. '!he battery would weigh 350-400 lbsj
the charging voltage would be 220V. The average discharge voltage would
be 201V at a 10kW load and 165V at a 55kW load. .
III
~~
-------
Foreword
This report presents the work performed by the US Army Electronics
Technology and Devices Laboratory (ECOM) in evaluating advanced nickel-
zinc battery concepts for a heat engine/battery hybrid for family car
propu1.sion. In this application the battery stxnes part of the energy
provided by the heat engine and supplies it when acceleration power is
needed. The engine thus operates essentially at steady state, and
thereby reduces exhaust emissions. This work, conducted for the
Enviromnental Protection Agency under an interagency agreement covered
by EPA Obligation Code 690403, extended from May 1971 to November 1972.
The assistance of Dr. J. Salihi of the Advanced Automotive Power
Systems Development Division, Office of Air and Water Programs,
Enviromnental Protection Agency, is gratefully acknowledged.
iv
..
-------
TABLE OF CONTmTS
PAGE NO.
INTRODUCTI ON . . . . . . . . . . . . . . . 0 8 . . ., e 8 0 . . . 8 . C' . . . . 0 D . . . . 0 . . . . . . . . . .. . . . . .1
DISCUSSION.o..........o...808ec80800GOGo..oeo8o....................1
Nickel .Electrode.......o................Q......o.............3
Zinc Electrode.........O....O..8.OOO...~.O....................3
Separator....o...:....o...e...................................4
EXPERIMENTAL
REStJLTS. . . . . . . . . . . . .. . . .. . .. . . o. . . II . . 0 ... . .. . . .. . . . . ..4
Storage Life of Standard Nickel Zinc Cells............o..o....4
Voltage-Current Characteristics of Standard and Advanced
Eagle-Picher Ni-Zn Batteries......o...o.....................5
Cycle Life Test Regtme.o.....o..o......o......o........o......7
Cycling Equipmento.........o.o.o..........o...................8
Cycle Life Data for Standard Eagle-Picher Ni-Zn Batteries.....8
Cycle Life Data for Advanced, High Rate Design, Eagle-
Picher Ni-Zn Batteries....... o. .:....0... ................... .10
Cycle Life Data for Advanced; High Rate Design, Energy
Research Corp., 4 Ah, Ni-Zn Batteries............oo..o......14
Cycle Life Data for Advanced, High Rate Design, Energy
Research Corp., 20 Ah, Sealed Ni-Zn Batteries...............16
Battery Heat Generation and Removal....o......o..............o17
Gas Generation..........o.....................................~9
Battery Failure Analysis on Standard and Advanced
Eagle-Picher Ni-Zn Batteries................................19
Battery Failure Analysis on Cells with Non-Sintered Positive
Electrodes...............o8o..o..o..........................21
Cells with Inorganic Separatorso..............................22
Recommended Battery Design...........o........................23
CONCLUSIONS...............................o........................24
REFERENCES.........................OG..............................25
v
-------
TABLE NO.
I
II
III
IV
v
VI
VII
VIII
IX
FIGURE NO.
1
2
3
4
TABLES
PAGE NO.
Construction Parameters for Eagle-Picher
Standard and Advanced Ni-Zn Batteries....................6
Capacity Maintenance of Advanced, Eagle-Picher,
High Rate Ni-Zn Battery #1H..............................ll
Capacity Maintenance of Advanced, Eagle-Pi cher ,
High Rate Ni-Zn Battery #2H..........................o...13
Battery Design Parameters for Energy Research Corp.,
High Rate 4 Ah Ni-Zn Cells..........o....................14
Performance of 5 Cell, Energy Research Corp.; 4 Ah
Ni-Zn Battery.....o....................o.................16
Battery Design Parameters for Energy Research Corp.,
20 Ah Sealed Ni-Zn Batteries.............................17
Calculation of Battery Heat Generation...................18
Failure Analysis of Eagle-Picher Ni-Zn Batteries.........20
Capacity Analysis of Nickel and Zinc Electrodes
after 16,500 Ah.........................................o21
FIGURES
Cycle Life Versus Depth of Discharge for Various
Battery System.s.................... .. .... .... .. .. ..... .. .26
Total Energy Output as a Function of Depth of Discharge..27
Charge Current Versus Charge Voltage as a Function of
State-of-Charge for Standard Eagle-Pieher Ni-Zn Battery..28
Charge Current Versus Charge Voltage as a Function of
State-of-Charge" for Advanced, High Rate, Eagle-Picher
Battery, . c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
vi
-------
FIGURE NO.
FIGURES (CONT)
PAGE NO.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Voltage-CUrrent Characteristics of High Rate Ni-Zn
Batteries as a Function of State-of-Charge.......................30-
Discharge Voltage for 350 Ampere Load............................3~
Cycling Histogram for Standard, Eagle-Picher Ni-Zn
Battery BB-674( )/U (#1).........................................32
Cycling Histogram for Standard, Eagle-Picher Ni-Zn
Battery BB-674( )/u (#2).........................................33
Deep Discharge Curves for Standard, Eagle-Picher Ni-Zn
Battery #1 After Pulse Cycling...................................34
Effect of Cycling on Deep Discharge Capacity and Voltage of
Standard Eagle-Picher Ni-Zn Battery #2...........................35
Cycling Histogram for Advanced, High Rate, Eagle-Picher
Battery #111.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
Power Test Performance as a Function of Life for Advanced,
High Rate, Eagle-Picher Battery #IH..............................37
Cycling Histogram for Advanced, High Rate, Eagle-Picher
Be. t tery f/2.H..... 0 . . . . . . . . . . . . . . . CI . . . . . . . . . II . . . . . . . . . . . .. . . . .,. . . . . .38
Power Test Performance for Advanced, High Rate, Eagle-Picher
Ba.ttery =#'2.H. . . . . 0 . . . . . . . . . . . . . . . . . . . . II . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Power Test Performance of Energy Research Corporation,
5 Cell, 4 Ah, N1-Zn Battery.......... . . . . . . . . . . . . . . . . . . . . . . . . . . . .40
Separator Removed from S'tandard Eagle-Picher
Battery #1 After 13,000 Ah.......................................41
Zinc Electrode Removed from Standard Eagle-Picher Battery #1
After 13,000 Ah..................................................14-2
Zinc Electrode Removed from Advanced, High Rate Eagle-Picher
Battery:/I!2H A~er 26,400 Ah......................................43
Nickel'Electrode Removed from Standard Eagle-Picher Batter.r #1-
After 13,000 Ah.................................................. 44
vii
-------
FIGURE NO.
20
21
22
23
FIGUR&S (COOT)
PAGE NO.
Nickel Electrode Removed from Advanced, High Rate, Eagle-
Picher Battery #2H After 26,400 Ah...............................45
Capacity Maintenance of Astropower Sealed 40 Ah Silver-
Zinc Cells[[[1J6
Charge-Discharge Characteristics of a Ni-Zn Cell with
Astropower Separator and Zinc Electrode..........................47
Capacity Maintenance of Ni-Zn Cells with Astropower Separator
-------
INTROWCTION
The goal of this study 'Was to evaluate and improve the characteris-
tics of nickel-zinc secondary batteries for use in hybrid electric vehi-
cles. The hybrid electric vehicle has a propulsion system consisting of
a heat engine and a battery. The heat engine supplies steady state power,
while the battery provides the transient power demands. The primary purpose,
for hybrid propulsion, is to reduce exhaust emissions fram street vehicles.
The nickel-zinc secondary battery possesses a unique combination of
properties which makes it a strong candidate for use in hybrid electric
vehicle propulsion systems. These properties include high energy density
(25 Wh/lb), a flat discharge, and the ability to charge and discharge at
high rates (greater than 100 Wh/lb) and low temperatures. Another feature
of interest for vehicle propulsion use, is the fact that electrode mate-
rials are plentiful and of reasonable cost. The overall cost of a nickel-
zinc battery can be projected with reasonable confidence to approximately
$O.15/Wh or $2-3/1b in large quantities. However, several possible
problem areas do exist regarding the use of nickel-zinc cells, as presently
designed, for this application; these are short cycle life and fairly
high maintenance.
DISCUSSION:
A nickel-zinc battery for hybrid vehicle applications should be capa-
ble of meeting the following leve1.s required by EPA:
Voltage:
200 - 220 V open circuit
150 V minimum closed circuit voltage
Maximum Power:
Discharge - 55 kW for 50 seconds
Recharge - 30 kW for 50 seconds
Maximum Capacity Required (At Maximum Power).:
5.0 Ah
Weight:
350 pounds, maximum
Cost:
-
$550 maximum
Maintenance:
Minimum maintenance during life is desired
Life: Should be capable of yielding above power after 5 years
and 400,000 cycles. The detailed breakdown of the total 400,000 cycles
is shown on the following page.
1
-------
Number Discharge ~ Ah Total
Cycles Rate (kWt Amps* Time Sec Out Ah Out
-
1,000 50 310 27.2 2.34 2, 340
5,000 55 343 8.6 0.82 4,100
5,000 55 333 5.2 0.48 2,400
389,000 10 55 10.8 0.164 64,100
*The currents shown are based on the average expected d1scharge voltage.
The number, length, and ampl.1 tude of the various cycles given - .
above are based on the power requirements for the Federal Driving Cycle for
a 4,000 lb family car (1). The Federal Driving Cycle consists of various
stops, starts, accelerations, decelerations, and steady state periods rep-
resenting actual driving practice in an urban environment. During the
course of this study, EPA modified the hybrid battery power requirements (2).
However, since the aim of this program was mainly to obtain information, rather
than hardware, no changes were made in the specified test cycles.
In order to properly size the battery for the required life, projections
were required of life versus depth of discharge. This information was not
available for the nickel-zinc system, at the hybrid electric vehicle
cycling regime. Therefore, first estimates were based on whatever informa-
tion was available for nickel-zinc batteries, plus any information for
nickel-cadmium and silver-zinc batteries. Figure 1 presents the cycle es-
timates for various battery systems as a function of the depth of discharge.
The curves presented represent averaging and interpolation of the data
available in Crane Naval Ammunition Depot Test Reports (3) and other reports.
Figure 2 presents the same data, converted to yield total Ah/Ah of rated
capacity versus depth of d1s charge. Conversion of data fran Figure 2 in-
dicates that a nickel-cadmium battery of about 12-15 Ah capacity would be
capable of yielding the 73,000 total Ah required for the hybrid vehicle:
regime. A silver-cadmium battery would have to be rated at 73 Ah, while
a silver-zinc battery might have to be l!jo Ah size. The silver systems,
however, would not function for the 5 year requirement because of silver
attack on the separator.
Projections of nickel-zinc cycle life, based on data obtained for 60~
depth of discharge, indicate that a 3J Ah nickel-zinc cell would meet
approximately 1/2 the EPA hybrid vehicle battery requirement of 400,000
cycles or 73,000 Ah previously described. This expected level of perfor-
mance would equal the EPA requirement for Pb-acid"batteries of 36,500 Ah
total. An improved nickel-zinc battery would be needed to meet full
requirements. However, because only one data point for nickel-zinc was -
available at the start of the program, considerable testing was required
to determine the actual life.
2
-------
Three main areas of concern in the design of nickel-zinc cells for the
hybrid vehicle applications were the nickel electrod~zinc electrode, and
separator.
Nickel-Electrode
The presently available sintered nickel electrode structures are more
than capable of achieving the desired life and rate capability. In fact,
if the normal ratio of nickel to zinc capacity (1/3) was used, the nickel
electrode would be considerably overdesigned. As the data for nickel cad-
mium cells in Figure 2 indicates, it would be possible to meet the require-
ments with a nickel capacity of less than 15 Ah. Because of plate size
considerations, with sintered electrodes, it was not feasible to reduce
the nickel capacity that low. However, a. nickel capacity of 20-25 Ah was
both feasible and desirable. The required nickel capacity was achieved
through the use of lightly loaded, thin nickel electrodes. This resulted
in improved high rate performance, potentially lower cost and weight.
Further small improvements in performance could have been achieved through
modifications of plaque preparation parameters. This might have included
such changes as increasing sintering temperature and time to achieve
stronger more conductive plaques. However, time did not permit investi-
gation of these variables.
The use of a non-sintered nickel-electrode structure permits tailoring
the plate thickness, loading, and conductive additives to optimize for
this application. Energy Research Corp. (4) has developed a non-sintered
nickel electrode that can be made thinner than plaque types and has a
lower weight per Ah. Cells using these non-sintered electrodes will be
more fully described later in this report.
Zinc Electrode
Teflonated, oversized, and cQntou.red zinc electrode structures have
yielded high cycle life in both silver-zinc (5,6) and nickel-zinc cells
(7) cycling at 62~ and lOO~ depth of discharge. Up to &>0 cycles, which
equals a total of 360 Ah/Ah of rated capacity, has been obtained in silver-
zinc cells at a 62~ depth of discharge regime. In most cases, failure was
due to separator deterioration. The zinc electrode was still capable of
yielding more than 62~ of rated capacity. In silver-zinc cells, the
theoretical capacity of the zinc electrode was 3 times the rated capacity.
Similar zinc electrode theoretical capacities have been used in nickel- .
zinc cells. Studies have indicated that increasing the amount of negative
electrode material will increase cycle life considerably. With a lower
capacity positive, it was possible to use the space and weight available
to further increase the zinc capacity.
Two minor areas still needing further investigation to optimize the
zinc electrode are the percent amalgamation and the grid material. Tests
3
-------
at ECOM have indicated that maximum cycle life and zinc stability are ob-
tained with minimum mercury contents, 0..0. 5~, rather than 2~ mercury which
is used for minimum gassing. However, for sealed operation 2~ mercury may
be required.
Silver-zinc and nickel-zinc cells are normally fabricated with silver
Exmet type grids since they have the highest conductivity and do not promote
self discharge of zinc. However, the cost of a silver grid is high; in
fact, it may represent a major portion of the cost of a finished zinc elec-
trode. Tests of cells with iron grids have been satisfactory to date from
the standpoint of performance. However, hydrogen generation is higher.
Nevertheless, the rate of self discharge of the negative would still be
below that of the positive. Additional experimentation in this area will
be required to obtain the lowest cost zinc electrode consistent with good
cycle life and mintmum hydrogen generation.
Separator
The weakest element in the nickel-zinc system is the separator.
Organic materials, such as cellophane, can be oxidized and weakened by
the attack of nascent oxygen generated during overcharge. This then re-
sults in premature shorting, as zinc grows through the cracks fonned.
However, over 200 deep cycles, equal to a total of 150 Ah/ Ah of rated ca-
pacity, have been obtained in cells with cellophane or sausage casing
separators if certain procedures are followed. These include coating the
exposed edges of the separator with vinyl (8) and the use of zinc elec-
trodes that are designed for minimum shape change.
Radiation grafted polyethylene has greater physical stability and
resistance to shorting than cellulosic separators. However, it has lower
conductivity which reduces power density and causes more rapid shape
change on the zinc electrode.
Non-oxidizable inorganic separators offer the best hope for increasing
the cycle life of the nickel-zinc system. It is estimated that cycle life
could be at least doubled, if separator failure is eliminated, before the
limitations of the zinc electrode begin to be felt. Inorganic separator
materials investigated include the materials originally developed by Douglas-
Astropower and now available through NASA-Lewis, and the l~er hydroxide
inorganic separator structures developed by Energy Research Corp.
EXPERIMENTAL RESULTS
Storage Life of Standard Nickel-Zinc Cells
The capacity retention of the nickel-zinc system on stand is approxi-
mately the same a~ that for the nickel-cadmium system. For example; data
4
-------
reported by Broglio (9) indicates that capa.ci ty retention of the 35 Ah
Nickel-Zinc BB-674()fu* Battery is 75~ after 28 days at 25°C and approxi-
mately 50~ after 7 days at 60°c. The ~sing rste of this battery on
stand at 25°C is approximately 5 cm3/Hr/cell. If this gas is all hydrogen,
as would be expected, then it would correspond to a capacity loss of .25
Ah/day or 7 Ah for 28 days. Seven Ah represent a capacity loss of
. 20~ over this period of time, compared with a 25~ capacity loss by the cell,
indicating that the nickel electrode controls capacity loss on stand.
The effect of prolonged wet uncontrolled storage on the nickel-zinc
system was detennined on a limited number of sample cells approximately
2-2 1/2 years old. Though several types fran different manufacturers were
available, only those fram General Telephone and Electronics (GT&E) were
available in adequate quantity and have sufficient reliability for a
meaningful evaluation. The construction of these cells, which were activated
April 1969, 'was given by Blickwedel (10). Seven-Ah cells containing
.' two different types of negative electrodes were charged and di scharged. for
up to 4 cy~les. The capacities obtained were then caupared with the initial
performance of cauparable cells. Ai'ter 2 years of storage the cells with
Tef~onated zinc ~ectrodes (25.7 Ah theoretical zinc capacity) yielded ca-
paci ties equal to or greater than 'WIE'4'e obtained initially, except for the
first cycle after storage. The capacity on this cycle was only 50~ of that
obtained before storage. Cells with sprayed zinc electrodes (13 Ah theoreti-
cal capacity) gave only 60~ of their initial capacity after 2 years of .
storage after 3 cycles. It is thought that poor initial capacity after dis-
charged stand is due to a partial passivation of the zinc. Therefore, with
a large excess of zinc oxide such as was used in the Teflonated zinc elec-
trode cells, and in cells designed for hybrid operation, no problems
should be encountered if the hybrid batteries are stored discharged for
several years prior to use.
Voltage-Current Characteristics of Standard & Advanced Eag1e-Pich~
Ni-Zn Batteries
Charge current versus potential as a function of state of charge, . and
discharge voltage for high rate discharges were determined for 3 types of
6.4v, nickel-:Unc batteries. They were the BB-674( )/U and two high rate
designs in the same size case manufactured by Eagle-Picher Industries (EPI).
The construction parameters of these three battery designs are given in
Table I.
* Military Standard nanenclature, use of "( )" in nanenclature
battery has not been formally procured.
indicates
5
-------
Table I.
Construction Parameters for Eagle-Picher Standard and Advanced
Ni-Zn Batteries
High Rate High Rate
~ ~ ~
Rated Capacity (Ah) 35 24 24
Initial Capacity (Ah) 38 ,24.5 24
Weight 8 lb 12 oz 8 lb 15 oz 9 lb 6 oz
Positive Electrode number 10 +2 16 16
size (10~ 2.81 x 5 x .037 2.81 x 5 x .020 2.81 x 5 x .020
(inches) (2 2.81 x 5 x .025
Negative Electrode number 11 17 17
size (in) 2.81 x 5 x .035 2.81 x 5 x .021 2.81 x 5 x .024
Theoretical Capacity (Ah) 150 125 155
Negative Tabs 2, .016" Dia Ag' Wire 2, .01" x 1/8" 2, .01 x 1/8"
Ag Foil Ag Foil
Separator (going from positive Pe1lon bag/2TFSC' 2TFSC/Pellon bag 2TFSC
to negative electrode) Pe1lon bag
Active Area (in2) ~ 436 436
*FSC (Fibrous Sausage Casing)
Voltage versus current was also determined for a battery consisting of 4 each
35 Ah cells manufactured by the General Electric Co. for another use. However,
these cells had an active area of only 165 in? and therefore were not able to
meet the requirements for the hybrid application.
The voltage-current characteristics of the batteries were determined
using a test method designed to simulate actual use conditions. The battery
was first run on the life cycle regime for several hours, as described later;
then following a series of discharge pulses, the charge voltage was varied and
the average charge current measured. The battery was then subjected to a series
of discharge pUlses, 10.8 seconds in length, at various currents to determine
discharge voltages with charge periods of 50.2 seconds interspersed to return
the battery to full charge. The battery was then discharged at 55 A until
25~ of rated capacity had been removed (8.75 Ah for the BB-614( ) /u and 6 Ah
for battery #2H). The charge and discharge regimES described above were repeated
and current-voltage measurements were taken.
6
-------
The amount of charge and dischaTge capacity entering and being with=
drawn f':rom the battery was measured on 2, HBS* dig! tal ampere=hou.r meters 0
Using this exact knowledge of the ba.ttery stateaOf-charge, additional
capacity vas withdrawn at 55 A to reduce the battery's state'3of-charge
'to 5O~ of rated. The current-voltage characteristics for both charge and
discharge were again determined. Figures 3 and 4 show the charging charac-
teristics of the BB-674( )/u and the high rate battery (2 H)..respectively,
as a function of state-of-charge, while Figure 5 canpares their discharge
voltage-current characteristics. The 50~ state-of-charge could not be
maintained because of the high charge acceptance of the batteries. There-
fore, the data for ~ state-of-charge represents a state-ot-charge range
iran ~ to 65~. Depending on the minimum acceptable voltage, power d~nsi-
ties exceeding 3X> W/lb can be obtained vith the high rate designed Ni-Zn
battery and up to 200 W/lb with the BB-674( )ju. Both battery types were
run at a 350A load for 50 seconds. to simulate the 55kW power test. The
vol tages obtained are shawn in Figure 6. The improvement in discharge
voltage with the high rate design ws due to the fact that it has 45~ more
electrode area, increased tabbing on both positive and negative electrodes,
heavier terminals, and intercell connectors.
Cycle Life Test Regime
The most important consideration in choosing a Ni-Zn battery is how
it will perform over a cycling reg~e representative of actual vehicle use
conditions. Obviously a cycling regime carried out over a period of 5
years is not practical, therefore, a simulated regime had to be devised. The
following regime was chosen as a reasonable compranise between the conflic-
ting demands of realism and testing time.
1. 1 pulse at 10 kW rate (551'. based on 2l0V battery discharging at
aa average l82V) for 10.8 seconds every minute. for 29 out of every 30 minutes.
2. 1 pulse at 55 kW rote (350A based on 2l0V battery discharging at
an average l57V) for 10.8 seconds every 30 minutes.
The following energy would be supplied by the battery each d~:
55A X 10.8 seconds X 58 pulses X 24 hours = 228 Ah
3600 hour
350A X 10.8 seconds X 2 ~lses X 24 hours = 50 Ah
3600 hour
Total EneTgy per Day
= 278Ah
To yield a t~tal 73,000 Ah would require 262 days.
* HBS Equipment Division, Los Angeles, California.
7
-------
The load, therefore, would be made up of~,400 10 kW pulses + 12,600
55 kW pulses. Recharge rates were high (120A maxtmum) in order to fully re-
charge the cells in the short time available. Maximum charge acceptance is
important in hybrid service in view of a t.actor such as regenerative braking.
In addition, periodic test discharges of 55 kW for 50 seconds (power test)
were to be performed approximately once a week to determine the ability
of the battery to meet the power requirements. Deep discharges at the C/2
rate were to be performed after every 3rd power test.
During the course of cycling, several variations to the above cycling
regime were made for a variety of valid reasons; tor example, the dis-
charge loads were run at fixed currents rather than fixed power, for easier
comparison of different cell designs. Thus, depending on the output voltage,
the load could exceed or be less than the specified power levels. In most
cases, except at the end of life, the batteries generally delivered more
power then the specified levels. Deep cycles were usually run more fre-
quently than scheduled in order to obtain additional data and to recondition
the batteries.
Cycling Equipment
The cycling equipment consisted of timing clocks, a Harrison Model
6456B power supply, and a water-cooled MetrodYne (800 Amp) load module.
Measurement of the total charge input to the battery in Ah was obtained
fran an HBS digital ampere-hour meter. Protection against battery failure
was provided by a voltage sensing Beede meter relay. Any sudden drop in dis-
charge voltage, such as wOuld occur if a short occurred, disconnected the
charging and discharging units. The 55 kW, 50 second power test 'Was run on
the same equipment with minimum modification. Charge current and battery
voltage were recorded on a Hewlett-Packard Model 7100B 2 pen recorder.
Cycle Life Data For Standard Eagle-Picher Ni-Zn Batteries
The initial batteries cycled were BB-674( ) /U' s . They were available
at the start of the program and were used to determine the general response
of the Ni-Zn system to the hybrid vehicle cycling regime, even though it
was recognized that they delivered insufficient voltage to pass the 55 kW
power test.
Figures 7 and 8 are histograms of the cycling run on 2, 6.4 v,
35 Ah BB-674( )/u :Batkries. The histograms include information on the
discharge voltages at 55 and 350 A, the battery external temperatures.. and
the percent of the charge current going into overcharge. The overcharge
percentage is determined as follows:
Ah-Ah
o Ah i x 100
i
Ahi - Total Ampere-hour input to battery during time (T).
Aho = Total Ampere-hour output to lCBd during time (T).
T = Time period for measurement (usually 24 hours or more)
~ Overcharge =
8
-------
Battery #1, shown in Figure 7, was filled with electrolyte in April
1971 and run for a total of 16, 650 Ab (equal to approximately 90,000 cycles)
before cycling was terminated January 1972 because of low voltage during the
350 A discharge and "the development of slow short circuiting. furing the
course of the cycling shown in Figure 7, there was a gradual decrease in
discharge voltage especially at the 350 A rate. Periodic deep discharges
increased the subsequent pulse voltage levels, but the effect was not
permanent. This battery was charged at 2 voltage levels, 7.4 and 7.6 volts,
or 1.85 and 1.90 V/cell, respectively. Since the lower potential effectively
maintained charge and minimized overcharge, 704 V was chosen as the preferred
constant potential limit. The typical percentage of overcharge at 7.4v was 2-3%
except when the zinc electrode was not fully charged, then the overcharge was
7-10%. This 7-10% overcharge was measured at the beginning of charge, when
uncharged zinc oxide was being charged, and at 6,500 Ah after a reversal, when
the zinc had been completely discharged. Excess rates of overcharge were also
measured after 14,000 Ah of cycling when shorting developed.
Battery temperature generally showed a slow increase with time and
can be related to a corresponding drop in battery discharge voltage. In
general, the temperatures measured were not excessive except at the end of
cycling, when shorting occurred and the excess charge was 13% of tbe output
current.
Battery #2, shown in Figure 8, was filled with electrolyte in January
1972 and run for a total of 13,000 Ah before cycling was terminated in March
1972 due to short circuiting after a deep cycle. During the cycling of this"
batter;y; one cell developed a short at 5,000 Ah and was by-passed, with cycling
continued on the remaining 3 cells. In order to simplify tbe data in Figure 8,
the voltage values obtained have been reported on the basis of a 4-cell mono-
block. The same general pattern of decreasing voltage with cycling was noted
with battery 12 as for battery #1. However, the voltage levels achieved for
the 350 A discharge were up to 25% higher for battery =/t2, compared to battery
#1. This may be due to the fact that battery #1 had been deep cycled 7 times
at varying environments and had stood for 6 months prior to the start of
cycling. Battery =/t2, on the other hand, was given only one deep cycle and
then started immediately on cycling.
The effect of pulse cycling, on deep cycle capacity, is shown in
Figures 9 and 10 for batteries #1 and /12, respectively. Tbe voltage degra-
dation for battery #1 is considerably more severe than the corresponding
capacity loss; for example, at 13,000 Ah output, battery #1 yielded 17 Ah
to 5.0 V while battery =/t2 yielded 17.3 Ah. The corresponding average
voltages for the 14 A discharges were 5.55 V for battery #1 and 6.35 V for
battery lie. The discharge curves in Figure 9 for battery #1 show a consid-
erable slope, indicative of an increase in internal resistance with de-
creasing state of charge. Reference electrode measurements made at 6,500 Ab
9
-------
indicated that most of the voltage loss was due to the zinc electrode.
By discharging to reversal, to convert all the active zinc to zinc-oxide,
the curve shown in Figure 9 for 6,540 Ah vas obtained. This curve is flat
and remains at a high level throughout the discharge and is typical of a
nickel limited discharge. The other curves on Figure 9, however, are
more indicative of zinc limited discharges. Figure 10 shows that battery
#2 did not suffer excessive voltage losses due to cycling. Failure analysis
was perfonned on batteries #1 and #2 to determine the cause of failure
and the reasons for the differences between them.
Cycle Life Data for Advanced,High Rate Design, Eagle-Picher
Ni-Zn Batteries
Three high rate designed 6.4v monoblocks were received from Eagle-Picher
Industries; two of them were started on cycling during March 1972.
The specific design features of these batteries were presented in Table I.
The dif:ference between them, was that Battery #1H was built with pellon
bags on the negative electrodes, While Battery #3H had no pellon bags, but
an increased weight o:f zinc oxide. Figure 11 is a histogram o:f the cycling
run on Battery #1H. The histogram includes in:formation on the discharge
voltages at 55 and 350 A, the battery external temperatures and the per-
cent o:f the charge current going into overcharge. Battery #3H was run
under the same conditions as IJJI but only lasted for 3700 Ah be:fore a
power test :failure indicated some di:f:ficul ty. A deep cycle was run and
upon recharge it was determined that one cell o:f :Be.ttery #3H had shorted.
It was :felt that the removal o:f all the pellon interseparator made this de-
sign more liable to shorting. No gain in voltage was achieved by elimina-
ting the negative pellon bag. There:fore, cycling was terminated on Battery
#3H at 3,700 Ah, but continued on Battery #1H to :failure. As Figure 11
shows, Battery Hill maintained good capacity and voltage levels until shorting
developed. During the course o:f cycling, this high rate battery
operated at a high voltage level :for almost 8,000 Ah when lack o:f electrolyte
became apparent. This extended period without water or electrolyte addi-
tions was made possible by the low average overcharge required by this bat-
tery. This also resulted in a lower operating temperature. After 12,000
Ah o:f operation, slow shorting apparently developed in one or two cells.
The battery was deep cycled, recharged, and then subjected to reversal in
an attempt to break the shorts. This attempt was unsuccessful, as shown
by the complete :failure within ~ Ah a:fter cycling was reinitiated. Ex-
cept for the shorting this battery appeared capable o:f considerably more
cycling. Table II gives the deep cycle capacity and average discharge
voltage.
10
-------
Table II.
Capacity Maintenance of Advanced,Eagle-Picher,
High Rate. Ni-Zn Battery 1t!!
Total Ah
Delivered
Initial
*
Capacity ~
24.5
Average Discharge
Voltage (V)
@ 3,700 Ah
@ 9,074 Ah
23.1
14.0
6.4
6.2
. @ 9,100 Ah after addition of KOH
21.0
6.3
6.3
@ 12,412 Ah
@ 12,433 Ah
21.0
~5.4 Ah to 5.0 V**
+50 Ah to L 0 v***
6.5
*Batteries discharged at 14 A to 5.0 V; charged at 4 A for 6.5 hrs.
** Some loss of capacity due to shorting on stand
***Battery reverse discharged at 3.0 A.
Figure 12 shows the voltage output during the power test of :Battery
#lH as a function of the ampere-hour output of the battery. Af'ter 9,800
Ah, when electrolyte had been added to the battery, it was still unable to
supply 50 seconds continuously at 350 A. However, it might be possible to
supp~ two 25 second periods at this current. At 12,300 Ah, the loss of
voltage on the power test reflects the development of shorting in several
cells.
High rate Battery #eH, which was identical in construction to #1lI,
was started on cycling af'ter the failure of #lH. The charge voltage for
Battery ff2H was reduced to 7.2 from the 7.4 V used to charge Battery #lH in
an attempt to reduce overcharge and lower the average battery temperature.
It was hoped that this step would preserve separator integrity, prevent
shorting, and extend cycle life. A histogram of the cycling behavior of
Battery #eH. is shown in Figure 13. When Figures 11 and 13 are canpared, it
can be seen that the reduction in charge voltage did have several effects,
first, the temperature was approximately 10°F lower; second, the amount
of overcharge was loweredj thlrd,the discharge voltages were somewhat
lowerjand fourth,the number of cycles to the first short was increased by
75'/0. The lower voltage at the 55 A discharge is caused primarily by the
lower starting voltage for the discharge, probably due to a reduced amount
of more active nickel-oxides. A reduction in the charging voltage of 0.2 V
(0.05 V/cell). caused a reduction of 0.15 V in the battery discharge voltage
11
-------
at 55 A, while the later increase in charging voltage of 0.1 V to 7.3 V
resulted in an 0.08 V increase.
The lowest possible charge voltage appears to give the greatest
efficiency. For the 350 A load, the discharge voltage was reduced 0.2 V by
the reduction of 0.2 V in the charging voltage, however, it is believed
that this is primarily due to the fact that the battery is cooler and
hence has a higher resistance. On cycling, after Batteries #ll! and {IeH
had been on stand and were at approximately the same temperature, the
voltage difference caused by the lower charge voltage was considerably
smaller. During the course of cycling Battery I!eH, occasional low voltages
and unusual voltage dips were recorded during power tests, and also start-
ing at about 5,200 Ah, on sane of the 350 A pulses. These low pulse
voltages occurred somewhat randomly and appeared to indicate a possible
reduction in battery state-or-charge. Accordingly, the charge potential
was raised from 7.2 to 7.3 V at 6,500 Ah. This increased the discharge
voltage, as previously mentioned, but did not completely eliminate low
voltage pulses.
Figure 14 shows typical power test output voltages for :Battery I2H
after 8,530 Ah of operation. The variation in performance for the 350 A
discharge is clearly seen from the difference between Gurves A and B in
Figure 14. The only difference in conditions between the curves 18 that the
higher one, Curve B, 'WaS run 2 hours a:f'ter the other. The ~oss 1n vo~tage
shown in Curve A was caused by one cell going to reversal. Appareritly
the reversal of a cell reconditioned it for a later cycle. The reversal
of a cell was also the cause of the periodic low voltages at the 350 A
load. However, during the course of cycling, the same cell was not always
the limiting one. Reference electrode readings were taken in an attempt
to determine if only one electrode was limiting. In general, the nickel
electrode reversed first, however, the zinc electrode also reversed voltage.
very soon thereaf'ter. However, because of the wide variability of condi-
tions under which the low pulse voltages occurred and also their unpredict-
ability, it is impossible to ascribe a cause for this most puzzling behavior.
Deep cycling was the only treatment that appeared to eliminate the low
voltage pulses for any length of time. However, its effect did not last
for more than liOO Ah of cycling.
The deep cycle capacity data for Battery #2H are shown in Table III.
Examination indicates that the battery did not suffer any significant ~oss of
capacity due to cycling at a_low charge potential. . However, charging at 7.3 V
resulted in a slightly higher state-of-charge than charging at 7.2 v.
Capacity maintenance with cycling was excellent for this battery, with
more than 85~ of initial capacity attained after 16,710 Ah output (equiva-
lent to 92,000 cycles), '
12
-------
Table III.
Total Ah
Delivered
Charge
Voltage
Initial
3,600
5,850
7.2
7.2
8,500
16,710
7.3
7.3
21,500
7.3
5.5
21,512
21,530
21,568
21,577
5.5
5.5
26, 370
26, 380
3.65
3.65
26,400
Capacity(Ah): Average
Discharge Voltage
24.0 6.3V
18.2 6.4v 8
20.3 6.3V 1
22.7 6.3V
20.4 6.2V
12.3 6.3V 0
18.2 4.7V 3 cells only
38.0 0.5V residual
zinc capacity
9.0 4.7V
11.3 3.0V 2 cells only,
2nd cell shorted
10.0 2.8V 0 II
12.0 2.9V
Cells shorted after overnight stand
*Batteries discharged at 14A to 5.0Vj charged at 4A for 6.5 hours.
8 Water added after cycle.
~ KOH added during charge, one cell apParently shorted.
The first failure in Battery #2H occurred when one cell was dis-
covered to have shorted at 21,500 Ah during the deep cycle run. A series
of deep cycles were then run in an attempt to:recondition the battery and
possibly break t,he short circuit. This did not prove feasible and, in .
fact, continued deep cycling only resulted in the short circuiting of another
cell.
13
-------
The two remaining cells. were then returned to the vehicle simu-
lation regime. However, the load module used for the vehicle cycling
test was unable to maintain \'i. constant current of 350 A at input voltages
of less than 2.5V. Therefore, in order for the total Ah delivered per
day with 2 cells to be th~ same as with 4 cells, it was necessary to in-
crease the frequency of the high rate pulse fran once every ~ minutes
to once every 15 minutes. This change maintained the total Ah . de-
livered per day relatively constant at approximately 260-280 Ah per day.
At 26,000 Ah the voltage at the 55 A load was 3.18 V (1. 59V/cell) , and
at the 350 A load (final current approximately 200 A) was 1.87 V (0.93
V/cell). These output voltages were considerably below those obtained
earlier in the life of the battery and below the voltage considered ac-
ceptable for hybrid vehicle use. Cycling was terminated at 26,370 Ah
and several de~ cycles were run, as indicated in Table III. At this
point Battery :fIeH had only half its initial capacity and had suffered
considerable loss in voltage. The battery was then subjected to failure
analysis. .
Cyc.le Life Data for Advanced, High Rate Design, Energy Research Corp.
4 Ah, Ni-Zn Batteries .
High rate 4 Ah Ni-Zn cells having a theoretical capacity of 5.8 Ah
and an initial capacity of 4.4 Ah were received fram Energy Research
Corp. (ERC). These cells had non-sintered nickel electrodes. Because
of the difficulty in manufacturing a pin-hole free, thin inorganic separa-
tor, these cells were made with a 1 turn sausage casing separator. The
specific construction details of these cells are given in Table IV.
Particularly interesting was ERC's ability to manufacture very thin posi-
tive and negative electrodes, which made a high surface area possible.
Table IV. Battery Design Parameters for Energy Research Corp.,
High Rate, 4 Ah Ni-Zn Cells
Rated Capacity (Ah)
Initial Capacity (Ah)
4.0
4.4
Positive Electrode:
(Non-sintered type)
number 16
size (inches) 2 3/8 x 1 7/8 x .0165
Negative Electrode:
i!1eoretical Capacity (Ab)
number 17
size (inches) 2 3/8 x 1 1/8 x .008
14
Separator (Pellon/Fibrous Sausage Casing)
(going tram positive to negative electrode)
A\';tive Ar~
Weight
133 in2
3/4 Ib
14
-------
A 5-cell battery was constructed and started on cycling after an
initial cycle. Because of the high surface area available with these
cells, it was decided to cycle them at an accelerated rate. The high
current pulse was 100 A, and the lower current pu1se was 16 A. These
currents are approximately 6f4 higher than those based on a straight
capacity ratio with the 24 Ah high rate design. At the hi~ rate
(100 A), the power density of this battery exceeded 200 Wh/lb.
The performance of this battery on the 50-second power test is
shown in Figure 15. The voltages obtained (1.29 V/cell at the end of
50 seconds, after 420 Ah of cycling) compared favorably with the
1.2 V / cell obtained fran Battery #2H after 2300 Ah of cycling. The
current density for the ERC 4-Ah cells was 0.75 A/in2 at 100 A, while
for Eagle-Picher Battery #2H it was 0.8 A/in2 at 350 A. This power
test data appears to indicate that the use of a non-sintered nickel
electrode for hybrid vehicle service would not cause any significant
voltage loss. However, after a total 1300 Ah output, on1y 15-20 seconds
were obtainable on the power test above 1.0 V / cell. Deep cycling was
effecti ve in improving the power test performance. Nevertheless, the
battery, after 1300 Ah output, was never capable of meeting the fu11
50-second power test requirement. These power test results, however,
were not significantly different from those obtained with the sintered
nickel electrodes.
The voltage characteristics of the battery during life cycling
are given in Table V. Based on the capacity available in the battery,
the life obtained was equivalent to 32,000 Ah for the fu11 sized
Eagle-Picher high rate battery which was cycled for a maximum of
26,400 Ah. At the completion of cycling, a failure analysis was
performed and will be discussed later.
15
-------
Tab1.e V. Performance of' 5 Cel1., Energy Research Corp., 4 Ah Ni-Zn Battery
Vo1.tage @ 100 J!r Vo1.tage @ 1.6 A*
Ah BatterYlO.75 A/in2) Cell. BattervlO.12A/in2) Cell.
65 7.28 1..46 8.9 1.78
1.40 7.23 1.45 8.83 1.77
500 7.08 1.42 8.78 1.76
650 6.97 1.39 8.77 1.75
875 6.90 1.38 8.77 1.75
936 7.01. 1.40 8.80 1.76 ~@
1010 6.96 1.39 8.80 1.76
1382 6.75-6.25 1.. 35-1. 25 8.73 1. 74 ~
1.814 6.35-6.00 1..27-1.20 8.60 1.72
2388 5.5 1.1 8.58 1.74
2540 6.90 1.. 38 8.72 1..74 @@
2760 5.48 1.37 6.92 1. 73 one cell.
shorted
2960 4.01 1.. 35 5.1. 7 1. 72 2nd cell.
shorted
3034 4.02 1.34 5.1.4 1..71.
3410 3.96 1.. 32 5.15 1.71.
3770 3.94 1.. 31 5.1.5 1..71.
4283 3.94 1.31 5.1.6 1.72
4456 3.90 1.30 5.12 1..71.
4535 4.ll. 1.37 5.1.5 1..71. @
4744 4.1.3 1.38 5.1.8 1..72
481.9 3.85 1.28 5.05 1.68 {High
51.08 204 0.80 5.02 1..57 overcharge
5313 2.1. 0.70 5.0 1.66 8-15~
*Batteries charged at 9.2 V (1.84 V/cell.)
o After KOH addition
@ After deep cyc1.e
G) Variation in EX>D vo1.tage at 100 A 1.oad
Cycle Life Data for Advanced, High Rate. Design, Energy Research Corp.
20 Ah, Sealed Ni-Zn Batteries
. Two f'ul.l. size 20-Ah. batteries constructed in the same cel.l. cases as the
BB-674( ) /u were ,received from Energy Research Corp. at the very end of this
project 0 These batteries incorporated sane of the design features that cou1.d
1.6
-------
result in a better nickel-zinc battery for hybrid service. Tab~e VI styes
their specific design parameters. The batteries were seued, with a ~ psi
v~ve, to eliminate maintenance. The separator system incorporated iMl"-
ganic materi~ as an interseparator. One J.ayer of cellophane was inQ1.uded
to stop zinc penetration, which would occur if o~y the ERC inorganic
materi~ was used. Two Ag-Hg third e~ectrodes of sufficient area to
permit a continuous overcharge of approximately 5 A were incorporated in
these batteries.
O~y initiu data were obtained with these batteries and result.
indicated possib~e prob~em areas. Initiu vo~tages obtained were co..1d-
erab~y ~ower than expected: 1.66 V/cell at 55 A, and 1.08 V/cell at 350 A.
The batteries were charged at 7.3 V (1.83 V/cell) and received approxlMte~y
~~ overcharge. The heavy potted construction, emp~oyed to prevent battery
swelling, has resulted in great~ increased weight and poor heat traufer
. characteristics. Further evuuation of these batteries could not be
carried out because of the comp~etion of the test period.
Tab~e VI.
Battery Design Parameters for Energy Research Corp. 20 .Ah
Seued Ni-Zn Batteries
In1ti~ Capacity (Ab)
20
Positive Electrode
(Non-sintered type)
Theoretic~ Capacity (Ab)
Number - ~6
Size - 5 ~/8" x 2 7/8" x 0.0~8"
24.5
Negative Electrode
Theoreticu Capacity (Ab)
Number - ~6 ~, 2 hail on ends
Size - 5 ~/8" x 2 7/8" X 0.023" (full)
0.0~3" (haJ.t)
~O
Separator - ERC 200l/PUDO 300/FI{C 200~
(going from positive to negative e~ectrode)
Third Electrode
Si~ver-Mercury Type
2 each 5 ~/2" x ~ ~/8"
50-60 psi
Pressure Relief
Active Area
472 in2
9.~ ~b (potted)
Weight
Battery Heat Generation and Removu
p
A key consideration in the design of a ~ size hybrid batte17
for vehic~e service is the amount of heat wi11ch will be generated and,
therefore, the amount of conductive structure which will be required
to remove it.
~7
-------
Battery heat generation may be closely approximated by taking
the difference between the electrical input and output. It will be
assumed, for the purpose of this calculation, that all the gas gener.
ated by overcharge is recombined. If any gas is vented from the battery,
then the amount of heat generated would be somewhat lower than that
shown by the calculation in Table VII.
Table VII.
Calculation of Battery Heat Generation
BB-674( )/U ~ (a)
@ 3,000 Ah output
Energy Input/Hour
Energy Output/Hour
7.4 V x 11.8 Ah . 87.4 Wh/h
6.75 V x 9.5 Ah . 64.1 Wh/h
4.7 V x 2.1 Ah. l.9
TOTAL 7 .0Wh/h
High ~te Battery #1lI (b)
@ 3,000 Ah output
7.4 V x 11.7 Ah . 86.6 Wh/h
6.9 V x 9.5 Ah . 65.5 Wh/h
5.6 V x 2.1 Ah . 11.8
77.3Wh/h
Heat Output 87.4 - 74.0 = 13.4 Wh/h = 45.1 ~u/J1 86.6 - 17.3 = 9.3 Wh/h =
31.'7Btu/h
( a) Data:f'rom Figure 8
(b) Data:f'ran Figure 11
NOTE:
Voltages are average discharge voltages.
At the calculation points, which have been chosen to represent
approximately steady state conditions, where all the uncharged zinc
oxide has been charged, a BB-674( )/U cell would evolve 11.45 Btu/h,
while the high rate design cell would evolve only 7.93 Btu/h. Heat
generation could be further lowered by increased plate area and con-
ducti ve members. An even more effective way of reducing heat was a
reduction in the charging voltage. When battery #e.H was charged at
7.2 V at 5,000 Ah output, the calculated heat generation was only
6.8 Btu/ce11/h. Therefore, based on a heat generation of 7.93 Btu/
cell/h, a full size l2O-cell battery would evolve 952 Btu/h on the
simulated regime. This amount of heat could be easily removed in ve-
hicle operation. As the battery ages, heat generation will increase
approximately 70~ at the point when the battery is no longer capable
of meeting its power requirements. This amount of heat (1700 Btu/h)
can still be easily removed.
Based. on' a very limited number ('jf runs of the ERC sealed 20 Ah
'Dattery, it would Gppear that a sealed battery with uncharged zinc-oxide
18
-------
may evolve twice as much heat as a comparable vented one. This is
due to the fact that the percent overcharge remains at approximately
loop for the sealed battery, while in a vented battery the overcharge
percentage drops fram 8-l~ to l-~ after all the reserve zinc oxide
is charged.
Gas Generation
Several measurements of gas generation, by individual cells in the
various batteries tested, were made during the course of this project.
Gas generation was found to be a function of the state-of-charge of
the battery (zinc electrode) and the percent measured overcharge.
Typical gas evolution rates were 50-100 cC/h/cell. In general, ex-
cept for the initial 1000-2000 Ah of cycling, the gas evolved is
1/3 oxygen and 2/3 hydrogen. During the initial cycles the gas
evolved is mainly oxygen. Calculations of actual vs theoretical gaa
have indicated that, in the typical vented nickel system, negligible
recombination of gases occur. With a low charging voltage, the per-
cent overcharge can be reduced, and "hence the amount of gas generati.
and water makeup required.
Battery Failure Analysis on Standard and .Advanced Eagle-Picher
Ni-Zn Batteries
Upon the completion of cycling, individual cells were dissected
and failure analysis was conducted. At a minimum, this analysis con-
sisted of a visual examination and measurements of the positive and
negative electrodes and separator, and a carbonate analysis of the
electrolyte. Table VIII lists the tests performed on each battery and
the results. In addition, electrodes were removed from BB-674( )/U *1
after 16,500 Ah and built into test batteries with new counter electrodes
and separator. Table IX lists the various electrodes evaluated and
their performance.
From the results shown in Table IX it can be seen that the nickel
electrode has suffered no apparent loss in capacity due to 16,500 Ah
of cycling, in fact, it may have even increased in capacity. The zinc
electrode on the other hand appears to be about 25~ below its initial
capacity and has some drop in voltage. Because of the large excess ot
zinc capacity originally available, sufficient zinc capacity should
exi st to enable the battery to deliver rated capacity.
19
-------
I\)
o
Table VIII.
Failure Analysis of Eagle Picher: Ni-Zn Batteries
Examined at
Final Batt.
Electrolyte
Separator
Zn Electrode
Nickel Electrode
Battery Ah output Capaci ty KOHl K2CX> ~I Appearance Appearance A'P'Dearance
BB-674 13,000 17 Ah 3 33 No zinc pene- Negligible shape Excellent condition.
#1 See Fig 9 tration. change. No blisters, cracks,
. Limited degra- See Figures l7a,b. or excessive expan-
dation of FSC sion.
at edges and See Figure 19.
above plates;
See Fig 16.
BB-674 13,000 17.3 Ah 10 23 As above Approximately ~ No cracks or bliste
lie (Shorted) See Fig 10 ~inc electrode Edge curling apposi
shape change. areas of zinc shape
change.
High 12,!joo l5.!~ Ah - - Zinc dendrites Approximately 10- As above
Rate (Shorted) in negative 15~ Zinc electrode
Battery pellon bag. ,shape change.
#1 FSC attacked a '
bottom, top an~
edges.
Cause of short
not apparent.,
High 26,400 12.0 Ah 6 20 Attacked at Less than l5~ No curling or
Rate (Shorted) edges and top. zinc electrode cracking.
Battery Poss ible short shape change. See Figures 20&, b
f2 at positive See Figures 1&, b
tab area.
re.
te
-------
Table IX.
Capacity Analysis of Nickel and Zinc Electrodes A:f'ter 16,500 Ah
Cycle 1 Cycle 2
Type Aho Vavg. Aho Vavg.
1 old Zn, 2 old Ni 5.1 1.55 5.0 1053
1 old Zn, 2 new Ni 5.1 1.l,8 4.2 1.47
1 new Zn, 2 old Ni 6.6 1.. 58 5.8 1.. 58
1 new Zn, 2 new Ni 6.0 1.. 56 5.3 1.56
NOTE:
Batteries charged at 1.0 A for 9 hours; discharged at 3.0 A to 1.0 V.
Electrode area = 23.4 in2; theoretical capacity 6.4 Ah,
BB-674 ( ) /U Battery #1
The failure analysis data given in Tables VIII and IX, as well as
Figures 16, 17, and 18, clearly indicate that life is limited by shorting
and separator degradation rather than zinc electrode shape change. Even
though the cellulosic separator is not completely oxidized when it is
between the electrodes , it is a major source of carbonate contamination
of the electrolyte. High carbonate contents in the electrolyte appear
to confer one benefit in that the amount of zinc shape change is reduced.
This occurs because zinc solubility is lower in carbonated solutions.
Zinc electrode shape change was found to be very low compared to the total
amount of energy delivered because of two other reasons. The first was
the Teflonated zinc electrode design and, second, the nature of the pulse
loading applied. .
The nickel electrode, as shown in Figures 19 and 20, suffered negli-
gible change during up to 26,450 Ah of cycling and mainta.ined full or
even increased capacity. Action must be taken to protect the tab area
(See Figure 20), particularly on thin plate cells to prevent high spots
'Which can accelerate shorting. In general, it can be stated that the
nickel electrode should present no problem in a. hybrid vehicle regime.
Battery Failure Analysis on Cells with Non-Sintered Positive Electrodes
Energy Research Corp. 4-Ah high rate cells were dissected a:f'ter 5,300
Ah of cycling. The design of these cells was given in Table IV. The
failure analysis indicated that the 1 turn of fibrous sausage casing used
as the separator was penetrated by zinc dendrites. The separator was
attacked at the top and sides. The nickel electrode was 11nQAmAged but had
swelled fran 0.0165" to 0.0265". This amount of expansion is considered
relatively normal for the non-sintered nickel hydroxide electrode. Subse-
quent to the bU;lding of these cells, Charkey (11) reported an improved
nickel hydroxide preparation process which resulted in less expansion of the
positive plate with cycling. The zinc electrodes in these cells had approxi-
mately 10% shape change. Based on these observations, it can be concluded
that a non-sintered nickel electrode can be used in a Ni-Zn battery for
hybrid vehicle service. One turn of fibrous casing, however, would not be
sufficient separation for the desired life.
21
-------
Cells with Inorganic Separators
Several 40 Ah sealed silver-zinc cells and zinc electrodes made
vi th Astropower inorganic separato:tB were made available for this program
by NASA-Levis. The silver-zinc cells had been activated for more than one
year and had also been heat sterilized. One battery was put on deep cycling
to determine its capacity maintenance characteristics. The regime con-
sisted of a charge at 2.0 A to 2.05 V followed by a 10 A discharge to
1.2 V. The results obtained are shown in Figure 21.
Because of the good results obtained for the first 25 cycles and
hence the apparent excellent shape change characteristics of the Astro-
power electrode and the stability of the separator, a 3 plate Ni-Zn cell
was constructed using the bagged Astropower zinc electrode and 2 cut
down nickel electrodes from a BB-674( )/U battery. The charge-discharge
voltages for this cell are shown in Figure 22. The discharge voltage
obtained was considerably lower than comparable standard cells. It had
previously been reported that the Astropower separator and electrode
might require several cycles and/or ster 1lization to be completely
wetted and therefore have a low internal resistance. One cell was built
with Astropower zinc electrodes and separator and subjected to 24 hours
sterilization at 110°C. Another cell was built using the Astropower zinc
electrode with a fibrous casing separator. Based on the discharge voltages
obtained fram the 3 cells described, it was determined that the sterili-
zation treatment had a negligible effect on voltage. The cell with the
casing separator had a slightly higher voltage than the other two cells,
but still less than normal. Reference electrode readings indicated that
the major portion of the voltage loss occurred in the Astropower zinc
electrode and was probably due to the addition of K.'l' (potassium titanate)
as an electrolyte absorber to the zinc. The 3 cells were then put on a
cycling regime consisting of 5 hours of charge at 1 A followed by a 1 A
discharge to 1.2 V / cell. The data obtained ate shown in Figure 23. Based
on the results, it would appear that the Astropower separator reduced
cycle life by 50~ compared with a sausage casing. Sterilization of the
Astropower separator and zinc electrode fUrther reduced cycle life.
After the Astropower 40 Ah Ag-Zn cell had completed 174 cycles and
had indicated short circuiting, cycling was terminated and the cell
dissected. The zinc electrode had been reduced to less than 4o~ of its
original area. The separator bags had cracked on the edges, permitting
the growth of zinc dendrites and therefore short circuiting. However,
the separator did not show penetration of either zinc or silver on areas
opposite the electrode faces.
The Astropower separator and zinc electrodes, while having good sta-
bility in shallow, depth of discharge cycling in silver-zinc cells, do
not appear to "have the low resistivity and stability required for hybrid
vehicle operation of nickel-zinc batteries.
22
-------
Sealed nickel-zinc cells containing inorganic separator material
built by Energy Research Corp. were made available from another program.
The inorganic separator consisted of metallic layer hydroxide inorganic
materials with a Teflon binder. Two different thickness and porosity
materials designated ERC100l and ERC2001 were prepared and built into cells
(4,11). When cells containing 2 layers of the thin ERC2001 seParator were
cycled,they shorted after approximately 40 deep cycles. Examination
revealed several small spots on the separator where zinc may have pene-
trated. It appears ,therefore, that 2 layers of the ERC2001 separator are
insufficiently hole-free to prevent shorting. In addition, if sealed
cells of this type are allowed to stand for even a day, in a partially
shorted state, they develop considerable hydrogen pressure. The ERC2001
material 1s, at present, inadequate by itself as the separator for hybrid
Ni-Zn batteries.
Recommended Battery Design
A proposed Ni-Zn battery for hybrid service in a 4,000 lb family type
vehicle,based on present technology, should have the following characteris-
tics:
Number of cells
120
Charge voltage (modified constant potential of
1. 835V/cell)
Open circuit voltage (after discharge)
220V
210V
10kW average discharge voltage
55kW average discharge voltage
207V
165V
Maximum charge acceptability
> 30kW (l30A)
Nominal capacity
Active area (in2/cell)
20-25Ah
>500
Cell Weight
2.4 lbs each
Estimated Battery weight
(Includes battery case & heat conductors)
350-400 lbs
Battery volume
2.7-3.0 ft3
$1, 500
Cost (estimate only)
Life exPected (present designs)
150,000 cycles
G
-------
CONCLUSIONS
High rate design conventional nickel-zinc batteries, capable of
delivering up to JJO W/lb, have demonstrated a capability of up to 26,000
Ah on the hybrid cycling regime. This amount of total capacity is equiv-
alent to approximately 150,000 cycles and is 3/8 of required life. Small
4 Ah cells have demonstrated a life equivalent to 1/2 of that required. Re-
liable operation above these levels cannot be attained without improved
separators.
Nickel-zinc batteries can operate' in hybrid service with energy effi-
ciencies of up to 8~, which is better than competing systems.
Separators containing cellulosic material, such as fibrous sausage
casing,are oxidized and convert the KOH electrolyte to carbonates under
the high temperature conditions existing on the hybrid simulation regime.
The oxidized cellulosic separator is then unable to prevent shorting.
Astropower type separators and zinc electrodes appear to offer no
advantages for hybrid vehicle operation of Ni-Zn batteries. The ERC
layer hydroxide inorganic separators tested have low electrical resis-
tance but do not prevent shorting by zinc dendrites when they are used
as the main separator.
The nickel electrode is basically unaffected by cycling even when
many thousands of ampere-hours are drawn from it.
The zinc electrode suffers considerably less shape change on the
hybrid regime than it would be with conventional cycling.
Modified constant potential charging, at potentials as low as 1.83V/
cell, adequately maintain cell capacity, while minimizing overcharge and
heat generation. This results in less frequent water additions. It is
estimated that for the average driver, water would only have to be added
every 6 months. A deep conditioning cycle would probably be required
at about the same frequency.
Insufficient data has been generated to fully assess the effect of
sealed operation. However, it does increase overcharge and temperature.
24
-------
REFERENCES:
1. Aerospace Corp., "Hybrid Heat Engine/Electrical Systems Study," Final
Report, Contract ro470l-70-C-0059 (EPA), July 1971.
2. J. R. Kettler, "Batteries for the Hybrid Heat Engine/Electric Vehicles,"
Proc. of the Symposium on Batteries for Traction and Propulsion, March
1972.
3. D. Mains, "Evaluation Program for Secondary Spacecra:f't Cells,!' Sixth
Annual Report, Contract Wl2,397 (NASA-Goddard), Naval Ammn'1,tion Depot,
Crane, March 1970.
4.
A. Cha.r~ey, "Sealed Rechargeable Nickel-Zinc Batteries," Final Report,
Contract DAAB 07-7l-C-0134 (ECCM), Energy Research Co~~ July 1972.
5. M. SUlkes, "Development of the Sealed Zinc-Silver Oxide, Secondary
Battery Systems, ":'Final Report, Contract DA-36-039 AMC-02238(E) (ECOM),
Yardney Electric CO?,July 1966.
6. J. GOodkin, J. McBreen and G. DalinJ "Long Life Stable Zinc Electrodes
for Alkaline Secondary Batteries," Final Report, Contract DAAB 07-67 -c-0185
(ECOM), Yardney Electric Co?~July 1969.
7. M. SUlkes, "Nickel-Zinc Secondary Batteries," Froc. 23rd Annual Power
Sources Conference, May 1969.
8. P. V. Popat, "Development of Lower Cost Nickel-Zinc Batteries," Final
Report, Contract DAAB 07-69-C-0172 (EC
-------
105
4
10
"-#"
'"
"""
~"
'I.
",
'"
",
'"
""
'"
"'"
",
"
",
""
"
"
'"
"
",
'(,"
'"
'"
'"
",
"#
""
""
'"
""
""
""
"
"
"
"
"
,
'. '..
'. ~
'., '..
.'. ~,
'. ,
'. ,
'. ,
'. ..
, . ~ ..
'., ~,
., ,
.'.
f2
H
~
o
103
1111111111 Ni-Cd
Ag-Cd
--- Ni-Zn
2 . - . _.' Ag-Zn
10
10
20
40
50
60
30
DEPrH OF DISCHARGE (;,)
?:'Z'J.~e L
CYCLE LIFE VERSUS DEPrH OF DISCHARGE FOR VARIOUS BATTERY SYSTEMS
26
-------
6000
~
u
~
cs
~
~
~
o
4000
~
--.......
fJ
ffi
>
H
tJ
A
~
~
2000
111111111
Sealed Ni-Cd
Sealed Ag-Cd
Ni-Zn (Estimate)
Ag-Zn
---
-.-
"
~,
~,
"
"
"
"
"
"
"
"
"
-:.
..,
-:.
-:.
."
."
~
-:.
-:.
."
."
.."
."
."
""
."
."
."
~,
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
#,
"
"
"
'"
'"
'"
'"
'"
'"
O",N1-zn Requirement
ifI'b:Cid Requirement
--
--
...-
-...
-...
-...
--...
-.-.- ......
.......-.-.---.- --..
. - . - . - . _.....
10
40
20
30
50
DEPl'H OF DISCHARGE (10)
Figur~ 2.'
TOTAL ENEmY OUTPUT AS A FUNCTION OF DEPl'H OF DISCHAroE
27
-------
8,0
7.8 .
,--
E;,
~.: 7.6 .
-...
~
~
t:)
~
....
~
7.4 ..
7.2 ~
7.0
J" . .4'~" ,"
, ~ I .
llOCJI, {I'~5-851> /1' 50-6CJ1,
.4,' I"';
, ,'"
Il~ {I'" {I,I'
l I" ",
,,1 ~/" ,I'
" I" ~,I
- ,,' "r.
~" I'
I' "
I" I'
,,' ""
. I'
I,I'
,,1'
,II'
"
.
-
-
Eagle-Pieher BB-674( )/U
Ni-Zn Battery (6.4V, 35AH)
-
I
.
100
.
.
.
.
20
40
60 80
CURRENT (AMPERE:»
120
Figure 3. CHARGE cummrr VERSUS CHAmE VOLTAGE AS A FUNCTION OF STATE-OF-CHAmE
FOR STANDARD EAGLE-PICHER Ni-Zn BATTERY
-------
. 7.8
,.....
tf)
E-<
~ 7.(, .
......
I\) ~
\0 ~
~
0
>
7.4 .
8.0
.
7.2
.
Figure 4.
,'.
"
"
..'
,
,
.
,
,
,
,
,
I
t
..
. looj
..
.
..
!
Eagle-Picher High Rate Battery
(6.4V, 24AH )
Battery Internal Temperature
approximately 120° F
.
.
40
"
.'
,. ~5J
.
, .'
,,' ,.'.
, ,
" ,~-65J
" ,,'
" "
, "
" ~,
. ,. , .....
"
"
"
"
,
-,' .
.
.
100
.
120
20
60 80
CURRmT ( AMPERES )
CHAroE CURRENT VERSUS CHAroE VOLTAGE AS A FUNCTION OF STATE-OF-CHARGE
FOR ADVANCED, HIGH RATE, EAGLE-PICHER Ni-Zn BATTERY
-------
6
-
tI)
H
...:I
o
W >
'-'
o
,~
C,!)
-<
H
...:I
o
>
Figure 5.
7
2.75 kW (306 W/1b)
5
" ...
.', ..
~ ;"" ,
..", ..
... " ~
, ", ~
~"~
'~~
.~
~'
"
"
"
.""
" '::""
BB-674 ~ . '"
(100%) ~, '."'"
.. '.'"
~ ~"
.. . '"
~ .. ",
, .... '"
.. "
~ "
, ' ~"'"
, ' ";'"
, ' '::'"
, ' '::'"
.. , ':""
.. ... "
, . :t"
.. ... "
~ . ':",
, ... "
.. . ':"""
~ .... "
, ~
~1. 6 kW
'190 W/1b)
2.45 kW (270 W/1b)
Eag1e-Picher Battery #2H
Battery Internal Temp ~ 120°,
Minimum Voltage for 10.8 Second
Pulse
Battery Charged at 7.4 V for 49.2
Seconds Before Next Discharge Pulse
BB-674( )/U Tested As Above
200
400
600
CURRENT (AMPERES)
VOLTAGE-CURRENT CHARACTERISTICS OF HIGH RATE Ni-Zn BA1'l'ERIES AS A FUNCTION OF STATE-OF-CHAmE
-------
8.6
1.0
-
i::
C 6.0
:;..
'--
w
......
r:.:I
t:)
c(
~
o
>
5.0
4.0
It
Initial Performance
1111111111111
High Rate Hickel-Zinc
Battery #lH
'.
..
Eagle-Picher BB-614( )/U
Nickel-Zinc Battery
-.-.-
.
~
~
I~
~ "
I ~
",
Po "
"
"
"
'..
"
~I
"
""
"
II"
"''''11
1111111111111111111111111111111111111111111111
11'11'11"11
Temperature: 80°F
Battery Fully-Charged
.
I-
~
.',
.
,
.
""
..
'.
'-8
.,
. ...... .
...... ..
-.-.
-.-.'a-.
---------,.-
. -
.
3.0 . , I . . .
10 20 30 40 50 60
TIME (S~ONOO)
Figure 6. DISCHAmE VOLTAGE FOR 350 AMPERE LOAD
-------
TEMP ON BATTERY
EXTERIOR ~ 90
WALL 0
~
~
BATTERY
CHARGED AT
. rz:I
7.4 and 7.6 V ~ 10
MAX CURRENT ~
120 A u
p:::
rz:I
~ 5
~
'"
"
DISCHARGE
VOLTAGES
(MINIMUM)
-
~ 7
6
>
'-'
rz:I 6
~
-------
TEMPERATURE 110
~
ON BATTERY 0
EXTERIOR ~ 90
WALL ~
70
15
BATTERY ~
c.:> 10
CHARGED AT g
7.4 VOLTS
MAX CURRENT ~ 5
120 A ~
>
o
w
w ~
0
7
DISCHARGE
VOLTAGES 6
(MINIMUM) '"'
tI)
~
6
. - V @ 55 A > 5
-
D
~
-V @ 350 A t!)
-< 4
D ~
6
>
3
.....
~ ...................--
--'--
--
1 - Water Added
2 - Deep Cycle Run
3 - Cell Shorted. Removed. Voltages Recalculated
11.
L
2000
4000
6000
8000
L
Cell Shorted
After Deep
Cycle
--
10.000
12.000
14.000
TOTAL AMPERE-HOURS DELIVERED
Figure 8.
CYCLIl«} HISTOORAM R>R STANDARD, EMLE-PICHER Ni-Zn BATTERY BB-674( )/U (#2)
-------
w
~
---
en
~
S 6.0
>
~
~
o
<
~
....:I
o
>
7..0
ID = 14 A
5.0
"
'I>.. .-.. 0
-~'" T = 80 F
'" ,
" """ '....
, "'" ....-...
.... - '" ---...
'-'1-. ---...
~~-'-7~ .
""" .~ -. -. -. lilt
'" , ' ....
I, -... ...
'" ---... ,
'" -- -... ..
'" -- -.. ~.
'" -- .-
""" ',\
"" ~ ~
'" 4IWl
"'" - .,. ,
""" " ,
'" ., \
".. \
'10 I \
\. \
~I
~'t.. \
~ \~
After 13,000 Ah and
Recharge at 9 A
.-----
After 4,500 Ah
11111111111.1111111
After 6,500 Ah and Recharge
at 9 A
- . - . - 1 - Af ter 6,540 Ah and Discharge
To 0 Volts Recharged at 9 A
10
20
30
Figure 9.
CAPACITY (AMPERE-HOURS)
DEEP DISCHARJE CURVES FOR STANDARD, EAGLE-PICHER Ni-Zn ~ #1
AF.rER PULSE CYCLING
-------
,..."
Cf)
E-<
....:I
0
W ::>
V1 ........
4J
c..')
<
E-<
....:I 6.0
o
::>
7.0
5.0
~
, ~,.
, ....
'.... ---
~-.-. ---------------
-.... ------.
.4I1U \'lll II.. II '--'~Ini t ia1
, ","",",". '
# II....... ,
,~ ~
# ,~ ~
. " """ .~ 3.900 Ah
.. " ~ ~
BB-674( )/U 112 ~ ", \
~ "
14 A to 5.0 V \"rA -13,000 Ah ""''-- 9 ,200 Ah \
5 Hours \ "" \
.. ", ,
, ~ ,
.. "
, ~ ,
"
.. ~,
"
, #.,
..
I ,
.
I =
D
I = 9 A for
c
T = BOoF
10
40
20
30
Figure 10.
CAPACITY (AMPERE-HOURS)
EFFECT OF CYCLING ON DEEP DISCHAmE CAPACITY AND VOLTAGE OF STANDARD EAGLE~PICHER
N1-Zn JlA1W.rERY #2
-------
TEMPERATURE
ON BATTERY /
EXTERIOR WALL ----
~
70
15 /,
BATTEN.Y
CHARGED AT ~
t.:) U, to
7.4 VOLTS ~
MAX. CURRENT ~ 10 50% O.C.
u
120 A .~
~
:>
0 5
N
VJ '-
0'\ ----
0 /
7
DISCHARGE ~- ............... - \
........ -- -
VOLTAGE en l '
~ 6
(MINIMUM) ...:I 3
0 1 2
:>
'-'
~ 5
t.:) All Cells
< -
~ Shorted
...:I
~ 4
3
VD @ 350 A
- - - V @ 55 A
D
1 ~ Dee~ Cycle Run
2 - 30 CC, 20% KOH Added
3 - Deep Cycle And Reversal Run
2000
4000
6000
8000
10,000
12,000
14,000
TOTAL AMPERE-HOURS DELIVERED
Figure li.
CYCLING HISTOGRAM FOR ADVANCED, HIGH RATE, EAGLE-PICHER BA1'l'ERY 11H
-------
.-..
en
E-<
....:I
o
>
'-"
~
~
\JJ -<
-.J E-<
....:I
o
>
5
#,
"#
, "'....
, "'" . ...
" -~.....
, ....~ -..-.
,~ ~~ -'-1_1_'-
~ ~~ '....
',~-~- ''''''''''
- " . ~~- ''''''(1
Eagle-Picher 4 Cell ..., ~
Monoblock 111 H '...,
"
"
"
"
,
6
4
10
20
"
,
-
"-
-
"-
"-
,
30
40
TIME (SECONDS)
At 500 Ah Output
.1"1. At 4,300 Ah Output
~~/~ At 9,800 Ah Output
--- At 12,300 Ah Output
.,
50
60
Figure 12. POWER TEST PERFORMANCE AS A FUNCTION OF LIFE FOR AINANCED, HIGH RATE,
EAGLE-PICHER BATTERY #1H
-------
TEMPERATURE 110
ON BATTERY ~
0
EXTERIOR ~90
WALL
Eo-<
70
15
BATTERY ~ Charge At
CHARGED AT 0
7.2 & 7.3 ~10
VOLTS u 7.2 V . .. 7.3 V
IJG
MAX CURRENT ~
:>
120 A 0 5 ~
N
~ 0 --
DISCHARGE
VOLTAGE
(MINIMUM)
. - . VD @ 55 A ~ 5
Eo-<
...:I
- VD @ 350 A ~ 4
7
-
U)
~ 6
o
:>
'-'
----.-.
1
2 * 1 2 1
~~-
1 - Deep Cycle R,n
2 - 20 cc H 0 Added
3 - 20 cc KOH Added
3
Figure 13.
4000
8000
- .................. - - '- -
1
T ",\!--
1 3 2< 2 3
R~
-'-..
4 - Deep Cycle and
Run
5 - 2 Cells Shorted (Max Current 200 A) Data Recalculated
* Individual Pulse Voltages Begin to Drop as Low as 3.0 V
12,000
16,000
20,000
2~, 000
2B.000
TOTAL AMPERE-HOURS DELIVERED
CYCLING HISTOORAM FOR ADVANCED, HIGH RATE, EAGLE-PICBER :BA181'ERY #2H
-------
""".. ~ -~ ---~ ~-
-.--
w
\D
-;;;- 5
E-o
S
:>
'-"
~
~
<
E-o
S 4
:>
~
~
E-o
E-o
<
!:Q
6
.,
",
-.--
-~# ---
.... - -
~, ---
'# -___~B
, ---
# . -.
~
#
\
.-
\
#
\
.-
\.
.-
,
.
,
.
,
.
,
-
3
10
20
A
30
40
ID = 350 A
Curve A at 8,530 Ah
Curve B at 8,550 Ah
50
60
TIME (SECONDS)
Figure 14. POWER TEST PERFORMANCE FOR ADVANCED, HIGH RATE,
EAGLE-PICHER BATTERY #2H
-------
~
o
-
U)
H
~
o
~ 7
~
c
<
E-<
~
o
>
6
8
,
# "
'# "
. '. "
'. "
. '., ,~
., ~~
., ~~...
., --...
. .........
, . ... ... ...
'. -----
'. -------
'. --------
'.
'.
'.
'.
'.
'.
'#
'#
"
'.
"
'.
"
.
I = 100 A
o
o
Temp = 80 F
- - -. Test Run at 65 Ah Output
Test Run at 420 Ah Output
I. - . - . I Test Run at 900 Ah Output
10
20
30
40
50
TIME (SECONDS)
Figure 15.
POWER TEST PERFORMANCE OF nfEOOY RFSEARCH CORPORA'l'ION, 5 CELL, 4 Ah, N1-Zn BA.~
-------
.J:'"
I-'
---
--
. .
. .
! "
"t"
" "\'.& .
. ,
'. . .
J~
"'f,
.~
i",-
I
,'w .,
. Q
"
,J# ..
, , 'II"
..
,
: .~
.J
~ '~
~..
,,-1".,
,I
.. '.Ii'
'. . . 1
,
.
", '.'
L~~~o: ~ Negative EJ.ectro~~~ r;'~rous- saus~e Casing (FSC)
o~~- "iIP.. -
--'~-... - II!I .....
~
\
'; Figure 16.
,.
-I
I
I
, ~e11on 'From Positive EJ.ectrodel
- ..-- -=- - .: -- =- - r:..=- -.. II
SEPARATOR REMOVED FROM STANDARD EAGLE-PICHER BATI'ERY #1 AFTER 13,000 Ah
-------
::
-,':1' .
f'f"I"'-
IJ
,
.:L
/ '" b
c-
O...',
.~
.
1<
rl-
I r.03Z
I
i
;
j
!
I
-- ..-------.---- 5'/
-..-.---- ----- ----..-------.-..--.---,
.03~
.03~
. OJ'::
.0+0
.032.
.035
. 03..,
THICKNESS DISTRIBUTION
.032
.0:'7
.033
.C3~
Figure ~ 7. ZINC ELECTRODE REMOVED FROM STANDARD EN}LE-PICHER
BATl'ERY #~ AFTER ~3, 000 Ah
42
Original
Thickness
.035 "
-------
Figure 18.
ZINC ~TRODE REMOVED FROM ADVANCED, HIGH RATE
EAGLE-PICHER BATrERY #2H AFrER 26,400 Ah
43
-------
I-
~
.r::-
.r::-
rr.
'.
Figure 19. NICKEL ELECTRODE REMOVED FROM STANDARD EAGLE-PICHER BATTERY #1 AFTER 13,000 Ah
-------
r---- -
I
I
.
~~.
~--, - .~~ ~... ~~
,. : -..,../ or,
"
:~ :i;"
.,
Closeup
.. . ~Of Tab
. ~ krea
.;'.tI
t4
'. I'
I
"1'
~
Figure 20. NICKEL ELECTRODE REMOVED FROM ADVANCED, HIGH RATE,
EAGLE-PICKER BATrERY #2H AFl'ER 26,400 Ah
45
-------
45
40
~
,.....
~ 35
~
o
::c
I
~
~
~
'-" 30
><
f-<
......
U
<
p...
<
u
25
20
IC = 2.0 A
ID = 10 A
T = 80°F
~ 2.05 V
a 1.2 V
50
;,
* Shorted
100
150
200
CYCLES
Figure 21.
CAPACITY MAINTENANCE OF ASTROPOWER SEALED 40 Ah SILVER ZINC CELffi
-------
,.....
U)
E-<
...:I
o
:>
'-'
~
~
~
t:) 1.
<
E-<
~
o
:>
2.
1.
.'
--.--
..-....-
..........
_1-'.'
"...'
"
,..
1.
1.
1
3
4
2
CAPACITY (AMPERE-HOURS)
-.-
Charge IC = 1.0 A
Discharge ID = 3.0 A
2 ea .038 Ni Electrodes
1 ea Zinc Electrode
Electrode Area = 21.5 in2
T = 800F
Cycle 113
5
6
Figure 22.
CHARGE-DISCHARGE CHARACTERISTICS OF A Ni-Zn CELL
WITH ASTROPOWER SEPARATOR AND zmc ELECTRODE
-------
'"' 6
en
~
=:>
0
::c
I
r:.:I
~
r:.:I
p..
~ 3
'-'
I:&; >c
E-<
H
U
<
p..
<
u
!
5
1
~
,
/#
.,
,
, ,
\ '
.,
\ '
\ "
\ ,
\ "
\ '
\ ~,
\ ~#
\ "
\ "
,
\ " .
\ #~#~
\ .~.~
\ .~.~
\ .~.
~
I = 1 A
c
Id = 1 A
T = 800F
.5 Hours
,.1.2 V
20
40
60
80
CYCLES
Astropower Zinc Electrode With
Fibrous Sausage Casing Separator
..... Astropower Separator and Zinc
... "IAstropower Separator and Zinc
Sterilized at 1100C
100
120
140
Figure 23.
CAPACITY MAINTENANCE OF Ni-Zn CELLS WITH ASTROPOWER SEPARATOR AND ZINC ELECTRODES
. .
-------
TECHNICAL REPORT DATA
(Please read IRUT.uctions on the reve1"$e before completing)
1. REPORT NO. 12. 3. RECIPIENT'S ACCESSIOf\llNO.
EPA-460/3-74-025
4. TITLE AND SUBTITLE 5. REPORT DATE
Nickel-Zinc Batteries for Hybrid Vehicle Operation December lQ72
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) 8. PERFORMING ORGANIZATION REPORT NO.
Martin J. Sulkes
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
U.S. Electronics Technology and Devices Laboratory
U.S. Army Electronics Command 11. CONTRACT/GRANT NO.
Fort Monmouth, New Jersey
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
U.S. Environmental Protection Agency
Office of Air and Waste Management 14. SPONSORING AGENCY CODE
Office of Mobile Source Air Pollution Control
Ann Arbor, Michigan 48105
15. SUPPLEMENTARY NOTES
Un-numbered interagency agreement
16. ABSTRACT
Nickel-zinc batteries have been evaluated against the requirements of a hybrid electric
vehicle regime. This regime consists of many short duration pulses at high rates of
discharge with rapid recharge. High rate designed conventional Ni-Zn batteries of a
nominal 24Ah capacity yielded up to 26,000 Ah total output (equivalent to 150,000
cycles) before failure; this compares to a desired output of 73,000 Ah or 400,000
cycles. The principal failure mechanisms were shorting and loss of voltage due to
high carbonate concentrations caused by degradation of the cellulosic separator. Zinc
shape change (10-20%) was surprisingly low in view of the large number of cycles. The
nickel electrodes were essentially unaffected by the cycling. Rapid and complete
recharge was accomplished with a modified constant potential of 1.83v/cell. The use
of this voltage resulted in low heat generation, approximately 7-8Btu/cell/h and a low
percentage of overcharge. Overall energy efficiency on the hybrid vehicle test regime c
is as high as 89%. Two types of inorganic separator material were evaluated, and were
found to be unsuitable. A full size battery, to power a hybrid electric vehicle,
would consist of 120 cells. The battery would weigh 350-400 lbs; the charging voltage
would be 220V. The average discharge voltage would be 207V at a 10kW load and l65V
at a 55kW load.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS b.IDENTIFIERS/OPEN ENDED TERMS C. COSA T I Field/Group
18. DISTRIBUTION STATEMENT 19. SECURITY CLASS (This Report) 21. NO. OF PAGES
Unclassified 57
Unlimited 20. SECURITY CLASS (This page) 22. PRICE
Unclassified
EPA Form 2220-1 (9-73)
49
------- |