Characterization of Evertroll Nickel-Zinc

Batteries
F. P. T r e d e a u a n d Z. M. Salameh, Senior Member. IEEE

practical solution for providing on-board electric power for an

Ahsrracl- This paper describes the performance of Nickel independent vehicle is the electro-chemical battery.
Zinc batteries and their suitability for use in electric vehicles. Batteries have been used in electric cars for over 125 years.
The goals of the evaluation were to find: the ratings of the pre-dating the internal combustion engine (although, notably.
battery at different temperatures (-20, 0, +20 and +4OoC):
not pre-dating external combustion engines, i.e., steam) 121.
determine the suitability for using this battery in an electric
vehicle: a n d compare the results of different battery discharp With the rapid improvements in power electronics (both
algorithms, i.e., Constant Current. Constant Power and Realistic power handlers and control systems), magnetic materials and
Load (described herein). 1-cstinp was done at the Battery electric motor topologies. the suitable battery is the last big
Evaluation Lab at URIL. which is described herein. The battery piece (outside of the support infrastructure) needed to
parameters measured were terminal voltage, rurrent. case successfully market an all-electric vehicle in direct
temperature and a m b i e n t air temperature. Results o f the
competition with IC automobiles.
experiments are as follows: the manufacturer's capacity ratings
were confirmed (i.e., 85 A-h): tlie battery is found to be suitable The U.S. Advanced Battery Consortium (USABC) hac
for use in EV's with superior specific energy: and the battery has been funding battery research since the early 1990's and has
higher energy capacity at higher temperatures while the set the following goals [3]:
charging efficiency is higher at lower temperatures.
TABLE I
Itidex 7rrms--Batteries, Road vehicle electric propulsion. USABC MID A N D LONGT E R M BATTERY PERFORMANCE GOALS
Nickel-Zinc (NiZn) Batteries
USABC SPECIFIC
COS1
GOALS POWER
1 . 1NTRODUCTlON
$150 or
Mid- 150-200 80- 100 w-h
S ome of the most advanced and successful vehicles in the
world are all-electric high-speed passenger trains in
Western Europe and Japan. The Bullet Trains of Japan
term
goals
warts per
kilogram
Pel
kilogram
5 years
less per
kilowatt
hour
regularly transport large numbers of passengers between the In a a driving
major cities of Honshu at speeds exceeding 300 Km/h in 0-50
typical range of battery cost a $4500-
clean, quiet vehicles that requires relatively easy maintenance acceleration
EV this 125 miles amortized 6000
in 12
with a nearly flawless safety record. The Bullet Train has would in normal over 5 years batten
mean ... seconds
been in regular service since I964 and is favorably thought of usage
by the traveling public [I]. $100 or
Lonp 200 wan
High speed rail systems like this, along with practically all 400 watt5 less per
term hours per I O years
ki lowat1
the subway and trolleys in the world share the attribute of per kilogram
goals kilogram
operating on electric power. Electric power has proven itself hour
over more than 100 years of mass transit service to be the In a a drivinp
0-60 battery cost
optimal solution for many key measures of performance. typical range of
acceleration amonized a $4000
It is widely acknowledged that adoption of electric EV this 200 miles
in nine over batten,
would in normal
vehicles (EV) for private use in free-roaming passenger second5 years
mean ... usage
vehicles and trucks would greatly help to alleviate many
pollution and resource problems. In all the cases referred to
above, the electric power is delivered by wires or othei The implementation dates for these goals have been
c,onductors or through electro-magnetic induction from an off- modified several times. A comparison to other contendinp
vehicle source. The problem with the private, free-roaming baner?, chemistries is found in the following table [4]:
vehicle is that the source of electric power must be carried as
part of the vehicle. Although there are several proposed
techniques for providing electric power. the most developed

165 0-7803-7863-6103/$17.00 02003 IEEL

 TABLE I I
CHEMISTRIES FOR ELECTRIC
LEADING VEHICLE BATTERIES

In this report. we will evaluate two samples of the Evenroll electrode. Both reactions ate mediated by the alkalirie
battery from Evercel Corp. and compare its characteristics to electrolyte in the separator (potassium hydroxide). At the
nickel-cadmium. lead-acid and nickel-metal hydride. W e will positive electrode the reaction is:
also evaluate NiZn batrery technology in light of the medium-
and long-term goals of the USABC 2NiOOH + 2H20 + 2e- ts 2Ni(OH), +20H
11. RATINGS AND CONSTRUCTIOW
The reaction at the negative electrode is:
Although the NiZn battery has been known to have many
attractive qualities for nearly 100 years IS], it has suffered
from the drawback of having shon cycle life. which was due
t o the solubility of the zinc electrode in the alkaline electrolyte
The above reactions proceed 10 the right during discharge
151, 163. The development of the plastic bonded calcium-
and proceed to the left during charge. The overall reaction is:
zincate (Ca(Zn(0H)J2) electrode by Evercel Corporation has
greatly increased the integrity of the zinc cathode and made
the prqjected cycle life of the NiZn battery competitive or 2 N i O O H + 2H,O + Zn H 2Ni(OH)? + Zn(OH), (3)
superior to contending, currently available battery
technologies. Again: the reaction proceeds to the right during discharge
One of the more compelling qualities of NiZn technology and t o the left during charge. These processes result in a
is the character of the basic chemistry. While other competing terminal voltage with a maximum galvanic potential of 1.’73
battery technologies employ mercury, lead, cadmium, etc., the volts. The Evertroll battery has 7 cells with a nominal
nickel-zinc battery metals are relatively benign. It is not terminal voltage of 12 volts.
wholly non-polluting, however, in that both zinc and nickel B. Batrety Evaluation Lab a1 UML
(and their compounds) ate considered hazardous in the EU
The investigation into the characteristics of the Evertroll
(71. NiZn batteries from Evercel was done at the Batiery
The Evertroll battery is nominally rated at 12 volts and 85
Evaluation Lab at UMass, Lowell (UML). The Battery
A-h. A power rating could not be ascertained. The Evertroll
Evaluation Lab presently has three, independent, custom built
battery measures 19.5 cm high, 17.5 cm deep and 31.0 cm
b a n e n exerciseridata recording systems. The exercisers have
wide for a volume of 10.6 liters. The battery has a mass of 14
different voltage and current capabilities and, in various
kf . combinations; are appropriate for large or small, 6 or 12 volt
The battery is comprised of 7 prismatic cells. Each cell has
batteries. Even single cell batteries can be exercised. Current
a self-resealing pressure relief valve but is otherwise sealed.
ratings range from lniA to 320 amps: charging or discharging.
Its nominal rating of 85 A-h gives it a nominal W-h rating of
The data acquisition system used was built and developed
12 x 85 = 1.020 kWh. This corresponds to a rated energy
by U M L staff and students 181. Figure 1 shows a block
density of 96.2 watt-hourSAiter and a specific energy of 72.9
diagram of the data acquisition system.
watt-hourslkg .
These ratings will be assessed and the energy capacity will
be established.
A . Chemistn.
The basic chemistry is the reaction o f nickel and water at
the positive electrode and zinc and water at the negative
 reflect a broad range of temperatures that an electric vehicle
would be expected to experience in real service.
C
.... * .
C
...... ......
C
......
0
A . Constant Curre nt Discharge Cwle Tests

.
IC :Ib i .i
0 . .
During the constant-current discharge phase. the batter!
. .
!Jp tu t
energy is drained at a constant 20 amps until the terminal
Emenes UlOC voltage reaches approximately 10.4 volts. Most of the stored
-
les: ..
._.. . -. .-.-
... . . -. energy is extracted from the battery during the constant
current phase
A typical Constant Current characterization run is shown
in Fig. 3. Battery # I is subjected to five charge-discharge
cycles at an ambient of 0.0"C
CO"Eta"! C",,.n! cycle'
6an.ryai @ooc.a~u~

ldultiplexer Unit

Ill
AID & D/A
Section

Fig 3 Constant Current Discharge Cycle 1est for Batten. # 1 at 0 0°C

Fig 1 Block Diagram of the Battery Exerciser and Data Acquisition System
at UMass. Lowell
In Fig. 3. the flat tops of the current trace corresponding to
I l l . BATTERY TESTS
CYCLE the constant current discharge during which the terminal
voltage drops to approximately 10.4 volts. That is followed
by a segment of constant voltage during which the current
falls to 2.0 amps. The charging phase then conimences
starting with a constant charging current of 20 amps followed
by the taper portion of the charge where the current falls
toward zero as the terminal voltage approaches 14 volts. Case
temperature [rofiles for the constant current cycle tests were
recorded. Below is a sample graph of temperature and current
for the constant current load test. The data is for battery #O at
+20"C.

Fig 2 The tvenroll Nickel Zinc Battery from Evercel

The batteries were subjected to three types of

characterization tests.
1 ) Constant Current discharge cycle tests z* U I ? Lm WT) u%
t. q v 7 %<Y

2) Constant Power discharge cycle tests lmr nrron,

3) Realistic Load discharge tests Fig 4 Current and 1emperature vs I ime for the Constant Current Cvclc
7 ests
Each cycle was repeated four times: once each at -20°C.
0°C. 20°C and 40°C. These temperatures were chosen to

167
 Note the sharp rise in temperature during the constant
current segment of the discharge and that the temperature
actually begins to fall during the constant voltage portion of
the discharge phase (in this case, most of the constant voltage
discharge is done at relatively low power). The temperature
falls at an even higher rate during the constant current charge:
indicating that charging is endothermic. This indicates that
the battery may need t o be cooled while under h e a w load in
an EV during very hot weather but that cooling will probably
riot be required in most cases.
B. Cotisrunt Power Cycle Tests
The difference between the Constant Current and Constant ..>
L mo' 003% soah IOLJL "0. :a"*,.
Power tests is, of course: the nature of the discharge profile. 4Wh

lim. (*.rmn*
?M#tA li.:u

In the constant power tests: the discharge regimen was to load

Fig. 6 lemperature and Current vs Time for Battery # I at + 4 0 T lor tht.
the battery at a constant 200 watts until the terminal voltage Constant Power Cvcle 1 esi
reached 10 volts. This was done to determine if a constant
power cycle test would be a more realistic determinant of the Figure 6 shows exothermic characteristic during discharge
battery's useful capacih in a constant power load like an and strong endothermic characteristic during charge. Also.
electric vehicle. A graph of the results of a typical cycle test note that at +4OoC. almost all the charging is done during the
can be seen in Figure 2-5. The graph shows the voltage and constant current phase of the charge and very little is done
current for battery # 1 at O.O°C. Five charge and discharge during the taper portion. The charging current falls quickly at
cycles were performed. the end indicating that the battery is fully charged.

Constant Power Cycle lesi

C. Realisric Load Tesrs
Banwy I1 @ 0 0 Cclsius In the realistic load tests, the batteries were subjected to a
cycle lest that involves discharging the batteries using data
gathered by U M L staff and students as described in 191. The
P
5 rest of the cycle test was the same as for the Constant Current
and Constant Power tests
During the realistic portion of the discharge cycle, the
battery is sub-jected to occasional discharge current as high as
120 amps (which was observed during heavy acceleration)
and recharge current as high as 1 10 amps (which was
observed during heavy regenerative breaking).
Realistic Dlschargc Characterization
Bancry D O at +mCelsius

7Ea mx Byn 93m

rime (Ssccndr)
- tmax 1"

I'ip 5 Voltape and Current vs Time for Batien, # 1 at 0 0°C

In Figure 5 . note that the discharpe current is constantl?

rising as the terminal voltage is dropping. thereby achieving
constant power.
Case temperature profiles for the constant power load test
were recorded. Below IS a sample of the temperature and
current vs time for the constant power load test. The data is
for batiery # I at +40°C. ,M lDI, 110 33% (XI ,X' 4%' I35 l.m
Time fMinutes)

Fig 7 Voltape and Current vs Time for the Realistic Load 1 est at 20 "C

Case temperature profiles for the realistic load discharge tests

were recorded. Below is a sample of a graph of case
temperature and battery current vs time for battery # I at 0.0"C
during the Realistic Load Test:
 Current VS lcmpcrature
Battetyll @O.O Celrlus D. Barrety Cvcle Tesl R e s u h
Realistic Loao
I From Table 111 below, we can see that, like most batteries,
the Evertroll NiZn has higher energy availability when it is at
or slightly above room temperature. The difference observed
between the CC. C W and RL tests indicate that under more
realistic conditions. the banery will deliver full power longer
100 when the battery case temperature is above +20°C. Better
charging efficiency can be achieved by keeping the battery
50
case temperature between 0.0 and +20°C. Active cooling
during discharge would only be needed if the battery were in
an ambient above 40°C and then only if the discharge rate
-50 were more than 20 amps (average) for a long period. At lower
ambient temperatures, the slight self-heatinp of the b a n e p
i 1ooc 1001 3000 ,000 5000 601" ,000 moll 90oc IOOOI during discharge would have a beneficial effect on b a t t e p
T". (Secondri
performance.
Fig 8 Current and 1 einperature vs Time for the Realistic Load 1 est

From Fipure 8. it can be seen that the temperature rose less

than 5°C over the course of the RL test. This indicates that
when the battery is subjected to the moderate average load of
a typical EV. extra battery cooling will probably not be
necessac .

TABLE 111
EVERTROLL NICKELZ W C BATTERY
CAPACITY AND EFFICIENCY
U N D E R DIFFERENT
DISCHARGE
ALGORJTHMS AN AT VARIOUS 1-ELIPERATURES

Constant Constant Realistic

Current Power Load

1- -20°C
0.0"C
+20°C
57.7
67.9
87.7
46.9
67.9
80.0
70.2
74.3
80.6
+4OoC 92.7 92.4 87.9
-20°C 659.9 522.3 777.1
0.0"C 782.5 632.0 860.0
+20"C 1 006. I 909.8 935.5
+40°C 1052.0 1049.4 1003.6
-20°C 0.90 0.95 1.04

1- 0.0"C

+20"C
+40"C
1 .oo

0.96

0.95
I .oo
0.96

0.97
0.98

0.96

0.93
-20°C 0.77 0.78 0.85
0.0"C 0.86 0.80 0.86
+20"C 0.84 0.83 0.86
+40"C 0.83 0.86 0.82

169
 1V. COMPARISON
TO USABC GUIDELINES

W e can fill in some of the requirenients in the USABC Guidelines using the findings above:

TABLE I\'
USABC GUIDELINES
FOR AVAILABLEA N D PENDING BATTERY CHEMISTRIE:

Market Range (miles Life Cycle life Cost Specific energy Energy density
availability per charge) (years) ($ I kWh) (watt-hours (watt-hours
(years) per kg) per L)
Midterm goal -- -- 5 600 -=$ 1 so 80- 100 135
Long-term goal _- -- 10 1000 4 1 00 200 300
Nickel-Zinc Available TBD 5-10' 600-1000' $385" 75 99
Nickel-cadmium Available 65-140 8 700-1200 $300- 45-55 100-1so
$500
Lead-acid Availablc SO-70 2-3 300-400 $ 1 SO- 3.5-40 100-130
$200
Available 100-120 8 700-1200 $300- Hiyh 60s-70s 1 SO-200
$700
1-4 1 SO-300 5-10 400-1200 $150- 100-150 100-200
I ithium-nolvmer $220

' Derived from expected cycle life and USABC goals above
From [SI
' From Cabela's on-line catalog (\\'\\U .cabela.com) as of 2/15/2003

The NiZn battery meets the USABC mid-term goals for observed during the + 4 0 T tests indicate that the battery will
cycle life and life expectancy and nearly meets the mid-term not need cooling during discharge. The NiZn is endothermic
goal for specific energy. It should be noted that the price during charging. which will make quick charging easier and
listed in Table IV is less than half the list price of only a feu can be used t o detect overcharging when the battery becomes
years ago. The price is expected to continue t o decline for at exothermic.
least a few niore years. Also. since the specific energy The NiZn battery shows some temperature dependency on
measured above is only about 22% of the theoretical limit. both charging and discharging. In particular. the battery has a
there is an expectation of large improvements in the future [SI depressed W-h capacity below 0°C. This will make it
may be used as a template for preparing your technical
necessary for the batteries to be heated to extract maximuin
work. When you open the file. select "Pape Layout" from the
power and range when used in an EV operating in cold
"View" nienu (View I Page Layout). which allows you to see
the footnotes. You may then type over sections of the weather. The battery also has depressed terminal voltage
document. cut and paste into it (Edit I Paste Special 1 when the case temperature is above 40°C.
Unforniatted Text). andlor use markup styles.
VI. RECOMMENDATIONS
V. CONCLUSIONS 1) The characterization tests should be performed on several
The characterization tests show that the Evertroll NiZn additional Evertroll batteries to build a larger statistical
batiery has specific energy that nearly meets the mid-term database.
goals of the USABC battery development specification. The 2) Cycle life for these batteries (or other. comparable NiZn
battery has good A-h and W-h capacity and efficiency: a. baneries on the market) should be established by multiple
good as NiMH and better than Lead-Acid and Ni-Cad. I n charge/discharge cycles (hundreds of cycles) t o confirm
particular. compared t o the NiMH banery, the NiZn bane? the projected useful battery life. This should be done in
has same energy density. better specific energy (implying con-junction with # S below.
longer driving range on a single charge). less toxic chemistry 3) NiZn batteries should be installed in an electric vehicle(s).
and greatly simplified cooling. With all these improvement. and an appropriate charger should be used. t o obtain real-
at less than 15% of the cost of NiMH. the NiZn battery is life performance data.
clearly the superior choice of the currently available battery 4) In the electric vehicle tests. a battery compartment thermal
technologies for use in electric vehicles. management system should be developed to investigate
The NiZn generates some heat during discharge, though ways of optimizing the temperature/perforniance of N i b
much less than NiMH 119. 201 The case temperatures batteries [lo].

170
 5) Cell-to-cell imbalance and banery-to-banery imbalance
should be monitored and a means for keeping the batteries,
or individual cells, in balance should be developed and it5 VI I I. BI O G M PI41E!:
efficacy should be evaluated. Prof Salamrh recieved his Diploma from Moscow Power Engineering
6) Means for rapid charging of NiZn batteries should be Institute in 1974 and his M.S and Ph.D from Universih o f Michigan. Ann
Arbor. in 1980 and 1982 respectively l i e is currently a Professor and
investigated, developed and evaluated. Dcpanment head at University o f Massachusetts Lowell He is also Director
'7) Computer battery models, such as circuit. fuzzy. neural or of Center for Electric Cars and E n e r p Conversion H i s areas o f inierest are
neuro-fuzzy should be developed. Electric Vehicles and renewable energy sources He has authored o r co-
authored over 80 reserch papers
VII. REFERENCES F r a n k l r c d c a u was born in 1954 in Framinpham. Massachusetts He
1I] Bvrii~B w n Shinhamen. An industry supponed web site for information received his A.A from Cape Cod Community College in 1976 and his
ahout N R T s Bullet Irain. Produced by Dave Fossett B.S.E E from Nonheastem University in 1986 He I S a life member o f 7bli
lillp /;\\I\\\
112 d1011 i l c . ~ ~ ~ . ' - ~ L~l ll l~h ~L l il' ~ ' l ~ \ Beta P I
He has held technical and management positions on the staffs o f GTL
[2] Hunt G L.: Thr Great Batierj Search Laboratories. MASSCOMP (later Concurrent Computer). American Power
IEEE Spectrum, vol. 35. N o 1 1 , pp 19-28. N o \ 199: Conversion. UB N e w o r k s (later Newbridpe Networks). Raytheon Compan!
USABC official infonnational website and Celox Networks
hiij> ;. \ \ \\I\' pcocitics cotii!h:iitcar2illi l;haitc~w*255 liiii He IS married and father of two
131 USABC official inf ormational websitc
iitip OCIIICS C(II~~;IW
I : hI i W
ti ~ ~

[8] Lynch, William A.. UMass. Lowell. Nickel Cadmium Batten

Evaluation. Modeling. and Application in an Electric Vehicle, 1997

[9] Youn, Kilyoung. UMass. Lowell. Realisric Elecrric Vehicle Barrety 7esi
1998

I IO] Alaoui. Chakib. UMass. Lowell. Elecmc I'ehrce Energy Maiiapenieni

S>,steni (sic). 1998lEEE Guide for Application of Power Apparat1i.c
Bushings. lEEE Standard C57.19.100-1995. Aug 1995

17: