Nickel-Iron

This all but forgotten technology has a very important place to occupy with users that desire
very long life and the ability to suffer abuse in their battery systems
Peter J. DeMar
Battery Research and Testing, Inc.
Oswego, NY, USA
[email protected]

AbstractThis paper is going to look at real life aged 80+ year

old Nickel-Iron cells that are still functional and will explain the
simple recovery techniques that were documented in an original
Edison Alkaline Storage Battery brochure from the 1920s. Some
of the cells had been charged intermittently, many had sat off
charge for many years, and some had sat off charge and all but
empty, but all made very substantial recoveries, and when
subjected to discharge testing that followed the guidelines of the
IEEE 1115 they all were able to pass load tests at their applicable
rate.

I.

INTRODUCTION

The aim of this paper is to introduce this very old battery

technology, which is over 100 years old, to those that have no
idea that such a battery exists, or ever did exist. The majority
of us are most familiar with various forms of lead acid, or
nickel-cadmium as they were and still are the batteries of
choice for most stationary applications, with Lithium Ion (in
various forms) and other technologies gaining acceptance in
many stationary applications.
The cells that are reported on in this paper are real life aged
cells with an average age of 85 years, and the conditions that
they were operated in and stored in were less than ideal. They
spent the last approximately 60 years in a wooden shed, at a
hunting lodge in the Adirondack Mountains, with temperatures
from below -18C to above 32C. They were charged
intermittently and often sat in a partial or full discharged state
for weeks or months or years, at a time. Their function was to
provide lighting to the lodge.
Waldemar Jungner of Sweden created the first Nickel-Iron
battery in the late 1890s and has multiple patents on the
design. However he found that by substituting Cadmium
instead of Iron that he could improve the performance and
efficiency of the cells, and he abandoned the development of
the Nickel-Iron cell in favor of Nickel-Cadmium. There are
two patents for the Ni-Fe technology and one for the Ni-Cd
technology in his name from 1899.
Thomas Edison believed that Ni-Fe could displace lead acid
as the battery of choice and in 1901 obtained both a US and a
European patent for his version of the technology. Edison

978-1-4577-1250-0/11/$26.00 2011 IEEE

performed some very extensive testing on his cell designs to

verify their hardiness for usage in RR applications, or electric
automobiles (which he thought would replace internal
combustion engines), or material handling (tuggers and such).
Two of my favorite tests that he created to demonstrate the
durability of his Ni-Fe batteries are as follows.
He mounted a battery system on a cart and then the cart was
rammed into a brick wall at 15 MPH and the battery had to
survive 1,000 such shocks, which it did. My favorite test
though was where he hooked a cell to a motor driven pendulum
and the device raised the cell and dropped the cell onto a
wooden platform. The cell survived 1, 776,000 such drops and
then following that it passed a load test. (1)
The Thomas Edison battery factory in West Orange New
Jersey USA produced cells from 1903 to 1972 when it was sold
to the Exide Battery Company (name at that time), which
continued production until 1975 when the factory was closed.
Presently there are two companies that are still manufacturing
Ni-Fe cells and they are Kursk Accumulator in Russia, and
ChangHong Battery in China.
It is our belief that this very old technology still has a place
in the current market, where the user has a need for a very long
life battery that can stand frequent cycling and abusive
conditions. In America these are being offered for usage in the
off grid market due to their long life and ability to withstand
the daily repeated cycling, and setting in a partial state of
discharge for extended periods.
II.

THE BOAST

It has been stated that Thomas Edison boasted of a 100 year

battery with his Nickel-Iron design, but I have not been able to
successfully locate those exact words. Now that sounds like a
pretty bold marketing statement, sort of like the original
marketing words Maintenance Free when referring to VRLA
cells. However our experience in testing these old Nickel Iron
cells convinced us that it may not have been just boldness or
marketing on his part.

III.

THE OPPORTUNITY TO PROVE OR DISPROVE

As can be seen in the following picture of the three

different cell sizes the two on the left are the A4H and the
A8H, and the one on the right is an A8 cell.

With our gaining access to a substantial number of Thomas

Edison Alkaline cells in two different amp hour sizes (150 and
300AH) at the five hour rate, we had an opportunity to find out
if there was any validity in a 100 year life statement.
Our first task was to locate documentation on these cells,
and we turned to the Internet to locate manuals, documents,
specifications, etc (1,2). While locating different manuals was
easy enough, we could not determine the serial number code
that was stamped into the top of each cell, so we did not know
the age of any of the cells. Luckily we finally reached out to
Ole Vigerstol of Saft who contacted their Railroad Group
people, who then provided us with the original Edison Date
codes. And yes we did have cells that were all built between
1924 and 1931.
We also utilized installation and maintenance manuals from
both Saft (3) and ChangHong (5) as guides or comparison
purposes, to see if there were any major differences in their
instructions from the Edison manuals. While there were some
differences none of the differences were of any great concern.
When we received the cells they were in various conditions
of charge, or fill, or just plain cleanliness. It must be
understood the majority of these cells had been setting off
charge in various states for many years.

Picture 1. This shows the general condition of some of the cells as we

received them.

The following shows the three size cells that we received.

These originally were coated with a rubber like paint
compound which was named Esbalite which is described in the
Edison manuals as a special insulating paint. This coating
covered the sides and the bottom, but none was on the top of
any cell. However during the cleaning process of the cells, the
coating which came off and we have not yet determined what
we will use as a coating, so for our experiments we used wood
spacers to maintain separation between the cells.

Picture 2. The three cell sizes.

The following picture shows the carbonate build up that we

found in some of the cells, which of course has a severe impact
on the cells performance.

Picture 3. Carbonate build up we found in some cells.

IV.

RECOVERY PROCEDURES

We randomly took cells of the same AH rating and made

up different battery strings, and in some cases we took single
cells and with each we boost charged and then float charged at
the voltages that were stated in a 1916 Thomas Edison manual
and then followed up with load testing at the full published five
hour rate. All of the cells or battery strings failed miserably.

Our as found individual cell voltages ranged from 0.06 of a

volt to 1.36 volts. We attributed this wide of a voltage spread
to the fact that some cells had been on charge just prior to our
receiving them and some had been off charge for months or
years. Some were filled with electrolyte and some were empty
or nearly so.

manufactures
Nickel-Cadmium
cells,
ChangHong
manufacturers both Nickel-Cadmium and Nickel-Iron, and of
course the Edison cells are all Nickel-Iron. The common
denominators are the Nickel and the Potassium Hydroxide
electrolyte. We decided to follow Edisons procedure since the
cells were Edison cells.

The following is from one string of A8 cells and is an

example of the age of the cells, and the as found open circuit
voltages. As can be seen in this battery the age of the cells
range from 1926 to 1930 with a voltage spread from 0.005 to
1.356. It is easy to see which cells have been setting around
the longest and which ones were recently on charge.

After the electrolyte replacement we placed the cells back

on float and then boost charged at 1.65 VPC followed by a
return to float at 1.49 volts per cell and then further load tests.
The following shows the same cells as the previous chart, but
with the respective float voltages following 100 hours of boost
charging, and then being on float charge for about six weeks.

Serial
Number
3404X
1587X
4101P
1089R
499W
1595X
565T
2164R
2050R
1610X
2080R
7330P

year
built
1930
1930
1926
1927
1929
1930
1928
1927
1927
1930
1927
1926

as found
voltage
8/4/10
1.344
1.341
0.105
0.005
1.353
1.356
0.856
0.720
1.279
1.250
0.051
0.044

While experimenting with these cells we realized that even

though the voltage would rapidly drop off in a matter of
minutes when we tried to run a load test at the full published
rate of the particular cell or battery, that if we lowered the
discharge rate, the battery would hold voltage for a
substantially longer amount of time, even though the best string
would only support a 10 amp load for 22 minutes to an end
voltage of 12.0 volts.
Throughout our testing we followed the instructions in the
Edison manuals, and following those instructions we decided
to replace the electrolyte. We obtained new electrolyte from
Saft as they are a major supplier of Nickel-Cadmium batteries
and the Potassium Hydroxide that they use is the same as what
is utilized in the Nickel-Iron batteries. Both Saft and
ChangHong also provide instructions that explain that when the
capacity drops off and boost charging does not return desirable
results, that the electrolyte needs to be replaced.
A discrepancy that we discovered between the three
manufacturers (Edison, Saft, and ChangHong) is that during
the electrolyte replacement procedures, Edison states to pour
out about half of the old electrolyte then to shake the cell
vigorously and then to pour out the remaining electrolyte, but
to not rinse with any water, and then to fill with new
electrolyte. Changhong says to pour out the old electrolyte and
to shake it, and if the electrolyte is dirty in color to rinse it with
distilled water two or three times, and then to fill with new
electrolyte. Saft says to carefully pour out the old electrolyte
and then to fill with new. This was the only real difference that
we found between the three manufacturers. Of course Saft

Serial
Number
3404X
1587X
4101P
1089R
499W
1595X
565T
2164R
2050R
1610X
2080R
7330P

float
voltage
9/26/10
1.473
1.477
1.482
1.469
1.477
1.471
1.467
1.463
1.470
1.469
1.470
1.443

V. PUBLISHED RATINGS
There were three different model cells that we had received
and played with. There were models A4H, A8, and A8H cells.
The H in any model just means that the cell is the same AH
rating but it has more electrolyte reservoir and is approximately
7.6 Centimeters taller than the cells that do not have the H in
their model number. The H designated cells were to be used in
applications where there longer time periods between
maintenance intervals.
With the three battery strings that we are reporting on here,
we utilized the five hour rating from the Edison manual, and
we used the end voltage of 1.0 VPC, also from the Edison
manual.
The published rate for the A4H cells is 30 amps for five
hours to an end voltage of 1.0 volt.
The published rate for the A8 and A8H cells is 60 amps for
five hours to 1.0 volt.
VI.

WHAT WERE WE TRYING TO UNDERSTAND

We are trying to learn if the Edison Alkaline cells that we

had would indeed function at their advanced ages. But there is
no existing standard to follow as a guide, so we decided to
utilize the IEEE 1106 (4) since it is for Nickel-Cadmium cells
and the only primary difference between the two types is the
Cadmium content in place of the Iron, otherwise they are
Nickel and Potassium hydroxide.

With our main goal being to determine if these cells or

batteries would work reliably at their extended ages, and not to
prove a specific capacity we decided to utilize the 1% per year
aging factor from annex E of the IEEE 1106 .
With cell ages ranging from 80 to 87 years of age and an
average of 85 years we decided to be conservative and used an
aging factor of 0.2 which would reflect a 1% per year de-rating
factor for an 80 year old cell. With that decision made we
made we settled on the following discharge rates. As you will
notice we used the same five hour rate for the two different
models, even though one was a 150 AH model and two were
300 AH models. We do not yet understand why the A4H cells
performed so much better than the A8 and A8H cells. We are
suspecting that it was due to the fact that the A4H string had
many more discharge/recharge sequences than either of the
other strings, but only time will tell if the A8 and A8H strings
continue to improve over time and cycling.
A8 and A8H cells used a 15 amp rate to 1.0 VPC

Chart 1 is comprised of A4H cells.

350
300
250
200

The following charts show the load test results at various

times over the past approximately twelve months on all three of
the strings. All of the load tests were run at the five hour rates
to 1.0 VPC. As can be seen, the load tests that were run before
we replaced the electrolyte were somewhat dismal, however as
you will see in the load tests that were run after the electrolyte
had been replaced were substantially improved, and then by the
last load tests which were all performed in July 2011there was
further improvement.

19Dec10

150

7Jul11
100
50
0
RuntimeinminutesoftheA4H
stringat15ampload

A4H cells used a 15 amp rate to 1.0 VPC, where as if we

used a 0.2 aging factor the rate would have been 6 amps.
We made up one twelve cell string from the A8 cells,
another twelve cell string from the A8H cells and an 18 cell
string from the A4H cells. Each was placed upon its own
charger. We utilized a varying range of float voltages at
different times as part of this experiment. Primarily we kept
the voltage between 1.47 and 1.5 but did sometimes go up to
1.57 volts per cell and 1.65 up to 1.85 when we equalized or
boost charged. These voltages came out of the Edison manual
and the float voltage corresponds to that recommended by both
ChangHong and by Saft.

16Aug10

Chart 1. This chart shows the increase in run time with this battery with a 15
amp load.

Chart 2 is the string that is made up of A8 cells.

350
300
250
200
17Sep10
150

20Dec10
9Jul11

100
50

VII. TEST RESULTS

0

As can be seen in the following charts, with each battery

string there was some amount of run time under load but it was
not until we replaced the electrolyte and then ran a number of
discharge and recharge scenarios that the run time really
returned. While we could not get any strings to recover to a
level where they could support their full published rates, it was
encouraging that they could support an age related reduced
discharge rate for a full five hours.
In each chart the left hand column is the original as found
run time, with boost and float charging but no electrolyte
replacements.
The middle column is after electrolyte
replacement and boost charging and from float voltage. The
green line is after some number of discharges and boost
charges and also is from a float condition.

RuntimeinminutesoftheA8
stringat15ampload
Chart 2. This chart shows the increase in run time with this battery with a
15 amp load.

Chart 3 is the string that is made up of the A8H cells.

That these 80+ year old cells are still functional proves
without any doubt that Nickel-Iron is a long lived design, now
it will just take another 15 years to see if they will still be
functioning at 100 years of age as Thomas Edison is supposed
to have declared.

350
300
250

IX.
200

15Sep10
21Dec10

150

14Jul11
100
50
0
RuntimeinminutesoftheA8H
stringat15ampload
Chart 3. This chart shows the increase in run time with this battery with a
15 amp load.

VIII. CONCLUSION
This find of these old Thomas Edison Nickel-Iron cells has
been quite an education for us at Battery Research and
Testing, as our work for the past 29 years has been primarily
with lead acid and some Nickel-Cadmium, but with nothing of
the age of these cells. In fact the oldest lead acid cells that we
have load tested and that were still functional were old Exide
Manchex strings that were 42 years old, and it appears that the
only existing lead acid cells that might be able to be functional
at 40 years of age are the Bell developed round cells for
Telecom applications.
What we have learned has opened up our minds to explore
possibilities for this design long life design cell. It would sure
seem that any site that has a requirement for a long life battery
that will tolerate abusive conditions would consider the total
life costs of these type cells and see which works out to be the
most cost effective.
I have approached the IEEE Battery Working Group to
have Nickel-Iron included in the IEEE 1106 and IEEE 1115
documents during the recent re-affirmation process, but it was
decided to not include Nickel-Iron in those documents at this
time. It would seem that since the IEEE 450 and IEEE 485
documents cover all of the different Vented Lead Acid designs
such as Lead Antimony, Lead Selenium, Plante, or Lead
Calcium which all use sulfuric Acid, that the IEEE 1106 and
IEEE 1115 which covers Nickel-Cadmium cells which also
uses Potassium Hydroxide as the electrolyte could easily have
been expanded to include the Nickel-Iron cells.

ACKNOWLEDGEMENTS

I would like to extend a special thank you to Weston

Mitchell of the Fayetteville Hunting Club for providing us a
chance to learn about Nickel-Iron cells and these specific
pieces of the Thomas Edison history. If it had not been for his
environmental consciousness, we would not have our eyes
opened to this very durable battery technology that is all but
forgotten here in the US. Also thanks need to go to Bob
Howland and Jim Miner of Battery Research for their
assistance and technicians time to perform the various
assembly of the strings and testing. And last but not least,
thanks to Ole Vigerstol and Jim McDowall from Saft for their
assistance and comments, as well as to Sam Zhow from
Sichuan Changhong Battery Company for his support.
X.

REFERENCES

1. General Information and Instructions for the Operation and

Care of the Edison Alkaline Storage Battery. Edison
Publication Bulletin 850X.
2. The Edison Alkaline Storage Battery. By the technical staff
of the Edison Storage Battery Company. For the National
Education Association Joint-Committee Series Monograph
III. Document 804. Copyright 1916. From the University of
Michigan Libraries. Digitized by Google.
3. Saft Installation and operating instructions for single cell
Ni-Cd, models SCL, SCM, and SCH plastic case cells.
4. IEEE Std. 1106. IEEE Recommended Practice for
Installation, Maintenance, Testing, and Replacement of
Nickel-Cadmium Batteries for Stationary Applications.
5. ChangHong Battery Performance Data Manual. ChangHong
Battery Operation and Maintenance Manual.