Vous êtes sur la page 1sur 4

James A.

McDowall
SubstationBattery Options:
Present and Future
F
or several decades, vented lead-acid batteries have been used
to power dc switchgear and other substation loads. While
they tend to have short lifetimes in the harsher environments of
small distribution substations, these batteries have, on the whole,
provided satisfactory service. However, the high cost of some
battery maintenance operations, such as water additions, specific
gravity checks, and connection maintenance, has driven some us-
ers to install valve-regulated lead-acid (VRLA) batteries in their
substations. VRLA technology offers low cost, high energy den-
sity, and freedomfromsome maintenance chores. What many us-
ers have not properly understood is that these benefits come at the
price of battery life, reliability, and overall cost of ownership.
Users have also seen laptop computer batteries evolve from
nickel-cadmium through nickel-metal hydride to lithium-ion in
the space of just a few years. Articles on electric vehicle (EV)
research mention similar batteries, along with more exotic vari-
eties. It is reasonable to assume that all this research work will
lead eventually to a better battery for substations. But, when will
that happen, and will you be able to afford it?
This article discusses the benefits and drawbacks of some of
the potential alternatives to vented lead-acid batteries in substa-
tion service. These include VRLA, nickel-cadmium (Ni-Cd),
nickel-metal hydride (Ni-MH), lithium-ion (Li-ion), and lith-
ium polymer (Li-polymer). The aim of the article is to provide
an overview of ongoing battery development work and an idea
of the timeframe for commercial availability.
Substation Battery Requirements
Substation batteries are required to provide reliable power to op-
erate circuit breakers and other protective devices, while also
providing for the continuous operation of various low-power in-
dication and control functions.
Although the loads required for circuit breaker operation are
of short duration (typically a fewcycles for tripping and closing,
to several seconds for spring charging) the batteries normally
provided for this application are not of the high-performance
type. There are two reasons for this. First, the mix of short dura-
tion, high current switching loads, and continuous low current
control loads often favors medium performance general-pur-
pose battery types, particularly in larger substations. Second,
and more important, high-performance lead-acid batteries have
thinner plates and, therefore, a shorter life than their medium
performance counterparts, and this shorter life would compro-
mise battery reliability.
The control buildings in many larger substations are either
climate controlled or have sufficient thermal inertia that the bat-
4 0272-1724/00/$10.002000 IEEE IEEE Power EngineeringReview, November 2000
J.A. McDowall is with Saft America, Inc., North Haven, CT 06473 USA.

D
i
g
i
t
a
l
S
t
o
c
k
,
1
9
9
6
teries are not subjected to temperature extremes. This is not the
case with smaller distribution substations, in which batteries are
frequently installed in small, poorly ventilated, metal structures.
Substation batteries must be capable of providing reliable oper-
ation within the expected operating temperature range, even if
that range is extreme.
The Drive to Lower Battery Costs
When sized and maintained in accordance with IEEE recom-
mended practices [1] [2], vented lead-acid batteries offer reli-
able service at a reasonable cost. However, many users view the
taking of specific gravity measurements and periodic water ad-
ditions as a burden, since this requires specialized training for
the maintenance technician and involves personnel risks in han-
dling hazardous materials (battery acid).
Battery discharge testing, long known to be the only absolute
method for determining a batterys condition, is simply not per-
formed by many utilities, since the cumulative cost of testing of-
ten exceeds the cost of the battery itself.
As operating budgets continue to be squeezed, the drive to re-
duce overall battery costs is likely to increase. In some cases,
this may involve a scaling back of maintenance and testing oper-
ations, thus increasing the risk of unexpected failures. In others,
it may lead a utility to consider alternative battery types.
Valve-Regulated Lead-Acid Batteries
VRLA batteries seem to offer an attractive alternative to vented
lead-acid batteries. They have typically been marketed with the
same 20-year warranty that applies to general-purpose vented
batteries, while offering the lure of maintenance-free operation.
It is certainly true that VRLA batteries are maintenance free
with respect to the electrolyte. Specific gravity measurements
are not possible, and these batteries are designed to be operated
throughout their lives with no periodic water additions. The acid
electrolyte is immobilized so that, even if a battery unit is
dropped, no spillage should result. This means that a VRLAbat-
tery in normal operation is inherently safer than a vented unit, so
it is reasonable to assign less-specialized maintenance techni-
cians to care for these systems.
Some utilities have chosen to use VRLA batteries in their
substations, pointing to the similarities between the distribution
system for electric power and that of telephone service. In the
local telephone infrastructure, VRLA batteries are routinely in-
stalled in outdoor cabinets and are subjected to much more ex-
treme conditions than in substations, with relatively few total
system failures.
There are, however, significant differences between tele-
phone and power distribution systems in terms of battery opera-
tion. Batteries in the telephone system are designed to support
their loads for long periods, typically 8 hours. Although normal
substation battery duty cycles are also 8 hours in duration, the
critical switching load is very short and involves much higher
currents. Abattery that has degraded to, say, 50%of its rated ca-
pacity will at least provide something close to 4 hours of backup
in telephone service, while a substation battery, if accurately
sized, may be incapable of providing its switching loads once it
falls below 80% of its rating.
In addition, telephone batteries are typically deployed in par-
allel strings. This is at least in part because of the tendency of
VRLA cells to fail in an open condition. This may be a result of
dryout, or it may be due to excessive internal corrosion. If a sin-
gle cell fails open, the entire string is lost. For a lowcurrent load
supplied by two parallel strings, the loss of one string causes the
loss of a few hours of reserve time, but does not completely dis-
able the system.
In substation operation, it is highly unusual to install parallel
battery strings. Indeed, the benefits of doing so would not be the
same as in telephone service. This is because battery efficiency
decreases rapidly as the discharge time approaches zero. For a
system employing two parallel battery strings that together are
just capable of supporting a specified 1-minute load, the loss of
one string would be catastrophic. The initial voltage drop that is
characteristic of lead-acid batteries would leave the system well
below the specified minimum voltage almost instantaneously.
To be sure, there are some potential mitigating factors here.
The battery sizing may have included an aging factor, a design
margin for load expansion, and/or a factor for low temperature
operation. If the battery is not operating under the worst case
sizing assumptions, there may be enough margin in a 50%string
to support the load. The battery may be sized to trip, say, ten
breakers simultaneously, while in reality it may only have to trip
a few of these. Finally, there may be considerable tolerance in
the system; batteries are often sized to a minimum voltage of
105 V, while trip coils can operate down to 70 V. The total mar-
gin available will vary from system to system, depending on the
sizing criteria.
Summarizing this issue for VRLA batteries, there is a consid-
erable risk involved in installing a single VRLAstring in a substa-
tion. If parallel strings are installed, to operate reliably, they must
be redundant, either by design or by a sufficient degree of conser-
vatism in the sizing calculation. In building in redundancy, how-
ever, the main aim of reducing battery costs is compromised.
Despite the early claims of maintenance-free operation,
VRLA batteries require considerable surveillance and testing to
maintain a high degree of reliability. IEEE 1188-1996 [3] rec-
ommends quarterly internal ohmic measurements and annual
discharge testing for VRLA. These measures are largely ignored
by the telephone operating companies because of their lowloads
and use of parallel strings, as detailed above. In substation oper-
ation, however, these practices are doubly important because of
the higher currents involved.
An additional issue with VRLA is the life expectancy. The
batteries that are sold in the United States as having a 20-year
life are sold elsewhere in the world as 10-year batteries. Worse
still, recent reports [4] of VRLA testing indicate serious prob-
lems with premature failures and an actual life, even under fa-
vorable conditions, of only 5-7 years. Obviously, a justification
for using VRLA that is based on a 20-year life would be in seri-
ous jeopardy on this basis, particularly if the end result is the
loss of a transformer.
Nickel-Cadmium Batteries
Despite being available since the turn of the century, Ni-Cd bat-
teries have not been extensively used in U.S. substations. This is
partly because they were not promoted to U.S. utilities before
about 20-25 years ago and partly because of their high initial
cost relative to vented lead-acid batteries. Indeed, for larger sub-
stations with reasonably mild operating temperatures, it is diffi-
cult to justify these batteries. However, for more extreme
conditions, Ni-Cd offers some useful benefits.
Probably the biggest advantage of Ni-Cd batteries is their re-
sistance to high temperature. There is a rule of thumb for
lead-acid batteries that their operating life is reduced by 50%for
every 8 C (15 F) increase in temperature above 25 C (77 F).
For the same temperature increase, Ni-Cd life is reduced by only
20%. Thus, a general-purpose vented lead-acid battery that has a
IEEE Power EngineeringReview, November 2000 5
15-year life at 25 Cwould give only 3.8 years at an effective av-
erage temperature of 41 C (106 F), while a 20-year Ni-Cd bat-
tery would give 12.8 years at the same temperature.
Whereas the structural components of lead-acid positive
plates degrade over time through a corrosion process, the steel
internal hardware of Ni-Cd batteries is protected by the alkaline
electrolyte used in these systems. This allows thin-plate
high-performance designs to be built without sacrificing life ex-
pectancy. This is an important factor in smaller distribution sub-
stations, in which the switchgear often requires relatively high
operating currents and has lower continuous loads, thus favor-
ing higher-performance batteries. Coupled with the more ex-
treme operating temperatures frequently found in smaller
distribution substations, this can give high rate Ni-Cd batteries a
distinct advantage in long-term cost effectiveness.
Future Options
Most of us have experienced a huge increase in the number of
battery-powered systems that affect our personal and business
lives. Whether it be cordless appliances, cellular telephones,
power tools, or laptop computers, we have seen rapid develop-
ments in the this field as manufacturers strive for lighter and
more powerful batteries. Many laptop users who are unfamiliar
with stationary batteries are quite comfortable discussing the
merits of Ni-Cd, Ni-MH, and Li-ion batteries.
In addition to the huge sums allotted to research in the porta-
ble battery field, government mandates on vehicle emissions are
fueling similar research in EV batteries, for which hundreds of
millions of dollars are being spent. Organizations such as EPRI
and the U.S. Department of Energy are also funding research
into battery energy storage for load leveling and power quality
applications.
From all this work, it is quite reasonable to assume that via-
ble stationary battery systems will be developed. What is less
obvious is the potential impact that these systems may have on
the substation battery market.
For the most part, the most favorable battery developments
fall into three categories with which many users are already fa-
miliar: Ni-MH, Li-ion, and Li-polymer. Each of these catego-
ries covers a broad range of battery systems, but it is possible to
generalize their characteristics and potential for substation ap-
plications.
Smart Batteries
One characteristic shared by these systems is that they all re-
quire electronic controls to prevent overcharging. While these
controls are required for safe operation and correct functioning
with existing chargers, it is easy to build in additional smart fea-
tures, such as communications interfaces that allow remote
monitoring and control of the system, right down to the level of
individual cells. These systems have already proved effective on
EV batteries consisting of 300+ cells.
The resulting ability of smart batteries to reduce routine sur-
veillance operations is very real, leading to potentially signifi-
cant cost savings.
Nickel-Metal Hydride
Ni-MH batteries are chemically rather similar to Ni-Cd, but
have much higher energy density, a factor that is vital in EVs,
but of much less importance for substations.
Although Ni-MHbatteries are maintenance-free with respect
to the electrolyte, this advantage is offset by increased sensitiv-
ity to high temperatures, compared with Ni-Cd. Since high tem-
perature operation is one of the essential factors in considering
batteries other than vented lead-acid, this seems to count against
Ni-MH. Having said this, there are continual developments in
this technology, and it is possible that some future development
will improve this situation.
Unfortunately, Ni-MH batteries are also more costly than
Ni-Cd. While it is possible to justify the use of Ni-Cd in high
temperature applications, this will not be possible with Ni-MH
unless the high-temperature aging characteristics are signifi-
cantly improved.
LithiumIon
Recent developments in Li-ion technology have led to this bat-
tery type supplanting Ni-MHin many high-end consumer appli-
cations, such as laptop computers. This battery type also shows
considerable promise in EV and energy storage applications.
However, one of the main attractions of Li-ion is its energy den-
sity, which is around 6-8 times as high as typical vented
lead-acid substation batteries. As mentioned previously, how-
ever, energy density is not a major factor in substations, so it
would be difficult to justify Li-ion on this basis alone.
6 IEEE Power EngineeringReview, November 2000
120 V vented lead-acid substation battery on a two-step seismic rack
High temperature operation is a definite plus for Li-ion. In
cycling tests at 60 C (140 F), there is very little difference in
the rate of aging compared with tests at 25 C (77 F).
Li-ion batteries are hermetically sealed, so they are com-
pletely maintenance free with respect to the electrolyte. Since
their chemistry dictates that no water be present in the system,
there can be no electrolysis or gas evolution. This makes them
safer to handle and gives them an ampere-hour charge effi-
ciency of virtually 100%.
To date, most testing of this battery type has been directed to-
wards the cyclic operation of charge and discharge, since this is
the normal usage mode in consumer and EV applications. Bat-
teries in substations, however, are subjected to continuous float
charging, and much work remains to be done to characterize
Li-ion for this type of operation. Initial indications point to a
float life of 15 years or more.
As many laptop computer users are aware, Li-ion batteries
are very expensive, compared with Ni-MH, which in turn are
more expensive than Ni-Cd. Their cost will come down as pro-
duction volume increases, but will still be significantly higher
than lead-acid.
LithiumPolymer
There are various types of Li-polymer batteries, with markedly
different characteristics. The type discussed here is a dry polymer
type that is being developed for telecommunications applications
and has been discussed in at least one technical paper [5].
The electrochemical basis for Li-polymer batteries is essen-
tially similar to that of Li-ion, except that the organic electrolyte
of the Li-ion types is replaced by a solid polymer that encapsu-
lates the electrodes. This construction offers a further improve-
ment in safety compared with Li-ion, by reducing flammability.
In order to achieve adequate conductivity in the polymer
electrolyte, these cells must be operated at high temperatures,
typically 60-80 C (140-176 F). This means that the ambient
temperature in substation use is not a factor, and the battery will
give its full life expectancy under all conditions.
Just what that life expectancy is, remains to be seen. As with
Li-ion, Li-polymer batteries are still under development, and little
is known about their life expectancy under float charging condi-
tions. There is also some question whether these batteries will be
cost effective for the capacities typically used in substations.
Availability
Larger capacity Ni-MHbatteries are nowin production for EVs.
Large Li-ion cells are being produced at a pilot plant level. To
qualify either type for stationary applications such as substa-
tions, however, will require additional work. At present, the best
estimate for availability of these battery types in large-scale pro-
duction for stationary applications is 3-5 years.
Importance of Life-Cycle Costing
Utility users are very familiar with the characteristics of vented
lead-acid batteries, and it is reasonable to compare competing
brands based on their initial cost.
As users begin to consider other battery types, a comparison
based on initial cost is no longer valid. For example, VRLA bat-
teries generally have a lower initial cost than their vented coun-
terparts. However, if all the IEEE recommendations for
maintaining VRLA batteries are followed, and the real life
(rather than the warranty) is factored in, VRLA can turn out to
be a much more expensive proposition.
The other battery types in this comparison are, or are likely to
be, more expensive than vented lead-acid. While some of their
characteristics may be desirable, the question has to be asked:
Are they worth it? The only accurate way to answer this ques-
tion is to perform a life-cycle cost analysis. Such an exercise
considers all the costs associated with battery ownership over a
certain period of time, including the replacement of shorter-life
batteries and all associated maintenance and testing activities.
Saft has developed software for life-cycle costing that is
freely available to battery users. This software allows estab-
lished battery types to be compared for a given application, with
a graphical analysis that clearly shows the relative cost benefits
of the different technologies. All costs are under the users con-
trol, so that an unbiased comparison can be made.
Conclusions
For many substations in which the operating environment is rel-
atively mild, vented lead-acid batteries will continue to be the
battery of choice for the foreseeable future. For those installa-
tions in which the battery is exposed to higher temperatures, op-
portunities exist for cost-effective deployment of other battery
types. Of the currently available technologies, nickel-cadmium
offers the best potential.
In the authors opinion, the current trend towards the use of
VRLA batteries in substations poses a significant risk of cata-
strophic failure. Utilities using, or considering the use of, this
technology should reevaluate their position based on the current
level of industry knowledge of this battery type.
While newer technologies, particularly lithium ion, offer
some interesting characteristics, the higher cost of these new
batteries will be difficult to justify in this application for at least
the next few years.
Whenever a new battery type is considered, it is important to
review all the costs involved in owning that battery. The best
way to do this is to perform a life cycle cost analysis.
References
[1] IEEE Recommended Practice for Sizing Lead-Acid Batteries for
Stationary Applications, IEEE Standard 485, 1997
[2] IEEE Recommended Practice for Maintenance, Testing, and Re-
placement of Vented Lead-Acid Batteries for Stationary Applica-
tions, IEEE Standard 450, 1995.
[3] IEEE Recommended Practice for Maintenance, Testing, and Re-
placement of Valve-Regulated Lead-Acid Batteries for Stationary
Applications, IEEE Standard 1188, 1996.
[4] D.O. Feder, Performance measurement and reliability of VRLA
batteries, in Proc. 1995 Int. Telecommunications Conf. (INTELEC
95), pp. 22-28.
[5] C. Letourneau, D. Wilmont, and D. Worboys, Lithium polymer
batteries: The next generation power source, in Proc. 1997 Int.
Telecommunications Conf. (INTELEC 97), pp. 87-91.
About the Author
James A. McDowall is business development manager for the
Industrial Battery group of Saft America, Inc. He is a graduate
of the Royal Institute of Chemistry in London, UK. He has
worked in the battery industry since 1977, providing technical
support services to customers and sales personnel. During this
time, he has worked extensively with nickel-cadmium, vented
lead-acid, and valve-regulated lead-acid batteries. He is very
active in the battery standards field and has been a working
group chair for a number of IEEE standards projects. He is
chair of the IEEE PES Stationary Battery Committee and is
chair of the IEEE Standards Coordinating Committee 29 on
Stationary Batteries.
IEEE Power EngineeringReview, November 2000 7

Vous aimerez peut-être aussi