Daniel W. O'Brien
BSEE from Purdue University and a MS in Management from Krannert
Graduate School, Purdue. He has worked the last eight years at NACC
and is currently the Director of Engineering with responsibilities
over all process and design engineering groups. In addition, he has
responsibilities as the applications engineer for the new
Electrochemical Capacitor product line. He has written several
articles and has worked with numerous companies to design this new
type of capacitor into products.
Of the three capacitor classifications
(electrolytic, electrostatic, and electrochemical), electrochemical
capacitors (EC capacitors) are the least understood. Like aluminum
electrolytics, EC capacitors utilize two metallic electrodes with
an electrolyte that functions as two capacitors in series connected
via the electrolyte. Instead of using a metal-oxide dielectric, EC
capacitors have a dielectric layer that forms naturally with the
application of voltage. This dielectric forms in a very thin double
layer on the surface of the capacitor's electrodes. Because of the
dielectric double layer, EC capacitors are also known as Double
Layer Capacitors (DLC). The charge stored in these capacitors is a
true electrostatic charge and not a result of chemistry, as the
name electrochemical implies. EC capacitors are perhaps best known
for their high capacitance values, and are specified in Farads (F)
or kilo-Farads (kF).
Although the theory behind the EC capacitor has been known for
over 100 years, it was not until the 1960s that SOHIO developed a
functional energy storage device. NEC developed the first
successful product for the memory back-up market in the late 1970s.
In the 1990s the technology scaled up and commercialized targeting
pulse power applications, engine starting applications, and
specialty energy storage applications. Because NEC and Pinnacle
were two of the first players in the market, EC capacitors are
often referred to by these company's trademarked
namessupercapacitor and ultracapacitor.
EC Capacitor Construction
EC capacitors are assembled by combining individual cells in
series. The decomposition potential of an EC capacitor's
electrolyte limits its cell voltage. For a capacitor with an
aqueous electrolyte, the cell voltage is limited to between 0.8 and
1.6 Vdc. The voltage for a cell using a non-aqueous electrolyte can
reach as high as 3 or 4 Vdc. However, the use of a non-aqueous
electrolyte does require the use of dry rooms, vacuum chambers, and
other expensive processing methods that result in significantly
higher production costs.
Figure 1 shows how EC capacitors fit within the energy /
power spectrum of batteries and conventional capacitors. EC
capacitors have up to 80x higher energy density than conventional
capacitors and up to 10x higher power density than batteries. By
combining batteries or conventional capacitors with EC capacitors,
an entire power system can be optimized both for power and energy
resulting in a more efficient system.
Figure 1: The Ragone plot demonstrates how EC
capacitors bridge the gap between batteries and capacitors. While
batteries have discharge times measured in minutes or hours and
conventional capacitors in pico to microseconds, EC capacitors
serve the nano-seconds to seconds range.
To gain the higher voltages required by applications, the cells
must be connected in series. For example, an automotive application
may require 14.5 Vdc. Assuming a capacitor design is capable of 1.5
Vdc per cell, the final design would consist of 10 cells connected
in series (1.5 Vdc per cell x 10 cells = 15.0 Vdc). The cell
voltage capability also depends on the specific capacitor design
and can often be adjusted depending on the application.
Activated carbon is a common electrode material due to its high
surface area (1000+ meters/gram), availability, chemical stability,
and relatively low cost. The two most common used forms are powder
or carbon cloth. However, research continues into other materials
such as carbon nanotubes, carbon fibers, carbon foams, conducting
polymers, and ruthenium oxide. EC capacitors that use the same
electrode material for both electrodes are called symmetric
capacitors. The alternative is the asymmetric design, which uses a
different material for one of the electrodes such as nickel
hydroxide or ruthenium oxide.
- Symmetric Type Design
Symmetric EC capacitors can be put in the application without
regard to polarity, and can be discharged to zero volts like a
conventional capacitor. In addition, symmetric capacitors typically
have lower internal resistance (ESR) than asymmetric type
capacitors, which results in better power density since for any
P = V² / (4 x ESR)
Where power (P) is measured in Watts, voltage (V) in volts, and
Equivalent Series Resistance (ESR) in Ohms ().
P is inversely proportional to ESR, so a smaller ESR results in
greater power potential.
- Asymmetric Type Design
Asymmetric capacitors are polar in nature because they do not have
a balanced electrode design. In addition, asymmetric capacitors
have a minimum discharge voltage that is half of the maximum
operating voltage. However, discharging to half of the voltage will
release three-quarters of the total energy since:
E = 1/2CV²
Where energy (E) is measured in Joules, Cap (C) in Farads, and
voltage (V) in volts.
E is proportional to V², so the higher voltage levels hold
more potential energy than the lower voltage levels.
The key advantage of asymmetric EC capacitors is they have four to
five times more energy density than symmetric capacitors. This can
result in a much smaller and cheaper solution if energy density is
a key parameter in the capacitor's application.
The packaging of EC capacitors varies greatly. Some designs
utilize a wound-type construction while others use a layered
prismatic construction. When comparing EC capacitors, it is
important to note the rated voltage of the package design. Some
packages consist of one single cell, which limits the voltage of
the package to between 0.8 and 1.6 Vdc (aqueous electrolyte) or 3
to 4 Vdc (non-aqueous). Other package designs are essentially
multiple cells already connected in series and can reach ratings of
up to 600 Vdc.
When voltage is applied across an EC
capacitor that consists of many different cells in series, the
total voltage will divide across each cell. It is critical that the
voltage across any one cell does not exceed its maximum allowed
voltage or a cascading failure could occur. How the voltage divides
depends upon many factors, including the cell design, the cell
capacitance, and the cell DCL (also called self-discharge rate). Of
these three influences, the main cause of voltage imbalance is the
variation in DCL from cell to cell. During use under voltage, the
cells will begin self-discharging at different rates causing
imbalances that add up over repeated charge and discharge cycles.
This voltage imbalance problem must be corrected with external
circuitry that keeps the voltage equal across the cells.
Asymmetric EC Capacitors
Voltage balancing is not a concern with asymmetrical EC
capacitors. One of the electrodes in each cell is a
very-high-capacitance electrode that keeps the voltage divided
evenly among the cells, countering the DCL discharge effect.
Symmetric EC Capacitors
Symmetric capacitors require active or passive voltage balancing
to prevent the failure of the capacitor due to an over-voltaged
cell. Some manufacturers will build in balancing resistors with
their capacitors while other manufacturers require the user to
design and add their own balancing circuits.
Passive balancing uses simple electronic components such as
resistors, which offer a low-cost and simple design solution.
However, adding resistors in parallel with the cells causes a
higher self-discharge rate (high DCL). Passively balanced
capacitors will discharge to zero volts in a matter of hours or
days. This could be a problem in low-power applications or those
applications where the capacitor needs to be connected in parallel
to a battery.
Active balancing uses smart electronics such as ICs or
microprocessors. EC capacitors with active balancing circuitry
could provide a good self-discharge rate, but the cost and
circuitry development time will increase significantly.
Capacitance Frequency Response and
Conventional capacitors are typically read on a RLC bridge, but
there are no commercial RLC bridges that will work with EC
capacitors. However, there are impedance bridges (phase-angle
bridges) that are designed specifically for EC capacitors
). MACCOR and Arbin Instruments are two
manufacturers of these impedance bridges.
Figure 2: A graph generated from an impedance
bridge showing capacitance versus frequency for a 90 Farad EC
capacitor. Notice that the capacitance decreases rapidly between 1
and 10 Hz. This means that these capacitors are not capable of
filtering at frequencies greater than a few Hertz.
Alternatively, since capacitance is directly related to charge
storage capability (C=Q/V), we can use a current charging method to
calculate the capacitance. You can measure time while charging a
capacitor from an initial voltage to a final voltage using a fixed
current. You can then calculate the capacitance using the
C = (T x I) / (Vfinal - Vinitial)
Where capacitance (C) is measured in Farads, time (T) in
seconds, current (I) in amps, and initial voltage
(Vinitial) and final voltage (Vfinal) in
For example, if it takes 185 seconds to charge an EC capacitor
from 8 Vdc to 12 Vdc using 9 amps of current, the capacitance is
(185 x 9)/(12-8) = 416.25 Farads. At this time there are no
standards established for measuring capacitance for EC
Manufacturers will typically rate their products -50 to
+60°C (with minor variations). Compared to conventional
capacitors, EC capacitors have a similar or slightly greater
temperature coefficient depending on the evaluated ratings.
However, compared to batteries, EC capacitors offer an outstanding
temperature response (Figure 3).
Figure 3: As the temperature decreases to
-40°C, the EC capacitor still retains 90% of its room
temperature capacitance. However, the battery quickly looses its
capacity below freezing. The battery is greatly affected since its
charge storage is based on a chemical technology, but the capacitor
uses a true electrostatic method of storing charge that is not
Equivalent Series Resistance
ESR is a measure of a capacitor's internal resistance. The ESR is
measured in Ohms, and can be modeled as a resistor in series with
an ideal capacitor. Like capacitance, there is no standard yet for
measuring the ESR of EC capacitors. However, it is not uncommon for
manufacturers to read the ESR at 1 KHz using standard measuring
The ESR of EC capacitors is higher than that of conventional
capacitors. As the cells are put in series, the ESR of the
individual cells are added together to obtain the overall capacitor
ESR. However, the ESR of EC capacitors is lower than that of
batteries. Figure 4 shows the ESR of an EC capacitor and a
SLI lead acid battery over the temperature range of -40 to
+50°C. Figure 5 shows a normalized ESR versus
temperature curve for a typical EC capacitor. Not only does the
capacity of the battery decrease with lower temperatures, but its
power capability is also severely diminished. This is one reason
why there is so much interest in using EC capacitors for engine
starting applications, especially in cold weather areas.
Figure 4: As the temperature approaches -40°C,
the ESR of the battery increases much faster than the capacitor, so
the power capability diminishes more quickly.
Figure 5: This figure shows the change in ESR from
its 20°C value as a function of temperature. For example, if
the ESR of an EC capacitor is 2.0 mOhms at 20°C, it will be
300% greater (6.0 mOhms) at -45°C.
An ideal capacitor will retain its charge forever. However, all
capacitors experience some self-discharge due to flaw sites or
other transfer mechanisms through the dielectric medium. A simple
way to measure this discharge rate is to put a resistor in series
with the EC capacitor and charge up the capacitor using a power
supply. After a period of time, you can measure and convert the
voltage across the resistor to a current reading to see how much
current is still flowing into the capacitor. This measurement of
the capacitor's DC-Leakage (DCL) is typically specified in
microAmperes (µA) or milliAmperes (mA).
The DCL of EC capacitors will age down over time, but because of
their very high capacitance values, it can take hours or days
before the DCL reaches a steady state level. For conventional
electrolytic capacitors, it is standard to measure the current
flowing after five minutes of charging. However, for EC capacitors,
there is no standard time limit, so you must examine the
manufacturer's data for method as well as the reported values.
Another important factor is that the reported DCL may not
include the balancing circuit. If you must add a balancing circuit
to a capacitor, this could significantly increase the
self-discharge rate. The DCL value is critical in applications such
as low-power or battery-assist applications where the capacitor is
in parallel with batteries.
Other EC Capacitor Characteristics
Life tests have shown EC capacitor technology to be very robust.
Most manufacturers will rate the life of the capacitor in terms of
cycle life. One cycle consists of one full charge and one full
discharge. EC capacitors are often rated for a minimum of 100,000
cycles, and manufacturers typically report that no significant
changes are seen after 1 million cycles. This is in contrast to
batteries, which are lucky to see a cycle life greater than 1500
Charging and Discharging
EC capacitors have no inherent limit on the amount of current they
can use for charging or discharging. For example, large EC
capacitors built for engine starting can supply several thousand
Amps for a short period of time. Likewise, you can charge them up
just as quickly. However, the size and amount of current collectors
used in a capacitor design may limit the amount of current that can
be used. An additional benefit of EC capacitors is that as they
fill up with charge, the current automatically begins to decrease.
This eliminates the need for complex charging circuits. This
quick-charging ability is in contrast to batteries, which require
complex charging and monitoring circuits that limit the amount of
current during the charging operation.
Another benefit of the EC capacitor technology is that no
hydrogen is emitted during the charging operation. In addition,
there is no danger of explosion during charging such as can happen
when batteries are over-charged. These last two benefits are
especially applicable in the aerospace and mining industries.
Power Versus Energy Tradeoff
For any electrical energy source, there is a tradeoff between
the amount of power and the energy that it can provide. This means
that the faster you try to pull energy from a device, the less
total energy you will get from it.
Batteries are good examples of this diminishing returns
principle. A battery manufacturer will rate its batteries at
different Ah (Amp-hours) depending on the time period the battery
is discharged. Typically, lead acid batteries are rated at a
20-hour discharge rate. If a battery is rated at 100 Ah with a
20-hour discharge rate, it may only be rated at 80 Ah with a
10-hour discharge rate. It is difficult to determine how batteries
will perform in very high power applications because battery
manufacturers often will not provide this data.
For EC capacitors, manufacturers will provide graphs showing
energy density versus power density for a capacitor design
(Figure 6). The EC capacitor shown in Figure 6a is
optimized for power as shown by the high specific power values. An
EC capacitor that is optimized for energy would have the profile
shown in Figure 6b.
Figure 6: (a) For the capacitor in the graph on the
left, at 500 W/kg, the capacitor can supply 0.9 Wh/kg. However, at
2000 W/kg, the capacitor can only supply 0.3 Wh/kg. (b) The
capacitor in the graph on the right is capable of supplying 7.0
Wh/kg at 50 W/kg. EC capacitors optimized for energy can supply up
to between 10 and 12 Wh/kg for low-power applications.
Applications that can benefit from EC Capacitors include medical
(x-ray and MRI), welding (spot and contact), audio line stiffening,
actuators, large electric motor starting, and power quality such as
UPS systems (initial pulse powernot battery replacement).
- EC vs. Aluminum (Al) Electrolytic Capacitors
Al electrolytic capacitors have excellent pulse power
characteristics as they can supply power in the microsecond
timeframe. However, large Al electrolytic capacitors are
reluctantly used by system designers because of their poor energy
versus cost ratio. Al electrolytic capacitors currently cost $200-
to $400-per-Farad. EC capacitors, however, cost $20- to less than
$1-per-Farad, giving system designers a new low-cost option for
pulse power applications.
EC capacitors do offer a higher ESR than Al electrolytic
capacitors. While Al electrolytic capacitors can supply power in
the microsecond timeframe, EC capacitors show better
characteristics in the nanoseconds to seconds timeframe. If power
is needed in the microsecond timeframe, one option is to combine a
smaller and cheaper Al electrolytic capacitor with an EC capacitor
to optimize the power versus cost trade-off.
- EC Capacitors vs. Batteries
Even though batteries show poor performance in pulse power
applications, they have been extensively used due to their low-cost
and the lack of options for designers. They have high internal
resistance (ESR), so in order to get the energy out quickly enough,
batteries must be vastly over-sized. In addition, high current
levels severely stress batteries, shortening their useful life.
Batteries also have a poor cycle life performance (1200 to 1500
cycles) and some battery technologies also require frequent
For pulse power applications requiring pulse power durations of a
few seconds or less, EC capacitors will provide a much smaller and
lighter solution. The 100,000+ cycle-life performance and zero
maintenance will significantly reduce the replacement and
maintenance costs. If the application requires a time duration of
greater than a few seconds, then the designer should combine EC
capacitors with batteries so that the power versus energy needs of
the system are optimized. The EC capacitor can supply the initial
power pulse and then let the batteries take over for the long-term
energy needs. The advantages include a much smaller and lighter
solution, longer battery life, and long-term cost savings.
Engine starting is a specialized pulse power application. The
advantages of using an EC capacitor in conjunction with a battery
for starting a vehicle include less space and weight, greatly
improved cold weather performance, improved breakaway torque,
increased battery life, and long-term cost savings.
Electric Vehicle Regenerative Braking/Acceleration
By using regenerative breaking in electric vehicles, a more
efficient overall power system is achieved since the braking energy
is recaptured. Batteries are not capable of handling the full
in-rush of current. By using EC capacitors, the vehicle can collect
and use the full braking energy upon acceleration resulting in a
smaller, more efficient system.
Although EC capacitors excel in pulse power situations, EC
capacitors are also finding use in energy storage applications. If
an application requires only a few seconds of energy, EC capacitors
can provide a smaller and cheaper solution because batteries are
usually severely over-sized for this short time span. In addition,
some industries such as aerospace and mining have very high
maintenance costs and do not like the very small explosion
potential of batteries. As the price of EC capacitors continues to
decline due to economies of scale, this technology will push into
other energy storage areas now serviced by batteries. Critical UPS
systems are one possible area as batteries can fail without warning
whereas you can monitor EC capacitors to tell when they are
approaching the end of their life.
One day, material handling vehicles and electric vehicles may use
EC capacitors rather than batteries. EC capacitors in traction
applications offer the benefits of simple on-board charging, quick
opportunity recharging, higher cycle life, maintenance-free 24-hour
operation, and superior cold weather performance. However, cost is
a main driver for this application.