SECTION 4:

ULTRACAPACITORS
ESE 471 – Energy Storage Systems
2 Introduction

K. Webb ESE 471

 Ultracapacitors
3

 Capacitors are electrical energy storage devices

 Energy is stored in an electric field
 Advantages of capacitors for energy storage
 High specific power
 High efficiency
 Equal charge and discharge rates
 Long lifetime
 Disadvantages of capacitors for energy storage
 Low specific energy

 Ultracapacitors (or supercapacitors) are variations of

traditional capacitors with significantly improved specific
energy
 Useful in high-power energy-storage applications
K. Webb ESE 471
 Ultracapacitors – Ragone Plot
4

K. Webb ESE 471

 Ultracapacitors - Applications
5

 Ultracapacitors are useful in relatively high-power, low-energy applications

 They occupy a similar region in the Ragone plane as flywheels

 Energy recovery and regenerative braking applications

 Cars
 EV, HEV, ICE (e.g. Mazda 6 i-ELOOP)
 Buses
 Trains
 Cranes
 Elevators

 Uninterruptible power supply (UPS) applications

 Fast-responding, short-term power until generators take over

 Wind turbine pitch control

 Put turbine blades in safe position during loss of power

K. Webb ESE 471

6 Capacitor Fundamentals

K. Webb ESE 471

 Fluid Capacitor
7

 Consider the following device:

 Two rigid hemispherical shells
 Separated by an impermeable
elastic membrane
 Modulus of elasticity, 𝜆𝜆
 Area, 𝐴𝐴
 Incompressible fluid
 External pumps set
pressure or flow rate at
each port
 Total volume inside shell is
constant
 Volume on either side of the
membrane may vary
K. Webb ESE 471
 Fluid Capacitor – Equilibrium
8

 Equal pressures
Δ𝑃𝑃 = 𝑃𝑃1 − 𝑃𝑃2 = 0
 No fluid flow
𝑄𝑄1 = 𝑄𝑄2 = 0
 Membrane does not
deform
 Equal volume on
each side
𝑉𝑉
𝑉𝑉1 = 𝑉𝑉2 =
2
K. Webb ESE 471
 Fluid Capacitor – 𝑃𝑃1 > 𝑃𝑃2
9

 Pressure differential
Δ𝑃𝑃 = 𝑃𝑃1 − 𝑃𝑃2 > 0
 Membrane deforms
 Volume differential
Δ𝑉𝑉 = 𝑉𝑉1 − 𝑉𝑉2 > 0
 Transient flow as
membrane stretches,
but...
 No steady-state flow
 As 𝑡𝑡 → ∞
𝑄𝑄1 = 𝑄𝑄2 = 0

K. Webb ESE 471

 Fluid Capacitor – 𝑃𝑃1 < 𝑃𝑃2
10

 Pressure differential
Δ𝑃𝑃 = 𝑃𝑃1 − 𝑃𝑃2 < 0
 Volume differential
Δ𝑉𝑉 = 𝑉𝑉1 − 𝑉𝑉2 < 0

 Δ𝑉𝑉 proportional to:

 Pressure differential
 Physical properties, 𝜆𝜆, 𝐴𝐴
 Total volume remains constant
𝑉𝑉1 + 𝑉𝑉2 = 𝑉𝑉
 Again, no steady-state flow

K. Webb ESE 471

 Fluid Capacitor – Constant Flow Rate
11

 Constant flow rate forced into port 1

𝑄𝑄1 ≠ 0

 Incompressible, so flows are equal and

opposite
𝑄𝑄1 = 𝑄𝑄2

 Volume on each side

proportional to time
𝑉𝑉
𝑉𝑉1 = + 𝑄𝑄1 ⋅ 𝑡𝑡
2
𝑉𝑉 𝑉𝑉
𝑉𝑉2 = − 𝑄𝑄2 ⋅ 𝑡𝑡 = − 𝑄𝑄1 ⋅ 𝑡𝑡
2 2
 Volume differential proportional to time
Δ𝑉𝑉 = 𝑉𝑉1 − 𝑉𝑉2 = 2𝑄𝑄1 ⋅ 𝑡𝑡
K. Webb ESE 471
 Fluid Capacitor – Capacitance
12

 Define a relationship between

differential volume and pressure
 Capacitance
Δ𝑉𝑉
𝐶𝐶 =
Δ𝑃𝑃

 Intrinsic device
property
 Determined by physical
parameters:
 Membrane area, 𝐴𝐴
 Modulus of elasticity, 𝜆𝜆

K. Webb ESE 471

 Fluid Capacitor – DC vs. AC
13

 In steady-state (DC), no fluid flows

𝑄𝑄1 = 𝑄𝑄2 = 0

 Consider sinusoidal Δ𝑃𝑃 (AC):

Δ𝑃𝑃 = 𝑃𝑃 sin 𝜔𝜔𝜔𝜔

 Resulting flow rate is

proportional to:
 Rate of change of
differential pressure
 Capacitance
𝑑𝑑𝑑𝑑
𝑄𝑄1 = 𝑄𝑄2 = 𝐶𝐶 = 𝜔𝜔𝜔𝜔𝜔𝜔 cos 𝜔𝜔𝜔𝜔
𝑑𝑑𝑑𝑑

K. Webb ESE 471

 Fluid Capacitor – Time-Varying Δ𝑃𝑃
14

 Equal and opposite flow at both ports

𝑄𝑄1 = 𝑄𝑄2

 Not the same fluid flowing at both ports

 Fluid cannot permeate the membrane

 Fluid appears to flow through

the device
 Due to the displacement of the
membrane
 A displacement flow

 The faster Δ𝑃𝑃 changes, the higher the flow rate

𝑄𝑄 ∝ 𝜔𝜔

 The larger the capacitance, the higher the flow rate

𝑄𝑄 ∝ 𝐶𝐶
K. Webb ESE 471
 Fluid Capacitor – Changing Δ𝑃𝑃
15

 A given Δ𝑃𝑃 corresponds to a particular membrane

displacement
 Forces must balance
 Membrane cannot instantaneously
jump from one displacement to
another
 Step change in
displacement/pressure is
impossible
 Would require an infinite
flow rate
 Pressure across a fluid capacitor cannot
change instantaneously

K. Webb ESE 471

 Fluid Capacitor – Energy Storage
16

 Stretched membrane stores energy

 Potential energy
 Stored energy proportional to:
 Δ𝑃𝑃
 Δ𝑉𝑉
 Energy released as
membrane returns
 𝑃𝑃 and 𝑄𝑄 are supplied

 Not a real device, but analogous to

other potential energy storage methods
 PHES
 CAES
 Electrical capacitors

K. Webb ESE 471

 Electrical Capacitor
17

 In the electrical domain, our “working fluid” is

positive electrical charge
 In either domain, we have a potential-driven flow

Fluid Domain Electrical Domain

Pressure – P Voltage – V

Volumetric flow rate – Q Current – I

Volume – V Charge – Q

K. Webb ESE 471

 Electrical Capacitor
18

 Parallel-plate capacitor
 Parallel metal plates
 Separated by an insulator
 Applied voltage creates
charge differential
 Equal and opposite charge
𝑄𝑄1 = −𝑄𝑄2
 Zero net charge
 Equal current
𝐼𝐼1 = 𝐼𝐼2
 What flows in one side
 Schematic symbol:
flows out the other
 Units: Farads (F)
K. Webb ESE 471
 Electrical Capacitor – Electric Field
19

 Charge differential results

in an electric field, 𝑬𝑬, in
the dielectric
 Units: 𝑉𝑉/𝑚𝑚
 |𝑬𝑬| is inversely proportional
to dielectric thickness, 𝑑𝑑
 Above some maximum
electric field strength,
dielectric will break down
 Conducts electrical current
 Maximum capacitor voltage
rating
K. Webb ESE 471
 Electrical Capacitor - Capacitance
20

 Capacitance
 Ratio of charge to
voltage
𝑄𝑄
𝐶𝐶 =
𝑉𝑉
 Intrinsic device property
 Proportional to physical
parameters:
 Dielectric thickness, 𝑑𝑑
 Dielectric constant, 𝜀𝜀
 Area of electrodes, 𝐴𝐴
K. Webb ESE 471
 Parallel-Plate Capacitor
21

 Capacitance
𝜀𝜀𝜀𝜀
𝐶𝐶 =
𝑑𝑑
 𝜀𝜀: dielectric permittivity
 𝐴𝐴: area of the plates
 𝑑𝑑: dielectric thickness

 Capacitance is maximized
by using:
 High-dielectric-constant
materials
 Thin dielectric
 Large-surface-area plates

K. Webb ESE 471

 Capacitors – Voltage and Current
22

 Current through a capacitor is

proportional to
 Capacitance
 Rate of change of the voltage

𝑑𝑑𝑑𝑑
𝑖𝑖 𝑡𝑡 = 𝐶𝐶
𝑑𝑑𝑑𝑑
 Voltage across capacitor results from an accumulation
of charge differential
 Capacitor integrates current
1
𝑣𝑣 𝑡𝑡 = ∫ 𝑖𝑖 𝑡𝑡 𝑑𝑑𝑑𝑑
𝐶𝐶
K. Webb ESE 471
 Voltage Change Across a Capacitor
23

 For a step change in voltage,

𝑑𝑑𝑑𝑑
=∞
𝑑𝑑𝑑𝑑
 The corresponding current would be infinite
 Voltage across a capacitor cannot change
instantaneously
 Current can change instantaneously, but voltage is the
integral of current
𝑡𝑡0 +Δ𝑡𝑡
lim ΔV = lim � 𝑖𝑖 𝑡𝑡 𝑑𝑑𝑑𝑑 = 0
Δt→0 Δt→0 𝑡𝑡
0

K. Webb ESE 471

 Capacitors – Open Circuits at DC
24

 Current through a capacitor is proportional to the

time rate of change of the voltage across the
capacitor
𝑑𝑑𝑑𝑑
𝑖𝑖 𝑡𝑡 = 𝐶𝐶
𝑑𝑑𝑑𝑑
 A DC voltage does not change with time, so
𝑑𝑑𝑑𝑑
= 0 and 𝑖𝑖 𝑡𝑡 = 0
𝑑𝑑𝑑𝑑

 A capacitor is an open circuit at DC

K. Webb ESE 471

 Capacitors in Parallel
25

 Total charge on two parallel

capacitors is
𝑄𝑄 = 𝑄𝑄1 + 𝑄𝑄2
𝑄𝑄 = 𝐶𝐶1 𝑉𝑉 + 𝐶𝐶2 𝑉𝑉
𝑄𝑄 = 𝐶𝐶1 + 𝐶𝐶2 𝑉𝑉
𝑄𝑄 = 𝐶𝐶𝑒𝑒𝑒𝑒 𝑉𝑉

 Capacitances in parallel add

𝐶𝐶𝑒𝑒𝑒𝑒 = 𝐶𝐶1 + 𝐶𝐶2

K. Webb ESE 471

 Capacitors in Series
26

 Total voltage across the series

combination is
𝑉𝑉 = 𝑉𝑉1 + 𝑉𝑉2
𝑄𝑄 𝑄𝑄
𝑉𝑉 = +
𝐶𝐶1 𝐶𝐶2
1 1 𝑄𝑄
𝑉𝑉 = 𝑄𝑄 + =
𝐶𝐶1 𝐶𝐶2 𝐶𝐶𝑒𝑒𝑒𝑒

 The inverses of capacitors in series add

−1
1 1 𝐶𝐶1 𝐶𝐶2
𝐶𝐶𝑒𝑒𝑒𝑒 = + =
𝐶𝐶1 𝐶𝐶2 𝐶𝐶1 + 𝐶𝐶2

K. Webb ESE 471

 Constant Current Onto a Capacitor
27

 Capacitor voltage
increases linearly for
constant current
𝐼𝐼
𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 =
𝐶𝐶

𝐼𝐼 𝑡𝑡−𝑡𝑡0
𝑣𝑣 𝑡𝑡 = , 𝑡𝑡 ≥ 𝑡𝑡0
𝐶𝐶

K. Webb ESE 471

 Electrical Capacitor – Energy Storage
28

 Capacitors store
electrical energy
 Energy stored in the
electric field
 Stored energy is
proportional to:
 Voltage
 Charge differential
1 1 2 1 𝑄𝑄2
𝐸𝐸 = 𝑄𝑄𝑄𝑄 = 𝐶𝐶𝑉𝑉 =
2 2 2 𝐶𝐶
 Energy released as E-field
collapses
 𝑉𝑉 and 𝐼𝐼 supplied
K. Webb ESE 471
 Energy Storage – Example
29

 A capacitor is charged to 100 V

 The stored energy will be used to lift a 1000 kg elevator car 10 stories (35 m)
 Determine the required capacitance

 The required energy is

𝑚𝑚
𝐸𝐸 = 𝑚𝑚𝑚𝑚𝑚 = 1000 𝑘𝑘𝑘𝑘 ⋅ 9.81 2 ⋅ 35 𝑚𝑚
𝑠𝑠
𝐸𝐸 = 343.4 𝑘𝑘𝑘𝑘

 Energy stored on the capacitor is

1 2
𝐸𝐸 = 𝐶𝐶 100 𝑉𝑉
2
 The required capacitance is
2 ⋅ 343.4 𝑘𝑘𝑘𝑘
𝐶𝐶 = = 68.7 𝐹𝐹
100 𝑉𝑉 2

K. Webb ESE 471

30 Ultracapacitors

K. Webb ESE 471

 Ultracapacitors - Introduction
31

 Energy stored by a capacitor

1 2
𝐸𝐸 = 𝐶𝐶𝑉𝑉
2
 Would like to maximize capacitance in order to maximize energy storage
 Recall the capacitance of a parallel-plate capacitor
𝜀𝜀𝜀𝜀
𝐶𝐶 =
𝑑𝑑
 To increase capacitance:
 Use a higher-permittivity dielectric
 Increase surface area of the plates
 Decrease dielectric thickness
 Traditional capacitors do all of these things
 𝜀𝜀 limited by available materials and dielectric strength
 𝐴𝐴 limited by practical overall device size
 𝑑𝑑 limited by dielectric breakdown field strength

K. Webb ESE 471

 Traditional Capacitors – Construction
32

 Let’s take a look at the construction of two high-capacitance

traditional capacitors
 Aluminum electrolytic
 Tantalum electrolytic

 Aluminum electrolytic capacitor:

K. Webb ESE 471

 Traditional Capacitors – Construction
33

 Tantalum electrolytic capacitor:

 In both of these types of capacitors, efforts are made to

maximize 𝐴𝐴 and minimize 𝑑𝑑
 But, a physical dielectric layer of non-zero thickness is used

K. Webb ESE 471

 Ultracapacitors
34

 In a previous example, we found we needed a

capacitance of 68.7 F
 Impractically large for a traditional capacitor
 Not so for an ultracapacitor

 Ultracapacitors or supercapacitors achieve very high

capacitance values by eliminating the solid dielectric
layer of traditional capacitors
 Energy is stored in an E-field
 Not in a dielectric layer
 In an electric double layer (Helmholtz double layer)
 Electric double-layer capacitors (EDLC)

K. Webb ESE 471

 Ultracapacitors
35

 Electric double-layer capacitor

K. Webb ESE 471

 Ultracapacitors
36

 Electrodes are rough and

porous
 Surface area is increased
 Activated charcoal
 Aerogel
 No charge transfer between
the electrolyte and the
electrode
 Separator is permeable
 Mechanical separation Source: Yves Brunet, Energy Storage

preventing contact between

electrodes
 Thickness of double layers is on the molecular scale
K. Webb ESE 471
 Ultracapacitors
37

 Two double layers

 Two capacitors in series

 Capacitance values in the range of 1 … 1000s of farads are common

 Ultracapacitors are polarized
 Positive electrode must be kept at a higher potential

 Maximum voltage determined by the electrolyte dissociation voltage

 Typically ~2.5 V
 For higher-voltage operation, multiple ultracapacitors are connected in series

K. Webb ESE 471

38 Equivalent Circuit Model

K. Webb ESE 471

 Equivalent Circuit Model
39

 Ultracapacitor equivalent circuit model

 Rs: equivalent series resistance (ESR)

 Primarily due to ionic conduction in the electrolyte
 C0: primary capacitance of the ultracapacitor
 Cv: voltage-dependent capacitance
 Associated with diffusion layers near the double layers
 Cv = k⋅v
 Rleak: leakage resistance
 Typically specified as a leakage current at Vmax
 R1, C1, … Rn, Cn: distributed resistance and capacitance of the porous electrodes
 Models multiple time constants

K. Webb ESE 471

 Equivalent Circuit Model
40

 We will typically simplify this model significantly

 Account for only capacitance and ESR
 Typical ESR values: 0.5 mΩ … 500 mΩ

 Account for leakage resistance, Rleak, when appropriate

 Self-discharge
 Typical leakage resistance: 100 Ω … 100 kΩ
 Typical leakage currents: 10𝜇𝜇A … 10 mA
K. Webb ESE 471
41 Charging and Discharging

K. Webb ESE 471

 Charging and Discharging
42

 The voltage seen across a capacitor is proportional to the

stored charge differential
𝑄𝑄
𝑉𝑉 =
𝐶𝐶
 So, unlike batteries, capacitor voltage does not remain
constant as a capacitor discharges
 Power electronic circuitry generally required to interface
between ultracapacitors and load
 DC-DC converters
 Inverters – DC-AC and AC-DC converters
 Interface circuitry also provides charge/discharge control
 Current/power control
K. Webb ESE 471
 Charging and Discharging
43

 Two primary modes of charging/discharging

 Constant current
 Constant power

 Unlike batteries, capacitors can be charged and

discharged at the same rates
 Constant-current charging is simple
 Both in terms of circuitry and analysis/design
 Constant-power charging useful in many applications,
such as regenerative braking
 Charging while drawing constant power from the vehicle
 Discharging while supplying constant power to the vehicle

K. Webb ESE 471

 Constant-Current Charging
44

 Voltage drop across

Rs during
charge/discharge
 Constant rate of
voltage change
 Power varies
K. Webb ESE 471
 Constant-Power Charging
45

 Varying rate of voltage

change
 Current varies depending
on state of charge
 Higher current at lower
state of charge
 Lower current near full
charge
K. Webb ESE 471
46 Cell Balancing

K. Webb ESE 471

 Cell Balancing
47

 Typically, 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 = 2.5 𝑉𝑉 … 3.0 𝑉𝑉

 Series-connected cells provide higher voltages
 Consider a series connection of four cells
 Equal charge differential, ±𝑄𝑄, on each cell
 The voltage across each capacitor is
𝑄𝑄 𝑄𝑄 𝑄𝑄 𝑄𝑄
𝑉𝑉1 = , 𝑉𝑉2 = , 𝑉𝑉3 = , 𝑉𝑉4 =
𝐶𝐶1 𝐶𝐶2 𝐶𝐶3 𝐶𝐶4

 Nominally, all capacitors are equal

𝐶𝐶1 = 𝐶𝐶2 = 𝐶𝐶3 = 𝐶𝐶4 = 𝐶𝐶

 Nominally, all voltages are equal

𝑄𝑄
𝑉𝑉1 = 𝑉𝑉2 = 𝑉𝑉3 = 𝑉𝑉4 = = 2.5V
𝐶𝐶
 But, capacitances may vary by as much as ±20%

K. Webb ESE 471

 Cell Balancing
48

 Consider the following scenario:

 Total equivalent capacitance
𝐶𝐶𝑒𝑒𝑒𝑒 = 0.24𝐶𝐶
 Stored charge
𝑄𝑄 = 10 𝑉𝑉 ⋅ 0.24𝐶𝐶
 Now, cell voltages are imbalanced
𝑄𝑄 10 𝑉𝑉 ⋅ 0.24𝐶𝐶
𝑉𝑉1 = 𝑉𝑉2 = = = 2 𝑉𝑉
1.2𝐶𝐶 1.2𝐶𝐶
𝑄𝑄 10 𝑉𝑉 ⋅ 0.24𝐶𝐶
𝑉𝑉3 = 𝑉𝑉4 = = = 3 𝑉𝑉
0.8𝐶𝐶 0.8𝐶𝐶
 If 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 = 2.5 𝑉𝑉, then
 𝐶𝐶1 and 𝐶𝐶2 are underutilized
 𝐶𝐶3 and 𝐶𝐶4 are overstressed

K. Webb ESE 471

 Cell Balancing
49

 Cell balancing circuitry

 Safely utilize each cell’s storage
capacity
 Two balancing approaches:
 Resistive balancing
 Resistors placed in parallel with the cells
 Slow – not for high-duty-cycle
applications
 Active balancing
 Cell voltages monitored and electronic
switches balance voltages
 Fast – good for high-duty-cycle
applications

K. Webb ESE 471

50 Efficiency

K. Webb ESE 471

 Ultracapacitors – Efficiency
51

 Ultracapacitors have small, but non-zero, ESR

 They are lossy devices
 Not all input energy is available for use
 Efficiency is less than 100%
 We will define round-trip efficiency as the efficiency
through an entire charge/discharge cycle
 Ratio of output energy to input energy
𝐸𝐸𝑜𝑜𝑜𝑜𝑜𝑜
𝜂𝜂𝑟𝑟𝑟𝑟 = ⋅ 100% (1)
𝐸𝐸𝑖𝑖𝑖𝑖

 Efficiency depends on how a capacitor is used

 Rate of charge/discharge
 Depth of discharge

K. Webb ESE 471

 Ultracapacitors – Efficiency
52

 Energy stored by a capacitor is proportional to the capacitor

voltage squared
1
𝐸𝐸𝑐𝑐 = 𝐶𝐶𝑉𝑉 2 (2)
2

 Capacitor’s effectiveness at storing energy depends on its

state of charge (SOC)
 Energy stored more quickly at high SOC
 Energy stored more slowly at low SOC

 Loss in ESR depends on the current

 Therefore, instantaneous efficiency, 𝜂𝜂 𝑡𝑡 , varies with SOC
 Total round-trip efficiency depends on depth of discharge
 Ultracapacitors are typically not discharged completely
K. Webb ESE 471
 Discharge Factor
53

 Discharge factor
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
𝑑𝑑 = (3)
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚

 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 : voltage at the lowest allowable SOC

 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 : maximum allowable (fully-charged) capacitor voltage

 We’ll now examine the round-trip efficiency for

capacitors operated at constant current and at
constant power

K. Webb ESE 471

 Efficiency – Constant Current
54

 For a capacitor operating at a non-zero discharge factor, only some

of the stored energy is usable
 Usable energy
1 2 1 2
𝐸𝐸𝑢𝑢 = 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 − 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
2 2
1 2 1 2
𝐸𝐸𝑢𝑢 = 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 − 𝐶𝐶 𝑑𝑑𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
2 2
1 2
𝐸𝐸𝑢𝑢 = 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑 2 (4)
2

 Power dissipated in the ESR at constant current, 𝐼𝐼, is

𝑃𝑃𝑅𝑅 = 𝐼𝐼2 𝑅𝑅𝑠𝑠
 Energy lost in the ESR is
𝐸𝐸𝑅𝑅 = 𝐼𝐼2 𝑅𝑅𝑠𝑠 ⋅ 𝑡𝑡

K. Webb ESE 471

 Charging efficiency – Constant Current
55

 During the charging cycle, the efficiency is

1 2 2
𝐸𝐸𝑢𝑢 2
𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑
𝜂𝜂1 = = 1
𝐸𝐸𝑢𝑢 +𝐸𝐸𝑅𝑅 𝐶𝐶𝑉𝑉 2
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑 2 +𝐼𝐼2 𝑅𝑅𝑠𝑠 ⋅𝑡𝑡𝑐𝑐

where 𝑡𝑡𝑐𝑐 is the duration of the charging cycle

 We can solve for 𝑡𝑡𝑐𝑐 as follows
𝐼𝐼⋅𝑡𝑡𝑐𝑐 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 − 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 = → 𝑡𝑡𝑐𝑐 =
𝐶𝐶 𝐼𝐼
 The efficiency then becomes
1
𝐶𝐶𝑉𝑉 2
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑 2
𝜂𝜂1 = 1 2 2 +𝐼𝐼𝑅𝑅 𝐶𝐶𝑉𝑉
2
𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑 𝑠𝑠 𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑

1
𝑉𝑉
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑 2
𝜂𝜂1 = 1 (5)
𝑉𝑉
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑2 +𝐼𝐼𝑅𝑅𝑠𝑠 1−𝑑𝑑

K. Webb ESE 471

 Round-Trip Efficiency – Constant Current
56

 Similar loss is incurred in the ESR during discharge

 Energy output is the stored energy minus resistive loss
 Round-trip efficiency is
𝐸𝐸𝑜𝑜𝑜𝑜𝑜𝑜 𝐸𝐸𝑢𝑢 −𝐸𝐸𝑅𝑅
𝜂𝜂𝑟𝑟𝑟𝑟 = =
𝐸𝐸𝑖𝑖𝑖𝑖 𝐸𝐸𝑢𝑢 +𝐸𝐸𝑅𝑅

1
𝐶𝐶𝑉𝑉 2
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑2 −𝐼𝐼𝑅𝑅𝑠𝑠 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 1 2
2
𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑2 +𝐼𝐼𝑅𝑅𝑠𝑠 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1−𝑑𝑑

1
𝑉𝑉 1−𝑑𝑑 2 −𝐼𝐼𝑅𝑅𝑠𝑠 1−𝑑𝑑
2 𝑚𝑚𝑚𝑚𝑚𝑚
𝜂𝜂𝑟𝑟𝑟𝑟 = 1 (6)
𝑉𝑉
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑 2 +𝐼𝐼𝑅𝑅𝑠𝑠 1−𝑑𝑑

K. Webb ESE 471

 Instantaneous Efficiency – Constant Current
57

 At lower SOC:
 Low rate of
energy storage
 Low efficiency

 At higher SOC:
 Higher rate of
energy storage
 Higher efficiency

K. Webb ESE 471

 Efficiency vs. Current – Constant Current
58

 As current increases, loss in ESR increases

K. Webb ESE 471

 Efficiency vs. Discharge Factor – Constant Current
59

 Greater depth-of-discharge corresponds to lower efficiency

K. Webb ESE 471

 Efficiency – Constant Power
60

 For constant-power charging/discharging, we start with the same

expression for efficiency
𝐸𝐸𝑢𝑢 − 𝐸𝐸𝑅𝑅
𝜂𝜂𝑟𝑟𝑟𝑟 =
𝐸𝐸𝐸𝐸 + 𝐸𝐸𝑅𝑅
 Usable energy is the same
1 2
𝐸𝐸𝑢𝑢 = 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑 2
2
 But, since current is now time varying, the energy lost in the
resistance (in one direction) is
𝑡𝑡𝑐𝑐
𝐸𝐸𝑅𝑅 = � 𝑖𝑖 2 𝑡𝑡 𝑅𝑅𝑆𝑆 𝑑𝑑𝑑𝑑
0

 Things get a bit more complicated, as we now need to solve a

differential equation to determine the capacitor voltage
K. Webb ESE 471
 Efficiency – Constant Power
61

 The input to the capacitor is now constant power

 Write a power-balance equation
𝑃𝑃 − 𝑖𝑖 2 𝑡𝑡 ⋅ 𝑅𝑅𝑠𝑠 − 𝑖𝑖 𝑡𝑡 𝑣𝑣𝑐𝑐 𝑡𝑡 = 0 i

 Current is given by +
Rs

𝑑𝑑𝑣𝑣𝑐𝑐 +
𝑖𝑖 𝑡𝑡 = 𝐶𝐶 v
C vc
𝑑𝑑𝑑𝑑
–
–
 The power balance becomes
2
2 𝑑𝑑𝑣𝑣𝑐𝑐 𝑑𝑑𝑣𝑣𝑐𝑐
𝑅𝑅𝑠𝑠 𝐶𝐶 + 𝐶𝐶 𝑣𝑣 𝑡𝑡 − 𝑃𝑃 = 0 (7)
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 𝑐𝑐

 An ordinary differential equation in quadratic form

K. Webb ESE 471

 Efficiency – Constant Power
62

 Applying the quadratic formula to (7), we get

𝑑𝑑𝑣𝑣𝑐𝑐 −𝐶𝐶𝑣𝑣𝑐𝑐 𝑡𝑡 ± 𝐶𝐶 2 𝑣𝑣𝑐𝑐2 𝑡𝑡 +4𝑅𝑅𝑠𝑠 𝐶𝐶 2 𝑃𝑃
=
𝑑𝑑𝑑𝑑 2𝑅𝑅𝑠𝑠 𝐶𝐶 2

 Simplifying, and keeping only the valid ‘+’ solution

𝑑𝑑𝑣𝑣𝑐𝑐 −𝑣𝑣𝑐𝑐 𝑡𝑡 + 𝑣𝑣𝑐𝑐2 𝑡𝑡 +4𝑅𝑅𝑠𝑠 𝑃𝑃
= (8)
𝑑𝑑𝑑𝑑 2𝑅𝑅𝑠𝑠 𝐶𝐶

 For discharging this becomes

𝑑𝑑𝑣𝑣𝑐𝑐 −𝑣𝑣𝑐𝑐 𝑡𝑡 + 𝑣𝑣𝑐𝑐2 𝑡𝑡 −4𝑅𝑅𝑠𝑠 𝑃𝑃
= (9)
𝑑𝑑𝑑𝑑 2𝑅𝑅𝑠𝑠 𝐶𝐶
 We don’t have a nice closed-form solutions to (8) or (9),
but we can solve them numerically
K. Webb ESE 471
 Instantaneous Efficiency – Constant Power
63

 At lower SOC:
 High current
2
 High 𝐼𝐼 𝑅𝑅 loss

 Low efficiency

 At higher SOC:
 Lower current
2
 Lower 𝐼𝐼 𝑅𝑅 loss

 Higher efficiency

K. Webb ESE 471

 Efficiency vs. Current – Constant Power
64

 As power increases, current increases and resistive loss

increases

K. Webb ESE 471

 Efficiency vs. Discharge Factor – Constant Power
65

 Greater depth-of-discharge corresponds to lower efficiency

K. Webb ESE 471

 Approximate Efficiency – Constant Power
66

 Sometimes we may want a quick way to approximate efficiency for

constant-power charging/discharging
 Calculate an average current
 Calculate efficiency as you would for constant-current charging/discharging
 Approximate average current as the average of the maximum and
minimum currents
𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 + 𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚
𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎 ≈
2
where
𝑃𝑃 𝑃𝑃
𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 = and 𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 =
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚

 Then, efficiency is approximately given by

1
𝑉𝑉
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑 2 −𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎 𝑅𝑅𝑠𝑠 1−𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 ≈ 1 (10)
𝑉𝑉
2 𝑚𝑚𝑚𝑚𝑚𝑚
1−𝑑𝑑 2 +𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎 𝑅𝑅𝑠𝑠 1−𝑑𝑑

K. Webb ESE 471

67 Series-Connected Capacitor Cells

K. Webb ESE 471

 Series-Connected Capacitor Cells
68

 To achieve higher working voltages, multiple capacitor cells

are connected in series
 How does this effect
 Energy storage?
 Efficiency?
 Consider a series connection of 𝑁𝑁 cells, each with a
capacitance of 𝐶𝐶0 and a maximum voltage of 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
 Equivalent circuit model:

K. Webb ESE 471

 Series-Connected Capacitor Cells
69

 Energy stored:
1 𝐶𝐶0 2
𝐸𝐸 = 𝑁𝑁 ⋅ 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
2 𝑁𝑁
1 2
𝐸𝐸 = 𝑁𝑁 ⋅ 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
2
 As expected, this is 𝑁𝑁 times the energy stored in a single cell
 For a discharge factor of 𝑑𝑑, the usable stored energy is
1 2 1 2
𝐸𝐸𝑢𝑢 = 𝑁𝑁 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 − 𝑁𝑁 𝐶𝐶0 𝑑𝑑𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
2 2
1 2
𝐸𝐸𝑢𝑢 = 𝑁𝑁 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 (11)
2

K. Webb ESE 471

 Cells in Series – Constant Current
70

 Energy lost in the resistance during charging (or discharging):

𝐸𝐸𝑅𝑅 = 𝐼𝐼2 𝑁𝑁𝑅𝑅0 𝑡𝑡𝑐𝑐

 For constant-current operation

𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 − 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 = 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑 = 𝑁𝑁𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑
𝐼𝐼 𝑡𝑡𝑐𝑐 𝐼𝐼 𝑡𝑡𝑐𝑐
𝑁𝑁𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 = =
𝐶𝐶 𝐶𝐶0 /𝑁𝑁

 The charging (or discharging) time is

𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑
𝑡𝑡𝑐𝑐 =
𝐼𝐼
 So, losses during charging (or discharging) are
𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑
𝐸𝐸𝑅𝑅 = 𝐼𝐼2 𝑁𝑁𝑅𝑅0
𝐼𝐼
𝐸𝐸𝑅𝑅 = 𝑁𝑁𝑁𝑁𝑅𝑅0 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 (12)

K. Webb ESE 471

 Cells in Series – Constant Current
71

 Round-trip efficiency is
𝐸𝐸𝑢𝑢 − 𝐸𝐸𝑅𝑅
𝜂𝜂𝑟𝑟𝑟𝑟 =
𝐸𝐸𝑢𝑢 + 𝐸𝐸𝑅𝑅

 Using (11) and (12), we get

1 2
𝑁𝑁 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝑁𝑁𝑁𝑁𝑅𝑅0 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 2
1 2
𝑁𝑁 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 + 𝑁𝑁𝑁𝑁𝑅𝑅0 𝐶𝐶0 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑
2
1
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅0 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 =2
1
1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅0 1 − 𝑑𝑑
2 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
 This is, of course, the same as for a single cell, but
 Current can be reduced at a higher maximum voltage
 Efficiency will improve

K. Webb ESE 471

 Cells in Series – Constant Power
72

 For constant-power
operation, we’ll again
investigate numerically
 Round-trip efficiency vs.
# of series-connected
cells:
 Assuming:
 𝐸𝐸 = 100 𝑘𝑘𝑘𝑘
 𝑃𝑃 = 200 𝑊𝑊
 R 0 = 3 mΩ
 𝑑𝑑 = 0.25
 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 2.5 𝑉𝑉

K. Webb ESE 471

73 Capacitor Bank Sizing

K. Webb ESE 471

 Ultracapacitor Sizing
74

 Sizing a capacitor bank involves determining the following parameters

 Stored energy: 𝐸𝐸
 Available power: 𝑃𝑃
 Maximum voltage: 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
 Minimum voltage: 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
 Discharge factor: 𝑑𝑑
 # of cells in series and/or parallel: 𝑁𝑁𝑠𝑠 , 𝑁𝑁𝑝𝑝
 Total capacitance: 𝐶𝐶
 Cell capacitance: 𝐶𝐶0
 Efficiency: 𝜂𝜂
 Sizing procedure depends on which of these parameters are specified
 We’ll outline a procedure assuming the specified requirements are:
 Energy storage
 Power
 Voltage range

K. Webb ESE 471

 Ultracapacitor Sizing Procedure
75

1. Discharge factor
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
𝑑𝑑 =
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
2. Number of series-connected cells
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚
𝑁𝑁𝑠𝑠 =
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
3. Total required capacitance
1 2 2𝐸𝐸
𝐸𝐸 = 𝐶𝐶𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑2 → 𝐶𝐶 = 2
2 𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑2
4. Cell capacitance
𝐶𝐶0 = 𝑁𝑁𝑠𝑠 ⋅ 𝐶𝐶
5. Determine the resulting efficiency using the required power
 Evaluate numerically or approximate
6. Iterate if necessary
 Adjust N as needed

K. Webb ESE 471

 Ultracapacitor Sizing – Example
76

 Size a capacitor bank for an energy recovery system for a tower crane with the
following specifications
 Height: ℎ = 50 𝑚𝑚
 Capacity: 𝑚𝑚 = 5,000 𝑘𝑘𝑘𝑘
 Time to lift max load: 𝑡𝑡𝑑𝑑 = 30 𝑠𝑠
 Let’s assume we have a power conversion system that can operate over the range
of 60 𝑉𝑉𝐷𝐷𝐷𝐷 … 150 𝑉𝑉𝐷𝐷𝐷𝐷 at an efficiency of 𝜂𝜂𝑝𝑝𝑝𝑝𝑝𝑝 = 97%

 The required energy to lift the maximum load is

1 𝑚𝑚 1
𝐸𝐸 = 𝑚𝑚𝑚𝑚𝑚 ⋅ = 5000 𝑘𝑘𝑘𝑘 ⋅ 9.81 ⋅ 50 𝑚𝑚 ⋅
𝜂𝜂𝑝𝑝𝑝𝑝𝑝𝑝 𝑠𝑠 2 0.97

𝐸𝐸 = 2.53 𝑀𝑀𝑀𝑀

 Performing that lift in 30 𝑠𝑠𝑠𝑠𝑠𝑠 corresponds to a required power of

𝐸𝐸 2.53 𝑀𝑀𝑀𝑀
𝑃𝑃 = = = 84.3 𝑘𝑘𝑘𝑘
𝑡𝑡𝑑𝑑 30 𝑠𝑠

K. Webb ESE 471

 Ultracapacitor Sizing – Example
77

 The discharge factor is

𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 60 𝑉𝑉
𝑑𝑑 = = = 0.4
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 150 𝑉𝑉
 The required number of cells in series is
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 150 𝑉𝑉
𝑁𝑁𝑠𝑠 = = = 60
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 2.5 𝑉𝑉
 The total required capacitance is
2𝐸𝐸
𝐶𝐶 = 2
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑 2
2 ⋅ 2.53 𝑀𝑀𝑀𝑀
𝐶𝐶 =
150 𝑉𝑉 2 1 − 0.42
𝐶𝐶 = 267.6 𝐹𝐹
K. Webb ESE 471
 Ultracapacitor Sizing – Example
78

 The capacitance of each individual cell is

𝐶𝐶0 = 𝑁𝑁𝑠𝑠 ⋅ 𝐶𝐶 = 60 ⋅ 267.6 𝐹𝐹
𝐶𝐶0 = 16.1 𝑘𝑘𝑘𝑘

 So, the capacitor bank would consist of sixty 16.1 𝑘𝑘𝑘𝑘 capacitors connected in series
 16.1 𝑘𝑘𝑘𝑘 is a large capacitance – likely unavailable
 Connect multiple capacitors in parallel
 Let’s say we have access to individual capacitor cells with the following
specifications
 𝐶𝐶0 = 3400 𝐹𝐹
 𝑅𝑅0 = 0.28 𝑚𝑚Ω
 Five capacitors in parallel will give
𝐶𝐶0𝑝𝑝 = 𝑁𝑁𝑝𝑝 𝐶𝐶0 = 5𝐶𝐶0 = 17 𝑘𝑘𝑘𝑘
𝑅𝑅0 𝑅𝑅0 0.28 𝑚𝑚Ω
𝑅𝑅0𝑝𝑝 = = = = 56 𝜇𝜇Ω
𝑁𝑁𝑝𝑝 5 5

K. Webb ESE 471

 Ultracapacitor Sizing – Example
79

 Placing 60 of the 5-capacitor groups in series, the total series resistance is

𝑅𝑅𝑠𝑠 = 𝑁𝑁𝑠𝑠 ⋅ 𝑅𝑅0𝑝𝑝 = 60 ⋅ 56 𝜇𝜇Ω = 3.36 𝑚𝑚Ω

 For constant-power charge/discharge, we can approximate efficiency

 The maximum current is
𝑃𝑃 84.3 𝑘𝑘𝑘𝑘
𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 = = = 1.4 𝑘𝑘𝑘𝑘
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 60 𝑉𝑉

 The minimum current is

𝑃𝑃 84.3 𝑘𝑘𝑘𝑘
𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 = = = 562 𝐴𝐴
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 150 𝑉𝑉

 The approximate average current is

𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 + 𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚
𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎 ≈ = 981 𝐴𝐴
2

K. Webb ESE 471

 Ultracapacitor Sizing – Example
80

 Using the average current, we can approximate the round trip efficiency for
the capacitor bank as
1
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎 𝑅𝑅0𝑝𝑝 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 ≈2
1
1 − 𝑑𝑑 2 + 𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎 𝑅𝑅0𝑝𝑝 1 − 𝑑𝑑
2 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐

1.25 𝑉𝑉 1 − 0.42 − 981 𝐴𝐴 ⋅ 56 𝜇𝜇Ω 1 − 0.4

𝜂𝜂𝑟𝑟𝑟𝑟 ≈
1.25 𝑉𝑉 1 − 0.42 + 981 𝐴𝐴 ⋅ 56 𝜇𝜇Ω 1 − 0.4

𝜂𝜂𝑟𝑟𝑟𝑟 ≈ 0.92 → 𝜂𝜂𝑟𝑟𝑟𝑟 ≈ 92%

 Note that 𝑅𝑅0𝑝𝑝 is used, because that is the resistance of each of the 60 series-
connected parallel combinations
 Total round-trip efficiency must include the power conversion system
𝜂𝜂𝑟𝑟𝑟𝑟 = 0.97 ⋅ 0.92 ⋅ 0.97 = 0.87 → 𝜂𝜂𝑟𝑟𝑟𝑟 = 87%

K. Webb ESE 471

 Ultracapacitor Sizing – Example
81

 Solving numerically, we
find that the efficiency
of the capacitor bank is
a bit higher
𝜂𝜂𝑟𝑟𝑟𝑟 = 94.6%
 𝐼𝐼𝑎𝑎𝑎𝑎𝑎𝑎overestimates the
time-average current
 Accounting for
conversion losses, we
have
𝜂𝜂𝑟𝑟𝑟𝑟 = 89%
K. Webb ESE 471
82 Ultracapacitor Efficiency Summary

K. Webb ESE 471

 Efficiency Summary
83

Configuration Round-trip efficiency

1
Single 2
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅0 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 =
capacitor 1
𝑉𝑉 1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅0 1 − 𝑑𝑑
2 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐

1
𝑁𝑁𝑠𝑠 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑁𝑁𝑠𝑠 𝑅𝑅0 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 2
1
𝑁𝑁 𝑉𝑉 1 − 𝑑𝑑 2 + 𝐼𝐼𝑁𝑁𝑠𝑠 𝑅𝑅0 1 − 𝑑𝑑
2 𝑠𝑠 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐

1
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅𝑠𝑠 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 2
𝑁𝑁𝑠𝑠 in series 1
𝑉𝑉 1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅𝑠𝑠 1 − 𝑑𝑑
2 𝑚𝑚𝑚𝑚𝑚𝑚

1
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅0 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 2
1
𝑉𝑉 1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅0 1 − 𝑑𝑑
2 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐

K. Webb ESE 471

 Efficiency Summary
84

Configuration Round-trip efficiency

1 𝑅𝑅
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼 0 1 − 𝑑𝑑
2 𝑁𝑁𝑝𝑝
𝜂𝜂𝑟𝑟𝑟𝑟 =
1 𝑅𝑅
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 + 𝐼𝐼 0 1 − 𝑑𝑑
2 𝑁𝑁𝑝𝑝

𝑁𝑁𝑝𝑝 in parallel
1
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅0𝑝𝑝 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 2
1
𝑉𝑉 1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅0𝑝𝑝 1 − 𝑑𝑑
2 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐

K. Webb ESE 471

 Efficiency Summary
85

Configuration Round-trip efficiency

1 𝑁𝑁
𝑁𝑁𝑠𝑠 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼 𝑠𝑠 𝑅𝑅0 1 − 𝑑𝑑
2 𝑁𝑁𝑝𝑝
𝜂𝜂𝑟𝑟𝑟𝑟 =
1 𝑁𝑁
𝑁𝑁𝑠𝑠 𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 + 𝐼𝐼 𝑠𝑠 𝑅𝑅0 1 − 𝑑𝑑
2 𝑁𝑁𝑝𝑝

1
𝑉𝑉𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅0𝑝𝑝 1 − 𝑑𝑑
𝑁𝑁𝑠𝑠 in series, 𝜂𝜂𝑟𝑟𝑟𝑟 = 2
1
𝑁𝑁𝑝𝑝 in parallel 𝑉𝑉
2 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅0𝑝𝑝 1 − 𝑑𝑑

1
𝑉𝑉𝑚𝑚𝑚𝑚𝑚𝑚 1 − 𝑑𝑑 2 − 𝐼𝐼𝑅𝑅𝑒𝑒𝑒𝑒 1 − 𝑑𝑑
𝜂𝜂𝑟𝑟𝑟𝑟 = 2
1
𝑉𝑉 1 − 𝑑𝑑 2 + 𝐼𝐼𝑅𝑅𝑒𝑒𝑒𝑒 1 − 𝑑𝑑
2 𝑚𝑚𝑚𝑚𝑚𝑚

 All of the expressions on this and the previous two pages are for
constant-current charging/discharging
 For constant power, use an approximate average current
K. Webb ESE 471