Lecture #7

Basics of energy storage

PHYS-E0483_ Peter Lund 2016

 Contents Lecture # 7
§  Basic principles of energy storage
§  Energy storage in an energy system context
§  Storage characterization
§  Energy storage technologies

PHYS-E0483_#7 Peter Lund

2016
Starting point for energy storage
considerations
§  Energy demand and supply matching
§  different time scales and response requirements

§  Storage as part of the “system”

§  Physical performance of the storage, e.g. electrochemistry
§  Balance of system (BOS) make a storage operational (storage unit +
ancillary components)

§  Different storage technologies available

§  Electrical, thermal and fuel type storage

PHYS-E0483_#7 Peter Lund

2016
 Energy storage schemes
•  Traditional energy
systems
–  storage of fuel (primary
energy storage)
–  between extraction and
conversion before end use

•  Fuel storage in natural

deposits
–  e.g. fossil fuels

•  Distributed (renewable)
systems
–  energy storage in the end-
use side (after conversion)
–  storage of final energy

PHYS-E0483_#7 Peter Lund

2016
 Basic analysis principles of
storage systems
Production Load φL(t)
φP(t) φin(t) φout(t) φL(t) φP(t)

energy storage Qs(t)

•  Charge φP(t) > φL(t);

φin(t)=φP(t)-φL(t)
•  Discharge: φP(t) < φL(t);
φout(t)=φL(t)-φP(t)
•  Balance: dQs/dt=φin(t)- φout(t)
•  Capacity: Qs,min <Qs < Qs,max
PHYS-E0483_#7 Peter Lund
2016
 Cyclical load and demand behavior-
classical solar energy storage case
storage

production load

Forming a storage balance equation and solving analytically:

•  Solar I(t)=I0cos ω(t+δ), Load L(t)=L0 cos ω’t,
ω=amplitude, t=time, δ=time shift between solar and load peaks

•  Storage balance =solar-load-losses C dQ/dt= I(t)-L(t)- k Q (t)

(C=unit storage capacity; k=efficiency or loss factor)

•  Using e.g. Laplace transform we get

Q(t) = Q(0) + I0/(ωC) sin ω(t+δ)- L0 /(ω’C) sin ω’t- I0/(ωC)*sin ωδ
PHYS-E0483_#7 Peter Lund
2016
Example of a cyclic storage behaviour
Solar=seasonal, load=daily Solar=seasonal, load=seasonal
ω(load)= 24 d, ω(prod)=365 d,δ=90 d ω(load)= 365 d, ω(prod)=365 d,δ=90 d

1.5 1.5
Prod Prod
C=120 units
C=120 units Load Load
1 1 Stor
Stor

0.5 0.5

0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12

-0.5 -0.5

-1 -1

-1.5 -1.5

2 2
Prod Prod
C=20 units C=20 units
1 Load 1 Load
Stor Stor
0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
-1 -1

-2 -2

-3 -3

-4 -4

-5 -5

-6 -6

PHYS-E0483_#7 Peter Lund 2016

 Key energy system benefits
from energy storage
•  Energy system stabilization
–  storage levels out peak or random behavior
–  same effects as when having a large group of
consumers (averaging effects)
–  storage helps DEGS to follow load, reduces
DEGS capacity needed
•  Ride-Through capability
–  e.g. with solar and wind energy systems
during down-time
–  storage vs. back-up power source
•  Dispatchability
–  increasing dispatchability through storage
–  meeting the production schedule
–  economic optimization/balancing
PHYS-E0483_#7 Peter Lund
2016
 Time variations
Mismatch between demand supply versus time

PHYS-E0483_#7 Peter Lund 2016 Source :PwC, October 2013

Magnitude of variability vs storage
demand

PHYS-E0483_ Peter Lund 2015

 How to analyse storage dynamics
in an energy system?
•  Storage involves stochastic start-ups and
shutdowns
•  Estimating the effects of storage requires
preserving of the chronology of events
•  Load duration models loses the chronology
of the load (c.f. Figure)
–  to estimate the chronology, identical days
have to be assumed
–  e.g. weekend with different load profile
(downtime 60 hours, but during weeknight
10 hours)
–  several load peaks difficult to analyze
–  shape of load duration curve does not give
information on the load gradient
•  Chronological models show the transition
from hour to hour
–  ramping effects in centralized power 1-3% of
capacity/minute (Δt < 10 min)
–  grid quality (Δt < 1s) PHYS-E0483_#7 Peter Lund
2016
 Economic optimization of energy
storage
•  Optimization problem with storage
–  minimize costs= min [Z=installation, fuel, O&M], Vmin<V<Vmax
–  maximize economic merit
–  maintain quality of electricity

•  Economic benefits from storage

–  Traditional benefits = Qs * N * (cdis-ccha/η) = ∑[φdis(t)*cdis(t)-
φcha(t)* ccha(t)] Δt
cdis= cost of discharging, ccha=cost of charging storage, N=number
of cycles, η= storage efficiency of a charge-discharge cycle
–  Non-traditional benefits

PHYS-E0483_#7 Peter Lund

2016
 Non-traditional benefits from
energy storage
•  Non-traditional benefits of
storage
–  not accounted for in
traditional power system
planning models
–  dynamic benefits (operational)
–  inherent benefits
(transmission, planning)
•  Examples of dynamic
benefits of storage
–  increase the thermal ramping
limits by 50%
–  reduce fuel costs of spinning
reserve by 50%
–  50% savings in load following

PHYS-E0483_#7 Peter Lund

2016
Characterization of energy storage
Technical storage parameters
•  storage efficiency (%)
•  energy density (MJ/kg,MJ/m3)
•  power density(W/kg,W/m3)
•  response time (s)

PHYS-E0483_#7 Peter Lund

2016
Main types of energy storage

PHYS-E0483 #7Peter Lund 2016

Available energy storage
capacity

§  Pumped hydro
dominates,
>99% of all
storage capacity

PHYS-E0483_ Peter Lund 2015

 Mechanical energy storage
(potential or kinetic)
•  Gravitational storage •  Kinetic energy storage (flywheel)
–  Potential energy Wpot=-GMm/r,
•  Movement= ½ mu2 ; rotation= ½
M=mass of earth, m=mass, r=
Iω2; I=moment of inertia;
distance from earth center, G=
gravitational constant ω=angular velocity (e.g. 3-5 rps)
–  –∆Wpot= - G M m [1/(r+∆r] -1/r) •  Flywheel W=½ Iω2, I=∫ρ(x)r2dx,
≈ mg ∆r, g= 9.81 m/s2 ρ(x)=body of mass distribution,
–  1ton of water elevated to 10m =
•  High storage capacity= high
9.81x104 J =0.0273kWh
velocity and massà material
–  High storage capacity= high
strength? tensile strength limited
height and mass

PHYS-E0483_#7 Peter Lund

2016
 Pumped hydro storage
•  99.5% of commercial electrical
storage is pumped hydro
•  2-way turbines
•  1. elevated reservoir or 2.
underground lower reservoir and
surface level upper reservoir
•  Artificial hydro storage capacity
typically <24 hours
•  Natural reservoirs may have up to
1-2 years capacity
•  Pelton impulse turbines: mg ∆z =
Wpot= ½ mu2+(m’-m) P/ρ= Wkin+H
•  Overall efficiency good >70%

PHYS-E0483_#7 Peter Lund

2016
 Compressed air energy storage
(CAES)
•  Gas is compressible
•  Ideal gas law: pV=nRT, p=pressure,
V=volume, T=temperature
•  Work done by pressure (e.g. piston)
W=A ∫ f dx=- ∫ P dV (V0à V) ; f=force,
A=cross-section area of cylinder
•  Energy could be charged as
compressed air is steel cylinders,
underground caverns à discharge of
air through a gas turbine
•  Pressure and temperature variations
may be large in a compressed air
storage system (cooling, slow charging)

PHYS-E0483_#7 Peter Lund

2016
 Adiabatic compressed air energy
storage
•  High pressure and temperature
variations à large thermal losses
•  If compression done adiabatically (no
heat exchange to environment:
–  Win= P0 V0/(γ-1)[(P/P0)(γ-1)/γ -1] ;γ=1.4
–  Wout=P1V1/(γ-1)[1 (P2P1)ηt(γ-1)/γ] ,
ηt=turbine efficiency
–  T= T0 (P/P0)(γ-1)/γ
–  P/P0 = 70 gives T>1000 K à air need be
cooled before entering the storage
•  η=Wout/(Win+H0)=41% ; H0=enthalpy
input
•  Cases: Huntorf Germany 300MW
(300,000 m3) Ohio planned 2000- 3000
MW; overall storage cycle efficiency >60%
(heat stored)
PHYS-E0483_#7 Peter Lund
2016
 Flywheel energy storage
•  W/M=σρ-1 Km , Km = geometric shape factor (0.3-1),
σ=maximum stress, ρ =mass density; W=energy density,
M=fly wheel mass
•  Light materials: Kevlar, carbon fibers, alloys, composites
(largeσ/ρ, but expensive)
•  W/M: plywood15, superpaper 27 kJ/kg, composites <200

PHYS-E0483_#7 Peter Lund

2016
Design of a Flywheel Storage (Ex)

Consider a solid disc flywheel of radius 50 cm and mass 140 kg.

How fast would it have to spin to have a store the equivalent
amount of energy that is stored in just 10 kg of gasoline when
burned in an engine?
10 kg of gasoline = 140 kWh
Engine has 15% efficiency à 21 kWh of useable energy
Flywheel has a conversion efficiency of 80%
Flywheel must therefore store 21/.8 = 26.25 kWh
Kinetic Energy goes as 1/2*I*w2, I =1/2MR2.
If we measure w in revolutions per second then the stored
energy of a flywheel is approximately 6MR2 x w2 (RPS)

For M=140 kg and R=50cm this yields a required w of 500 RPS or

30,000 RPM
The required energy storage is 26 kWh/140 Kg = 0.18 kWh/kg
Such a flywheel requires construction out of carbon fiber.

PHYS-E0483_#7 Peter Lund 2016

 Electromagnetic energy storage

•  Electric and magnetic fields can

store energy
•  Electric field storage = capacitor
•  Magnetic field storage =
electromagnet
•  Energy density of magnetic fields
>> electric fields
•  Storage applications:
Supercapacitors,
Superconductive magnetic
storage (SMES)

PHYS-E0483_#7 Peter Lund

2016
 Capacitive energy storage
•  Two plates separated with a dielectric material V
–  Electric field E= external voltage between plates E
divided by distance = V/d + -
–  Energy stored W= ∫w(r) =∫½ ε0 E(r)2=½ ε0 V2Ad-1 -
–  ε0= vacuum permittivity, A=plate area, w= energy + -
density of the electric field
–  C= ½ ε0 Ad-1=constant=capacitance à W=½ C V2
-
+ d -
- small capacitors W=40-50 μJ, but P=1000 MW/m3

•  Supercapacitors: +
–  No dieletric material but a double-layer (both plates
soaked in electrolyte with very thin insulator in +
between); very low d, large A
–  Energy density 0.5-15 Wh/kg, power density up to 15
kWh/kgVery short term storage, high power

Source: wikipedia
PHYS-E0483_#7 Peter Lund
2016
Nanoporous supercapacitors

•  where k is the dielectric

constant, e0 is the permittivity
of free space, A is the total
effective surface area of the
electrode structure, and d is
defined as the separation
distance between the
electrode surface and the
ions;
•  High specific capacitance can
be achieved through the use
of porous electrode
•  Power densities up to 2 kW/
cm3 = 2000 MW/m3 !

PHYS-E0483_#7 Peter Lund 2016

 Other types of energy storage
•  Electromagnetic storage
–  W=½ μ H2, μ=permeability, H=magnetic field intensity
–  Superconducting coils (e.g. Nb-Ti) < 1000 kg/MJ
•  Hydrogen
–  H20à H2+1/2 O2
–  electrolyzer+hydrogen storage+fuel cell <50% efficiency
•  Electrochemical energy storage
–  Free energy change (∆G) of a chemical reaction is
converted into electricity ; open circuit voltage (max V)
E0=-∆G/nF, F=Faraday constant
–  Me à ßMen++ne- (e.g. Li++e-àLi, E0=-3.05V)
–  Many electrode reaction available
–  Batteries, fuel cells, etc. based on this principle
•  Thermal storage
–  Sensible heat, phase change, thermochemical
PHYS-E0483_#7 Peter Lund
reactions 2016
 Power range and
discharge time of
storage technologies

PHYS-E0483_ Peter Lund 2016

Electricity storage vs time

PHYS-E0483_#7 Peter Lund 2016

Storage vs energy system requirements

PHYS-E0483_ Peter Lund 2016

Overview of electricity
storage technologies

PHYS-E0483_ Peter Lund 2015