Training2022 - VRE Utilization in Power Grids - Final

Variable Renewable Energy (VRE)
Utilization in Power Grids

(การใช ้พลังงานหมุนเวียนชนิ ดผันแปรในโครงข่ายไฟฟ้ า)
WIJARN WANGDEE
Center of Excellence in Electrical Power Technology (CEPT)
Faculty of Engineering, Chulalongkorn University
March 2022
Course Content
Chapter 1: Renewable Energy (RE) Overview
 RE Technologies
 RE Portfolio in Germany vs. Thailand
 Variable RE (VRE) Properties
Chapter 2: High VRE from Reliability Perspectives
 Uncertainty in Power Grid
 High VRE Required Probabilistic Analysis
 Capacity Credit (Equivalent Firm Capacity) for VRE
 Generation & Transmission Adequacy for VRE
 Items Required for Value-Based Reliability Analysis
Chapter 3: Power Grid Flexibility for VRE
 Impact of Variability & Uncertainty
 Need for Grid Flexibility
 Increase in Grid Flexibility
2
INSTRUCTOR : WIJARN WANGDEE
Chapter 1
Chapter 1: Renewable Energy (RE) Overview
 RE Technologies
 RE Portfolio in Germany vs. Thailand
 Variable RE (VRE) Properties
3
RE Technologies
Relatively Firm
(except some marine)
Relatively Non-Firm
(Variable Renewable Energy)
4
RE Technologies : Solar Energy
Sun Radiation for Power Generation
There are two types of solar radiation:
 Direct radiation: Solar radiation traveling on a straight
line from the sun down to the surface of the earth.
 Diffuse radiation: Scattered radiation by molecules
and particles in the atmosphere but still reaches
earth's surface.
Clear sky: Direct 85%, Diffuse 15%

 As the sun goes lower, diffuse radiation keeps going
up until it reaches 40%.
 Clouds and pollution also increase diffuse radiation.
 Ratio of Direct and Diffuse varies with latitude and
climate. (Source: Amornrat Limmanee, 2015)
5

Sun Radiation for Power Generation
Pyranometer:
for measuring
total (direct and
diffuse) radiation
Pyrheliometer:
for measuring
only direct
(Source: Amornrat Limmanee, 2015)
radiation
 The solar cell or collector should be oriented directly
to the equator, facing south (in Northern hemisphere).
 The optimum tilt angle of the solar cell (or collector) is
equal to the latitude of the location.
 Thailand: longitude east 100° latitude 14~15°
https://medium.com 6
Concentrating Solar Thermal Power Plants
Solar Dish
Fresnel Reflector
Parabolic Trough
Solar Tower
7

Concentrating Solar
Power System
Concentrating solar thermal
power plant consists of four
main functionalities:
 Concentrator
 Receiver
 Heat transfer media or
storage
 Power conversion
(Source: O.E.K. Onwunta, 2014)8

RE Technologies : Solar Thermal
Concentrating Solar
Power (CSP) System
There are four basic types
of CSP plants:
a) Parabolic Trough
a) b)
b) Fresnel Reflector
c) Solar Tower
d) Solar Dish
CSP technologies are cost-
effective for areas with
Direct Normal Irradiance
(DNI) > 2000 kWh/m2-yr
Thailand DNI ≈ 1400 kWh/m2-yr

Thailand GHI ≈ 1900 kWh/m2-yr
c) d)
( GHI = Global Horizontal Irradiance )
9
INSTRUCTOR : WIJARN WANGDEE (Source: https://www.energy.gov)
RE Technologies : Solar Thermal

Concentrating Solar Power (CSP) System Enhancement (Hybrid System)
Stand-Alone CSP
CSP with Fossil Integrated Solar Combined

Fuel Backup Cycle System (ISCC) 10
INSTRUCTOR : WIJARN WANGDEE (Source: http://www.greenrhinoenergy.com/solar/technologies/cst_systems.php.)
(Source: K.D.G. Imalka Jayawardena, 2013)
RE Technologies : PV
Timeline of Four Generations of
Solar PV Cells
 1st Gen: based on crystalline silicon
technologies (mono- & poly-)
 2nd Gen: amorphous silicon (a-Si), μc-
Si thin films, cadmium telluride (CdTe),
cadmium sulfide (CdS) and copper
indium gallium selenide (CIGS)
 3rd Gen: based on new compounds
including nanocrystalline films, tandem or
stacked PV, organic (polymer)-based,
dyed-sensitized solar cells (DSSC)
 4th Gen: known as “inorganics-in-
organics” combining the low cost/flexible
polymer thin films with the stability of
novel inorganic nanostructures or
organic-based nanomaterials such as
carbon nanotubes, graphene, etc.
11
RE Technologies : PV (Source: NREL, 2021)

Efficiency of Four Generations of Solar PV Cells 12
RE Technologies : PV (Source: NREL, 2021)

Efficiency of Emerging Solar PV Cells 13
RE Technologies : Solar PV
Crystalline Si takes 40 years to double its efficiency, while Perovskite takes 5 years to do so!!
In 2020,
SwiftSolar puts
perovskite
(attracting
blue light) on
the top of
standard Si cell
(attracting red
light) as a
multi‐layer PV
cell to improve
an efficiency
up to 28%.
(Source: https://www.ossila.com/pages/perovskites-and-perovskite-solar-cells-an-introduction)

Efficiency of Some Selected Solar PV Cells 14
Efficiency of the First Three Generations of Solar PV Cells and Trending
Trending: Improve an efficiency, then reduce the cost
Trending
(Source: Gavin Conibeer, 2007)
15
Widely Used Solar PV Cells
Thin Film Si
Thin Film Si
Mono c-Si
Mono c-Si
Poly c-Si Poly c-Si

16
Newer Types of Concentrating PV
(for direct sunlight)
Solar PV Cells
Perovskite Solar Cells

(Organic PV)
Dye Sensitized (Organic PV)
CIGS
17
PV Cell Efficiency vs. Cost
Efficiency > 40% for Niche market,

i.e. Mars Solar Power by NASA
(using GaAs solar cells)
18
INSTRUCTOR : WIJARN WANGDEE NOTE: Old information (before 2015)
Worldwide Solar Energy Costs Compared to Other Generation Sources
“Solar power is not only cheaper

than new fossil fuel and nuclear
power generation, but it is even
cheaper than wind power.”
19
(Source: https://www.dw.com/en/desert-large-solar-plants-also-pay-off-in-countries-with-less-sun/a-58284114)
Price Learning Curve
of Solar PV Module
 2006 was the end of
628 GWp (2019 end)
Silicon shortage as China

756 GWp (2020 end)
began the c-Si PV mass

production.
Shipments at Avg. price at

Year year end year end
(GWp) (US$/Wp)
2016 76 0.37
2017 105 0.35
2018 109 0.25
2019 105 0.23
2020 128 0.21 (Source: ITRPV, 2021)
20
Trending Cost of Solar PV vs Battery (EV)
 Falling battery prices (improving battery range) is
the biggest drivers of EV.
 Cheap “Solar PV + Battery” : What does it mean?
Battery cost
Solar PV cost
(Source: Susana Almeida de Graaff, 2018) (Source: Mark Wehling, 2021) 21

Different Grid-Connected PV Inverter Structures
a) Module
Inverter
b) String
Inverter
c) Multi-String
Inverter
d) Central
Inverter
22
INSTRUCTOR : WIJARN WANGDEE (Source: O.E.K. Onwunta, 2014)
Different Grid-Connected PV Inverter Structures
Structure Advantages Disadvantages
Module ⁻ Maximizing power production (no mismatch losses) ⁻ Low power ratings of PV modules
Inverter due to the use of single solar panel with its own ⁻ High cost per watt
inverter and MPPT control ⁻ Lifetime of an inverter smaller than a PV module
String ⁻ Individual MPPT for each string --> independent ⁻ Higher cost per watt compared to central inverter
Inverter ⁻ Improved system reliability (one inverter/string) structure
⁻ Flexible to build with different orientations & no. of
PV modules for each string
Multi-String ⁻ Combining the advantages of both string-inverters ⁻ Two-stage system composed by more than one
Inverter (high power due to individual MPPT control) and DC-DC stage
central-inverters (low cost). ⁻ Higher cost per watt compared to central inverter
⁻ Flexible solution due to individually controlled of structure
each PV string
Central ⁻ Typically, most widely use for large-scale PV. ⁻ Need for high DC-link voltage (550–850V) & very
Inverter ⁻ High efficiency (low losses in a conversion stage) long DC cables (between PV strings to inverter)
⁻ Low cost due to the use of only one inverter ⁻ Entire system reliability depends on only one
inverter
(Source: O.E.K. Onwunta, 2014) 23
Solar PV Cell Interconnection
 Case: identical PV cells
24
INSTRUCTOR : WIJARN WANGDEE (Source: Raeey A. Regassa, 2017)
Solar PV Cell Interconnection (Case: non-identical PV cells)
25
Partial Shading Effect on V-I Curve
 Case: without bypass diode
26
V-I Characteristic Curve of a PV Panel
 Effects of Ambient Temperature and Solar Irradiation
Effect of Ambient Temperature @1000W/m2 Effect of Solar Irradiation @25°C

27
INSTRUCTOR : WIJARN WANGDEE (Source: http://www.emrwebsite.org/photovoltaic-system-using-emr-and-simuliktm.html)
Effect of Temperature on V-I Characteristic of Solar PV
 Efficiency of solar PV reduces by 1% for every 2.5 °C increase in ambient temp. (vice versa)
(Source: K. Lappalainen, 2017)
NOTE: Rated values at 25 °C. The irradiance on the surface of the cell is 1000 W/m2.
28
Power Output of a PV Panel with Maximum Power Point Tracking (MPPT)
 Effects of Ambient Temperature and Solar Irradiation
Effect of Ambient Temperature @1000W/m2 Effect of Solar Irradiation @25°C

29
INSTRUCTOR : WIJARN WANGDEE (Source: http://www.emrwebsite.org/photovoltaic-system-using-emr-and-simuliktm.html
Power Output of a PV Panel with Maximum Power Point Tracking (MPPT)
 Some more details about MPPT (Perturb & Observe Method)
(Source: Raeey A. Regassa, 2017)
30
Power Output of a PV Panel with Single- and Dual-Axis Tracking Systems
 Tracking the sun maximizes solar power generation,

providing between 30-50% more power compared to
untracked panels.
(Source: https://jaizianenergy.com/products/)
31
INSTRUCTOR : WIJARN WANGDEE (Source: http://apdico.net/solar_power)
Solar PV Output Depends on:
 Orientation
 Tilt
 Tracking
(Source: EIA, 2014)
32
(Source: O.E.K. Onwunta, 2014)
Wind Power System

 Basic components of a
modern, horizontal-axis
wind turbine with a
gearbox.
 The rotor is mounted
directly on the generator
shaft. The rotor is the
heart of a wind turbine
and consists of multiple
rotor blades attached to
a hub.
33
RE Technologies : Wind Power

Wind Turbine : Gearbox vs. Gearless
Wind Turbine with Gearbox Gearless Wind Turbine

34
INSTRUCTOR : WIJARN WANGDEE (Source: Ghaeth Fandi, 2017)
Typical Power Output vs.
Wind Speed Curve (blue curve)
 Cut-in speed ≈ 2.5 to 4 m/s
 Rated speed ≈ 11 to 15 m/s
 Cut-out speed ≈ 20 to 25 m/s
 Betz limit: theoretical limit up

to 59.3% on the amount of
wind energy that can be
extracted. (yellow curve)
 In Region II, variable speed
operation allows energy
extraction at peak efficiency
over a wider range of wind
speeds. (green curve)
35

Wind Turbine
Power Curve and
Control Zones
Variable-speed pitch-
controlled wind turbine,
with 4 zones and 2 areas:
 Below rated power,
the wind turbine
produces only a
fraction of its rated
power, and thus an
optimization control
strategy needs to be
performed.
 Above rated power, a
limitation control
strategy is required to
prevent the damage.
36
Wind Turbine Power Curve and Control Zones
 Zone 1: Rotor blades are fixed at a specific angle, the generator reaction torque regulates rotor
speed to maximize energy capture.
 Zone 2: Power delivered by the rotor is subject to a particular design of the blades. The blade is set
to a fine pitch position to yield higher power capture values while the generator torque itself regulates
the rotor speed.
 Zone 3: Power delivered by the rotor is regulated

either by pitching the blades toward the wind to
maximize energy capture or by pitching to
discard the excess power to be within the
mechanical limitations (the torque is held
constant and the pitch is continually changed).
 Zone 4: Active stall control is a combination of
stall and pitch control. Above rated operation, the
control system pitches the blades to induce stall
instead of feathering.
37

Evolution of Power Electronics Coverage and Wind Turbine Size
 The growth of power electronics in
wind turbine systems with its power
rating coverage and function role is
shown in the figure.
 Blue circle = the power coverage
by power electronics
38
INSTRUCTOR : WIJARN WANGDEE (Source: Ghaeth Fandi, 2017)
Relationship between Wind
Turbine Output Power and
Respective Diameters
 If a rotor diameter is doubled, a
swept area will increase by four
times (𝐴 𝜋𝑟 ).
 Wind speed (𝑣) depends on the
height (H), i.e. 𝑣 𝑣 .
 Power output:
Where:
α = terrain friction (≈1/7), ρ = air density
Cp = Coefficient performance (<0.59)
39

Wind Energy Conversion Systems
In consideration of speed, wind energy systems
are either fixed or variable speed while the
coupling between the mechanical and electrical
parts could be with or without a gear-box.
 SCIG = Squirrel-Cage Induction Generator
 WRSG = Wound Rotor Synchronous
Generator
 PMSG = Permanent-Magnet
Synchronous Generator
 WRIG = Wound Rotor
Induction Generator
 DFIG = Doubly-Fed
Induction Generator
40
Wind Turbine Type 1 : Fixed Speed Wind Turbine
 The first type is a constant-speed wind turbine system with a standard squirrel-cage
induction generator (SCIG) directly connected to the electric grid using a step up power
transformer. To limit the output power, this turbine uses currently the active stall control.
 Fixed speed systems have the advantage of simplicity and low cost. However, the main
drawbacks of this concept include the inability of supporting speed control, the requirement
of a stiff grid (nearly constant voltage and frequency).
 A major drawback of the
induction machine is the reactive
power that it consumes for its
excitation field and the large
currents the machine can draw
when started. Therefore, this
turbine type typically employs a
soft starter and discrete steps of
capacitor banks. 41
INSTRUCTOR : WIJARN WANGDEE (Source: M.G. Molina, 2011)

Wind Turbine Type 2 : Variable Speed Wind Turbine
 The second topology corresponds to the limited variable speed controlled wind turbine with
variable rotor resistance. It uses a wound rotor induction generator (WRIG).
 The stator windings of the generator are directly connected to the grid. The rotor winding of
the generator is connected in series with a controlled resistance that defines the range of
the variable speed (typically 0 – 10% above synchronous speed). Thus, by varying the total
rotor resistance the generator speed and thus the output power are controlled.
 The turbine would have to
spin faster to create the
same output power, for
an added rotor resistance.
 However, power is lost as
heat in the rotor resistance.
42
Wind Turbine Type 3 : Doubly-Fed Induction Generator (DFIG)
 This type is a variable speed wind turbine system with a DFIG. The power electronic
converter feeding the rotor winding has a power rating of approximately up to 30% of the
rated power; the stator winding of the DFIG is directly connected to the grid.
 However, its main drawbacks are the use
of slip rings, which needs brushes and
maintenance, and the complex protection
schemes in the case of grid faults.
 The greatest advantage of the DFIG is that it
offers the benefits of separate real and reactive
power control, much like a traditional synchronous
generator, while being able to run asynchronously.
 Such control makes the rotor flux response fast
enough to remain stable during severe grid
disturbances.
43

Wind Turbine Type 4 : Full-Converter Wind Turbine
 The fourth type is a variable speed wind turbine with full-rated power electronic conversion
system and a synchronous generator or an induction generator (SCIG). The synchronous
generator can be electrically excited (WRSG) or permanent magnet excited type (PMSG).
 A type 4 WTG has similar characteristics to type 3 and, since it is completely decoupled
from the grid, it can provide an even wider range of speed variation as well as reactive
power and voltage control capability.
 As entirely decoupled from the grid, grid
codes such as fault ride through
and grid support are easier to
be accomplished, as required
from modern applications.
 Also, its output current can be
modulated to zero, thus limiting
the short-circuit current
contribution to the grid. 44
RE Portfolio in Germany vs. Thailand
45
INSTRUCTOR : WIJARN WANGDEE (Source: https://www.cleanenergywire.org/factsheets/germanys-energy-consumption-and-power-mix-charts)

Installed Generation Capacity and Capacity Expansion Pathways in Germany
(Source: Thobias Sach , 2019)
46
RE Geographical Development in Germany
47

in Germany
48
Renewable Energy Outlook: Thailand
(Source: กรมพัฒนาพลังงานทดแทนและอนุ รก
ั ษ์พลังงาน (พพ.) กระทรวงพลังงาน, https://www.dede.go.th/download/stat63/Fiscal_Year_renewabel_Energy_October_2564.pdf)
49

High solar contribution at midday peak load whereas wind ramps up at evening
(Source: IEA, 2018)
50
Solar PV Comparison: Thailand vs. Germany
(Source: http://globalsolaratlas.info)
Country Thailand Germany
Size (km2) 513,120 357,022

Irradiation
1900 1100
(kWh/m2 per yr)
Calculated Temp.
27.5 9.5
(Degree Celsius)
PV Output
1500 1000
(kWh/kWp per yr)
PV Installed Cap.
3 53
(GW) by 2020
Percentage of
<10 >70
Solar Rooftop (%)
Percentage of
>90 <30
Solar Farm (%)
(Source: IRENA, Renewable Capacity Statistics 2021)
51
Variable RE (VRE) Properties

Approximate Ranges of VRE Penetrations
 Three qualitative levels of maximum instantaneous VRE penetrations are classified in to low,
medium and high.
 If at any point in time the share of VRE generation stays below about 10–15 % of the total
instantaneous load, the VRE penetration can be described as low and no significant
integration issues are generally expected.
(Source: IRENA, 2018a)

52
Relevant Time-Scales of Electric Power System Planning and Operations
 From grid operation standpoint, wind and solar power output variations are in the time-scale
ranging from second to days.
Operations Planning
“Dynamics”
(Automation)
(Source: Andreas Ulbig, 2014a) 53

VRE Properties (Source: PIER, 2007)
Analyses Required for VRE at

Different Time Scales in Grid
Planning & Operations
 Four key study steps needed to be
conducted across the operational
and planning time horizons.
Resource Adequacy
Unit Commitment
Economic Dispatch
Frequency Regulation
54
VRE Properties : Technical Challenges for VRE Integration
 Non-synchronous: power electronics interfaced, which decouple their frequency and
voltage from the grid, and therefore lead to different dynamic response to disturbances, i.e.
inertial, short circuit current, etc.
 Location: system constrained by the availability of VRE in the area. Modularity implies to
VRE being deployed at very small scales and rapidly in distribution grids.
 Uncertainty: related to their generation, or the unexpected change in their power output
compared to forecasted values.
 Variability: the expected change in power output due to variability in the available primary
resource (solar irradiation or wind speed) occurring over time.
55

Key Links between VRE Properties, System Properties and Planning Stages
 Figure shows the key areas in which VRE deployment has potential to influence the
planning of power systems beyond simple changes to operational practice.
56
INSTRUCTOR : WIJARN WANGDEE (Source: IRENA, 2017b)
Planning Solutions for VRE Integration (System Properties)
 Planning for Firm Capacity: VRE makes the concept of "capacity credit", known as "firm
capacity". Essential for generation resource planning to have sufficient supplies to cover
periods when low amounts of VRE are available. The concept of Effective Load Carrying
Capability (ELCC) based on reliability analysis is widely used to estimate "capacity credit".
 Planning for Flexibility: the higher VRE, the greater variability and uncertainly of supply.
Thus, the flexibility of a system becomes more important. High shares of VRE are likely to
require more investment in flexibility measures to maintain balance at all times.
 Planning for Transmission Capacity: the availability of VRE resources depends on their
location, and new capacity may need to be planned to transmit power from VRE sources.
 Planning for Stability: improving operational practices (frequency and voltage responses)
to maintain the capability to respond to contingency events.
57

Grid Impacts & Challenges Arising from VRE Integration
Firm
Capacity
Flexibility
Transmission
Capacity
Stability
(Source: IRENA, 2017c) 58

Power Grid Reliability: From Planning to Operation
 The key system properties (firm capacity, flexibility, transmission capacity, and stability) are
placed in the reliability matrix in the table below.
 Power system reliability can be divided into two aspects: adequacy and security.
 Each reliability aspect can then be addressed from generation and network viewpoint.
(Source: IRENA, 2017b)
59
Chapter 2
Chapter 2: High VRE from Reliability Perspectives
 Uncertainty in Power Grids
 High VRE Required Probabilistic Analysis
 Capacity Credit (Equivalent Firm Capacity) for VRE
 Generation & Transmission Adequacy for VRE
 Items Required for Value-Based Reliability Analysis
60
Uncertainty in Power Grids
Reliability Study
Consideration
 Hierarchical Level I (HL-I)
involves the total generation
to meet the total load.
 Hierarchical Level II (HL-II)
involves the generation and
transmission facilities to deliver
power to supply points.
 Hierarchical Level III (HL-III)
involves the complete system.
(In practice, HL-II and distribution
system are separately analyzed.)
(Source: Wijarn Wangdee, 2017)

61

Power Grid Reliability Fundamentals
Reliability
Adequacy Security
Existence of sufficient Ability to respond to system

generation and disturbances such as N-1
transmission facilities. contingency events.
(Probabilistic Approach) (Deterministic Approach)
(System Planning) (System Operation)
62
Types of Uncertainty in Power Grids
 Reliability analysis is needed for coping with uncertainties.
 Two different forms of uncertainty in power grid reliability analysis:
1. Aleatory Uncertainty : random process of equipment failure
2. Epistemic Uncertainty : knowledge-based limitations
 In aleatory uncertainty, this is the uncertainty due to equipment outage and repair processes
that are random and create variability on available capacity/capability.
 In epistemic uncertainty, this is the uncertainty due to lack of information that can be
reduced by better information such as:
 Load forecast enhancement
 Renewable energy forecast enhancement
63

Uncertainties in Power Grids for the Good Old Days (Past)
 High aleatory uncertainty, but low epistemic uncertainty
 Time-varying (changing) loads are easily handled by large centralized generators.
Aleatory Uncertainty
Epistemic Uncertainty 64
Uncertainties in Power Grid for the Challenging New Days (Future)
 Even higher aleatory uncertainty, but much more epistemic uncertainty
 “Net” loads (DGs embedded) & high portion of variable RE cause more fluctuations.
Epistemic Uncertainty
+ Aleatory Uncertainty

Epistemic Uncertainty
For Reducing
Forecast
Center
More Epistemic Uncertainty (both loads & generations) + Aleatory Uncertainty 65


Uncertainties in Power Grid for the Challenging New Days(Future)
(Aleatory Uncertainty) (Epistemic Uncertainty) (Epistemic Uncertainty) (Epistemic Uncertainty)
(Source: adapted from Lion Hirth, 2013)

66
High VRE Required Probabilistic Analysis
Reliability Criteria Applied to Power Grids
 Redundant capacities in generation and network facilities have been inbuilt in order to
ensure adequate and acceptable continuity of supply.
 The main question arisen is “how much redundancy and at what cost?”
 To satisfy both economic and reliability constraints, the planning and operating criteria have
been developed and applied in reliability analysis:
 Deterministic Criteria
 Probabilistic Criteria
 Both criteria above are used to deal with randomly equipment failure events, but cannot
guarantee the same degree of reliability can be achieved.
 For example: the weakness of the deterministic criteria is in the difficulty to handle variability
and uncertainty factors.
67

Reliability Criteria Applied to Power Grids: Deterministic vs. Probabilistic
Deterministic Criteria Applied: Probabilistic Criteria Applied:
 Straightforward, easy to  Need of more knowledge, more efforts
implement, no new computing in studies, and new computing tools
tool required needed
versus
 Consequence analysis only, no  Both consequence and probability are

probability considered.
 The “worst case” study but the  All possible system states and load
worst case may be missed. levels which can occur.
 Not dealing with multiple outages  Dealing with multiple outages
 When variable generation (wind, solar, etc.) and DGs are greatly penetrated into
power systems, probabilistic methods becomes much more important for both
planning & operation applications to handle large variability and uncertainty factors.
68
Reliability Criteria Applied to Power Power Grids: Deterministic vs. Probabilistic
Examples of Deterministic Criteria:
 A fixed percentage of reserve margin
in planning (i.e. 15%)
 Spinning reserve = a largest online
generator capacity
 N-1 criterion
Examples of Probabilistic Criteria:

 Loss of Load Expectation (LOLE), i.e.
LOLE = 1 day in 10 years
 SAIDI and SAIFI targets
 Severity Index (SI)
69

Reliability or Risk Index (Indices)
n
LOLP   pk (Loss of Load Probability)
k 1
n
LOLE   pk tk (Loss of Load Expectation)
k 1
n
LOEE   pk Ek (Loss of Energy Expectation)
k 1
Where  n is the total no. of capacity outage states.

 pk is the individual probability of the
capacity outage state k.
 tk is the no. of time units when there is a
loss of load.
 Ek represents the energy that cannot be
supplied in a capacity outage state k. 70
Reliability or Risk Index (Indices)
LOLE: 0.1 day/year ≠ 2.4 hours/year !!
 Daily peak load variation curve is

used for calculating LOLE in day/year.
 Hourly load duration curve is used for
calculating LOLE in hours/year. 71
High VRE Required Prob.

Chronological Load vs. Load
Duration Curve (LDC)
(Source: Falko Ueckerdt, 2015)
72
VRE Modelling: Solar PV and Wind Power Output “Duration Curves”
 Non-chronological plot in descending order (from max to min)
 Be Careful !!! that these plots CANNOT
directly use to subtract from the native
(original) “load duration curve” to obtain
the “net load” duration curve (due to non-
chronological characteristic).
(Source: Wolf-Peter. Schill, 2013)
73

VRE Modelling: Impact of Variable Renewable Energy on Net Load
 Abundance: high VRE output
decreases the net load and may
reduce non-VRE plant output
 Scarcity: low contribution during
peak hours
 Deep ramping: flexibility
required during peak periods
Utilization for non-VRE plants (0% VRE)
Utilization for non-VRE plants (20% VRE)

(Source: Simon Müller, 2015) 74
VRE Modelling: Renewables Subtracted from Native Load in LOLP Calculation
 Renewable production is subtracted from native load (original load) to obtain “net
load”, which is always lower. LOLP reduced.
(Source: Ame Olson, 2015)
75

VRE Modelling: Multi-State Model for Wind Turbine Generator (WTG) Outputs
Wind Turbine
Power Curve
(Source: Jaeseok Choi, 2014)

Wind Speed
Profile
76
VRE Modelling: Actual Wind Power Curve (A Wind Turbine & A Wind Farm)
 Zone 1 is caused by shade effect. Zone 2 is abnormal working states (pitch control
misoperation). Zones 3 & 4 are affected by quality of data; sensor, commu. link, etc.
(Source: E. Sainz, 2009)
(Source: Dinh-Duong Le, 2013)
77

Non-Sequential (Multi-state): Sequential (Time-series):
VRE Modelling: Two Methods Can Be Applied
Peace Region wind speed profile (winter) 60

0.05
50 Year 1 Year 2
W ind S peed
0.04
Probability
40
(km /h)
0.03 30
0.02 20
0.01 10
0 0
0 2 4 6 8 10 12 14 16 18 20 22 24 1 25 49 73 97 121 145
Wind Speed (m/s) Hour
Wind Farm Power Curve Wind Farm Power Curve

1.0 1.0
Power Output
Power Output
0.8 0.8
(p.u.)
(p.u.)
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
0 4 6 8 10 12 14 16 18 20 22 24 26 28 30 0 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Wind Speed (m/s) Wind Speed (m/s)
Peace Region wind power output (winter)

0.20 50
Capacity Factor = 0.33 Year 1 Year 2
P ow er O utput
40
Probability
0.15
30
(M W )
0.10
20
0.05 10
0.00 0
0
48
192
418
734
1077
1385
1551
1600
1 25 49 73 97 121 145
MW Time (hour)
78
VRE Modelling: Diversity in Solar Power for Geographically Dispersed Locations
 19 sites across Georgia State (>3 times smaller in size compared to Thailand)
 Red curves = hourly data, while black curves = minute-resolution data
(Source: I. Kim, 2017)

79

Correlation of Solar Irradiances vs. Distance between Them (km)
Greater Geographical Diversity = Less Correlation (between sources)  Better System Reliability
 Solar farms located further apart from each other are more independent (less correlated).
(Source: Richard Perez, 2016)
80
Correlation of Solar Irradiances vs. Distance between Them (km)
Greater Geographical Diversity = Less Correlation (between sources)  Better System Reliability
(Source: Tom Hoff, 2011)

81

VRE Modelling: Correlation of Wind Farm Output vs. Distance between Them (km)
Geographical Diversity = (Source: M. O’Malley, 2011)
Less Correlation:
 Wind farms located
further apart from
each other are more
independent (less
correlated).
 Thus, they help to
smooth out the
overall MW output as
well as enhancing the
system reliability.
82
VRE Modelling: Wind Speed Correlations (Geographical Diversity)
High correlation (less independent):
two wind farms are located very
close to each other. (Source: Wijarn Wangdee, 2006)
P(0) = 0.25 Mean= 9.59 MW Mean= 9.75 MW
Median= 4.58 MW Median= 5.36 MW
P(0)= 0.19
S.D.= 11.68 MW S.D.= 11.23 MW
P(0)= 0.14
Mean= 9.53 MW
P(0)= 0.12 Median= 6.00 MW
0.3 S.D.= 10.15 MW
P(0) = 0.10
0.25 Rxy = 0.94
Probability
0.2 Rxy = 0.75
0.15 Rxy = 0.48
0.1 Rxy = 0.25

0.05
Rxy = 0.05
0
Mean= 9.58 MW 0 4 8 12 16 20 24 28 32 36 40 Mean= 9.44 MW
Median= 7.25 MW Median= 6.52 MW
Power Output (MW)
S.D.= 8.75 MW (2×20MW) S.D.= 9.41 MW
Low correlation (more independent):

two wind farms are located far away
from each other. 83

VRE Modelling: System Reliability with Different Wind Correlations
 Two wind farms with 20MW installed at each farm
 Two wind farms with 40MW installed at each farm
 Note: low correlation (near zero) means outputs from two wind farms are more independent.
2.5 0.8
LOLF (occ./yr.)
LOLE (hrs./yr.)
2.0 0.6
1.5 2×20MW 0.5 2×20MW

1.0 2×40MW 0.3 2×40MW
0.5 0.2
0.0 0.0
0.94 0.75 0.48 0.25 0.05 0.94 0.75 0.48 0.25 0.05
Cross-correlation (Rxy) Cross-correlation (Rxy)
LOLE = Loss of Load Expectation (hours/year)

LOLF = Loss of Load Frequency (occurrences/year) 84
Capacity Credit (Equivalent Firm Capacity) for VRE
Effective Load Carrying
System Peak Load (MW)
Capability (ELCC)
10028
10078
10128
10178
10228
10278
10328
10378
9878
9928
9978
Calculation:
 To determine capacity 6.00
Severity Index (sys.mins)

credit of wind power. 10199 MW
5.00 Base Case (No Generation Addition)
 By increasing system load Adding 1600 MW Wind Power
4.00
while keeping generation
capacity unchanged. 3.00
2.00 Specified reliability level
1.00
0.20
Capacity Factor = 0.33
0.00
Probability
0.15
0.10 ELCC = 321 MW
0.05
0.00
SI = Severity Index (system.minutes)
0
48
192
418
734
1077
1385
1551
1600
MW Where: SI = 60*(Expected Energy Not Supplied)/(System Peak Load)85


ELCC at Different Wind Power Penetration
 ELCC values depend on the specified risk index used in the calculation.
 VRE outputs frequently change, so ELCC based on LOLF is less than that of LOLE.
50% 50%
ELCC(LOLF) ELCC(LOLF)
installed WECS
installed WECS
40% 40%
ELCC in % of
ELCC(LOLE)
ELCC in % of
ELCC(LOLE)
30% 30%
20% 20%
10% 10%
Regina site Swift Current site
0% 0%
0% 5% 10% 15% 20% 25% 0% 5% 10% 15% 20% 25%
Wind power penetration (%) Wind power penetration (%)
LOLF = Loss of Load Frequency (occurrences/year)

LOLE = Loss of Load Expectation (hours/year)
86
Factors That Affect the Capacity Credit (Value) of VRE
 Coincidence with load
 Locations with better resources and better correlation with high load periods will have
higher ELCC values.
 Coincidence with existing VRE

 Common resource types show diminishing marginal returns; each additional plant
has less value than the previous one.
 Production variability
 Statistically, the possibility of low production during a peak load event reduces the
value of a resource.
 Reliability target
 Effective capacity does not have linear relationship with system LOLE.
87

Example of ELCC Calculation for “High Load Weekday in August”
88
Example of ELCC Calculation for “High Load Weekday in August”
 Within each bin, choose each (load, wind, and solar) daily profile randomly, and
independent of other daily profiles
89
INSTRUCTOR : WIJARN WANGDEE (Source: Ame Olson, 2015)

Example of ELCC Calculation for “High Load in August” & “Entire Year”
 System LOLE can be

obtained by summing
all the bins (weighted)
throughout the entire
year.
 System LOLE is
concentrated in
summer afternoon
hours (highest LOLE
during the high load
hours in August).
90
 ELCC of the wind site

for all the bins
throughout the year.
 Wind site has relative
low output on summer
afternoons, resulting
in low ELCC.
 Coincidence of low
wind power output
and high system
LOLE will result in a
lower ELCC
91

 ELCC of the solar site

for all the bins
throughout the year.
 Solar PV has high
output on summer
afternoons, resulting
in high ELCC.
 Coincidence of high
solar power output
and high system
LOLE will result in a
higher ELCC.
92
Example of ELCC Calculation for “Seasons” & “Entire Year” (Source: Ame Olson, 2015)
 Incremental ELCC measures the additional ELCC

provided by adding new resources to the portfolio.
 Wind resources have higher ELCC in winter than
in the summer.
 Solar PV resources have high summer value due
to coincidence of output with peak load needs, but
very low winter value due to night-time peak loads.
93

ELCC from Loss of Load Expectation (LOLE) for Different Wind/Solar Additions
 Example of reliability study results obtained from a generation adequacy study. The targeted
LOLE criterion is set at 24 hours per year (i.e. a reference case without VRE addition).
 Figure shows how the
addition of VRE
capacities (convention
generations and load are
kept the same for all
cases) would decrease
the LOLE of the system.
 However, the lower
marginal decrease can
be clearly observed at
higher penetrations
(saturation effect).
94
ELCC of PV Translated to Peak Load Reduction = Capacity Credit
 Figure illustrates how PV can reduce the net load by generating electricity during the hours
of peak demand (typically 4 to 6 pm).
 This results in a
capacity credit; the
ability of PV to
replace conventional
generation.
 However, the figure
also shows how the
capacity credit of PV
can decrease as PV
penetration increases.
(Source: Paul Denholm, 2016)
95

Capacity Credit (ELCC) of Solar PV Declines as Penetration Increases
 Solar PV has impact on reducing peak and shifting the “net peak” to a later hour in the day.
 This shift reduces the coincidence of the solar profile and the net peak such that additional
solar resources have a smaller impact on the net peak.
96
Wind Power Capacity Credit Reduces with Increase in Wind Power Penetration
 Meaning that every
MW of new wind
plant will replace
less conventional
generation capacity.
 Capacity credit is
quantified by
determining the
capacity of
conventional plants
displaced by wind
power while
maintaining the (Source: EWEA, 2010)
same reliability level.

 @ MW power 97

Wind Power Capacity Credit Reduces with Increase in Wind Power Penetration
 Meaning that every
MWh of new wind
plant will replace
less conventional
generation capacity.
 Capacity credit is
quantified by
determining the
capacity of
conventional plants
displaced by wind
power while
maintaining the (Source: IEA, 2016)
same reliability level.

 @ MWh energy 98
Solar Power Capacity Credit Reduces with Increase in Solar PV Penetration
 In figure, the dashed
line is average capacity
credit, solid line is
incremental capacity
credit @ MWh energy.
 Large range of capacity
credit estimates due to
both methodological
differences and the
regional variation in the
coincidence of solar
generation with peak
(Source: Andrew Mills, 2012)
load. (Note: this is also

true for wind power
Note: Dashed lines indicate total capacity credit, while solid lines indicate marginal capacity credit.
capacity credit estimate)
99

Reliability Values (ELCC) of Different Resources vs Penetration Levels
(Source: Nick Schlag, 2020)
 ELCC
captures
saturation
effects with
increasing
resource
penetration.
100
Illustration of Reliability Values (ELCC) of Different Resources (ERCOT in US)
 Reliability value (%) for different resources with factors considered in their reliability values.
(Source: E3, 2021) 101


Increase Capacity Credit (ELCC) of Wind Power by Using Hydro Coordination
 Pumped hydro coordination or use of large-scale energy storage can help to increase the
capacity value of wind power generation.
321 321
Wind power installed capacity of 1600 MW is equivalent to ELCC of 321 MW (no storage consideration)
102
Increase Capacity Credit (ELCC) of Solar PV by Adding Energy Storage
PV + Storage: Independent PV + Storage: Loosely Coupled PV + Storage: Tightly Coupled
(Florida Municipal Power Pool) (Jacksonville municipal utility)

Some of the JEA peak hours occur in the winter, when solar
production is lower and less able to fully recharge storage.
Installed Capacities: Installed Capacities:

Solar PV 100MW, Storage 100MW Solar PV 100MW, Storage 100MW
103
INSTRUCTOR : WIJARN WANGDEE (Source: Andrew D. Mills, 2020)

Reliability Contribution (from ELCC Perspective) in Adding Energy Storage
104
INSTRUCTOR : WIJARN WANGDEE (Source: Paul Denholm, 2018)
ELCC Captures Saturation Effects at Increasing Solar and Storage Penetrations
 Solar PV and other variable RE

resources (e.g. wind) exhibit
declining ELCC value due to
variability of power generation
profiles.
 Energy storage and other energy-

limited resources (e.g. DR, hydro)
exhibit declining ELCC value due
to limited ability to generate power
output over sustained periods.
(Source: E3, 2020) 105

Diversity Benefit (ELCC Increase) by Interactive Effect of “Solar + Storage”
 Resources with complementary characteristics produce the interactive effect, synergistic
interactions described as "diversity benefit".
 As penetrations of VRE and energy storage
grow, the magnitude of these interactive
effects (diversity benefit) will increase and
become non-negligible.
(Source: E3, 2020)
106
Diversity Benefit (ELCC Increase) by Interactive Effect of “Solar + Storage”
 0.6 GW ELCC for

solar plus 1.6 GW
ELCC for storage
equals 2.8GW!
 That’s an extra
0.6 GW!
(diversity benefit).
107
INSTRUCTOR : WIJARN WANGDEE (Source: Direct Testimony of Nick Schlag in Nevada Public Utilities Commission Docket 20-07023, 2020)

Capacity Credit (ELCC) of Energy Storage with Different Duration Capabilities
 ELCC of a 4-hr storage begins to decline at 1,200 MWh (or 300 MW, i.e. 30% of peak load).
 ELCC of a 8-hr storage begins to decline at 3,200 MWh (or 400 MW, i.e. 40% of peak load).
(Source: PNM 2020-2040 Integrated Resource Plan, 2020)
(Example of 4‐hr Storage from the left figure)
(Source: Derek Stenclik, 2018)
Capacity value of storage with increasing penetration Resource variability requires an improved forecasting
108
Capacity Credit for VRE: Adequacy vs. Security
Security Constrained Adequacy Evaluation System Well-Being Analysis
Healthy: Degree of Security

All equipment and operating constraints are
within limits and there is sufficient margin
to serve the total load demand even with the
loss of any element (i.e. deterministic N-1
criterion).
Marginal: Degree of Insecurity

The system is still operating within limits, but
there is no longer sufficient margin to satisfy
the accepted deterministic criterion.
At Risk: Degree of Inadequacy

Equipment or system constraints are violated
Well-Being Framework and some load is curtailed. (Source: Wijarn Wangdee, 2012) 109
Healthy State Marginal State At Risk State
Cliff of “Loss of Load”
“It is difficult if not impossible to determine where you want to go if you do not know where you have been.”
– Dr. Roy Billinton, Canada
110
INSTRUCTOR : WIJARN WANGDEE (Source: Wijarn Wangdee, 2012)
Security Constrained Adequacy Evaluation System Well-Being Analysis
RBTS
PH = Probability of Healthy State

PM = Probability of Marginal State
PR = Probability of At-Risk State
Considerations:
PH + PM + PR = 1 • Random generator outages plus an N-1 criterion for online generators
• All scenarios can maintain the same adequacy level (PR = 0.00043)
 Wind generation cannot secure the system in
the same way as conventional generation.
 Greater wind diversity required less installed
capacity to maintain system adequacy, but
adversely lowering system security.
111

“Adequacy-based ELCC” vs. “Security-based ELCC” for RBTS (small system)
Adequacy‐Based Contribution
Conventional unit: similar capacity con-

tributions for adequacy and security.
Wind power: different contributions for
adequacy and security.
Both adequacy and security capacity con-

tributions can be increased by increasing
diversity (more wind farm locations).
Security‐Based Contribution
112
“Adequacy-based ELCC” vs. “Security-based ELCC” for IEEE-RTS (large system)
Adequacy‐Based Contribution
Adequacy contribution dominates the
security contribution at low wind penetration.
However, adequacy contribution declines
while the security contribution increases at
high wind penetration.
Implying that the system capacity reserve

(adequacy-based) contributed at high wind
penetration is transformed to play a role in the
system operating reserve (security-based).
Security‐Based Contribution
113
Generation & Transmission Adequacy for VRE

Geographical Diversity:
 Generation Resource Planning
 Identify sources of renewable
energy for all regions
 Data Collection (i.e. monitor
wind speed for a few years)
 Create Renewable Energy
Zones (clusters).
 Transmission System Planning

 Identify POI for RE clusters.
 Static & Dynamic Simulation
to identify system limitations
(Source: www.bchydro.com) 114
Firm Capacity Estimation : Generation Adequacy 0.05
Peace Region wind speed profile (winter)
0.04
(Wind Power Modelling Required)
Probability
0.03
0.02
 Wind speed time series 0.01
0
 Probability distribution of wind speed 0 2 4 6 8 10 12 14 16 18 20 22 24
Wind Speed (m/s)
 Wind farm power curve Wind Farm Power Curve

1.0
 Multi-state wind power output
Power Output
0.8
(p.u.)
0.6
0.4
25 Mean(x) = 7.4 m/s, Mean(y) = 8.8 m/s, Cross Correlation in December = 0.64 0.2
Bear Mountain Area (x) 0.0
Wind Speed (m/s)
20 Mount Wartenbe Area (y) 0 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Wind Speed (m/s)
15
10
0.20
5 Capacity Factor = 0.33
Probability
0.15
0
0.10
Dec.01
Dec.02
Dec.03
Dec.05
Dec.06
Dec.08
Dec.09
Dec.11
Dec.12
Dec.13
Dec.15
Dec.16
Dec.18
Dec.19
Dec.21
Dec.22
Dec.24
Dec.25
Dec.26
Dec.28
Dec.29
Dec.31
0.05
Time Stamp in December (10 minute interval) 0.00
48
192
418
734
1077
1385
1551
1600
MW 115

Firm Capacity Estimation : Generation Adequacy (Using ELCC Concept)
 ELCC approach is based on incremental load perspective.
 ELCC concept is to evaluate how much
additional load can be served after a new
generation resource is added to the system
without exceeding the specified reliability level.
System Peak Load (MW)
10028
10078
10128
10178
10228
10278
10328
10378
9878
9928
9978
6.00
10199 MW
5.00 Base Case (No Generation Addition)
4.00 Adding 1600 MW Wind Power
3.00
2.00 Specified reliability level
1.00
0.00
ELCC = 321 MW
116
Transmission Capacity Requirement : Transmission Adequacy
 Using deterministic N-1 planning criterion, but augmented by a probabilistic method.
 Using ELCC value (or higher) as an equivalent
conventional generation, and then gradually
increase transfer capability until the reliability
is no longer changed.
1.40
1.30
1.20
1.10
1.00 Base Case (no generation and load added)
0.90 Load plus ELCC, adding 1600 MW wind
0.80 Load plus ELCC, adding 340 MW hydro
0.70
0 340 680 1020 1600
Increased TC (MW)
117

Transmission Capacity Requirement : Transmission Adequacy
 Chronological Characteristics of Transfer Path Considering 1,600MW Wind Power.

118
Transmission Capacity Requirement : Cost-Benefit Study for VAr Compensation
119
Items Required for Value-Based Reliability Analysis

3 Must-Have Items for Value-Based Reliability Applications (Planning&Operation)
1. Relevant Data Sets (Database)
 Equipment Models
 Forced & Planned Outages
 Forecast & Historical Data (load, VRE)
2. Reliability Analysis Tools
 Reliability Planning (i.e. HL-I, HL-II)
 Operational Reliability (i.e. Unit Commitment Risk,
Spinning Reserve Requirement, Response Risk)
3. Interruption Cost (Customer Damage Cost)
120
Reliability Database Management System (RDMS) @ BC Hydro
 Control Room Operating Window (CROW):
feeding the outage information commented by
system operators, i.e. start/restored times,
no.of customers affected, causes of failure, etc.
 Portal Data Warehouse (PDW): providing asset
information, i.e. name, type, vendor, ISD, etc.
 E-meter and Plant Information (PI): providing
substation load curve (from SCADA/EMS).
 Site Information System (SIS): providing single-
line diagrams for substations and networks.
 Delivery Point (DP) model, equipment model,
and load curve model are needed for calculating
reliability indices. (Source: Wenyuan Li, 2007)
121

Example of Historical Outage Data (Transmission Lines)
(Source: CEA, 2010)
122
Analysis Tools: Reliability Software Integrated Package (RSIP) @ BC Hydro
 MECORE – Bulk electric system (generation & transmission)
 MCGSR – Generation system (load & resource balance)
 SDREP – Station and Distribution systems
 NETREL – Radial and parallel system
 MEANLIFE – Equipment end-of-life
 SPARE – Spare equipment
 PLOSS – Network energy loss
RDMS
Reliability Decision
Management System
123

Analysis Tools: Reliability Analysis Functions in DIgSILENT PowerFactory
124
Analysis Tools: Framework for Operational Reliability Management Applications
RE Forecast
Center
Wind, Solar, DGs
EMS/SCADA
(Equipment status,
Automated
Main Engine Risk &
Gen outputs, loads) DIgSILENT
Script App.
PowerFactory
(Combine data, Cost
Create scenarios, Indices
RDMS (SE, OPF, ED,
Collect & combine
Reliability Database Unit Commitment,
results, Calculate Reporting
Management System Reliability indices)
additional indices)
(Gen & Load Models)
(Interruption Cost)
125
Chapter 3
Chapter 3: Power Grid Flexibility for VRE
 Impact of Variability & Uncertainty
 Need for Grid Flexibility
 Increase in Grid Flexibility
126
Impact of Variability & Uncertainty
Change in Daily Load Curve (Net Load) with High Solar PV Integration
from “Camel Curve” to

“Duck Curve”
(Source: Andreas Poullikkas, 2016)
 With High PV generation, the net daily

load curve will be reduced during the
daytime as it is displaced by solar PV
generation. Thus, conventional
generation could be at minimum
output during the peak PV generation.
127

Significant Ramping Capability Required for High Solar PV Penetration
 The ramping challenge is well illustrated by the "duck curve" for California ISO, where it
expects a ramping need of over 13,000 MW in 3 hours (5pm-8pm) by year 2020.
 Moreover, the
behind-the-meter
residential solar,
which today cannot
be dispatched or
even accurately
measured by the
system operator,
will increase the
ramping uncertainty.
(Source: Brendan Pierpont, 2017)
128
Example of Variability and Uncertainty
 Variability is the expected changes in VRE output.
 Uncertainty is the change in VRE output that is unexpected.
 The variability and uncertainty on systems is what causes the need for Operating Reserves.
(Source: Erik Ela, 2011)

129

Are There Any Generation Patterns from VRE Sources?
 Wind and solar sources is weather-dependent, but typical patterns in wind and solar
availability exist.
 Wind seasonal generation patterns vary by region. Solar daily generation patterns for
photovoltaic panels can vary based on the direction they face.
 In a typical day, wind generation
peaks overnight and during early
morning hours, and solar generation
peaks in early afternoon.
 Over the year, wind generation
typically peaks in winter, while solar
generation typically peaks in summer.
 Note: Load demand also shows
variability in a form of daily and
seasonal patterns.
(Source: Ashley J. Lawson, 2019)
130
Daily Load and Wind (Source: MIT, 2011)
Generation Profiles in
the California System
Compared to Average
Profiles
 Challenges are that
wind power output can
change more rapidly
and over a wider range
than load, and its output
does not conform to
daily cyclic patterns
(unlike solar PV output).
131

Seasonal Wind and Solar Generation Profiles in Thailand
 There is very good seasonal diversification across these two forms of renewable energy.
(Source: IES, 2016)
132
Variability of VRE : Aggregation of Solar Power
 The lack of correlation in
changes solar over short time
scales means that the variability
of the aggregated multiple sites
is significantly smoother than the
variability of an individual site.
 Five closest sites: 50 – 170 km
apart
 All 23 sites: 20 – 440 km apart
(Source: A. D. Mills, 2011)

133

Variability of VRE : Aggregation of Solar Power
 For the aggregated output based on geographical diversity of VRE, system operators can
respond to a much “smoother” generation profile with less amount of operating reserves.
134
Seasonal PV
Daily Output
Profile
 Shows max,
min, median,
and middle
50% range
for a Georgia
Site (in US)
(Source: Tom Key, 2014) 135


Solar Resource
Calendar –
1MW Output
Power
 December
2011:
Tennessee
1MW PV
System
Power
(Source: Tom Key, 2014)
136
Categories for
Daily Variability
Conditions
 Sandia’s
variability
index (VI)
and clearness
index (CI) to
classify days
(Source: Tom Key, 2014)

137

Variability of VRE : Aggregation of Wind Power
(Source: W. Krewitt, 2011)
138
Variability of VRE : Aggregation of Wind Power
 Aggregation and geographic diversity reduces the variability of wind power.
(Approximately 8 hours)
(Source: Michael Milligan, 2016)
139

Wind & Solar PV Output Variability from EGAT (3-5 October 2018, 6-second resolution)
140
Major Differences between Wind and Solar PV
141
INSTRUCTOR : WIJARN WANGDEE (Source: ESMAP, 2019)
Need for Grid Flexibility

The Role of Flexibility on VRE Integration
(Source: Emanuele Taibi, 2018) 142

What is Baseload? : Economically Inflexible Generation
(Source: Andrew Barbeau, 2017)
Traditional Approach to Planning High Renewable Penetration Scenario
143

Flexibility Issues in a System with High Penetration of Solar PV
144
What is Flexibility?
 Flexibility is the ability of a power system to respond to change in demand and supply.
 In other words, flexibility is the ability of a system to respond to a range of uncertain future
states by taking an alternative course of action within acceptable cost threshold and time
window.
 Four key elements
 Time
 Action
 Uncertainty
 Cost
 Serve as a basis for constructing measures of flexibility
145

Flexibility with regard to “Time”
 Indicate how fast the system re-acts to a disturbance and restore the system to its normal
state (i.e. seconds, minutes, hours, days, months, and years).
 A system might have sufficient capacity to cope with demand growth in a year, but not
enough capability to adapt to hourly load fluctuations.
Flexibility with regard to “Action”

 The corrective actions
that can be taken within
the response time window.
(Source: Jinye Zhao, 2016) 146

Flexibility with regard to “Uncertainty”
 The lack of complete information of the future system state
 Component failure, load forecast errors, renewable generation variability
 The magnitude of uncertainty determines how much flexibility a system requires to handle
uncertainty.
Flexibility with regard to “Cost”

 The response cost depends on the corrective action.
 Cost threshold
 Some corrective actions become uneconomical, and will not be considered.
 Minimize response cost
 The most economic corrective actions are sought in response to uncertainty
147

Types of Power Grid Flexibility
 The ability of a power system to accommodate the changes in electricity demand is often
expressed in terms of its “flexibility”.
 It refers to the ability of the grid and generation fleet to balance supply and demand over
multiple time scales, which is important with increased variability and uncertainty of net load.
 There are 4 types of power grid flexibility.

148
The Need for Power Grid Flexibility
Sources of Flexibility:
 Grid infrastructure (Source: John Moura, 2018)
 Dispatchable gen &

ancillary services
 Large-scale
energy storage
 Demand side
integration
149
Need for Grid Flexibility : Assessment

Future Generation Planning Process Shall Include System Operation Impacts
 Flexibility analysis must be additionally considered under significant variable RE.
• Generation model • Upward ramping capability
• RE, DG & Load models • Downward ramping capability
• Forced outage rates • Response risk • Unit commitment model
• Risk index, i.e. LOLE • Minimum generation levels • Unit commitment risk
• ELCC • Number of starts • Economic dispatch
reliable operable least-cost generation
(Source: IEC, 2012)
150
Future Generation Planning Process Shall Include System Operation Impacts
(Source: IEC, 2012) 151


Key Transition Planning Steps and Time Horizon
1. Generation
Expansion
Planning
2. Geo-Spatial
Planning
3. Dispatch
Simulation
4. Technical
Network
Studies
152
INSTRUCTOR : WIJARN WANGDEE (Source: Asami Miketa, 2017)
Key Transition Planning Steps and Time Horizon
153

Relevant Time-Scales of Electric Power Systems
 From grid operation standpoint, wind and solar power output variations are in the time-scale
ranging from second to days.
Operations Planning
“Dynamics”
(Automation)
(Source: Andreas Ulbig, 2014) 154

Different
Time-
scales of
Power
Grid
Flexibility
(Source: IEA, 2018b) 155

Need for Flex.

Impacts of VRE
at Various Time
Scales and
Relevant
Flexibility
Solutions
 Operation side
 Demand side
 Supply side
 Energy storage
156
INSTRUCTOR : WIJARN WANGDEE (Source: IRENA, 2018b)
Need for Grid Flexibility : Operating Reserve
Range of Relevant Time-Scales for Real-Time Operation of Power Grids
 VRE such as wind and solar power are the source of operational challenges due to the fact
that the time-scales of their VRE output variability, i.e. seconds to hours, coincides with the
time-scales of generator dynamics as well as frequency and power control.
(Source: Alexandra von Meier, 2006)
157

Time Scale for Different Operation Mechanisms and Reserves
 The goal of power system operation is to balance supply and demand at all times while
maintaining reliability and system security standards. The process is categorized as follows:
 A few seconds for inertial
response and governor,
 1–5 minutes for
automatic generation
control (AGC) and
regulation,
 5–30 minutes for
dispatch or load
following, and
 30 minutes to some
hours for replacement
and supplementary (Source: The World Bank, 2013)
reserve.
158
Ancillary Services Differentiated by Response Timeframe
 Normal Operations (Source: Brendan Kirby, 2013)
 Regulation
 Load Following
 Contingency Operations
 Spinning Reserve
 Non-Spinning Reserve
 Replacement or
Supplemental Reserve
159

Operating Reserve Categorization : Non-Event vs Event Reserves
(Source: Paul Denholm, 2016) 160

Non-Event Reserves
(Source: Erik Ela, 2011)
 Regulation (regulating) is done by

centralized control center sending
out control signals to generators
equipped with AGC to adjust their
dispatch set points.
 Load following is the action to
follow a trend of load pattern
within the day. This is usually
performed by economic dispatch
and may involve the starting and
stopping of quick-start combustion
turbines or hydro facilities.
161

Event Reserves: Loss of Generator
 The sequential actions of primary,
secondary, and tertiary frequency
control restore system frequency
following the sudden loss of
generation.
 Primary frequency control = (Source: Joseph H. Eto, 2010)
governor droop control response

 Secondary frequency control =
generators on automatic
generation control (AGC)
 Tertiary frequency control =
generators through generation
dispatch
162
Actual Example of Event Reserves
Event: The loss of a large power plant (1,200 MW) in France with the contingency reserve
activations.
 Primary reserve:
all Europe
 Secondary reserve:
all AGC in France
 Tertiary reserve:
some units in/near
France
163
Need for Grid Flexibility : Ramping Capability

Technical
Studies and
How They
Address Key
VRE Integration
Challenges
 The purpose of
these studies is
to support
planning
decisions and to
avoid technical
issues in real-
time operation or
frequent
activation of
remedial actions. (Source: IRENA, 2018a)
164
Different Time Scales for Flexibility Requirements
165

Time scales and Functions of the Three Different Frequency Controls
(Source: Shin-ichi Inage, 2019)
Note: • Governor Response or Governor‐Free (GF) Control

• Load Frequency Control (LFC) = AGC
• Economic Dispatch Control (EDC)
166
Regulation Requirement
 Solar PV output can drastically
change within a matter of minutes
due to cloud cover.
 Generator under AGC is referred
to as “Regulating Reserve".
 Automatic generation dispatches
through AGC are applied in every
4-6 seconds.
 Although the regulation dispatch
is done every 4-6 seconds, the
regulation margin must be adequate
to meet deviations within a 5-minute
dispatch interval as illustrated in the
red shaded area. (Source: NERC, 2013)
167

Load-Following Flexibility Requirement
 Load following requirement is
shown as the blue shaded area.
 High penetration of VRE often
results in large intraday ramps.
 More VRE leads to steeper
ramps, and thus require flexible
resources to follow the ramps.
 Two key components are required to
ensure that the net load can be met:
̶ Accurate VRE forecast
̶ Fast-acting capability resources
(Source: NERC, 2013)
168
Ramping Capability to Cope with VRE Outputs
 System operators must accurately follow load and minimize inadvertent energy flows.
(Source: NERC, 2013)
 This requires
the ramping
capacity in
both speed
and quantity,
which is
dictated by
how fast and
how much
the VRE
patterns
change.
169

Probabilistic Assessment of Capacity and Ramping Requirements
Assessment of
the capacity
and ramping
requirements is
performed using
a specially
developed
probabilistic
algorithm based
on a histogram
analysis
incorporating all
sources of
uncertainty and
parameters of a
continuous and
discrete nature. (Source: P.V. Etingov, 2012)
170
Impacts of VRE on Coal-Fired Power Plants
(Source: L.L. Sloss, 2016)
171
Need for Grid Flex.

Key Characteristics of
Power Generation
Sources
(Source: Power Advisory LLC, 2017) 172

Comparison of the Ramp Rates of Different Power Plants (Supply Adjustability)
(Source: Shin-ichi Inage, 2019)
 In general, hydro plants have
the fastest response times
and are able to change from
full power to zero and vice
versa within one minute.
 On the other hand, coal
thermal power plants
respond comparatively slowly.
Using a suitable combination
of these power plants yields
the optimal frequency.
173

Flexibility Characteristics of Controllable Generators
 Technical minimum: % of nameplate capacity below which the generator cannot be operated.
 Note: Heat-rate curves (showing how fuel consumption is impacted by the output power level) or
start-up costs are also required when cost estimations are expected.
174
Technical Comparison between Gas-Fired and Coal-Fired Power Technologies
175

Impacts of VRE on Coal Unit Cycling
 Cycling
(referred to
thermal
generators)
implies to
changing
the output
of a power
plant by
starting up,
shutting
down,
ramping up
or ramping (Source: L.L. Sloss, 2016)
down.
176
Increase in Cycling and Ramping Costs
(Source: L.L. Sloss, 2016)
177

How Many MW of Conventional Regulation Resource Could 1 MW of Fast
Regulation Replace?
 Conventional generators takes time to either increase or decrease their power outputs according
the ramp rates.
 In contrast, energy storage resources like batteries or flywheels have nearly instantaneous ramp
rates, i.e. effectively changing 100 percent of total power capacity per minute.
 This implies that a 1 MW BESS can equivalently provide the same regulation service as 1.7 MW
of an average hydropower plant, 2.7 MW of an average combustion turbine, or 29 MW of an
average steam turbine.
(Source: Thomas Lee, 2017) 178

Summary and Findings for Ramping Capability with High VRE
 Specific VRE integration issues depend on complex interactions between load profile, generation profiles,
and system capacity and transmission constraints.
 For example, solar is more predictable than wind, but also more intermittent (quicker change).
Solar generation increases ramping needs more than wind generation
 VRE increases ramping needs  California’s infamous “duck curve” features midday solar over-
generation followed by steep ramping requirements for evening peak.
 Flexible capacity  Dispatchable hydro, gas turbines, curtailable VRE, energy storage, demand response,
and enhanced coal/nuclear can all provide fast-ramping capacity.
 System flexibility constraints:
 Increased cycling of existing flexible generation (hydro, gas) is cheapest way to meet ramping
requirements
 Retrofit of existing baseload generation (coal, nuclear) may increase flexibility
- Lower minimum unit output (Pmin), increased ramp rates, reduced start-stop times, etc.
 Over-generation:
 At low levels, curtailment has minimal cost. It can also be used to minimize system ramping needs
 Exporting to neighboring grid is cheaper than building new storage capacity
179
Increase in Grid Flexibility

Interaction of VRE Properties and Power Grid Flexibility
 Key point: The integration challenge is determined by the balance between properties of
VRE and grid flexibility.
(Source: IEA, 2016)
180
Relative Economics of
Grid Flexibility Options
for High VRE Integration
 Flexible solutions by
system operation and
service from VRE are
relatively cheaper than
infrastructure and
technological options
(i.e. load, flexible
conventional generation,
networks, and storage).
(Source: Sandro Benz, 2015)

181

Technical
Options to
Increase
Grid Flexibility
 Supply side
 Demand side
 Energy storage
 Grid infrastructure
 Operation side

182
Increase in Grid Flexibility : Grid Expansion
Expanded Balancing Areas = Increasing Diversity (Reducing Variability)
(Source: Han Jiang, 2019)
(Source: Tirathai, 2017)

183
Increase in Grid Flexibility : What Operator Can Do

Sources of Power Grid Flexibility from Operators Perspective
 Operational
flexibility can be
obtained on the
generation side
and on the
demand side.
 In addition, VRE
in-feed can also
be curtailed or
limited below its
max output.
 Furthermore,
stationary
storages and EV
are well suited
for providing
operational
flexibility.
184
INSTRUCTOR : WIJARN WANGDEE (Source: Andreas Ulbig, 2014b)
Increase in Grid Flexibility : From Operator Side
High Penetration of VRE Could Be Achieved IF:
 Increase use of sub-hourly scheduling
 Increase utilization of transmission
 Enable coordinated commitment Flexibility Resource Costs
and dispatch over wider regions Associated with Increased VRE
Penetration
 Use forecasts in operations
 Increase flexibility of
dispatchable generation
 Commit additional operating
reserves as appropriate
 Implement/expand demand
response programs
 Require wind to provide down
reserves (Source: Aidan Tuohy, 2019)
185

How Does Faster Scheduling Support VRE Integration?
 Making scheduling and dispatch decisions closer to real-time reduces uncertainty and the
need for expensive ancillary services.
 Increase flexibility and reduce system costs
 Better alignment with the VRE timescale, enabling better utilization of wind & solar forecasts.
 Reduce wind and solar curtailment
Hourly scheduling Sub-hourly (5-min) scheduling

(Source: Michael Milligan, 2016) 186
Importance of Forecast: Effect of
Distributed Solar PV (DPV)
Generations on the Load Forecast
 Figure shows the impact of cloud
cover on load for similar days (in
terms of season, day of week, and
temperature).
 DPV has a significant impact on load
during the middle of the day.
 These effects can make it difficult for
load forecasting models to select
appropriately similar days.
(Source: Debra Lew, 2017)187


Importance of Forecast: Resource Modeling and Forecasting Time Scales
 Solar and wind power forecasting strongly relates to the ability to predict temperature,
sunshine and wind conditions.
 Weather services can offer reasonably good forecasts for larger areas within a resolution
of hours to days, uncertainty ranges however increase significantly for very local forecasts.
 Real-time forecasting tools for wind speed, temperature, solar insolation, down to the time
scale of minutes. This will facilitate operator action items to compensate for variations.
188
Importance of Forecast: Wind Forecasting Method
The forecast process
uses information from
multiple systems,
which include:
 Site geo-location,
Meteorological tower
geo-location
 EMS (i.e. resource status,
telemetered site specific
meteorological data)
 Outage Scheduler (i.e. derate
start/stop dates, derate values)
(Source: Sandip Sharma, 2017)

189

Importance of Forecast: Solar Prediction
for Different Time Scales
 Overview of key modelling steps
in solar PV power prediction
 Solar forecasting techniques
for different timescales
NWP = Numerical Weather Predictions

SCADA = Supervisory Control And Data Acquisition
190
INSTRUCTOR : WIJARN WANGDEE (Source: ARENA, 2016)
Importance of Forecast: Cloud Pattern Recognition and Optimization (Expert EMS)
The forecast process
framework is as follows:
 Characterize Cloud
Cover Days.
 Criteria is built to identify
each defined day type.
 Optimization routine
 Weather forecasts to
identify next day cloud
state
 Identify day type and
adopt optimal values
(Source: Johan Enslin, 2017) 191


Importance of Forecast: Solar Forecast Performance as Mean Absolute Error (%)
(Source: ARENA, 2016)
192
Wind Forecast Performance
(Day-Ahead & Hour-Ahead)
 Shorter time forecast (i.e. hour
ahead) can be more accurate
(less error).
(Source: Julia Matevosyan, 2017)
 With better information and improved

tool, the forecast tends to be more and
more accurate over time.
193

Importance of Forecast: Probabilistic Display Concept of Capacity Requirements
 The light and dark orange areas represent the evaluated capacity ranges for different
confidence levels. The blue curve is the generation schedule.
 The gray area is the available
balancing reserve, calculated
from the online generators.
 Adequate balancing reserve
achieved if the gray area covers
the entire net load uncertainty
range (the orange area).
 If deficiency of balancing reserve
occurs, it is necessary to commit
or additional generators to achieve
the desired confidence level for the
balancing reserve. (Source: P.V. Etingov, 2012)
194
Importance of Forecast: Capacity Requirements Uncertainty Evaluation (CAISO)
 The scheduling process for
real-time economic dispatch
(RTED), real-time unit
commitment (RTUC), day-
ahead market (DAM) utilizes
different forecasts depending
on different resolutions and
time horizons.
 Therefore, these forecasts
have different accuracies,
statistical characteristics,
and uncertainty ranges
associated with them
 Different shades of each
color indicate different (Source: P.V. Etingov, 2012)
levels of confidence.
195

Increase in Operating Reserve Requirement as More Wind Power Added
 Dena and
California (Source: EWEA, 2010)
considers
day-ahead
uncertainty.
 Others
considers
four-hour
ahead
uncertainty.
196
Increase in Operating Reserve Requirement as More Wind Power Added
 Reserve depends
on the time (Source: IEA, 2016)
horizon.
 A 24 hour ahead
forecast requires
a larger reserve
increase than
that of a 4 hour
ahead forecast.
 Small reserve
increase required
for an hour ahead
uncertainty.
197

Probabilistic Operating Reserve Assessment Associated with VRE
(Source: M. A. Matos, 2011)
198
 Example of Generator Outage Statistics*
All Generating Unit Types, 2002-2006 Statistics Fossil Units – Coal, 2002-2006 Statistics
Failure Rate Repair Time Year of Failure Rate Repair Time
Unit Type FOR (%) FOR (%)
(λ in f/yr) (r in hrs) Service (λ in f/yr) (r in hrs)
Hydro Unit 1.97 2.30 42.22 6th – 10th 2.00 6.23 25.97
Fossil Unit 7.32 10.36 52.64 11th – 15th 2.06 6.71 24.91
Nuclear Unit 7.64 2.24 229.33 16th – 20th 3.76 8.24 37.26
CTU (Gas Turbine) 29.78 10.82 141.44 21st – 25th 4.26 9.36 37.51
Note: 26th – 30th 6.61 10.32 47.89
FOR = Forced Outage Rate (i.e. Unavailability) 31st – 35th 9.26 13.97 48.70
Definition of FOR used by CEA is not equal to . 36th – 40th 12.90 16.03 56.86
41st – 45th 12.69 11.85 67.12
*Source: Canadian Electricity Association (CEA) : ERIS – Generation Equipment Status Report 2006
199

 Convolution Method
for Forecast Errors
 "Native Load
Forecast Error" is
convoluted with
"VRE Forecast Error"
to Obtain "Net Load
Forecast Error".
 Then, "Net Load
Forecast Error“ is
used in Probabilistic
Operating Reserve
Calculation
(Source: adapted from S. E. Haupt, 2019)

200
Unit Commitment Schedule

 Net Load = Native Load – VRE
 Longer forecast period (lead
time) = More forecast error
Outage Replacement Rate (ORR)
Unit Commitment Risk

INSTRUCTOR : WIJARN WANGDEE (Source: adapted from Y. V. Makarov, 2010)
Assessment 201

Example: At a given period, 7 generators Type Size No. of λ ORR for 2-hour lead time
shown in the table (200 MW in total) are (MW) Units (f/yr) (ORR = λt/8760 = λ×2/8760)
committed to serve the net load of 160 MW.
Thermal 10 2 3 0.000685
The outage replacement rates (ORR) in the
table are calculated for a 2-hour lead time. Thermal 20 3 3 0.000685
(Note: Net Load = Native Load – VRE) Thermal 60 2 4 0.000913
Case 1: Without forecast error on the net load, the risk (probability) = 0.001826
Case 2: With forecast error on the net load (normal distribution, SD of 8%), the risk = 0.008508
Case 3: To reduce the risk by committing another generator shown below (now 7+1 = 8 units),
with forecast error on the net load (normal distribution, SD of 8%), the risk = 0.001279
Type Size No. of λ ORR for 2-hour Total online generation = 200+15 = 215 MW
(MW) Units (f/yr) lead time (λt/8760) (increase in the operating reserve as a result
Hydro 15 1 2 0.000457 of VRE uncertainty)
202
Ramping Requirements vs. System Capability (Ramping Cap. Adequacy)
 Figure on the Right: A comparison of system ramping capability (blue curve) against
ramping requirements (red curve) for some operating hours.
 An area with insufficient system ramping
capability is highlighted in red. In other
words, there is certain probability that the
generation fleet does not have sufficient
ramping up capability to meet the demand.
Max. available system

capability in terms of
time and ramp rate
(Source: P. V. Etingov, 2012)

203

Ramping Capability Requirements  Response Risk
 The ability to respond to system changes and to pick-up load on demand depends very
much on the type of units being used as spinning reserve. Typically, the response rate may
vary from about 30% of full capacity output per minute for hydro-electric plant to only 1% of
full capacity per minute for some types of thermal plant.
 Therefore, the essence of allocating spinning reserve between units is to decide which of
the committed units should be dispatched and which should be held as reserve. These
decisions can be assisted by evaluating the probability of achieving a certain response or
regulating margin within the required response time. This assessment is know as response
risk evaluation.
204
Response Risk : Evaluation Technique
 The evaluation of response risk is similar to the evaluation of the risk in any generating
system. The problem is to evaluate the probability of achieving within the required response
time, the various possible output states of the units held in reserve.
 The evaluation should include the effect of the response rate of a unit and the probability
that the responding unit fail during the response period and therefore cannot contribute to
the required response.
 The evaluation method is best described by numerical example by evaluating the response
risk of the system with 7 committed units described in this example at various dispatch
patterns. Let the system load be 140 MW and assume the required response time of 5
minutes. The failure probability and response rate data are as follows:
Size No. of λ Probability of Failure in 5 minutes Response Rate
Type
(MW) Units (f/yr) (ORR = λt/8760 = λ×(5/60)/8760) (MW/minute)
Thermal 10 2 3 0.0000285 1
Thermal 20 3 3 0.0000285 1
Thermal 60 2 4 0.0000381 1
205

Response Risk : Generation Dispatch Pattern A
From the response rate data, only 5 MW can be obtained from each unit within the 5-minute
response period, given that the responded units do not fail during this period.
Dispatch A (Effect of relying on a single large unit spinning reserve): If a 60 MW of
spinning reserve has been committed and assume all this reserve be allocated to one of the
60 MW thermal units. These operating circumstances are shown in the table below
(where G = available generation, L = load dispatched on each unit, and R = the available 5
minute response of each unit. All values in MW).
Element Unit #1 Unit #2 Unit #3 Unit #4 Unit #5 Unit #6 Unit #7 Total
G 10 10 20 20 20 60 60 200
L 10 10 20 20 20 60 0 140
R 0 0 0 0 0 0 5 5
Only 5 MW of 60 MW spinning reserve can respond Response Risk of Dispatch A:

in the time period of interest when allocating all the Response Cumulative Prob.
spinning reserve to one thermal unit. If the response Individual Prob.
(MW) (Risk)
risk is defined as achieving a given capacity response,
5 0.9999619 1.0000000
the risk associated with this dispatch is very high.
0 0.0000381 0.0000381 206
Response Risk : Generation Dispatch Patterns B and C
Note: Unit commitment risk level of all Dispatches A, B and C are the same.
Dispatch B (Effect of distributing spinning reserve to two units):
G 10 10 20 20 20 60 60 200
L 10 10 20 20 20 30 30 140
R 0 0 0 0 0 5 5 10
Dispatch C (Effect of distributing spinning reserve to all units):

G 10 10 20 20 20 60 60 200
L 0 0 0 15 15 55 55 140
R 5 5 5 5 5 5 5 35
Response Risk of Dispatch B: Response Risk of Dispatch C:
Response (MW) Cumulative Risk Response (MW) Cumulative Risk Great advantage is gained
10 1.0000000 35 1.0000000 by distributing the spinning
5 0.0000762 30 0.0002187 reserve between available
0 0.0000000 25 0.0000000 units.
The distribution is not only increases the available response capacity, but also decreases the risk associated with a given response requirement.
207

Response Risk : Generation Dispatch Patterns D and E
60MW hydro unit replaces a 60MW thermal unit. Response rate of hydro unit = 20 MW/minute.
Dispatch D (Effect of using a hydro unit for spinning reserve):
G 10 10 20 20 20 60 (Therm) 60 (Hydro) 200
L 10 10 20 20 20 60 0 140
R 0 0 0 0 0 0 60 60
Dispatch E (Effect of using hydro and thermal units for spinning reserve):
G 10 10 20 20 20 60 (Therm) 60 (Hydro) 200
L 10 10 20 20 20 30 30 140
R 0 0 0 0 0 5 30 35
Response Risk of Dispatch D: Response Risk of Dispatch E:
Response (MW) Cumulative Risk Response (MW) Cumulative Risk
Hydro unit can respond to
60 1.0000000 35 1.0000000
change very rapid and
0 0.0000095 30 0.0000476 therefore is useful as
A greater response can be achieved using Dispatch D, 5 0.0000095 spinning reserve unit.
but one hydro unit carries all spinning reserve shall
not be used as the risk of failing is high.
0 0.0000000 208
Increase in Grid Flexibility : From VRE Services
Solar Power Plant Evolution (Three Grid Phases)
 Basic Solar (Grid 1.0): this works as long as solar penetration is low of solar in a region.
 Grid Flexible Solar (Grid 2.0): high penetration of solar requires solar plants to contribute to grid
flexibility and reliability services.
 Firm Dispatchable Solar
(Grid 3.0):
combination of
solar with
stationary
batteries,
becoming
Swiss army
knife of the grid.
209
INSTRUCTOR : WIJARN WANGDEE (Source: SolarPower Europe, 2019)
Increase in Grid Flexibility : From VRE Services

Confidence in Solar Forecasts & Resulting Forecast Error Reserve Requirements
 Solar plant can provide power regulation up/down, spinning reserve for real-time operation.
 Confidence in lower and upper bounds of hour-ahead solar forecast can be used for:
 Setting headroom and footroom needs to cope with system-wide solar forecast error.
 Setting limits on how much the solar plant could be dispatched as a dispatchable plant.
(Source: E3, 2018)
(Headroom for Upward Ramping)
(Footroom for Downward Ramping)
210
Increase in Grid Flexibility : From Energy Storage
Energy Storage Application Time Scales
Energy storages have different time-scale targets that serve for different purposes. The
spectrum of time scales for different storage applications is shown in the figure.
 Months: seasonal energy storage  15-30 minutes: up- and down-regulation
 4-8 hours: demand shifting  Seconds to minutes: solar & wind output smoothing
 2 hours: supplemental energy dispatch  Sub-milliseconds: power quality adjustment

211

Classification of Energy Types and Storage Techniques
(Source: WWF Poland, 2019)
212
Energy Storage Technology Capabilities (in Time Scales)
(Source: Johan Enslin, 2017)
213

Energy Storage Application Requirements (in Time Scales)
(Source: Johan Enslin, 2017)
214
Examples of Suitable Storage Technologies for Power System Applications
(Source: Sandro Benz, 2015) 215


Energy Storage System Applications
(Source: Andy Tang, 2018) 216

Energy Storage System Applications : Ramp Rate Control
(Source: Andy Tang, 2018)
217

Energy Storage System Applications : Renewables Smoothing
218
Energy Storage System Applications : Renewables Firming
219

Energy Storage System Applications : Renewables Shifting
220
Ancillary Service Comparison: Pumped Hydropower Storage vs. Battery Storage
(Source: Ingela Tietze, 2017)
221

(Source: Ingela Tietze, 2017)
222
 Batteries are particularly well suited to fast response short term balancing requirements.
Larger storage capacities for longer term services are not currently common.
 On the other hand, Pumped hydropower energy stores tend to hold large volumes, have far
higher energy-to-power ratios and thus are able to provide longer term services, even
bridging prolonged periods of low VRE output.
 With high VRE penetration, the longer term services of energy storage are more critical.
 On the one hand, modern batteries will last longer if charging and discharging is done
incrementally, avoiding maximum charge and depletion.
 On the other hand, if pumped hydropower power is running on partial load, its efficiency is
being compromised.
 However, any storage technology will have to weigh up their technically preferred running
mode against grid requirements and related economic impacts.
 Thus a trade-off has to be made between maximum operating hours and optimum operational
loads.
223

Enhanced Flexibility Can Accommodate More PV Penetration (Less PV Curtailment)
 Original system is limited by “minimum generation level” from nuclear, coal, etc. (California)
 Considering a 1,290MW pumped-hydro storage addition with 11% of DR (shifting demand
to times of lower energy prices, at the period of low net demand and high PV output)
 This flexibility enhances the grid leading to a decrease in PV curtailment. This means that
additional 3,000MW PV generation can be further penetrated to the grid (allowing more PV).
224
Enhanced Flexibility Can Increase PV Capacity Credit (More Firm)
 Impact of high PV generation on net load shape may make it easier for load-shifting
technologies by "DR and storage" to provide firm capacity.
 The net load shape (with high
PV) has the narrowing of the
peak period, meaning the no.
of peak load hours is shorter.
 Instead of relying on pumped
hydro storage for carrying
many hours of peak load,
narrowing of the peak
demand period could allow
shorter-duration batteries to
provide the same level of (Source: Paul Denholm, 2016)
service under high PV
penetration. 225

Hydrogen Economy
 Electrolysis
 Methanation
Efficiency of
Power-to-Gas
Technologies
(Source: Michael Sterner, 2021)
226
Flex by Storage
Hydrogen Economy
(Power-to-Gas)
Note:
PtG = Power-to-Gas
PtH = Power-to-Hydrogen
PtM = Power-to-Methan
(Source: Michael Sterner, 2021)
227

Energy Flow in ESS & Efficiency
Charging eff. =
(Source: Omid Palizban, 2016)
Storing period eff. =
Discharging eff. =
Total energy storage efficiency =

228
Typical Performance Parameters and Investment Costs of Different Storage Technologies (Source: WWF Poland, 2019) 229
Training Course on “High Penetration of Renewable Energy and Impacts

on Power Grid Operation”
References :
1. Amornrat Limmanee (2015), "Solar Energy Technologies: Present and Future Direction", NECTEC, 2015.
2. O. E. K. Onwunta, (2014) "Modelling and Simulation of the Impacts of Distributed Generation Integration into the Smart Grid", Dotoral Thesis, Cape Peninsula University
of Technology, 2014.
3. K. D. G. Imalka Jayawardena et al. (2013), "Inorganics-in-Organics: recent developments and outlook for 4G polymer solar cells", Nanoscale, Vol.5, 2013, pp.8411–
8427.
4. NREL (2021), "Best Research Solar Cell Efficiencies", National Renewable Energy Laboratory, July 2019.
5. Gavin Conibeer (2007), "Third-Generation Photovoltaics", Materials Today, Vol.10, No.11, 2007, pp.42–50.
6. Aurélie Beauvais et al. (2018), "Grid Intelligent Solar: Unleashing the Full Potential of Utility-Scale Solar Generation", SolarPower Europe, 2018.
7. ITRPV (2021), "International Technology Roadmap for Photovoltaic (ITRPV): 2020 Results", 12th Edition, 2021.
8. Susana Almeida de Graaff (2018), "System Operations Future Challenges due tothe Energy Transition" SEPOPE, 2018.
9. Mark Wehling (2021), "Global Energy Storage Outlook", BloombergNEF, 2021.
10. Raeey A. Regassa (2017), "Impact of Increasing Amounts of Distributed Generation on Transient Behavior of the Distribution System", Ph.D. Thesis, Georgia Institute of
Technology, May 2017.
11. K. Lappalainen (2017), “Output Power Variation and Mismatch Losses of Photovoltaic Power Generators Caused by Moving Clouds”, Publication 1506, Tampere
University of Technology, 2017.
12. EIA, "Solar photovoltaic output depends on orientation, tilt, and tracking", Energy Information Administration (EIA), November 19, 2014.
13. Ghaeth Fandi (2017), "Intelligent Distribution Systems with Dispersed Electricity Generation", Doctoral Thesis, Czech Technical University in Prague, 2017.
14. M.G. Molina et al. (2011), "Modelling and Control Design of Pitch-Controlled Variable Speed Wind Turbines" in Wind Turbines, 2011.
15. Thobias Sach et al (2019), "Auctions for the support of renewable energy in Germany", December 2019.
230
References :
16. IRENA (2017a), “Renewable Energy Outlook: Thailand”, International Renewable Energy Agency, Abu Dhabi, 2017.
17. IEA (2018a), “Thailand Grid Renewable Integration Assessment: Final Report 2018”, International Energy Agency.
18. IRENA (2018a), "Transforming Small-Island Power Systems: Technical Planning Studies for the Integration of Variable Renewables", International Renewable Energy
Agency, 2018.
19. Andreas Ulbig (2014a), "Operational Flexibility in Electric Power Systems", Doctoral Thesis, ETH Zurich, 2014.
20. PIER (2007), "Intermittency Analysis Project: Final Report", Public Interest Energy Research (PIER) Program for California Energy Commission, July 2007.
21. IRENA (2017b), "Planning for the renewable future: Long-term modelling and tools to expand variable renewable power in emerging economies", International
Renewable Energy Agency, 2017.
22. IRENA (2017c), "Grid codes as enablers of the energy transition: Scaling up Variable Renewable Power", World Future Energy Summit, 2017.
23. Wijarn Wangdee (2017), “Distributed Generation Systems”, Textbook Publishing Center KMUTNB, 2017. ISBN: 978-616-455-208-1.
24. Lion Hirth et al. (2013), “Control Power and Variable Renewables: A Glimpse at German Data”, SSRN Electronic Journal, January 2013.
25. Falko Ueckerdt et al. (2015), "Analyzing major challenges of wind and solar variability in power systems", Renewable Energy, Vol.81, 2015, pp.1-10.
26. Wolf-Peter Schill (2013), “Residual Load, Renewable Surplus Generation and Storage Requirements in Germany”, Deutsches Institut für Wirtschaftsforschung (DIW
Berlin), 2013.
27. Simon Müller (2015), “Flexible Power Systems to Integrate Large Shares of Renewables: The Value of Interconnections”, IEA-SGCC Dialogue on Global Energy
Interconnection, 21-22 July 2015, Beijing, China.
28. Arne Olson et al. (2015), “Capacity Contribution of Dispatch-Limited Resources”, Portland General Electric IRP Public Meeting #3, August 13, 2015.
29. Jaeseok Choi (2014), “Web Based Online Adequacy Reliability Information System for Power Systems Including Intermittent Resource Generators”, IEEE PES General
Meeting, July 27-31, 2014.
231

References :
30. E. Sainz et al. (2009), “Robust filtering for the characterization of wind turbines: Improving its operation and maintenance”, Energy Conversion and Managaement,
50:2136–2147, 2009.
31. Dinh-Duong Le (2013), “Impact of Wind Power Penetration on Power System Security by a Probabilistic Approach”, Doctoral Dissertation, Politecnico Di Milano, 2013.
32. Wijarn Wangdee et al. (2009), “Assessing Transfer Capability Requirement for Wind Power Generation Using a Combined Deterministic and Probabilistic Approach,”
Proceedings of 2009 IEEE PES General Meeting, July 2009, Calgary, Canada.
33. I. Kim et al. (2017), “Impact of Short-Term Variations in the Generation Output of Geographically Dispersed PV Systems”, Proceedings of the 50th Hawaii International
Conference on System Sciences, 2017.
34. Richard Perez et. al., "Spatial and Temporal Variability of Solar Energy", Foundations and Trends in Renewable Energy, Vol.1, No.1, 2016.
35. Tom Hoff, "CSI RD&D Project: Advanced Modeling and Verification for High Penetration PV", High Penetration Solar Forum, 2011.
36. Mark O’Malley (2011), “Harvesting Renewable Energy - The Grid Integration Challenge”, IEEE DLP, 2011.
37. Wijarn Wangdee et al. (2006), “Considering Load Carrying Capability and Wind Speed Correlation of WECS in Generation Adequacy Assessment,” IEEE Transactions
on Energy Conversion, Vol. 21, No. 3, September 2006, pp. 734–741.
38. Paul Denholm et al. (2016), "On the Path to SunShot: Emerging Issues and Challenges in Integrating High Levels of Solar into the Electrical Generation and
Transmission System", National Renewable Energy Laboratory, 2016.
39. EWEA (2010), “Powering Europe: wind energy and the electricity grid”, European Wind Energy Association, November 2010.
40. IEA (2016), "Design and operation of power systems with large amounts of wind power", IEA Wind Task 25, 2016.
41. Andrew Mills et al. (2012), An Evaluation of Solar Valuation Methods Used in Utility Planning and Procurement Processes, Lawrence Berkeley National Laboratory,
December 2012.
42. Nick Schlag (2020), "Resource Adequacy and ELCC Overview", PNM 2020-2040 Integrated Resource Plan, 2020.
43. E3 (2021), "The Load-Serving Entity Reliability Obligation: A Market Design Reform to Ensure Electric Reliability in Texas", Energy and Environmental Economics Inc.
(E3), September 2021. 232
References :
44. Wijarn Wangdee et al. (2010), “Coordinating Wind and Hydro Generation to Increase the Effective Load Carrying Capability,” the 11th International Conference on
Probabilistic Methods Applied to Power Systems (PMAPS 2010), June 2010, Singapore.
45. Andrew D. Mills et al. (2020), "A simple and fast algorithm for estimating the capacity credit of solar and storage", Energy, 210, 2020.
46. Paul Denholm et al. (2018), "The Potential for Energy Storage to Provide Peaking Capacity in California under Increased Penetration of Solar Photovoltaics", National
Renewable Energy Laboratory, 2018.
47. E3 (2020), "Capacity and Reliability Planning in the Era of Decarbonization: Practical Application of Effective Load Carrying Capability in Resource Adequacy", Energy
and Environmental Economics Inc. (E3), August 2020.
48. Derek Stenclik et al. (2018), "Energy Storage as a Peaker Replacement", IEEE Electrification Magazine, September 2018.
49. Wijarn Wangdee et al. (2012), “Probing the Intermittent Energy Resource Contributions from Generation Adequacy and Security Perspectives,” IEEE Transactions on
Power Systems, Vol. 27, No. 4, November 2012, pp. 2306–2313.
50. Wenyuan Li et al. (2007), “Reliability Decision Management System: Experiences at BCTC”, the 2007 Canadian Conference on Electrical and Computer Engineering,
22-26 April 2007, Vancouver, Canada.
51. CEA (2010), “Forced Outage Performance of Transmission Equipment”, Canadian Electricity Association, 2010.
52. Wijarn Wangdee et al. (2013), “Probabilistic Transmission Planning at BC Hydro: Method and Example”, the 2013 IEEE PES General Meeting, July 21-25, 2013,
Vancouver, Canada.
53. Andreas Poullikkas (2016), "Fundamentals of Electricity Systems and Economics", IET, 2016.
54. Brendan Pierpont et al. (2017), "Flexibility: The path to low-carbon, low-cost electricity grids", Climate Policy Initiative, 2017.
55. Erik Ela et al. (2011), "Operating Reserves and Variable Generation", National Renewable Energy Laboratory, 2011.
56. Ashley J. Lawson (2019), "Variable Renewable Energy: An Introduction", Congressional Research Service, 2019.
57. MIT (2011), "The Future of the Electric Grid", Massachusetts Institute of Technology, 2011.
233

References :
58. IES (2016), "Alternatives for Power Generation in the Greater Mekong Sub-Region Volume 5: Power Sector Vision for the Kingdom of Thailand", 2016.
59. A. D. Mills et al. (2011), "Implications of geographic diversity for short-term variability and predictability of solar power", IEEE Power and Energy Society General Meeting,
2011.
60. Tom Key (2014), "Grid Integration of Distributed Generation and Storage", IEEE PES General Meeting, 2014.
61. W. Krewitt et al. (2011), "Integration of Renewable Energy into Present and Future Energy Systems", Cambridge University Press, 2011.
62. Michael Milligan (2016), "Big, Fast, and Flexible: Grid Operations for Efficient Variable Renewable Integration", Greening the Grid, 2016.
63. ESMAP (2019), “Grid Integration Requirements for Variable Renewable Energy” Energy Sector Management Assistance Program (ESMAP), World Bank, 2019.
64. Emanuele Taibi (2018), "Flexibility assessment of future generation mix using the IRENA FlexTool", Copenhagen, 25 May 2018.
65. Andrew Barbeau (2017), "Panel: Glimpse into the Future", NextGrid Illinois, 2017.
66. IRENA (2018b), "Power System Flexibility for The Energy Transition - Part 1: Overview For Policy Makers", International Renewable Energy Agency, 2018.
67. Jinye Zhao et al. (2016), "A Unified Framework for Defining and Measuring Flexibility in Power System", Optimization and Equilibrium in Energy Economics Workshop,
2016.
68. John Moura (2018), “Accommodating High Levels of Distributed Energy Resources: A Bulk Power System Reliability Perspective”, EPRI Transmission & Distribution
Coordination Workshop, February 8, 2018.
69. IEC (2012), “Grid Integration of Large-Capacity Renewable Energy Sources and Use of Large-Capacity Electrical Energy Storage”, International Electrotechnical
Commission, 2012.
70. Asami Miketa (2017), "Planning tools for long-term power system expansion with a higher share of variable renewable energy", International Renewable Energy Agency,
2017.
71. IEA (2018b), "Status of Power System Transformation 2018: Advanced Power Plant Flexibility", International Energy Agency, 2018.
72. Alexandra von Meier (2006), "Electric Power Systems: A Conceptual Introduction", Wiley-IEEE Press, 2006.
234
References :
73. The World Bank (2013), “A Guide to Operational Impact Analysis of Variable Renewables: Application to the Philippines”, May 2013.
74. Brendan Kirby (2013), "The Value of Flexible Generation", PowerGen 2013.
75. Joseph H. Eto et al. (2010), "Use of Frequency Response Metrics to Assess the Planning and Operating Requirements for Reliable Integration of Variable Renewable
Generation", Lawrence Berkeley National Laboratory, 2010.
76. Shin-ichi Inage (2019), "The Role of Large-Scale Energy Storage Under High Shares of Renewable Energy", in Advances in Energy Systems: The Large-scale
Renewable Energy Integration Challenge, John Wiley & Sons, 2019.
77. NERC (2013), "Maintaining Bulk Power System Reliability While Integrating Variable Energy Resources – CAISO Approach", November 2013.
78. P.V. Etingov et al. (2012), "Online Analysis of Wind and Solar Part I: Ramping Tool", Pacific Northwest National Laboratory, 2012.
79. L.L. Sloss (2016), "Levelling the intermittency of renewables with coal", IEA Clean Coal Centre, 2016.
80. Power Advisory LLC (2017), "Integration of Variable Output Renewable Energy Sources – The Importance of Essential Reliability Services", Energy and Mines Ministers’
Conference, 2017.
81. Thomas Lee (2017), "Energy Storage in PJM", Kleinman Center for Energy Policy, July 27, 2017.
82. IEA (2016), "System Integration of Renewables: Implications for Electricity Security", International Energy Agency (IEA), 2016.
83. Sandro Benz (2015), "Grid-Scale Battery Storage for Variable Renewable Electricity in Sweden", Master Thesis, Lund University, 2015.
84. Tirathai (2017), "ASEAN Power Grid", Tirathai Journal, Vol.6, Issue 18, Apr-Jul 2017.
85. Han Jiang et. al. (2019), "Study of future power interconnection scheme in ASEAN", Global Energy Interconnection, Vol.2 No.6, December 2019, pp.549-559.
86. Andreas Ulbig et al. (2014b), "Role of Power System Flexibility", a book chapter in Renewable Energy Integration, Elsevier, 2014.
87. Aidan Tuohy (2019), "Storage and Demand-Side Options for Integrating Wind Power", in Advances in Energy Systems: The Large-scale Renewable Energy Integration
Challenge, John Wiley & Sons, 2019.
235

References :
88. Debra Lew et al. (2017), "The Power of Small: The Effects of Distributed Energy Resources on System Reliability", IEEE Power & Energy Magazine,
November/December 2017.
89. Alexandra von Meier (2014), "Challenges to the Integration of Renewable Resources at High System Penetration", California Institute for Energy and Environment, 2014.
90. Sandip Sharma (2017), "Challenges and Solutions for System Operations with High Share of RE– ERCOT (Texas)", Electric Reliability Council of Texas, 2017.
91. ARENA (2016), "Australian Solar Energy Forecasting System", Australian Renewable Energy Agency, 2016.
92. Johan Enslin (2017), "Center for Advanced Power Engineering Research", Clemson University, 2017.
93. Julia Matevosyan (2017), "ERCOT Renewable Integration", 2017.
94. M. A. Matos et. al. (2011), "Setting the Operating Reserve Using Probabilistic Wind Power Forecasts", IEEE Transactions on Power Systems, Vol. 26, No. 2, May 2011,
pp.594-603.
95. S. E. Haupt et al. (2019), "The Use of Probabilistic Forecasts", IEEE Power & Energy Magazine, November/December 2019, pp.46-57.
96. Y. V. Makarov et al., "Incorporating Wind Generation and Load Forecast Uncertainties into Power Grid Operations", Pacific Northwest National Laboratory, January 2010.
97. SolarPower Europe, "Grid Intelligent Solar: Unleashing the Full Potential of Utility-Scale Solar Generation in Europe" 2019.
98. E3 (2018), "Investigating the Economic Value of Flexible Solar Power Plant Operation", Energy and Environmental Economics, Inc., 2018.
99. WWF Poland (2019), "Available and Future Methods of Energy Storage: Report 2020", World Wide Fund for Nature (WWF), 2019.
100. Andy Tang (2018), "Creating Value with Energy Storage", Flexible Power Symposium 2018, 13 June 2018.
101. Ingela Tietze et al. (2017), “Comparing Pumped Hydropower Storage and Battery Storage – Applicability and Impacts”, Euro-Asian Journal of Sustainable Energy
Development Policy, Volume 5, Issue 2, 2017, pp.16-29.
102. Michael Sterner et al. (2021), "Power-to-Gas and Power-to-X—The History and Results of Developing a New Storage Concept", Energies, Vol.14, 2021.
103. Omid Palizban et al. (2016), "Energy storage systems in modern grids - Matrix of technologies and applications", Journal of Energy Storage, Vol.6, 2016, pp.248–259.
236
My Contact Information
WIJARN WANGDEE (e-mail: [email protected])
Center of Excellence in Electrical Power Technology (CEPT)
Building 1 Room 107, Faculty of Engineering, Chulalongkorn University

Training2022 - VRE Utilization in Power Grids - Final

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Training2022 - VRE Utilization in Power Grids - Final

Uploaded by

Copyright:

Available Formats

Variable Renewable Energy (VRE)

Utilization in Power Grids

 RE Portfolio in Germany vs. Thailand

 Variable RE (VRE) Properties

Clear sky: Direct 85%, Diffuse 15%

RE Technologies : Solar Energy

RE Technologies : Solar Energy

(Source: O.E.K. Onwunta, 2014)8

Thailand DNI ≈ 1400 kWh/m2-yr

RE Technologies : Solar Thermal

CSP with Fossil Integrated Solar Combined

RE Technologies : PV (Source: NREL, 2021)

INSTRUCTOR : WIJARN WANGDEE

INSTRUCTOR : WIJARN WANGDEE

INSTRUCTOR : WIJARN WANGDEE

Trending: Improve an efficiency, then reduce the cost

(Source: Gavin Conibeer, 2007)

Poly c-Si Poly c-Si

Perovskite Solar Cells

Dye Sensitized (Organic PV)

Efficiency > 40% for Niche market,

“Solar power is not only cheaper

Silicon shortage as China

began the c-Si PV mass

Shipments at Avg. price at

(Source: Susana Almeida de Graaff, 2018) (Source: Mark Wehling, 2021) 21

Effect of Ambient Temperature @1000W/m2 Effect of Solar Irradiation @25°C

Effect of Ambient Temperature @1000W/m2 Effect of Solar Irradiation @25°C

(Source: Raeey A. Regassa, 2017)

 Tracking the sun maximizes solar power generation,

(Source: EIA, 2014)

Wind Power System

RE Technologies : Wind Power

Wind Turbine with Gearbox Gearless Wind Turbine

 Betz limit: theoretical limit up

RE Technologies : Wind Power

 Zone 3: Power delivered by the rotor is regulated

RE Technologies : Wind Power

RE Technologies : Wind Power

RE Technologies : Wind Power

RE Technologies : Wind Power

RE Portfolio in Germany vs. Thailand

(Source: Thobias Sach , 2019)

RE Portfolio in Germany vs. Thailand

RE Portfolio in Germany vs. Thailand

(Source: IEA, 2018)

Country Thailand Germany

Size (km2) 513,120 357,022

Variable RE (VRE) Properties

(Source: IRENA, 2018a)

(Source: Andreas Ulbig, 2014a) 53

VRE Properties (Source: PIER, 2007)

Analyses Required for VRE at

Variable RE (VRE) Properties

Variable RE (VRE) Properties

(Source: IRENA, 2017c) 58

(Source: IRENA, 2017b)

 High VRE Required Probabilistic Analysis

 Capacity Credit (Equivalent Firm Capacity) for VRE

 Generation & Transmission Adequacy for VRE

 Items Required for Value-Based Reliability Analysis

(Source: Wijarn Wangdee, 2017)

Uncertainty in Power Grids