Lecture #2
1/February/2024
The Traditional Electricity Grid
• Transmission Lines typically run long distances required for high voltages, with complex
meshing to enhance power-flow control, increasing power transfer margins, flexible
compensation, with some degree of freedom for frequency and voltage variations.
• Distribution Systems have conventionally been configured in radial configuration with short
lines and simple controls, with power flowing in only one direction: from the substation to
the client.
• T&D are shifting away from traditional power system towards distributed generation based
on power electronics. In order to achieve a smarter grid, we have to design or redesign T&D
to face complex challenges, as generators of all types and sizes with increasing amounts of
power electronics are installed in a distributed fashion.
Hosting Capacity Approach
• Amount of generation that can be connected to a distribution feeder is called “hosting
capacity”
• The hosting capacity is the maximum amount of generation that can be integrated into the
power system, while still maintaining its performance within acceptable limits.
• The hosting capacity approach uses the existing power system as a starting point and considers
the way in which distributed generation changes the performance of the system when no
additional measures are taken.
• For this, a set of performance indicators is needed. It is a normal procedure in the power quality
area, recently adopted for dispersed generation.
• Usually, the grid operator sets certain requirements for the power quality injected by the DG.
• Steady-State AC analysis for sizing, power system management with use of Thévenin Equivalent in several PCCs
• Power Flow Analysis uses optimization with large linear systems algebraic solutions
• Transient Analysis and Real-Time Control were not important for the traditional power grid (big change for 21st century !)
• Unidirectional power flow would make the system centralized and totally managed by utilities and their subsidiaries
• Users were loads and should follow imposed quality metrics by the power companies
Overloading and Increased Losses for Transmission
• In order to estimate the amount of distributed generation that will cause
overloading and increased losses, hosting capacity must be identified.
• The first hosting capacity is when the maximum generation is more than the sum
of maximum load and minimum load. When this level is exceeded for the load and
generation downstream of a feeder section, a more detailed study is needed. Note
that in most cases no immediate mitigation methods are needed after there is a
certain margin between the maximum consumption and the ampacity (current
carrying capacity) of the feeder section.
• The second hosting capacity is reached when the sum of generation and minimum
load exceeds the ampacity of the feeder section. In this case, mitigation actions are
needed, such as enforcement of the feeder section or curtailment of the generation
during periods with low consumption.
Minimum Consumption
• The amount of generation that can be connected to a distribution feeder (the
“hosting capacity”) is directly dependent on the minimum consumption.
-- Every 1 W reduction in minimum consumption gives 1 W reduction
in the hosting capacity --
• However, minimum consumption in a traditional distribution network is mostly not
known. Before the introduction of DG the minimum consumption was never
considered, even for the first design of the grid. The maximum consumption used to
be the most important consideration for a distribution feeder without any DG.
• “After-Diversity Maximum Consumption” per customer plays an important role in
determining the required rating of the components.
• For DG it is important to make analysis of issues such as “after-diversity maximum
production,” “after-diversity minimum consumption,” and “after-diversity
minimum production.”
Redundancy and Meshed Operation
• Many distribution networks are meshed in their structure but operated radially. This
allows for interruptions to be limited in duration by means of switching actions.
• To ensure that backup operation is possible, the network should also in that case not
be overloaded.Consider the simple network below, during backup operation, the
network is also operated radially, so the expression:
still holds, with generation and load the sum of the two feeders:
Redundancy and Meshed Operation – Cont.
• Transmission and distribution networks are operated meshed, so that the loss of a
component does normally not result in an interruption for any of the customers.
• Radially operated meshed network: distribution feeder with
possibility for backup supply
Redundancy and Meshed Operation – Cont.
• Meshed operation has some potential impacts for the risk of overloading. In a
radial network, distributed generation will always result in a reduction in the
network loading, as long as the amount of distributed generation is less than the
“first hosting capacity.”
• For a meshed network, as below, when a large power flow takes places between
two regions; it may result in a large power flow in the connection A–B and a
lesser flow in the loop A–C–D–B.
Meshed Networks - common in transmission and
distribution grids - problem of overloading and losses
can become complicated
• The connection of production to the distribution network will reduce the load. The
same holds for the connection of a large wind park to a substation to which mainly
consumption is connected. In general, the reduction in loading for one or more
individual buses will result in a reduction in the power flows through the network.
• There are situations where the reduction in minimum load at certain buses directly
limits the ability of the subtransmission or transmission grid to transfer power. When
power transfer through the region is important, it may not be possible to connect any
distributed generation to those buses. The same may be the case when already a
large amount of generation is connected to a subtransmission bus. That generation
may have taken up all hosting capacity, leaving no further space for distributed
generation.
• Consider the network of a meshed subtransmission
network. A linearized DC loadflow is used to obtain an
expression for the power flow through connection C–D due
to a large power transfer through the region. Neglecting
the resistance, the active power transferred over a line is a
function of the angular difference between the voltages on
both sides of the line :
Example of a meshed subtransmission network
𝒔𝒔𝒔𝒔𝒔𝒔 𝜽𝜽𝒊𝒊𝒊𝒊
𝑷𝑷𝒊𝒊𝒊𝒊 = 𝑽𝑽𝒊𝒊 𝑽𝑽𝒋𝒋
𝒛𝒛𝒊𝒊𝒊𝒊
• with zij the impedance of the line. For not too large angles, we can
linearize this expression into the following one:
• For connection C–D of the previous meshed network, the power can be
calculated as :
• For the other connections, similar expressions are obtained, where it has
been used that the flow through A–C is the sum of the flow through C–D
and the load at bus C, and that the flow through D–B is the difference
between the flow through C–D and the load at bus D:
When the flow through C–D is limiting the power transfer through the region. The
equation above shows a maximum permitted power through A–B that will not result
in overloading of connection C–D.
• Increased load at bus C will increase the transfer capacity;
• Increased load at bus D will decrease the transfer capacity.
• Connection of generation to bus C will reduce the transfer capacity;
• Connecting generation to bus D will increase the transfer capacity.
• Power transfer through a region might take place in both directions.
• The result is that the hosting capacity at both buses can be low, or even zero.
Generation System
• Generates three-phase power at 4–25 kV,
• Interfaces with transmission system through the generating station where
voltage is stepped up to 115–765 kV range.
Transmission System
• Meshed network transports 3-phase power at 115–765 kV interfaces with
other transmission lines or sub-transmission lines at switching stations,
• switching stations connecting to sub-transmission system step down voltage to
the 20–69 kV range.
Distribution System
• Mostly radial system transports power through primary (3-phase) or secondary
(1-phase) feeders to customers,
• Connects to the sub-transmission system through the distribution substation
where the voltage is stepped down to 33 kV and below.
Distributed Generation System Control and Data Acquisition
• Generally, problems caused by renewable DGs arise because of the intermittent
nature of their output or the uncertainty of information at the PCC. Not particularly
because they are renewable energy sources !
• Most of DG units are connected to low voltage distribution networks and not to high-
voltage transmission grids.
• Most of these units are not subject to central system control, partly because of the
high cost of conventional telemetry.
• System control and data acquisition (SCADA) has low penetration at voltages below
132 kV.
• At the sub-transmission or distribution levels, the traditional electricity grid operate
largely with incomplete and uncertain information.
Power Quality
• PQ concerns the electrical interaction between the network and its customers.
• It consists of two parts: the voltage quality concerns the way in which the supply
voltage impacts equipment and the current quality concerns the way in which the
equipment current impacts the system.
• Most of the traditional power systems view of PQ emphasizes the voltage quality
- voltage dips and interruptions being dominant.
• The customers may be consumers (“end users”), producers, or both, called by
prosumers.
• The customers in this course are primarily considered to be distributed
generators.
Voltage Quality
• The voltage quality at the terminals of the generator is determined by
other equipment connected to the grid and by events in the grid.
• Any generator unit is affected by the voltage quality in the same way as
any other equipment on that feeder.
Voltage Quality – Cont.
• The impact of voltage disturbances includes a reduction of the lifetime of
equipment, erroneous tripping of the equipment, and damage to equipment.
• An important difference between generator units and other equipment
connected to the grid is that the erroneous tripping of generator units may pose
a safety risk: the energy flow is interrupted potentially leading to machine
overspeed and large overvoltages with electronic equipment.
• Guidelines for the immunity of generator units against voltage disturbances
requires a distinction between variations, normal events, and abnormal events.
voltage variations and their duration with corresponding
effect on the power system
Current Quality
• The generator current impacts the network. This current disturbance can affect
other customers.
• Distributed generation can also be used to improve current quality in a
controlled way ---- This requires to control the current magnitude, the phase
angle of the current compared to the voltage, and their current waveform.
Current Quality – Cont.
• For high power applications, synchronous machines can be used for improving
voltage and current quality.
• We have nowadays the area of Grid-Forming Converters
• For the distribution system, fast control of distributed generation with power
electronics interface can improve:
• the current quality, with active power filters (shunt)
• the voltage quality, with dynamic voltage restorers (series)
How Can We Visualize the Power
Grid Evolution ?
• The power system and electricity needs were simpler in the last 100
years.
• The grid was designed to deliver unidirectional electricity.
• Power Systems Perspective
• equivalent linear load
• power-factor
• Thèvenin Equivalent for analysis
• Control systems based on steady-state
• Droop Control
• frequency as related to power flow
• Utilities and substations assume aggregated load
• Timespan in on the order of minutes to hours
Power Electronics Perspective
• Assume an equivalent sinusoidal voltage source at the point of
common coupling
• Study how power electronics circuits affect the current supplied by
the utility
• The control systems system design is for the instantaneous
operations of electronic switches (transistors, diodes, thyristors)
• Timespan in on the order of milliseconds down to microseconds
Research Paradigm Unified Smart-Grid: Power Systems, Power
Electronics, Power Quality, and Renewable Energy Systems
Power Electronics Power Systems
Smart-Grid
• Core competencies • Overlapping Skills • Core competencies
• Advanced Circuits, Electronics and Control • Steady-State and • Power Theory, Electrical Power Circuits
• Power Electronics and Power Quality Transient Analysis • Power Quality, Network Theory
• Computer Project Based Analysis and Simulations Design • Advanced Control • Computer Oriented PowerFlow, Voltage Stability
• Artificial Intelligence Analysis
1) Steady-state simulations:
•Power flow
•State estimation
•Steady-state security assessment
•N-1, N-1-1 contingency analysis 5) Power, energy management (in
•Fault analysis, fault location analysis seconds)
•Volt/VAr optimization •Automatic generation control (AGC)
•Symmetrical components and unbalanced faults •Operator training simulator (OTS)
•Induction motor analysis •Reactive power control
•Device coordination and selectivity 3) Scheduling (in hours) •Capacitor and inductor storage for power
•Reliability assessment •Economic dispatch quality
•AMR/AMI integration •Optimal power flow (OPF)
•Power system markets •Unit commitment 6) Fast dynamic simulations (in
•Hydro-thermal Coordination milliseconds)
2) Long-term planning (in years) •Preventive Security Constrained OPF •Modal analysis of oscillations
•Production costing •Renewable energy storage •Oscillation damping control
•Model validation of generators •Weather modeling and predictions •PSS, FACTS controller tuning
•Reliability evaluation •Transient Stability
4) Slow dynamic simulations (in •Synchronous coupling
3) Scheduling (in hours) minutes) •Active power control
•Economic dispatch •Voltage stability •Speed control of electrical machines
•Optimal power flow (OPF) •Loadability limit calculations •Flux optimization
•Unit commitment •Thermal analysis •Voltage quality metrics
•Hydro-thermal Coordination •Battery storage •Islanding and reconnection control
•Preventive Security Constrained OPF
•Renewable energy storage 7) Transient control (in microseconds)
•Weather modeling and predictions •Pulse-width-modulation (PWM)
•Current quality metrics
4) Slow dynamic simulations (in minutes) •Synchrophasor integration
•Voltage stability •d-q, p-q , CPT control of machines
•Loadability limit calculations •Torque control of electrical machines
•Thermal analysis •Harmonic mitigation, active filtering
•Battery storage
Electrical Circuits with Hardware-in-the-Loop Real Time
Analysis, Modeling and Control
Slow Dynamics Fast Dynamics
“smart-inverter” or “multi-functional inverter”
for smart-grid applications there are additional needs
on top of control systems
It is necessary to incorporate bidirectional data communications with
the utility -- user friendly interfaces -- net-metering -- lead-follower multi
agents management control.
Flexible applications of inverter-based systems have
huge and big data. Such data is approached by
computer science, machine learning, and artificial
intelligence applications.
Data Pre-
Processing
Cyber Data
Security Compression
Power Quality
Big Data Engineering
Dynamic Cloud
Sampling Computing
Parallel
Processing
• The smart-grid is a two-way dialogue of electricity and information
exchanged between the customers and the utility.
• The future grid must be efficient, more reliable, more secure and
greener.
• A smart-grid enables new technologies to be integrated. Further wind
and solar energy production. Plug-in electric vehicle charging. The smart-
grid is the new infrastructure of today's grid.
• Enhanced services and functions make the definition of
multifunctional inverters or multi-task inverters
• These inverters are powerful devices, applied to small-scale
power generation
• Flexible power requires smart inverters
enhancing power quality and system reliability
AI in Power System Protection for SmartGrids is to achieve Self-Healing
• The system should have sensors, automated controls, and advanced software that
uses real-time distribution of data to detect and the isolate faults and to
reconfigure the distribution network to minimize the power outage and customer
impact.
• A self-healing grid has improved reliability. This can be accomplished by
reconfiguring the breakers, relays, and reclosers to quickly isolate the faulty
section of the feeder and re-establish the service to as many customers as possible
from alternate sources or feeders.
• Self-healing will have to reconfigure the system within a couple minutes.
Intelligent Control, fast real-time processing, high bandwidth communications,
machine learning on big data, synchronized measurements, hardware, software,
all must be engaged and designed for improved self-healing of SmartGrids.
• Big Data Engineering
• Parallel Processing
• Cloud Computing
• Cyber Security
Data Pre-
Processing
Cyber Data
Security Compression
Power Quality
Big Data Engineering
Dynamic Cloud
Sampling Computing
Parallel
Processing
Cybersecurity
Digital Twin – virtual model based RTS, allows a
dynamic, evolving, and an ‘intelligent’ entity so
that it changes over time as the physical system
evolves.
Physical Power System Digital Twin
A Digital Twin has the following attributes: U
Digital Control Room GUI
(i) a digital model in a simulated environment, Scenario 1
Scenario 2
(ii) the physical entity in real space, and Data
Exchange
U0 Data
Exchange
(iii) a connection between the virtual model
t
tpast DT Monitoring t actual DT Prediction t predict
and the physical entity for the data flow.
Malfunction Data Malfunction Data
Training Malfunction 1
U U
The figure shows a Cyber-Physical Surveillance Sensor
Malfunction 2
Malfunction 3 Malfunction U 0
Scenario 1
U0
Scenario 3
and Security Assessment, where a DT aided by Malfunction 4 data Scenario 2 Scenario 4
neural networks allows continuous data
...
tk tk+1 tk tk+1
Malfunction n
Application Malfunction Data
exchange between the cyber and physical Malfunction
Data Detection Operator
U
Scenario n
world. The DT is characterized by its ability to U0
monitor the physical system accurately and Maintenance action
adaptively on different scales of time, it can tk t k+1
be a part of the cyber-physical system, which
interacts with the physical entities, equipment,
environment, and humans.
Scientific Method - observing the real-world phenomena with data and analysis, gives a
model for the scientist or engineer, based on equations, graphs, diagrams. This
understanding will lead a decision-making process to handle the variables on the real-
world, in order to have a constructive and operational functional design.
Such ad-hoc observation is the BASIS for AI Applications in Power Systems
real
world
2
observation
decision
1 making
data
analysis
model
Questions, Thoughts, and Reflection
1.How can you compare the traditional (old) grid with the modern (smart) grid ?
2.Describe how you use steady-state AC analysis power system analysis, management, and
control (do you know Droop Control ?)
3.What is Power Flow Analysis, and the use of optimization with linear systems solutions ?
4.Why is the need of transient analysis and real-time control for microgrids and smartgrids ?
5.In a simple Meshed Network (A-B-C-D) describe the limits of power transfer through the
region. If you apply the same concepts for a Microgrid, what are the controllability issues and
possibly intelligence required for best performance ?
6.Regarding Power Quality, what is Voltage Quality, Current Quality, and how integrated
power electronics may enhance such functions ?
7.Discuss all the possible case studies, scenarios, and simulations possible for integrating slow
dynamics versus fast dynamics of electrical circuits with their possible hardware-in-the-loop,
analysis, modeling, and control.
8.How can you define Cybersecurity ? How do you define a Digital Twin ?
9.Based on all the eight answers above, write one page to describe what is the use and
application of Artificial Intelligence for Smart Power Systems.
