AI in Smart

Energy
Systems
Lecture 1

Sohail Khan
Contents
AI in Smart Energy Systems .................................................................. Error! Bookmark not defined.
1.1 Course Overview ................................................................................................................ 2
1.2 Instructor Overview ........................................................................................................... 5
1.3 Learning objectives ............................................................................................................ 5
1.4 What we expect from you .................................................................................................. 6
2 Course structure ........................................................................................................................ 6
2.1 The Digital Transformation of the Energy System ............................................................... 6
2.2 Computational Methods for Energy Networks.................................................................... 6
2.3 Decision Support in Integrated Energy Systems .................................................................. 7
2.4 AI-Based Data and Machine Learning Approaches .............................................................. 7
What is your Dream/Image of Integrated and Digitalized Energy System in your Opinion?......... 7
3 Lecture 1: Digital Technology And Digital Transformation .......................................................... 8
"Clearly, the thing that's transforming is not the technology - It's the technology that is
transforming you"...................................................................................................................... 8
3.1 Analogue vs Digital ............................................................................................................. 8
3.2 Digitalizing the power grid.................................................................................................. 9
4 Smart Cities ............................................................................................................................. 11
4.1 Introduction ..................................................................................................................... 11
4.2 Stakeholders .................................................................................................................... 12
4.3 Challenges........................................................................................................................ 12
4.4 Smart city example........................................................................................................... 12
4.5 Summary.......................................................................................................................... 13
5 Assignment 1 ........................................................................................................................... 13
1.1 Course Overview
This course will teach you how to digitalize the 'conventional' energy system and which digital
technologies shall play a key role. The digitized grid operations then open ways of the
application of AI, machine learning, blockchain and computer simulations. The objective of
this transformation is to make the energy system sustainable, affordable, available, and
secure. The transition is not easy. We need to incorporate renewable energy, electric
transport, heat and gas into our energy system.
The transformation has already begun with more people installing solar panels every day,
social adoption of the hybrid vehicles paving the way to electric vehicles and the advent of
new technologies like heat pumps. The primary driver behind is the economic factor but in
addition climate change is key driver.
Why do we hear more and more about the blackouts, energy shortages and capacity problems
in the energy system nowadays? And how do we leverage digital technology to tackle such
challenges?
If you are eager to discover the answers to these questions then you're in the right place this
course, the artificial intelligence in smart energy systems will provide you with insights to
overcome these challenges.
What are the two keywords of the digital transformation?
In this course you will learn about the opportunities of a digital grid and the various digital
technologies to achieve this. You will see how artificial intelligence and machine learning can
make the grid operations more efficient and more autonomous. Transforming human
operators into supervisors.
We discuss numerical simulators, virtual system models and digital twins, all helping in testing
the effects of these digital technologies on the real world, which eventually leads to better
design choices. You will learn where the grid might get vulnerable and how to protect it from
cyber-attacks.

The course consists of an introductory and four-course content modules, each touching upon
a different facet of the digital transformation of the energy system:

1. The Digital Transformation of the Energy System

2. Computational Methods for Energy Networks
3. Decision Support in Integrated Energy Systems
4. AI-Based Data and Machine Learning Approaches

All modules follow a similar structure. We start with a small introduction to the module. Then,
we dive deep in the contents of that module. Near the end of the module, we see industrial
applications and use-case students. Each module is closed with a module assignment.
1.2 Instructor Overview

Dr. Sohail Khan

Sohail Khan holds a Doctoral degree with distinction in the field of Electric Power Systems
from Technical University of Vienna, Austria, obtained in 2016. He is presently working as
senior faculty at Sino-Pakistan Center for Artificial Intelligence (SPCAI) at Pakistan-Austria
Fachhochschule-Institute of Applied Sciences and Technology (PAF-IAST), since October 2020.
Before that he worked as Scientist from 2016 to 2020 in the leading applied research institute
of Austria called Austrian Institute of Technology (AIT). There, he worked in the co-funded
European and customer-oriented projects in the areas of renewable energy integration in
medium and low voltage (MV, LV) distribution systems, frequency stability analysis of
transmission systems, virtual inertia support from bulk energy storage systems and integrated
energy systems (combining the district heating networks with power distribution networks).
He is actively working on using AI for energy forecasting, machine learning for dealing with
the uncertainty in energy systems and cross cutting applications using AI for decision support
in urban planning, agriculture, transport, and security related areas.

1.3 Learning objectives

In this course, you will see how digitalization will affect the energy system of the future. More
specifically, you will learn:

 To recognize the digital transformation in the energy sector, identify challenges and
solutions, and evaluate its impact on both the power system and society itself.
 To compare the different model types used for numerical simulations of energy
systems and evaluate the influence of individual parameters and system models on
the simulation performance.
 To explain the different objectives of decision-making in energy systems, and the
influence of different units and their properties on the decision making.
 To apply and evaluate machine learning methods for prediction and control in energy
systems
1.4 What we expect from you

As a Masters student, we expect you to be an active participant in this course by contributing

to a positive and inclusive atmosphere. We want you to question, share and help others by
engaging in meaningful discussions.

We expect you to keep on track with deadlines to benefit from learning within a community.
This course is meant to be a place where you learn with and from others. In this sense, we
would like you to experience collaboration through the exchange of ideas, so please make
sure you follow along with other participants to enrich the overall learning experience.

2 Course structure
The course is organized into 4 modules. A brief summary of each module is presented below.
Detailed instructions and resources will be provided during the course.

2.1 The Digital Transformation of the Energy System

In this module, you will learn how digitalization and the digital transformation is penetrating
the electricity grid and reinventing the way we operate our power system. We will discuss the
following elements:

 Digitalization and digital transformation

 Smart cities
 Digital twins
 Transactive energy
 Blockchain
 Data ownership in the digital power system

What are the new business opportunities in the digitally transformed energy system.

Finally, you will investigate how digitalization is penetrating and impacting your working
environment.

2.2 Computational Methods for Energy Networks

This module will present numerical simulators, computer programs that allow simulations of
power system behavior. You will learn the following:

 What is numerical simulation, and why do we need it?

 What methods are there for simulating the electrical power system?
 How can we couple different energy domains?
 How do different modeling approaches impact simulation performance?

There are multiple simulation exercises where you will be able to see for yourself how
different simulation types and methods, solver settings, and network models impact the
performance and result of the numerical simulation.
2.3 Decision Support in Integrated Energy Systems

The power system has different types of units, and operators have distinct goals when
operating their system. Traditionally, this was low cost, but nowadays, a greater emphasis is
placed on emissions and reliability. However, these are at a trade-off with each other,
meaning that no golden combination exist. In this module, you will be taught:

 What trade-offs there are in energy systems.

 What influence different units and their properties have on decision objectives.
 How capacity planning and other methods can be used in decision making.
 How to use the optimal scheduling problem to meet your decision objectives.

At the end of the module, you will get hands-on experience with the optimal scheduling
problem. You will see how the optimal schedule of a small system changes with different
model parameters and investigate how to achieve some objectives.

2.4 AI-Based Data and Machine Learning Approaches

In this module, you will learn how machine learning can be applied in the energy system,
allowing for more accurate predictions and automatic control of the power system. We will
discuss the following topics:

 Machine learning basics

 Forecasting with regression models, ensemble methods and neural networks
 Dynamic security assessment with classification and decision trees
 Identification of customer types using k-means clustering
 Anomaly detection using auto-encoders
 Surrogate modeling
 Learning control actions with reinforcement learning

Which sector of the economy do you think is most digitalized?

Primary Sector: This sector deals with the extraction and harvesting of natural resources such
as agriculture and mining.

Secondary Sector: This sector comprises construction, manufacturing, and processing.

Basically, this sector comprises industries that relate to the production of finished goods from
raw materials.

Tertiary Sector: Retailers, entertainment, and financial companies make up this sector. These
companies provide services to consumers.

Quaternary Sector: The final sector deals with the knowledge or intellectual pursuits
including research and development (R&D), business, consulting services, and education.

What is your Dream/Image of Integrated and Digitalized Energy System in your

Opinion?
3 Lecture 1: Digital Technology And Digital Transformation
"Clearly, the thing that's transforming is not the technology - It's the technology
that is transforming you"

-Jeanne W Ross, MIT Sloan Center For Information Systems Research

Digitalization is one of the biggest and most important revolutions of human life. But, ever
wondered what exactly is this digitalization..??

Consider an example, let's take the electricity bill payment. Before the digitization era, we
used to stand in a long queue for hours to pay the bill. But now, it can be paid with a touch
on your smartphone or even be automatically deducted from your account.

Not just in IT fields, but almost all the fields including the energy systems have been impacted
by digitization.

Let's take a deeper look at this...

3.1 Analogue vs Digital

Analogue: If you want to measure the size of a particular object, you could go near that object,
measure with your hands, and then return and – or better show – it's this size. Here, you
transform what it would like to know into something else that you can see or feel.

Digital: You could also use measuring equipment like scale, measure the object, and find out
that it is 'x cm'. This option is to convert it into numbers i.e. digital.

Key differences:

Analogue measurements are, in theory, continuous: you can do everything with your hands.
Contrastingly, Digital is discretized - the scale goes to millimeters, so 20.3, but there is no
20.314. You cannot make a continuous measurement.
The resolution of the analogue measurement is, in theory, infinite, and for digital, it is finite
because it is discretized. On the other hand, the robustness of an analogue measurement is,
of course, not that good.

There are always distortions, especially when measuring with my hands. When you walk
around with the measurement in your hand, your 20.3 centimeters have already changed to
something else. But the number 20.3, is intact. Only massive distortions might change one
digit of your measurement value during transmission.

Analogue measurements are also not that flexible. You cannot measure everything with your
hands. For example; how would you measure volume or speeds? With numbers (digital), you
can represent everything.

Finally, the costs favor digital measurements. Digital scales better and using digital
technology, you can transmit the data over a long distance for a low price.

3.2 Digitalizing the power grid

Let us analyze what this means for the power grid.

Previously, in the power grid, all measurements were analogue with 20mA current loops. They
were brought either to the substation control unit or to the control room with a dedicated
analogue line. In modern substations, every piece of equipment in the substation, for
example, transformers and circuit breakers, has a digital element that sends a digital packet
with the measurement numbers to the controllers, easy to change in software. So, the gain
in flexibility, efficiency and cost-effectiveness is evident with digital technology.

Compared to digital technology, the digital transformation is much more than just replacing
analogue sensors and measurement cables with digital protocols. It’s a transformation of the
businesses and processes enabled by digital technology. It’s also a change of roles, rules, and
mindset. In the end, It’s a change of society itself.

The fundamental property of such a digital transformation is that it uses digital platforms.
These platforms make everything easier: they bring people together, they align the data and
decisions, and by that, they tear down barriers. Some things are very complicated in the
analogue world, such as meeting 10,000 people, asking their names, making agreements and
contracts, or identifying people making a payment. These things are very tedious and don’t
scale well in the analogue world, but they are just a click away with a digital platform.

Examples of digital transformation can be found in almost all sectors, and now it’s also hitting
the power sector. It was difficult (and tedious) to become a taxi or a hotel company in the
past.

Now, you can just sign up to one of these platforms, and you are done. All the worries of
authentication, customer care, payment, documentation, taxes, or auditing are done and
solved by the digital platforms. The traditional barriers are gone. This is the digital
transformation: using digital technology to change an entire process or an entire sector.
However, for the power system and power grids, it’s a bit different. The blunt expectation is
that if we would have a digital transformation in the power system, its barriers will be torn
down again. In the extreme case, this would mean that the solar energy produced on my roof
with my photovoltaic panels can be sent tomorrow to you by email.

That’s what the digital transformation usually does: it frees you from all these analogue
boundaries. Unfortunately, electricity is a physical thing: we cannot fool physics. Thus, we
must see how far digital transformation can change the power system.

On one side, we could lower our expectations and say ok, this is just not possible: physics is
against it. But some of these constraints are because of laws and safety, which could be
changed with the digital transformation. So maybe, having these expectations of a digitally
transformed power sector will also change reality by making us think out of the box about
new solutions.

Summary

In this session:

 Analogue and digital measurements differ in many aspects, together with the pros and
cons of these two concepts.
 Digital substation is replacing all the analogue cables with just one digital one.
 Digital transformation is much more than just using digital cables and digital numbers.
It has platforms: it tears down barriers, it brings things together that were not so easily
together before and it all scales very well.
 Applying phenomenon of digital transformation to the power sector is challenging.
Still, it might lead us to an entirely new power system that is more capable and flexible
than the existing one.

Questions

Q1: In future substations, the majority of equipment will work with:

 analog sensors
 digital sensors
 hybrid sensors
 none of the above

Q2: The digital transformation changes:

 rules
 roles
 mindsets

Q3: Why is the digital transformation of power grids more difficult than in other sectors?

 there are more things that need to be transformed, resulting in a slow change
  the core product (electricity) is physical, meaning that the transformation might not go
as expected
 there are insufficient technologies available to facilitate the transformation
 there is no need for the digital transformation in the power grid

Discussion:

Future of Digital Meters, how it can support digitization of the power grid.

Which digital technologies you think can transform the energy sector

4 Smart Cities

4.1 Introduction

A smart city is a city where its assets and processes are equipped and enhanced with digital
technology. We put sensors in our garbage carriers, provide the city with a smart grid, and
improve all other environmental elements with digital technology. This in turn means that we
get a lot of data, and we want to use that data to make our lives more comfortable, more
efficient, and safer. In short, we want to make our lives better.

However, it is not only individual assets that become smart. In the end, this smart city
connects sectors that were probably not connected before. We hope to discover and enable
synergies between those processes and sectors.

Indeed, risks and threats can be found, understood, and hopefully mitigated in time, if the
right data is available.
4.2 Stakeholders

To get an overall idea of smart cities, we should know who is the most important stakeholders
and what their needs are.

The people who live in that city are the important stakeholder. Let’s not forget that the city
is made for humans to live in. They are the ones constantly interacting with the smart city;
they are permanently in the loop. It’s not just technology and objects that are part of the
smart city. No, people are equally, if not more important to look at. Besides, they also co-own
the assets and the data, which brings us to data ownership and privacy questions. In short,
we need a solution that serves and satisfies the needs of the people.

4.3 Challenges

There are some challenges in building a smart city. The root challenge starts with the
important stakeholder itself. We need to create awareness among the people—a challenge
we face when introducing the smart city concept. We got to have the people on board; they
have to accept it, like it, and be part of it.

To achieve this, we have to clearly show them the benefits out of the concept. Otherwise, we
don’t get acceptance for it. It is a fact that if you try to introduce some innovative technology
in a smart city without the people on your side, you can forget about it!

Finally, the participation of the city inhabitants, of the Public Administration, of everybody in
the loop is of utmost importance for smart city projects. 'It’s not a technology project; it’s a
people’s project'.

4.4 Smart city example

Aspern: It is a part of Vienna, Austria, wherein the next 20-30 years, an extensive smart city
development program is executed. It’s already running for a while, so the first phase of the
building is already there. They built apartments, schools, shops, an underground line, and
even smart electrical grids.

These buildings there have smart meters and sensors measuring temperature, air quality, and
power consumption. Moreover, there are lots of technology underground: batteries for
intelligent balancing of rooftop photovoltaics and other resources, and even heat storage.
And you, as an inhabitant, can participate in many things via your phone.

It’s an actual smart city with smart, digital technology, but the focus is really on the people.
The people living there are always involved, they are asked feedback on how they like it, and
they can participate and co-create the entire smart city.

Keep in mind: 'Smart cities are often full of technology such as sensors, actuators, wireless
networks and a lot of interconnected things, but the essential players in this entire story
are the people living and working there.'
4.5 Summary

 Smart cities are cities where we equip processes and assets in the city with digital
technology. By that, we get lots of data that can be used to optimize things or to
discover synergies between the elements of the city.
 The smart city connects things that were not connected before, and ideally, we use
this for something worthwhile. We use it to make the city more efficient, convenient,
and safer.
 Noteworthy are the people who do not only permanently interact with the city and all
its processes but who also co-own the assets. Thus, we have to have permission or
consensus to equip things with digital technology and to use potentially private
information.
 In the end, we have to involve the people from the very first moment when we do a
smart city project. Let them participate, let them share their ideas, and also let them
share their concerns. Otherwise, we will not get the required acceptance and
participation needed for a smart city. This is one of the most important challenges
faced in building a smart city.

5 Assignment 1
What is digitalization in your perspective and how it has or can transform you as individual in
a society?

How can digitalization support your energy requirements for daily use, mobility and work?

Please write a paragraph on both questions