D S S P P: Evelopment of A Upervisory Ystem For A Ower Lant

DEVELOPMENT OF A SUPERVISORY SYSTEM FOR A POWER
PLANT
FERNANDA MOURA QUINTÃO SILVA
MASTER’S THESIS DEFENSE DATE: 06/04/21

UNIVERSIDADE FEDERAL DE MINAS GERAIS – UFMG
ESCOLA DE ENGENHARIA
PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA ELÉTRICA - PPGEE
DEVELOPMENT OF A SUPERVISORY SYSTEM FOR A SMART

POWER PLANT
FERNANDA MOURA QUINTÃO SILVA
Master’s Thesis submitted to the Examination

Board assigned by the Colegiado do Programa
de Pós-Graduação em Engenharia Elétrica da
Escola de Engenharia da Universidade Federal
de Minas Gerais, as a requirement for the
acquisition of the Title of Master in Electrical
Engineering.
Concentration Field: Power Engineering

Line of Research: Power Electronics
Advisor: Prof. Dr. Thales Alexandre Carvalho

Maia
Co-Advisor: Prof. Dr. Igor Amariz Pires
BELO HORIZONTE
APRIL – 2021
Acknowledgments
I would first like to thank my thesis advisor Dr. Thales Alexandre Carvalho
Maia and my thesis co-advisor Dr. Igor Amariz Pires. They always steered me in the
right direction throughout the development of this thesis and offered the guidance
needed whilst allowing this work to be my own.
I would also like to thank Petrobras for the opportunity to join the project that
allowed me to deepen my knowledge in promising areas. I would also like to thank
Petrobras for the financial support.
I would also like to acknowledge the team involved in the project at

Universidade Federal de Minas Gerais for without their efforts I would not be able to
make progress in my own work.
Finally, I would like to express my gratitude to my parents, my sister, and to

André for always providing me with support and encouragement throughout my studies.
This accomplishment would not be possible without them. Thank you.
iv
Abstract
Throughout the years, there has been a steady technological advance in the
industrial automation field, that has propelled a constant evolution of the supervisory
and control systems employed to monitor and manage plants, including power plants.
Nowadays, with the coming of the fourth industrial revolution which has brought the
concept of the Internet of Things, it’s possible to apply this concept to create an
information sharing environment that simplifies and increases the effectiveness of
supervisory and control systems.
This work applies this research field by implementing a new supervisory system
using the ideas of the concept of Internet of Things in a real solar power plant. The
supervisory system developed in this work is capable of providing to a user a friendly
environment to monitor any desired measurement and/or status of the devices that
compose this power plant.
This dissertation presents the development of this supervisory system that uses
devices integrated with the Internet of Things that can be implemented in any type of
power plant. This work goes in detail about the tools and services that are available
when using these novel concepts and how it can be beneficial for the plant having a
supervisory system developed employing this technology and strategy. The benefits are
even more highlighted when a comparison between traditional supervisory systems and
the system developed here is presented.
v
Summary
1 Introduction .................................................................................................... 1
1.1 Objectives ............................................................................................... 2
1.2 Methodology ........................................................................................... 3
1.3 Scientific Contributions .......................................................................... 7
1.4 Text Organization ................................................................................... 7
2 Bibliographic Review ..................................................................................... 9
2.1 Micro-grids ............................................................................................. 9
2.2 Traditional Supervisory Systems .......................................................... 12
2.3 Internet of Things .................................................................................. 16
2.3.1 Smart Devices .................................................................................. 18
2.3.2 Key Challenges in IoT ..................................................................... 19
2.4 Open-Source IoT Tools for Supervisory System .................................. 20
2.4.1 Node-Red ......................................................................................... 20
2.4.2 Emoncms ......................................................................................... 23
2.5 Micro-grid Tesla ................................................................................... 24
2.5.1 Photovoltaic Solar Power Plant Tesla ............................................. 25
2.5.2 Energy Storage System .................................................................... 28
3 Supervisory System ...................................................................................... 31
3.1 Data Acquisition and Processing .......................................................... 32
3.1.1 Solar Power Plant ............................................................................ 32
3.1.2 Energy Storage System .................................................................... 40
3.2 Data Monitoring, Storage and Control.................................................. 42
3.2.1 Node-Red ......................................................................................... 42
3.2.2 Emoncms Dashboard ....................................................................... 42
3.3 Final Architecture ................................................................................. 45

vi
4 Results .......................................................................................................... 47
4.1 Solar Power Plant.................................................................................. 47
5 Conclusions .................................................................................................. 55
5.1 Proposed Continuation .......................................................................... 56
Appendix A – Emoncms Feeds .......................................................................... 57
Bibliographic References ................................................................................... 61

1 Introduction
The main problems with renewable energy sources are cost and availability, for
example, wind and solar power are not always available when and where needed – the
energy is not “dispatchable”. Daily and seasonal changes affect greatly the generation of
energy. Oscillations in wind speed and solar radiation would translate directly to
variations in energy production, reducing its predictability and resulting in an
intermittent generation (CAMACHO, SAMAD, GARCIA-SANZ, & HISKENS, 2010).
Control and energy storage have been used to mitigate this type of problem and
significant progress has been made throughout the years, making it possible to employ
this form of energy as a reliable and viable source for a cleaner generation. Therefore,
there has been an effort in order to better understand the possibilities of this energy
source besides the possible enhancements which can be done due to the advancement of
technology.
As an interested party in developing clean energy, the department of Electrical

Engineering of UFMG (Universidade Federal de Minas Gerais) concluded the
construction of the Usina Experimental Fotovoltaica TESLA Engenharia de Potência
with capacity of 37 kWp (FOUREAUX, MARRA, & ANTHONY, 2014). The
production of generated energy by this power plant is connected to the grid. The solar
power plant is described in greater details in a future section of this dissertation.
Alongside the rise in interest in renewable energy sources, it’s also noticeable a
technological advance in the industrial automation field. This field plays a vital part to
the monitoring and control of photovoltaic power plants (data processing and
commands sending to the plant). A number of acquisition and processing data systems
have already been developed to the measurement, processing and acquisition of
environmental data, evaluation of the performance of the power plant, amongst others
(FORERO, HERNÁNDEZ, & GORDILLO, 2006).
CHAPTER 1 – INTRODUCTION
With the advancement of technology in wired and wireless networks, there is a

tendency nowadays to renew the industrial automation concepts and to associate them
to the Internet of Things concept. The Internet of Things (IoT) is an information sharing
environment, in which the devices are connected to wired or wireless networks, making
easy remote access viable. Considering the growing interest in clean energy and the
increasingly recurrent construction of renewable energy sites and the necessity of
constant monitoring of variables related to the application (environmental and electrical
data), applying the power of the Internet of Things to micro-grids seems extremely
promising (ADHYA, SAHA, DAS, JANA, & SAHA, 2016).
1.1 Objectives
This dissertation work is part of the multidisciplinary project under development

by the department of Electrical Engineering of UFMG. This project is currently funded
by the Petrobras P&D Project “Desenvolvimento de Sistema Smart Battery e Planta
Piloto de Armazenamento de Energia Associado à Geração Distribuída de Energia
Elétrica” (development of a smart battery and a pilot plant of energy storage associated
to distributed generation of electric energy in free translation), ANEEL 21/2016, PD-
00553-0046/2016. The goal of this dissertation is to provide the control and monitoring
of the photovoltaic system, applying the concept of the Internet of Things to the plant
automation for the acquisition and processing of data as well as plant control, i.e., this
concept is used for the supervisory system development.
Currently, the analyzed solar power plant is monitored by a supervisory system

developed by a third party company. The contract for this service must be renewed
every three years; if not, the user doesn’t have access to the real-time nor to the
historical data. This supervisory system is capable of presenting an overview of
environmental and electrical variables of the system. It’s possible to monitor graphically
these field variables along any specified period of time. However, since this system is
made by a third party company, it’s not easily customizable to add or delete variables;
meaning that further expansions - such as the inclusion of a hybrid inverter and the
battery banks - would require further developments by this company. Also, this system
doesn’t provide any information in regards to the current state of the devices nor does it
show a schematic of the circuitry. And finally the current supervisory system doesn’t
2
allow the sending of commands to the field devices, preventing the control of power
flow.
The main idea is to provide a custom-made monitoring and control system of

easy remote access of all environmental and electrical variables available by the
photovoltaic panels, inverters and battery banks. This supervisory system should be
flexible – easy to incorporate expansions and/or changes -, adaptable to research, and
cheaper. That means, the main idea is to provide a free supervisory system that is able to
surpass the limitations cited above presented in the supervisory system currently in use
in the power plant.
The specific goals are:
 Develop a low-cost open-source flexible supervisory system for the

monitoring of the measured values of the analyzed photovoltaic power
plant, using IoT devices and adapting non-IoT devices to the new
generation of current automation (IoT);
 Develop a system that allows communication between IoT devices in
different networks in order to incorporate data from different networks;
 Develop a communication structure for a battery management system
and for the power flow control;
 Develop a graph-plotting feature that allows sampling rate configuration
in order to provide graphics able to track all the variations in the
variables throughout the day.
1.2 Methodology
Due to the limitations of the current supervisory system - its structure’s rigidity,
non-adaptability, and difficulty in further expansions - and due to its incapacity of
fulfilling the specific goals, another set of tools has to be used in order to provide the
power plant with a flexible supervisory system that is easily expandable. Also, another
motivation for developing a new supervisory system is being the design authority of the
system, removing the necessity of further development by a third-party company. Since
it’s interesting to mesh the concept of the Internet of Things and renewable energy
generation, the idea is to find tools that are already aligned with IoT ideas. Also, it’s
3
important to find software that is easily programmable in order to create a supervisory

system that fulfills any requirement (monitoring and controlling) in order to make any
necessary expansions that may occur.
There is an array of open-source software options that would fulfill the

requirements. The one that is the most popular and the most versatile is Node-Red.
Another tool that would be a good complement to Node-Red to provide the power plant
with a complete supervisory system is Emoncms, that is frequently used to monitor
generated and consumed energy and to develop dashboards. Both are free Web-Service
platforms already integrated with the IoT concept. Node-Red is a tool based on visual
programming flow to connect hardware devices, APIs and online services. Emoncms is
an application to process, register and visualize energy and other environmental and
electrical data. In later sections of this master’s thesis, the advantages of employing
Node-Red and Emoncms in this project when comparing to other tools is further
detailed.
With the platforms to program the software already chosen, it’s necessary to
select the hardware that acts as a server for necessary services for the development of
the monitoring and control system of the power plant. The chosen hardware is the
microcomputer Raspberry Pi 3 B+. The Raspberry Pi is a series of compact single-board
computers. To establish the supervisory system, it is necessary to implement some
services in order to gather the data, process it and display it in the screens of the
supervisory system. To this end, it’s necessary to establish communication between the
field equipment and the application (supervisory system), data storage, and a backup
system. Three servers are established using the Raspberry: communication server,
database and application server and the backup server as seen in Figure 1-1.
4
Raspberry Pi 1
Cloud Computing
Router
Switch Switch
Raspberry Pi 2 Raspberry Pi 3
Computer
Figure 1-1 – Schematic of the proposed servers architecture for the supervisory system showcasing the modularized
architecture
In this schematic, it’s possible to see that the services are separated in different
servers; the system is modularized. If one of the servers is not working properly or if a
server becomes unavailable, the remaining servers are not affected by this failure; the
supervisory system would not be completely affected and the maintenance would be
simpler since the defect would be more traceable to the defective server and it would be
easier to isolate it and correct the error.
Some devices in the power plant can be easily incorporated in the network
established in the power plant, such as the inverters and the energy meter that are
already equipped to communicate with a wired/wireless network. But there are some
gauges that are not; for these, it`s necessary to adapt them to communicate with the
power plant network established for the project. For this purpose, another device has to
be connected to the gauges so that their data is also available for the development of the
supervisory system.
The proposed scheme can be seen in Figure 1-2. The figure is not on scale – the
elements are depicted in a way to show the proposed scheme.
5
Figure 1-2 – Proposed structure for the data acquisition and supervisory system including the data flow path
Figure 1-2 shows the proposed schematic for the development of the supervisory
system since the data acquisition from the field devices going through the
communication and application servers, ending in the user terminals, local and remote
via the Internet. Some field devices are already configured to establish communication
with a network whilst others aren’t; the latters have to be connected to an intermediary
device in order to provide their data to the network. There are devices that are not in the
same network as the supervisory system servers. These devices communicate with the
supervisory system via the Internet; this communication is shown in Figure 1-2 with the
cloud symbol.
6
It’s important to note all the devices and servers are wired connected to the
routers and switches. The wireless connection is negatively impacted when the distance
between the router and the device is long. Besides, wireless connections can be
obstructed by items and structures, such as walls, ceilings, and furniture. Finally, the
network security is a greater issue in wireless connections, meaning that the information
is less secure and the data can be more easily compromised.
1.3 Scientific Contributions
Throughout the development of this dissertation, some scientific contributions

were made. These contributions were manifested as scientific papers, cited bellow:
 “A Long Term Evaluation of Photovoltaic Systems under Power Quality

Problems” in collaboration with Prof. Dr. Thales Maia and Prof. Dr. Igor
Pires – this paper analyzed the impact of power quality problems in solar
power devices, motivated by the failure of one photovoltaic inverter. The
paper was published in the XXIII Congresso Brasileiro de Automática;
 “Design of a SCADA System Based on Open-Source Tools” in
collaboration with Prof. Dr. Braz Cardoso Filho, Prof. Dr. Thales Maia,
and Prof. Dr. Igor Pires – this paper presented the development of a
supervisory system employing techniques related to the Internet of
Things. This paper is submitted for approval in the IEEE/IAS
International Conference on Industry Applications 2021.
1.4 Text Organization
The text was organized in the following manner: the current chapter presents the
motivation, the specific goals and the used methodology to reach them.
In Chapter 2, there is a bibliographic review on micro-grids and the systems

that have been used to monitor and control plants traditionally, in addition to detailing
the specific plant for this work. Also, the Internet of Things (IoT) concept is presented
in more detail and the two programs used for data acquisition and control (Node-RED)
and for supervising variables (Emoncms) are explained in depth.
7
In Chapter 3, the schemes for the data acquisition and control developed in
Node-Red and for the supervisory system developed in Emoncms are proposed.
Furthermore, the proposed logic proposed for the integration of the Arduino to the
power plant is outlined.
In Chapter 4, the results are shown. There is a comparison between the current
supervisory system provided by a third party company and the supervisory system
developed along this dissertation.
The main conclusions and proposed continuation are presented in Chapter 5.

The bibliographic references cited throughout the text are listed at the end.
8
2 Bibliographic Review
In this section, the subjects related to the research are introduced in order to
establish important definitions that are used in subsequent chapters to develop the
supervisory system for the power plant.
The techniques used to develop the supervisory system can be integrated to any
plant. However, in order to deepen the amassed knowledge, it’s integrated to a specific
power plant that is built as a micro-grid. So, the first topic of interest to be introduced is
micro-grids. It’s also interesting to present the techniques traditionally used when
developing a supervisory system in order to compare them with the novel techniques
employed here in order to build a new type of supervisory system.
One of the goals of this project when developing the supervisory system is to
follow the current trend in the automation field, which means employing the ideas of the
new wave of industrial revolution: the Internet of Things (IoT). In this chapter, a short
overview of this new concept, the devices related to it and its key challenges are
discussed in order to prepare the groundwork for the development of the supervisory
system.
After this bibliographic review of concepts and traditional methods, the tools
selected for the building of the supervisory system in this project are introduced as is the
power plant of interest.
2.1 Micro-grids
For the context of the project analyzed in this dissertation, the concept of micro-
grids implies a group of loads and micro-sources that operate as a single unity that
provides energy to the area. The components can be assemblies of micro-turbines, fuel
cells, photovoltaic panels and other small generators, storage systems and controllable
loads. To the final consumer, the micro-grid can be seen as a means to satisfy specific
requirements, such as improvement of the local reliability, reduction of losses, amongst
others (MARINHO, 2011).
CHAPTER 2 – BIBLIOGRAPHIC REVIEW
Micro-grids combine modular energy sources, such as wind, photovoltaic, with

energy storage systems and controllable loads, forming a low-voltage distribution
system. A micro-grid is a small-scale, self-supporting network by on-site generation
energy, able to disconnect from the main grid due to sustainability and/or energy
security (VENAYAGAMOORTHY G. K., SHARMA, GAUTAM, & AHMADI, 2016).
An example of a generic micro-grid can be seen in Figure 2-1.
Figure 2-1 – Illustration of a generic micro-grid, highlighting the type of data that can be processed by its control and
supervisory systems (STADLER & NASLÉ, 2019)
Figure 2-1 depicts a generic micro-grid that is able to connect and disconnect to
a main power plant. It also depicts the type of information that can be monitored when
interacting with the supervisory and control system of a micro-grid, such as: data from
renewable power plants, energy storage systems, weather forecast, markets, and other
micro-grids.
Distributed energy is the main generation system in micro-grids architectures.

According to the International Energy Agency (IEA), distributed generation is the
energy production from wind and/or solar, amongst others. The energy generation
technology is close to the consumer unity. Therefore, this type of generation doesn’t
require high voltage transmission lines for the energy transportation to the consumer
(NARUTO, 2017). Micro-grids can operate in two distinct modes: connected mode, the
micro-grid is connected to the main grid, and island mode, the micro-grid is not
10
connected to the main grid (ÁLVAREZ, CAMPOS, GARCÍA, GONZÁLEZ, & DÍEZ,
2010).
Micro-grids are an effective tool to employ in the use of local sources of

renewable energy (non-renewable as well). There are some advantages when using
micro-grids: improved energy efficiency, reduced energy consumption and
environmental impact, higher reliability and resilience and more cost efficient
maintenance (ÁLVAREZ, CAMPOS, GARCÍA, GONZÁLEZ, & DÍEZ, 2010). Micro-
grids also reduce long-distance electricity losses as electricity transmission accounts for
approximately 4% of total power consumption (HU, CHEN, CHEN, & CHANG, 2013).
Due to the fact that micro-grids are small-scale and they are usually provided
with a robust communication system, the control and monitoring are decentralized,
explaining the reason they have higher reliability and resilience (NARUTO, 2017).
The output of renewable generation may vary according to the weather condition
of the day (wind speed, solar radiation to name a few factors) and time of day. This
means that the majority of the renewable resources can’t guarantee a continuous and
steady power generation on their own. Since this may have a destabilizing effect, it’s
important to consider it when there is a desire to integrate renewable energy sources into
traditional grids; one of the tools to mitigate this concern is to incorporate energy
storage systems into the micro-grid. (VENAYAGAMOORTHY G. K., SHARMA,
GAUTAM, & AHMADI, 2016).
Since control and supervision are vital to energy generation and distribution
especially for micro-grids that can be disconnected from the main grid and are related to
energy storage, it’s important to establish an information flow between the micro-grid
and the user in order to enable tracking and monitoring of the processes and electrical
and environmental variables. For this reason, it’s interesting to review the most
common techniques employed in the development of supervisory systems (not only for
micro-grids). The traditional supervisory system is presented next in order to familiarize
oneself with the commonly used tools so that the innovations in the supervisory system
developed in this dissertation become clearer.
11
2.2 Traditional Supervisory Systems
SCADA (supervisory control and data acquisition) is a system of hardware and

software components that enable process control. Automatic data acquisition system has
been used for monitoring system performance and control. SCADA is traditionally a
central control that consists of controller’s network interfaces, input/output,
communication equipment, and software. It’s used to monitor and control the equipment
in processes, gathering data from sensors and instruments to showcase an overview of
the whole process.
Generally, the SCADA includes local processors, operating (field) equipment,

PLCs (programmable logical controller), instruments, sensors, remote terminal unit
(RTU), intelligent electronic devices (IEDs), and a computer with a human interface
(RAGHVENDRA, 2019). Figure 2-2 shows an overview of the structure.
Figure 2-2 - Representation of the hardware architecture of a SCADA system. Modified from (Inductive Automation,
2020)
The supervisory system acts as a link between the user in the control room and
the equipment like PLCs, RTUs, IEDS, and sensors. The RTUs are connected to field
devices; they are controlled by microprocessors and are used for transmitting recorded
data to the supervisory system. They also receive commands in order to control the field
devices. The PLCs are connected to sensors in order to convert the sensor output signal
into digital data (RAGHVENDRA, 2019).
12
Traditionally, a common characteristic of such structure is the use of data

loggers or microcontrollers for measuring and acquiring data and transmitting them to a
user terminal (personal computer, microcomputer) via a serial port (RS232) (FORERO,
HERNÁNDEZ, & GORDILLO, 2006), (MUKARO & CARELSE, 1999). A system is
developed in order to collect, record, and transmit the measured data to a computer. The
system developed in (MUKARO & CARELSE, 1999) is seen in Figure 2-3 in a
simplified representation.
Figure 2-3 - Block diagram of a traditional data acquisition system using serial communication interface
As it can be seen, this acquisition scheme relies on an I/O (input/output)

microcontroller in order to retrieve the measured data and it transmits this data to a
portable computer via wired connection through a serial port. The data is then
downloaded to the computer via the communication protocol and then it can be
accessed for further analysis. The supervisory and control system are developed on this
specific portable computer using assembly language and it’s only accessible in it
(MUKARO & CARELSE, 1999).
In (BENGHANEM & MAAFI, 1998), this methodology is also employed even

though another microcontroller is used. This system uses bidirectional communication;
the system receives data from the sensor and sends commands through this serial
protocol. This data is treated and converted to physical values and used by the central
computer to study the performance of the photovoltaic system (BENGHANEM &
13
MAAFI, 1998); this study can only be done in this computer where the supervisory
system was programmed and the data is only accessible inside the local network.
Due to some limitations when using serial port, such as limited data transmission
when data acquisition at high sampling rates is required, some research has been done in
order to minimize the impacts of this limitation. In (FORERO, HERNÁNDEZ, &
GORDILLO, 2006), to avoid this, it`s proposed to use I/O modular devices (with field
point analog modules) and data acquisition cards with high-speed analog to digital
conversion. This project also relies on a local supervisory system with wired-serial
connections between the user terminal and the field devices, enabling only local users
the monitoring of the power plant. In (KOUTROULIS & KALAITZAKIS, 2003), the
authors use the same methodology utilizing high-speed cards and I/O modules.
More recently, with the technological advancements in wired/wireless networks,

the serial communication has been replaced by the establishment of a local network. In
(DUMITRU & GLIGOR, 2012), the authors present the more popular approach
nowadays for the development of a SCADA (supervisory control and data acquisition)
system for any plant. The scheme discussed in this study, implements a modernization
in comparison to the previous ones discussed in this dissertation up to this point by
establishing a local network. Most operations are automatically executed by the local
equipment of data acquisition and control (RTU) or by the programmable logic control
unit (PLC). A generic SCADA system can be seen in Figure 2-4.
14
Switch
Figure 2-4 – Illustration of a SCADA, showcasing the data-acquisition devices, servers, workstations and user
terminals
Data acquisition begins at the RTU or PLC and involves reading the
measurements and the state of gauges that are transmitted to SCADA application. The
monitoring and control systems are made by dedicated devices. The values of the
measured parameters are presented graphically in real time or historically in the users
terminals connected to the local network; in Figure 2-4, by serial communication, router
of the network or by the communication server. The control means sending commands
towards the system from the user interface (DUMITRU & GLIGOR, 2012). This
structure shows progress in regards to the other discussed here since it considers a LAN
(local area network), which can be wired or wireless, for the building of a
communication path between the field devices (RTU, Local Terminal, PLC) and the
SCADA serves, the communication server and the user interfaces.
In (PAPAGEORGAS, PIROMALIS, ANTONAKOGLOU, VOKAS, &

TSELES, 2013), the authors propose a supervisory system that uses a similar structure
15
presented in Figure 2-4 for a photovoltaic solar plant. The field devices communicate
with a local gateway through a serial bus. The gateway provides wireless
communication services for the devices using wireless sensor network (WSN)
technology with a central terminal, where the data is processed. It’s interesting to notice
in this research work that the devices are not IoT devices – they are merely wireless;
they can’t access the Internet. However, after the data is processed in the central
terminal, it’s published on the Internet to remote users using publishing services such as
Cosm, that is used in (PAPAGEORGAS, PIROMALIS, ANTONAKOGLOU, VOKAS,
& TSELES, 2013). This work shows a step towards the concept of the Internet of
Things.
2.3 Internet of Things
Due to the advancements in wired/wireless networks and in sensing, the Internet

of Things has been identified as one of the emerging technologies in IT (GUBBIA,
BUYYAB, MARUSIC, & PALANISWAMI, 2013). It is a global Internet-based
information architecture, facilitating the exchange of goods and services in global
supply chain networks (WEBER, 2010). The Internet of Things is a new technology
paradigm that enables the setup of a global network of machines and devices that are
capable of interacting with each other. The main differential of this new technology is
the possibility to enable devices to communicate with each other and interact with
inventory systems, customer support systems, and business applications. Also, the
Internet of Thing allows the remote control of devices since the communication is
performed via the Internet. Due to IoT, a whole slew of applications emerge, especially
when considering industry and customer oriented applications, such as smart houses,
smart neighborhoods, and smart cars (LEE & LEE, 2015). However, embedded in this
new technology, there are a lot of new challenges: data management (due to the high
volume of data), privacy and security, and standardization (WORTMANN &
FLÜCTHER, 2015).
The devices and gauges capable of interacting with an Internet of Things

environment are smart devices. Smart devices are equipment capable of communication
and computation; they range from simple sensing gauges to home appliances and
smartphones. The key features of smart devices are autonomy, connectivity, context-
16
awareness, user-interaction, mobility, and data storage (SILVERIO-FERNÁNDEZ,

RENUKAPPA, & SURESH, 2018).
The Internet of Things connects a vast amount of objects, creating an extremely

large data flow and a necessity for large data storages. Besides these two factors, the
Internet of Things has to deal with new challenges, such as data security and privacy.
The architecture for IoT needs to incorporate features that address the new challenges
like scalability, interoperability, and reliability, amongst others. The basic architecture
for the Internet of Things is seen in Figure 2-5 (KHAN, KHAN, ZAHEER, & KHAN,
2012).
Figure 2-5 – Architecture of the IoT structure, showcasing each fundamental block. Modified from (KHAN, KHAN,
ZAHEER, & KHAN, 2012)
The initial block depicting the physical objects and sensors is related to the
identification of each individual component when inserted in the context of the network;
each device has its own IP address for example. As it’s represented by the arrows, the
physical objects can send/receive data to/from the network. The communication
infrastructure is related to the communication protocols used to establish the link
between devices, devices and services, devices and users. The decision making inside
the computation and processing unit is related to the ability of extracting information
from the devices data in order to activate certain services and to execute local
algorithms (KHAN, KHAN, ZAHEER, & KHAN, 2012). The basic structure for IoT is
composed of 5 layers as seen in Figure 2-6.
17
Figure 2-6 – IoT 5 structural layers – layers applied in the IoT environment from sensing and data acquisition until
system management. Modified from (KHAN, KHAN, ZAHEER, & KHAN, 2012)
The bottom layer is the perception (or device) layer; it consists of the physical
equipment and the sensing devices. This layer is responsible for the identification of the
devices and data acquisition. This information is then passed to the network (or
transmission) layer for its secure transmission to the data processing system. The
transmission can be wired or wireless. The data is then transmitted to the middleware
layer. This layer has links to the database; it’s capable of storing the data. The
middleware layer processes the data and activates services based on the results of the
data processing. The application layer is responsible for global management of the
application based on the data processing on the previous layer. And lastly, the business
layer is responsible for the management of the IoT system, including applications and
services (KHAN, KHAN, ZAHEER, & KHAN, 2012).
2.3.1 Smart Devices
Smart devices are able to connect to the Internet via wired/wireless connections,
being capable of establishing data exchange with other smart devices and allowing
control and access to remote users and services. A smart device is a context-aware
18
equipment that can be used as a service provider and can communicate with devices in
different networks (SILVERIO-FERNÁNDEZ, RENUKAPPA, & SURESH, 2018).
Smart devices implement three steps: data acquisition, data process, and
communication. In the first step, data acquisition, the device acquires the data measured
by its sensing unit. Then, the raw data is processed by the processing unit. And, finally,
the processed data is transmitted to the network via a communication protocol
(JIMÉNEZ-GARCÍA, BASELGA-MASIA, POZA-LUJÁN, MUNERA, POSADAS-
YAGÜE, & SIMÓ-TEN, 2014).
The key features considered by (SILVERIO-FERNÁNDEZ, RENUKAPPA, &

SURESH, 2018) for a device to be considered smart are: autonomy, context-awareness
and connectivity. Autonomy would relate to the feature of smart devices of being
capable of performing tasks without external interference – no direct user command
necessary to perform the task. Context-awareness means the device is able to perceive
environmental information via sensors or sensing units. The connectivity is the feature
of establishing a connection to a wired/wireless network of any size in order to achieve
a goal: access the Internet, share information with other devices, amongst others.
With technological advancements, the devices processing power and storage

abilities increased whilst their sizes were reduced. Due to its features, the smart device
technology can lead to various possibilities: smart homes, smart cities, industry
applications, smart metering, and smart security to name a few (KHAN, KHAN,
ZAHEER, & KHAN, 2012). However, there are still some challenges that must be
investigated regarding smart devices specially when considering the popularity of this
technology in different fields nowadays.
2.3.2 Key Challenges in IoT
There are some challenges when it comes to IoT; some of them are (KHAN,
KHAN, ZAHEER, & KHAN, 2012):
 Naming and identity management: Each device connected to the network

must have a unique identity over the Internet – an efficient naming and
identity system that is capable of managing a large number of devices is
required;
19
 Interoperability and standardization: Many manufactures have their own

protocols – it’s important to standardize them to guarantee
interoperability;
 Devices safety and security: It’s necessary to prevent security breach in
order to protect the integrity of the devices;
 Network security: The data transmission should be secure and capable
enough to protect the integrity of the data and to prevent external
interference or monitoring.
Even with these challenges, it’s important to note that this new technology can
provide significant benefits (KHAN, KHAN, ZAHEER, & KHAN, 2012). As discussed
previously, the smart devices are capable of connecting to the Internet. Therefore, it’s
interesting to develop a supervisory system for smart plants with Web-Services that are
already built with the IoT concept embedded.
2.4 Open-Source IoT Tools for Supervisory System
There are a variety of Web-Service tools available for the development of

supervisory systems of processes equipped with smart devices. As discussed previously,
the tools chosen for this project are the free Web-Services Node-Red and Emoncms due
to their flexibility, popularity - since they have a mature, dependable, and stable
structure at their core -, reliability, and to their friendly and robust user interface based
on flow programming. Also, since they are both Web-Services, they can be used in any
web browser (Google Chrome, Safari, Firefox, Chromium, Opera, etc.), providing the
desired versatility and portability to the supervisory system developed in this work.
Since their structure supports a lightweight runtime environment, they can easily fit on
devices like the Raspberry PI, which is the chosen microcomputer for this work.
2.4.1 Node-Red
Node-Red is an open source tool developed initially by IBM for flow

programming using a local host, meaning it’s a browser based flow editor
(SUMANGALI & KIRAN, 2016).
It’s interesting to use flow-based programming (FBP) when connecting to

hardware due to a number of reasons. The main reason is that FBP allows parallelism,
20
providing performance benefits in some situations. In FBP, the dataflow is the main
driving force of the program. The logical execution flow is expressed by the block
diagram created: when a node receives all necessary inputs, the block produces an
output that is transmitted to the next node in the path. The transmission of data through
the nodes establishes the execution order of the functions. Each dataflow path runs
parallel to each other, meaning that if there is a bottleneck in the code, this “hotspot”
will not interfere in other processes. Also, FBP treats the applications as black-boxes:
the important component for this type of programming is the connections between
components that are done externally to the processes. Due to this style of programming,
FBP lends itself to a plug-and-play approach, connecting all the hardware present in the
plant. This type of programming allows the programmer to coordinate components into
entire systems.
Node-Red consists of nodes that can be connected by wires (SICARI,

RIZZARD, & COEN-PORISIN, 2018). This application is able to connect to hardware
devices, such as microcontrollers and the raspberry pi amongst many others, and to the
cloud environment. Node-Red runs on Java Script and it uses the Node.js framework
(SUMANGALI & KIRAN, 2016). Figure 2-7 shows an example of a code in the
environment.
Figure 2-7 – Example of the structure of Node-Red and its interface showcasing an example of a Node-Red code –
showcasing its flow-based programming structure
The visual representation provides the users a better understanding of the

interactions happening in the IoT network architecture as well as the messages being
transmitted. Different kinds of hardware, e.g. sensors, and software, e.g. services, can be
21
modeled (SICARI, RIZZARD, & COEN-PORISIN, 2018). Also, the user can program
customized functions and can develop new nodes in order to fulfill the requirements of
the designed application. Node-Red natively provides nodes for developing a well-
structured dashboard with charts, gauges, amongst others, such as the one seen in Figure
2-8.
Figure 2-8 - Example of Node-Red Dashboard developed in the Node-Red platform using available nodes for
creation of dashboards
Node-Red makes available simultaneously many flows acting as clouds-nodes or

Raspberry PIs, Arduino platforms and so on. It enables the real connection of hardware
devices and APIs (SICARI, RIZZARD, & COEN-PORISIN, 2018).
Due to the lightweight nature of Node.js and the simplicity of the Node-Red
execution engine, Node-Red can be deployed with good performance on smaller
computers and microcontrollers, such as the Raspberry Pi or Arduino (SICARI,
RIZZARD, & COEN-PORISIN, 2018).
The advantages of using Node-Red when comparing it to other open-source

tools, such as Eclipse Kura and Flogo, come especially from the usage of Node.js.
Node-js is a lighter and agile open source tool. Node-Red with its mature framework
also presents a friendlier user interface for programming. Due to the employment of
JSON for its metadata, Node-Red has a more protected metadata structure by not
relying on XML. Besides, due to its flow-based programming and the ability of building
22
custom functions, Node-Red is able to fulfill the requirement of flexibility. Due to the
nature of the project, a supervisory system for monitoring electrical and environmental
variables, the processing rate of messages inside the code doesn’t need to be extremely
fast. For this reason, the speed of 1 message/second of Node-Red is sufficient and more
suitable than a tool as Crosser (10 messages/second), whose performance would surpass
the requirement.
2.4.2 Emoncms
Emoncms is a powerful open-source web-app developed initially by Open

Energy Monitor for processing, logging and visualizing energy-related data. The web
application provides an API key in order to allow reading and writing requests to its
data. It enables the user to do the setup of the inputs, feeds, graphs, and of the dashboard
(BENAMMAR, ABDAOUI, AHMAD, TOUATI, & KADRI, 2018).
Emoncms uses the Hyper Text Markup Language (HTML), Cascading Style
Sheets (CSS) and Javascript to build its user interface. The user communicates with the
Application Programming Interface (API) of the server through Asynchronous
Javascript and XML (AJAX), exchanging JSON messages. The API HTTP of the server
activates the internal models to perform functions as data registration, processing and
validation (MARIANO, 2016).
This application is divided in modules, enabling individual function

development – the main modules are Feeds and Inputs, which can be seen in the
Emoncms environment in Figure 2-9.
Figure 2-9 - Example of Emoncms environment showcasing its user interface and its tabs: Inputs, Feeds, Graphs, and
Visualization
23
The raw data that is received by the application forms the Inputs module. They
need to be treated before being stored in the database in order to optimize its processing
for the eventual visualization. The tool provides various functions to treat the input and
convert it (if necessary) to the desired value to be stored in the database. The input can
be attached to different functions that produce an output that is linked to a Feed; the
Feeds module then is composed of processed raw data that is suitable for storage in the
database (MARIANO, 2016). This configuration is made manually by the user in the
Emoncms user interface.
The tool also allows the user to do the setup of the database when publishing
new inputs and configuring them as feeds. This database is not based on MySQL; the
data is stored in folders in the local server based on a NoSQL approach to store an
arbitrary number of time points without an index since it’s more lightweight and it’s
more efficient when it comes to the process involved in the application. Figure 2-10
shows an example of a dashboard developed in the Emoncms environment.
Figure 2-10 - Example of Emoncms dashboard developed in the Emoncms platform using the gadgets for creation of
dashboards
2.5 Micro-grid Tesla
The micro-grid analyzed in this project is composed by a battery energy storage

system (battery banks and three hybrid inverters capable of disconnecting the micro-
grid from the main grid), a photovoltaic power plant (Usina Solar Fotovoltaica Tesla –
USF Tesla), loads (programmable loads and feeders). The schematic for the whole
system can be seen in Figure 2-11. This project focuses on the power plant USF Tesla
and in the energy storage management system.
24
Figure 2-11 - Micro-grid Tesla single-line diagram
2.5.1 Photovoltaic Solar Power Plant Tesla
The power plant has been operational since July, 2016; its capacity is 37 kWp.
The construction project was funded by the P&D 0047-0060/2011, 13/2011 ANEEL,
project. The photovoltaic panels can be seen in Figure 2-12; there are 152 silicon
polycrystalline cells, each 245Wp. And the control room can be seen in Figure 2-13.
Figure 2-12 - Photovoltaic panels installed in the photovoltaic solar power plant Tesla
25
Figure 2-13 - Control room of the solar photovoltaic power plant, highlighting the inverters, DC and AC panels, and
data panel
As it can be seen, there are two power inverters of two distinct manufacturers:
SMA and Fronius. The main characteristics of each one are described in Table 2.1. Both
of them have a WLAN and an Ethernet interface which enables the connection to a
wireless/wired network; both of them are IoT ready, capable of delivering data to the
network, data about its own variables or the grid’s. The inverters SMA and Fronius can
be seen respectively in Figure 2-14a and Figure 2-14b.
Table 2.1 - Power inverters main electrical characteristics.
SMA Sunny Tripower

Inverter Fronius IG Plus 150V-3
120000TL
Peak Power (kWp) 12 12
MPP Voltage 440-800 230-500
Range (V)
Max. Intput 1000 600
Voltage (Vdc)
Max. Input Current 55.6 36
(A)
Number of MPP 2 1
Output Voltage 400 400
(Vac)
26
Number of 15 10
Modules/String
Number of Strings 3 5
in Parallel
Installed Peak 11.025 12.25
Power (kWp)
a) b)
Figure 2-14 – a) SMA Sunny Tripower 12000TL b) Fronius IG Plus 150V-3
There are sensors throughout the power plant: energy meter Janitza UMG604
seen in Figure 2-15, current sensors, pyranometer, thermometer, barometer, humidity
sensor and anemometer. The sensors make up the solar meter station and the
measurement details can be seen in Table 2.2.
27
Figure 2-15 - Energy Meter Janitza UMG604
Table 2.2 – Characteristics of the measurements of the solar meter station
Measurement Measuring Range Accuracy Sampling Rate

Air temperature -50°C ... +60°C +/-0.2°C 1 minute
Humidity 0 ... 100% RH +/-2% RH 1 minute
Air Pressure 300 ... 1200hPa +/-1.5hPa 1 minute
Wind Speed 0 … 60m/s +/-0.3m/s 10 seconds
Wind Direction 0 … 359.9° < 3° 10 seconds
Global Radiation 0.0 … 1400.0 W/m² < 1 W/m² 1 minute
2.5.2 Energy Storage System
This system is not connected to the solar power plant yet; it’s still in its testing
stages. This system will enable power dispatch in the connection point between the
micro-grid and the main grid. The batteries are also vital for the operation in island
mode, when disconnecting the micro-grid from the main grid. The system is composed
by three independent single-phase hybrid inverters and three battery banks – each one a
different technology.
The first bank is made by 20 lead-carbon batteries seen in Figure 2-16a

distributed in 5 blocks in parallel, each block containing 4 batteries in series. The
second one is the molten-salt battery bank seen in Figure 2-16b, with 2 batteries in
parallel. And, lastly, the lithium-ion bank battery seen in Figure 2-16c has 6 batteries in
parallel. The main characteristics of each technology are presented in Table 2.3.
Table 2.3 - Batteries in the Energy Storage System.
Battery Lead-Carbon: Narada Molten-Salt: FZSoNick Lithium-Ion: Unipower
28
REXC70 48TL200 UPLFP48-100

Nominal Voltage 12 48 48
(V)
Capacity (Ah) 70 200 100
a) b)
c)
Figure 2-16 – a) Battery Narada REXC70, b) Battery FZSoNick 48TL200, c) Battery Unipower UPLFP48-100
Each battery bank is connected to the correspondent hybrid inverter, enabling

the monitoring and control of each electrical variable through the battery management
system.
The hybrid inverters are 3 single-phase independent SMA Sunny Island 8.0h
inverters seen in Figure 2-17. They have a WLAN and an Ethernet interface which
enable them to connect to a wireless/wired network. The hybrid inverters are connected
to the battery banks and are capable of enabling dataflow from/to the battery
management system to/from the control room via a local network or via the Internet.
They make any variable available for the user in this network: their own, the ones of the
main grid and the variables related to the connected battery bank. The hybrid inverters
enable the possibility of data monitoring and power flow control. The variables
measured by the DC-side related to the batteries and to the AC-side related to the grid
29
are important to the development of a supervisory system. These inverters also allow
the sending of commands to control the charge and discharge of the battery banks in
order to adjust the power flow.
Figure 2-17 - SMA Sunny Island 8.0h
30
3 Supervisory System
In this chapter, the methodology for developing the supervisory system is
detailed. The supervisory system is developed through the usage of Raspberry Pi 3b+
microcomputers seen in Figure 3-1. There are 3 servers that constitute the supervisory
system: communication server, application and database server and the backup server.
The communication server consists of data acquisition and processing. The data
acquisition and its processing, e.g. converting the raw data to engineering unit, are
performed in the Node-Red environment. Then, the processed data can either be sent to
the Emoncms environment in the application server in order to build the supervisory
dashboard (a schematic of the power plant that shows the real-time measurements), or
be used in the Node-Red dashboard environment. The Node-Red dashboard can be built
to monitor variables in graphs or it can be used to send commands to the field devices,
such as sending commands to control the active and reactive power flow. The data is
stored in the application and database server through the configuration of the input data
received by Emoncms.
Figure 3-1 – Hardware: Microcomputer Raspberry Pi 3 b+
There are two distinct networks that can communicate with each other: the solar
power plant with the power inverters and the environment measurements, and the
energy storage system with the hybrid inverters and the battery banks.
CHAPTER 3 – SUPERVISORY SYSTEM
3.1 Data Acquisition and Processing
3.1.1 Solar Power Plant
The solar power plant has two power inverters and an energy meter that are
smart devices: the SMA Sunny Tripower 12000TL, the Fronius IG Plus 150V-3 and the
Janitza UMG604 energy meter; they can connect to the TCP/IP. The inverters are
capable of measuring electrical data from the DC-bus (the solar panel side) and from the
AC-bus (the main grid side) as the energy meter. This data is available via
communications protocol for data processing.
The communication protocol used in this dissertation for the communication

between the communication server and the smart devices (the inverters and the meter) is
the Modbus TCP/IP since they are already configured for this protocol; the meter and
each inverter can be configured for a specific fixed IP determined by the user and has its
own Modbus map.
The Modbus protocol is a highly used flexible open message structure used for
the communication between master (client) and slave (server). The communication is
only initiated by the master; the master is responsible for initiating the communication
between devices by sending a request to the slave for it to send its data as seen in Figure
3-2. The master may request/receive data to/from the slave at any moment since the
communication is not initiated by the slave. The slave receives the request and responds
by sending the required data; the sent data may be binary or numeric. Figure 3-3 shows
how the message is built in the master and in the slave and how the Modbus message
exchange happens; it’s possible to see that this protocol requires a polling structure
(FREITAS, 2014).
Figure 3-2 – Illustration of the communication between master and slave in the Modbus protocol (NI, 2019a)
32
Figure 3-3 – Schematic of the Modbus Message Exchange. Modified from (FREITAS, 2014)
The device address corresponds to the IP address of the device that can be
configured by the user. The function code refers to the type of request the master is
performing: read/write analog/digital registers. The TCP port 502 is the standard port
for the connection with Modbus TCP servers (NI, 2019a). The Ethernet standard
(TCP/IP) in the Modbus protocol may have some variations when it comes to the
available speed: it can reach 100Mbps or even 10Gbps (NI, 2019b).
Modbus can also use the serial RS-485 standard to establish communication
between devices. This standard allows communication rates that may reach 12 Mbps or
50 Mbps (NI, 2019b). Ethernet is faster and easier to troubleshoot; however, sensors,
transmitters don’t need to report data very fast and Modbus RS-485 may be less
expensive since it doesn’t need a switch when having various Modbus servers in the
network (RINALDI, 2013). A generic Modbus RS-485 connection can be seen in
Figure 3-4. The physical connection is done with a twisted wire pair plus ground
(GND).
Figure 3-4 - An example of a wired-connection between devices using the RS-485 Modbus protocol. Modified from
(BB Smart Worx, 2020)
33
The Modbus protocol doesn’t need an intermediary (such as a broker) since the
connection is done point-to-point. The way this communication is performed guarantees
that the final recipient receives the message: the master knows the request has reached
the slave and the slave knows the master has received the response. However, since the
protocol requires a request-response structure, the network may be sub utilized or the
application code may become extremely complex and the consumption of network
resources is high. One important disadvantage of Modbus is that it doesn’t incorporate
security resources (SILVA, 2019).
Node-Red is equipped with nodes that can be configured to establish a Modbus

connection between the communication server (Raspberry installed with Node-Red) and
the inverters and the energy meter. That way, the raw data from these devices can be
acquired by the request made by Node-Red (acting as a master) to them (acting as
slaves) for certain electrical variables (e.g. AC/DC voltage, current, power amongst
other).
The raw data obtained for each device has to be processed in order to obtain the
data in engineering units, meaning that each data has to be converted to a float value
that represents the real value of the electrical variable. For this reason, some conversion
functions are developed in Java Script using the Node-Red environment as seen in
Figure 3-5. In Figure 3-5, it’s also possible to see that the processed data is then sent to
the Node-Red Dashboard as a graph and to the Emoncms environment to be a part of
the supervisory dashboard.
Modbus link between

Fronius inverter and Data conversion to
Communication Server engineering units
Figure 3-5 – An example of the acquisition and processing of data in the Node-Red environment
Besides the meter and the power inverters, there are gauges that are responsible
for measuring the relative and absolute (calculated from relative humidity and air
34
temperature) humidity, the relative (referenced to sea level, calculated using the
barometric formula with the aid of local altitude) and absolute pressure, the ambient
temperature, the wind direction and the global irradiance – the solar meter station. There
are also gauges to measure the temperature and the radiance of a reference cell.
These gauges are not smart devices; however, they are currently being
monitored by a data-logging device provided by the third party company that is
currently supplying the supervisory system. This device communicates with the gauges
via Modbus RS485. Therefore, it’s possible to retrieve this data from the third party
company device and upload it to the application server through the file transfer protocol
(FTP).
FTP is a secure connection between devices that allows file exchange as seen in
Figure 3-6. The file transfer protocol serves as a bridge between devices connected to
the same network or connected to the Internet. FTP also relies in a client-server
structure; the client requests access to a certain file and the server grants this access.
After the file transfer is completed, the communication is automatically terminated. This
protocol is a simple tool that allows high volume of data transfer through a network and
it allows various directories to be transferred at the same instant (L, 2020).
Figure 3-6 – Illustration of the process related to the File Transfer Protocol. Modified from (L, 2020)
The idea is to upload the files with the gauges data to the communication server
and post it to the application server to incorporate it to the supervisory system. As
explained, the FTP client needs a program to act as the FTP link and make requests to
the FTP server and receive access to the desired files. In order to supply this link for the
35
current application to retrieve the sensors data from the data-logging device, a Python
program is developed. This program is uploaded to the communication server as a
background service that starts running with the booting of the Raspberry. The upload to
the communication server is performed via FTP through the Python service.
In order to establish a link between the communication server and the

application server to post the data from the solar meter station to the application, it’s
necessary to integrate another communication protocol in the Python service. The
protocol selected is the MQ Telemetry Transport (MQTT) protocol. MQTT is a
communication protocol based on a publish-subscribe architecture. In this type of
architecture, there are devices that generate and publish data (publishers) and other
devices that consume this data (subscribers). This architecture demands an extra entity
that acts as a centralizer in the data exchange, the Message Broker. The devices that
generate data publish it to the broker which organizes the data in topics. The devices
that consume the data subscribe to the desired topic and are able to retrieve the data
assigned to the topic. Every instant new information reaches the topic in the broker, this
information is automatically sent to the device that subscribed to the topic. It’s
important to note that the publishers and subscribers never interact directly; the link is
done via the broker (SILVA, 2019). The process is summarized in the scheme seen in
Figure 3-7.
Figure 3-7 – Illustration of the communication performed in the MQTT protocol (SILVA, 2019)
There is no need for the subscribers to know the address of the publishers; they
only have to connect to the broker and subscribe to the desired topics to receive the
data. The publisher doesn’t have to wait for a request from any device to publish its
36
data; the publisher determines the best moment to send its data to the broker, enabling a
more efficient energy management. This protocol has implemented security layers; it
requires login and password for broker access and for message exchange. And, even
though it can implement complex security measures, it’s a lightweight protocol
(SILVA, 2019).
In addition to the aforementioned advantages, the installation of Emoncms also

installs a lightweight MQTT broker (the open source message broker Mosquitto).
Therefore, the MQTT protocol is a well-suited option for this project.
By using the MQTT protocol in the Python service, it’s possible to publish the
data retrieved from the files obtained by FTP to topics to which Emoncms (and,
therefore, the application server) is already subscribed to. The environmental data is
sent directly as an input to the Emoncms environment in the application server; it
doesn’t need to be processed since it’s already in engineering units, so it isn’t sent to
Node-Red.
Figure 3-8 - Schematic of the communication established between the solar meter station and the application server,
showcasing the FTP and MQTT links
In order to send the measurements of the inverters from the communication

server to the application server, the protocol MQTT is also used. However, to send the
data from the energy meter to the application server, another communication protocol is
used: HTTP. The HTTP (Hypertext Transfer Protocol) is a protocol that enables
hypermedia documents transfer (MDN Web Docs, 2020). It was developed in order to
enable communication between Web browsers and Web servers. It is based on a client-
server structure seen in Figure 3-9; the client initiates the link by making a request and
waits until the server provides a response. After the request is responded by the server,
the connection between the client and the server is severed. A new connection must be
done for every new request (DE SOUZA, 2019).
37
Figure 3-9 – Illustration of the process related to the HTTP Client-Server Communication. Modified from (DE
SOUZA, 2019)
Node-Red has an Emoncms node that uses this protocol to establish a path
between Node-Red and Emoncms for the exchange of data. This node provides
resources to post data to Emoncms via PUSH transaction. This means that it’s possible
to initiate the request to publish data to the client with no initial request from the
receiver – in this case, Emoncms (MDN Web Docs, 2020). This is possible by using the
API key available for reading/writing found in the Emoncms administrator page.
There are also current sensors that measure the DC input current seen in Figure
3-10. These current sensors aren’t smart devices as well. In order to make these
measurements available for the network, the output of each current sensor (there are
three) is connected as an analog input to an Arduino Nano with an Ethernet Shield
(ENC28J60) seen in Figure 3-11.
Figure 3-10 - DC-Current Sensor used in the power plant for the measurement of the DC input current in the power
inverters (3-K, 2020)
38
Figure 3-11 – Microcontroller with the Arduino Nano framework with Ethernet Shield
Arduino is an embedded open source platform in which various controllers may

be used with a microcontroller Atmel AVR that is capable of interacting with the
environment via hardware and software (AQEEL, 2018). The device is built with input
and outputs and the programming language is based on C/C++. Arduino boards are used
in robotics, embedded systems, and electronic projects in which automation is necessary
to the process. The Arduino Nano is a small, compatible, flexible microcontroller and
ideal to smaller projects due to its reduced size (SoldaFria, 2020). Besides this, the
microcontroller selected has more than one analog input – there are three current
sensors – unlike another viable option that would be the ESP8266. Other advantages of
the selected microcontroller are its price, protection, and resilience that are considerably
more suitable for the application than another option such as the ESP32 that has a
higher cost, a slew of protection problems, and it’s more fragile. However, the selected
microcontroller has a limited RAM memory (32kb) and it doesn’t have embedded
resources as WiFi or Bluetooth; therefore, it’s necessary to connect a shield to it in
order to incorporate it in IoT projects (AQEEL, 2018). So, the Ethernet Shield is used to
provide a network connection to the Arduino Nano. A fixed IP for the Arduino Nano is
determined via an Ethernet library – possible due to the Ethernet Shield.
Each analog input of the Arduino that is connected to the DC-current sensors has
its value read in the Arduino. These values need to be converted to engineering units
since they are presented in bits as digital outputs. After they are processed, they’re
transmitted from the communication server to the application server via MQTT. Then,
they are stored in the database and are available for the development of the monitoring
39
screens of the supervisory system. The schematic shown in Figure 3-12 summarizes this
process.
Figure 3-12 - Data acquisition for the DC-Current sensors using a microcontroller – developing a smart sensor
3.1.2 Energy Storage System
There are three hybrid smart inverters, SMA Sunny Island 8.0h, in this network
that are connected to the battery banks detailed in Chapter 2 of this dissertation. These
inverters also communicate with a communication server via Modbus TCP/IP.
These inverters not only provide data but also receive data – commands – via
Modbus TCP/IP in order to enable the user to control the reactive and active power flow
of the grid by managing the charge and discharge of the battery banks. The inverters are
capable of providing data regarding the measurements of the grid (AC-side) and data
regarding the measurement of whichever battery bank is connected to it (DC-side); it’s
important to note that each inverter is connected to a phase conductor and they operate
separately and don’t communicate amongst themselves. As with the power inverters in
the solar power plant, the data sent by these inverters need to be further processed in
order to convert the measurements to engineering units. After this processing, the data is
sent to the application server, to be transmitted and/or to the Emoncms dashboard or the
Node-Red dashboard as seen in Figure 3-5.
3.1.2.1 Secure Protocol

In order to send the data to the Emoncms in the application server, the MQTT
protocol is also used. However, since the networks are different, it’s necessary to
establish this link via the Internet. Therefore, it’s preferable to use the secure protocol:
MQTTS. The MQTTS is the implementation of the MQTT protocol over an additional
security layer that uses the SSL/TLS protocol, using encryption to protect the transfer of
40
data and information. The SSL/TLS protocol secures communications between a server
and a client regardless of the chosen communication protocol, meaning it can be applied
to HTTP or to MQTT or to FTP, amongst others.
The secure communication starts with a TLS handshake: the server and the
client exchange messages to acknowledge and verify each other, they establish the used
encryption algorithms and agree on session keys that are used to encrypt and decrypt all
communications in this session after the TLS handshake (for each new session, a new
set of random keys is generated). Two keys are used during the session: a public key
(client side) and a private key (server side). An example of the exchange between client
and server can be seen in Figure 3-13.
Figure 3-13 - Example of a TLS link with 2 encryption keys in the SSL/TLS secure protocol communication
(CLOUDFARE, 2020)
The identity verification happens via the authentication of the SSL certificates
that have to be stored in the server and can be stored in the client (if the communication
requires client certificates), ensuring each party of the link is who they claim to be. The
SSL certificate is a file containing the public key and other information; without the
certificate, the communication cannot be encrypted by TLS. TLS also ensures that the
data from the server has not been altered, since a message authentication code
accompanies the data during the exchange, preventing on-path attacks.
To use this protocol, it’s necessary to generate certificates for the server (in this
case, the communication server that sends the data to the Emoncms) and for the client
(in this case, the application server that houses the Emoncms). These two sets of
certificates must be applied to the configuration in the server side and also in the client
side so that the link between them via MQTTS works properly.
Since the Emoncms environment doesn’t allow communication via the MQTTS
protocol, the link between the server in the energy storage system and the solar power
41
plant system is done using the Node-Red environment at both ends. The data from the
energy storage system is transmitted to the Node-Red in the communication server of
the solar power plant and then the data is transmitted to Emoncms via MQTT.
3.2 Data Monitoring, Storage and Control
3.2.1 Node-Red
As detailed in the previous section, the processed data can be sent to the Node-
Red Dashboard to be monitored in a graphic format. The processed data from the
communication server is sent to a dashboard node that presents the real-time
information in a graphic format.
Node-Red Dashboard is also used to build a supervisory system to send

commands to the energy storage system by allowing the user to configure the active and
reactive power flow in the grid. The dashboard is assembled in a way that allows the
user to enter with the desired value for 4 parameters for the active and reactive power
flow control. These values are then transmitted to the hybrid inverter that is responsible
for managing this information by charging or discharging the battery bank connected to
it.
3.2.2 Emoncms Dashboard
As detailed in the previous section, the processed data can be sent to the
Emoncms Dashboard to be monitored. As detailed in Chapter 2, the Emoncms
environment gives the possibility to create custom dashboards in order to provide
flexibility regarding the type of dashboard to be generated.
For this research, the idea is to generate a dashboard in Emoncms that represents
the single-line diagram of the system, including the photovoltaic solar power plant and
the energy storage system. Also, this dashboard has to be able to provide further
information about the electrical variables in a graphic format if the user wants to
analyze real-time and historical data.
The communication server is responsible for sending all the necessary data to
the application server with Emoncms. The data is received by the Emoncms
42
environment as inputs. In order for these inputs to be available for usage for the
supervisory system, the user has to configure them as feeds to the database. The
assembly of the database and its manipulation is performed in the application server via
Emoncms, since it has a database tool embedded into it that doesn’t rely on MySQL.
The problem with MySQL in this application would be that, with the increase of stored
data, the process for retrieving historical data would be significantly slowed. So, the
solution implemented by Emoncms developers was to create database engines that don’t
rely on the creation or manipulation of tables on MySQL, but generate files that can be
easily perused by the software when a historical request is made.
The data is recorded at a fixed interval determined by the user when configuring
the feeds. Emoncms provides different processes that can be applied to the input in
order to produce the feed, such as: logging, converting power to energy, performing
mathematical or logical operations with one or more inputs.
In Figure 3-14, some inputs can be seen; in this case, all the inputs are only
logged to the database. In Table 3.1, a sample of the feeds can be seen; the whole list of
configured feeds can be seen in Appendix A.
Figure 3-14 – Sample of inputs in the Emoncms environment – showcasing electrical measured values from the
Fronius inverter
Table 3.1 – Sample of feeds configured in the Emoncms environment
Variable Feed Database Sample Time
43
Total
FaseA Fronius-
FaseB Corrente
FaseC
FaseA
Fronius-
FaseB
Tensao
FaseC
Corrente
PHPFINA 10s
Tensao Fronius-CC
Potencia
Frequencia
Potencia Reativa Fronius-
Potencia Ativa Medicoes
Energia Ativa
Geracao_Fronius
-
Geracao_CA_Fronius
Note: Highlighted variables indicate virtual feeds – not measured values.
With all the necessary feeds already configured, it’s possible to start the
assembly of the supervisory system in the Emoncms dashboard configuration
environment. Each new dashboard is a new screen; the environment for the building of
the screen is seen in Figure 3-15.
44
Figure 3-15 – Emoncms Dashboard assembly screen, showcasing the toolbox with the gadgets to develop the desired
screens
The Toolbox provides a vast variety of gadgets that can be used in order to
create the desired screen. Each gadget is used in conjunction with one or more of the
configured feeds. In this work, the gadgets mostly used are graphs – real time and
historical – and LEDs that represent the status of the energy transmission and status of
charge of the battery banks. There is also the possibility to program HTML functions in
order to further customize the dashboard. In this research, the single-line diagram is
built using this feature in order to upload to the Emoncms Dashboard an image that
represents the diagram.
3.3 Final Architecture
In order to present the whole structure - from the data acquisition until the data
display via the developed screens -, Figure 3-16 is shown.
45
Figure 3-16 - Final architecture of the data acquisition and data monitoring of the developed supervisory system,
showcasing the communication protocols
The schematic shown in Figure 3-16 summarizes the connections made

beginning with the field devices and ending in user terminals in order to acquire, store,
and monitor the data from the solar power plant and from the energy storage system. All
the communication protocols implemented and detailed in the present chapter are also
seen in Figure 3-16.
46
4 Results
In this chapter, the complete supervisory system with its screens in Emoncms
and in Node-Red are presented. The screens were developed as detailed in the previous
chapter and the interactions between the supervisory system and the user are also
described in this chapter. The screens developed are compared to the current
supervisory system in order to better understand the benefits of the new developed
system.
4.1 Solar Power Plant
The dashboard is primarily developed in the Emoncms platform since it provides

a user-friendly interface to build a custom-made supervisory system, providing some
built-in gadgets and allowing customization. Besides, due to the features that Emoncms
offers to its users related to its inputs and database storage, the platform is extremely
adaptable to fulfill any requirements and it’s easily expandable – to insert new variables,
devices and/or features.
This versatility of Emoncms is an important benefit of the platform specially

when comparing it to the current supervisory system available in the solar power plant.
For the latter, it’s cumbersome to add any new features and/or variables to it, since the
system is developed by a third party company and the source is not available. Besides,
the available structure of the current supervisory system is more rigid while Emoncms
provides an environment that allows for any type of required dashboard.
The current supervisory system can be seen in Figure 4-1 and Figure 4-2. As it
can be seen, this supervisory system has two tabs for the user. The main page seen in
Figure 4-1, Cockpit, depicts an overview of the current main electrical and
environmental measurements of the solar power plant. The Evaluation page seen in
Figure 4-2 provides the user with a graphical visualization of the available variables for
a customizable period of time.
APPENDIX A – EMONCMS FEEDS
Figure 4-1 - Main page of the current supervisory system
Figure 4-2 - Evaluation page of the current supervisory system
The current supervisory system doesn’t allow for the creation of an overview of
the schematic of the plant, which would be important for the user specially when
tracking the availability of the plant throughout the day. This overview would also be
interesting to understand the structure of the power plant and to easily detect any
potential problems and/or failures.
The new supervisory system developed in Emoncms allows the inclusion of this
overview besides allowing the creation of any necessary graphics and visual gadgets
required to fully inform the user of the current status and operation of the power plant.
The screens created in the dashboard can be seen in Figure 4-3, Figure 4-4 and Figure
4-5.
48
Figure 4-3 - Main page of the new supervisory system developed in Emoncms
Figure 4-4 - Solar Meter Station page of the new supervisory system developed in Emoncms
Figure 4-5 - Grid measurements page of the new supervisory system developed in Emoncms
The main screen seen in Figure 4-3 shows the some electrical variables: DC/AC
voltage, current of each inverter and the AC voltage/current in the AC side being
measured by each inverter. In this overview, it’s also interesting to note that it’s
49
possible to detect whether the solar power plant is generating power by following the
connections to and from the inverters. If the connection is active, it’s shown as green. If
it’s not, it’s shown as red. Also, the solar radiance and the ambient temperature can be
seen. In this screen, there are also buttons that redirect the user to detailed data of the
energy meter seen in Figure 4-5 and of the solar meter station seen in Figure 4-4. The
user can monitor all the UMG604 variables and all the data from the solar meter station.
The detailed screens are built with graphs. These graphs are also capable of displaying
historical information – if the user wants to see historical information, it’s only needed
to set the desired dates and the graphic will be updated to show the desired time frame.
Any variable that is available in the database accessed by the Emoncms

environment can be added to the dashboard easily graphically or via a gauge. So, if
there’s a requirement to include any variable, the environment is readily able to fulfill
this requirement.
Besides the dashboard, the Emoncms offers a Graphs tab that allows the user to
plot any variable available in the database in a customizable period of time. This tab can
be seen in Figure 4-6. In comparison to the Evaluation tab available in the current
supervisory system seen in Figure 4-2, the Emoncms Graphs tab integrates new data
more easily and effectively since it doesn’t require any other configuration as long as
the data is stored in the database. If the data is stored in the database, the data is
automatically available in the Graphs tab and ready to be plotted.
50
Figure 4-6 - Graph tab available in the Emoncms environment built-in the new supervisory system
Both supervisory systems provide the option of downloading in a CSV format

any available data in its environment for a customizable period of time. The current
supervisory system allows this feature in the Evaluation tab and the new supervisory
system allows this feature in the Feeds tab seen in Figure 4-7.
Figure 4-7 - Feeds tab available in the Emoncms environment built-in the new supervisory system
It’s interesting to compare the graphs built in both systems in order to

understand the difference between the smaller polling time (10 seconds) in the
developed supervisory system and the bigger polling time (5 minutes) in the original
supervisory system. Figure 4-8 shows the graph comparing the acquisition of the
51
measurement of active power of the SMA inverter by both systems. Figure 4-9
highlights the difference between the two supervisory systems by zooming in the data
shown in Figure 4-9.
12
Original System
10 Developed System
8
Power (kW)
0
06:00 08:24 10:48 13:12 15:36 18:00
Figure 4-8 - Active AC power of SMA Inverter - Data acquired by the original and the developed system
12
Original System Developed System
10
8
Power (kW)
0
09:36 09:50 10:04 10:19 10:33 10:48 11:02 11:16 11:31 11:45 12:00
Figure 4-9 - Zoom of active AC power of SMA Inverter - Comparison analysis of original and developed systems
52
It’s possible to see that due to the bigger polling time, the original system loses a
considerable amount of data that relates to variation in the power generation measured
by the SMA inverter. By having a smaller polling time, the developed system is able to
track all the variations that occur throughout the day, including higher peak values and
lower valley values. It’s important to note that due to the polling time, the original
system may lose important data for analysis especially cloudy days or when there are a
lot of weather changes during the day. As the user can configure the system for lower
polling times, the system can overcome this particular challenge. Also, this developed
system provides the choice of determining the polling time for each feed. That means
that different variables can each have a polling time that is more suitable for the type of
data. In the original system, all the variables have the same polling time of 5 minutes.
The new supervisory system also includes graphs in the user interface available
in Node-Red in order to present the user in the communication server with
measurements to facilitate an eventual debugging of any potential communication
failure. The data that is displayed in the Node-Red environment is shown via graphics to
provide the user with a view of the behavior of certain variables during the last hour.
The electrical variables that are on display in the Node-Red dashboard are the DC/AC
voltage, DC/AC current, DC power and AC reactive and active power of each inverter.
Measurements from Janitza UMG604 are also on display in this interface. These screens
can be seen in Figure 4-10, Figure 4-11 and Figure 4-12.
Figure 4-10 - Node-Red User Interface depicting data from the Janitza UMG604 meter
53
Figure 4-11 - Node-Red User Interface depicting some of the data on display of the Fronius inverter
Figure 4-12 - Node-Red User Interface depicting some of the data on display of the SMA inverter
As seen in Figure 4-10, Figure 4-11, Figure 4-12, the user can track the variable
of interest for the last hour by selecting the desired device and searching for the
correspondent graphic. With this feature, it’s possible to trace a potential
communication failure by checking if the communication server is connected correctly
to the field devices.
54
5 Conclusions
The main goal of this work was to develop a supervisory system for a power
plant using the concepts related to the Internet of Things, employing devices that were
network ready and adapting the ones that weren’t. In order to exemplify the usage of
this supervisory system, it was applied to an existing solar power plant installed in the
Escola de Engenharia at the Universidade Federal de Minas Gerais. However, it’s
important to notice that all the knowledge and the techniques used during the
implementation of the supervisory system could be easily transferred not only for any
power plant (with any type of source) but also for any industrial/residential plant as long
as it’s possible to establish a network in the plant.
Using smart devices (such as the power inverters used in the power plant and the
hybrid inverters connected to the energy storage system) and adapting the devices that
are not smart (the solar meter station connected to an intermediary device for example),
it was possible to collect all the available data of the solar power plant and of the energy
storage system. With this data accessible in the network, making it available for the
selected microcomputer, Raspberry Pi, it was possible to develop different servers:
communication and database servers that could help assembly the application server
that houses the supervisory system per se. That way, the supervisory system could be
populated and the screens depicting the entire system could be developed, providing the
user with the means to monitor the operation of the power plant and of the energy
storage system, and also the control of the latter.
The supervisory system was developed with Internet of Things concepts since
this new technology provides a number of advantages. One of these advantages is
potential increased autonomy for the plant since the devices are connected to a network
that provides an information sharing environment (i.e. the devices can exchange data
amongst them in order to operate more efficiently and accurately). Other advantages
include how this technology facilitates the establishment of remote monitoring since it
uses web platforms, such as the ones used here: Node-Red and Emoncms, and increased
reliability, productivity, and efficiency.
55
5.1 Proposed Continuation
One idea for a continuation of this project would be to integrate a prediction of

the energy generation in order to be able to schedule the operation of the energy storage
system. This added capability would be able to predict the power output of the system
based on the weather forecast for the next day and schedule when to charge/discharge
the battery bank based on the forecast, adjusting in real time during the actual operation.
This type of application is in research currently and the preliminary results are
promising for renewable power plants (solar, wind).
In this vein, it would also be interesting to develop an optimization system for

the automatic control of the plant in order to maximize the efficiency of the energy
generation and minimize costs and losses. Currently, it’s possible to see that many
research works are developing optimization controls concurrently with the development
of supervisory system for renewable power plants that are integrated with the Internet of
Things.
Another proposed continuation would be to incorporate the status of any

switches and circuit breakers that are installed in the plant in order to get an overview of
the whole operation of the plant, not only of the measurements and status of the
inverters. It would also be interesting to provide for an authorized user the possibility to
control these devices in order to manage the plant.
56
Appendix A – Emoncms
Feeds
The following tables, Table A.1, Table A.2, Table A.3, Table A.4 , show the
complete feeds configured in the Emoncms environment detailing their name, number
identification and sampling rate. The feeds are categorized by each individual device.
The feeds that are highlighted indicate virtual feeds.
Table A.1 - Fronius Feeds
Variable Feed Database Sample Time Feed ID

Total 89
FaseA Fronius- 90
FaseB Corrente 91
FaseC 92
FaseA 93
Fronius-
FaseB 94
Tensao
FaseC 95
Corrente 96
PHPFINA 10s
Tensao Fronius-CC 97
Potencia 98
Frequencia 119
Potencia Reativa Fronius- 120
Potencia Ativa Medicoes 121
Energia Ativa 123
Geracao_Fronius 109
-
Geracao_CA_Fronius 114
57
Table A.2 - Janitza Feeds

L3 45
Janitza-
L2 46
Corrente
L1 47
L3 56
L2 57
L1 58
Janitza-Tensao
L1L2 59
L3L1 60
L2L3 61
L1 Janitza- 48
L2 Potencia 49
L3 Reativa 50
L1 51
L2 Janitza- PHPFINA 10s 52
L3 Potencia Ativa 53
Total 54
L1 62
Janitza-Energia
L2 63
Reativa
L3 64
L1 65
Janitza-Energia
L2 66
Real
L3 67
Consumida
Total 68
L1 69
L2 Janitza-Energia 70
L3 Real 71
Total 72
58
Table A.3 - SMA Feeds

Total 128
L1 SMA Potencia 129
L2 Ativa 130
L3 131
L1 132
SMA Tensao
L2 133
CA
L3 134
Corrente 125
SMA
Tensao 126
Medicoes CC
Potencia 127
Corrente PHPFINA 10s 142
SMA
Tensao 143
Medicoes CC 2
Potencia 144
Frequencia SMA 135
Frequencia
L1 136
SMA Corrente
L2 137
CA
L3 138
Geracao_CC 140
Geracao_CA SMA 141
Geracao_CC 2 145
Table A.4 - Solar Station Feeds

Umidade_rel 148
Pressao_rel 149
Vento_direcao Estacao_Solar PHPFINA 10s 150
Irradiancia_global 151
Temp_ambiente 152
59
Vento_velocidade 153
Umidade_absol 154
Pressao_absol 155
Irradiancia 156
Celula_referencia
Temperatura 157
60
Bibliographic References
3-K Electric. Datasheet: DMU51 + 60-V02 Spannungsmessumformer. 2020.
Available: https://www.3-k-elektrik.de/index.php?page=shop_product&id=13&lng=de.
Access: 12 January 2021.
ADHYA, S.; SAHA, D.; DAS, A.; JANA, J.; SAHA, H. An IoT Based Smart Solar
Photovoltaic Remote Monitoring and Control Unit, In: International Conference on
Control, Instrumentation, Energy & Communication (CIEC), 2, 2016, Kolkata. p. 432-
436. Available: https://ieeexplore.ieee.org/document/7513793. Access: 12 January
2021.
ÁLVAREZ, E.; CAMPOS, A. M.; GARCÍA, R.; GONZÁLEZ, S.; DÍEZ, C.; Scalable
and Usable Web Based Supervisory and Control System for Micro-grid
Management, In: International Conference on Renewable Energies and Power Quality,
1, 2010, Granada. p. 763-768.. Available: http://www.icrepq.com/icrepq'10/467-
Alvarez.pdf. Access: 12 January 2021.
AQEEL, A. Introduction to Arduino Nano, In: The Engineering Projects. 2018.

Electronic Archive. Available:
https://www.theengineeringprojects.com/2018/06/introduction-to-arduino-nano.html,
BENAMMAR, M.; ABDAOUI, A.; AHMAD, S.; TOUATI, F.; KADRI, A. A Modular
IoT Platform for Real-Time Indoor Air Quality Monitoring, Sensors, v. 18, n. 2, p. 1-
18, Feb. 2018. Available:
https://www.researchgate.net/publication/323181052_A_Modular_IoT_Platform_for_R
eal-Time_Indoor_Air_Quality_Monitoring. Access: 12 January 2021.
BB Smart Worx. Frequent Asked Questions: RS-485 Connections FAQ. 2020.

Electronic Archive. Available: https://www.bb-elec.com/Learning-Center/All-White-
Papers/Serial/RS-485-Connections-FAQ.aspx. Access: 12 January 2021.
BENGHANEM, M.; MAAFI, A. Data Acquisition System for Photovoltaic Systems

Performance Monitoring, IEEE Transactions on Instrumentation and Measurement,
REFERÊNCIAS BIBLIOGRÁFICAS
v. 47, n. 1, p. 30-33, Feb. 1998. Available: https://ieeexplore.ieee.org/document/728784.

CAMACHO, E.; SAMAD, T.; GARCIA-SANZ, M.; HISKENS, I. Control for

Renewable Energy and Smart Grids, The Impact of Control Technology, v. 6, p. 19-
25, 2011. Available:
https://www.researchgate.net/publication/284025192_Control_for_renewable_energy_a
nd_smart_grids. Access: 12 January 2021.
CARNEIRO, A. V. Projeto, Desenvolvimento e Implementação de Microrrede em

Campus Universitário com Tecnologia Solar Fotovoltaica e de Armazenamento,
2017, 177 f. Dissertação (Mestrado em Engenharia Elétrica) – Faculdade de Engenharia,
Universidade Federal do Ceará, Fortaleza, 2017.
Cloudfare. Blog Post: How Does SSL Work? SSL Certificates and TLS. 2020.
Electronic Archive. Available: https://www.cloudflare.com/pt-br/learning/ssl/how-does-
ssl-work/. Access: 12 January 2021.
DE SOUZA, I. Blog Post: Enteda o que é HTTP e o quão importante esse protocolo
é para o seu site, In: Rockcontent. 2019. Electronic Archive Available:
https://rockcontent.com/br/blog/http/#:~:text=HTTP%20%C3%A9%20um%20protocol
o%20de,do%20ingl%C3%AAs%20Hypertext%20Transfer%20Protocol. Access: 12
January 2021.
DUMITRU, C.; GLIGOR, A. SCADA Based Software for Renewable Energy

Management System, Procedia Economics and Finance, v. 3, p. 262-267, Oct. 2012.
Available: https://www.sciencedirect.com/science/article/pii/S2212567112001505.
FORERO, N.; HERNÁNDEZ, J.; GORDILLO, G. Development of a Monitoring

System for a PV Solar Plant, Energy Conversion and Management, v. 47, n. 15-16, p.
2329-2336, Jan. 2006. Available:
http://www.sciencedirect.com/science/article/pii/S0196890405003134. Access: 12
January 2021.
62
FOUREAUX, N.; MARRA, F.; ANTHONY, A. Usina Experimental Fotovoltaica

TESLA Engenharia de Potência: technical report, p. 1-35. Belo Horizonte, May,
2014.
FREITAS, C. M. Blog Post: Protocolo Modbus: Fundamentos e Aplicações, In:

Embarcados. April. 2014. Electronic Archive. Available:
https://www.embarcados.com.br/protocolo-modbus/. Access: 12 January 2021.
GUBBIA, J.; BUYYAB, R.; MARUSIC, S.; PALANISWAMI, M. Internet of Things

(IoT): A vision, architectural elements, and future directions, Future Generation
Computer Systems, v. 29, n. 7, p. 1645-1660, Sept. 2013. Available:
https://www.sciencedirect.com/science/article/pii/S0167739X13000241. Access: 12
January 2021.
HILL, C.; SUCH, M. C.; CHEN, D.; GONZALEZ, J.; GRADY, W. M. Battery Energy
Storage for Enabling Integration of Distributed Solar Power Generation, IEEE
Transactions on Smart Grid, v. 3, n. 2, p. 850-857, Jun. 2012. Available:
https://ieeexplore.ieee.org/abstract/document/6198748. Access: 12 January 2021.
HU, M.; CHEN, Y.; CHEN, Y.; CHANG, Y. Optimal Operating Strategies and
Management for Smart Micro-grid Systems, Journal of Energy Engineering, v. 140,
n. 1, p. 04013011-1 - 04013011-9, Mar. 2014. Available:
https://www.researchgate.net/publication/264365782_Optimal_Operating_Strategies_an
d_Management_for_Smart_Microgrid_Systems. Access: 12 January 2021.
Inductive Automation. Blog Post: What is SCADA? In: Inductive Automation, 2020.
Electronic Archive. Available:
https://www.inductiveautomation.com/resources/article/what-is-scada. Access: 12
January 2021.
JIMÉNEZ-GARCÍA, J. L.; BASELGA-MASIA, D.; POZA-LUJÁN, J. L.; MUNERA,

E.; POSADAS-YAGÜE, J. L.; SIMÓ-TEN, J. E, Smart device definition and
application on embedded system: performance and optimi-zation on a RGBD sensor,
ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal,
vol. 3, n. 1, p. 46-55, Oct. 2014. Available:
https://doi.org/10.14201/ADCAIJ2014384655. Access: 13 April 2021.
63
KHAN, R.; KHAN, S.; ZAHEER, R.; KHAN, S. Future Internet: The Internet of
Things Architecture, Possible Applications and Key Challenges, In:International
Conference on Frontiers of Information Technology, 10, 2012, Islamabad: p. 257-260.
Available: https://ieeexplore.ieee.org/document/6424332. Access: 12 January 2021.
KOUTROULIS, E.; KALAITZAKIS, K. Development of an Integrated Data-

Acquisition System for Renewable Energy Sources Systems Monitoring, Renewable
Energy, v. 28, n. 1, p. 139-152, Jan. 2003. Available:
https://www.sciencedirect.com/science/article/pii/S0960148101001975. Access: 12
January 2021.
L., A. Tutorial: FTP: o que é, como funciona e qual o tipo para gerenciar arquivos
na internet, In: Hostinger Tutoriais. Nov. 2020. Electronic Archive. Available:
https://www.hostinger.com.br/tutoriais/ftp-o-que-e-como-
funciona#:~:text=FTP%20%C3%A9%20a%20sigla%20para,dois%20computadores%2
0conectados%20%C3%A0%20internet. Access: 12 January 2021.
LEE, I; LEE, K. The Internet of Things (IoT): Applications, investments, and

challenges for enterprises, Business Horizons, v. 58, n. 4, p. 431-440, Jul. 2015.
Available: https://www-
sciencedirect.ez27.periodicos.capes.gov.br/science/article/pii/S0007681315000373?via
%3Dihub. Access: 13 April 2021.
MARIANO, F. A. Integração de Tecnologias da Internet das Coisas a um

Supervisor de Energia Através da Programação Orientada à[sic] Fluxos, 2016, 71
f. Monografia (Curso Técnico em Sistemas de Telecomunicações) – Tecnologia em
Sistemas de Comunicação, Instituto Federal de Santa Catarina, São José, 2016.
MARINHO, G. S. P. Apresentação de uma Indústria Sucroalcooleira sob os

Conceitos de Microrrede de Energia Elétrica, 2011, 195 f. Dissertação (Mestrado em
Engenharia Elétrica) – Engenharia Elétrica, Universidade Estadual Paulista, Ilha
Solteira, 2011.
MDN Web Docs. Tutorial: HTTP, In: Developer Mozilla. 2020. Electronic Archive.
Available: https://developer.mozilla.org/pt-BR/docs/Web/HTTP. Access: 12 January
2021.
64
MUKARO, R.; CARELSE, X. A Microcontroller-Based Data Acquisition System for

Solar Radiation and Environmental Monitoring, IEEE Transactions on
Instrumentation and Measurement, v. 48, n. 6, p. 1232-1238, Dec. 1999. Available:
https://ieeexplore.ieee.org/document/816142. Access: 12 January 2021.
NARUTO, D. T. Vantagens e Desvantagens da Geração Distribuída e Estudo de

Caso de um Sistema Solar Fotovoltaico Conectado à Rede Elétrica, 2017, 97 f.
Monografia (Engenharia Elétrica) – Escola Politécnica, Universidade Federal do Rio de
Janeiro, Rio de Janeiro, 2017.
NI, Blog Post: Desenvolvimento de aplicações com Modbus, In: National

Instruments. 2019a. Electronic Archive. Available: https://www.ni.com/pt-
br/innovations/white-papers/14/application-development-with-modbus.html. Access: 12
January 2021.
NI, Blog Post: O protocolo Modbus em detalhes, In: National Instruments. 2019b.
Electronic Archive. Available: https://www.ni.com/pt-br/innovations/white-
papers/14/the-modbus-protocol-in-depth.html. Access: 12 January 2021.
OLIVEIRA, G.; RIBEIRO, R.; ROCHA, T.; COSTA, F.; NETO, J., Controle
Cooperativo do Fluxo de Potência com uso de Bateria, In: Seminar on Power
Electronics and Control – SEPOC 2019, 12, 2019, Natal. Available:
http://sepoc2019.ct.ufrn.br/sepoc2019/images/arquivos/papers/track2/17.-Controle-
Cooperativo-do-Fluxo-de-Potncia-com-uso-de-Bateria.pdf. Access: 12 January 2021.
PAPAGEORGAS, P.; PIROMALIS, D.; ANTONAKOGLOU, K.; VOKAS, G.;

TSELES, D.; ARVANITIS, K. G. Smart Solar Panels: In-situ Monitoring of
Photovoltaic Panels Based on Wired and Wireless Sensor Networks, Energy Procedia,
v. 36, p. 535-545, 2013. Available:
https://www.sciencedirect.com/science/article/pii/S187661021301148X. Access: 12
January 2021.
RAGHVENDRA, N. Blog Post: SCADA System – Components, Hardware, &

Software Architecture, Types, In: Electricalfundablog. 2019. Electronic Archive.
Available: https://electricalfundablog.com/scada-system-components-architecture/.
65
RINALDI, J. S. Blog Post: Modbus TCP vs. Modbus RTU, In: Real Time
Automation. 2013. Electronic Archive. Available: https://www.rtautomation.com/rtas-
blog/modbus-tcp-vs-modbus-rtu/. Access: 12 January 2021.
SICARI, S.; RIZZARD, A.; COEN-PORISIN, A. How to Evaluate an Internet of

Things System: Models, Case Studies, and Real Developments, Software Practice and
Experience, v. 49, n. 7, p. 1663-1685, Jul. 2019. Available:
https://www.researchgate.net/publication/334593703_How_to_evaluate_an_Internet_of
_Things_system_Models_case_studies_and_real_developments. Access: 12 January
2021.
SILVA, F. C. de S. Blog Post: Artigo: Modelos Cliente-Servidor do Modbus TCP e

Publish –Subscribe do MQTT: quando utilizar cada um deles e quais suas
vantagens e desvantagens? In: Novus. 2019. Electronic Archive. Available:
https://www.novus.com.br/site/default.asp?Idioma=55&TroncoID=053663&SecaoID=0
&SubsecaoID=0&Template=../artigosnoticias/user_exibir.asp&ID=618088#. Access:
12 January 2021.
SILVERIO-FERNÁNDEZ, M.; RENUKAPPA, S; SURESH, S. What is a smart

device? - a conceptualisation within the paradigm of the internet of things,
Visualization in Eng, vol. 6, n. 3, p. 1-10, May 2018. Available:
https://doi.org/10.1186/s40327-018-0063-8. Access: 13 April 2021.
SoldaFria. Blog Post: O que é um Arduino, para que serve, como funciona, onde
comprar? In: Blog do Soldafria. 2020. Electronic Archive. Available:
https://www.soldafria.com.br/blog/o-que-e-um-arduino-para-que-serve-como-funciona-
onde-comprar. Access: 12 January 2021.
STADLER, M.; NASLÉ, A. Planning and Implementation of Bankable Microgrids, The

Electricity Journal, v. 32, n. 5, p. 24-29, Jun. 2019. Available:
January 2021.
SUMANGALI, K.; KIRAN, P. Comparative Study of Open Source Tools for Internet of
Things, International Journal of Pharmacy and Technology, v. 8, n. 4, p. 26235-
26241, Dec. 2016. Available:
66
https://www.researchgate.net/publication/316886994_Comparative_study_of_open_sou
rce_tools_for_internet_of_things. Access: 12 January 2021.
TIGGEMANN, H. Sistema de Monitoramento Remoto e Controle de Microrrede

Híbrida Isolada, 2015, 84 f. Dissertação (Mestrado em Engenharia Mecänica) –
Faculdade de Engenharia, Universidade Federal de Minas Gerais, Belo Horizonte, 2015.
VENAYAGAMOORTHY, G. K.; SHARMA, R. K.; GAUTAM, P. K.; AHMADI, A.

Dynamic Energy Management System for a Smart Micro-grid, IEEE Transactions on
Neural Networks and Learning Systems, v. 27, n. 8, p. 1643-1656, Aug. 2016.
Available: https://ieeexplore.ieee.org/document/7387757. Access: 12 January 2021.
WEBER, R. H. Internet of Things – New security and privacy challenges, Computer

Law & Security Review, v. 26, n. 1, p. 23-30, Jan. 2010. Available:
January 2021.
WORTMANN, F.; FLÜCTHER, K. Internet of Things, Bus Inf Syst Eng, v. 57, p.221-
224, Mar. 2015. Available: https://doi-
org.ez27.periodicos.capes.gov.br/10.1007/s12599-015-0383-3. Access: 13 April 2021.
67

D S S P P: Evelopment of A Upervisory Ystem For A Ower Lant

Uploaded by

Copyright:

Available Formats

D S S P P: Evelopment of A Upervisory Ystem For A Ower Lant

Uploaded by

Document Information

Original Description:

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

D S S P P: Evelopment of A Upervisory Ystem For A Ower Lant

Uploaded by

Copyright:

Available Formats

DEVELOPMENT OF A SUPERVISORY SYSTEM FOR A POWER

FERNANDA MOURA QUINTÃO SILVA

MASTER’S THESIS DEFENSE DATE: 06/04/21

DEVELOPMENT OF A SUPERVISORY SYSTEM FOR A SMART

FERNANDA MOURA QUINTÃO SILVA

Master’s Thesis submitted to the Examination

Concentration Field: Power Engineering

Advisor: Prof. Dr. Thales Alexandre Carvalho

I would also like to acknowledge the team involved in the project at

Finally, I would like to express my gratitude to my parents, my sister, and to

1.1 Objectives ............................................................................................... 2

1.2 Methodology ........................................................................................... 3

1.3 Scientific Contributions .......................................................................... 7

1.4 Text Organization ................................................................................... 7

2 Bibliographic Review ..................................................................................... 9

2.1 Micro-grids ............................................................................................. 9

2.2 Traditional Supervisory Systems .......................................................... 12

2.3 Internet of Things .................................................................................. 16

2.3.1 Smart Devices .................................................................................. 18

2.3.2 Key Challenges in IoT ..................................................................... 19

2.4 Open-Source IoT Tools for Supervisory System .................................. 20

2.4.1 Node-Red ......................................................................................... 20

2.4.2 Emoncms ......................................................................................... 23

2.5 Micro-grid Tesla ................................................................................... 24

2.5.1 Photovoltaic Solar Power Plant Tesla ............................................. 25

2.5.2 Energy Storage System .................................................................... 28

3 Supervisory System ...................................................................................... 31

3.1 Data Acquisition and Processing .......................................................... 32

3.1.1 Solar Power Plant ............................................................................ 32

3.1.2 Energy Storage System .................................................................... 40

3.2 Data Monitoring, Storage and Control.................................................. 42

3.2.1 Node-Red ......................................................................................... 42

3.2.2 Emoncms Dashboard ....................................................................... 42

3.3 Final Architecture ................................................................................. 45

4.1 Solar Power Plant.................................................................................. 47

5.1 Proposed Continuation .......................................................................... 56

Appendix A – Emoncms Feeds .......................................................................... 57

Bibliographic References ................................................................................... 61

As an interested party in developing clean energy, the department of Electrical

With the advancement of technology in wired and wireless networks, there is a

This dissertation work is part of the multidisciplinary project under development

Currently, the analyzed solar power plant is monitored by a supervisory system

The main idea is to provide a custom-made monitoring and control system of

The specific goals are:

 Develop a low-cost open-source flexible supervisory system for the

important to find software that is easily programmable in order to create a supervisory

There is an array of open-source software options that would fulfill the

1.3 Scientific Contributions

Throughout the development of this dissertation, some scientific contributions

 “A Long Term Evaluation of Photovoltaic Systems under Power Quality

1.4 Text Organization

In Chapter 2, there is a bibliographic review on micro-grids and the systems

The main conclusions and proposed continuation are presented in Chapter 5.

Micro-grids combine modular energy sources, such as wind, photovoltaic, with

Distributed energy is the main generation system in micro-grids architectures.

Micro-grids are an effective tool to employ in the use of local sources of

2.2 Traditional Supervisory Systems

SCADA (supervisory control and data acquisition) is a system of hardware and