Smart and Secure Energy Solutions for Future Mobility

Results of the Energy ECS Project at Lapland University of Applied Sciences

Editor:

• Leinonen Jukka, MEng, Specialist, New Industry, Lapland University of Applied Sciences

Authors of the articles:

• Autioniemi Juha, BEng, Specialist, Smart Built Environment, Lapland University of Applied Sciences
• Etto Jaakko, M.Sc. (El. Eng), Lecturer, New Industry, Lapland University of Applied Sciences
• Hienonen Mirka, BEng, Specialist, Digital solutions Expertise Group, Lapland University of Applied Sciences
• Juustovaara Jouko, M.Sc. (El. Eng), University student, Lappeenranta-Lahti University of Technology LUT
• Kenttä Kari, MEng, Lecturer, New Industry, Lapland University of Applied Sciences
• Lehti Lauri, Project Manager, Unikie Oy
• Leinonen Jukka, MEng, Specialist, New Industry, Lapland University of Applied Sciences
• Narkilahti Aleksi, BEng, Specialist, Digital Solutions Expertise Group, Lapland University of Applied Sciences
• Sinisalo Tuomas, Embedded Software Engineer, Aurora Powertrains Oy
• Suopajärvi Jussi, M.Sc. (El. Eng), Specialist, New Industry, Lapland University of Applied Sciences
• Öörni Jyri, M.Sc. (Eng), Chief Technology Officer, Merus Power Oyj

Preface:

• Leinonen Jukka, MEng, Specialist, New Industry, Lapland University of Applied Sciences

Summary

This collection of articles presents the main results of the Energy ECS project in which Lapland University of Applied Sciences participated. The content of the articles is related to smart energy solutions, such as micro- and nano-grids, bi-directional charging and visualisation of energy flows, as well as simulation of autonomous driving. The authors of the articles participated in the project and most of them are specialists and lecturers from Lapland UAS, but also from other organizations.

The Energy ECS project brought together 28 partners from eight European countries to revolutionise smart and secure energy solutions for future mobility. The project consisted of six use cases and was funded by the European Union Chip Joint Undertaking. In addition, national funding in Finland was provided by Business Finland. The project was implemented between 1.6.2021 and 30.11.2024.

---

Preface

Jukka Leinonen

The future of mobility is rapidly evolving, driven by innovations in electric vehicles (EVs), autonomous driving technologies, and a growing need for sustainable, intelligent energy systems. The future of mobility is not just about cleaner vehicles or more efficient transportation networks, it is about creating intelligent systems that seamlessly connect energy production, storage, and consumption. This requires advanced technologies that can optimise energy flow, improve grid stability, and enhance the overall efficiency of transportation.

At the heart of smart energy solutions for future mobility are microgrids, localised energy systems that integrate energy production, energy storage and consumption. The microgrid can operate independently, drawing energy from renewable energy sources or energy storage, or connected to an electricity distribution network, later called a grid.

Energy storage is a crucial component of smart energy systems for future mobility. As renewable energy sources like solar and wind are intermittent, it is essential to store energy when production exceeds demand and release it when there is a shortfall. In a microgrid, energy storage serves as a buffer, ensuring a continuous power supply even when renewable energy generation is low or fluctuates.

A groundbreaking solution that enhances the interaction between electric vehicles and the grid is bidirectional charging technology. While traditional EV chargers only allow for energy to flow into the vehicle, bidirectional chargers enable energy to flow both ways. This capability, known as Vehicle-to-Grid (V2G), allows EVs to not only draw power from the grid for charging but also return excess energy back to the grid or directly supply power to homes, businesses, or other infrastructure.

This flexibility offers multiple benefits. For EV owners, bidirectional charging can lower energy costs by allowing them to sell surplus energy back to the grid or use it for their own needs, such as powering their home during peak hours. For the broader energy ecosystem, it enhances grid resilience and reduces pressure on centralised power stations, making the system more sustainable and responsive to changes in energy demand.

Artificial intelligence (AI) based control plays a crucial role in managing the flow of energy within microgrids. By analysing real-time data, including energy consumption, storage capacity and weather forecasts, AI can predict energy production and consumption, adjust charging and discharging rates and optimise the use of renewable energy. For instance, if solar power generation is high during midday, AI systems can direct excess energy to charge EVs or store it in batteries for later use. Conversely, when solar power is low or demand increases, the system can release stored energy or draw it from the grid.

Together, these smart energy solutions create an integrated, dynamic system that enhances energy efficiency, reduces environmental impact and energy costs, and supports sustainable and efficient mobility of the future.

Smart and secure energy solutions for future mobility project

A large consortium project, smart and secure energy solutions for future mobility (Energy ECS), aimed to develop a set of technologies to improve the digitalisation of e-mobility systems and related energy solutions (Energy ECS, 2025).

Lapland University of Applied Sciences participated in the Energy ECS project, which started in June 2021 and ended in November 2024, together with 27 project partners. The partners were from eight European countries, and the total project budget was about 33 M€. The project was funded by the European Union Chip Joint Undertaking and national funders. The project was based on the six use cases shown in Figure 1. (Energy ECS, 2025)

Figure 1. Six use cases of the Energy ECS project (Energy ECS project).

UC1 aimed at extending the operational range of drones by enabling the drone battery charging on the rooftop of an electric bus. The project implemented, among other things, a dock for the drone with wireless charging, sensors for landing on the dock and a photovoltaic cell for charging. UC2 aimed to develop smart containers for more energy efficient, safer, and smarter transport and was separated into two dedicated use cases. UC2a focused on the tracking and monitoring of intermodal containers on ships and UC2b on temperature-controlled air cargo containers. For both use cases, the aim was to develop energy management for the systems and photovoltaic units for energy harvesting and for use case 2b also aimed at minimising the overall mass of the container. (Energy ECS, 2025)

The objective of use case 3 was to develop a microgrid controller. The microgrid is connected to renewable energy production, a battery storage and electric vehicles with bidirectional charging capability, whose operation is controlled and predicted by the microgrid controller. The objective of use case 4 was to develop equipment for electric vehicles compliant with the bidirectional charging standard and to demonstrate Vehicle-to-grid (V2G). Electric vehicles, related charging technologies and their impact to electric grids play a central role in the mobility and energy related transformations taking place in society. (Energy ECS, 2025)

The objective of use case 5 was to develop energy harvesting and the necessary electronics to power the car tyre sensor. The sensor will continuously monitor the condition of the car tyre, both for improving safety and for autonomous driving. The objective of use case 6 was to develop autonomous driving technology and demonstrate it on an electric bus by driving it autonomously from the parking slot to the charging station and back in a closed bus depot area. (Energy ECS, 2025)

Lapland University of Applied Sciences participated in the implementation of the project in use cases 3, 4 and 6 with three competence groups: New Industry, Digital Solutions and Smart Built Environment.

This collection of articles is based on the main project results of the Energy ECS project of Lapland UAS. The articles are written by the project staff of Lapland UAS and four external contributors.

Source

Energy ECS, 2025. The project. Referenced 10.3.2025. https://energyecs.eu/project/

---

Energy Measurements at Demonstration Site

Juha Autioniemi

Introduction

This article introduces the energy measurement system that was installed at the demonstration site in Energy ECS project. The project aimed to develop smart technologies to improve the digitalisation of e-mobility systems and related energy solutions. The energy measurement system was part of Energy ECS project Use Case 4 (UC4), that concentrated on developing and researching bidirectional onboard and offboard charging technologies for electric vehicles to enable smart V2G applications (Energy ECS, 2025).

Demonstration site was located in a Finnish Lapland in a small village called Pyhätunturi. Pyhätunturi is known for Pyhä Ski Resort. The local travel industry company Sunday Morning Resort operates at the demonstration site. Sunday Morning resort is a holiday resort that offers accommodation in various hotel rooms and cabins and many kinds of activities, for example electric snowmobile safaris (Experience Pyhä, 2024).

In Energy ECS project this site was an excellent demonstration environment for small scale microgrid-system and other electrical components related to this microgrid. The site had invested in energy efficiency and green energy. There were for example a solar power plant, a geothermal heating system, various electric vehicles, LED lighting and a smart building automation system that controls heating and ventilation based on sensor data (Experience Pyhä, 2024).

During the project, the aim was to install around ten bidirectional charging stations developed in the project for eSleds use at the demonstration site. The energy meters to be installed at the demonstration site were designed to measure bidirectional energy from EV charging stations and other significant electricity consumption points. Web-based energy flow visualisation software was implemented to display the real-time measurement data from the energy meters. Measurement data also allowed to perform statistical analyses and reports on energy consumption and production and to detect anomalies in the energy data.

Hardware used in energy measurements

The energy meter used at the demonstration site was Shelly 3EM (Figure 1.). This energy meter was chosen among many other energy meters because it is cheap, has small physical size and has multiple ways to connect into existing network. Small size was an important property because many of the existing electrical cabinets at the site were already full or almost full.

Figure 1. Shelly Energy Meter. Source: https://www.shelly.com/en-fi/products/product-overview/shelly-pro-3em-120a.

Shelly´s energy meter could be connected into two different wireless networks, and it also had a wired network connector. Device can be commissioned using Bluetooth connection and mobile phone or through network connection. It supported many modern IoT-protocols. One of the supported protocols was MQTT (Message Queuing Telemetry Transport), which was also used in this case. MQTT is a lightweight, safe, and reliable protocol for bi-directional communication between device and cloud. (MQTT.org, 2024).

Installation of energy meters

There were two versions available of the energy meter. The first version was able to measure electric current up to 120 amperes (A) and the second version up to 400 A. Both versions were used on the demonstration site. 400 A meters were used to measure the main grid, main building and maintenance hall supplies. Smaller meters were used to measure all other consumption or production. A total of 25 energy meters were installed in the site. In figure 2 is an installation diagram of meters. From the diagram can be seen relations of the meters, for example, if meter data was included in another measurement point.

Figure 2. Pyhä demonstration site energy meter installation diagram.

All the installed energy meters were bi-directional, which means they were able to measure both consumption and production. Solar power plant was the only one that could produce electricity, but in the future some of the EV (Electric Vehicle) chargers and especially electric snowmobile chargers might be upgraded to support vehicle to grid energy transfer (V2G). Using V2G chargers in the future with electric snowmobiles makes it possible to use snowmobile batteries to balance microgrid and cut power peaks.

Smaller energy meter (120 A) measurement points were chosen based on the best knowledge that those points could have continuous or significant seasonal electrical load or production now or in the future. Esled chargers were a good example of seasonal load as they consume energy when charging in the winter season, but nothing happens in the summer season. This seasonality could change to continuous behaviour, if chargers are upgraded to support V2G in the near future. 400 A energy meters gave the bigger picture of energy flow at the site.

Energy meter data acquisition system architecture

Data acquisition system consisted of energy meters and industrial router. Energy meters were connected to the local network of the demonstration site. Most of the meters were connected using wireless connection, but few were connected using wired connection because of the poor wireless signal in the edges of the demonstration site. Local wireless network was password protected and separated from public wireless network, which was meant for resort guests, for security reasons.

There was a single Advantech ICR-3231 industrial router connected to the same network with energy meters. Router management software was remotely accessible via mobile data connection using OpenVPN software. Mobile data connection was separated from local network, so that local network was not reachable via mobile connection. The purpose of the router was to act as the central unit of the data acquisition system. The router was programmed using Node-RED and the router also had integrated MQTT server software.

Energy meters used MQTT to transfer real time data to MQTT server located at the router. In the router Node-RED program connected to this same local MQTT server and processed all data received from different energy meters. Energy meter MAC-address was used as unique identifier in database and program. The identifier was also mapped into more user-friendly numbering starting from number 1. Exact numbering can be seen in Figure 2, installation diagram.

Data processed at router was sent to Lapland UAS cloud platform, where another MQTT server received the data. Received data was saved into database and shared in real time to energy data visualization system. Historical data in database could be accessed to create statistical reports, charts, analyses and energy usage reports. Main components and high-level architecture of the data acquisition system is graphically described in figure 3.

Figure 3. Architecture of the data acquisition system.

Data acquisition system was developed to be modular. New energy meters are easy to add into the system if new buildings or devices are added to the site or even more detailed energy data is needed at some location. Data acquisition system also supports adding other kinds of sensors directly to industrial router or using suitable bus adapter. One example could be a local weather station that measures microclimate of the site, and which data could be used to optimize and validate heating and ventilation control of the buildings even more precisely.

Energy data interface

Energy data was shared for external systems at the demonstration site in multiple ways. The primary interface to access data was Modbus TCP interface and secondary was MQTT. Modbus is quite old but a simple and reliable protocol to transfer data between automation and control devices. Modbus protocol uses registers to store data. Single register can hold 16-bits data. Modbus TCP practically just means that Modbus protocol works using IP network. (Real Time Automation, 2024)

An external system could be building automation system or in this case it was microgrid controller. The microgrid controller acted as Modbus TCP master unit and regularly polled current power data from Modbus interface. Data from different energy meters could be useful to optimize energy usage at the demonstration site.

Modbus TCP interface at router had 50 Modbus input registers. Each register in Modbus protocol has the size of 16-bits. A single register can for example hold 16 ON/OFF values or a single numerical value between 0-65535 or combination of two registers can for example hold 32-bit floating point number.

Demonstration site energy meters actual power values were stored into input registers. Each value was stored as datatype float, so each value took two (2) registers. Modbus TCP energy data interface is described as Modbus interface register table. Part of this table is shown in Table 1. Meter numbering is linked into a diagram shown in Figure 2.

Table 1. Example of Modbus register table.

Register	Meter number	Sensor	Unit	Datatype	Scaling
30000	27	Power	W	float (BE)	1
30002	26	Power	W	float (BE)	1
…	n	Power	W	float (BE)	1

Sources

Energy ECS, 2025. About the project. Referenced 8.5.2025. https://energyecs.eu/project/

MQTT.org, 2024. MQTT-protocol. Referenced 24.9.2024. https://mqtt.org

Experience Pyhä, 2024. Home site of Experience Pyhä. Referenced 26.8.2024. https://experiencepyha.fi/en

Real Time Automation, Inc, 2024. An introduction to MODBUS TCP/IP. Referenced 24.9.2024. https://www.rtautomation.com/technologies/modbus-tcpip

---

Development of Energy Flow Visualization Software

Mirka Hienonen

Introduction

As a part of Energy ECS and UC4 projects, a prototype web application was created in React with Typescript, for user friendly and accessible display of consumption (positive values) and energy yield (negative values), to visualize energy flows of Pyhä demonstration site in real time. In the future, the prototype application could also be used for other similar sites, as it was made customisable to meet the needs of different real-world changes, such as changes to devices or new locations on the site.

The demonstration site -a local travel industry company called Sunday Morning Resort- consists of numerous buildings that have been referred to as locations in this article. Each location can hold multiple energy meters, with devices connected to them. Some meters have multiple devices, while others have only one. For example, energy meters for eSleds, electric snowmobiles, have three chargers connected to them. This means a single meter can show data of three eSleds at once, in phases one to three or the total of the phases in phase four or occasionally referred as ‘Phase Total’.

The visualisation software consists of three main pages, that are home page, map page and history data page. In addition, there are control pages that allow users to add, delete and edit energy meters, device groups and devices connected to the software.

Home page of the visualisation software

Energy flows of the demonstration site can be studied in an overall display at home page (figure 1), which shows data from device groups, such as buildings, eSled and car chargers, solar powerplant, main grid and total reading of all devices at microgrid at the center of the meter network.

The visualisation of energy flows is colour coded so, that energy consumption is shown by orange and red spheres and energy yield is shown by green and white spheres. The spheres change size depending on the reading on the meter and can also change direction to travel from or towards microgrid or between meters. The reason for direction change is change in the energy flow, turning from positive (consuming) to yielding (negative) or vice versa. The feature was implemented to show flow change in scenarios, such as eSleds yielding energy back to the microgrid.

Figure 1. Home page view of meters and their dataflow.

Device groups have been defined during the process of adding devices (Figure 2) and their energy meters in the application and each meter has been programmed to show data for specific group and its devices.

Figure 2. Menu of phases for a device.

By clicking a meter, user can choose to view detailed flow of a specific device, belonging in a device group a meter represents (Figure 3). The values are updated in real time, giving a more detailed explanation of the total sum of the meter.

Figure 3. Detail window of devices meter.

For more specific consumption and yield study, a so-called master meter (Figure 4) was implemented in the menu. The meter’s capacity has been set to a maximum value of 315 by default, based on possible overall consumption the site can handle and to maximum yield of -30, which is the site’s current maximum yield based on solar powerplant.

When the meter displays consumption the progress bar will move clockwise, and change colour from yellow to orange and finally red, the closer it gets to maximum value. While showing positive values, the progress bar is green and will move counterclockwise.

When closing in on maximum value, the meter’s progress bar will change its colour from white, to yellow and finally red, and while closing in on -30, the meter will turn green instead.

Figure 4. Different simulation stages of the master meter.

The application can also detect anomalies in data, such as long nan values and informs the user about this by popping a warning icon next to the info-button. By clicking the icon, the user can view on what device the anomality was detected on and act accordingly.

When the message has been read, it and the warning icon can be closed by simply clicking the OK button. Anomaly detection consists of both –simulated and real data versions. The simulated version was made for testing purposes only but was left in code in case need for future testing rises. Simulation can be switched off from code, so anomaly detection handles real data anomalies instead.

In Figure 5 below, a simulated example tells a long Nan value has been detected at Main Grid, meaning no data is coming through. Same message applies to real data, where user gets informed about anomaly type and device it was detected on.

Figure 5. Anomality detection warning.

Map Page of the visualisation software

To study energy flows based on specific locations on the test site, user can view them on a map page that has been locked on the test site by default, but can be zoomed in, out, or moved to any direction by dragging the displayed view by mouse.

In addition to this, the user can place new meters on map by simply clicking a desired spot (Figure 6). Once a new meter has been added, it can be edited by clicking the meter to open a meter specified detail window.

Figure 6. User has clicked map to add a meter.

From the window (Figure 7), the user can change the meter’s name, select and unselect device groups and devices, tick suppression for negative values and remove the meter from map and database, should need for this arise.

From the image below, we can see the differences between a window that has been opened for the first time, selection stage, and when window is opened after saving the changes.

Figure 7. Different states of the meter specified window.

After the window state has been saved, selected devices are passed onto the meter.

Each meter on the map (Figure 8) shows the summary of the location’s energy usage (selected devices) and colour codes it by meter to inactive (white), consuming (yellow) and yielding (green).

Figure 8. View of the map page.

To scale the application to meet different scenarios and possible changes to the demonstration site’s device structure, several simulations have been added to the application during development to ease the testing and inform developers and users about the results of changes.

Battery simulation has been added as part of the meter network, where in joins as part of the active meters by a click of the switch. When the switch is on, the battery meter is active and starts to yield energy for microgrid. Changes in battery happen every 2000 seconds, updating the value by 0.5 till it runs down to zero. When the battery meter is turned off from the switch, it stores the latest displayed value and simulation continues where it left off, when battery is turned on again. If the computer’s cache is cleaned, simulation resets and starts from 0.0kWh.

The battery’s size has been set to 75.0kWh by default, but user can scale it up and down by precision of tens by mouse wheel or clicking from the buttons (Figure 9). If battery size needs to be reset mid simulation, it can be done by clicking the reset button from the menu.

Users can also choose to give the battery a pre-set value, to simulate the amount of stored energy. Once a value that is above zero has been set and the switch is turned on, spheres begin to move towards microgrid, growing smaller in size the smaller the ‘amount of stored’ energy gets. Once simulation reaches value of zero, spheres stop rendering.

Figure 9. Battery simulation control window with pre-set value.

The demonstration site’s overall energy simulation can be studied from the menu’s master meter. The meter works as a button that opens simulation menu (Figure 10), where user can select values from maximum consumption and yield. The meter reacts to changes in real time and the progress bar changes colour and direction based on selected values and microgrid’s current reading. Selected values are saved when the simulation window is closed and will stick until reset to default values from menu, or until computer’s cache is cleaned to end the current simulation.

Figure 10. Master meter simulation settings, displaying default values.

Presentation of history data

To study history data from the demonstration site, users can navigate to data charts via device menu. Each device has the option to open a unique page of history data, specified for the selected device (Figure 11).

Figure 11. History data displayed in bar-chart.

From the chart page, the user can select a specific time-period and whether they want data to be displayed by hours or days. In addition, the user can choose to select what sort of history data they want to view from the device from seven options (Figure 12) total. By default, solar powerplant has been set to show values of energy yield and other devices of energy consumption.

Figure 12. Data type selector for charts.

Besides these features, the user can also control which phases they want to study, by deselecting unwanted ones from the phase menu above the chart. Data can also be viewed in real time by selecting the data type and real time option (Figure 13).

Figure 13. Real time data-chart.

The energy flow visualization software helps monitor and analyse real-time and historical energy usage, enabling users to identify inefficiencies and trends. This aids in optimizing energy consumption, planning future needs, and implementing cost-saving strategies.

Data analysis and its uses based on history data

Along with the option to view historical data, a data analysis was conducted in Jupyter Notebook using Pandas—a common Python library for handling datasets—to identify the hours when multiple electric vehicles were charging simultaneously. The analysis consisted of measurement data from 22 charging stations from October 2023 to May 2024. The primary goal of this analysis was to collect data for evaluating the power quality measurements of the demonstration site. Additionally, the results can aid in planning cost-efficient charging times.

To find the active chargers, the dataset was cleaned of everything, but devices with specific IDs and the four phases of filtered devices L1, L2, L3 and phase total L4. It was also confirmed in code that dataset was set to use Europe/Helsinki time, to make sure the time stamps would show the correct hours.

For eSled chargers, the three phases represented different devices, connected to a single energy meter, meaning one meter could have three devices connected to it at a time. In the analysis, this meant that meters with eSleds connected to them were counted as one, even though more than one phase was active at the same time.

To know which of the selected devices and meters had been active, a refiltering was made to find phase and datetime, when reading of the phase had been 1000 watthours or over. Also, negative readings were converted to positive, due to some known errors in gathered data, that caused non-existing energy yielding periods for some of the devices. After this, data was cleaned again to remove inactive periods and rows with duplicate hours for the same device.

Next step involved counting the number of active chargers during each hour. This was done based on device ID and phase connected to it. A second refiltering connected the active charges by hour, to find out the most active hours based on the number of chargers. Finally, the data was sorted first by number of devices per hour and then by the largest consumption per hour, to find out what devices were consuming the most power at certain times (Figure 14).

Figure 14. Phases consuming 1000Wh or over

In the future, information gained from the analysis could be combined with forecasts of exchange electricity prices to optimize the use of vehicle-to-grid (V2G) chargers—determining whether they should be used for charging vehicles or supplying energy to the microgrid and battery.

In conclusion, the analysis offered plenty of information, both regards the analysing of power quality and future planning of energy usage of the demonstration site, helping parties involved to make the most ecological choices while planning their use of electricity.

---

Environmental Testing of Bidirectional Charging Equipment

Juha Autioniemi

Introduction

Environmental testing of bidirectional charging equipment was carried out as part of the Energy ECS project. The Energy ECS project aimed to develop smart technologies to improve the digitalisation of e-mobility systems and related energy solutions. This environmental testing was part of Energy ECS project Use Case 4 (UC4), that concentrated on developing smart bidirectional onboard and offboard charging technologies and applications for electric vehicles (Energy ECS, 2025).

Main goal of the environmental testing was to find out how Aurora Powertrains electric snowmobile and specifically its bidirectional charging related peripheral systems work in extreme conditions. Peripheral systems include battery pack, battery management system (BMS), vehicle charging control unit (VCCU) and battery pack thermal management system. VCCU was developed in the project and other peripherals were upgraded or enhanced to fully support bidirectional charging.  The testing facility was Lapland University of Applied Sciences environmental room (Figure 1) that is located in the university’s campus area in Rovaniemi.

Figure 1. Lapland UAS environmental room. (Photo: Henna Harmoinen)

Finnish company InterControl Latinki 11 V2G charging station was used for bidirectional charging testing and typical charging tests were carried out using Juice Booster 2 electric vehicle charger. Latinki charger supplies direct current (DC) and Juice charger supplies alternating current (AC).  Third party electric vehicle charging equipment was used, because the development of bidirectional charging devices in the project was delayed.

Environmental room technical data

The environmental room is built in the year 2023 and put to use on early 2024. Room temperature can be controlled between -40°C…+30°C using three condenser units. Each unit can be controlled separately, that makes it possible to create test scenarios, for example to test different kinds of building materials. Room relative humidity can be controlled in temperatures above 0 °C using humidifier.

The floor size of the room is 10 x 7 meters, and its height is approximately 3 meters. Height can vary in different parts of the room. It has a sliding door (Figure 2.) which doorway is approximately 3 m wide and 2.5 m tall. It is possible to drive a car into the room.

Figure 2. Front side and sliding door of environmental room. (Photo: Henna Harmoinen)

The environmental room is remote controlled and runs on time-based schedules. The room is equipped with different kinds of sensors and measurement devices including temperature and humidity sensors, energy meters, thermal camera and multiple IP-cameras. Sensors and measurements devices can be customized depending on test setup using Lapland UAS construction and energy technology laboratory equipment.

Test setup and test scenarios

For testing all of the devices were placed into the environmental room. Basic test setup and both electric snowmobile (eSled) and Latinki V2G charger can be seen in Figure 3 located in environmental room. During testing environmental room temperature and energy parameters between grid and charger among many others were constantly recorded. Testing was also visually monitored and recorded using three IP-cameras.

Figure 3. Test setup in environmental room.

The original test plan was to study the following questions:

1. Does the eSled cold start after 72 hours at –20 °C, does it move and does it start charging with a DC and AC charger?
2. How well does the eSled’s battery thermal management system work at -20 °C to -40 °C and how much energy does it consume?
3. Does bi-directional charging work in a cold environment (-20 °C)? Is the eSled suitable for use as energy storage at different temperatures (-20 °C and +30 °C)?

To find out answers to these questions a total of three different tests were run. Test number 1 gave answer to question 1, test number 2 showed how thermal management system performed and finally test number 3 was carried out to find out how bi-directional charging worked in different environmental conditions.

Test results

Test 1 – Cold start and AC/DC charging

Environmental testing started with test number 1. The test setpoint temperature was set to -20 °C and the test time period was from 3^rd October 17:00 to 6^th October 2024 17:00. The test length was 72 hours. Recorded test cycle is presented in figure 4. In the beginning of the test electric snowmobile battery state of charge (SOC) was 87.5% and battery cell average temperature was +25 °C. Electric snowmobile was turned off after these values were read.

Figure 4. Room temperature and setpoint during test number 1.

After 72 hours the electric snowmobile was turned on and it started with no problems. Electric snowmobile moving capability was tested and it moved forward and backward. When the electric snowmobile was turned on its battery SOC was 86.5 % and battery cell average temperature -17 °C.

Both AC charger and DC charger were tested and plugged into electric snowmobile. Charging successfully started in both cases. Charging power was low due to cold batteries.

Test 2 – Battery thermal management system

In test 2 the electric snowmobile was plugged into AC charger, so that it can charge its battery and use electricity from grid to run battery thermal management system to heat battery pack. The test time period was from 11^th October 16:10 to 14^th October 04:00. The test length was a total of 60 hours. The recorded test cycle is shown in figure 5. The setpoint was set to -20 °C for the first 24 hours and then set to -40 °C for 24 hours and finally set to -20 °C for 12 hours. As can be seen from figure 4 the environmental room did not reach -40 °C in this time period. The lowest value recorded was -35 °C.

Figure 5. Room temperature and setpoint during test number 2.

In the beginning of test 2 electric snowmobile SOC was 100% and battery cell average temperature was +27 °C.

Electric snowmobile thermal management system worked mostly as expected during the test and its heating power is enough to keep battery pack temperature inside predefined limits. Currently electric snowmobile battery heating starts when any of the battery cells drops down to +5 °C and stops when any of the cells reaches +15 °C. In figure 6 can be seen that thermal management system works. Battery pack temperature rises as expected when temperature drops below limits.  The heating cycle is longer when the room temperature is lower.

It can be seen from figure 6 that thermal management system had some software related bugs that occurred in lower temperatures. This can be seen to happen in graph after midnight 00:00 13^th October. Temperature curve reacts differently to heating compared to time before or after this period. This bug found during test 2 was fixed after this test session.

Figure 6. Room temperature, setpoint and battery temperatures during test 2.

In figure 6 it can be seen power that in the beginning of test electric snowmobile drawed higher power from grid when plugged in to AC charger to fully charge its battery. When battery pack SOC reached 100%, there is a ripple in the curve that was caused by AC charger. The heating cycle can also be seen from figure 6 when power draw from grid rises to 550 watts (W) level, this level is marked in the graph using blue constant line. Power drawn from the grid was at the expected level, because nominal power of heating element is 550 W.

From figure 7 can be read that single cycle (heating + cooling down) length is about three hours in -20 °C. When the room temperature is -35 °C single cycle takes about 4 hours.

Figure 7. Power drawn from grid during test 2.

In table 1 are energy meter watt-hour (Wh) readings 12^th October at 00:41 and at 12:40. This almost 12 hours long time period included four cycles, and four cycles took approximately 2384 Wh of energy. One day energy consumption in static -20 °C ambient temperature is approximately 4,8 kilowatt-hours (kWh).

Table 1. Measured heating energy consumption at temperature -20°C.

Room temperature -20°C
Cycles	Start time	End time	Energy at start	Energy at end	Length	Energy [Wh]
4	12.10.2024 0:41	12.10.2024 12:40	779805,39	782189,14	11:59:00	2383,75
1					2:59:45	595,9375
					Day	4767,5

In table 2 are energy meter watt-hour readings 13^th October at 09:00 and at 16:59. Time period between these readings is almost 8 hours long and includes 2 cycles and these cycles took approximately 2207 Wh of energy. Calculated one day energy consumption in static -35 °C ambient temperature is approximately 6,6 kWh.

Table 2. Measured heating energy consumption at temperature -35°C.

Room temperature -35°C
Cycles	Start time	End time	Energy at start	Energy at end	Length	Energy [Wh]
2	13.10.2024 9:00	13.10.2024 16:59	786913,8	789120,71	7:59:00	2206,91
1					3:59:30	1103,455
					Day	6620,73

Test 3 – Bi-directional charging testing

In beginning of test number 3 bi-directional charging was tested with electric snowmobile together with InterControl Latinki 11 DC charger. The charger can handle power up to 11 kilowatts (kW) in both directions. In test 3 electric snowmobile maximum charging power was 4 kW and discharging power was 2,5 kW. The electric snowmobile battery pack capacity was approximately 21 kWh. Discharge capacity in this test was much lower due to DC voltage safety limits in Latinki charging station power electronics. Aurora Powertrains electric snowmobile battery pack nominal voltage is 120 VDC which is much lower than a regular electric car battery pack voltage.

Currently an electric snowmobile thermal management system does not run when connected to DC charger using bi-directional charging session. There was no energy involved that was used to heat battery pack in test number 3.

Bi-directional charging between charger and electric snowmobile was set up using ISO15118-20 standard. It took quite a lot of effort to get electric snowmobile and charging station to work together according to standard. Bi-directional charging was first tested using charger´s back-end system to set different charging and discharging power for charging session. Backend provided only manual control of charging session power. In figure 8 and 9 are successful charging and discharging sessions running between charger and electric snowmobile on both devices displays.

Figure 8. Successful charging and discharging sessions on charger display.

Figure 9. Successful charging and discharging sessions on electric snowmobile display.

Due to voltage safety limit, a specific script was developed and run in Latinki charger. This script made it possible to test bi-directional charging for longer period automated and without back-end control. At the start script sets mode to charge. Then it starts to follow electric snowmobile battery pack voltage and sets mode to discharge when voltage reaches 118 VDC. When voltage drops down to 105 VDC, mode is set back to charge. During the testing a short pause before mode change and other minor tweaks were added to the script. Script continues until the session is stopped.

After communication and bi-directional session were established and tested successfully, actual environmental testing was started. Temperature cycle used in test number 3 is shown in Figure 10. The test started with an overnight period of 17 hours in room temperature -5 °C. After some validation that everything was working, testing was continued for 48 hours in -20 °C, then there was approximately 12 hours pause. The environmental room was shutdown to let the room temperature rise naturally. Finally, the room temperature was set to +30 °C for 21 hours. The test was finished 21st October at 09:00.

Figure 10. Room temperature and setpoint during test number 3.

Figure 11 shows single charge and discharge cycle during test number 3 in room temperature -20 °C. Start of charge and discharge modes are marked in the diagram and calculated efficiency for charging session and discharging session is shown in diagram using second y-axis. Efficiency in charging mode was approximately 86 % and in discharging mode a little bit higher, approximately 86-87 %. Total energy charged in this cycle was 4129,03 Wh and discharged energy 2 674,02 Wh.

Figure 11. Single charge and discharge cycle 18th October 21:00 – 19th October 00:10 in room temperature -20 °C.

Figure 12 shows single charge and discharge cycle during test number 3 in room temperature +30 °C. Start of charge and discharge modes are marked in the diagram and calculated efficiency for charging session and discharging session is drawn using second y-axis. Efficiency in charging mode was approximately 85 % and in discharging mode a little bit higher, approximately 86-88 %. Total charged energy in this cycle was 6491,25 Wh and discharged energy 4712,69 Wh.

Figure 12. Single charge and discharge cycle 18th October 21:00 – 19th October 00:10 in room temperature +30 °C.

Discharge capability from electric snowmobile battery pack to grid rose approximately 76% when room temperature was +30 °C compared to -20 °C. Charging and discharging efficiencies were close to each other in both temperatures. To give accurate results about overall efficiencies, more testing would be required.

During test 3 a total of 34 charge and discharge cycles were run. Total energy discharged to grid during these cycles was 114 kWh and average efficiency of discharging was 86.7 %. The total energy charged from grid was 165 kWh and efficiency of charging 85.5 %.

From figure 13 it can be seen that electric snowmobile battery pack average temperature stayed around +12 °C when room temperature was -5 °C, approximately +5 °C when room temperature was -20 °C and rose to +35 °C when room temperature was +30 °C. Battery pack temperature in room temperature +30 °C did not have enough time to stabilize, so it might still rise over more time.

Figure 13. Battery pack temperature during test 3.

Summary

These testing sessions were not carried out following any official standards but nevertheless testing indicated valuable information about how electric snowmobile and its peripheral systems operated in extreme conditions.

Test number 1 indicated that Aurora´s electric snowmobile worked and started in cold conditions and its peripheral systems were operable even when it had been outside for a few days with no attention to it.

The electric snowmobile thermal management system was tested in test number 2 and test showed that when electric snowmobile was plugged in, it needed somewhat of energy to keep battery pack warm in cold temperatures. This could be improved in the future using predictive heating logic and the development of battery pack insulation.

Most interesting test was testing number 3 where bi-directional charging was tested. It took quite a lot of effort and time to get the DC charger and electric snowmobile to talk to each other and get everything to run smoothly. Bi-directional charging tests indicated that bi-directional charging worked in temperature range -20 °C to +30 °C with tested devices.

In the future it is possible to use the electric snowmobile as energy storage and charge and discharge it with custom logic from the back-end system or building automation system. During the test session, there were very few DC chargers available on the market that could make this possible and with tested charger there were some minor software related issues that must be considered and taken care of before electric snowmobiles can be easily and safely used as energy storage.

Sources

Energy ECS, 2025. About the project. Referenced 8.5.2025. https://energyecs.eu/project/

InterControl, 2024. Latinki 11. Referenced 25.9.2024. https://www.intercontrol.fi/en/products/dc-chargers/latinki-11/

---

Final Demonstration of Bidirectional Charging in a Microgrid

Tuomas Sinisalo, Jukka Leinonen and Jyri Öörni

Introduction

The Energy ECS project developed several components and systems for microgrid control and bidirectional charging of electric vehicles. These were demonstrated at the Sunday Morning Resort in Pyhätunturi in the Finnish Lapland. At the resort, the electrical load included several accommodation buildings, a restaurant kitchen, a laundry, a geothermal heat pump and several electric vehicle charging stations for cars and electric snowmobiles (eSleds). In addition, the resort had a 30 kWp solar power plant. These made the resort a microgrid and an excellent demonstration site for the components and systems developed in the project.

Energy ECS use case 3 developed a microgrid controller to be connected to electric loads, renewable energy production, battery storage and electric vehicles with bidirectional charging capability. To control the microgrid, cloud software was also developed to optimise the use of electricity and also to predict near future production and consumption.

For UC4, which Aurora Powertrains was the lead of, a smart Vehicle-to-Grid (V2G) charging system was developed. Aurora Powertrains developed a Vehicle Charge Control Unit (VCCU) that handles the control of charging on the electric vehicle (EV) side. Lapland University of Applied Sciences installed 25 energy meters at the demonstration site and implemented energy flow visualisation software to measure, record and analyse electricity consumption under different electricity load conditions.

Installing and testing microgrid systems

Merus Power, Aurora Powertrains and Lapland UAS together carried out a demonstration on the microgrid systems at the Pyhätunturi demonstration site. Merus Power installed the microgrid controller it developed, and Aurora Powertrains connected the eSled with the VCCU (Figure 1) to the microgrid, enabling bidirectional charging.

Figure 1. Vehicle Charge Control Unit.

The microgrid controller controls the stationary energy storage system (ESS) located in the same cabinets, but also the ISO15118-20 V2G capable eSled. An InterControl Latinki 11 DC charger with a maximum output of 11 kW was used as the bidirectional charger. Lapland UAS tested the implemented energy flow visualization software. Figure 2 shows a diagram of the final bidirectional demonstrator components and system.

Figure 2. Final bidirectional demonstrator components and system.

Due to ongoing construction works at the demonstration site, the microgrid controller could not be installed in the building where the main electrical panel was located, but in the adjacent maintenance building. This meant that the current transformers designed to be installed in the main electrical panel could not be utilised for real-time power and power quality measurement and their control purposes.

The real-time power measurements had to be replaced by an energy meter installed by Lapland UAS for the energy flow visualisation software, which measured the power of the main grid connection point of the site to the electricity distribution network. The frequency of this energy meter was about once every 5 seconds so the response time of the microgrid controller had to be slowed down to achieve stable operation. This causes power fluctuations in peak shaving operation as the load on the demonstration site varies.

This demonstration event took 3 days in total and during that time all the subsystems were connected, tested and finally demonstrated. Merus Power first tested their UC3 system by adjusting the power output of the stationary ESS according to simulated microgrid frequency fluctuations (Figure 3).

Figure 3. Preparing the microgrid controller.

Later Aurora Powertrains arrived and built up the UC4 system using the eSled and a 3rd party 15118-20 standard supporting 11kW DC Wallbox (Figure 4). Both systems were first operated in isolation from each other.

Figure 4. eSled and the V2G wallbox.

After successful tests with both charging and discharging, the UC4 system was ready for the next step. Displays from both the wallbox and the eSled presented the expected corresponding discharging power measurements which were 2.4kW in this case (Figure 5). After the charging seemed to work great, the UC3 and UC4 systems were ready to be connected together.

Figure 5. Wallbox and eSled discharging displays side by side.

Final demonstrator

First, both V2G and microgrid controller systems were connected to the same network through ethernet cables. UC4 wallbox and eSled combo were set up to a mode where the eSled charging would be controlled by the wallbox and the wallbox charging on the other hand would be controlled by commands sent by the UC3 microgrid controller.

After all the needed connections were made and the communication between each endpoint was working as intended, the microgrid controller was ready to demonstrate frequency regulation together with the stationary ESS and the eSled battery (Figure 6). The frequency regulation demonstration went as planned.

Figure 6. Microgrid controller dashboard.

Finally, a peak shaving test of the microgrid was performed. The purpose of the peak shaving test was to limit the main grid power draw from the distribution grid to the setpoint. If the power consumption of the demonstration site was higher than the setpoint, the microgrid controller would draw the required power from the stationary ESS and V2G capable electric vehicle.

Figure 7 shows the peak shaving operation period where the combination of ESS and V2G power is controlled so that the main grid power remains at the set point of 45 kW. The figure shows in orange the power drawn from the distribution network, in blue the power drawn from the stationary ESS and in green the power drawn from the eSled. The average power drawn from the main grid was 45.03 kW. Despite the low measurement frequency and the relatively high communication delay and the resulting fluctuations in main grid power, the average power remained extremely well within the set point.

Figure 7. Demonstration of microgrid peak shaving with ESS and V2G.

Tracking of all the energy flows (both real-time and historical data) between different endpoints was made easy by the energy meters installed and the energy flow visualisation software (Figure 8) developed by Lapland UAS.

Figure 8. Energy flow visualisation software.

Summary

The microgrid at the Pyhätunturi demonstration site was controlled with the microgrid controller and the stationary ESS and eSled battery connected to it. Both frequency regulation and peak shaving demonstrations were successfully carried out on the microgrid. Due to ongoing construction works at the demonstration site, the microgrid controller could not be installed in the planned building and therefore the current transformers could not be installed in the main electrical panel. For this reason, it was not possible to demonstrate the power quality improvements of the microgrid.

The demonstration was also a success showcasing possible future usage of EVs as a tool for the stabilization of a microgrid and other new economic opportunities. Aurora Powertrains will be using the VCCU as a default charge controller in its products.

Energy flow visualisation software was used to visualise the electricity consumption at different locations in the microgrid during the demonstrations. The software can be used in the future to monitor and optimise electricity consumption in the microgrid.

---

Implementing a Nanogrid Testing Environment

Jaakko Etto and Jussi Suopajärvi

Introduction

The nanogrid testing environment was built as part of the Energy ECS project related to use case three. Use case three investigated smart grids and their interaction with electric mobility. The increase in electric mobility and the timing of electricity use based on energy prices cause challenges to the operation of the public electricity distribution network and also to the formation of electricity prices in the electricity market.

In the project, the test site was in Pyhätunturi, where the electrical effects of electric cars, electric snowmobiles, and solar power production on the property’s electrical network and the grid connection could be well investigated. The Pyhätunturi test site is in continuous use, which limits the performance and timing of functional tests. For this reason, a nanogrid corresponding to a microgrid in terms of functional features was implemented as a testing environment in the building electrification and automation laboratory of Lapland University of Applied Sciences’ Kemi unit. In this nanogrid environment, functional and electrical tests can be performed in a house-like environment without limiting factors. The nanogrid was designed to have a similar load to the project’s actual test site in Pyhätunturi, but on a smaller scale.

Objectives of the Testing Environment

The goal of the testing environment implemented at Lapland University of Applied Sciences is to provide versatile opportunities to test the features of the nanogrid both as an independent, islanded power network and when connected to the public electricity distribution grid. This objective influenced the selection and sizing of devices and components used in the nanogrid’s implementation. The connection and interfacing methods of the electrical and automation devices were designed to accommodate the various operating modes of the nanogrid and to enable the integration of different measurement devices for testing purposes in different parts of the nanogrid. The essential components and functions of the nanogrid and microgrid are illustrated in Figure 1.

The implementation of the test network aimed to enable the following measurements and tests in various operating modes of the nanogrid:

• Power quality measurements
• Coupling and change state phenomena
• Operation of the hybrid inverter
• Performance and characteristics of the battery system
• Features and operation of the solar panel simulator
• Functionality of fixed measurements (connection point measurements and smart meter) and data transmission
• Operation of overload protection
• Functioning of short-circuit protection both when grid-connected and in island mode
• Operation of surge protection
• Performance of load control automation in the electrical distribution board
• Connection, operating characteristics, and effects of an external generator
• Data transmission between different devices
• Features and functionality of the home automation system and user interface
• Connections to data networks, such as electricity price information.

Figure 1. General Structural Diagram of Microgrid and Nanogrid (Kukkola M. and IEEE)

Components and Implementation of the Nanogrid

The implementation of the nanogrid was based on the typical electricity consumption and production of a small house and the selection of a suitable hybrid inverter. The hybrid inverter includes a backup power function, allowing the testing environment to form an independent island with its own energy production and battery system, which is a prerequisite for the definition of a nanogrid or microgrid (Figure 1). Remote-controlled relays operating on WLAN and Zigbee networks were used for load control. The Home Assistant software (Figure 2), a third-party program capable of controlling devices from various manufacturers, was used as the control software.

Figure 2. User interface of home assistant program

The system’s functionality was tested during the initial implementation phase using a portable installation frame located in a separate, enclosed test room (Figure 3). At this stage, the focus was on evaluating and testing the communication between different devices within the system, the suitability of the home automation setup, and the network connectivity.

Figure 3. a) Manufacturer’s schematic diagram of the hybrid inverter assembly (Fronius 2023), b) Initial phase test installation in spring 2023. (Etto J., … 2023)

Due to the requirements of the testing environment, actual solar panels were not connected to the nanogrid for energy production. Instead, a DC power supply was procured to simulate solar panel output, with adjustable power ranging from 0 to 6 kW. The DC power supply is controlled by the SAS software (Solar Array Simulator), which allows for the simulation of solar energy production under various panel configurations, geographic locations, seasons, and weather conditions (Figure 4). The software and DC power supply enable consistent and repeatable solar energy production simulations for each test, ensuring identical conditions when required.

Figure 4. a) SAS (Solar Array Simulator) software interface, b) ITECH DC power source

The testing environment was implemented to resemble a traditional single-family house as closely as possible. Typical consumption devices were chosen as loads, such as an air source heat pump, an air-to-water heat pump, a water heater, and various household electrical appliances and lights. Additionally, single-phase and three-phase charging devices for electric vehicles are used as loads and in measurements.

The hybrid inverter with a backup power function has a power of 8 kW, which is just under half of the testing environment’s maximum power, which is about 17 kW with a three-phase 25 A (connection) supply. This means that when operating in island mode without a generator, the maximum load of the testing environment must be limited to 8 kW.

Key components of the testing environment (Figure 5):

• Hybrid inverter 8 kW Fronius Symo Gen24 plus
• Smart meter (Fronius)
• Fronius interface
• Battery system BYD 10.2 kWh
• DC power source Itech IT6000B and SAS simulation software
• Home automation system and interface
• Workstation for device installations and measuring instruments
• Measurement center
• Network connection isolation equipment
• Distribution board
• Fixed wired controllable loads
• Flexible connection possibility for additional loads
• Electric vehicle charging points on the outer wall
• Connection for a backup generator on the outer wall.

The testing environment can be used for research and testing with different equipment configurations, automation implementations, and loads.

Figure 5. Nanogrid testing environment in the building electrification and automation laboratory of Lapland University of Applied Sciences in Kemi. (Suopajärvi Jussi, 2024)

Implementing the research and testing environment in an educational institution requires more detailed safety considerations than the implementation of a nanogrid in a single-family house. The educational environment requires compliance with the supplementary requirements of the standard SFS 6000-Low Voltage Installations, specifically SFS 6000-8-803:2022 Supplementary Requirements, Electrical Repair Shops, and Electrical Engineering Teaching Facilities.

When designing a testing and research environment for an educational institution, the most important consideration is safety. Educational institutions and electrical repair shops are subject to their own annex in the standard (SFS 6000), specifically SFS 6000-8-803, which defines supplementary requirements. These supplementary requirements include, among other things, that all supplies must be protected by residual current devices and that equipment connections must be implemented with touch-protected connectors. Additionally, as a safety measure, the PE conductor of the equipment was doubled, ensuring the continuity of the protective earth between the main components of the equipment with fixed protective earth conductors in addition to banana plugs. (Suopajärvi Jussi, 2024)

Operation of the Nanogrid

In normal operation, the nanogrid is connected to the public electricity grid, and the system simulating solar panels produces power depending on the time of day or the test program. The battery system can be charged or discharged depending on the price of purchased electricity and other set parameters guiding the operation. Electricity usage can be controlled, limited, and timed according to the price of electricity and the characteristics of the usage points.

In test use or during a public electricity grid outage, the nanogrid operates in island mode. In this mode, electricity is produced as needed by a backup generator and energy storage systems (battery system and, in the future, the battery of an electric vehicle). In island mode, it is important to limit the load, and part of the load limitation is fully automatic, while part of the load can be controlled.

In the testing environment, the island mode is activated automatically when the public electricity distribution network becomes de-energized. The activation of the island mode takes about 30–60 seconds. During this activation time, the hybrid inverter monitors the public electricity distribution network to see if the voltage returns to normal. If this does not happen, the hybrid inverter disconnects the testing environment to form an independent island and starts supplying power to the island. This is because a nano- or microgrid must not supply voltage to the public electricity grid when it is de-energized. When the public electricity distribution network returns to operation, the hybrid inverter monitors the network’s operation and stability for about 30 seconds before reconnecting the nanogrid to the public electricity distribution network.

The battery system in the testing environment is a 10.2 kWh property battery. This size was chosen because, with this hybrid inverter and battery configuration, the duration of a full load test is just over an hour. Additionally, this configuration could also correspond well to the size of a system installed in a typical single-family house. The testing environment is implemented so that it can be used in full configuration, but also in such a way that, for example, the backup generator, electric vehicle charging points, or battery system are not used in all usage and testing situations.

Loads and Load Protection of the Nanogrid

The testing environment includes a distribution board typical of a single-family house (Figure 6), where some of the loads are controlled by home automation, and some loads are also measured with separate measurements.

Figure 6. Distribution center of the small house in the testing environment, with the measurement center on the left and the distribution board on the right.

The loads in the testing environment include an air source heat pump, an air-to-water heat pump, electric heaters, household electronics, lighting, refrigeration appliances, and kitchen devices and fixtures. The electrical installations and cabling are carried out according to standards. Figure 7 shows some heating loads, and Figure 8 shows the connection possibilities for loads and measurements.

Figure 7. An air-to-water heat pump, an air source heat pump and heat recovery unit.

Figure 8. Load connection options in the electrical panel.

Summary

In addition to the measurement center and distribution board of a single-family house, the testing environment includes a PV production simulator, a property battery, a home automation system, and typical single-family house loads. The system was implemented according to the plans, and its operation was verified and put into research use during commissioning. The implementation of the system has proven to be a flexible and adaptable research environment that meets the objectives.

References

Suopajärvi Jussi, Älykäs nanoverkko ja sen toteutus oppilaitosverkkoon (in Finnish), diplomityö, Lappeenrannan tekninen yliopisto, 2024, https://lutpub.lut.fi/handle/10024/167395

Juustovaara Jouko, Nanoverkon ohjausjärjestelmä ja älykäs kuormanhallinta (in Finnish), diplomityö, Lappeenrannan tekninen yliopisto, 2024, https://lutpub.lut.fi/handle/10024/167385

Etto Jaakko, Suopajärvi Jussi, Juustovaara Jouko, Leinonen Jukka, Asumisen nanogrid-microgrid ympäristöt, Automaatiopäivät-Automation Days, 2023, Finnish Society of Automation (FSA), ISBN 13 978-952-5183-62-7. (in Finnish) https://www.automaatioseura.fi/site/assets/files/3870/asumisen_microgrid_-_nanogrid_testiymparistot.pdf

”IEEE Standard for the Specification of Microgrid Controllers”, IEEE Std 2030.7-2017, ss. 1– 43, huhti 2018, doi: 10.1109/IEEESTD.2018.8340204.

M. Kukkola, ”Mikroverkon suojaus – yleiset periaatteet ja toteutus opetusympäristöissä”, Sähkötekniikan diplomityö, vsk. 2022, s. 121, 2022. https://lutpub.lut.fi/handle/10024/164320

Fronius https://www.fronius.com/fi-fi/finland/aurinkoenergiaa/asentajat-ja-kumppanit/tuotteet-ja-ratkaisut/omakotitalon-energiaratkaisut/gen24-plus-invertteri-joustavalla-varavoimalla

---

Nanogrid as Part of Home Automation

Jouko Juustovaara and Jussi Suopajärvi

Introduction

The control system for the nanogrid has been implemented as part of the Energy ECS project, where the use case 3 focuses on a microgrid. As part of the microgrid research, a separate nanogrid environment was established at Lapland University of Applied Sciences. This nanogrid environment includes a control system for the nanogrid implemented through a home automation system, which features energy management solutions that enable load management, measurements, and different operational modes of the equipment.

A nanogrid refers to a smaller system than a microgrid. The general structure and operating principles are mostly the same as those of a microgrid. A microgrid can consist of multiple nanogrids, as shown in Figure 1. The definition of a nanogrid states that it includes its own electricity production.

Figure 1. Multiple nanogrids form a microgrid.

In Finland, fluctuations in electricity prices have recently been greater than before. This is primarily due to the strong increase in renewable energy sources, such as wind power. In the Nordic countries, the hourly price of electricity is determined by supply and demand in the Nordpool electricity market, which uses a ”closed auction” trading model (Figure 2). This trading model provides the spot electricity price for each hour of the day. The variation in electricity prices throughout the day allows consumers to take advantage of the cheapest hours for their electricity consumption.

Figure 2. The impact of supply and demand on electricity prices.

Home automation system

A nanogrid can be part of a home automation system, and its energy consumption control and load management can be implemented through home automation system. In energy consumption control, hourly electricity prices can be utilized to adjust usage accordingly. Additionally, if the nanogrid generates surplus electricity, it can be sold to the market at the most profitable times.

There are several options for implementing nanogrids and home automation, depending on the hardware and protocols used. Figure 3 illustrates an example of the control protocol for a home automation system. The home automation system receives electricity price information via a web application, which enables the management of electricity distribution and storage.

Figure 3. Example of protocols used for home automation systems.

Figure 3 illustrates an example of the control protocol for a home automation system, where electricity distribution and storage represent the nanogrid system.

Examples of energy management in home automation systems

Figure 4 presents the hourly electricity prices for the day. This figure shows the price fluctuations at different times of the day, which helps consumers and home automation systems optimize their electricity usage.

Figure 4. Electricity hourly price 2.9.2024

One effective example of home automation control is the management of water heaters. The home automation system intelligently activates the water heater during the most cost-effective hours of the day, optimizing the heating of household’s domestic water. Figure 5 illustrates how the temperature of the water heater rises during these hours. This optimization reduces electricity costs by taking advantage of lower electricity prices while ensuring that the demand for hot water is met efficiently.

Figure 5. Example case of water boiler temperature

The production system of nanogrid can consist of, for example, wind or solar power, or a combination of both. Figure 6 presents an example case showing the hourly electricity production from the solar power system and the electricity consumption of the household. In such a system, electricity usage can be optimized so that the self-generated power covers consumption as efficiently as possible.

Figure 6. An example of solar energy production and the electricity consumption of a detached house.

In Figure 6, the electricity production of the solar power system is shown on a green background, while the electricity consumption of the household is represented on a yellow background. As solar electricity production increases, the household’s electricity consumption also rises when the water heater is activated for heating (Figure7). This optimizes electricity usage so that the water heater takes advantage of the peaks in self-generated solar power, reducing the need to purchase electricity from the grid and improving the overall energy efficiency of the system. Thus, the household can maximize its use of renewable energy while also saving costs.

Figure 7. Example case of water heater temperatures.

In Figure 7, solar energy production exceeded the household’s electricity needs around 09:00. The temperature of the water heater rose until approximately 14:00, after which surplus solar energy production was sold back to the grid. This enables both the optimization of self-energy usage and the financial benefit from selling surplus energy.

Summary

In Finland, recent fluctuations in electricity prices, driven by increasing renewable energy sources, allow consumers to optimize their energy usage based on hourly price variations. A nanogrid can leverage these price changes through a home automation system, enabling efficient control of energy consumption and the potential sale of surplus electricity.

The implementation options for nanogrids and home automation systems vary depending on the hardware and protocols used. The control system can manage various devices, such as water heaters. Integrating renewable energy sources, such as solar energy, with smart energy management enables the use of self-generated electricity and the sale of excess electricity to the grid, improving both energy efficiency and cost savings.

References

Suopajärvi Jussi, Älykäs nanoverkko ja sen toteutus oppilaitosverkkoon (in Finnish), diplomityö, Lappeenrannan tekninen yliopisto, 2024, https://lutpub.lut.fi/handle/10024/167395

Juustovaara Jouko, Nanoverkon ohjausjärjestelmä ja älykäs kuormanhallinta (in Finnish), diplomityö, Lappeenrannan tekninen yliopisto, 2024, https://lutpub.lut.fi/handle/10024/167385

”IEEE Standard for the Specification of Microgrid Controllers”, IEEE Std 2030.7-2017, ss. 1– 43, huhti 2018, doi: 10.1109/IEEESTD.2018.8340204.

M. Kukkola, ”Mikroverkon suojaus – yleiset periaatteet ja toteutus opetusympäristöissä”, Sähkötekniikan diplomityö, vsk. 2022, s. 121, 2022. https://lutpub.lut.fi/handle/10024/164320

D. Burmester, R. Rayudu, W. Seah, ja D. Akinyele, ”A review of nanogrid topologies and technologies”, Renewable and Sustainable Energy Reviews, vsk. 67, nro C, ss. 760–775, 2017.

---

Tests and Measurements at Nanogrid Test Environment

Jaakko Etto and Jussi Suopajärvi

Introduction

As part of the Energy ECS project, a nanogrid research environment was implemented in Kemi, in the building electrification and automation learning environment of Lapland University of Applied Sciences’ electrical and automation engineering education, to complement the Pyhätunturi test site. The small house-scale nanogrid implemented in the laboratory environment allows for free testing of the grid’s features without adverse effects on electricity users. The testing equipment supplements the results obtained from the Pyhätunturi test site.

The testing environment allows for versatile research and testing of the nanogrid’s operation, usage, controls, changes in operating modes, electrical characteristics, power quality, loads, energy storage, and electricity production, as well as the clarity and operational features of various automation and interface implementations.

Testing system of the nanogrid

The nanogrid in the laboratory of Lapland University of Applied Sciences can be configured to operate with different equipment setups and loads in addition to the full configuration, according to testing objectives. The full configuration of the nanogrid used for research is as follows:

• Hybrid inverter 8 kW Fronius Symo Gen24 plus
• Smart meter (Fronius)
• Fronius interface, HMI
• Battery system BYD 10.2 kWh for energy storage
• Simulation of solar power production: Itech IT6000B DC power source and SAS (Solar array simulation) -software
• Home automation system and it’s interface (Home Assistant)
• Electrical connection to the distribution network, measurement center, network connection isolation equipment, distribution board, and safety switches
• Electric vehicle charging point and backup generator connection point in the yard area
• Electricity usage: various electrical loads typical of a small house, such as an air source heat pump, air-to-water heat pump, water heater, electric heaters, household electronics, kitchen electrical appliances, indoor and outdoor lighting, etc.
• Workbench, interfaces, displays, and ready measurement connections for various measuring devices and analyzers.
•  

Figure 1. Main devices of the nanogrid testing environment. The electric vehicle charging and backup generator connection point is outside the window in the yard and loads in the laboratory premises.

The objective of the nanogrid research environment (Figure 1) is to conduct versatile operational tests, investigate and measure the electrical characteristics of the nanogrid in various usage and operating modes. Testing and research can be done with different equipment configurations and selected electrical loads. Solar power production can be simulated as needed using a DC power source and SAS simulation software. Research, measurements, and tests can focus on various features and functions of the nanogrid in the testing equipment:

• Measurements and analyses of power quality in different parts and operating situations of the nanogrid: connection point to the distribution network, electrical center, various components of the network, loads, and protections
• Phenomena during switching and transition states: different connection situations of loads, production, and energy storage, their functions, changes, and fault situations
• Control, operation, and characteristics of the main components of the nanogrid: hybrid inverter, battery system, solar power production, electric vehicle charging, backup generator
• Operation of fixed electricity meters and data transmission of measurements: network operator’s meter, smart meter, load measurements in the center
• Operation of electrical protections when connected to the grid and in island mode: overload protection, short-circuit protection, residual current protection, overvoltage protection
• Control of loads and production: hybrid inverter, battery system, PV simulator, electric vehicle charging, backup generator
• Features and operation of the home automation system and interface
• Data transmission between different devices, connections to the data network, e.g., electricity price information, remote control, and monitoring
• Compliance with standards, protection, and features.

Start-up Testing

After planning and procurement, the testing of the nanogrid’s operation began with the acceptance inspections of the devices. The devices were installed according to the plans (Figure 2) and carried out visual and measurement-based commissioning inspections required by electrical safety regulations. The parameters required for connection and operation were set for the devices, the electrical and data transmission connections between the devices were checked, the functionality of the research equipment was tested, and trial runs were conducted. The implementation, connections, operation, and interface of the home automation system were tested.

Figure 2. 3D model from the preliminary design phase of the equipment (Etto J, Suopajärvi J., …)

The testing of the equipment’s operation for research use has included the following tests and inspections:

• Acceptance inspections of devices and components
• Commissioning inspections of devices (visual, measurements, and functional)
• Inspection and testing of cabling and connections according to the plans, and inspection and testing of device parameter selections
• Inspection and testing of electrical protections
• Inspection of the functionality of I/O connections and data transmission buses between devices
• Functionality of interfaces (HMI)
• Commissioning, test productions, and functionality inspection of the solar power production simulation device

Testing of the Operation

In research use, the operation of the equipment and the nanogrid is tested in various electricity production, purchase, sale, load, and storage situations. The tests focused on verifying the operation of the equipment according to specifications in different operating situations and transitions between operating modes. Examples of operating modes are shown in Figure 3, where the system’s solar panel production, electricity usage, and battery charge status can be seen, as well as whether the system is connected to the public electricity grid or operating in island mode.

Figure 3. System operating modes a) PV production covers consumption and the surplus are stored in the battery b) Island mode, where controlled consumption is covered by PV production and the battery.

The operation of the system in different operating modes and the transitions between operating modes were tested. Such verifiable functions included, for example:

• Operation of the hybrid inverter during a power outage of the supplying electricity grid (island mode)
• Operation of battery charging and discharging according to the set mode
• Backup generator usage situations and operation in island mode
• Protections
• Control of loads and the battery (manual, home automation, and hybrid inverter)
• Tariff-based operational controls (home automation).

Figure 4. PV production simulation with SAS (Solar Array Simulator) software and DC power source, showing the programmed operation and operating values in the interface.

The solar power production and the operation of the SAS simulation software were tested with different solar panel configurations, panel location data, and operation time scaling. An example of the simulation program interface is shown in Figure 4, where the operating values at a specific time during the test period and the characteristics of the selected PV operation can be seen.

Power Quality Measurements

The standard SFS-EN 50160:2022 defines the power quality at the customer’s connection point under normal operating conditions. The standard specifies the limits within which the voltage must remain. The measurement period is one week. The standard provides limits for total harmonic distortion, individual harmonic overtones, flicker, and voltage level variations (SFS-EN 50160:2022).

The compliance with the voltage quality specified in the standard can be assessed through measurements. For power quality measurements in different points and operating modes of the nanogrid system, a Fluke 438 power quality analyzer was used for instantaneous and long-term measurements. A three-phase power quality data logger, Fluke 1748, which meets international power quality standards such as IEC 61000-4-30 and records numerous measurement data during the recording period, was also used in the tests. Figure 5 shows the measurement connection of the device, and the load of the distribution board in the test environment is measured from the measurement point. The next Figure 6 shows an example of a measurement result when the public electricity grid supply was interrupted and switched to island mode. Figure 7 shows an excerpt from the results of a long-term measurement.

Figure 5. Nanogrid testing environment, power quality measurement, with the measurement point being the supply of the distribution board (Suopajärvi J., 2024).

Figure 6. Nanogrid testing environment, frequency, voltage, and current measurements in island mode (Suopajärvi J., 2024).

Figure 7. The example of long-term measurement results, phases and neutral current waveforms.

In the research environment, measurements of voltage and current harmonics, instantaneous values, and waveforms were conducted. Figure 8 shows the effective values of current and voltage and the values of harmonics of the waveform when the load is mainly heating.

Figure 8. Measurement results at the connection point of the test network with heating loads a) phase currents, their indicators, and neutral current, b) voltages, c) harmonics of phase currents, d) harmonics of phase voltages, e) currents with heating load, and f) currents without heating load.

Figure 9 shows the characteristics of voltage and current with an asymmetric load. With an asymmetric load, a voltage difference between phases and a high neutral current can be observed, as well as the impact of power electronics and lighting load on the current waveform.

Figure 9. Measurement results of voltage and current at the connection point of the test network with an asymmetric load (phases L1, L2, L3, and N).

When the public grid is not available, the system automatically switches to island mode through a power outage. In island mode, the total load is limited as needed. Electricity production is based on the battery and solar power produced by the PV simulator, modelling real situations. Both power sources are DC and connected to the hybrid inverter, which converts DC to three-phase AC. If necessary, a three-phase generator can be connected to the system as an additional production method to supply power to the load and charge the battery. Figure 10 shows operational displays and some measurements during island mode.

Figure 10. Island mode, measurements from the distribution board: a) Operation in island mode, b) public grid re-energized and inverter switching mode, c) phase L2 with a small load in island mode, d) phase L3 with heating and lighting, e) harmonic currents of phase L2, and f) harmonic currents of phase L3.

Summary

Based on the use of the equipment and the tests performed, it has been observed that the equipment performs its task well and operates according to the settings. The automation of the nanogrid has been tested in identical energy production conditions, and changes made to it have been easy and quick to detect. Additionally, the quick and easy changes in loads have been significantly beneficial for the tests. More detailed results and analyses of the operation and power quality measurements will be presented later in other publications.

References

Etto Jaakko, Suopajärvi Jussi, Juustovaara Jouko, Leinonen Jukka, Asumisen nanogrid-microgrid ympäristöt, Automaatiopäivät-Automation Days, 2023, Finnish Society of Automation (FSA), ISBN 13 978-952-5183-62-7. (in Finnish) https://www.automaatioseura.fi/site/assets/files/3870/asumisen_microgrid_-_nanogrid_testiymparistot.pdf

Suopajärvi Jussi, Älykäs nanoverkko ja sen toteutus oppilaitosverkkoon (in Finnish), diplomityö, Lappeenrannan tekninen yliopisto, 2024, https://lutpub.lut.fi/handle/10024/167395

SFS-EN 50160:2022 Voltage characteristics of electricity supplied by public electricity networks. 30.12.2022.

---

Power Quality Measurements at the Demonstration Site

Kari Kenttä

Introduction

The power quality measurements in the microgrid were part of Use Case 3 in the Energy ECS project. The demonstration site was the Sunday Morning Resort, a boutique hotel located in Pyhätunturi. This site was selected for the demonstration because, in addition to conventional electricity consumption, it featured charging stations for electric cars and snowmobiles as well as a solar power plant. The site also included electric heating (a geothermal heat pump), an on-site laundry facility, a restaurant kitchen, outdoor and area lighting, and several accommodation buildings equipped with saunas.

One of the interesting aspects of charging the electric snowmobiles was that it was conducted using single-phase charging. The goal was to determine whether single-phase charging, which occurs randomly on different phases either simultaneously or at different times, could lead to overloading of the neutral conductor or problems caused by the summation of harmonic distortions on the neutral. Additionally, the study aimed to assess whether vehicle charging caused any power quality issues and if the capacity of the electrical connection was sufficient for the charging needs.

Electrical connection of the site

The site’s electrical connection was implemented using two AXMK 4 x 185 S cables. The phase conductors and the combined protective earth and neutral conductor (PEN) of the connection had the same cross-sectional area. The main fuses at the site were rated at 315 A, resulting in a calculated connection power of 218 kVA. The connection cable runs from a pole-mounted transformer located approximately 100 meters from the site. The transformer is rated at 315 kVA (Figure 1). No other connection cables run from the transformer, making the hotel the only load on the transformer.

Figure 1. A 315 kVA pole-mounted transformer is connected to the overhead power line network.

Power quality data collection

The measurement point was installed on the building’s rising cable, located just before the main distribution board (Figure 2). The progress of the measurements was monitored remotely, ensuring the correct operation of the measurement device throughout the extended measurement period.

Figure 2. The main distribution board at the measurement site supplies several downstream sub-distribution boards. The measurement device, shown in the lower left of the image, is connected to record data.

Figure 3. The data collection device was connected to the connection cable of the test site, positioned before the main switch.

The power quality measurements were conducted using a three-phase power quality data logger, the Fluke 1748. This device complies with international power quality standards such as IEC 61000-4-30 and is capable of recording multiple parameters simultaneously during the same event. The logging function is secured by a built-in battery, ensuring data retention even during power outages. The reporting standard used was EN50160:2010+A2, 3, which was the latest standard supported by the analysis software (Fluke Energy Analyze Plus 3.11.2.0) at the time. (Fluke, 2024)

The data logger measured voltage via direct connection to the potential of the site’s conductors. For current measurements, Fluke iFlex1500-24IP65 current transformers were used, wrapped twice around the phase conductors. This method aimed to ensure the highest possible measurement accuracy. In Figure 3, the data logger is shown installed near the main fuses of the building.

Analysis of measurement data

The measurement period commenced on 21.11.2023, and concluded on 16.4.2024, lasting a total of 146 days. During this period, the maximum instantaneous power recorded at the site was 196 kVA, while the connection capacity of the facility was 217 kVA. This indicates that the electrical connection is appropriately sized based on the peak power usage. The average power consumption during the measurement period was 71 kVA, approximately one-third of the peak power. Thus, it can be concluded that, despite the measurement period falling within winter and the peak tourist season, there remains a considerable amount of unused capacity in the connection. Implementing load balancing and intelligent load control could promote more even usage of the capacity.

Notably, the connection is not underpowered, even with the addition of electric vehicle (EV) charging points, and it is anticipated that the number of these will increase in the future. Some asymmetry in consumption between the phase conductors was occasionally observed due to single-phase charging of EVs and the usage rate of charging, but overall, there were no significant concerns. The average consumption asymmetry during the measurement period was 11.7 %, which can be considered good.

During the identified operating times of the charging devices, no voltage sags or significant harmonic distortions were observed (see Figure 4). Several sources of harmonic distortions were present at the site, as lighting, office equipment, geothermal heat pump controls, and single-phase charging stations can contribute to harmonic currents.

From Figure 4, it is evident that harmonics 3, 5, and 7 are prominent in the results, which are common problematic harmonics. The horizontal axis represents harmonic orders 0-25, while the vertical axis shows the current as a percentage of the fundamental wave.

However, the measured results are below the permissible limits; for example, the 3rd harmonic was at 4.5 %, which is below the allowable limit of 5 %. The total voltage distortion (THD) was approximately 6%, remaining under the permissible limit of 8 % (SFS-EN 50160:2022).

Significant peaks in current harmonic distortions were not observed during the operation times of the charging devices.

Figure 4. The figure illustrates the current harmonic distortions observed during the measurement period. The colors represent the following phases: black for phase L1, red for phase L2, and blue for phase L3.

During the measurement period, a total of 36 voltage sags were recorded. A closer examination revealed that these voltage sags were caused by power supply disturbances in the overhead power line serving the site. On windy days, snow-laden trees can create short circuits on the electrical line, which manifest to end users as voltage sags or brief power outages. A total of three power outages were observed during the measurement period.

The analysis of the measurement data also investigated the current flowing through the neutral conductor of the connection cable under various load conditions. The site utilized single-phase charging stations for electric vehicles (EVs), and their simultaneous use had not been specifically coordinated. If charging loads were to occur on the same phase at the same time, this could result in an unbalanced load, causing the currents on the same phase in the neutral conductor to sum together. Such usage could lead to excessive currents exceeding the neutral conductor’s capacity.

However, no significant currents in the neutral conductor were observed during the measurement period. The highest recorded neutral current was 136.1 A on 17.12.2023, during a voltage outage, after which the electrical devices restarted simultaneously, resulting in this current spike. The average current in the neutral conductor during the measurement period was 23.2 A. Based on these observations, it can be concluded that there was no significant current imbalance between the phase conductors at the site. A phase current imbalance would have been reflected in a higher average current in the neutral conductor.

An examination of the current peaks observed in the neutral conductor yielded the following findings:

• 2 peaks over 90 A
• 3 peaks over 80 A
• 12 peaks over 70 A

The peak currents for the phase conductors were recorded as follows: 305.2 A for phase L1, 296 A for phase L2, and 286.6 A for phase L3. The corresponding average phase currents were 98.2 A, 107.3 A, and 92.8 A. These measured results strongly support the conclusion that good current symmetry existed among the phases.

Summary

The charging of electric vehicles was found to have no significant impacts on the electrical connection of the site (such as voltage sags, harmonic distortions, current imbalances among phase conductors, or neutral conductor overloads). The sizing of the electrical connection was adequate but not oversized. The situation is also facilitated by the fact that the site’s connection cable is sufficiently large, and the combined neutral and protective earth conductor has the same cross-sectional area as the phase conductors. Additionally, the electrical connection is supplied by a sufficiently sized transformer with no other connections.

Sources

Fluke 1748 Three-Phase Power Quality Loggers manual. Referred to 5.9.2024 1746-b-eus-datasheet.pdf (fluke-direct.com)

SFS-EN 50160:2022 Voltage characteristics of electricity supplied by public electricity networks. 30.12.2022.

---

Autonomous Driving of EV to Charging Station

Lauri Lehti and Jukka Leinonen

Introduction

Energy ECS project use case 6, led by Unikie, was based on a real-life need. Operating a bus depot efficiently is crucial in today’s fast-paced world. Daily challenge of managing hundreds of bus movements 24/7, reducing costs related to damages, waiting times, and inefficient use of space. There are also other reasons which can cause problems like unexperienced drivers, tiredness and bad weather conditions. By automating depot driving tasks such as driving the buses to charging, cleaning, or maintenance stations, one could overcome the challenges of driver shortages and skill gaps, paving the way for a seamless transition to all-electric and autonomous bus fleets.

The objective of UC6 was to build a proof of concept (POC) for an automated bus depot. At a practical level, the aim was to automate the driving of the electric bus (EV) from the parking lot to the charging station and back. In this concept sensors, driving logic or any intelligence was not on the bus. Instead, sensors would be mounted in the infrastructure of the bus depot such as on lamp posts and walls. In this approach buses can be as they are and there is no need to do any installation on the buses, which will be an advantage when scaling the solution. Sensor data would be used to create situation awareness from the bus depot area. All the control logic also containing the object detection and path planning would be in the cloud. The concept architecture is illustrated in Figure 1.

A network diagram that includes a cloud server connected to electric buses, charging stations and external sensors. Connections operate over 4G/5G networks.

Figure 1. The high-level architecture of UC6.

Simulation phase

The implementation of the POC was divided into two phases: a simulation phase and a real-life phase. First, an automated driving solution was developed and tested in a simulation environment. After the solution was validated in the simulation environment, it was set up in the real-life environment for final validation.

The CARLA simulator was selected as the simulation environment. CARLA turned out to be a very versatile and flexible environment. It was not limited to supporting automated driving operations, but Lapland University of Applied Sciences implemented a realistic project-specific LiDAR sensor support for CARLA. It was valuable to have sensors modelled in the simulation environment, which enabled validation of the infrastructure-based automated driving solution as it is.

The basis for the CARLA simulation was the digital model of the bus depot. A realistic, GPS accurate model of Straeto’s bus depot in Reykjavik was created by Lapland UAS with help of Fixposition, and Scantinel. The model included several details, which enabled for example mounting sensors in light poles. Lapland UAS implemented ROS-based interfaces for the created digital-twin in CARLA which enabled the integration of the Unikie’s automated driving solution. Figure 2 shows how automated driving solution was visualized in simulation environment. This environment was used for the development of automated driving, and it corresponds very well to the real vehicle and sensors.

Figure 2. Visualisation of the automated driving solution in CARLA.

In addition to Straeto bus depot, the Unikie’s test center was modelled in UC6. This model included the inside and outside area of the test center and is illustrated in Figure 3. This Digital Twin is one of the key results for Unikie in the project. It was used to demonstrate the driving of two buses by automated driving system developed, which would not have been possible in real life. This model has been used intensively for different development and testing activities outside the Energy ECS project. Unikie further develops this model and continues to use it as part of the daily development activities, because it significantly speeds up development and testing cycle.

Figure 3. Digital-Twin of Unikie’s test center.

Real-life phase

After successful validation in CARLA simulation environment developed POC system was deployed to real-life environment. The original plan was to utilize Yutong bus which are used by Straeto. Eventually Yutong bus was not available and Linkker EV bus was used instead as test vehicle. Due to issues with the Yutong bus, the final demonstration was relocated to Turku, at Unikie’s test center. Figure 4 illustrates automated driving of the EV bus in outside area.

Figure 4. The Linkker bus automatically drives in the Unikie’s test center area.

Summary

Deployed POC showed that there are no technical limitations to implement systems that enable automated driving of the bus in the bus-depot area. The efficiency of the simulation environment was also proven. There was no need to adjust or relocate the LiDARs in the Unikie’s test center area once they were installed, as the simulation results had accurately determined their placement. In addition, the same driving SW was used as developed in the simulation environment. When comparing the validation results to the objectives set at the beginning of the project, one can say that all of them were reached. All in all, the partners of the UC6 can be proud of what they have achieved.

---

Creating and Enhancing Autonomous Driving Simulation Environment

Aleksi Narkilahti

Introduction

Within the Energy ECS project, one of the use cases—Autonomous Driving of EV to Charging Station—focused on enabling electric buses to navigate autonomously within a bus depot. The idea was for the buses to move from parking slots to charging stations and back during the night. Since the bus depot operates dozens of buses, equipping each vehicle with sensors would be costly. To address this, the project aimed to mount sensors in the infrastructure, such as on lamp posts and walls, reducing the need for sensors on the buses themselves.

To test, develop, and validate the system before moving to real-life tests, simulation was required. The FrostBit Software Lab team at Lapland UAS was responsible for developing the simulation, utilizing CARLA Simulator, an open-source simulator for autonomous driving research built with Unreal Engine 4, which the team has used in several previous projects.

Creation of bus depot digital twin

Lapland UAS developed a digital twin of Straeto’s bus depot located in Reykjavik, Iceland. To model the area, point clouds were obtained from project partners DTT and Svarmi. DTT used a laser scanner to capture the bus depot from the ground level, while Svarmi used a drone to scan the entire bus depot from above. These materials served as references for the modelling and map creation work. The point cloud data also included georeferenced information, which was integrated into the CARLA Simulator to ensure accurate GPS sensor readings. Figure 1 shows the Straeto bus depot digital twin from above view.

Figure 1. Digital twin of the Straeto bus depot from above view.

The Straeto digital twin map includes everything the real bus depot has, parking spots, lamp posts, buildings, charging stations and even some of the small objects such as bushes, asphalt paintings, waste pins, signs and so on. It was chosen to focus solely on the bus depot area, as external roads, buildings and other areas were deemed unnecessary for the use case. Figure 2 shows the modelled objects in more detail.

Figure 2. A more detailed view of the modelled bus depot.

The electric bus was also modelled and made controllable in the CARLA Simulator (Figure 3). It was created using photos and point clouds provided by DTT, along with additional photos and blueprints sourced from the internet.

Figure 3. Modelled electric bus.

Lapland UAS also modelled the washing station at the bus depot. Like the rest of the bus depot area, the washing station was modeled based on photos and point clouds obtained from DTT. In the simulator, the washing station can be driven through, and its doors can be controlled to be either open or closed. Figure 4 shows the washing station from a rear view.

Figure 4. Bus depot washing station from the inside.

Enhancing simulator tools and performance

By default, CARLA Simulator doesn’t include any user interface tools, and only offers functionality via their Python API (Application programming interface). Several tools with easy-to-use user interfaces were created to simplify setting up the simulation environment. For example, a tool was implemented to spawn vehicles and props anywhere on the map. More detailed settings and weather control interfaces were developed as well. CARLA’s weather system was expanded to include more realistic-looking clouds, and an accurate sun position based on the real-world location and time of day was implemented.

Additionally, a tool was developed for setting up Lidar sensors anywhere on the map. All these tools made it easier to test various simulation setups and find optimal sensor installation spots. Figure 5 shows the user interface for placing and configuring Lidar sensors.

Figure 5. User interface for placing the Lidar sensors and configuring the parameters.

In addition to these user interface tools, CARLA Simulator’s Lidar sensor performance was optimized. By default, CARLA Simulator could handle 2-3 Lidar sensors simultaneously before performance dropped significantly. However, this use case required the ability to simulate many Lidar sensors simultaneously. With the implemented optimizations, the simulator could handle up to 10+ Lidar sensors simultaneously, depending on the sensor type and its parameters. Scantinel, one of the project partner’s Lidar sensors, was also integrated into the simulator. Furthermore, Lidar sensor point cloud visualization was added, along with an optional Lidar noise model from a previous simulation project.

Lapland UAS also integrated CARLA’s ROS (Robotic Operating System) Bridge more closely to the simulator. ROS is an open-source, flexible framework commonly used for data communication between different systems and applications. This closer integration made it easier to use by automatically launching the ROS Bridge when the simulator starts. It also simplified stopping, restarting, and changing the ROS Bridge settings. Additionally, the performance of Lidar point cloud ROS message creation was optimized.

Figure 6 features a video trailer showcasing the Straeto digital twin map and simulator features.

Oho! Tämä YouTube-upotus ei näy, koska et ole hyväksynyt markkinointievästeitä. Hyväksy markkinointievästeet.

Figure 6. Video trailer showcasing the Straeto digital twin map and simulator features.

Demonstration and validation

Initially, the plan was to conduct the final testing with the electric bus at Straeto’s bus depot in Reykjavik. However, due to issues with accessing the electric bus electronics, the final demonstration was relocated to Unikie’s test center in Turku, where a different bus was used. Upon learning of this change, Unikie and other partners agreed to model and implement the Turku test center into the simulator as well. A visit to Unikie’s Turku test center took place, where the facility was laser scanned and photographed for modelling purposes. Fortunately, the area was not too complex, consisting mainly of the main hall and two outside areas. Additionally, all the tools developed were adaptable for this digital twin map with minimal modifications.

Validation of the simulation’s Lidar sensors was carried out in collaboration with Unikie, focusing on Lidar sensor’s ability to detect smaller objects, such as a person walking. Unikie provided point cloud data from real-life tests where a person walked through the test hall, and this scenario was replicated in the simulation environment. While some differences between the real-life and simulation data were observed, the results were generally similar, demonstrating that the simulation’s Lidar sensors effectively detected the walking person. This validation and overall feedback from Unikie confirmed that the simulation is a valuable tool for developing, testing, and validating autonomous driving systems before conducting real-life tests. Figure 7 shows the digital twin map of Unikie’s test center, with Lidar sensors set up and a person walking through the hall.

Figure 7. Turku test center digital twin map with Lidar sensors and walker.

Summary

Overall, the simulation development aspects of the Energy ECS project were successful despite some challenges and unexpected changes. The adaptability of the FrostBit Software Lab team at Lapland UAS and project partners enabled the creation of two digital twin maps and an effective simulation tool for the use case. The team significantly improved CARLA’s Lidar sensor performance and gained valuable insights throughout the project. These learnings have already been applied in new projects, contributing to ongoing advancements in autonomous driving research and simulation development.

---

Exploiting the Results of the Energy ECS Project

Jukka Leinonen

Lapland University of Applied Sciences participated in the Energy ECS project in three use cases and achieved the objectives of the project plan. The results and knowledge gained were already exploited in teaching and RDI projects during the Energy ECS project and will be exploited further in the future. The four most significant results achieved during the Energy ECS project and how they have been and will be exploited are described below.

In use case 3, the objective was to develop a microgrid controller and demonstrate its operation in a resort at Pyhätunturi, Finnish Lapland. The Lapland UAS aimed to measure the power quality at the demonstration site. Before measurements were carried out at the demonstration site, a nano-grid (Figure 1) was implemented to allow the power quality to be investigated without causing production interruptions at the demonstration site.

Figure 1. Implemented nanogrid. (Suopajärvi Jussi, 2024)

The nano- and microgrid expertise gained during the project was exploited with new expertise in an updated course on building automation, and a completely new course on smart grids will start in autumn 2025. Around 30 students will participate in both courses each year. In addition, in autumn 2025, an Interreg Aurora-funded project on distributed energy production will be launched at Lapland UAS, where the nanogrid will be one of the test environments. The nanogrid will also be exploited in other RDI projects and education in the future.

The objective of use case 4 was to develop equipment for electric vehicles compliant with the bidirectional charging standard and to demonstrate Vehicle-to-grid (V2G). Lapland UAS was tasked with testing electronics, components and systems in harsh environments. Figure 2 shows an eSled connected to a bidirectional DC charger in an environmental room where the temperature dropped to -35 degrees Celsius during the test.

Figure 2. Cold conditions test underway in the environment room.

The harsh environment testing methods and skills developed in this project will be exploited in education, RDI projects and commercial testing services. The knowledge gained from the cold testing has already been used in a Climatic and Reliability Testing course with 5 students.

In addition, Lapland UAS implemented software for the visualisation of energy flows in the microgrid of the Pyhätunturi demonstration site. The software can be configured to work also in other microgrids. Figure 3 shows the home page of the implemented software.

Figure 3. The home page of the energy flow visualisation software.

The software development expertise developed in this project will be exploited in education and RDI projects. Already during the project, around 75 students learned how to read the price of electricity from open data sources in their programming project using the interface implemented in the Energy ECS project.

The objective of use case 6 was to develop autonomous driving technology and demonstrate it on an electric bus by driving it autonomously from the parking slot to the charging station and back in a closed bus depot area. To test this system in a simulator before real-world tests, Lapland UAS created and enhanced an autonomous driving simulator environment for the bus depot. Figure 4 shows a digital twin of the bus depot.

Figure 4. Digital twin of the bus depot.

The knowledge accumulated in the development of the autonomous driving simulator has already been exploited in two projects, AGRARSENSE and CHARM, both funded by CHIPS JU and Business Finland. In spring 2028, a course on Digital Twins will start, which will also exploit the knowledge gained from this project. This knowledge will be exploited in other RDI projects and education in the future.

In addition, it is estimated that at least eight new RDI projects will be launched by the end of 2030, based on the results and themes of the Energy ECS project. The Energy ECS project, its results and their exploitation can therefore be considered a great success.