This paper presents a two-stage, voice-driven pipeline for hotel housekeeping that converts multilingual speech into English text and an intent label. An in-house dataset of 400 short housekeeping utterances is introduced, available in Croatian, English, and Italian, and annotated with three intent classes and a detailed maintenance taxonomy. The Whisper Large model is evaluated for both transcription in the source language and speech-to-English translation, with the translation setting further adapted using parameter-efficient fine-tuning with LoRA. Output quality is measured using standard error and overlap metrics, and the resulting texts are used for intent classification. For classification, BERT and ELECTRA models are compared using grid search and early stopping, reporting accuracy, F1 variants, top-k accuracy, and confusion matrices. Results indicate improved translation performance after LoRA adaptation and stronger intent classification performance for ELECTRA compared to BERT.

This study examines the causes and implications of the unprecedented electricity consumption observed in Croatia during an intense heatwave in July 2024. On the evening of 17 July 2024, power demand reached an all-time high of 3381 MW, significantly surpassing the average demand of around 2000 MW. More concerningly, during these peak hours, 35% of the electricity had to be imported due to insufficient domestic generation capacity. As a result, average monthly electricity prices for July and August 2024 exceeded 250 EUR/MWh in the evening hours. Looking ahead, Croatia and Southern Europe are expected to face increasingly hotter summers, pushing power systems to accommodate even higher peak loads. As the energy transition progresses toward a greater reliance on intermittent renewable energy, enhancing power grid flexibility will become essential. Flexible power generation will play a critical role in bridging gaps in renewable energy output. Solutions such as pumped hydro storage and battery systems can store excess renewable energy and release it during peak demand periods. Additionally, demand response strategies—encouraging the shift of electricity usage to times of higher wind and solar availability—offer another effective way to adapt to the intermittent nature of renewable energy sources.

This study analyzes the record electricity consumption in Croatia during the July 2024 heatwave and evaluates how the increased deployment of onshore wind and solar photovoltaics (PV) could mitigate a similar event in the future. Electricity demand and generation patterns under current (2024) and projected (2030) scenarios have been simulated using a sub-hourly power system model. The findings show that during the July 2024 heatwave, Croatia imported 35% of the electricity, with prices exceeding 400 €/MWh during peak hours. By 2030, the expanded wind and solar PV sectors (1.5 GW each) will increase the renewable share from 38.8% in July 2024 to 54.7% in July 2030. On the annual level, renewable energy generation increases from 53.8% in 2024 up to 66.9% in 2030. As result, the carbon intensity of the power sector will reduce from 223 gCO2eq/kWhel in 2024 to 197 gCO2eq/kWhel in 2030. The share of fossil fuel generation will increase slightly, from 19.7% in 2024 to 22% in 2030, but more significantly in the summer to meet the heatwave-induced electricity demand. Besides that, short-term energy storage of 2 GWh (400 MW discharge over 5 h) could effectively manage evening peak demands after solar PV ceases production.

The application of carbon dioxide (CO2) in enhanced oil recovery (EOR) projects is becoming increasingly more interesting in the petroleum industry because it can boost oil production rates while potentially reducing greenhouse gas emissions. The injected CO2 causes oil swelling and viscosity reduction, making it easier to flow through the reservoir fractures. This article studies the performance of the first EOR CO2 project in Croatia, which started operations in 2014 with a capacity of 0.41 million tons per year (Mtpa). The CO2 source is a natural gas processing plant, which released the CO2 into the atmosphere prior to the EOR project. Now, this CO2 is continuously captured, compressed, transported, liquefied, and injected into the EOR fields of Ivanić and Žutica. Tertiary (EOR) oil recovery at these two oil fields is expected to raise the oil recovery factors up to 55% from the previously achieved 9% in the primary stage and 36% in the secondary stage. Besides the EOR project, this article reviews other carbon capture, utilization, and storage (CCUS) projects in Croatia, for the cement industry, power generation, and biofuel production. All these projects combined could bring the total CCUS capacity up to 1.843 Mtpa by 2030.

In this paper, energy efficiencies of two heating systems with heat pump, latent thermal energy storage (LTES) and solar collectors that are working in moderate continental climate conditions were numerically analysed and compared. The two considered LTES roles are the heat source for the heat pump (LTES on the evaporator side) and high temperature heat storage for directly supplying the consumers (LTES on the condenser side). In the first system, the purpose for using LTES is to provide more favourable working conditions for the heat pump by using previously collected and stored heat from solar collectors and increase its average coefficient of performance. In the second system, LTES is used for storing the collected solar heat on a sufficiently high temperature so it can be used directly for heating, avoiding the use of heat pump in those periods. The numerical investigations were performed for the whole heating period and the results were presented in terms of the share of renewable energy in total delivered energy, seasonal electricity consumption, system seasonal performance factor, total stored energy and average solar collector efficiency. System with LTES on the condenser side achieved a higher share of renewable energy and higher seasonal performance factor and also consumed less electricity than the system with LTES on the evaporator side.

The thermoeconomic performance of geothermal power plants is influenced by a variety of site-specific factors, major economic variables, and the type of the involved technology. In addition to those, ambient conditions also play a role in geothermal power generation by acting on the cooling towers. This study focuses on the performance analysis of a binary cycle with isobutane for geothermal power generation under the impact of climate change. Long-term temperature variations in ambient air are described by temperature anomalies under two shared socioeconomic pathways (SSP). These are the intermediate SSP2-4.5 scenario and the extreme SSP5-8.5 scenario, over the period from 2021 to 2100. Different climate models from the most recent Climate Model Intercomparison Project (CMIP6) are compared against each other and against the observed temperature data. The predictive power of the CMIP6 climate models is evaluated using the root mean square error (RMSE) and the Kullback–Leibler (KL) criteria. The thermoeconomic performance of the geothermal power plant is expressed in terms of net power output, annual electricity generation (AEG), and levelized cost of electricity (LCOE). The geothermal power plant achieves a net power output of 10 MW and an LCOE of 79.2 USD/MWh for an ambient air temperature of 12 °C. This temperature is the average temperature over the reference period of 1991–2020 in Bjelovar, Croatia (45.8988° N, 16.8423° E). Under the impact of climate change, the same geothermal power plant will have the AEG reduced by between 0.5% and 2.9% in the intermediate (SSP2-4.5) scenario and by between 2.0% and 8.7% in the extreme (SSP5-8.5) scenario. The LCOE will increase between 0.4% and 1.8% in the intermediate scenario and from 1.3% to 5.6% in the extreme scenario.

Heating, ventilation, and air conditioning (HVAC) systems are a popular research topic because buildings’ energy is mostly used for heating and/or cooling. These systems heavily rely on sensory measurements and typically make an integral part of the smart building concept. As such, they require the implementation of fault detection and diagnosis (FDD) methodologies, which should assist users in maintaining comfort while consuming minimal energy. Despite the fact that FDD approaches are a well-researched subject, not just for improving the operation of HVAC systems but also for a wider range of systems in industrial processes, there is a lack of application in commercial buildings due to their complexity and low transferability. The aim of this review paper is to present and systematize cutting-edge FDD methodologies, encompassing approaches and special techniques that can be applied in HVAC systems, as well as to provide best-practice heuristics for researchers and solution developers in this domain. While the literature analysis targets the FDD perspective, the main focus is put on the data-driven approach, which covers commonly used models and data pre-processing techniques in the field. Data-driven techniques and FDD solutions based on them, which are most commonly used in recent HVAC research, form the backbone of our study, while alternative FDD approaches are also presented and classified to properly contextualize and round out the review.

Heating, ventilation, and air conditioning (HVAC) systems are essential for maintaining a comfortable indoor environment in modern buildings. However, HVAC systems are known to consume a lot of energy, which can account for up to 50% of a building's energy consumption. Therefore, it is important to detect and troubleshoot problems in HVAC systems timely. Fault detection and diagnosis (FDD) techniques can help with HVAC monitoring and optimizing system performance for efficient use of energy. In this paper, we demonstrate how to create efficient fault detectors using physics-based modeling and machine learning. We show how to build a simulation model of a hotel building, which we then use to sample augmented data with typical faults commonly found in HVAC systems. We train predictive models using random forests (RFs). The results suggest that RFs can be used as stand-alone detectors for FDD, albeit their performance depends heavily on the data quality.

Optimizing the performance of heating, ventilation, and air-conditioning (HVAC) systems is critical in today's energy-conscious world. Fan coil units (FCUs) play a critical role in providing comfort in various environments as an important component of HVAC systems. However, FCUs often experience failures that affect their efficiency and increase their energy consumption. In this context, deep learning (DL)-based fault detection offers a promising solution. By detecting faults early and preventing system failures, the efficiency of FCUs can be improved. This paper explores DL models as fault detectors for FCUs to enable smarter and more energy-efficient hotel buildings. We tested three contemporary DL modeling approaches: convolutional neural network (CNN), long short-term memory network (LSTM), and a combination of CNN and gated recurrent unit (GRU). The random forest model (RF) was additionally developed as a baseline benchmark. The fault detectors were tested on a real-world dataset obtained from the sensory measurement system installed in a hotel and additionally supplemented with simulated data via a physical model developed in TRNSYS. Three representative FCU faults, namely, a stuck valve, a reduction in airflow, and an FCU outage, were simulated with a much larger dataset than is typically utilized in similar studies. The results showed that the hybrid model, integrating CNN and GRU, performed best for all three observed faults. DL-based fault detectors outperformed the baseline RF model, confirming these solutions as viable components for energy-efficient hotels.

With the expected increase in world population and economic growth, the demand for energy is steadily increasing worldwide. To support the growing exploitation of energy resources, buildings must be made smart enough to use energy efficiently without wasting it. Heating, ventilation, and air conditioning (HVAC) systems consume a lot of energy and hence require effective control and analysis capabilities, which is one of the most important aspects of current smart buildings. If HVAC systems are not properly controlled, this can lead to poor user comfort, high utility bills due to excessive energy consumption, and a negative impact on the environment. This problem is aggravated in the context of hotel buildings. To support the overall energy consumption tracking, we propose a system integration with an analytical web application for data visualization that allows the user to perform an in-depth analysis of the sensory data measured. The motivation for developing the application is primarily to monitor the regulation of the HVAC system by providing an interactive analytical dashboard needed for future smart buildings. The data visualization tool provides a quick and easy way to visualize and compare time-series data. The visualizations provide insight into how different hotel rooms compare, deviations of room temperature from setpoint temperature, data distribution, correlations between sensory data parameters, and the degree of correlation of room temperature compared to other parameters. The application is aimed at a wide range of users, from lay people and hotel staff to data analysts and scientists conducting research in this area.

The scavenging process is an important part of the two-stroke engine operation. Its efficiency affects the global engine performance such as power, fuel consumption, and pollutant emissions. Slow speed marine diesel engines are uniflow scavenged, which implies inlet scavenging ports on the bottom of the liner and an exhaust valve on the top of the cylinder. A CFD model of such an engine process was developed with the OpenFOAM software tools. A 12-degree sector of the mesh was used corresponding to one of the 30 scavenging ports. A mesh sensitivity test was performed, and the cylinder pressure was compared to experimental data for the analyzed part of the process. The scavenging performances were analyzed for real operation parameters. The influence of the scavenge air pressure and inlet ports geometric orientation was analyzed. The scavenging process is analyzed by means of a passive scalar representing fresh air in the cylinder. Isosurfaces that show the concentration of fresh air were presented. The variation of oxygen and carbon dioxide with time and the axial and angular momentum in the cylinder were calculated. Finally, the scavenging performance for the various operation parameters was evaluated by means of scavenging efficiency, charging efficiency, trapping efficiency, and delivery ratio. It was found that the scavenging efficiency decreases with the engine load due to the shorter time for the process. The scavenging efficiency increases with the pressure difference between the exhaust and scavenging port, and the scavenging efficiency decreases with the increase in the angle of the scavenging ports. It was concluded that smaller angles than the industry standard of 20° could be beneficial to the scavenging efficiency. In the investigation, the charging efficiency ranged from 0.91 to over 0.99, the trapping efficiency ranged from 0.54 to 0.83, the charging efficiency ranged from 0.78 to 0.92, and the delivery ratio ranged from 1.21 to 2.03.

This paper presents an experimental research of melting of paraffin inside a latent thermal energy storage. Experimental setup consists of a water-water heat pump, hot and cold water tanks and latent storage tank. Latent heat storage is a shell-and-tube type tank, consisting of 19 concentric tubes with eight equidistant longitudinal fins. A set of 30 thermocouples has been placed at various positions longitudinally and radially throughout the paraffin-side of the storage to monitor the timewise temperature changes during melting, while additional two thermocouples have been placed at water inlet and outlet positions. Data acquisition system, linked with a personal computer, was used for acquiring, processing and storing measured data. Experimental research has been performed in the Laboratory for thermal measurements at University of Rijeka, Faculty of Engineering. A set of experiments, with different water inlet temperatures, has been performed. Transient temperature changes of phase-change material (PCM) for different water inlet temperatures, and the amount of stored thermal energy have been analyzed and compared.

This paper presents a numerical investigation of heat and mass transfer inside a wet cooling tower with forced air draft, which find application in energy process industries and oil refineries. The mathematical model consists of mass, momentum and energy conservation equations, water droplet trajectories and their interaction with the gas phase, the computational domain and boundary conditions. Numerical distributions of air velocity, air temperatures, water vapor fractions and evaporation rates are shown and discussed. The wet cooling tower achieves an efficiency of around 80%, which can be improved by optimizing the value of the water droplet size, nozzle spray angle and water-to-air flow rate ratio. The water droplet size has a dominant effect on the cooling tower efficiency, whereas small droplets improve the efficiency up to 10%. On the other hand, the spray angle and the water-to-air ratio lead to slight improvements, about 2-3% in the best case.

This paper investigates on the space heating and cooling energy demand as well as on the thermal comfort in energy efficient detached single-family houses with mechanical ventilation heat recovery system. The necessary energy calculations are based on a test reference year with one-hour-step weather data taken for the City of Rijeka which can be seen as an average location in Croatia in terms of weather conditions. The analysis consists of the comparison of three houses with different energy performance levels: a reference single-family house, a low-energy single-family house and a passive single-family house. As expected, the space heating energy demand is lowest in the passive house and highest in the reference house. The thermal comfort is best in the passive house and worst in the reference house. It is concluded that the mechanical ventilation heat recovery system is an indispensable part of the energy system in the passive house and that it can ensure significant energy savings in the low-energy house, too.

This paper investigates on the space heating and cooling energy demand as well as on the thermal comfort in energy efficient detached single-family houses with mechanical ventilation heat recovery system. The necessary energy calculations are based on a test reference year with one-hour-step weather data taken for the City of Rijeka which can be seen as an average location in Croatia in terms of weather conditions. The analysis consists of the comparison of three houses with different energy performance levels: a reference single-family house, a low-energy single-family house and a passive single-family house. As expected, the space heating energy demand is lowest in the passive house and highest in the reference house. The thermal comfort is best in the passive house and worst in the reference house. It is concluded that the mechanical ventilation heat recovery system is an indispensable part of the energy system in the passive house and that it can ensure significant energy savings in the low-energy house, too.

CFD analysis of heavy fuel oil combustion in a 6.7 MW cylindrical vertically-fired furnace has been carried out. The furnace supplies process heat in the Oil refinery of the National oil company (INA) in Rijeka-Croatia. The motivation of the work was to improve the performance of the combustion process by changing fuel and burner parameters. The commercial CFD-code Fluent is used to model transport and reaction in the furnace. The chosen CFD models for heavy fuel oil spray combustion are compared with measurement data found in the literature and good agreement is achieved. The combustion process is investigated through the influence of different parameters: air-fuel ratio, fuel oil droplet diameter, spray cone half-angle and burner swirl number. It is determined that the best performance of the combustion process is achieved for an air-fuel ratio of 1.15. The droplet diameter should be neither too small nor too large: medium-sized droplets (~100 μm) mix well with air and reside in the furnace the right amount of time for complete burnout to occur. Small spray cone half-angles deteriorate mixing between air and fuel which is reflected in higher CO concentrations. The burner’s swirl number influences the mixing rate of air and fuel: a small swirl number is not desired as it elongates the flame, increases flue gases temperatures and reduces the furnace heating output.

One of the aims of the presented investigation, carried out as a part of a PhD thesis, was to develop a reliable numerical model that can be used for assessments of thermal comfort in open – plan offices. The office under the investigation is air – conditioned by two parallel ceiling air terminal devices (ATDs). During simultaneous operation of the ATDs, an interaction of produced turbulent wall jets can occur, resulting in creation of complex airflow patterns within the occupied zone that influence on the achievable level of thermal comfort. A three – dimensional steady – state mathematical model of the fluid flow and heat transfer problem was solved using a commercial CFD software Fluent, based on the control volume method. The numerical results were compared to experimental measurements, carried out in a full – scale test room situated in the Laboratory of Heating, Ventilating and Air – Conditioning at the Faculty of Engineering, University of Rijeka. The comparison confirmed the model as a promising tool for assessments of thermal comfort in similar rooms.