Skip to content

Development, investigation and Graywolf algorithm-assisted tri- objective optimization of a hybrid renewable-driven HVAC/power system.

The rising energy demand of the building sector has led to a critical need for innovative solutions to curb its 37 % share of global CO₂ emissions in 2020. This study introduces a novel hybrid energy system integrating solar photovoltaic-thermal (PVT) panels with a biofuel-driven boiler, designed to meet residential heating, cooling, and electricity needs. Dynamic simulations were conducted using TRNSYS to evaluate the system s real-world performance across varying climatic conditions. Optimization using the greywolf algorithm achieved a balance among three competing objectives: thermodynamic efficiency, environmental sustainability, and economic viability. The system delivered a thermal efficiency of 28.5 %, a levelized energy cost of 0.163 $/kWh, and a CO₂ emission index of 121 kg/MWh, reflecting significant improvements of 13.5 %, 23.1 %, and 15.4 %, respectively, over the baseline design. Unlike prior studies, this work uniquely combines dynamic transient modeling with multi-objective optimization for a solar-biofuel energy system, offering actionable insights into sustainable energy integration for the building sector. These findings underscore the system s potential to enhance energy performance while reducing environmental and economic costs. https://www.sciencedirect.com/science/article/abs/pii/ S2352710225001512

Heat recovery integration in a hybrid geothermal-based system producing power and heating using machine learning approach to maximize outputs

This study offers an in-depth thermodynamic analysis and optimization of an integrated renewable energy system that merges a double-flash geothermal system with a transcritical carbon dioxide Rankine cycle, utilizing machine learning algorithms. The innovative design aims to maximize the concurrent generation of heat and electricity, ultimately benefiting environmental sustainability and energy security. By employing regression machine learning algorithms, the research evaluates and enhances system performance, achieving remarkable R-squared accuracy levels of 98.86 % for heating output and 99.89 % for power output predictions. The thermodynamic modeling, which has been validated against recognized benchmarks, confirms the accuracy of the system s design. Optimization findings indicate that operating pressures between 840 and 870 kPa and pressure ratios of 1.56–1.60 deliver optimal outputs, with power production between 2582 and 2585 kW and heating output ranging from 12260 to 12280 kW. The system reaches its maximum performance at a pressure of 850 kPa and a pressure ratio of 1.57, resulting in a power output of 2583.97 kW and a heating output of 12279.3 kW. These results highlight the potential of combining advanced thermodynamic systems with machine learning methodologies to improve the efficiency and effectiveness of renewable energy sources. https://www.sciencedirect.com/science/article/pii/ S2214157X24012413

Optimizing and investigating the charging time of phase change materials in a compact-latent heat storage using pareto front analysis, artificial neural networks, and numerical simulations

There s a growing emphasis on adopting eco-friendly energy sources to mitigate greenhouse gas emissions and foster sustainability. While renewable energy sources like solar power offer numerous benefits, they have some limitations. Additionally, storing electricity generated from solar panels can be costly and challenging due to limitations in battery technology and storage capacity. Therefore, phase change material-based thermal energy storage systems offer a promising solution to these challenges. These systems also play a crucial role in enhancing buildings energy efficiency and sustainability. The dimensionality of these systems affects their performance. Traditional systems often rely on fixed dimensions, leading to non-uniform thermal conditions within the storage medium. To address these challenges, adopting a compact-latent heat storage (C-LHS) mechanism was recommended herein. Besides, some fins were inserted into the C-LHS system to alter the heat transfer dynamics within the device. Three of the critical parameters of the fins were changed to examine their impact on the charging time. Artificial neural networks and genetic algorithms were employed to determine the optimal positioning of fins to minimize the duration of material to get both part (melt fraction of 0.8) and full (melt fraction of 1) melting capacities. Here, the primary aim was to present a predictive model capable of accurately forecasting the charging time of the material. In all the presented finned samples, the entire PCM melted in less than 15,110 s (4 h and 12 min) and was able to fully utilize the energy absorption capacity in latent form, whereas in the non-finned system, only 68.6 % of the material managed to melt within the entire duration of the investigation (5 h). After analyzing the data, two optimal configurations of OS1 and OS2 with the minimum time for melt fractions of 0.8 and 1 were introduced. In the OS2 configuration, it took 16,200 s (4 h and 30 min) to absorb 4929 kJ of energy. The non-finned sample absorbed only 3250 kJ of energy during the same period. In General, OS2 achieved total energy absorption 51.6 % faster than the non-finned sample. Therefore, the introduced system has the potential to increase absorbed energy in the rest of the daylight hours to further amounts by increasing the number of units and employing a larger volume of material. https://www.sciencedirect.com/science/article/abs/pii/ S2352152X24035527

Can injecting additional green hydrogen result in environmentally friendly solar-biomass integration? Comprehensive comparison and multi-objective optimization

The present work proposes an innovative system for decarbonizing the energy mix and accelerating the worldwide green transition process. The system is driven by a biomass digester integrated with the supercritical carbon dioxide cycle for power generation and a multi-effect desalination unit for drinkable water production. At the heart of this concept is additional hydrogen injection through a proton exchange membrane electrolyzer based on photovoltaic panels. The suggested innovative model s techno-environmental, sustainability, and economic aspects are assessed and compared with a similar system without hydrogen injection. Then, a comparative multi-criteria optimization is applied to find the most optimal conditions from various facets based on the genetic algorithm with machine learning techniques. Afterward, the system’s performance at different optimal conditions is analyzed and compared by evaluating the most significant techno-economic, environmental, and sustainability indicators. The parametric assessment comparing different models indicates that the proposed novel model, including increased hydrogen injection, surpasses the basic system in terms of performance efficiencies, emissions, and energy costs. In the first optimization scenario, the proposed method demonstrates robustness by achieving higher water production of 1456 kg/s, a lower total cost of 118 $/h, and a higher net power of 1.1 MW than the design condition. When considering the sustainability index, energy cost, and emission metric as the optimization objective, their values are altered from 0.81 to 0.85, 92.5 $/MWh to 89.7 $/MWh, and 64.2 kg/MWh to 53.6 kg/MWh. The results further show that when prioritizing the sustainability index, energy cost, and emission as objectives, all components perform better from the energy conversion quality aspect compared to the scenario where water production, total cost, and net power are the optimization objectives. Finally, it is observed that the combustion chamber and solar panels are the worst components from the irreversibility aspect because of the highest exergy destruction rate. https://www.sciencedirect.com/science/article/abs/pii/ S0957582024004786?via%3Dihub

Thermal and Entropic analyses of free convection of nanofluid in a partially heterogeneous porous chamber using the two-phase mixture model

Abstract Entropy generation and convection heat transfer in a partially porous chamber with different side wall temperatures using CuO–H2O have been investigated. The importance of this issue is wide application of the results in solar collectors, thermal extrusion systems, heat exchangers, bio-medicine, nuclear waste disposal, etc. The innovation of the present work is related to the investigation of fluid and heat fields and entropy generation by using a matrix with subordination of porosity to the vertical axis and permeability, thermal conductivity, and viscositywith subordination of porosity. To obtain accurate results, the two-phase mixture model was used, and thermal conductivity and viscosity of nanofluid were simulated by experimental models by temperature and volume fraction dependence. Governing equations are solved by the FVM. The main findings indicate that the best and worst optimization factor will occur in the porous matrix ε = ε(y2) and ε = -ε(y2), respectively, which is 113 % and 86 % of NH of the homogeneous matrix, respectively. Also increasing the filling of the cavity, highly improves NH, so that the NH will reach from 1.19 to 1.82 with the increase of S from 0.25 to 1. https://www.sciencedirect.com/science/article/pii/ S2214157X24005161?via%3Dihub

Heat absorption/release efficiency betterment of phase change material inside a shell-and-tube latent heat storage system under six different conditions of tube and fins

Researchers are exploring innovative solutions for thermal energy storage to address the challenges posed by intermittent renewable sources, enhance energy efficiency, and contribute to the global shift towards cleaner and more sustainable energy practices. In the pursuit of an optimal system to improve the heat release/absorption efficiency of phase change materials (PCMs), a unique shell and tube latent heat storage system with four rectangular fins was designed. The melting and solidification behaviors of the material in this device were examined by manipulating the tube s position within the shell and fins around the tube. Six different cases were considered such as case A (tube in the center of the shell), case B (tube at the top of the shell), case C (tube at the bottom of the shell), case D (tube in the center of the shell with fins on its sides), case E (tube at the top of the shell with fins located in its bottom section), and case F (tube at the bottom of the shell with fins located in its top section). Cases D and E were the best options for absorbing and releasing heat in the shortest time. However, it should be noted that case F was faster during the melting process and dropped behind in the final stages. The authors proposed that if achieving a balanced result without incurring additional costs is essential, case D is a suitable option since it offers reasonable performance in melting and solidification processes. However, suppose researchers and developers of energy storage systems are seeking higher performances where heat absorption and release occur much more rapidly. In that case, it is suggested to construct one of the cases, E or F, and implement a rotational mechanism to enable access to the other case. Based on the outcomes, cases D and F needed 6235 s and 7552 s, respectively, to fully melt. While all cases even required more than 3 h to solidify 80 % of the PCM. The complete melting speed of Case F is 21.12 % faster than that of Case D. Additionally, the time required for 50 % solidification is 14.79 % faster for Case E compared to Case D. During 3 h, this system could absorb 1172 kJ of energy (cases D and F) and release 893 kJ of energy (cases D and E). https://www.sciencedirect.com/science/article/abs/pii/ S2352152X24014658?via%3Dihub

Artificial intelligence application for assessment/optimization of a cost-efficient energy system: Double-flash geothermal scheme tailored combined heat/power plant

Utilizing the capabilities of artificial intelligence can lead to the development of energy systems and power supply chain that are more efficient, sustainable, and resilient. The integration of machine learning techniques within these systems provides substantial benefits and is essential for enhancing overall performance. As the global community confronts challenges like climate change and rising energy demands, machine learning will play an increasingly vital role in defining the future of energy systems. This research examines how effective regression-based machine learning techniques are for analyzing and optimizing the performance of a geothermal combined heat and power system. It focuses on creating both linear and quadratic models to assess electricity generation, heat production, and the efficiency of the entire system. The evaluation of these models is performed through residual analysis and R-squared statistics. Results indicate that quadratic models surpass linear ones, with linear model achieving an R-squared value of 88.56 % for power generation, while the quadratic model reaches an impressive R-squared level of 99.88 %. Furthermore, the study demonstrates that quadratic machine learning models hold significant promise for optimizing system performance, shown by desirability metrics exceeding 0.99. This research highlights the importance of regression-based machine learning methods in analyzing and improving geothermal combined heat and power systems. https://www.sciencedirect.com/science/article/abs/pii/ S0360544224033723

Heat energy utilization of a double-flash geothermal source efficiently for heating/electricity supply through particle swarm optimization method

The principal aim of this article is to optimize the thermal and electrical efficiency of a geothermal combined heat and power system through metaheuristic particle swarm optimization (PSO) method. The objective of this research is to conduct a thorough analysis of the incorporation of metaheuristic PSO technique, with a specific emphasis on the potential advantages and obstacles associated with the utilization of metaheuristic approaches in improving the effectiveness of geothermal energy systems. The utilization of a double-flash geothermal system in conjunction with a transcritical carbon dioxide Rankine cycle is utilized for the co-generation of electricity and thermal energy. The research utilized a PSO method to enhance power generation, heating capacity, and overall system efficiency. The PSO algorithm was employed to determine the optimum operational parameters for a pressure level of 820 kPa and a pressure ratio of 1.59, leading to the maximization of power output to 2591.4 kW The PSO algorithm effectively identified the optimal operational parameters as a pressure of 820 kPa and a pressure ratio of 1.59, resulting in the achievement of a peak power output of 2591.4 kW. The methodology has determined that a pressure of 916.4 kPa and a pressure ratio of 1.5 represent the optimal parameters for achieving a maximum heating capacity of 12329.1 kW. https://www.sciencedirect.com/science/article/pii/ S2214157X24013741