Investigation of Power Consumption of Refrigeration model and its Exploratory Data Analysis (EDA) by using Machine Learning (ML) Algorithm

Investigation of Power Consumption of Refrigeration model and its Exploratory Data Analysis (EDA) by using Machine Learning (ML) Algorithm

1. 13th International Advanced Computing Conference on 15th & 16th Dec'23 @ Sharad Institute of Technology, College of Engineering, Kolhapur, Maharashtra    Mr. Avesahemad S. N. Husainy Investigation of Power Consumption of Refrigeration model and its Exploratory Data Analysis (EDA) by using Machine Learning (ML) Algorithm Conference paper title: Presented by: Co-Authors : Suresh M. Sawant , Sonali K. Kale, Anirban Sur, Sagar D. Patil , S. V. Kumbhar , V. V. Patil

2. OUTLINE OF PRESENTATION INTRODUCTION DIFFERENT HEAT TRANSFER ENHANCEMENT TECHNIQUES PREPARATION OF NANO-PCM FRAMEWORK OF ML IMPLEMENTATION EXPERIMENTATION AND OBSERVATIONS ML BASED RESULTS CONCLUSION REFERENCES

3. INTRODUCTION Source: India Cooling Action Plan The market size of the HVAC industry is projected to grow from $7,820 million in 2021 to $29,282 million in 2030 . The HVAC industry is expected to see a POSITIVE outlook in 2023, with a projected compound annual growth rate (CAGR) of 15.8 % from 2021 to 2030. Cold chain technology Problem statement (Power consumption) Need thermal storage in refrigeration industry Applications of Nano-PCM Heat transfer enhancement by NPCM Fig.1

4. Postharvest refrigerated supply chain of fruits and vegetables.

5. Applied Energy  283:116277 Sustainability  8(10):1046 CLASSIFICATION THERMAL ENERGY STORAG AND PHASE CHANGE MATERIAL Fig.2 Fig.3

6. Framework of machine learning implementation

7. PREPARATION OF NANO-PCM Preparation of Nano-PCM Fig.6

8. CHARACTERIZATIONS OF Grnp NANO-PCM The microstructure and elemental composition of graphene-phase change material composites were examined using a scanning electron microscope (SEM) and EDAX and are depicted in figure (a) and (b). The homogeneity of the composite material was investigated after the addition of graphene-PCM at different proportions (1% and 5 %) to determine whether the graphite particles have dispersion with PCM. It is observed from figure, that the phase change material is unevenly distributed on graphene surface or may be inside the graphene surface and adhered as a thin layer on the graphene surfaces. The agglomerated flakes of graphene are observed in SEM images. The compositional analysis of samples shows the concentration of elements like C, K, S, O present in both the samples (1% and 5 %). Volume 72, Part 3 , 2023, Pages 1510-1516, Published in Materialstoday Proceedings Fig.7

9. EXPERIMENTATION AND OBSERVATIONS Experimental set up Modified Evaporator section Cooling cabinet Fig.8 Fig.9

10. Sr. No. Name of TES Percentage 1 Eutectic Mixture of PCM ( KCl+NaCl ) 10 % each 2 Eutectic Mixture of PCM ( KCl+NaCl ) + Cuo (  80 nm) nanoparticles 10 % each PCM + 1 % CuO Nanoparticles 3 Eutectic Mixture of PCM ( KCl+NaCl ) + Grnp (  70 nm) nanoparticles 10 % each PCM + 1 % Grnp Nanoparticles TES COMBINATIONS FOR EXPERIMENTATIONS

11. The experiment is conducted with three different combinations of TES materials and comparison is made without consideration of TES used in evaporator section. During power OFF mode it is observed that without use of TES in evaporator section, cabinet temperature will reach at 24 C/normal temperature of cabinet in 13 hrs . By using phase change material as Eutectic mixture of 10 % KCl and 10% NaCl as TES will maintain 13 to 15 C in same time interval. By using Copper oxide ( CuO 1 %) as nanoparticles and Eutectic mixture of 15 % NaCl and 15 % KCl as PCM will maintain constant temperature 13 to 14 C in same time interval. Whereas by using graphene (1% Gnp ) as nanoparticle and Eutectic mixture of 15 % NaCl and 15 % KCl as PCM will reach 10 to 11 C COMPARISON OF TIME VS CABINET TEMPERATURE WITH DIFFERENT TES MATERIALS Fig.10

12. From above results it is observed that, as compared with eutectic mixture of PCM and CuO+PCM , Grn+PCM will give better results and slow phase change/solidification rate and it will helps to keep the cabinet temperature constant for longer duration of time. Also the fraction (f) of different thermal energy storage materials is studied during charging and discharging by using Nano-PCM ( Grn+PCM ) combination as shown in fig.11 and comparison of liquid fraction is studied by using PCM, CuO+PCM , Grn+PCM combinations during discharging process DISCHARGING TEMP OF TES AT DIFFERENT COMBINATIONS Fig.11

13. Proposed methodology

14. Fig.11. Difference of temperatures distribution by use and without use of NPCM during power ON mode by using displot

15. Fig.13. Pair plot for NPCM mixture result during power ON mode Fig.14. Pair plot for without NPCM mixture result during power ON mode

16. Scatter plot of implemented linear regression algorithm with NPCM Model Accuracy 96 % Scatter plot of implemented linear regression algorithm without NPCM Model Accuracy 82 %

17. Freezer electrical energy prediction service

18. Sr. No. TES Combination Cabinet Temp during power off TES material temperature during discharging/melting COP Enhancement during power ON Power consumption 1 With TES 22 to 24 degree after 4 to 5 hrs ----- ----- ----- 2 Eutectic Mixture of PCM (10% KCl + 10% NaCl) 15 to 16 degree after 12 to 13 hrs 18 degree in 12 hrs Increased by 10 % Reduced by 7 % 3 Eutectic Mixture of PCM (10% KCl + 10% NaCl ) + 1 % Cuo (  80 nm) nanoparticles 13 to 14 degree after 12 to 13 hrs 16 degree in 12 hrs Increased by 12 % Reduced by 9 % 4 Eutectic Mixture of PCM (10% KCl + 10% NaCl) + 1 % Grnp (  70 nm) nanoparticles 10 to 11 degree after 12 to 13 hrs 15 degree in 12 hrs Increased by 15 % Reduced by 12 % CONCLUSION The integration of NEPCM has demonstrated promising results, showcasing enhanced heat absorption and release capabilities during faster heat transfer during both the charging (melting) and discharging (solidification) phases .

19. Investigating how much power refrigeration systems use and combining Machine Learning (ML) methods with Exploratory Data Analysis (EDA) have produced insightful findings with real-world applications. Correlations and patterns found by EDA help us comprehend how different elements, such usage patterns and temperature settings, interact to effect power consumption . In conclusion, the application of machine learning technology has a lot to offer the refrigeration sector. Machine learning lowers costs, lessens environmental impact, and improves product quality and safety by maximizing energy efficiency, enabling predictive maintenance, and improving overall system performance.

20. REFERENCE   Loisel , J., Duret , S., Cornuéjols , A., Cagnon , D., Tardet , M., Derens-Bertheau , E., & Laguerre, O. (2021). Cold chain break detection and analysis: Can machine learning help? Trends in Food Science & Technology, 112, 391-399. Lundqvist , J., De Fraiture , C., & Molden , D. (2008). Saving water: from field to fork: curbing losses and wastage in the food chain. Bustos, C. A., & Moors, E. H. (2018). Reducing post-harvest food losses through innovative collaboration: Insights from the Colombian and Mexican avocado supply chains. Journal of Cleaner Production, 199, 1020-1034. Dos Santos, S. F., Cardoso, R. D. C. V., Borges, Í. M. P., e Almeida, A. C., Andrade, E. S., Ferreira, I. O., & do Carmo Ramos, L. (2020). Post-harvest losses of fruits and vegetables in supply centers in Salvador, Brazil: Analysis of determinants, volumes and reduction strategies. Waste Management, 101, 161-170. Food Wastage Footprint (Project). (2013). Food wastage footprint: impacts on natural resources: summary report. Food & Agriculture Organization of the UN (FAO). Chavan , S., Rudrapati , R., & Manickam , S. (2022). A comprehensive review on current advances of thermal energy storage and its applications. Alexandria Engineering Journal, 61(7), 5455-5463. Bertoldi , P., & Atanasiu , B. (2007). Electricity consumption and efficiency trends in the enlarged European Union. IES–JRC. European Union.