2795-1529 (Print) ISSN 2795-1537 (Online)
JOURNAL OF ENGINEERING ISSUES AND SOLUTIONS
Volume 1, Issue 1 April 2021
Official Journal of Nepal Engineers’ Association (NEA) 2795-1529 (Print) ISSN 2795-1537 (Online)
JOURNAL OF ENGINEERING ISSUES AND SOLUTIONS
Volume 1, Issue 1 April 2021
Official Journal of Nepal Engineers’ Association (NEA) Journal of Engineering Issues and Solutions (JoEIS), an official Journal of Nepal Engineers’ Association (NEA), is managed by “R&D, Research Journal Publication Committee” of NEA. The Committee comprises of following members;
• Er. Dr. Indra Prasad Acharya, Chair • Er. Dr. Bhesh Raj Thapa, Co-Chair • Er. Prof. Vishnu Prasad Pandey, PhD, Co-Chair • Er. Hari Lal Kharel, Co-Chair • Er. Dr. Binod Yadav, Member • Er. Dr. Pawan Kumar Bhattarai, Member • Er. Dipendra Gautam, Member • Er. Naresh Dhami, Member • Er. Prakash Gaire, Member • Er. Dilip Kumar Bishwkarma, Member Secretary Editorial Board
Advisory Editors • Prof. Dr. Tri Ratna Bajracharya, Institute of • Prof. Dr. Subarna Shakya, Computer Engineering, Engineering, Tribhuvan University, Nepal Institute of Engineering, Tribhuvan University, Nepal. • Prof. Dr. Narendra Man Shakya, Institute of Engineering, Tribhuvan University, Nepal. • Prof. Dr. Binod Tiwari, California State University, Fullerton, USA. • Prof. Dr. Rajesh Dhakal, University of Canterbury, New Zealand
Editor-in-Chief • Prof. Dr. Vishnu Prasad Pandey, Institute of Engineering, Tribhuvan University, Nepal.
Editors I. Architecture/Agriculture/ Civil/Environmental Tribhuvan University, Nepal. 1. Dr. Alka Sapkota, Whitehouse International 20. Dr. Rocky Talchabhadel, Texas A&M University; College, Kathmandu, Nepal. USA. 2. Dr. Anir Upadhyay, UNSW, Australia. 21. Dr. Rojee Pradhananga, Institute of Engineering, 3. Dr. Bhesh Raj Thapa, Universal Engineering and Tribhuvan University, Nepal. Science College, Affiliated with Pokhara University, 22. Dr. Sanjay Uprety, Institute of Engineering, Nepal. Tribhuvan University, Nepal. 4. Prof. Dr. Binaya Kumar Mishra, Pokhara 23. Dr. Sheela Katuwal, University of Arkansas, USA. University, Nepal. 24. Dr. Shreemat Shrestha, Nepal Agricultural 5. Dr. Bishnu Prasad Gautam, Province Planning Research Council (NARC), Government of Nepal. Commission, Lumbini Province, Government of 25. Shukra Raj Paudel, Institute of Engineering, Nepal. Tribhuvan University, Nepal. 6. Dipendra Gautam, Interdisciplinary Research 26. Dr. Surendra Adhikari, NASA, USA. Institute for Sustainability, Nepal. 27. Dr. Yadu Nath Pokhrel, Michigan State University, 7. Hari Lal Kharel, Himalaya College of Engineering, USA. Kathmandu, Nepal. 8. Dr. Indra Prasad Acharya, Institute of II. Computer/ICT/Information Technology Engineering, Tribhuvan University, Nepal. 1. Dr. Bishesh Khanal, Nepal Applied Mathematics 9. Dr. Keshab Sharma, BGC Engineering Inc., Canada. and Informatics Institute for Research (NAAMII), 10. Dr. Kirti Kusum Joshi, City Planning Commission, Nepal. Kathmandu Metropolitan City, Nepal. 2. Prof. Dr. Danda B. Rawat, Howard University, USA 11. Prof. Dr. Netra Prakash Bhandary, Ehime 3. Dr. Gajendra Sharma, Kathmandu University, University, Japan. Nepal. 12. Prof. Dr. Padma Bahadur Shahi, Department of 4. Dr. Gyanendra Prasad Joshi, Sejong University, Transportation Management, Government of Korea. Nepal. 5. Dr. Pradeep Poudel, Nepal Telecommunication 13. Dr. Pawan Bhattarai, Institute of Engineering, Authority, Nepal. Tribhuvan University, Nepal. 6. Dr. Rajeev Kanth, Savonia University of Applied 14. Dr. Pradip Kumar Shrestha, Institute of Sciences, Kuopio, Finland. Engineering, Tribhuvan University, Nepal. 7. Dr. Ved Kafle, National Institute of Information 15. Dr. Rajan Bhattarai, Water Resources Research and and Communications Technology, Japan. Development Center (WRRDC), Government of Nepal. III. Electrical/Electronics 16. Prof. Dr. Rajesh Rupakhety, University of Iceland, 1. Dr. Arvind Kumar Mishra, Institute of Iceland. Engineering Tribhuvan University, Nepal. 17. Dr. Rajit Ojha, Ministry of Water Supply and 2. Prof. Dr. Bhanu Shrestha, Kwangwoon University, Sewerage; Government of Nepal. Korea. 18. Dr. Rajkumar GC, Virginia Tech, USA. 3. Dr. Madhusudan Kayastha, Pokhara University, 19. Dr. Ram Chandra Tiwari. Institute of Engineering, Nepal. 4. Dr. Nirmal Paudel, Institute of Engineering, Management, KTH Royal Institute of Technology, Tribhuvan University, Nepal. Sweden. 5. Dr. Surendra Shrestha, Institute of Engineering, 3. Prof. Dr. Hari Prasad Neopane, School of Tribhuvan University, Nepal. Engineering, Kathmandu University, Nepal. 4. Dr. Nawraj Bhattarai, Institute of Engineering, IV. Mechanical/Automobile/Energy Tribhuvan University, Nepal. 1. Prof. Dr. Bim Shrestha, Kathmandu University, 5. Dr. Prajal Pradhan, Postdam Institute for Climate Nepal. Impact Research (PIK), Germany. 2. Dr. Dilip Khatiwada, Department of Energy 6. Dr. Shree Raj Shakya, Institute of Engineering, Technology, School of Industrial Engineering and 7. Tribhuvan University, Nepal.
Journal Manager • Dilip Kumar Bishwkarma, Member Secretary, NEA R&D, Journal Publication Committee. Acknowledgements for Reviewers
The voluntary service from the esteemed reviewers duly governs the quality of scientific disseminations. JoEIS adopted double blind peer review process and in the due course many reviewers relentlessly volunteered to assure the greatest possible quality of the articles. The advisory editorial board, the editor-in-chief, editors, and Nepal Engineers’ Association acknowledge the great service from the following researchers who served as the reviewers for the current issue. The list comprises the names of all known reviewers of both accepted and rejected manuscripts.
Anca Jurcut Khagendra Thapa Rishi Ram Parajuli Bharat Raj Pahari Khem Nath Poudyal Roshan Chitrakar Bhim Kumar Dahal Kishan Jayasawal Runxiao Zhang Bhoj Raj Ghimire Madhu Sudan Kayastha Sailesh Chitrakar Chuanlin Hu Manish Pokharel Satish Manandhar Hari Bahadur Dura Mohammad Aqel Shyam Sundar Khadka Hari Prasad Neopane Nimesh Chettri Surya Prasad Adhikari Hemant Tiwari Nirmal Paudel Sushil Bajracharya Iswor Bajracharya Rabindra Adhikari Utsav Bhattarai Jhabindra Ghimire Rabindra Maharjan Vishnu Prasad Pandey Kamal Darlami Rajesh Khadka Khada Nanda Dulal Rajyaswori Shrestha . Table of Contents
Evaluation of weibull parameter estimators for wind speed of Jumla, Nepal...... 1 Ayush Parajuli
Assessment of the impact of COVID-19 on transportation and its inter-linked sectors of Nepal...... 8 Dipendra Bahadur Singh, Deepak Kumar Sah
Shear and tensile bond strengths of autoclaved aerated concrete (AAC) masonry with different mortar mixtures and thicknesses...... 20 Raghav Tandon, Sanjeev Maharjan, Suraj Gautam
Voltage stability analysis of wind power plant integration into transmission network of Nepal...... 32 Sagar Dharel, Rabindra Maharjan
A comparative study of structural parameters of a RCC T-girder bridge using loading pattern from different codes...... 45 Sulav Sigdel
Analyzing effectiveness of active learning through project-based learning approach in university level ICT courses...... 59 Dhiraj Shrestha, Satyendra Nath Lohani, Roshan Manjushree Adhikari
Influence of structural irregularities on seismic performance of RC frame buildings...... 70 Krishna Ghimire, Hemchandra Chaulagain
Numerical modelling of the sand particle flow in pelton turbine injector...... 88 Tri Ratna Bajracharya, Rajendra Shrestha, Ashesh Babu Timilsina
Effects of source digital elevation models in assessment of gross run-off-river hydropower potential: A case study of West Rapti Basin, Nepal...... 106 Sunil Bista, Umesh Singh, Nagendra Kayastha, Bhola NS Ghimire, Rocky Talchabhadel
Window to wall ratio and orientation effects on thermal performance of residential building: A case of Butwal Sub-Metropolis...... 129 Sanjaya Uprety, Shiva Kafley, Barsha Shrestha
Design and simulation of components of vacuum forming machine using household vacuum cleaner...... 138 Navaraj Adhikari, Nirajan Sharma Timilsina1, Sanskar Gautam, Snehraj Kaphle1, Pratisthit Lal Shrestha
Development of rainfall – runoff model for extreme storm events in the Bagmati River Basin, Nepal..... 158 Nirajan Devkota, Narendra Man Shakya
A study on FTTH implementation and migration in Nepal...... 174 Naba Raj Khatiwoda, Babu R Dawadi . Journal of Engineering Issues and Solutions
Evaluation of weibull parameter estimators for wind speed of Jumla, Nepal
Ayush Parajuli1,* 1 Department of Mechanical Engineering and Science, Kyoto University, Kyoto, Japan *Corresponding email: [email protected] Received: January 10, 2021; Revised: February 22, 2021; Accepted: February 26, 2021
Abstract Weibull Probability Distribution Function (PDF) is widely used across world for estimation of wind power. Weibull function is a two parameter probability distribution function. The methods employed for the evaluation of these two parameters are critical for the efficient use of Weibull PDF. In the present study, three different Weibull PDF parameter estimators have been evaluated. For this purpose, the daily averaged wind speed data of Jumla Station, Nepal for period of 10 year (2004 – 2014: 2012 excluded) is studied. The parameter estimator evaluated in this study are Method of Moments (MoM), Least Square Error Method (LSEM) and Power Density Method (PDM). It has been found that Method of Moments (MoM) is the best estimator for evaluating Weibull Parameters.
Keywords: Least square error method; method of moments; power density method; weibull distribution
1. Introduction To address the demand of the clean and renewable energy, wind energy is getting broader attention even in remote terrain. Wind farm design analysis has a key objective to reduce the Levelised Cost of Energy (LCoE) while addressing the pressing environmental and social concerns. LCoE is used in various studies which gives the cost to generate 1 MWh energy incorporating all lifetime cost. Primarily, it is function of initial investment, revenue generated through farm operation, routine operational and maintenance cost and salvage value. To reduce LCoE, wind farm modeling/design analysis is done. The fidelity of such model will hinge on selection of objective function, optimization method, and definition of wind characteristics, turbine modeling and wind farm interaction. This study is limited to improving the definition of wind characteristics. The wind characteristics is defined using large data sets of available wind speeds and direction in a certain period of time. So, the probability distribution of different wind speeds is used as a representation of these data sets. Further, the probability distribution will assist: a) To retrospectively characterize past conditions. b) To predict future power generation at one location.
1 Journal of Engineering Issues and Solutions 1 (1): 1-7 [2021] Parajuli c) To predict power generation within a grid of turbines. d) To calibrate meteorological data.
Figure 1: Site location of synoptic station As seen from the literature, much concentration has been given to Weibull function because it is found to give best fit to the observed wind speed data at flat terrains, mountains and off shore (Abbas et al., 2012; Ahmed, 2013; Baseer et al., 2017; Deep et al., 2020; Hulio et al., 2019; Katinas et al., 2017; Odo et al., 2012; Oner et al., 2013; Oyedepo et al., 2012; Pishgar-Komleh et al., 2015; Rehman et al., 1994; Shoaib et al., 2017). Further, the author have also shown Weibull function to perform better than Rayleigh distribution in earlier research (Parajuli, 2016). Weibull distribution is a two parameter function. The two parameters are shape parameter (k) and scale parameter (c). Shape parameter is a dimensionless parameter and scale parameter is a dimensional parameter (Wais, 2017). Several methods can be deployed to estimate the parameters and the efficiency of these estimators determine the accuracy of the PDE (Chaurasiya et al., 2018). Therefore, this study is done with a single agenda i.e. to find the best Weibull parameter estimator for the given site. The reader should refer to Section 2.1 for details of the site.
2. Materials and Methods 2.1 Site Location and Data Collection Jumla, Nepal is a potential site for wind energy generation and has been shown to be a class III site in the previous study done by the author (Parajuli, 2016). Department of Hydrology and Meteorology (DHM) has a synoptic station in Chandannath Municipality of Jumla. The site is located at at 2300m above sea level. Wind speed was measured at height of 10m from ground level and average daily wind speed was available. Wind speed from 2004 to 2014 (2012 excluded) was used for analysis. The location of site is as shown in Figure 1. Previous studies to find best Weibull parameter estimators were done on sites at lower altitudes and flat terrains (Ahmed, 2013; Azad et al., 2014; Guarienti et al., 2020; Saeed et al., 2020; Tiam Kapen et al., 2020). We know that the effect of shear and mountain wakes are critical in higher altitude as atmosphere is a density stratified medium. Therefore, the application of findings at lower altitude shall be extended to higher altitude with caution. Therefore this study, which has been done with wind data of a site at 2300m altitude,
2 Parajuli Journal of Engineering Issues and Solutions 1 (1): 1-7 [2021] will provide insights whether the findings at lower altitude stay valid at higher altitude. 2.2 Vertical Extrapolation of Wind Speed To calculate the total wind energy potential, the measured wind speed must be modified for turbine hub height above ground level for small and moderate sized turbine. With the assumption of use of moderately sized turbine for difficult terrain, the single point extrapolation to mean hub height is done using the following equation (Abbas et al., 2012):
(1)
Where, u is the wind speed at measured wind speed (m/s), y is the measuring equipment height above ground (m) and z is the turbine hub height (m). The exponent a can be termed as shear parameter and depends on several factors such as atmospheric stability and surface roughness. According to the literature, for neutral stable condition, a is approximately 0.143, which is commonly assumed to be constant in wind resource assessments (Pishgar-Komleh et al., 2015). The author notes that an extensive study is also required for value of shear parameter in mountainous terrain at high altitude. Because of lack of available data of shear parameter for this region, the author cautiously uses the value estimated in previous study for different terrain. 2.3 Wind Speed Probability Distribution and Parameter Estimators Researches have shown that Weibull function fits the wind probability distribution more accurately compared to others. Weibull distribution is a two parameter function characterized by scale parameter c (m/s) and shape parameter k (dimensionless) and is given by (Ahmed, 2013; Weibull, 1951):
(2)
Shape parameter k and scale parameter c can be calculated using several methods. This study will evaluate following estimators:
2.3.1. Method of Moments (MOM) This technique is commonly used in field of parameter estimation. The method is used as an alternative to the Maximum Likelihood Method which has its associated computational cost. The shape and scale factor can be calculated as (Azad et al., 2014):
(3)
(4)
where is the gamma function.
Г
3 Journal of Engineering Issues and Solutions 1 (1): 1-7 [2021] Parajuli
(5)
2.3.2 Least Square Error Method (LSEM) This method tends to minimize the square of the error in probability plotting. The equation of CDF is linearized with natural log and least square method is employed to estimate the value of k and c:
(6)
(7) 2.3.3 Power Density Method (PDM) This method was suggested by Akdag et al (Akda & Dinler, 2009). The energy pattern factor is ratio of average of cube of velocity to cube of average velocity and k can be estimated using energy pattern factor. Value of c can be estimated using relationship of k and c in Equation 4.
(8)
(9)
2.4 Evaluation of Estimators In order to check how accurately the estimator predicted theoretical probability density fits measured data, in this paper, we employ root mean square error (RMSE) parameter (Azad et al., 2014; Chang, 2010; Ouarda et al., 2015).
(10)
Where N is number of observations, fm is frequency of observation or ith calculated value from measured data, fw is frequency of weibull or ith calculated value from the weibull distribution.
3. Results and Discussion The Weibull parameters k and c were estimated using the three methods employed. Instead of evaluating the parameters of every year and evaluating a large data set, the data for all ten years were considered as a single data set and the parameters were calculated. The same is tabulated in Table 1. The readers interested in parameter trend and seasonal fluctuations of the wind speed at the location can refer to previous article
4 Parajuli Journal of Engineering Issues and Solutions 1 (1): 1-7 [2021] by the author (Parajuli, 2016). The probability densities for the actual wind speed data and the fits are presented in Figure 2. It can be noted that all three methods underestimated the peak frequency. However, the underestimation of the peak by Power Density Method (PDM) is larger compared to other two. Further, it can be noted that, for high and low wind speeds, the performance of all three methods are similar. To evaluate the overall error, root mean square error method has been employed in this research and the same is tabulated in last column of Table 1. From the error analysis, we can see that Method of Moment (MoM) is the best estimator for evaluating Weibull Parameter for this site. Similarly, Power Density Method has been noted to have poor performance out of three estimators.
Figure 2: Observed histogram and weibull fits of various method Table 1: Parameters Estimation and RMSE
Method Shape Factor (k) Scale Factor (c) RMSE Method of Moments (MoM) 3.03 6.69 0.391 Least Square Error Method (LSEM) 3.16 6.68 1.036 Power Density Method (PDM) 2.86 6.70 1.457
In the research done by Ahmed et al. (Ahmed, 2013) for Halabja City, Iraq, Power Density Method has been shown to under predict the peak frequency and not favourable method for prediction of Weibull parameters. It is to be noted that the altitude level of Halabja City is 692m. Similarly, Azad et al. (Azad et al., 2014) showed for three sites of Bangladesh, which are below 10m altitude level, that Method of Moments (MOM) is the best estimator for Weibull Function. The previous researches corroborate with the finding of this analysis despite the differences in the altitude and terrain. In the previous study done by the author at the same site, Method of Moments (MoM) was used to estimate the Weibull parameters. Therefore, the reader interested in power estimation and wind speed characteristics of the site can refer to the article (Parajuli, 2016).
4. Conclusion Based on the wind speed data from Jumla, Nepal which is located at around 2300m above sea level, it has been found that Method of Moments (MOM) is the best estimator for prediction of Weibull Parameters. The finding is similar to studies done by other researchers at low altitude. To further strengthen the claim, the future studies shall be done on multiple locations at similar altitudes. Similarly, the variation of shear parameter shall also be duly noted in future studies.
5 Journal of Engineering Issues and Solutions 1 (1): 1-7 [2021] Parajuli
Acknowledgements The authors would like to thank the Department of Hydrology and Meteorology for providing wind speed data of the site.
Conflict of Interests Not declared by authors.
References Abbas, K., Alamgir, K., Ali, A., Khan, D., & Khalil, U. (2012). Statistical analysis of wind speed data in Pakistan. World Applied Sciences Journal. Ahmed, S. A. (2013). Comparative study of four methods for estimating Weibull parameters for Halabja , Iraq. International Journal of Physical Sciences, 8(5), 186–192. https://doi.org/10.5897/IJPS12.697 Akda, S. a., & Dinler, A. (2009). A new method to estimate Weibull parameters for wind energy applications. Energy Conversion and Management, 50(7), 1761–1766. https://doi.org/10.1016/j.enconman.2009.03.020 Azad, A., Rasul, M., & Yusaf, T. (2014). Statistical Diagnosis of the Best Weibull Methods for Wind Power Assessment for Agricultural Applications. Energies, 7(5), 3056–3085. https://doi.org/10.3390/en7053056 Baseer, M. A., Meyer, J. P., Rehman, S., & Alam, M. M. (2017). Wind power characteristics of seven data collection sites in Jubail, Saudi Arabia using Weibull parameters. In Renewable Energy (Vol. 102). https://doi.org/10.1016/j.renene.2016.10.040 Chang, T. P. (2010). Wind Speed and Power Density Analyses Based on Mixture Weibull and Maximum Entropy Distributions. In International Journal of Applied Science and Engineering (Vol. 8, Issue 1, pp. 39–46). Chaurasiya, P. K., Ahmed, S., & Warudkar, V. (2018). Study of different parameters estimation methods of Weibull distribution to determine wind power density using ground based Doppler SODAR instrument. Alexandria Engineering Journal, 57(4), 2299–2311. https://doi.org/10.1016/j.aej.2017.08.008 Deep, S., Sarkar, A., Ghawat, M., & Rajak, M. K. (2020). Estimation of the wind energy potential for coastal locations in India using the Weibull model. Renewable Energy, 161, 319–339. https://doi.org/10.1016/j.renene.2020.07.054 Guarienti, J. A., Kaufmann Almeida, A., Menegati Neto, A., de Oliveira Ferreira, A. R., Ottonelli, J. P., & Kaufmann de Almeida, I. (2020). Performance analysis of numerical methods for determining Weibull distribution parameters applied to wind speed in Mato Grosso do Sul, Brazil. Sustainable Energy Technologies and Assessments, 42. https://doi.org/10.1016/j. seta.2020.100854 Hulio, Z. H., Jiang, W., & Rehman, S. (2019). Techno - Economic assessment of wind power potential of Hawke’s Bay using Weibull parameter: A review. Energy Strategy Reviews, 26. https://doi.org/10.1016/j.esr.2019.100375 Katinas, V., Mar iukaitis, M., Gecevi ius, G., & Markevi ius, A. (2017). Statistical analysis of wind characteristics based on Weibull methods for estimation of power generation in Lithuania. Renewable Energy, 113, 190–201. https://doi. org/10.1016/j.renene.2017.05.071č č č Odo, F. C., Offiah, S. U., & Ugwuoke, P. E. (2012). Weibull distribution-based model for prediction of wind potential in Enugu , Nigeria. Advances in Applied Science Research, 3(2), 1202–1208. Oner, Y., Ozcira, S., Bekiroglu, N., & Senol, I. (2013). A comparative analysis of wind power density prediction methods for Canakkale, Intepe region, Turkey. Renewable and Sustainable Energy Reviews, 23, 491–502. https://doi.org/10.1016/j. rser.2013.01.052 Ouarda, T. B. M. J., Charron, C., Shin, J.-Y., Marpu, P. R., Al-Mandoos, A. H., Al-Tamimi, M. H., Ghedira, H., & Al Hosary, T. N. (2015). Probability distributions of wind speed in the UAE. Energy Conversion and Management, 93, 414–434. https://doi.org/10.1016/j.enconman.2015.01.036 Oyedepo, S. O., Adaramola, M. S., & Paul, S. S. (2012). Analysis of wind speed data and wind energy potential in three selected locations in south-east Nigeria. Int. J. Energy Environ. Eng., 3(1), 7. https://doi.org/10.1186/2251-6832-3-7 Parajuli, A. (2016). A Statistical Analysis of Wind Speed and Power Density Based on Weibull and Rayleigh Models of Jumla, Nepal. Energy and Power Engineering, 08(07), 271–282. https://doi.org/10.4236/epe.2016.87026 Pishgar-Komleh, S. H., Keyhani, A., & Sefeedpari, P. (2015). Wind speed and power density analysis based on Weibull and Rayleigh distributions (a case study: Firouzkooh county of Iran). Renewable and Sustainable Energy Reviews, 42(0),
6 Parajuli Journal of Engineering Issues and Solutions 1 (1): 1-7 [2021]
313–322. https://doi.org/http://dx.doi.org/10.1016/j.rser.2014.10.028 Rehman, S., Halawani, T. O., & Husain, T. (1994). Weibull parameters for wind speed distribution in Saudi Arabia. Solar Energy, 53(6), 473–479. https://doi.org/10.1016/0038-092X(94)90126-M Saeed, M. A., Ahmed, Z., Yang, J., & Zhang, W. (2020). An optimal approach of wind power assessment using Chebyshev metric for determining the Weibull distribution parameters. Sustainable Energy Technologies and Assessments, 37. https:// doi.org/10.1016/j.seta.2019.100612 Shoaib, M., Siddiqui, I., Amir, Y. M., & Rehman, S. U. (2017). Evaluation of wind power potential in Baburband (Pakistan) using Weibull distribution function. Renewable and Sustainable Energy Reviews, 70, 1343–1351. https://doi.org/10.1016/j. rser.2016.12.037 Tiam Kapen, P., Jeutho Gouajio, M., & Yemélé, D. (2020). Analysis and efficient comparison of ten numerical methods in estimating Weibull parameters for wind energy potential: Application to the city of Bafoussam, Cameroon. Renewable Energy, 159, 1188–1198. https://doi.org/10.1016/j.renene.2020.05.185 Wais, P. (2017). Two and three-parameter Weibull distribution in available wind power analysis. Renewable Energy, 103, 15–29. https://doi.org/10.1016/j.renene.2016.10.041 Weibull, W. (1951). A Statistical Distribution Function of Wide Applicability. In Journal of Applied Mechanics (Vol. 103).
7 Journal of Engineering Issues and Solutions
Assessment of the impact of COVID-19 on transportation and its inter-linked sectors of Nepal
Dipendra Bahadur Singh1,*, Deepak Kumar Sah1 1 Khwopa Engineering College, Liwali, Bhakatpur, Nepal *Corresponding email: [email protected] Received: January 01, 2021; Revised: February 20, 2021; Accepted: March 01, 2021
Abstract Nepal being a landlocked country is completely dependent on the roadways and airways for means of transportation however, the railway has not been started in Nepal to date. Transportation is interlinked with mobility and due to lockdown every sector related to mobility has been affected. Consequently, the aviation sector has been worst hit as airlines were prohibited to operate. The aviation industry of Nepal has foreseen significant decadence in the mobility of passengers and cargo (international and domestic) which has affected revenue generation. Similarly, the imposed lockdown has influenced the ongoing nation's pride and the long-term investment projects which have been considered as a milestone in the infrastructural development of Nepal and those projects need to be reprogrammed and reprioritized. Moreover, the sharp decrease in import of petroleum products has decreased the revenue paid to the Indian Oil Corporation (IOC) which will help to reduce trade loss. The lockdown induced due to COVID-19 has also affected the agricultural sector as the supply chain has been disrupted due to travel restrictions. The overall Gross Domestic Product (GDP) that the transport and its inter-linked sector contribute has been reduced during lockdown than the preceding years. Identifying the paucity of research in the transportation sector of Nepal this paper is focused on the comprehensive study of the impact of the COVID-19 transportation sector along with its inter-connected areas.
Keywords: COVID-19; GDP; lockdown; Nepal; modes of transportation.
1. Introduction Late in December 2019, a novel infectious disease was recognized in the Wuhan province of China which later on was named COVID-19 (Wang et al., 2020). After its outbreak in most of the provinces in China, WHO in January 2020 confirmed it as a communicable disease transferring via respiratory droplets, and on January 30th WHO declared a worldwide public health emergency (Zhou et al., 2020). Spreading in most of the European countries like Italy, France, COVID-19 turns into a pandemic, and to avoid its spread, most of the countries imposed lockdown (Euronews, 2020). As of 16 February 2020 Globally 108,822,960 people have
8 Singh and Sah Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] been infected whereas 2,403,641 have died from this virus (WHO, 2020). In Nepal, till 16 February 272,945 were infected whereas 269,303 people were recovered and 2,055 have died (MOHP, 2020). To stop the spread of infection lockdown was imposed around the world including Nepal. As lockdown has been imposed it has disrupted the mobility of goods and people around the globe ultimately affecting the GDP contributed by the transportation sector. Alike transportation, the construction industry is also affected thus increasing the unemployment rate. Due to the pandemic, the U.S. reported the least number of jobs in April 2020 associated with the construction industry (BLS, 2020). The construction industry which provided 7.7% of employment globally before the pandemic is expected to fall significantly (ILO, 2020). In India, a survey conducted by the Centre for Monitoring Indian Economy (CMIE) revealed that the weekly unemployment rate associated with daily wage laborers dropped from 39% to 26.4 % between March to May of 2020 (Barbate et al., 2021). In the development of transport sector of Nepal where many constraints such as lack of integrated coordination, the need for huge investment, uneven changes in topography were pre-existed, this virus has even worsened the pace of development activities. This study was undertaken to dig-out the repercussion of COVID-19 in transport and its interlinked sectors.
2. Methodology The methodology for this study incorporates four stages: Literature review, Collection of Primary and Secondary data, Analysis and graphical representation of available data, and Impact assessment. During the literature review, previously published journal articles and reports from World Health Organization (WHO), United Nations Development Programme (UNDP), International Labour Organization (ILO), Asian Development Bank (ADB), International Civil Aviation Organization (ICAO), and International Air Transport Association (IATA) have been reviewed and many information has been extracted. After the literature review, it is found that no study has been conducted to analyze the impact of COVID-19 on transportation and its-interlinked sectors of Nepal. So this research aims to fill the literature gap.
2.1. Primary Data Collection For this study, primary data were collected from the official governing bodies like the CAAN (data of international and domestic passengers, international and domestic Cargo and revenue generated), Nepal Oil Corporation (data of petroleum products and revenue paid to IOC). Moreover, primary data related to agriculture (dairy, vegetables, and poultry) and its revenue were extracted from concerned ministries and departments such as the Ministry of Agriculture and Livestock Development (MoALD), Department of Customs (DoC), Central Bureau of Statistics (CBS), and Nepal Rastra Bank (NRB).
2.2. Secondary Data Collection Different newspaper articles and interviews with the concerned personnel from Nepal Oil Corporation (NOC), Department of Roads (DoR), and Tribhuvan International Airport (TIA) have been taken as a secondary source of many facts in this paper.
3. Literature Review Transportation and mobility are interlinked with the economy: from tourism to hospitality, industry to agriculture, all are badly affected thus leading to a significant downfall in the overall economy of the world. In the best-case scenario of the COVID-19 outbreak for 2020, which is characterized as two months of a travel ban and a dramatic decrease in domestic demand, the monetary loss of global GDP is estimated to be around USD 76.7 billion. Moreover, the global GDP is expected to lose around USD 346.98 billion in the
9 Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] Singh and Sah worst-case scenario, described as the six months of the travel ban (Duffin, 2020). Asian Development Bank (ADB) states that if the effect of this pandemic persists for two months in the transportation sector, there will be a sharp decline in domestic demand, and GDP will fall between $77 billion to $347 million (ADB, 2020). On five continents represented by the International Road Transport Union (IRU) and its affiliates, nearly 3.5 million road transport operators are facing an extremely serious financial impact. Likewise, the Asia Pacific region has lost 379 billion revenues from road transport which accounts for a 20% reduction than in 2019. Similarly, Europe, North America, South America, and Eurasia have lost revenue from the transport sector by 20%, 12%, 20%, and 18% respectively (IRU, 2020). The aviation industry contributes 3.6% of the GDP of the world economy and most aviation experts afraid of its downfall (IATA, 2021). Before COVID-19, more than a hundred thousand commercial flights used to be scheduled and around 12 million passengers travel around the globe but this pandemic has completely halted scheduled flights except for evacuation, medical supplies, and emergency. There will be a 35% to 65% decline in passengers around the world in 2020 (IATA, 2021). Alone in Asia and Pacific, there will be a downfall of around 234 to 414 million passengers which accounts for 38% to 70% and revenue of 48 to 86 billion will be lost (ICAO, 2020). Travel restriction has also affected road transport. In Bolivia, there is a 20.7%, 18.8%, 5.37%, and 98.59% of reduction in movement of the passenger by road transport in January, February, March, and April 2020 respectively (Rivera, 2020). Whereas in Spain, as travel restriction was adopted, a 65% of reduction in traffic volume was observed (ETSC, 2020). Likewise, in Greece, the traffic in the first week of March has been decreased by around 10% and till the first week of April, it has been decreased by 75%. Whereas, in the Kingdom of Saudi Arabia till the first week of March, traffic have been plummeted to 20% and to 50% till the first week of April (Katrakazas et al., 2020). As soon as the countries around the world impose lockdown petroleum and oil industries in Saudi Arabia (leading exporter of petroleum products) have faced the steepest one-day price crash nearly in 30 years (Nicola et al., 2020). Whereas in United States (U.S.) crude oil price dipped to negative from $18 per barrel to -$38 per barrel1. International Energy Agency (IEA) reported that the revenue generated from the Oil and Gas will fall between 50% to 85% than in 2019 (IRE, 2020). The transportation and construction sectors in Nepal have an overall contribution of 15% to the national GDP. Furthermore, the construction industry provides13.87% of employment while transportation provides 4.5% of employment but this year the data will fall than in preceding years and (UNDP, 2020).
4. Findings Nepal, a landlocked nation trying to elevate from the status of the Least Developed Country, is highly vulnerable to the emerging COVID-19 pandemic. The effect of the global pandemic is immense and is now taking a significant toll on an economy that depends heavily on the construction and transportation sector, agriculture, and import of petroleum products.
4.1. Road and Construction Most suitable and productive months for construction work in Nepal have been adversely affected by this pandemic and mostly all the construction works are halted. As Nepal already holds a financial gap of 12.77% of GDP (MoF, 2020), the aim of Nepal to graduate from Least Developed Country (LDC) by 2022 and aim to achieve Sustainable development Goals (SDG’s) by 2030 is affected by COVID-19 leading to an expansion in duration. The extension (most probably from 6 months to 1 year) of undergoing nation pride projects such as Kathmandu-Terai/Madesh expressway, Nagdhunga tunnel construction, Kali-Gandaki Corridor, Pushpalal highway, Postal highway, Madan Bhandari highway is likely to occur. However, even
1 Based on news published on The Guardian. (https://www.theguardian.com/world/2020/apr/20/oil-prices-sink-to-20-year-low-as-un-sounds-alarm-on-to- covid-19-relief-fund ).
10 Singh and Sah Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] during pandemic adopting the full safety and precaution in the construction site and using limited materials and manpower, construction of these projects were continued2. Despite the lockdown Department of Road (DoR) blacktopped 77 Km roads in Kathmandu Valley and underlaid 121.6 Km of gravel road. In addition to that, the Department of Roads opened 36.75 km of new road, and maintenance of 52.30 Km of the road was conducted across the country. During the lockdown, one-third of the total (about 2000) projects targeted by the Department of Roads were under construction and 21 bridges were constructed successfully during the lockdown period3. In 2018, the construction sector of Nepal recruited 978,000 employees and has been increasing at an average rate of 9% per year (CBS, 2019). Thus it can be estimated that the pre- pandemic employment figures may have been around 1.2 million. Moreover in between February to May around 500,000 contractual and Indian workers are employed in the construction industry. Based on this data nearly 1.7 million people were employed in the construction sector when lockdown started. Among 1.7 million, 300,000 were hired on regular payrolls and 5% of them were hired by contractors that received the government permission to run on even during the lockdown and remaining 95% have lost the job during lockdown. A total of 1.4 million people were unemployed in the construction sector during lockdown (Prasai S., 2020).
4.2. Aviation Due to COVID-19, in the fiscal year 2075/76, The Civil Aviation Authority of Nepal has experienced a reduction of 42.5% on its targeted revenue of 8.33 billion. As of November 2020, domestic and international passengers have been declined by 65% and 72% respectively than in 2019. Due to this reduction in movement of passengers (domestic and international) CAAN has lost 5.96 billion in revenue which accounts for a reduction by 80% than in 2019. Furthermore, the fund worth of 2.21 billion which was targeted for Airport development was reduced by 37.55%. (CAAN, 2020).
4.2.1. International passengers and revenue collected From January 2017 to October 2018, the outgoing number of international passengers has increased by 17.55% and the incoming number of international passengers has been increased by 54.49%. But after October 2018 to July 2019 the number of outgoing passengers has been decreased by 13.39% whereas incoming passenger has been decreased by 18.58%. From July 2019 to November 2019, outgoing international passengers were increased by 49.54% and incoming international passengers were increased by 22.06% (Fig. 1)). Whereas the number of incoming and outgoing international passengers between March to July 2020 i.e. during lockdown time has been decreased (Fig. 2). Till April there is no movement of the incoming international passenger whereas 18,178 more passengers were recorded in July than in May. Moreover, the outgoing number of international passengers has fluctuated which is because of evacuation flights operated from different Embassies for respective citizens. From March to June 4,553 outgoing passengers were recorded which has been decreased by 3,218 in July 2020. The revenue collected in March 2020 has been decreased by 52.20% than in March 20191 and the revenue collected in July 2020 has been decreased by 96.85% than in July 2019.
4.2.2. Domestic passengers and revenue collected From 2017 to March 2020 the incoming domestic passenger has been increased by 34.56% and the outgoing passenger has been increased by 52.68%. After March 2020 there has been a drastic decrease in the movement
2 Based on articles published on different newspaper. (https://thehimalayantimes.com/nepal/nagdhunga-tunnel-construction-begins-in-full-swing-amid-covid- 19-crisis ), (https://www.newbusinessage.com/Articles/view/12802 ). 3 Based on interview published on The Himalayan Times with Er. Keshab Sharma, Director General, Department of Road. (https://thehimalayantimes.com/ business/department-of-roads-blacktops-77-kilometre-road-during-lockdown/ ).
11 Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] Singh and Sah of domestic passengers. The number of outgoing domestic passengers between February and March has been decreased by 23.21%, and incoming domestic passengers in the same period have been decreased by 27.02% (Fig. 3). Due to the decrease in movement of domestic passengers the revenue generated has also been decreased by 91.77%. After 24th March, there is no movement of domestic passengers as the complete lockdown was imposed and because of which no revenue was generated during that period. The revenue collected in March 2020 has been decreased by 72.43% than in March 2019 and the revenue collected in July 2020 has been decreased by 98.89% than in July 2019.
Figure 1: Movement (in and out) of international passengers before lockdown. (Source: CAAN, 2020)
Figure 2: Movement (in and out) of the international passenger during the lockdown. (Source: CAAN, 2020)
12 Singh and Sah Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021]
Figure 3: Movement of domestic passengers before lockdown (till March) and during lockdown (Till July). (Source: CAAN, 2020)
Figure 4: Movement of international cargo from January 2017 to March 2020 (before lockdown) (Source: CAAN, 2020)
4.2.3. International cargo and revenue collected From January 2017 to September 2018 the outgoing cargo has been increased by 21.94% and incoming cargo has been increased by 46.77%. As per the data available from CAAN no incoming cargo was noticed during February, March, and October of 2019. International cargo movement from April 2019 to March 2020 has been decreased by 52.62% whereas the incoming cargo in the same period has been increased by 15.43% (Fig. 4)). The movement of international cargo during the lockdown period i.e. from April 2020 to July 2020 has been fluctuated (Fig. 5). The incoming cargo movement (in and out) from April to June has been increased
13 Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] Singh and Sah whereas in July it has been decreased. The incoming cargo from April to June has been increased by 155.9 tons which again has been decreased by 68.2 tons in July 2020. Whereas the outgoing cargo from April to June has been increased by 317.387 tons and it has been decreased by 143.55 tons in July 2020. The revenue collected from the international cargo in March 2020 has been decreased by 57.47% than in March 2019 and has been decreased by 32.44% in July 2020 than in July 2019.
Figure 5: Movement (in and out) of the international cargo during the lockdown. (Source: CAAN, 2020)
Figure 6: Domestic cargo movement (in and out) before lockdown. (Source: CAAN, 2020)
4.2.4. Domestic cargo and revenue collected The impact of lockdown has affected domestic cargo movement as it is significantly decreased (Fig. 6). As per the analysis, the movement of outgoing domestic cargo from 2017 to 2018 has been decreased by 2.51%
14 Singh and Sah Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] whereas incoming cargo in the same period has been increased by 9.29%. In January 2019 the outgoing cargo has been increased by 10.82% and incoming cargo has been increased by 14.14%. Whereas, in March 2020 outgoing cargo has been decreased by 36.52% than in March 2019 (Fig. 6). Due to the lockdown, the revenue generated from the movement of domestic cargo in March 2020 has been decreased by 94.95% than in March 2019. During the lockdown (Fig. 7), the incoming cargo has been increased by 15,593 tons which have fallen in July by 68.15 tons. Likewise, the outgoing cargo also has been increased by 317.38 tons and in July it has been decreased by 143.13 tons. Due to the fluctuation in the movement of domestic cargo the revenue generated in July 2020 has been decreased by 24.02% than in July 2019.
Figure 7: Domestic cargo movement (in and out) during lockdown. (Source: CAAN, 2020)
Figure 8: Trend of revenue (in NPR) paid to IOC to import petrol (in Kilo liters, KL). (Source: Nepal Oil Corporation)
15 Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] Singh and Sah
4.3. Petroleum and Oil Industry 4.3.1. Petrol The effect of lockdown can be easily noticed in the import of petrol and revenue paid to the Indian Oil Corporation (IOC) (Fig. 8). As per the data, 137.42% more petrol was imported in FY 2075/76 than in 2072/73. As the import of petrol has been increased, the revenue paid to IOC has also been increased by 22 billion in Fiscal Year (FY) 2075/76. Whereas, due to Pandemic in FY 2076/77 only 513,984 KL of petrol was imported in Nepal which is a decrease by 9.3288% than in FY 2075/76. The decrease in import of petrol has also decreased the revenue paid to IOC in FY 2076/77 to 6 billion which accounts for the decrease of 20.04%. If the Covid-19 would not have persisted, with the on-going increasing trend in previous years, the import in FY 2076/77 would have been nearly about 13.38% more, and revenue to be paid to IOC would have been 15.35% more than in 2075/76.
4.3.2. Diesel The import of diesel has been increasing steadily from Fiscal Year 2072/73 to 2075/76 (Fig. 9). But the import and revenue paid to IOC have been reduced significantly after 2075/76 which is due to the effect of lockdown. From the quantitative analysis of data, it is found that in FY 2075/76 NOC imported 1,717,517KL of diesel which is 117.42% more than import in 2072/73. The increase in import of diesel has been increased the revenue paid to IOC by 79 billion in FY 2075/76. Whereas due to Pandemic in FY 2076/77 only 1473427.58 KL of diesel was imported in Nepal which is a decrease by 14.21% than in FY 2075/76. The decrease in import of diesel has also decreased the revenue paid by NOC in FY 2076/77 to 30 billion which accounts for the decrease of 26.94%. If the Covid-19 would not have persisted, the import in 2076/77, with the increasing trend, would be nearly about 11.59 % more, and revenue to be paid would have been increased by 15.18% than in 2075/76.
Figure 9: Trend of revenue (in NPR) paid to IOC to import diesel (in Kilo liters, KL). (Source: Nepal Oil Corporation)
4.3.3. Aviation fuel The import and revenue paid have been increasing from FY 2072/73 to 2075/76 (Fig. 10). But after 2075/76
16 Singh and Sah Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] the import of aviation fuel has been significantly reduced which is because of lockdown. From FY 2072/73 to 2075/76 the import of aviation fuel has been increased by 152.267% and because of which revenue paid to IOC has been increased by 307.645 %. While due to the pandemic in FY 2076/77 a decline of 31.25% has been noticed than in FY 2075/76. The drop in the import of aviation fuel has also decreased the revenues paid to IOC by 4 billion in FY 2076/77, which is a decrease of 36.74%. Based on the increasing rate of the previous year, if the pandemic would not have arrived, import in 2076/77 would have been 10.58 % more than in 2075/76. Moreover, the import in FY 2074/75 and 2075/76 is nearly similar i.e. only 1084 Kl more was imported in 2075/76. Despite the import of only 1084 Kl more in 2075/76, the revenue paid to IOC is 2 billion more in 2075/6. This is because the price of aviation fuel was hiked by IOC during the year 2075/764 .
Figure 10 : Trend of revenue (in NPR) paid to IOC to import aviation fuel (in Kilo liters, KL). (Source: Nepal Oil Corporation) 5. Agricultural Sector The food supply chain is the extensive network that interlinks the agriculture system to consumer tables through manufacturing, packing, storage, and distribution, and this extensive network is mobilized only with the different means of transportation. Travel restriction has disabled the movement of laborers, distributors, and supply of agricultural products to market, and because of this, farmers in Nepal are highly affected. In Nepal, the Dairy industry represents 12.4 % of the agriculture sector and contributes to 3.3% of GDP (CBS, 2018). In the first 8 months of FY 2018/19 Nepal produced 1.475 million metric tons of milk and exported dairy items worth $1.71 million to India, China, and Japan (DoC, 2019). Due to travel restrictions and failing consumption dairy industry has to bear a loss of 80% accounting for $30 million and this has impacted the 0.5 million farmers and 10,000 staff involved in the dairy industry (UNDP, 2020). Moreover, In the first nine months of FY 2019/2020, vegetable exports plunged, as Nepal exported vegetables by just $13.3 thousand which was $10.05 million in the same period of FY 2018/2019 (NRB, 2020). Not only vegetables and dairy farming but poultry farming which has been supplying 0.1 million metric tons of meat and 1.8 billion eggs (as of 2019) has been disrupted due to difficulties in transport and declination in consumption. During the lockdown private sector have faced a shortage of chickens, eggs and poultry feeder which have resulted to a major decrease in sales by 80% and 50% chickens and eggs respectively (MoALD, 2020).
4 Based on information interpreted by Hari Dutta Joshi, Deputy Manager-Finance Department, and NOC.
17 Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021] Singh and Sah
6. Conclusions The transportation industry, one of the good contributors to the GDP of Nepal, has been affected due to lockdown. Due to disruption in transportation overall loss of the GDP contributed by transport and its inter-related sector (petroleum products, agriculture, and construction) has been plummeted limiting the growth of the economy to 0.2% only. Impact on the transportation sector directly or indirectly affected the overall development as there have been difficulties in the mobility of manpower and construction materials. In Nepal construction of various projects targeted by the Department of Roads has been affected however with the limited supply of construction material one-third of construction work was continued. Moreover, the aviation industry of Nepal is the hardest hit due to the pandemic. The movement of both international and domestic flights, passengers, and cargo has been drastically declined as a result of which CAAN has to bear a loss of 8.33 billion in revenue collection. Besides this, the import of petroleum products has also been decreased due to imposed lockdown. Undoubtedly the recession in the import of petroleum products will help to compensate the country’s trade loss. To overcome the trade loss due to the import of petroleum products, the governing body could shift the focus from petroleum-based vehicles to electric vehicles which will not only help to lessen the dependence of Nepal on India for petroleum products but will also promote hydropower generation. Apart from this, during the data collection, it is found that the storage, update, and management of data in the governing authority is extremely poor. In the post-pandemic period, transport planning, policy management, and operating facilities need to be assessed and re-evaluated.
Acknowledgements We would like to thanks Er. Hemant Tiwari (Chairman, Safe and Sustainable Travel Nepal – SSTN) for helping us with the collection of data. Moreover, we would like to express our thanks and gratitude to Dr. Prasanna Humagain. (Utah State University, USA) and Er. Shubhechchha Bhatta (Toronto Transit Commission, Canada) for providing us with exclusives comments and reviews.
Conflict of Interests Not declared by authors.
References ADB. (2020). The Economic Impact of the COVID-19 Outbreak on Developing Asia. https://doi.org/10.22617/BRF200096 Barbate, V., Gade, R. N., & Raibagkar, S. S. (2021). COVID-19 and Its Impact on the Indian Economy. Vision: The Journal of Business Perspective, April, 097226292198912. https://doi.org/10.1177/0972262921989126 BLS. (2020). Current Employment Statistics - CES (National) : U.S. Bureau of Labor Statistics. https://www.bls.gov/ces/ CAAN. (2020). CAAN Report (2019-20). In CAAN. CBS. (2018). Central Bureau of Statistics. https://cbs.gov.np/ CBS. (2019). Report on the Nepal Labour Force Survey 2017/18. Central Bureau of Statistics. https://nepalindata.com/media/ resources/items/20/bNLFS-III_Final-Report.pdf DoC. (2019). FTS F/Y 2076/77 - || Department of Customs ||. https://www.customs.gov.np/page/fy-207677 Duffin, E. (2020). • COVID-19: monetary global GDP loss by scenario 2020 | Statista. Statista. https://www.statista.com/ statistics/1102971/covid-19-monetary-global-gdp-loss-scenario/ ETSC. (2020). The impact of COVID-19 lockdowns on road deaths in April 2020. July, 1–21. Euronews. (2020). Coronavirus: European countries tighten controls amid COVID-19 second wave fears | Euronews. https://www. euronews.com/2020/07/28/coronavirus-germany-and-austria-battle-local-outbreaks-as-italy-sets-1-000-face-mask- fine IATA. (2021). Economic Performance of the Airline Industry (Issue April 2009).
18 Singh and Sah Journal of Engineering Issues and Solutions 1 (1): 8-19 [2021]
ICAO. (2020). Effects of Novel Coronavirus (COVID-19) on Civil Aviation: Economic Impact Analysis Air Transport Bureau Contents. June. https://www.icao.int/sustainability/Documents/COVID-19/ICAO_Coronavirus_Econ_Impact.pdf ILO. (2020). ILO Sectoral Brief. 2019(March), 1–8. IRE. (2020). The impact of Coronavirus (COVID-19) and the global oil price shock on the fiscal position of oil-exporting developing countries. 1–18. https://www.oecd-ilibrary.org/development/development-assistance-committee-members-and-civil- society_51eb6df1-en IRU. (2020). COVID-19 impacts on commercial road transport. June, 7. Katrakazas, C., Michelaraki, E., Sekadakis, M., & Yannis, G. (2020). A descriptive analysis of the effect of the COVID-19 pandemic on driving behavior and road safety. Transportation Research Interdisciplinary Perspectives, 7, 100186. https:// doi.org/10.1016/j.trip.2020.100186 MoALD. (2020). Ministry of Agriculture and Livestock Development. https://www.moald.gov.np/publication/Agriculture Statistics MoF. (2020). Economic Survey 2019 / 20 Government of Nepal Ministry of Finance Singh Durbar , Kathmandu. 1–144. MOHP. (2020). CoVid19-Dashboard. https://covid19.mohp.gov.np/ Nicola, M., Alsafi, Z., Sohrabi, C., Kerwan, A., & Al-jabir, A. (2020). Since January 2020 Elsevier has created a COVID-19 resource center with free information in English and Mandarin on the novel coronavirus COVID- 19 . The COVID-19 resource center is hosted on Elsevier Connect , the company ’ s public news and information. January. NRB. (2020). Monthly Statistics Archives - Page 2 of 12. https://www.nrb.org.np/category/monthly-statistics/ page/2/?department=bfr Prasai S. (2020). the Impact of Covid-19 Lockdown on Nepal ’ S Construction Sector : May. Rivera, A. (2020). The impact of COVID-19 on transport and logistics connectivity in the landlocked countries of South America Thank you for your interest in this ECLAC publication. UNDP. (2020). RAPID ASSESSMENT OF SOCIO ECONOMIC IMPACT OF COVID-19 IN NEPAL. Wang, Y., Wang, Y., Chen, Y., & Qin, Q. (2020). Unique epidemiological and clinical features of the emerging 2019 novel coronavirus pneumonia (COVID-19) implicate special control measures. Journal of Medical Virology, 92(6), 568–576. https://doi.org/10.1002/jmv.25748 WHO. (2020). WHO Coronavirus Disease (COVID-19) Dashboard | WHO Coronavirus Disease (COVID-19) Dashboard. https:// covid19.who.int/ Zhou, B., She, J., Wang, Y., & Ma, X. (2020). WHO situation report-1 (2019-nCoV). JANUARY, 1–14.
19 Journal of Engineering Issues and Solutions
Shear and tensile bond strengths of autoclaved aerated concrete (AAC) masonry with different mortar mixtures and thicknesses
Raghav Tandon1*, Sanjeev Maharjan1, Suraj Gautam1 1 Department of Mechanical and Aerospace Engineering, Pulchowk Campus, IoE, Nepal *Corresponding Author email: [email protected] Received: January 11, 2021; Revised: February 26, 2021; Accepted: March 03, 2021
Abstract Autoclaved aerated concrete (AAC) blocks are commonly used for masonry walls. In order to understand the strength of AAC masonry, it is essential to assess the tensile and shear bond strengths of the AAC block-mortar interface for various mortar combinations. This research investigates the bond strength of AAC block mortar interface made up of a) polymer modified mortar (PMM) and b) ordinary cement sand mortar of 1:4 or 1:6 ratio with thickness of 10mm, 15mm or 20mm. A thin cement slurry coating was applied on the block surface before placing the cement sand mortar in the masonry. For all types of interface, shear bond strength of masonry was studied using a triplet test, while the tensile bond strength was determined through a cross- couplet test. Among the cement sand mortar used in this study, cement sand mortar of ratio 1:4 and thickness 15mm showed the maximum shear strength of 0.13MPa with the failure of blocks as the predominant failure while the PMM had shear bond strength of 0.12MPa with the failure of blocks as the predominant failure type. However, in case of the tensile bond strength testing, PMM showed the tensile bond strength of 0.19MPa, which was highest among all the test specimens used in this study. Considering both the tensile and shear bond strengths of the AAC masonry and based on the observed failure pattern, among all the combinations used in the experiment, either PMM or cement-sand mortar of ratio 1:4 and thickness of 15mm can be chosen for the AAC masonry.
Keywords: AAC blocks, cement sand mortar, PMM, failure pattern, polymer modified mortar, shear bond strength, tensile bond strength.
1. Introduction Autoclaved Aerated Concrete (AAC) block masonry is one of the most widely used construction materials for the residential and contemporary building considering its unique thermal properties, low density and high fire resistance (Andlsun, 2006; Radhi, 2011). It has been evolving as a potential alternative to the clay as well as fly ash bricks. There has been a successful history of the use of AAC blocks in different types of environments for all types of building (Wittmann et al., 1983; Concrete & Wittmann, 1992). Similarly, the availability of blocks in large sizes makes the construction works of AAC blocks masonry easy and rapid.
20 Tandon et al. Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021]
The preparation of AAC is possible through the wide range of cementitious materials; however, in common, Portland cement, fly ash and sand are used. Hamad (2014) suggested the addition of sand can contribute to achieve adequate fineness. Besides, a small amount of aluminum powder is also added in the mix to give the cellular structure of the block and on varying the amount of aluminum powder changes the density of the final block (Aroni et al., 1993; Fudge et al., 2019). AAC possesses porous structure with lightness and insulation properties due to the presence of aluminum paste in the composition; thereby making it a substantially different product as compared to the other light weight concrete materials (Aroni et al., 1993).
The compressive strength of AAC ranges from 1.5 to 10MPa while its density varies from 300 to 1000kg/m3. The density and porosity of the AAC block determines the compressive strength of the block. Alexanderson (1979) summarized that the increase in porosity and decrease in density results in the decrease of compressive strength. The splitting tensile strength tests was carried out and the failure mechanism were identified by the Ma yszko et al. (2017). For the adequate bond strength, there should be sufficient amount of cementitious material at the interface between the blocks. Different types of mortar joints such as cement-sand mortar and PMMł are used. For instance, a thin layer (2-4mm) of PMM has been used in constructing AAC masonry (Thamboo et al., 2013; Thamboo & Dhanasekar, 2015). Thamboo & Dhanasekar (2015) worked on the concrete masonry using thin layer of polymer-based mortar of thickness 2mm. Ferretti et al. (2015) used thin cementitious gray glue joints of 1.5mm thickness in the AAC Masonry and studied the compressive and flexural strengths of the AAC masonry. Mallikarjuna (2017) studied the bond strength of AAC masonry using thick sand-cement mortar joints. Similarly, Ferretti et al. (2015) investigated the compressive and flexural strength of AAC masonry focusing the thin glue joints of thickness 0.5 to 1mm, neglecting the effect of joint strength on the overall performance of AAC masonry. Bhosale et al. (2019) examined the bond strengths and compressive strengths of AAC masonry using polymer-based mortar of 2-5mm thickness. Generally, in practice, cement-sand mortar thickness varies from 10-18mm (IS:2250-1981 Reaffiremed 2000, 1981). However, little research exists on the optimum thickness of the cement-sand mortar joint in AAC- block masonry. The aim of this research is to identify the bond strength of the AAC masonry by using 1:4 and 1:6 cement sand mortar mix ratios with various thicknesses of 10mm, 15mm and 20mm.
2. Materials and Methods 2.1 AAC Blocks In this study, 108 AAC eco-blocks of dimension 600mm x 200mm x 100mm of a single lot were collected from a local industry. The specimens were brought to the Central Material Testing Laboratory of Institute of Engineering, Tribhuvan University for testing. Three blocks were tested for compressive strength, 63 blocks were tested for shear strength and 42 blocks were tested for tensile strength. 2.2 Joint Materials Before starting the evaluation of shear and tensile bond strengths of AAC masonry, properties of cement, sand, and AAC blocks used in the test were determined. Vicat apparatus with a 10mm diameter plunger was used to determine the normal consistency of cement paste in accordance with IS 4031 - 4 (2005). Similarly, particle size distribution (grading) of sand was analyzed in accordance with IS 2386- Part I (1963). For the study of the bond strength of AAC masonry, two types of joint materials were used in our study: they are, PMM and cement-sand mortar (CSM). PMM is the composite prepared by using polymer with cement and aggregates. A thin layer of PMM with thickness of 2-3mm is generally used in AAC block masonry (Thamboo & Dhanasekar, 2015). In this study, PMM was prepared by adding 300ml of water to 1kg of dry mortar mix.
21 Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021] Tandon et al.
Cement sand mortar was prepared with two ratios of 1:4 and 1:6. For each cement sand mortar mix, the thickness was varied as 10mm, 15mm and 20mm. It was then applied on the AAC block surface to study the bond strength. Cement-water slurry was initially applied on the block surface before applying the cement sand mortar as suggested by Raj et al. (2020). Compressive strength of cement sand mortar of ratio 1:4 and 1:6 and PMM was determined in accordance with IS:2250 (1981 Reaffirmed 2000, 1981).
2.3 Methods The overall study was carried out to investigate the bond strength of AAC masonry with regards to the PMM mortar with 3mm thickness and cement sand mortar ratios of 1:4 and1:6 with varying thickness of 10mm, 15mm and 20mm. The overall method can be represented in the Fig. 1.
Figure 1: Overall flow of the study
2.3.1 Properties of AAC blocks The physical properties like bulk density, and moisture content of the AAC blocks were investigated in accordance with IS 6441 (2001). The testing procedure for the bond strength were carried out as per ASTM (1991).
22 Tandon et al. Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021]
The test for compressive strength of AAC blocks was carried out from the blocks which were used to test the bulk density and moisture content. The test was carried out in accordance to IS 6441 (2001) and 3 sample blocks were used for the test. The samples were cut into three equal pieces such that each cut piece had the dimension of 200mm x 200mm x 100mm. Then the 9 pieces of AAC obtained from 3 samples were tested for compressive strength. Compressive strength for each piece was obtained by dividing the peak load with its area normal to the load.
2.3.2 Test for tensile bond strength of AAC masonry The cross-couplet specimen was prepared using AAC blocks and mortar bed joints. The specimen preparation and the testing procedure for the tensile bond strength were carried out as per ASTM (1991). The test was carried out in accordance with the procedures followed by Alecci et al. (2013) and Mallikarjuna (2017) as shown in Fig. 2.
Figure 2: Setup for AAC tensile bond strength Using a cross-couplet test, tensile bond strength of AAC block and mortar interface was determined as shown in Fig. 2. The tensile bond strength was computed corresponding to the peak load at failure which is given by:
(1)
Where,
Pmax Peak load recorded at failure A Contact area of the joint where is the tensile bond strength, (Pt)max is the peak load recorded at failure and A is the contact area between the two blocks joined by mortar layer. The failure of the block-mortar interface can take place in any of the following four patterns: complete block-mortar interface failure (Type I), partial block-mortar interface failure (Type II), partial tensile failure of the block (Type III), complete tensile failure of block (Type IV).
23 Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021] Tandon et al.
2.3.3 Test for shear bond strength of AAC masonry Using a triplet test, the shear bond strength of the AAC block and mortar interface was determined as shown in Fig. 3.
Figure 3: Setup of AAC shear strength test The shear bond strength is given by
(2)
Where,
Pmax Peak shear load recorded at failure Ac Contact area of the joint
The failure of the block– mortar interface using triplet test can take place in any one of the following patterns including: failure of block (Type A); failure of mortar (Type-B); and failure of block-mortar interface (Type C).
3. Results and Discussion The physical properties of AAC block (bulk density, moisture content and compressive strength), and the properties of the joint materials were observed initially. Shear and tensile bond strength of the AAC masonry using PMM were determined thereafter. Similarly, results for the AAC masonry with cement sand mortar of ratios 1:4 and 1:6 and of thickness 10mm, 15mm and 20mm were observed as discussed below:
3.1 Physical Properties of AAC Block From the experiment, average value of bulk density of the AAC blocks was 510kg/m3 as shown in Table 1. Similarly, moisture content of the AAC blocks were observed to be 37.17% as shown in Table 2. During the compressive strength test, the compressive load was applied with a loading rate of 0.05 - 0.196N/mm2 until the sample couldn’t take more load. Thus, the average compressive strength of AAC block samples was observed to be 3.19MPa as shown in Table 3.
24 Tandon et al. Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021]
Table 1: Bulk density of AAC blocks
Weight before Length of Breadth of Thickness of Volume of Weight after Bulk density drying (kg) block (cm) block (cm) block (cm) block (cm3) drying (kg) (g/cm3) 8.18 59.93 19.80 9.83 11669.09 5.98 0.51
Table 2: Moisture content of AAC blocks
Weight of block before drying- Weight of block after drying- Moisture content F (%)
W1(kg) W(kg) 8.18 5.98 37.17
Table 3: Compressive strength of AAC blocks (average of 9 tests)
Weight of Area of block Thickness of Ultimate load (L)- Compressive strength block (Kg) (mm2) block (mm) KN (N/mm2) 2.08 39138.00 99.33 125.00 3.19
3.2 Determination of the Properties of Joint Materials From the experiment the normal consistency of cement was observed to be 29% (Fig. 4a). Similarly, from the sieve analysis, the fineness modulus of sand was found to be 2.74 (Fig. 4b) which means the average size of particles of the fine aggregate was between 0.3 to 0.6mm and falls under the limit of sand used in mortar as per BIS (2116) .
(a) (b) (c) (d)
Figure 4: Properties of joint materials: (a)normal consistency of cement, (b)sieve analysis of sand, (c) mortar cube samples, (d)compressive test of an AAC block The compressive strength of the cement sand mortar of ratio 1:4 and 1:6 used in the experiment (Fig. 4c) was observed to be 14.97N/mm2 and 8.67N/mm2, respectively, while the PMM had the compressive strength of 11.56N/mm2 as shown in Table 4.
25 Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021] Tandon et al.
Table 4: Results of compressive strength test of cement sand mortar after 28 days of curing
Weight of Cross- sectional Maximum Mortar Water cement Compressive mortar cube area of mortar applied ratio ratio (w/c) strength (N/mm2) samples (kg) cube sample (mm2) load (N) 1:4 0.78 0.67 4900.00 73333.33 14.97 1:6 0.75 0.91 4900.00 42500.00 8.67 PMM 0.65 0.33 4900.00 56666.67 11.56
3.3 Shear Bond Strength of Masonry Triplet Triplet specimens were prepared and tested as shown in Fig. 5a and Fig. 5b. Three different failure patterns of the triplet specimen were observed during the test. As expected, the joint failure in shear was sudden and brittle. Most of the triplet specimens exhibited the block failure mode. The failure of the block-mortar interface using the triplet test occurred in either of the following patterns: 1. Failure of block (Type A as shown in Fig. 6a, 2. Failure of mortar (Type B as shown in Fig. 6b, 3. Failure of block-mortar interface (Type C as shown in Fig. 6c). From the triplet test results as shown in Table 5, the values of the shear bond strength of AAC masonry using the cement-sand mortar were found to be in the range of 0.06-0.13MPa while the AAC masonry with PMM had the highest shear bond strength of value 0.12MPa. For the cement sand mortar mix of 1:6, majority of the failure pattern exhibited was either type B or type C or both. However, in case of cement sand mortar mix of 1:4 ratio, the joint with 15mm mortar thickness exhibited the highest shear bond strength of 0.13MPa with failure type A being pre-dominant. Hence, cement-sand mortar of ratio 1:4 with mortar joint thickness of 15mm seems to be the best option for the shear bond strength among all the mortar joint samples used in this study. Table 5: Results from the triplet test of AAC masonry
Cross Thickness Load Shear bond Mortar sectional area Failure type (mm) (Kg) strength (N/mm2) (mm2) 10 59000.67 1073.33 0.09 1 in Type A, 2 in Type C 1:4 15 59090.45 1510.00 0.13 2 in Type A, 1 in Type C 20 58856.89 750.00 0.06 1 in Type A, 2 in Type C 10 59256.33 1076.67 0.09 2 in Type B, 1 in Type C 1:6 15 59234.44 1013.33 0.09 1 in Type A, 2 in Type B 20 58934.33 976.67 0.08 2 in Type B, 1 in Type C PMM 2-3 59278.33 1385.00 0.12 2 in Type A, 1 in Type C
26 Tandon et al. Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021]
(a) (b) Figure 5: AAC triplet sample: (a) preparation of triplet sample, (b) triplet test
(a) (b) (c)
Figure 6: Different failure patterns of AAC triplet specimen: (a) failure of block (type-A), (b) failure of mortar (type-B), (c) failure of block-mortar interface (type-C)
3.4 Tensile Bond Strength of Masonry Cross-Couplet The cross-couplet specimens were prepared and tested as shown in Fig. 7a and Fig. 7b, respectively. The failure patterns observed during the test are shown in Fig. 8a and Fig. 8b. The joint failure in tension was sudden and brittle. The failure of the cross-couplet specimens occurred in either of the following four patterns: 1. Complete block-mortar interface failure (Type I), 2. Partial block-mortar interface failure (Type II), 3. Partial tensile failure of the block (Type III),
27 Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021] Tandon et al.
4. Complete tensile failure of block (Type IV).
(a) (b)
Figure 7: AAC cross-couplet specimen: (a) cross-couplet samples, (b) tensile bond strength test of a cross- couplet specimen Results from the cross-couplet test (as shown in Fig. 7a and Fig. 7b) are presented in Table 6. Tensile bond strength of AAC block masonry were found in the range of 0.02- 0.19MPa. Masonry from PMM had the tensile strength of 0.19MPa with predominant Type IV failure. However, the tensile bond strength of the cross-couplet specimens using cement-sand mortar of ratio 1:6 with 20mm thickness predominantly exhibited Type II failure. All other combinations of cement sand mortar showed the Type IV failure. Thus, any of the studied cement-sand mortar combinations shown in Table 6 can be used for AAC block masonry with the exception of 1:6 mortar ratio of 20mm thickness. Most of the AAC masonry showed Type IV failure suggesting that the tensile strength of the AAC masonry joint is higher than the tensile strength of the block itself. Thus, all the possible mortar combinations (except cement sand mortar 1:6, 20mm thickness) can be recommended. Table 6: Result of cross-couplet test of AAC masonry
Cross Thickness Load Tensile bond Mortar sectional Failure mode (mm) (Kg) strength (N/mm2) area (mm2) 10 33000.00 96.67 0.03 3 in Type IV 1:4 15 36666.67 140.00 0.04 3 in Type IV 20 34833.33 116.67 0.03 3 in Type IV 10 34866.67 73.33 0.02 3 in Type IV 1:6 15 33600.00 106.67 0.03 3 in Type IV 20 34233.33 83.33 0.02 2 in Type II, 1 in Type IV PMM 2-3 26812.00 170.00 0.19 3 in Type IV Partial interface failure (Type II as shown in Fig. 8a) was mainly observed using the 1:6 mortar of joint thickness 20mm. In this type of failure, a portion of either block or mortar gets stuck to each other. In case of complete tensile failure of block (Type IV as shown in Fig. 8b) the block completely failed in tension and the joint remained intact. This type of failure occurs when the bond strength of block-mortar interface
28 Tandon et al. Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021] exceeds the tensile strength of the block. The failure pattern of type (IV) was observed mainly using PMM mortar, 1:4 mortar of all joint thicknesses and 1:6 mortar with joint thicknesses of 10mm and 15mm.
(a) (b)
Figure 8: Different failure patterns of AAC cross-couplet specimens: (a) partial block-mortar interface failure (Type-II), (b) complete tensile failure of block (Type IV)
3.4 Comparison of Bond Strengths From Table 5 and Table 6, both shear and tensile bond strengths of the AAC masonry using cement-sand mortar was less compared to the shear and tensile bond strength of PMM. Although the 1:6 mortar had low shear bond strength, its tensile bond strength for joint thickness of 15mm was similar as compared to the tensile bond strength of 1:4 mortar. From the experiment, either PMM or cement sand mortar ratio of 1:4 with thickness of 15mm was found the best for shear bond strength as compared to other combinations. However, in case of tensile bond strength, all the combinations (except cement sand mortar of ratio 1:6 & 20mm thickness) were found to be satisfactory. 4. Conclusions This study investigated the shear and tensile bond strengths of AAC masonry using triplet and cross- couplet specimen. In order to study the masonry bond strength, the AAC masonry was constructed using either ordinary sand-cement mortar or PMM in combination with cement slurry coating. The following are the main findings: • 1:4 mortar mix of thickness 15mm showed the maximum shear bond strength of 0.13MPa while the PMM mortar showed it to be 0.12MPa. In both of these mortar mixes, failure of blocks was the predominant failure type. • PMM mortar showed the tensile bond strength of 0.19MPa which was the highest among all types of mortar mix. Among the cement sand mortars, 1:4 mortar mix with 15mm thickness showed the highest tensile bond strength of 0.04MPa. • Considering both the tensile and shear bond strengths of the AAC masonry as well as the failure patterns of all the combinations used in this experiment, either PMM or cement-sand mortar of ratio 1:4 and thickness of 15mm can be used for the AAC masonry.
29 Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021] Tandon et al.
Acknowledgements The authors would like to thank the Central Material Testing Laboratory of Institute of Engineering, Tribhuvan University for providing the platform to conduct this research.
Conflict of Interests Not declared by authors.
References Alecci, V., Fagone, M., Rotunno, T., & De Stefano, M. (2013). Shear strength of brick masonry walls assembled with different types of mortar. Construction and Building Materials. https://doi.org/10.1016/j.conbuildmat.2012.11.107 Alexanderson, J. (1979). Relations between structure and mechanical properties of autoclaved aerated concrete. Cement and Concrete Research. https://doi.org/10.1016/0008-8846(79)90049-8 Andlsun, S. (2006). A study on material properties of autoclaved aerated Concrete (AAC) and its contemporary and historical wall sections. In Middle East Technical University. Aroni, S., 78-MCA., R. T. C., & 51-ALC., R. T. C. (1993). Autoclaved aerated concrete : properties, testing, and design : RILEM recommended practice. E & FN Spon. https://search.ebscohost.com/login. aspx?direct=true&scope=site&db=nlebk&db=nlabk&AN=105893 ASTM. (1991). American standard test method for bond strength of mortar to masonry units, ASTM C 952-91. Bhosale, A., Zade, N. P., Davis, R., & Sarkar, P. (2019). Experimental Investigation of Autoclaved Aerated Concrete Masonry. Journal of Materials in Civil Engineering. https://doi.org/10.1061/(asce)mt.1943-5533.0002762 BIS 2116. (1980). Specification for Sand for masonry mortars (1st revision). Bureau of Indian Standards, New Delhi, India. Concrete, R. I. S. on A. A., & Wittmann, F. H. (1992). Advances in autoclaved aerated concrete : proceedings of the 3rd international symposium on autoclaved aerated concrete, Zrich, Switzerland, 14-16 October 1992. Ferretti, D., Michelini, E., & Rosati, G. (2015). Mechanical characterization of autoclaved aerated concrete masonry subjected to in-plane loading: Experimental investigation and FE modeling. Construction and Building Materials. https://doi. org/10.1016/j.conbuildmat.2015.08.121 Fudge, C., Fouad, F., & Klingner, R. (2019). Autoclaved aerated concrete. In Developments in the Formulation and Reinforcement of Concrete. https://doi.org/10.1016/B978-0-08-102616-8.00015-0 Hamad, A. J. (2014). Materials, Production, Properties and Application of Aerated Lightweight Concrete: Review. International Journal of Materials Science and Engineering. https://doi.org/10.12720/ijmse.2.2.152-157 IS:2250-1981 Reaffiremed 2000. (1981). CODE OF PRACTICE FOR PREPARATION AND USE OF MASONRY MORTARS. Bureau of Indian Standards, New Delhi, India. IS 2386- Part I. (1963). Method of test for aggregate for concrete. Part I - Particle size and shape. Bureau of Indian Standards, New Delhi, India. IS 4031 - 4. (2005). Methods of Physical Tests for Hydraulic Cement, Part 4: Determination of Consistency of standard cement paste. In Bureau of Indian Standards, New Delhi. IS 6441. (2001). Indian Standard Code of Practice [IS: 6441-1972, Reaffirmed 2001] For testing autoclaved cellular concrete products (Fifth Revision). Mallikarjuna, S. (2017). Experimental determination of parameters for a micro-modeling based failure criterion for AAC block masonry shear wall. Indian Institute of Technology, Guwahati, India. Ma yszko, L., Kowalska, E., & Bilko, P. (2017). Splitting tensile behavior of autoclaved aerated concrete: Comparison of different specimens’ results. Construction and Building Materials. https://doi.org/10.1016/j.conbuildmat.2017.09.167 Radhi,ł H. (2011). Viability of autoclaved aerated concrete walls for the residential sector in the United Arab Emirates. Energy and Buildings. https://doi.org/10.1016/j.enbuild.2011.04.018 Raj, A., Borsaikia, A. C., & Dixit, U. S. (2020). Bond strength of Autoclaved Aerated Concrete (AAC) masonry using various joint materials. Journal of Building Engineering. https://doi.org/10.1016/j.jobe.2019.101039 Thamboo, J. A., & Dhanasekar, M. (2015). Characterisation of thin layer polymer cement mortared concrete masonry bond. Construction and Building Materials. https://doi.org/10.1016/j.conbuildmat.2014.12.098
30 Tandon et al. Journal of Engineering Issues and Solutions 1 (1): 20-31 [2021]
Thamboo, J. A., Dhanasekar, M., & Yan, C. (2013). Flexural and shear bond characteristics of thin layer polymer cement mortared concrete masonry. Construction and Building Materials. https://doi.org/10.1016/j.conbuildmat.2013.04.002 Wittmann, F. H., Structures., I. U. of T. and R. L. for M. and, & Concrete, R. I. S. on A. A. (1983). Autoclaved aerated concrete, moisture and properties.
31 Journal of Engineering Issues and Solutions
Voltage stability analysis of wind power plant integration into transmission network of Nepal
Sagar Dharel1, Rabindra Maharjan2,* 1Advanced College of Engineering and Management, Institute of Engineering, 2Pulchowk Campus, Institute of Engineering, Tribhuvan University, *Corresponding email: [email protected]* Received: January 28, 2021; Revised: February 27, 2021; Accepted: March 03, 2021
Abstract Government of Nepal has realized that wind energy could become a major source of alternative energy to solve energy crisis in the country as well as serve the purpose of energy mix. Various studies have identified several locations with potential for wind power generation in Nepal. The integration of wind power plant to the national grid, however, raises concerns regarding the power system stability. The voltage stability of the grid is a key issue, the effect on which increases with the increase in wind power penetration in the grid. This study performs voltage stability analysis due to high penetration of wind power in Integrated Nepalese Power System (INPS). Both steady state and dynamic stability study is performed using the power system simulation software DigSILENT/PowerFactory for different types of wind turbine generators.
Keywords: Dynamic study, INPS, voltage stability, wind turbine
1. Introduction Wind energy technology has developed and improved a great deal in last two decades with increase in wind turbine sizes but reduction in manufacturing cost (Ackerman, 2012). The reduction in cost has resulted in high penetration of wind energy in the transmission network. For example, wind energy has provided 15 per cent of the EU’s electricity demand in 2019 and constitutes the third largest energy source in China (Lee & Zhao, 2019). According to Global Wind Energy Council (GWEC) report 2019, a total of 651 GW of wind energy capacity has already been installed globally with 60.4 GW of wind energy capacity installed in 2019 (Lee & Zhao, 2019). The global wind statistics 2019 by World Wind Energy Association (WWEA) suggested that there are now 112 countries with grid connected wind turbines, out of which 36 countries have more than 1000 MW installed and, 10 countries with more than 10,000 MW. It is the most installed renewable technology all over the world. (IRENA, 2020) Nepal has high potential of various kind of renewable energy. The Government of Nepal (GoN) has been supporting promotion and development of renewable energy technologies since last two decades. Alternative Energy Promotion Center (AEPC), established by the Government, performed the wind resources assessment
32 Dharel and Maharjan Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] under the Solar and Wind Energy Resource Assessment (SWERA) project. The SWERA report, estimated Nepal’s gross wind power potential to be 3000 MW. In Nepal, wind potential for power generation is favorable in surrounding hills of Kathmandu Valley, Annapurna range, Tansen of Palpa, Lomangthang of Mustang and Khumbu regions of Nepal (SWERA, 2006; 2008). Wind turbine technology is constantly evolving and it can be fixed speed or variable speed type, employed with induction generator, synchronous generator or permanent magnet generator, and with or without power electronics converter. Most of the wind turbine technology employ induction generators, which contribute asynchronous power to the system, and as such, a large penetration of wind generation will impact the stability of the system, particularly the voltage stability of the system (Chi et al., 2006; Chen et al., 2009). Voltage stability is the ability of a power system to maintain steady acceptable voltages at all buses in the system under normal operating conditions and after being subjected to a disturbance. Voltage instability problems arise in power systems with heavy loads, disturbance or with shortage of reactive power. (Kundur, 1994) The interconnection of induction generator to the grid either demands reactive power from the grid or is provided with less reactive power supply capability in the form of power electronics converters, contributing to voltage instability problems (Pena et al., 1996; Maharjan & Kamalsadan, 2013). This paper studies the impact of wind turbine integration into Integrated Nepalese Power System (INPS). Both steady state voltage stability analysis and transient voltage stability analysis is conducted using simulation tool DigSILENT/ PowerFactory. Power Voltage (P-V) curves and Voltage-Reactive Power (V- Q) curves are obtained for various points of interconnection in the network for steady state analysis. The voltage recovery characteristic of the wind generator after a three-phase fault is studied in transient voltage stability analysis.
2. Wind Energy Technology A variety of wind energy conversion systems (WECS) are being used all over the world with different conversion system topology. The most commonly used wind turbine configuration is classified by their ability to control speed. Fixed speed induction generator (FSIG) or Type A and doubly fed induction generator (DFIG) or type C are the two most common wind turbine installed around the world (Ackerman, 2012; Chen et al., 2009).
2.1 Fixed Speed Induction Generator (FSIG) Fixed speed induction generator (FSIG) are the wind turbines with the induction generator whose rotor speed is fixed regardless of the wind speed. These generators were predominantly employed during early installations of wind generators. These units require reactive power support from the grid and are usually equipped with capacitor banks to provide the necessary reactive power. (Ackerman, 2012)
2.2 Doubly Fed Induction Generator (DFIG) Doubly fed induction generator (DFIG) is a variable speed wind turbine with partial scale frequency converter. These generator uses wound rotor induction generator and a partial scale frequency converter on the rotor circuit. The partial scale frequency converter performs the reactive power compensation and smoother grid operation Ackerman, 2012; Chen et al., 2009). Variable speed wind turbines equipped with doubly fed induction generator are becoming more widely used for its advanced reactive power and voltage control capability (Chen et al., 2009). Contrary to fixed-speed
33 Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] Dharel and Maharjan wind machines, these are capable of providing reactive power to the grid. The converter and control schemes associated with these machines permit controlling the active and reactive power output to the desired level. The DFIG can be operated in one of two control modes; firstly, fixed power factor (PF) control, where the turbine controls reactive power production in order to achieve a specified power factor; secondly, terminal voltage control, where the reactive power is controlled to meet a target voltage However, the reactive power capability of the PWM converter is not on par with the synchronous generator and when the voltage control requirement is beyond the capability of the DFIG, the voltage stability of the grid is affected (Chi et al., 2006; Srawan KVS, 2014; Maharjan & Kamalsadan, 2017).
3. Voltage Stability Voltage stability is the ability of a power system to maintain steady acceptable voltages at all buses in the system under normal operating conditions and after being subjected to a disturbance. (Kundur, 1994) The main factor causing instability is the inability of the power system to meet the demand for reactive power. Voltage stability of a power system is affected by various system elements and parameters. Strength of transmission network and power transfer level, generator reactive power/voltage control limits, load characteristics, characteristics of reactive compensation devices, action of voltage control devices such as on-load tap changers, etc. affect the voltage stability. For the purpose of analysis voltage stability is classified into two classes: small disturbance voltage stability and large disturbance voltage stability.
3.1 Small Disturbance Voltage Stability Small-disturbance voltage stability is concerned with a system’s ability to control voltages following small perturbations such as incremental changes in system load. This concept is useful in determining the change in system voltage, at any instant, in response to the small changes in the system. The basic processes contributing to small-disturbance voltage instability are essentially of a steady-state nature. There are various methods for analysis of voltage stability such as Power Voltage (P-V) curves and Voltage-Reactive Power (V-Q) and static analysis (Kundur, 1994). The PV curves and VQ curves can be effectively used to determine stability margins and stability limits of a power system. (Sulaiman, 2015)
3.1.1 PV curves PV curve is a relationship between bus voltage and the injected power. The PV curve in an interconnected network is drawn with the help of series of load-flow solutions (Kundur, 1994). After a bus is identified as point of interconnection (POI) of wind turbine-, the wind active power injection at that bus is gradually increased and the system load is adjusted according to the new generation to obtain the PV curve. The process is performed for three different types of wind turbine.
i Wind turbine 1- FSIG no load compensation ii. Wind turbine 2- FSIG with load compensation iii. Wind turbine 3- DFIG
3.1.2 VQ curves VQ curve is a powerful tool for steady state voltage stability analysis. It gives the relationship between bus voltages and the reactive power injections or absorptions. It helps in determining the steady state voltage stability limits, reactive power margins of the system and the sensitivity of the bus voltages with respect to the reactive power injections (Taylor, 1994). VQ curves at the POI for the three wind turbines presented in
34 Dharel and Maharjan Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] section 3.1.1 are obtained by a series of load flow solutions for different active power outputs. 3.2 Large Disturbance Voltage Stability (Transient Voltage Stability) Large disturbance voltage stability is concerned with a system’s ability to control voltages following a large disturbance such as system faults, loss of generation or circuit contingencies. This ability is determined by the system load characteristic and the interactions of controls and protections. The study period of interest may extend from a few seconds to tens of minutes. The typical approach to study the large disturbance voltage stability is to run time domain simulation under different scenarios. (Cutsem, 1998) During the simulation of transient analysis, generic dynamic data of generators are used. Built-in model of exciter and governor are used as controllers of the generators. IEEE type 1 and HYGOV are the built-in models of automatic voltage regulators and governors used respectively. The dynamic model of wind generators is also implemented in the DigSILENT/ PowerFactory.
4. Description of Studied System Analysis is performed in three different buses in INPS as the point of interconnection which are Suichatar (132 kV), Pokhara (132 kV) and M. Marshyangdi (132 kV). Fig. 1 shows portion of INPS with wind integration in the above-mentioned buses shown inside red circle.
Figure 1: INPS network
35 Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] Dharel and Maharjan
Analysis of wind penetration with different type of wind turbines are performed. For each type of wind turbines presented in section 3.1.1, PV curves and VQ curves are obtained through a series of load flow solutions. For transient voltage stability analysis, voltage recovery characteristics of FSIG and DFIG wind turbine is studied in each of the three locations.
5. Steady State Voltage Stability Analysis
5.1 PV Curve Analysis Wind turbine based on different generators are connected into the transmission grid. When the wind turbine output is low the POI voltage is not significantly affected. As the wind turbine output goes on increasing the POI voltage reduces significantly and decreases fast near the voltage collapse point. The PV curves of the buses as the wind turbine active power increased are shown in Fig. 2 to Fig. 4.
Figure 2: PV curves at Suichatar The base voltage of Suichatar bus is 131.9 kV (1 pu). In the case of wind turbine-1, the bus voltage is maintained around 1 pu for low wind penetration i.e. around 36 MW. As the penetration is higher the rate of decrease in the voltage is increased. The voltage drop is sharp around the nose point which is 374 MW, 114.1 kV (0.86 pu). In case of wind turbine--2, 75 MVAR compensation is used and the voltage has risen to 134.9 kV (1.02 pu) and the voltage stability limit has slightly improved. The nature of the curve remains similar to the wind turbine-1. In case of wind turbine-3, the voltage stability limit is a very high value, 614 MW; nearly double than in case of wind turbine--1. The voltage of the bus remains 1 pu for a large part of the curve, moreover, voltage profile of the bus has slightly increased for penetration up to 200 MW. The nose point (active power limit) for the three different turbines vary as they have different reactive power capability. Wind turbine-1, do not have reactive power generation capability hence, requires reactive power support from the grid and the active power limit is the lowest. The active power injection limit can be increased by increasing the reactive power capability of the reactive power compensating devices or by increasing reactive power capability of PWM converters for wind turbine-3. The base voltage of Pokhara bus is 131.4 kV (0.995 pu) similar to Butwal bus. However, the voltage stability limit of Pokhara bus is much lesser than Suichatar bus. The impact of wind integration is higher in this bus
36 Dharel and Maharjan Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] voltage than the Suichatar bus indicating Pokhara bus is a weak bus. The voltage stability limit is 184 MW, 219 MW and 372 MW for wind turbine--1, wind turbine--2 and wind turbine--3 respectively.
Figure 3: PV curves at Pokhara The peculiar nature of PV curves in M. Marshyangdi bus is attributed to the assignment of the bus as a voltage-controlled bus. M. Marshyangdi generator is directly connected to this bus and Upper Marshyangdi, Midim and Khudi Khola generators are connected by radial lines. M. Marshyangdi and Upper Marshyangdi are operated on voltage control mode set at 1 pu. Thus, Wind integration via this bus does not affect the bus voltage until the reactive power capabilities of the generators are not exceeded. Once the reactive power of the generators has been exceeded, further wind integration makes the system unstable. The voltage stability limit is 257 MW, 314 MW and 440 MW for wind turbine--1, wind turbine--2 and wind turbine--3 respectively.
Figure 4: PV curves at M. Marsyangdi
37 Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] Dharel and Maharjan
5.2 VQ Curve Analysis The VQ curves of the buses for different types of wind turbines are shown in Fig. 5-10. The reactive power margin in all three types of wind turbines decreases as the wind turbine active power output increases. However, the reactive power margin decreases significantly in case of wind turbine-1 and wind turbine-2 whereas wind turbine-3 has lesser impact. The reactive power margin without wind penetration is 790.83 MVAR in Suichatar bus which decreases to 362.95 MVAR in case of wind turbine-1, 398.9 MVAR in case of wind turbine-2 and 675.5 MVAR in case of wind turbine-3, when the active power output of each wind turbine is 252 MW. The reactive power margin in wind turbine-3 is about twice that of wind turbine-1.
Figure 5: VQ curves at Suichatar for Wind Turbine 1
Figure 6: VQ curves at Suichatar for Wind Turbine 2
38 Dharel and Maharjan Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021]
Figure 7: VQ curves at Suichatar for Wind Turbine 3
Figure 8: VQ curves at M Marsyangdi for Wind Turbine 1
Figure 9: VQ curves at M Marsyangdi for Wind Turbine 2
39 Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] Dharel and Maharjan
While constructing the VQ curves of M. Marshyangdi bus, the generator is operated in power factor mode. Construction of VQ curve requires a voltage-controlled element to be connected to the bus whose voltage setting is varied to get various results of voltage and reactive power. If the bus is already a voltage-controlled bus the load flow cannot be performed as there are more than one voltage-controlled elements with different settings.
Figure 10: VQ curves at M Marsyangdi for Wind Turbine 3 The initial reactive power margin of 318.14 MVAR is decreased to 26.7 MVAR, 69.2 MVAR and 233.2 MVAR in case of wind turbine-1, wind turbine-2 and wind turbine-3 respectively when the active power output of each wind turbine is 180 MW. Similarly, the VQ curves of Pokhara bus are also obtained for various active power output of wind turbine based on different generators. The initial reactive power margin in Pokhara bus is around 514.39 MVAR. When the wind turbine active power output is 144 MW, the reactive power margin decreases to 176.97 MVAR in case of wind turbine-1, 253.14 MVAR in case of wind turbine-2 and 446.36 MVAR in case of wind turbine-3. Even if the active power output of wind turbine-3 is increased to 288 MW (twice of 144MW), the reactive power margin is greater than in the case of wind turbine-1 with 144 MW output.
6. Transient Voltage Stability Analysis In transient voltage stability analysis, the voltage response of wind generator is studied after the application of an external three-phase fault on a line. The voltage recovery characteristics i.e., whether the wind generator can acquire a stable voltage after application of three phase fault is studied. The switching time i.e., tripping of the faulted line is taken as 150 ms.
6.1 Transient Voltage Stability of FSIG Fig. 11-12 show the generator voltage response and reactive power consumed by a single generator of 6 MW capacity in a wind farm connected. Generally, a wind farm contains numerous wind turbine and power is fed to common point to the POI. Since all the wind generators in the wind farm are identical with same voltage
40 Dharel and Maharjan Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] response and reactive power characteristics., study of only one generator is enough.
Figure 11: Voltage recovery characteristic of FSIG
Figure 12: Reactive power consumption of FSIG At Suichatar bus, the short circuit fault is defined in Suichatar-Marshyangdi line at time T= 0s and the line is tripped after 150 ms. The voltage recovery characteristics of the wind generator for varying amount of wind penetration is studied. The voltage reaches as low as 0.4 pu but ultimately recovers and attains a steady value of 0.98 pu. It can be seen that before fault the reactive power consumption of the wind generator is 3 MVAR and during transient period the consumption is as high as 9 MVAR. It is observed if the installed capacity of FSIG is larger than 96 MW in Suichatar bus, then the system does not regain steady state even if the fault is cleared in infinitely small time. When all the generators in the wind turbine are considered, there will be large reactive power consumption which affects the stability of the system. At Pokhara bus and M. Marshyagndi bus, the fault is defined in Lekhnath-Damauli line and Marshyangdi line respectively. Fig. 11 and 12 show the voltage and reactive power response of FSIG of 72 MW and 36 MW installed capacity in Pokhara bus and M. Marsyangdi bus respectively. The system fails to converge if the installed capacity of the wind farm is increased further.
41 Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] Dharel and Maharjan
6.2 Transient Voltage Stability of DFIG Fig. 13 to Fig. 16 shows the voltage characteristics and the reactive power consumption of DFIG for different buses. The fault occurs at time T=0s and the time frame of study is taken as 2 s. The DFIG is on PQ control mode with Q=0 MVAR prior to the fault. However, during fault the DFIG also consumes reactive power in transient period and affects the generator voltage and hence, the stability of the system.
Figure 13: Voltage recovery characteristic of DFIG at Suichatar
Figure 14: Reactive power consumption of DFIG at Suichatar Figure 13 and Figure 14 shows the voltage and reactive power response of DFIG with capacity of 120 MW connected to Suichatar bus. The system fails to converge for more than 120 MW installed capacity of the wind turbine. The step size for the time axis in the transient simulation when wind farm is connected to Suichatar bus is different than Pokhara bus and M. Marshyangdi bus, and hence the result is shown in separate figures.
7. Discussion Integration of wind farm of different technologies in three different buses in INPS showed that the Pokhara bus of INPS is weaker among the three buses as the penetration limit is the lowest and wind integration has noticeable impact on the voltage level, even for small penetration. Suichatar bus as POI is the best location for wind penetration in INPS among the three locations. VQ curves of various location showed that the reactive power margin of the POI is largely affected by FSIG whereas DFIG has lesser impact.
42 Dharel and Maharjan Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021]
Figure 15: Voltage recovery characteristic of DFIG
Figure 16: Reactive power consumption of DFIG Fig. 15 and Fig. 16 shows the voltage recovery characteristics and reactive power response of DFIG connected to Pokhara bus and M. Marshyangdi bus. The installed capacity of the wind farm is 90 MW and 66 MW for Pokhara and M. Marshyangdi respectively. The system fails to converge for DFIG of larger installed capacity. The steady state voltage stability limit showed that the limit for wind penetration in a network is large but such penetration was not possible due to transient voltage stability. In case of Suichatar bus the steady-state limit is 374 MW and 614 MW for wind farm-1 and wind farm-3 respectively. However, the transient voltage limit is only 96 MW for wind farm-1 and 120 MW for wind farm-3. Similarly, for other buses, the transient voltage limit is significantly lower than steady state limit. The voltage recovery characteristics of the wind farm are lost if the wind penetration is higher than the transient voltage limit. The transient voltage limit for M. Marshyangdi bus is 36 MW for wind farm-1 and 66 MW for wind farm-3, which is the lowest among the four buses.
8. Conclusions The study demonstrated the voltage stability limit of wind penetration of INPS. The wind power penetration
43 Journal of Engineering Issues and Solutions 1 (1): 32-44 [2021] Dharel and Maharjan capacity at locations with high power potential was determined for different kind of wind turbines in existing INPS system. Further study on the impact of the wind turbines on the frequency stability and control coordination of active and reactive power for stability enhancement can be done.
Conflict of Interests Not declared by authors.
References Ackerman, T. (2012). Wind Power In Power Systems (2nd edition). Wiley. Chen, Z., Guerrero, J., & Blaabjerg, F. (2009, August). A review of the State of the art of Power electronics for Wind Turbines. IEEE Transactions on Power Electronics, 24, 1859-1875. doi:10.1109/TPEL.2009.2017082 Chi, Y., Liu, Y., Wang, W., & Dai, H. (2006). Voltage Stability Analysis of Wind Farm Integration into Transmission Network. International Conference on Power System Technology, IEEE. Cutsem, T. V. (1998). Voltage Stability of Electric Power Systems. Springer Science+Business Media Dordrecht. doi:10.1007/978-0-387-75536-6 IRENA. (2020). Renewable Energy Statistics 2020. International Renewable Energy Agency. doi:978-92-9260-246-8 Kundur, P. (1994). Power System Stability and Control. New-York: McGraw-Hill. Lee, J., & Zhao, Z. (2019). Global Wind Report 2019. Brussels: Global Wind Energy Council. Retrieved from https://gwec. net/wp-content/uploads/2020/08/Annual-Wind-Report_2019_digital_final_2r.pdf Maharjan, R., & Kamalasadan, S. (2013). Real-time simulation for active and reactive power control of doubly fed induction generator. North American Power Symposium, 1-6. doi:10.1109/NAPS.2013.6666957 Maharjan, R., & Kamalasadan, S. (2017). Secondary voltage control of power grid using voltage stability index and voltage control areas. 2017 North American Power Symposium (NAPS) (pp. 1-6). Morgantown, WV: IEEE. doi:10.1109/ NAPS.2017.8107330 Pena, R., Clare, J., & Asher, G. (1996). Doubly fed induction generator using back-to-back PWM converters and its application to variable speed wind-energy generation. Electric Power Applications, IEE Proceedings, (pp. 231-241). Sravan K.V.S., R. K. (2014). Coordination of Reactive Power in Grid-Connected Wind Farms for Voltage Stability Enhancement. IEEE Transactions on Power Systems, 2381-2390. Sulaiman, M. F. (2015, January-February). Voltage Instability Analysis on PV and QV Curves for Radial Type and Mesh Type Electrical Power Networks. International Review of Electrical Engineering (I.R.E.E.), 10(1), 109-115. SWERA. (2006). UNEP/GEP, Solar and Wind Energy Resource Assessemnt Nepal. Alternative Energy Promotion Center. SWERA. (2008). Solar and Wind Energy Resource Final Report (GIS Part). Lalitpur: UNEP/GEF, Alternative Energy Promotion Center. Taylor, C. (1994). Power System Voltage Stability. New York: McGraw Hill.
44 Journal of Engineering Issues and Solutions
A comparative study of structural parameters of a RCC T-girder bridge using loading pattern from different codes
Sulav Sigdel1,* 1Department of Civil Engineering, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Nepal *Corresponding email: [email protected] Received: December 24, 2020; Revised: February 18, 2021; Accepted: March 06, 2021
Abstract Nepal is an under-developed country; it is on the threshold of becoming a developing country. With new highways and railroad projects launching, construction of bridges is likely to increase. Bridges improve connectivity across the country and provide support to the country's overall economic growth. While designing a bridge, concrete properties, reinforcement properties, superstructure and substructure sections, traffic movements and loading conditions are specified. Bridges in Nepal are designed based on criteria enumerated by Indian Road Congress (IRC) code provisions. But, there are different bridge design codes used by different countries. Although these provisions follow the same basic principles, they may yield different results. The study on various structural parameters' variation is significant while selecting the code provision for the design and analysis of the bridge. In this study, a T-Girder Bridge is considered and is modelled and analyzed by vehicular loading patterns from IRC Codal Provision, AASHTO Codal Provision, and Chinese Codal Provision. This study uses CSiBridge computer software to perform the analysis.
Keywords: AASHTO loadings, Chinese loadings, IRC loadings, modelling, structural parameters, T-girder bridge
1. Introduction The bridge's structural design consists of understanding structural members' behavior subjected to forces and loads and designing them with economy and elegance to give safe, serviceable, and durable bridge structures. The structural design of bridges of any country relies on specific codes of practices that provide essential data and standards in analyzing and designing the bridge for safe, economic and required strength criteria. The code of practices depends upon many factors such as desirable strength, economical value, environmental impacts, topological factors, soil types, seismic zones and hydrological characteristics. Till date, Nepal do not have its own code of practices for bridge design. All permanent road bridges in Nepal shall be designed as per IRC loadings or AASTHO loadings. Hence it is requisite to do thorough analysis of the results obtained from different codes before selecting the suitable code of practice. This can be done either
45 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel by manual method where bridge responses of bridge are computed manually following the code of practices. Another way is to model the bridge in the software. For this study, modeling approach is adopted and IRC loading standards for bridge, AASTHO specifications for bridge and Chinese loading are contemplated. For carriageway width less than 5.3m, IRC: 6-2010recommends use of IRC Class A loading [Clause 204.3, Table 2]. AASTHO suggests HSn-44 truck loading for rural and permanent bridges. This loading is practiced in Nepal. For Chinese loading, Chinese JTG Truck is selected as it is recommended by JTG D60-2015and Yan Dai (2016). CSiBridge Software is developed by Computers and Structures Inc., an American based company. Using CSiBridge, users can easily define bridge geometries, boundary conditions, dead and live load cases. The software creates a spine, shell, or solid objects models transformed into a mathematical finite element model. The software allows a user to perform static and dynamic analysis, linear and non-linear analysis, segmental lane analysis, default analysis and parameters analysis. The software provides a way for approximate bridge modelling with provisions of layout lines, spans, abutments, piers, pier caps, slab decks, diaphragm, and load cases (vehicle load, moving load, parapet load, material load, wearing course load, wind loads, seismic loads, hydrological loads, hydrodynamic loads, friction loads, braking loads, collision loads, and others). Software modelling permits us to specify the structure or behavior of a system which can be used efficiently for modelling and analysis of RC T-Girder Bridge. A study conducted in STAAD Pro concludes IRC A Class loading is the most economical and optimum loading for the bridge's design compared to AASTHO specifications and Euro codes. Another research paper entitled Comparative Study of the Analysis and Design of T-Beam Girder and Box Girder Super Structure concludes that the T- beam girder is economical than the box girder. In essence, software analysis can be used effectively and efficiently for modelling and analysis of bridge structure. The influence line is the function that provides the variation of a structural parameter at a specific point on a beam caused by a unit load moving across the beam. This function depends upon structural properties and support conditions and is different for each location and load effect. Influence lines are calculated in the software by matrix method which calculates it on the basis of known weights of a truck. This study uses dynamic moving load, and hence influence lines for each of the moving load shall be used to obtain bridge responses. This software generates influence lines for each case of moving load, and uses it to draw maximum envelopes curves. To the best of the author's knowledge, a comparative study of RCC T-Girder Bridge's structural parameters using loading pattern of bridge design codes as per AASTHO specifications, Chinese loading, and IRC loading is not reported in the existing literature.
2. Methodology The research methodology involves designing and analyzing an RCC T-Girder Bridge. This study adopts following processes. 1. Preliminary design of the superstructure of the bridge are performed. This includes fixing bridge span length, carriageway width, deck width, depth of deck, number of longitudinal girders, section of cross girder, number of cross girder and railing posts. These properties are provided as inputs to the software. 2. The self-weight load of different members of the bridge superstructure are calculated. Under this section, self-weight load of wearing course, curb and railing posts and their acting locations on the bridge are obtained. These loads are provided as inputs to the software.
46 Sigdel Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021]
3. The bridge model is analyzed using CSiBridge v22.1.0 software to obtain maximum and minimum envelopes of bridge responses: axial forces, shear forces, bending moments, vertical displacements, and longitudinal displacements under dead loads and vehicular loading pattern of AASTHO, Chinese, and IRC code of practices. 4. Bridge responses are compared for three code of practices.
3. Dimensioning and Modelling
3.1 Preliminary Dimension of Super Structure The aim of the model is to simulate dynamic truck loading passing over a bridge considering properties of model and realistic conditions. For the modelling, the bridge length is fixed considering the maximum value of linear waterway width and cross-section of the river at the bridge site. Linear waterway width is calculated following a detailed hydrological analysis of the river at the bridge site. The hydrological analysis includes the rational method, slope-velocity method, and empirical method. For a single lane carriageway, Nepal Road Standard (NRS-2070) guides the width of the carriageway, and then the width of the deck is calculated, providing curbs on both sides of the deck. Depth of deck is fixed using allowable span to depth ratio for T Girder, which satisfies serviceability condition of the girder. The thickness of the main girder and cross girder is fixed following IRC codes. The thickness of the bulb of the main girder and that of cross girder is also selected similarly. The preliminary dimensions of superstructures are presented in Table 1. Table 1: Preliminary Dimensions
Bridge member Data Provided Span Length 20m Number of Spans 1 Type of Lane Single Width of the Deck 4.4m Spacing between Main Girders 2.5m Number of Cross Girders per Span 6 Grade of Concrete M25 Grade of Reinforcement HYSD500 Dimension of Railing Post 1m x 225mm x 225mm Spacing of Railing Post 1.333m c/c Depth of Cross Girder 1m Thickness of Cross Girder 0.2m Wearing Course Asphalt Concrete Thickness of Wearing Course at the edge 40mm Thickness of Wearing Course at the Centre of the bridge 80mm Modulus of Elasticity of Concrete 5000√25 = 25000 Mpa Poisson's Ratio 0.2
Fig. 1 shows bridge cross section. Fig. 2 shows bridge longitudinal section, along with cross girders.
47 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel
all dimensions are in meters. Figure 1: Bridge deck section
all dimensions are in meters Figure 2: Bridge cross girder and railings
48 Sigdel Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021]
3.2 Modeling Modelling of the bridge is done in CSiBridge v22.1.0 software. Modelling involves creating a bridge window, adding a layout line and defining bridge component. Layout line is the first step to define a bridge object i.e., line object, and lanes. A layout line of end station at 20m is selected. The inputs for the analysis are dead loads from super-structure, material types, deck section and dynamic vehicular loads from each of the three codes. The supports of the bridge are considered as simply supported. The particular bridge input properties values are: deck section (Fig. 1), cross beam (Fig. 2), 2.64 KN/m2 area load of wearing course (Fig. 3a), 2.3 KN point load of railing posts spaced at 1.33m center to center (Fig. 3b), 2.53 KN/m line load of curb (Fig. 3c) which acts through the center of gravity of curb at a distance of 0.225m from the edge of the deck, and live loads (Fig. 4, Fig. 5, and Fig. 6).
(a) (b) (c) Figure 3: Dead loads in super-structure: (a) area load of wearing course, (b) point load of railing post, (c) line load of curb The structural model is updated as an area model with a preferred maximum sub-mesh size of 2.5 units. Load discretization is at every 0.05 seconds and the output time step size is 0.05 seconds. The speed of vehicular loading is 30 KMPH. Analysis type is linear, and the time history type is direct integration. Damping and vibrations are not considered. Lateral loads are not applied externally but after modelling, lateral loads are developed. This is due to the formation of frictional forces between vehicle tire and bridge surface. Hydrodynamic loads, brake loads, wind loads and accidental loads are not used. Live load defined in IRC code as IRC Class A (Fig. 4), in AASTHO code as HSn-44 (Fig. 5), and in Chinese code as Chinese Truck JTG 2015 (Fig. 6) are added.
Figure 4: IRC A loading
49 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel
Figure 5: AASTHO HSn-44 loading
Figure 5. AASTHO HSn-44 Loading
Figure 6: Chinese truck JTG loading From Fig. 4, Fig. 5, and Fig. 6, it can be seen that total vehicular load and length of vehicular loading of IRC Class loading, AASTHO HSn-44 loading and Chinese Truck JTG loading are 320.2 KN and 2.8m, 550 KN and 12.8m, and 543.4 KN and 18.8m respectively. Length of vehicular loading has effect on influence line diagram and vehicular load impacts the value of bridge responses. Since AASTHO HSn-44 loading and Chinese Truck JTG loading have comparable vehicular loadings which are greater than IRC Class A loading, maximum bridge responses can be developed due to these two vehicular loadings. The length of vehicular loading of Chinese loading is greater than the other two. Hence it can be predicted that the maximum bridge response may be developed by Chinese loading. The output of the dynamic vehicle-bridge simulation are the bridge responses: axial forces, shear forces, bending moments, vertical displacements, and longitudinal displacements. 3D model of the bridge before the analysis is shown in Fig. 7(a) and Fig. 7(b). The deformed model of the bridge obtained after the analysis is presented in Fig. 8. Fig. 9 shows the deformed model of the bridge under AASTHO HSn-44 loading after 1.4 seconds of the application of truck load.
50 Sigdel Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021]
Figure 7(a): 3D model of bridge Figure 7(b): L-Section of bridge model
Figure 8: Bridge model after analysis Figure 9: AASTHO Hsn-44 loading
4. Results and Discussions Bridge object response of the deck to the combination of dead load and live load along the span of the bridge is shown in the following table (Table 2) and figures.
Figure 10: Axial force envelope (Hsn-44) Figure 11: Axial force envelope (Chinese Truck)
51 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel
Table 2: Bridge object response
HSn-44 Truck Reactions Class A (IRC) (AASTHO) JTG(Chinese) Max Value 0.423 (Right) 0.117 (Right) 0.145 (Right) Axial Force (KN) Distance 0.000 5.714 0.000 Min Value 0.122 (Left) 0.023 (Right) 0.052 (Right) Distance 14.286 20.000 0.000 Max Value 795.088 (Up) 887.122 (Up) 842.429 (Down) Distance 0.000 0.000 20.000 Shear Force (KN) Min Value 87.930 (Down) 85.958 (Down) 89.974 (Up) Distance 11.429 11.429 8.571 Max Value 4048.189 (CW) 4454.947 (CW) 4338.307 (CW) Bending Moment Distance 8.571 8.571 11.429 (KN m) Min Value 1.0509 (CW) 0.1751 (CCW) 0.3195 (CW) Distance 20.000 20.000 20.000 Max Value 21.840 (Down) 24.378 (Down) 23.271 (Down) Vertical Distance 11.429 11.429 11.429 Displacement (mm) Min Value 0.160 (Down) 0.175 (Down) 0.156 (Down) Distance 0 20 0 Max Value 1.304 (Left) 1.470 (Left) 1.395 (Left) Longitudinal Distance 20 0 20 Displacement (mm) Min Value 0.299 (Left) 0.312 0.308 (Right) Distance 11.429 11.429 8.571
Due to dead loads and live loads on the longitudinal girder, the axial force envelopes are presented (Fig. 10, Fig. 11, and Fig. 12). The analysis shows longitudinal girder produced maximum axial force due to AASTHO combination loading at the start of the span (Fig. 10) but near mid-span, the longitudinal girder had more axial force due to the Chinese truck combination loading (Fig. 11). For IRC combination loading (Fig. 12), the longitudinal girder produced more axial force at the start of the span. Among these three loadings, AASTHO combination loading had a maximum axial force on the longitudinal girder. Shear force envelopes are presented (Fig. 13, Fig. 14, and Fig. 15). The analysis shows longitudinal girder produced maximum shear force due to AASTHO combination loading at the start of the span (Fig. 13) and minimum shear force at a distance of 11.429 meters. Due to Chinese combination loading (Fig. 14), the longitudinal girder produced more shear force at the start of the span and minimum shear force at a distance of 11.429 meters. Due to IRC combination loading (Fig. 15), the longitudinal girder produced more shear force at the start of the span and minimum shear force at a distance of 8.571 meters. Among these three loadings, Chinese combination loading produced maximum shear force on the longitudinal girder.
52 Sigdel Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021]
Figure 12: Axial force envelope (IRC A)
Figure 13: Shear force envelope (HSn-44) Figure 14: Shear force envelope (Chinese Truck)
53 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel
Figure 15: Shear force envelope (IRC A) Figure 16: B.M. envelope (HSn-44) Bending moment envelopes are presented (Fig. 16, Fig. 17, and Fig. 18). The analysis shows longitudinal girder produced maximum bending moment due to AASTHO combination loading at a distance of 8.571 meters (Fig. 16) and minimum bending moment at the end of the span (Fig.16). Due to Chinese combination loading (Fig.17), the longitudinal girder produced maximum bending moment at a distance of 8.571 meters and minimum bending moment at the end of the span. Due to IRC combination loading (Fig. 18), the longitudinal girder produced maximum bending moment at the start of the span and minimum bending moment at a distance of 8.571 meters. Among these three loadings, Chinese combination loading produced maximum bending moment on the longitudinal girder. Vertical displacement due to dead loads and live loads on longitudinal girder are shown (Fig. 19, Fig. 20, and Fig. 21). The analysis shows that longitudinal girder produced maximum vertical displacement due to AASTHO combination loading (Fig. 19) at a distance of 11.429 meters and minimum vertical displacement at the start of the span. Due to Chinese combination loading (Fig. 20), the longitudinal girder produced maximum vertical displacement at a distance of 11.429 meters and minimum vertical displacement at the end of the span. Due to IRC combination loading (Fig. 21), the longitudinal girder produced maximum vertical displacement at a distance of 11.429 meters and minimum vertical displacement at the start of the span. Among these three loadings, Chinese combination loading produced maximum vertical displacement on the longitudinal girder. Longitudinal displacement envelopes are presented (Fig. 22, Fig. 23, and Fig. 24). The analysis shows that longitudinal girder produced maximum longitudinal displacement due to AASTHO combination loading at the end of the span (Fig. 22) and minimum longitudinal displacement at a distance of 11.429 meters. Due to Chinese combination loading (Fig. 23), the longitudinal girder produced maximum longitudinal displacement at the start of the span and minimum longitudinal displacement at a distance of 11.429 meters. Due to IRC combination loading (Fig. 24), the longitudinal girder produced maximum longitudinal displacement at the end of the span and minimum longitudinal displacement at a distance of 8.571 meters. Among these three loadings, Chinese combination loading produced maximum longitudinal displacement on the longitudinal girder.
54 Sigdel Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021]
Figure 17: B.M. envelope (Chinese Truck) Figure 18: B.M. envelope (IRC A)
Figure 19: Vertical displacement (HSn-44) Figure 20: Vertical displacement (Chinese Truck) From these data, it can be seen that Chinese combination provides maximum values of shear forces, bending moments, vertical displacement and longitudinal displacement. This is due to the fact that axle load and total load of Chinese JTG Truck is greater than that of IRC Class A loading and AASTHO HSn-44 loading.
5. Conclusions Bridge design codes adopted by different countries may indicate variation in structural parameters for the design of bridges. Thus, to obtain comparative statement structural parameters, three bridge design codes are considered, and detailed modelling of the T-Girder Bridge is done in CSiBridge 2020 v21.1.0. In the design of bridge longitudinal girders with Chinese codes shear forces, bending moment, vertical displacement, and longitudinal displacement are greater (Table 2) than the other two, i.e. AASTHO specifications and IRC codes. The longitudinal girder design using IRC codes acquired axial force, shear force, bending moment, vertical displacement, and longitudinal displacement values in between the other two codes, i.e. AASTHO
55 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel specifications and Chinese codes. This depicts that IRC Class A loading provides the right balance between safety and serviceability for the design.
(HSn-44) (Chinese Truck) Figure 21: Vertical displacement (IRC A) Figure 22: Longitudinal displacement (AASTHO)
(HSn-44) Figure 23: Longitudinal displacement (Chinese) Figure 24: Longitudinal displacement (IRC A) (Chinese Truck) (IRC A) This study presents the results based on three bridge design codes only. Future studies may incorporate Euro codes, Great Britain codes, Canada codes, and International design codes. Variation of steel reinforcement in main girder, cross girder, and abutment can be studied which may provide a basis for economic comparison of the bridge design codes. Since CSiBridge uses a transformed mathematical finite-element model by meshing the material domain and assigning material properties, decreasing the mesh size, reducing load discretization,
56 Sigdel Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] and reducing output time step size gives more accurate results. For more detailed analysis, researchers can vary the width of the deck and length of the span.
Conflict of Interests Not declared by authors.
References AASHTO (2010), "AASHTO LRFD Bridge Design Specifications". American Association of State Highway and Transportation Officials, 5th Edition. Washington DC. Bhadauria D. S. and Patel, R. (2017), "Comparative Study of RCC Bridge for Central Zone of India for Different Sections of Girder". International Journal for Scientific Research & Development, Volume 5, and Issue 4, 2017. Bhat S.B. (2004), "Development of Highway Bridge loading for Nepal". Final thesis report, in the partial fulfilment of the requirements of the degree of Master of Science in Structural Engineering, IOE, Pulchowk Campus, Nepal. Buckland, P.G. and Sexsmith, R.G. (1980), "A comparison of design loads for highway bridges". Buckland and Taylor Ltd, North Vancouver, B.C, Canada V7P 2Y4. Buckle I.G. (1996), "Overview of Seismic design methods for bridges in different countries and future directions". Department of Civil Engineering and National Center for Earthquake Engineering Research, the State University of New York at Buffalo, New York, United States of America 14261. Chan, T.H.T., and O'Connor, C. (1990), "Wheel Loads from Highway Bridge Strains: Field Studies." Journal of Structural Engineering, ASCE, Vol. 116, No. 7. Chan, T.H.T., and O'Connor, C. (1990b), "Vehicle Model for Highway Bridge Impact." Journal of Structural Engineering, ASCE, Vol. 116, No. 7. Dai, Y. (2016, “Study on highway bridge vehicle load standard value”. International Conference on Civil, Transportation and Environment (ICCTE). Shaanxi, 710065, China. G.B. 50011(2010), "Code for Design of Concrete Structures", National Standard of the People's Republic of China, Beijing. Gautam, M. (2000), "Bridge Loadings of different countries and in the context of Nepal". Final thesis report, in partial fulfilment of the requirements of the degree of Master of Science in Structural Engineering, IOE, Pulchowk Campus, Nepal. I.S. 456:2000, "Code of Practice for Plain and Reinforced Concrete". Bureau ofIndian Standards. New Delhi, India. IRC 6-2010, "Standard Specifications and Code of Practice for Road Bridges". Section II, loads and stresses. The Indian Roads Congress. New Delhi, India. IRC: 21-2000, "Standard Specifications and Code of Practice for Road Bridges, Section III, Cement Concrete (Plain and Reinforced)". The Indian Roads Congress. New Delhi, India. 2000. IRC: S.P 54-2000, "Project Preparation Manual for Bridge". The Indian Roads Congress. New Delhi, India. 2000. JTG D60-2015, “General code for design of highway bridges and culverts[S]”.People's traffic press. Beijing Nepal Road Standard-2070, Government of Nepal, Ministry of Physical Planning and Works, Department of Roads. 2070 Paval, P. (2015), "Analysis of Multi-Cell Prestressed Concrete Box-Girder Bridge". International Journal of Engineering Technology Science and Research IJETSR (ISSN) 2394 – 3386. Vol 3. Pokharel, M. (2013), "Comparative Study of RCC T-Girder Bridge with Different Codes". Final thesis report in the partial fulfilment of the Master of Science in Structural Engineering requirements. IOE, Pulchowk Campus. Lalitpur, Nepal. Qaqish, Fadda, and Akawwi. (2008), "Design of T-Beam Bridge by Finite Element Method and AASHTO Specification". KMITL Sci. J. Vol.8 No.1. Raina V. K. (2007), "Concrete Bridge practice analysis, Design, and Economics". 2nd Edition. Rajamoori, A.K. and Krishna, B.V. (2014), "Design of Pre-Stressed Concrete T-Beam Bridges". International Journal of Bridge Engineering (IJBE), Vol. 2, No. 3, (2014), pp. 01-14. Raju, N.K. (2010), "Design of Bridges". 4th edition. Ravi K and Jagdish Chand (2019), "Design and Analysis of Bridge Girders using Different Codes". International Journal of Engineering Research & Technology (IJERT). Vol. 8, Issue 07. (July-2019). Saxena, A. and Dr Maru S. (2013), "Comparative Study of the Analysis and Design o T-Beam Girder and Box Girder Super Structure". International Journal of Research in Engineering & Advanced Technology (IJREAT), Vol. 1, Issue 2,
57 Journal of Engineering Issues and Solutions 1 (1): 45-58 [2021] Sigdel
(April-May, 2013). Victor, D.J. (2007), "Essentials of Bridge Engineering". 6th edition. Vijaykumar, S.P. and Mohan, K (2016), "Analysis of Bridge Girder with Beam and without Beam. International Journal of Civil Engineering and Technology, 7(5), 2016, pp.337–346. Vishal, Gore, and Salunke. (2014), "Analysis and Design of Prestressed Concrete Girder". International Journal of Inventive Engineering and Sciences (IJIES) Volume-2, Issue-2, January 2014. Wang, T.L. and Huang, D.Z. (1992), "Computer Modeling Analysis in Bridge Evaluation." Report No. FL/DOT/RMC/0542(2)- 4108. Structural Research Centre. Department of Transportation, Tallahassee.
58 Journal of Engineering Issues and Solutions
Analyzing effectiveness of active learning through project-based learning approach in university level ICT courses
Dhiraj Shrestha1, Satyendra Nath Lohani1,*, Roshan Manjushree Adhikari1 1School of Engineering, Department of Computer Science & Engineering, Kathmandu University, Dhulikhel *Corresponding email: [email protected] Received: January 26, 2021; Revised: March 05, 2021; Accepted: March 16, 2021
Abstract The concept of Active Learning (AL), which has journeyed through multiple research studies over the years, is an important part of the teaching learning process at academic institutions. The present study applies active learning via project-based approaches where students engage in real life projects and solve associated complications with their research, communication, and technical skills. As a case study of effectiveness of project-based learning (PBL), especially in engineering project contexts, the present research is conducted among students studying computer science and engineering at Kathmandu University (KU), Nepal. The key findings of the study suggest that PBL assignments have helped students in their active learning processes. This paper also compares teaching and learning approaches of KU with other IT institutions of Nepal.
Keywords: Active learning, Project based learning, Student centered learning, ICT education
1. Introduction According to Kokotsaki, Menzies, & Wiggins, (2016), Project-based learning (PBL) has been compared with similar other pedagogical approaches like problem based learning, experiential or collaborative learning. They have recommended six essential characteristics for successful adoption of PBL: effective guidance and support for students, regular teacher support through professional development opportunities, effective group work, balancing didactic instruction with independent inquiry method, continuous monitoring of progress and finally a sense of student choice and autonomy. Hence, PBL has been practiced a long way in modern education and success stories of PBL have been shown in almost every study that follows the active learning procedures. Active Learning (AL) is a process where students engage in activities that promote higher order learning skills (Bonwell & Eison, 1991). AL has received a good amount of attention from researchers and academicians for several years, and replaced the traditional teaching-learning method. AL differs from traditional learning
59 Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021] Shrestha et al. methods where students passively receive information from the instructor, does his/her homework or lab work whereas in active learning, the engagement of learners in the learning process is of central focus. PBL is a practical teaching methodology in which teaching is focused around activities that help students gain practical knowledge together among other learners while creating and testing an industry related or close to industry related project (Sanger &Ziyatdinova, 2014). In engineering context, a project can basically be defined as a work, generally closer to the professional world that will take a specific time scale for its completion and are more directed to application of the gained knowledge (Lacuesta, Palacios, & Fernández, 2009). As PBL is learner centric approach, learners are collaboratively involved in planning, designing, implementing and testing the project in real life situations due to which they create their own knowledge rather than depending on knowledge imposed to them (Giri, 2016). According to Stripling, Lovett & Macko (2009), “Project-based learning is the instructional strategy of empowering learners to pursue content knowledge on their own and demonstrate their new understandings through a variety of presentation modes.” According to Thomas (2000), the characteristics of PBL practices are that they: are central and not peripheral to the curriculum, are focused on driving problems, make students involved in a constructive investigation and are closer to reality. The adoption of PBL is gradually increasing throughout universities of the world for producing graduates that are capable of applying practical application engineering (Sanger &Ziyatdinova, 2014; Thomas, 2000) . In this study, the authors present their experience of introducing PBL at Kathmandu University, Nepal. Students have developed projects for solving problems of Internet quality measurement and File management system for Namobuddha Municipality. The key findings of the study suggest that PBL has helped students envision how real-life projects work, improve their research, communication, time management and networking skills. These findings have led the researchers to recommend a possible integration of PBL in future teaching-learning sessions especially in engineering project contexts.
2. Methodology This research used both quantitative and qualitative methods to analyze efficacy of project-based learning in ICT courses. The quantitative model helped the researchers to measure the differences of students' perception in various questionnaires before and after the projects. The qualitative approach helped in perceiving the experiences of students and faculty members in overall completion of this research project. The steps taken to carry this project are discussed in the sections below. Step 1: Participants Identifications and Initial Data Collection Firstly, two students’ groups, out of 100 students, have been chosen to carry out two different research projects. The students engaged in this process have studied in the fourth semester of the Department of Computer Science and Engineering at Kathmandu University. The questionnaires have been distributed to measure their views on project motivation factors, faculty and students bonding, student abilities, and difficulties faced by students in their academic activities. Likert-scale of 5 points have been used to measure their responses for a particular statement. The detailed questionnaire is attached in the appendix section of this paper. The response set for our questionnaire is listed below. Step 2: Implementing PBL Secondly, two courses namely Communication and Networking (COMP 204) and System Analysis and Design (COMP 302) have been introduced where project-based learning methods have been applied to enhance
60 Shrestha et al. Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021] students' knowledge. Student groups have been taken to Namobuddha municipality where they investigated the current problems that could be solved using IT skills and resources. Two problems have been identified and they were turned into projects namely ‘Internet Health Nepal’ and ‘File Management System’. Students have been supervised by principal investigator and co-principal investigator of this research project who were also course instructors of two aforementioned subjects. Table 1: A summary of response to the questionnaire
Response Set 1 2 3 4 5 Agreement Strongly Disagree Neither Agree nor Agree Strongly Agree Disagree Disagree Frequency Never Rarely Sometimes Often Always Importance Not at all Slightly Moderately Very Extremely important important important important Important Satisfactions Not at all Slightly Moderately Very Satisfied Completely satisfied satisfied Satisfied Satisfied
Step 3: Final Data Collection After successful completion of projects, students have been evaluated by various faculties and officials of Namobuddha municipality and students' efforts have been highly acknowledged. Participants have been questioned again with a previous list of questionnaires to measure their views regarding specific topics this time. Similarly, participants and course instructors submitted their written reflection about their learning and teaching experiences, respectively. Step 4: Data Analysis & Reflection Analysis The variation of data in our Likert scale (before and after) is our subject to analyze and is discussed in results sections. Qualitative analysis has been done to measure the effectiveness of project-based learning where students and course instructors both expressed their individual learning and teaching experiences. Step 5: Comparison with other Institutions The same questionnaire set (from step 1) have been distributed among students from other institutions of Nepal who are involved in IT education, namely Islington College, Kantipur City College, Lumbini Engineering College and Everest Engineering College. This process has been carried out in order to understand the teaching and learning approach of various IT institutions of Nepal. The data obtained from this survey have been compared with Kathmandu University initial data.
3. Results and Discussion 3.1 Students’ Perception The study has shown that the technical as well as non-technical skills of the students have been improved after their involvement in PBL. Based on the reflection obtained from the students, the following were some common responses:
‘The project gave me an opportunity to work directly for the community and face real life problems.' 'The experience taught me how to face real life problems to develop products that actually provide solutions to the society's
61 Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021] Shrestha et al. problems.' 'Till this project I didn't know how a real-life system was developed.' ‘We regularly visited our supervisor for guidance and to improvise our work’ ‘It gave me an insight on how to deal with people in real world field as we have just been sitting and looking on the internet without knowing people’ ‘By the end of the whole process, I got in touch with my skills I hadn’t yet discovered and got to sharpen the ones I was familiar with.’ These results indicate that the students had a significant experience working on a real problem-solving environment that helps them increase team coordination, ability to communicate, ability to relate technical aspects to real life problems.
3.2 Questionnaire Distribution and Analysis Here are some of the major conclusions drawn from the student’s responses from questionnaires distributed to them.
• As shown in Fig. 1, ‘Solving real world problems and challenges’ have been rated the most important motivating factor while selecting the projects whereas factors ‘Enhancing programming skills’ and ‘Fulfilling academic requirements’ have received a decrease in priority after being involved in PBL. • As shown in Fig. 2, frequencies of meeting their respective project supervisor have increased after being involved in PBL. • As shown in Fig. 2, both the influence of supervisor and an independent student to change the features of the project during the development phase have decreased. This fact was also reflected in students’ reflections where students being involved expressed that as they were being involved in real projects solving real world problems, the scope of flexibility became a little restricted compared to working on an academic project for which the main motivation was to enhance programming skills. • As shown in Fig. 3, student’s ability to self-learn, written and oral communication skills and team work coordination improved significantly. One of the main strengths of the PBL process is to improve skills of its learners and judging by the responses of the participants, they have successfully been able to incorporate some of these skills (Thomas, 2000). • As shown in Fig. 4, the data also reflected that PBL has helped students develop more efficient time management skills. Learning new technical skills and creating a project obviously takes time but PBL creates an environment where students work in groups, share ideas, and learn to break down tasks which makes the time management easier. The serious approach of PBL also enforces students to learn the importance of time management and act accordingly (Mergendoller & Thomas, 2003). • As shown in Fig. 4, students felt more responsible towards the project than they did on their previous academic project involvement as PBL demands more and more from its participants involved in the project. Students were encouraged to think, brainstorm, work together and come up with new ideas. • As shown in Fig., individual student's freedom of action was restricted after being involved in the PBL process. Even though students at their early phases of learning would like to experiment and play
62 Shrestha et al. Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021]
around with new ideas regarding the project and add creativity, PBL restricts them to do so at a certain level as they have been dealing with real world problems.
Figure 1: Comparison of before vs after Likert scores of the students regarding project motivation factor
Figure 2: Comparison of before vs after Likert scores of the students regarding faculty and student’s bonding
3.3 Faculties’ Perception The course instructors have used almost similar kinds of approaches to involve their students in PBL with the intention to measure the activeness of their students in each phase. In a nutshell, the phases have been divided into requirement, analysis and implementation, along with the testing phase. It has been seen from the faculty perspective that each group has been dedicated and enthusiastic because they have been dealing with real life problems. Unanimously, it can be concluded that students and faculty bonding have been strong and harmonious during the course conduction. And students’ self-abilities have also been increased due to involvement in PBL.
63 Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021] Shrestha et al.
Figure 3: Comparison of before vs after Likert scores of the students regarding student abilities
Figure 4:Comparison of before vs after Likert scores of the students regarding difficulties remarked by students while doing the project
3.4 Comparison with Other Institutions In Figures (5-17), the comparisons between average Likert scores have been made between KU students and students from other institutions. As shown in figures, the average difference between the Likert scores turned out to be very low (±0.5 out of 5 max Likert score). So, due to this 10% of difference, it has been concluded that the teaching methodology and learning approach followed in KU and other institutions are similar in nature.
64 Shrestha et al. Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021]
Figure 5:Comparison of average Likert scores of all Figure 6: Comparison of average Likert scores of institutions regarding “Solving real world problems all institutions regarding “Enhancing programming and challenges” as motivation factor for project skills” as motivation factor for project
Figure 7: Comparison of average Likert scores Figure 8: Comparison of average Likert scores of all institutions regarding “Fulfilling academic of all institutions regarding their supervisor visit requirements” as motivation factor for project frequency
Figure 9: Comparison of average Likert scores of all Figure 10: Comparison of average Likert scores institutions regarding supervisor influence to change of all institutions regarding individual student's the features of the project in development phase influence to change the features of the project in development phase
65 Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021] Shrestha et al.
Figure 12: Comparison of average Likert scores Figure 11: Comparison of average Likert scores of of all institutions regarding their oral and written all institutions regarding their self-learning abilities communication skills
Figure 13: Comparison of average Likert scores of all Figure 14: Comparison of average Likert scores institutions regarding their team work coordination of all institutions regarding time management
Figure 15: Comparison of average Likert scores of Figure 16: Comparison of average Likert scores all institutions regarding supervisor enforced plan of all institutions regarding lack of individual responsibility
66 Shrestha et al. Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021]
Figure 17: Comparison of average Likert scores of all institutions regarding student freedom of action
4. Conclusions and Recommendations Active learning through a project based approach have been introduced and implemented for ICT courses at the Department of Computer Science and Engineering, Kathmandu University. The teacher-centric course delivery mode in comparison to PBL has been discussed as deprecated for meeting the 21st century learner skills. In general, both students and faculty members responded positively towards this new approach. Students’ enthusiasm for solving real world problems has been visible through self-learning abilities, oral and written communication skills, teamwork coordination and time management skills have been increased due to implementation of project based learning approach. On the other hand, fulfilling academic requirements, supervisor influence and enforced plan, lack of responsibility towards individual student and student freedom of action have been given less priority after implementing PBL. Also, it has been seen that the level of challenges that the students have to tackle are almost the same while comparing with other IT institutions of Nepal. In this regard, we can recommend the use of PBL for increasing the effectiveness of learning approach in academic contexts where the universities and educational institutions should be open towards change, exploring new ways of how to improvise teaching learning so that they can produce more confident and competent ICT graduates.
Acknowledgements The success and final outcome of this research project required a lot of guidance and assistance from many people and the authors feel extreme gratitude for their endless guidance and assistance. To begin with, we would like to thank University Grant Commission (UGC), Nepal for providing financial support to carry on our research project. Furthermore, we would also like to thank the DoCSE (Department of Computer Science and Engineering) family of Kathmandu University for their continuous feedback and mutual support to complete this research project. We are also extremely thankful to various IT colleges of Nepal for providing valuable data for comparison and platform to interact with their students. Lastly, we would like to thank our undergraduate students for their active participation in our research project.
Conflict of Interests Not declared by authors.
67 Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021] Shrestha et al.
Appendix: Questionnaire (All the questions expect answers in Likert scale: 1-5. The description of response set is explained in Methodology section)
Project Motivation Factors
• How would you rate "Solving real world problems and challenges" as the motivation factor while selecting the projects? • How would you rate "Enhancing programming skills" as the motivation factor for your projects? • How would you rate "Fulfilling academic requirements" as motivation factor of your projects?
Faculty & Students Bonding
• How frequently do you meet with your project Supervisor during project? development? • Rate your supervisor influence to change the features of your project in development phase. • Rate the role of individual student to change the features of your project in development phase (convincing the group and supervisor).
Student Abilities
• Rate your self-learning abilities • Rate you oral and written communication abilities • Rate your team work coordination
Difficulties remarked by students while doing project
• Time management • Supervisor enforced plan • Lack of Responsibility of individuals in team • Student freedom of action
References Bonwell, C. C., & Eison, J. A. (1991). Active Learning: Creating Excitement in the classroom. Association for the Study of Higher Education.
Giri, D. R. (2016). Project-Based Learning as 21st Century Teaching Approach: A Study in Nepalese Private Schools. 487-497.
Kokotsaki, D., Menzies, V., & Wiggins, A. (2016). Project-based learning: A review of the literature. Improving Schools .
Lacuesta, R., Palacios, G., & Fernández, L. (2009). Active Learning through Problem Based Learning Methodology in Engineering Education. 39th ASEE/IEEE Frontiers in Education Conference. San Antonio: IEEE.
68 Shrestha et al. Journal of Engineering Issues and Solutions 1 (1): 59-69 [2021]
Mergendoller, J. R., & Thomas, J. W. (2003). Managing Project Based Learning: Principles from the Field. California: Buck Institute for Education.
Sanger, P. A., &Ziyatdinova, J. (2014). Project Based Learning: Real World Experiential. International Conference on Interactive (pp. 541-544). Dubai: IEEE.
Stripling, B., Lovett, N., &Macko, F. C. (2000). Project-based learning: Inspiring middle school students to engage in deep and active learning.
Thomas, J. W. (2000). A review of research on Project Based Learning. San Rafael, California: TheAutodesk Foundation. Retrieved from http://www.bobpearlman.org/BestPractices/PBL_Research.pdf
69 Journal of Engineering Issues and Solutions
Influence of structural irregularities on seismic performance of RC frame buildings
Krishna Ghimire1, Hemchandra Chaulagain1,* 1 School of Engineering, Faculty of Science and Technology, Pokhara University, Kaski, Nepal *Corresponding email: [email protected] Received: January 05, 2021; Revised: March 17, 2021; Accepted: March 21, 2021
Abstract Irregular building structure is frequently constructed across the globe for fulfilling aesthetic as well as functional requirements. The structures with irregularities are the common building type in earthquake-prone country like Nepal. However, a post-earthquake reconnaissance survey reports revealed the high seismic vulnerability of the building with structural irregularities. In this context, the present study explores the influence of structural irregularities on performance of reinforced concrete (RC) frame structure. To this end, the structural irregularities are created in in the building structures. The geometrical irregularities are created by removing the bays in different floor levels. Likewise, the effect due to mass irregularities are studied by considering the swimming pool and game house at different floor levels. Furthermore, the stiffness irregularities are formulated by removing the building columns at different sections. All these irregularities are studied analytically in finite element program with 3-D structural models. The numerical analysis is done with non-linear static pushover and time history analysis. The results are analyzed in terms of fundamental time period, storey shear, storey displacement, drift and overturning moment. The results indicate that the level of irregularities significantly influenced the behavior of structures.
Keywords: Nonlinear analysis; plastic hinge; storey drift; structural irregularity.
1. Introduction The behavior of a civil engineering structure during strong ground shaking depends on the level of irregularities in structures (Lee and Ko, 2007). It mainly occurs due to the irregular distribution in their strength, stiffness, mass and uneven plan configuration along the height of the structure and its combined effects. Past scenarios of the damage patterns of the building indicated that the seismic response of irregular building subjected to ground motion tends to be significantly stronger due to torsional effects. It arises from the non-uniform distribution of mass and stiffness of the structure. Torsion has been the cause of major damage to buildings subjected to strong shaking. It occurs under the action of earthquake forces when the centre of mass of the building does not coincide with the center of rigidity. The distance between them is
70 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] called eccentricity. Lateral force multiplied with this eccentricity causes a torsional moment that must be resisted by the structure (Gautam and Chaulagain, 2016). To perform well against seismic force, structure should be subjected to adequate lateral strength, simple and regular configuration, sufficient stiffness and ductility. Buildings with simple geometry and uniformly distributed mass and stiffness in plan and elevation are less vulnerable in comparison to the structure with irregular configuration (Kostinakis and Anthanatopoulou, 2020). In reality, a large number of building structures are in irregular in some sense. Some have been initially so designed and others have become so by accidently. The main vertical irregularities examined by the researchers are stiffness irregularity (soft- storey), mass irregularity, vertical geometric irregularity and in-plane discontinuity. Similarly, the horizontal irregularities are basically due to asymmetrical plan shapes, re-entrants’ corners, diaphragm discontinuity and torsional irregularities (Varadharajan et al., 2012). Nowadays, irregular structures are quite frequently built in Nepal. These constructions are popular in multi-storied building because of its both aesthetic architecture and functional use. Due to the irregular nature of the structure stress concentration and ductility demand is localized in the structure. On the other hand, regular structures have uniform distribution of mass and stiffness and resulting the improved level of performance. In this context, this study highlights the effect of irregularities by comparing the results with regular structure. The results are analyzed in terms of fundamental time period, storey shear, storey displacement, drift and overturning moment.
2. Classification of Irregularities 2.1 Mass Irregularity In structural system, if there is a variation of more than 150% of mass between the adjacent story then it is considered as mass irregularity (see Fig.1). Researchers highlighted the effect of several irregularities such as strength, mass, discontinuity in capacity and restrained corner in their study (Sadashiva et al. 2009). Several building structures were damaged during Bhuj, Chili and Gorkha earthquake due to the mass irregularities. The higher amount of mass leads in the reduction of ductility of vertical load resisting elements and leads to the collapse of structures. The heavy mass on upper story leads the structure to the vulnerable condition than those at lower story level. From the analytical study of different regular and irregular building, it is noticed that a type, magnitude and location of irregularities had strong influence on collapse capacity of the structures. The buildings having stiffness, setback and strength irregularity at the bottom storey has less collapse capacity (Chaulagain et al., 2016). For mass irregular building, the maximum impact on collapse response was observed for the case when mass irregularity was present at the top story.
Figure 1: Representation of mass irregular structure.
71 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
2.2 Stiffness Irregularity In structural system, if the lateral stiffness is less than 70% of that in the storey above or less than 80% of the average stiffness up to 3 storey then it is said to be soft story (Dya and Oretaa, 2015). During the earthquake in Chili, several number of buildings around the alto-Rio building were badly damaged but safe while the alto-Rio building got completely collapsed due to vertical irregularities in the stiffness (Rahman and Salik, 2016). The performance of the structure also depends on the lateral shear stiffness or flexural stiffness. The lateral shear stiffness of the story can be found by using following relation. The representation of stiffness irregularity is shown in Fig. 2.
Where,
Nc total number of continuum columns in the ith story Nstrut the total number of struts in ith story Ej, modulus of elasticity of materials Ij moment of inertia of the member Lj, length of column I direction of interest Em elastic modulus Am axial area Lm Length m angle of inclination with respect to the horizontal axis of strut
Θ
(a) (b) (c)
Figure 2: Stiffness irregularity: (a) stiff and strong upper floors due to masonry infills, (b) the columns is one storey longer than those above and (c) soft storey caused by discontinuous column.
72 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
2.3 Vertical Geometric Irregularity In the structural system, if the horizontal dimension of the lateral force-resisting system in any story varies by more than 130 percentage of adjacent story in both above and below level, then it is said to be vertical geometric irregularities (Amiri and Yakhchalian 2020, Sarkar et al 2010). This type of irregularity exists in elevation (see Fig. 3a)
2.4 Horizontal Irregularity These types of irregularity exist if any element of the lateral load resisting system is not parallel to one of the orthogonal axes of the lateral load resisting system of the entire structure (see Figs. 3b and 4) (Raheem et al. 2018; Varadharajan 2014). Among the different horizontal irregularities, torsional irregularity is one and can be removed by increasing column sizes by bracing and adding the shear wall.
(a) (b)
Figure 3: a) Vertical geometric irregularity, b) horizontal irregularity.
Figure 4: Condition of stress concentration in the structure
73 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
2.5 Irregularities Limits as per Various Codes The irregularity limits for both horizontal and vertical irregularities based on Indian Standard Code IS 1893:2016 (Part 1), Eurocode (EC8:2004), Uniform Building Code (UBC 97), National Building Code of Canada (NBCC, 2005), International Building Code (IBC, 2003), Turkish Earthquake Code (TEC, 2007) and American Society of Civil Engineers (ASCE 7.05) standard can be summarized in the following Tables 1 and 2.
3. Structural Details and Modelling Approach 3.1 Description of the Buildings In this study, one regular and eight irregular RC building structures are taken for analysis. Among eight irregular building; four of them have geometrical irregularities, two have stiffness irregularities and rest of two buildings have mass irregularities. The details of information have been collected from the drawing by consultants, municipality drawing and a field survey of existing buildings in Pokhara Metropolitan city. The typical building model used in the study is the real model. The different irregularities in this study are formulated by modifying the real regular structure. Table 1: Irregularity limits prescribed by IS 1893:2016 (Part 1), EC8:2004, UBC 97, NBCC 2005
Type of irregularity IS 1893:2016 EC8 2004 UBC 97 NBCC 2005 Horizontal a) Re-entrant Ri ≤ 15% (Fig.2) Ri ≤ 5% Ri ≤ 15% corners rx > 3.33 eox b) Torsional dmax ≤ 1.2 davg ry > 3.33 eoy dmax ≤ 1.2 davg dmax ≤ 1.7 davg irregularity rx and ry > ls, Oa > 50% rx2 > ls2 + eox 2 Od > 50% c) Diaphragm Ry 2 > ls 2 + discontinuity Sd > 50% Sd > 50% eoy 2 Vertical Should not reduce a) Mass Mi < 2 Ma Mi < 1.5 Ma Mi < 1.5 Ma abruptly Si < 0.7Si+1 Or Si Si < 0.7Si+1 Or Si < Si < 0.7Si+1 Or Si < 0.7Si+1 Or Si < 0.8 b) Stiffness 0.8 (Si+1 + Si+2 + Si < 0.8 (Si+1 + < 0.8 (Si+1 + Si+2 (Si+1 + Si+2 + Si+3) Si+2 + Si+3) + Si+3) Si+3) Si < 0.7Si+1 Or Si Si < 0.7Si+1 or Si < < 0.8 c) Soft-storey 0.8 (Si+1 + Si+2 + Si < Si+1 (Si+1 + Si+2 + Si+3) Si+3) d) Weak story Si < 0.8Si+1 Si < 0.8Si+1 e) Setback Rd < 0.3Tw < 0.1 SBi < 1.5 SBa SBi < 1.3 SBa SBi < 1.3 SBa irregularity Tw at any level
74 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
Table 2: Irregularity limits prescribed by IBC 2003, Tec 2007 and ASCE – 7.05
Irregularity limits prescribed by IBC 2003, TEC 2007 And ASCE – 7.05 Type of irregularity IBC 2003 TEC 2007 ASCE – 7.05 Horizontal a) Re-entrant corners Ri ≤ 20% Ri ≤15% b) Torsional dmax ≤ 1.2 davg dmax ≤ 1.2 davg irregularity dmax ≤ 1.4 davg c) Diaphragm Oa > 33% Oa > 50% S > 50% discontinuity Vertical a) Mass Mi < 1.5 Ma Mi < 1.5 Ma Si < 0.7Si+1 Or Si < 0.7Si+1 Or b) Stiffness Si < 0.8 (Si+1 + Si+2 + Si < 0.8 (Si+1 + Si+2 + Si+3) Si+3) Si < 0.7Si+1 Or [ ki = ( i / hi) avr / Si < 0.7Si+1 Or c) Soft-storey Si < 0.8 (Si+1 + Si+2 + ( i+1 / hi +1) avr > 2.0 Si < 0.8 (Si+1 + Si+2 + Si+3) orη Δ Si+3) Δ Si < 0.6Si+1 Or d) Weak story Si < Si+1 [ ci = (Ae)i / < 0.80] Si < 0.7 (Si+1 + Si+2 + Si+3) e) Setback η SSBi < 1.3 SBa SBi < 1.3 SBa irregularity
The structural information such as the size and detailing of RC elements (beam and column), inter-storey height, type of steel reinforcement and grade of concrete is same for all building models. The material properties of the building are considered to be same in all the buildings as; a) compressive strength of concrete fc=20Mpa, b) reinforcing steel yield strength fy=415 MPa, c) roof live load =1.5 kN/m2 (nil for earthquake), d) roof and floor finish =1 kN/m2, e) floor live load = 2 kN/m2 (25% for earthquake). In this study, building models used in the analytical study are considered to have 7 bays with 4m width in X direction and 3 bays of 4m width in Y direction with 3m storey height. The regular building is kept regular throughout the seven story whereas some bays are removed in different story in case of irregular building. In IRR1 type irregular building one bays in X- direction is removed in each story of the buildings. In IRR2 type irregular building two bays in X- direction is removed from each two story of the building respectively. In IRR3 building 3 bays in X- direction are removed from G+ three story of the buildings while in IRR4 type irregular building 4 bays in X- direction are removed from G+ four story of the buildings. For IRR5 building weight equal to water of swimming pool is kept at the top floor followed by game house weight at (G+3) building in IRR7 building to create mass irregularity. Similarly, in IRR6 and IRR8 column to create stiffness irregularity there is removal of parking column at two different section C-C and E-E respectively (see Fig. 5). The parameters used for design of regular and irregular building models is presented in Table 3.
75 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
Table 3: Parameters used for design of regular and irregular building models.
Description of building model Parameters Data Unit Remarks Size of column 450x450 mmxmm Size of beam 350x350 mmxmm Slab thickness 150 mm Specific weight of concrete 25 kN/m3
Modulus of elasticity (infill) Em 5310 MPa
Modulus of elasticity (concrete) Ec 25000 MPa
Thickness of shear wall 250 mm
Figure 5: Plan of study building model.
3.2 Numerical Analysis The numerical analysis in this study is performed through pushover analysis. A pushover analysis is performed by subjecting a structure to a monotonically increasing load until structure become unstable or predefined displacement reached. Under incrementally increasing loads various structural elements may yield sequentially. Consequently, at each event, the structure experiences a loss of stiffness. Pushover analysis generate static pushover curve which plots an applied lateral load against displacement. The value of the lateral force incrementally increases with the transition of structure in the nonlinear zone, plastic hinge is formed. When analyzing frame structure, material non-linearity is assigned to discrete hinge location where plastic rotation occurs according to the FEMA 356 (2000), ATC-40 (1996), or other set of code-based or user defined criteria. Numerical analysis based on the bare frame building modelling with three dimensional models (see Fig. 7-10). Modelling of the structure is carried out by using finite element program SAP2000 (SAP 2000). Nonlinear behavior occurs within the frame elements at the location of plastic hinge (Nahavandi, 2015). Plastic hinges are the points on a structure where one expects cracking or yielding.
76 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
Figure 6: Force deformation curve with different performance level A generic component behavior curve is represented in figure 6. The points marked on the curve is expressed by the software SAP 2000 as follows:
. Point A is the origin . Point B represents yielding. No deformation occurs in the hinge up to point B, regardless of the deformation value specified for point B, the deformation (rotation) at point B will be subtracted from the deformations at points C, D, and E. Only the plastic deformation beyond point B will be exhibited by the hinge. . Point C represents the ultimate capacity for pushover analysis. However, a positive slope from C to D may be specified for other purposes. . Point D represents a residual strength for pushover analysis. However, a positive slope from C to D or D to E may be specified for other purposes. . Point E represents total failure. Beyond point E on the horizontal axis, if it is not desired that the hinge to fail this way, a large value for the deformation at point D may be specified. In the present study, the structures are modelled using default and user defined hinged properties. In the beam section, the moment curvature relation established which gives ultimate moment, yield moment, ultimate curvature and yield curvature and the values were normalized with respect to yield moment and yield curvature. The plastic hinge length is taken as half of the depth of beam (ATC-40, 1996). All the analysis is performed based on displacement-controlled procedure. The procedures adopted in this study can be summarized as:
. Application of 10% static lateral load induced due to earthquake at the CG of the building. . Developing (M- ) relationship for critical region of beam and column.
. Select control pointθ to see the displacement. . Apply full gravity load as a nonlinear static load pattern and gradually increasing lateral load, until the targeted displacement reached. . Developing hinge formation sequences and the base shear vs roof displacement (pushover curve) table.
77 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
(a) (b)
Figure 7: REG Model with a) front elevation and b) side elevation
(a) (b) (c) (d)
Figure 8: Irregularities in the buildings a) IRR1 model (up to 2nd floor), b) IRR2 model (up to 3rd floor), c) IRR3 model (up to 4th floor) and d) IRR4 model (up to 5th floor)
(a) (b)
Figure 9: Mass irregularities in the different floor level of building a) IRR5 model and b)IRR7 model
4. Analysis and Interpretation of Results 4.1 Pushover Curves From the pushover analysis, it is noticed that regular buildings have immediate occupancy level before the performance point whereas irregular building reached life safety level before the performance point. In regular building, plastic hinges are evenly distributed from bottom to top storey level whereas in irregular building plastic hinges are formed in some of the beam only in the same storey level reaching the plastic limit earlier. The column of irregular building reached life safety and collapse prevention earlier than the regular building. From the pushover curve, it is clearly seen that irregular building has slightly higher base shear capacity.
78 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
(a) (b)
Figure 10: Building with floating column a) IRR6 model and b) IRR8 model In regular building, the life safety level is reached from lower to higher storey in regular pattern whereas in irregular building (IRR1 and IRR2) the column of G+3 story reached life safety level. It is seen that when G+3 story column reached life safety level; G+1 and G+2 story is only in immediate occupancy level. Similarly, the results have shown that among the studied building types, regular building seem to have more capacity than any other steeped buildings. Regular building has higher stiffness compared to the buildings with floating columns. Irrespective of mass irregular building both of them have almost same capacity and have slightly less capacity than the regular building (see Fig. 11-12).
Figure 11: Comparison of base shear versus displacement of different regular and irregular building
4.2 Displacements Due to Pushover Analysis In figures 13, it is seen that the maximum displacement of regular building has more than that of the other irregular building. It is due to the higher mass up to top storey in regular building. The same condition is applied for the maximum top displacement in building model IRR3 and IRR4. These results justify that as the irregularity percentage increases in maximum displacement will decrease. However, due to torsional effects, the building model IRR1 has more displacement than IRR2 building model.
79 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
(a) (b)
Figure 12: Base shear versus displacement curve for: a) regular and b) irregular building structures
(a) (b)
Figure 13: Comparison of storey displacement (mm) of regular and geometric irregular building both in X and Y direction of loading, respectively.
As presented in figure 14, IRR5 and IRR7 building models have top displacement of 36.8mm and 35.4 mm, respectively. The displacement of regular building has maximum as compared to the irregular one. It is due to the unequal distribution of mass (due to swimming pool and game house) in higher floor level in irregular building. Similarly, building model IRR6 and IRR8 have the top displacement of 41.02 mm and 38.08 mm respectively. The building model of IRR6 have higher maximum displacement value at roof compared to REG and IRR8 building model. It is due to the fact that building model IRR8 have floating column at middle of the building showing symmetric while building model IRR6 have the floating column are apart from middle and resulting the torsional moment and increases deflection.
4.3 Comparison of Story and Story Drift of Structures As indicated in figure 15, story drift at the location of the steps building is changing abruptly compared to the regular building. The change in story drift is noticed in the location of change of steps. The maximum story drift of irregular stepped building has lower value compared to regular building. The story drift of
80 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] mass irregular building has almost same pattern. In case of floating column, the story drift of building model IRR 6 buildings is less than regular building and more than that of IRR8 buildings.
(a) (b)
Figure 14: Comparison of storey vs storey displacement in regular and irregular building model
(a) (b) (c)
Figure 15: Storey versus storey drift of irregular building models in push X.
4.4 Time History Analysis The most accurate procedure for structure subjected to strong ground motion is the time-history analysis. The pushover analysis is less onerous than nonlinear dynamic analysis since it does not require the monitoring of cyclic inelastic response of structural member and it avoids the dependence on the input motion (Landi, et al., 2014). Pushover analysis ensure a reliable structural assessment or design subjected to seismic loading in simplest and fastest way (Chaulagain et al., 2014; Themelis, 2008). The earthquake time history data is important for dynamic analyses of the structures. In the context of Nepal, real accelerograms records are not available sufficient for time history analysis. Due to lack of actual time history data in Nepal, the dynamic time history analysis is performed with El Centro time history data (Fig. 16). The analysis is good to represent the realistic behavior of structure (King 1998). From non-linear time history analysis, it is observed that the maximum top displacement of the regular building is 126.6 mm. The one step irregular building (IRR1) has displacement of 89.52 mm at the top while IRR2 have 94 mm and IRR3 have 95.98mm at the top, respectively. In this study, all the presented time-history results are peak-values. While comparing the result between the pushover and non-linear time history analysis the value of displacement of roof of the building given by non-linear time history is higher as compared to pushover
81 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain analysis but the pattern of displacement of both the regular and geometric irregular building is same that is REG building had more displacement followed by IRR4, IRR3, and so on (see Fig. 17).
Figure 16: Time history data for El Centro Earthquake From figure 18, it is seen that maximum displacement of the IRR5 at the top is more as compared to the regular and IRR5 building structure. The result shows that storey versus roof displacement curve has the same pattern but the value is more in time history analysis. The non-linear dynamic analysis shows the building model IRR6 have maximum displacement of 149.9 mm followed by IRR8 building with 130.8 mm. The pattern is same as that of pushover analysis. The value of displacement with time history analysis have higher as compared to non-linear pushover analysis. The building model IRR6 have higher deflection value. It is due to the removal of column for creating floating column. The removal of column for creating floating column is in unsymmetrical placed causing more torsion moment compared to IRR8 buildings.
Figure 17: Story versus displacement curve from non-linear time history analysis
82 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
(a) (b)
Figure 18: Story versus displacement curve of regular and irregular building model using non-linear analysis.
4.5 Comparison of Moment of Regular and Geometric Irregular Building At section DD, the moment of regular building model is greater than IRR1 building upto the 2nd story and slightly higher at 3rd story level. But, after the 3rd storey level the moment of IRR1 building is about 52% higher than regular building model. This variation is due to the higher level of irregularity in IRR1 building model. Similarly, IRR2 building model has about 13% less moment at lower storey as compared to regular building (Figs. 19-22).
(a) (b)
Figure 19: Plan and percentage increased or decreased of moment IRR1 with respect to regular building.
83 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
(a) (b)
Figure 20: Plan and percentage increased or decreased of moment IRR2 with respect to regular building.
(a) (b)
Figure 21: Plan and percentage increased or decreased of moment IRR3 with respect to regular building.
4.6 Torsion Effect on Irregular Building Structures The torsion in the building structure during earthquake is generated due to the unsymmetrical distribution of mass and stiffness along the height of building (Gokdemir et al., 2013; Cai and Pan, 2007; Neelavathi et al., 2017). In recent years, codes have given special provision to counter these effects by introducing accidental eccentricity which has to be considered during analysis and design. The torsional factor of studied building structure is presented in Table 4. From table, it is observed that regular building has almost same torsion factor in all the storey level. On the contrary, the high level of irregularity is clearly seen in the form of torsion appeared in the building models IRR1, IRR2, IRR3 and IRR4. The torsional effect is not observed in the structure with swimming pool (IRR5) and Game hose (IRR7).
84 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
(a) (b)
Figure 22: Plan and percentage increased or decreased of moment IRR4 with respect to regular building. (Note: D11, D32, D42 likely D represents the value of moment at section DD (see figure 5) second place numerical value represents the place of moment taken as per plan of the building and third place numerical value represents story levels. Here, the negative value represents that irregular building have more moment than regular building in percentages). Table 4: Torsion factor of studied building structures
Storey REG IRR1 IRR2 IRR3 IRR4 IRR5 IRR6 IRR7 IRR8 1 0.559 0.669 0.519 0.575 0.531 0.629 0.566 0.627 0.544 2 0.558 0.564 0.524 0.572 0.526 0.629 0.566 0.626 0.543 3 0.557 0.683 0.714 0.565 0.776 0.629 0.565 0.627 0.542 4 0.554 0.765 0.732 0.748 0.785 0.632 0.563 0.630 0.538 5 0.552 0.826 0.820 0.756 0.790 0.635 0.561 0.633 0.538 6 0.552 0.871 0.826 0.757 0.794 0.635 0.564 0.646 0.541 7 0.553 0.874 0.829 0.758 0.796 0.633 0.568 0.631 0.547
Note: Torsion factor =Deflection Umax / (Deflection U1max +Deflection U2max)
5. Conclusions This study highlights the effect of structural irregularities on seismic response of reinforced concrete building structures in Nepal. The geometrical irregularities are created by removing the bays in different floor levels and mass irregularities are studied by considering the swimming pool and game house at different floor levels. The results are analyzed analytically in terms of storey shear, storey displacement, drift and overturning moment. The effect of different irregularities is highlighted by comparing the results with regular structure. The main conclusions of the study can be summarized as:
. Based on the formation of plastic hinges, the columns of an irregular building reached life safety and collapse prevention level earlier than a regular building. The storey wise distribution of plastic hinges
85 Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021] Ghimire and Chaulagain
in beam and columns are distributed evenly in regular building model.
. The results indicated that the maximum top displacement of the regular building is 126.6 mm. Building model IRR1 have top displacement of 89.52 mm while IRR2 have 94 mm and IRR3 have 95.98mm at the top respectively. It reflects that higher the structural irregularities lower the storey displacement and vice versa. The displacement of regular building is more than the irregular building. It mainly depends on the amount of reduction of mass and stiffness in irregular structure.
. It can be observed that the moment of regular building model is greater than IRR1 building up to the 2nd story level. After 3rd storey level, the IRR1 building has about 52% higher moment than regular model. The regular building model generates the higher moment in lower stories. As the result of torsional effects, irregular building models induced higher moment in top stories.
Acknowledgements The authors would like to thank the School of Engineering, Pokhara University for providing the platform to conduct this research.
Conflict of Interests Not declared by authors.
References Amiri M and Yakhchalian (2020), “Performance of Intensity Measures for Seismic Collapse Assessment of Structures with Vertical Mass Irregularity”, Structures, Vol 24, 728-741. ASCE, “Minimum Design Loads for Building and Other Structures (ASCE/SEI 7-05)”, American Society of Civil Engineers, New York, 2005, U.S.A. ATC 40 (1996), Seismic Evaluation and Retrofit of Concrete Building, Applied Technical Council, California Seismic Safety Commission, Redwood City, California. Cai J and Pan D (2007), “New Structural Irregularity Assessing Index for Seismic Torsional Vibration”, Advances in Structural Engineering, 10(1): 73-82. Chaulagain H (2016), “Common Structural Deficiencies of RC Buildings in Nepal”, BSMC Journal of Local Development, Bharatpur Sub-Metropolitan City, Bharatpur, Chitwan 1(1). Chaulagain H, Rodrigues H, Spacone E, Guragain R, Mallik RK and Varum H (2014), “Response reduction factor of irregular RC buildings in Kathmandu Valley”, Earthquake Engineering and Engineering Vibration, 13 (3), 455–470. Dya AFC and Oretaa, AWC. (2015), “Seismic Vulnerability Assessment of soft story irregular building using pushover analysis”, Procedia Engineering, 125, 925–932. EC8. “Design of structures for earthquake resistance. General rules seismic actions and rules for buildings (EN 1998-1:2004), European committee for Standardization, Brussels, 2004. FEMA 356 (1997), Prestandard and Commentary for the Seismic Rehabilitation of Building, Federal Emergency Management Agency, Washington, DC. Gautam D and Chaulagain H (2016), “Structural performance and associated lessons to be learned from world earthquake in Nepal after 25 April 2015 Gorkha earthquake”, Engineering failure analysis, 68, 222–243. Gokdemir H, Ozbasaran H, Dogan M, Unluoglu E and Albayark U (2013), “Effects of Torsional Irregularity to Structures During Earthquakes”, Engineering Failure Analysis, 35, 713-717. IBC. “International building code 2003”, Illiniosis, International code council (ICC), 2002 Inc. IS 1893 (Part1):2016. Indian standard criteria for earthquake resistant design of structures, 5th revision. Bureau of Indian Standards, ManakBhavan, 9 Bahadur Shah ZafarMarg, New Delhi; 2016. Kostinakis K and Athanatopoulou A (2020), “Effect of In-plan Irregularities Caused by Masonry Infills on the Seismic
86 Ghimire and Chaulagain Journal of Engineering Issues and Solutions 1 (1): 70-87 [2021]
Behavior of RC Buildings, Soil Dynamics and Earthquake Engineering, Vol 129. Lee, H.-S., & Ko, D.-W. (2007). Seismic response characteristics of high-rise RC wall buildings having different irregularities in lower stories. Engineering Structures, 29(11), 3149–3167. Nahavandi H. (2015) Pushover Analysis of Retrofitted Reinforced Concrete Building, M.sc. Project Reports, Portland State University U.S.A. NBCC, “National Building Code of Canada 1995”, National Research Council of Canada, Ottawa, Ontario 1995. Neelavathi S, Shwetha KG and Mahesh Kumar CL (2017), “Torsional Behavior of Irregular RC Building under Static and Dynamic Loading”, Material Science Forum, 969, 247–252. Rahman S and Salik A (2016), “Seismic Analysis of Vertically Irregular building”, Current Science, 111(10). Raheem SEA, Ahmed MMM, Ahmed MM and Abdel-shafy AGA (2018), “Evaluation of plan configuration irregularity effects on seismic response demands of L-shaped MRF buildings”, Bulletin of Earthquake Engineering, 16, 3845–3869. Sadashiva V, Macrae G and Deam B (2009), “Determination of Structural Irregularity Limis- Mass Irregularity Example”, Bulletin of the New Zealand Society for Earthquake Engineering, 42(4). SAP 2000 V-20 (2009), Integrated Finite Element Analysis and Design of Structure Basic Analysis, Reference Manual, Berkeley (CA, USA): Computers and Structure Inc. Sarkar, P., Prasad, A. M., & Menon, D. (2010). Vertical geometric irregularity in stepped building frames. Engineering Structures, 32(8), 2175–2182. TEC 2007 Turkish Earthquake Code, Ministry of Public Works and Settlement, Specification Structures to Be Built In Disaster Areas, Part III Earthquake Disaster Prevention, Government of Republic of Turkey, Turkey, 2007. Themilis S. (2008), Pushover analysis for seismic assessment and design of structure, Ph.D. Dissertation, Heriot-Watt University, Edinburgh, Scotland. UBC, “Uniform building code (UBC 97)”, Intl. Conf. of Building offi cials (ICBO), Whittier, California, 1997. Varadharajan S (2014), “Study of Irregular RC Buildings under Seismic effect”, PhD Thesis, Department of Civil Engineering, National Institute of Technology, Kurukshetra, India. Varadharajan S, Sehgal VK and Saini B (2012), “Review of Different Structural Irregularities in Buildings”, Journal of Structural Engineering, 39 (5), 538–63.
87 Journal of Engineering Issues and Solutions
Numerical modelling of the sand particle flow in pelton turbine injector
Tri Ratna Bajracharya1, 2,*, Rajendra Shrestha 1, Ashesh Babu Timilsina 1, 2 1 Department of Mechanical Engineering, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Nepal 2Center for Energy Studies, Institute of Engineering, Tribhuvan University, Nepal *Corresponding email: [email protected] Received: January 05, 2021; Revised: March 17, 2021; Accepted: March 21, 2021
Abstract Pelton turbine is commonly employed high head impulse type turbine. Pelton turbine injector is an integrated part of the Pelton turbine machine which serves the purpose of converting entire pressure energy of water to kinetic energy and also regulates the water flowrate, with partial opening hence governing the power production. Severe erosion in Pelton turbine injector is reported from field setting research studies. Since the jet is atmospheric pressure jet, there is few chances of occurrence of cavitation hence it can be understood that impact of sand particles is the major cause of erosion. Furthermore, with turbine operating in partial flow condition, more erosion is reported in the needle of injector. For a long spear type injector, this study explores the cause of erosion by modeling the motion of the sand particle flow in steady state jet. For numerical modeling of the flow, the realizable k-epsilon model is used and for modeling the particle flow, the Discrete Phase Model (DPM) is used. Three different operating condition of the injector is considered and 77000 particles were injected to the flow domain. It is observed from the numerical simulations that the more sand particle hits the nozzle-needle surface with partial opening of the injector.
Keywords: Injector, needle, pelton turbine, sand particle motion.
1. Introduction One of the operational problems being faced by South-Asian hydropower plants is the heavy flow of sediments through the generating units causing severe erosion of the hydro mechanical parts (Thapa, 2004). The main cause of this heavy sediment inflow during a particular season (May–October) is the weak and fragile geometry of the Himalayas and its hills (Bajracharya et al., 2008). The heart of the power plant, the turbine’s performance deteriorates thus affecting the overall efficiency of the power plant as it erodes. The erosion is thus not only associated with financial losses due to less efficiency but also with increase in maintenance cost (Felix et al., 2016).
88 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
There are many perineal water sources, the rivers in Nepal and it is estimated that the economically feasible power potential of Nepal from these fresh flowing rivers is about 43,000 MW (Thapa, 2004). However, the rivers are loaded with enormous sediment during May-October season, due to heavy rainfall and the weak and fragile geometry of the Himalayas. The incessant and stochastic nature of the monsoon rain in the mountainous catchment area of these rivers are also responsible for the generation of an enormous amount of sediment in these rivers (Pandit, 2005; Poudel et al., 2012). A specific case study (Chhetri et al., 2016) of sediment loading of the Langtang river, one of the rivers located in central Nepal, reveals the sediment loading data as follow:
• Pre-monsoon: 37.69 tons per day • Post-monsoon: 11.52 tons per day, • Monsoon: 872.86 tons per day • Lowest value: 5.54 tons per day (winter season) The sediment loading is composed of minerals like quartz, feldspar, mica, granite, tourmaline, clay, organic matters etc. Detailed mineralogical analysis of the sediment content reveals the domination of hard minerals (hardness > 6 on Moh’s scale) like quartz and feldspar (Bajracharya et al., 2008; Neopane & Sujakhu, 2013; Poudel et al., 2012). Generally, the hardness of turbine material is about 6 on Moh’s scale. Hence, these hard minerals possess higher erosion potential. In addition, the sharp nature of their particles further adds up the erosion potential. Hydropower projects have well engineered desanding (or desilting) arrangements like gravel trap, settling basin. Despite such arrangements, sand particles of size < 2 mm (maximum size) passes through the turbine parts causing severe erosion. Despite heavy investments of time and finances to understand the erosion mechanism and its prevention techniques, the phenomenon is not yet clearly understood. It is well understood that beforehand prediction of such failures due to eroded parts will help engineers to predict the optimal repair time thus preventing damages, hazards and ensure operational safety of hydropower projects. Solid particle erosion is a term in tribology referring to the loss of solid materials containing fluid flow due to impingement of suspended solid matters in the flowing fluid. Among several variables affecting the solid particle erosion of slurry erosion, Javaheri et al. (2018) has classified the variables to four headings and have constructed a fish bone diagram (Javaheri et al., 2018). Broadly speaking, the erosion due to sand particles in hydro turbines are due to following causes (Brekke et al., 2010)
• Turbulence erosion: Erosion due to high turbulence in the boundary layers • Accelerated flow: Due to flow acceleration, particles are separated from flow and collide with the walls. • Vortex erosion: Formation of vortices due course of flow will trap particles within it thus causing erosion in region of vortex formation due to repeated impact on the wall Pelton turbine holds special place in Nepal for its application in both major hydropower plants as well as in micro hydro schemes for rural electrification. Review of field setting researches of Pelton turbine erosion reveals that the erosion prone areas in Pelton turbine are:
• Needle surface • Needle seat in nozzle • Bucket Surface • Bucket Splitter
89 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
Fig. 1 summarizes the erosion from two hydro power plants.
Figure 1: Typical erosion in Pelton turbine parts (On left: Mel Norway On Right: Bucket of Khimti Powerplant, Nepal) Image Courtesy: Neopane et al. 2011.
1.1 Erosion Studies of Pelton Turbine Needle Bajracharya et al. (2008) did a field setting research and laboratory research to reveal the empirical relation between loss of efficiency due to sand particle led erosion of Pelton turbine. The authors measured the eroded needle profile after fixed operational hours of the runner (see Fig. 2).
Figure 2: Erosion of Pelton turbine needle at Chilime HPS, Nepal (Adapted from Bajracharya et al. 2008) The authors observed that the needle erodes more between the nozzle exit and needle tip. Needle tip itself is free from erosion. The authors explain the observed phenomenon based on relation between erosion and 3 particle velocity ( Vp ). The authors explain that with the decreasing cross-sectional flow area, the flow velocity increases (continuityα equation) thus increasing the rate of erosion as flow approaches the needle tip and near the tip, theε αforce on the needle surface is low and jet flows under atmospheric pressure hence needle tip doesn’t erode much. The actual eroded needle is shown in Fig. 3. The eroded needle is seen to be have in depth erosion.
90 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
Figure 3: Needle damage due to erosion at Chilime HPS, Nepal Neopane et al. (2011), in review of erosion of hydraulic turbines have revealed the images of eroded nozzle and needle of Andhikhola HPS and Khimti HPS of Nepal. From the Fig. 4, it can be observed that there is uniform wear of the nozzle (in the needle seat) and scaly erosion is seen in the Needle surface.
Figure 4: Erosion in nozzle and needle (On left, erosion in nozzle ring of Andhi Khole HPP, Nepal and needle erosion in Khimti HPP, Nepal (Adapted from Neopane et al. 2011) Morales et al. (2017) experimentally studied about the erosion of Pelton turbine injector of Chivor hydroelectric plant, Columbia. The authors have also developed a test rig to study and develop empirical relation of nozzle-needle coatings and wear. The needle erosion of Chivor hydropower station has in depth erosion and is similar to that observed by Bajracharya et al. 2008. ANDRITZ hydro (Karandikar, 2015) performed the field test of HVOF Coatings to combat hydro abrasive erosion. The study reveals images (see Fig. 5) of coated and uncoated needles in a two nozzle Pelton turbine and concludes that the coated needle has four times longer life than the coated needle.
Figure 5: Erosion study in coated needle (Image Courtesy: Andritz Hydro)
91 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
The nature of erosion observed in injector is not similar as large scale erosion is seen in the needle surface and there is breakage of the needle tip. As classified in literature, all types of erosion are observed in the needle. It can be understood that the difference in erosion nature is due to site specific condition of the sediment, operating condition and of course the injector design. The Pelton turbine injector not only converts the pressure energy of the water to the kinetic energy but also controls the flowrate of the water, the governing action. Among the turbine components, the first one to interact with the sediment-laden water is the injector which thus suffers from erosion wear. The direct effect of the erosion is change in shape of the nozzle and needle thus altering the jet dispersion, jet velocity thus the energy content of the jet. In an experimental investigation about the jet dispersion, it has been concluded that poor jet quality is major cause of cavitation in bucket. In this study, the motion of sand particle in the jet is modeled using a commercial computational fluid dynamics package for correlating the motion of sand particle and erosion prone areas as observed from field setting researches.
2. Methodology The flow region (flow domain) is modeled in Three-Dimensional Computer Aided Design (3D CAD) software and then discretized (meshed) using ICEM. The structured mesh was then exported to a commercial CFD code where, the boundary condition and solver settings for fluid and flow properties, convergence criterion were defined. The numerical simulation was carried out using SIMPLE algorithm for pressure velocity coupling, realizable k epsilon model for modeling turbulence. The realizable k epsilon model is modification to the traditional k epsilon model by giving a benefit of better performance in simulation of planar and round jets (ANSYS Inc, 2013), rotation, recirculation and streamline curvature (Wasserman, 2021). The Pelton turbine jet has streamline curvature in at nozzle exit and is expanding jet (Bajracharya et al. et al., 2019; Zhang, 2016).
Figure 6: Flow domain.
92 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
Fig. 6 is the flow domain, the dimensions of the injector are those of a 2 kW model Pelton turbine test rig (Bajracharya et al., 2007) located at Center for Energy Studies, Institute of Engineering. To better capture the viscous layer effects, the fine mesh was used in the needle wall and in the nozzle wall. The axially symmetric nature of the flow domain made the choice of O-grid optimum in this case. Fig. 7 and 8 give the boundary conditions and longitudinal section of the mesh.
Figure 7: Boundary and boundary conditions
Figure 8: Meshing details The numerical value at pressure inlet was set at 3 bar (gauge) while the pressure outlet was assigned value of 0 bar (gauge). The boundary values were chosen based on the measured value from Pelton turbine test rig at Center for Energy Studies (Bajracharya et al., 2007). The solver settings used is summarized in table 1. Table 1: Summary of solver settings
Solver Three-Dimensional, Steady State, Pressure Based Time discretization Steady State Discretization scheme Second Order Turbulence model Realizable k-epsilon Flow Single Phase, Discrete Phase Modeling for Modeling Particle Flow Convergence criteria RMS, Scaled residual target = 10–6 Hardware Processor: Intel Core i7 @ 3.6 GHz (8 CPUs). Memory: 32 GB
As the flow simulation results were converged, then 77000 sand particles from were introduced normally from inlet to the flow. In the validated flow model, particles with maximum diameter 2 mm were introduced and nature of its flow along with the jet was studied for three different condition. The particle size distribution of the sand sample used for the study is shown in Fig. 9.
93 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
• Fully-open injector • Quarter-closed injector • Half-closed injector
Figure 9: Particle size distribution of sand sample (Rosin Rammler Fitted Curve) For different injector openings, the motion of sand particles was studied. For estimation of the number of particles that hit a particular wall, the wall was set to trap the particles and number of particles that were trapped were counted and subtracted with the number of particles trapped with wall set to rebound condition however after detailed observations of the particle flow, it was seen that few numbers of particles were rebounding after hitting the nozzle and then hitting the needle in partial opening condition.
3. Results and Discussion
3.1 Flow Modeling The steady state jet was modeled and the flow velocity, velocity profile and jet geometry are compared with experimentally observed one (Fig. 10). The details are published in Bajracharya et al. (2019) and is briefly summarized here.
94 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] 3.1.1 Velocity profile 3.1.1 Velocity
95 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
The obtained velocity profile (summarized in Fig. 10) reveals that at needle tip, the jet velocity is zero and as the jet moves forward the velocity in the region ahead of the tip increases. The jet velocity is maximum between the needle tip area and jet surface making the overall velocity profile of “B” shape. Zhang et al. (2016) experimentally studied the jet velocity distribution (using Laser Doppler Anemometry. The jet velocity obtained from computational simulations and experimental modeling are in good agreement
3.1.2 Jet geometry An ideal jet is supposed to have uniform velocity distribution in each cross section and the jet diameter is constant without any contraction or expansion (Zhang, 2016). Experimental investigations of the jet (Staubli & Hauser, 2004; Zhang & Casey, 2007) reveal that the real jet deviates from this ideal form. The real jet first contracts after it exits from the nozzle and then it expands. The geometry of jet flow obtained also matches with the experimental observation. Fig. 11 shows the jet geometry for half open injector. The jet contraction ahead of the nozzle exit can be seen. This contraction is associated with jet energy loss.
Figure 11: Jet geometry for half open injector (Right image courtesy of Staubli et al. 2004)
3.2 Particle Flow Modeling 3.2.1 Fully-open injector Fig. 12 shows the motion of the sand particles in fully open injector. A close observation of the nozzle exit area reveals none particle interacting with the needle surface while few particles exiting near the nozzle surface are hitting the needle seat area. Hence, the direct impingement of the particles in the needle seat causes its wear. Further, it can be said that the uniform hitting of the particles will cause even wear of the needle seat. The simplified representation of the trajectories is presented in Fig. 13.
Figure 12: Particle motion for fully open injector
96 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
Figure 13: Particle Trajectories for fully open injector 3.2.2 Quarter-closed injector Fig. 14 shows the motion of the sand particles in quarter closed injector. The particles in addition to hitting the needle seat partly hits the needle surface just upstream of the tip. The impingement angle is seen to be between 0 to 20 degrees. The simplified representation of the trajectories is presented in Fig.15.
Figure 14: Particle motion for quarter closed injector
Figure 15: Simplified particle trajectories for quarter closed injector
97 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
3.2.3 Half-closed injector In case the injector is half closed, the particle strikes in region between the needle tip and nozzle exit not uniformly but randomly on a specific region (see Fig.16 and 17). The needle surface-sand particle interaction increases more in comparison to the quarter closed condition. This shows that more erosion can be expected if the turbine is operated on partial load. This supports the findings of Bajracharya et al. (2008) who stated that erosion is directly related to the loading condition of the machine and it is dangerous with regards to erosion to operate the machine in part load.
Figure 16: Particle motion for half-closed injector
Figure 17: Simplified particle trajectories for half-closed injector The detailed flow of the particle in the flow domain along with flow region classification is shown in Fig.18. With manual counting, it was seen that among 100 particle trajectories, 19 particles hit the nozzle wall and among these 19, 12 particles were hitting the needle after rebound and 10 particles were hitting the needle wall without rebound from the nozzle exit.
98 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
Figure 18: Flow velocity and particle motion As seen from the Fig. 18, The heavier particles are settling in the low velocity region just upstream of nozzle beginning. The particle motion converges to a point ahead of the needle tip then disperses. Table 2: Summary of particle – wall interactions for different cases
Number of Particles hitting (among 100 particles) S.N. Case Nozzle Needle Needle-Nozzle 1. Fully Open Injector 15 - - 2. Quarter Open Injector 17 - 14 3. Half Open Injector 19 10 12
3.3 Erosion Erosion models are deduced by either theoretically by energy balance for particle-target wall collision or experimentally by formulating empirical relations or by dimensional analysis (Bajracharya et al., 2020). The first proposed erosion model to evaluate progressive loss of material was by Finnie in 1960 (Finnie, 1960). The erosion model proposed by Finnie is more suited for ductile materials, where the major variables are impact angle and velocity. Among several theoretical erosion models proposed thereafter, impact angle and velocity are those variables which are mostly considered for evaluation of progressive loss of material. Figure 19 shows the erosive behavior of ductile and brittle material. The Pelton turbine injector is manufactured by steel, a ductile material. From Fig. 12-17 it can be observed that the impact angle of most particles hitting either the nozzle or needle is 10 to 30 Degrees for which Fig. 19 shows maximum erosion.
99 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
Fig. 18 shows the particle trajectories for half open injector. As observed from field setting research there is asymmetrical nature of erosion of the needle (Bajracharya et al., 2008) i.e., needle erosion is mostly not uniform and less erosion is seen in the back side of the needle (the part attached to the needle stem). It can be deduced from the particle motion that a greater number of particles (also heavier) travel from the downward side of the needle causing more erosion on the downward part of the needle. Upon division of the flow domain on the basis of flow velocity (hence particle velocity), it can be seen that the backward side of the needle lies in low velocity region hence few or less erosion is observed in the backward side. In addition, the impact angle of particles is near to right angle thus causing minimal erosion damage. Table 2 summarizes the number of particles (among 100 particles) that hit selected surfaces. For fully open injector, the particles hit only the nozzle at exit only. The region is also maximum velocity region (see Fig. 18), these high speed particles when hit the nozzle at the exit, cause mass of nozzle to erode. Among 100 particles, 15 particles hit the nozzle at exit area. The number of particles hitting nozzle is fairly constant for other two cases with slight increase for partially open injectors. For the case of needle, no particles hit the needle for fully open injector while with partial opening, some particles that hit the nozzle and rebound towards needle were hitting the needle before leaving the injector domain. The number of particles hitting the needle has increased with partial opening and for half open injector, 22 particles were found to be hitting the needle. The hitting area was area ahead of nozzle exit. The impact angle of the particle was around 10 to 30 Degrees for region near tip and was greater than 30 Degrees for region farther from the tip. Hence as seen from field setting researches (Neopane et al., 2011; Thapa, 2004) more erosion can be expected at near tip region with less erosion upstream of the needle tip.
Figure 19: Erosion behavior of materials (Hutchings, 1981)
4. Conclusions With single phase steady state approach, the motion of sand particles as it passes from long spear type Pelton turbine injector is studied with DPM Model. The major objective was to study the trajectory of particles to understand erosion behavior of the injector. With nozzle and needle combining to form an injector assembly,
100 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] as seen from field setting research, there is ring type erosion around the periphery of nozzle and for needle, more erosion is seen towards the tip with decreasing erosion rate in upstream region. It was seen that number of particles hitting the nozzle exit is uniformly distributed hitting with fairly constant (15, 17 and 19 for fully open, quarter open and half open injector respectively) and the impact angle is 10 to 30 degrees. From velocity distribution, it was seen that the nozzle exit is the maximum velocity area. Hence for nozzle erosion is relatively constant no matter of operating position. For needle, it was observed that number of particles hitting the needle surface in increasing (0, 14 and 22 for fully open, quarter open and half open injector respectively) and the impact angle is 10 to 30 degrees for particles near the tip area and more than 30 degrees for upstream area. From velocity distribution, it was seen that the nozzle exit is the maximum velocity area. Hence for nozzle erosion is relatively constant no matter of operating position. Further heavier particles were flowing from the downward part of the injector which may cause much erosion damage on downside thus asymmetrical erosion of the needle. The present work is only limited to numerical modeling of the particle flow and relating the motion of particles seen and erosion behavior from field setting research studies. Further studies should consider secondary flows, and particle-particle interactions for better modeling of particle trajectories. Experimental modeling of the particle flow shall further enable to better understand the particle motion. Conclusions from present study shall form a base for further advancement of studies relating to particle motion modeling and geometrical modification and optimization of injector.
Acknowledgements University Grants Commission, Sanothimi, Bhaktapur funded this research under faculty research grant (FRG-74/75-Engg-04). Technical support from Center for Energy Studies (Institute of Engineering), Chilime Hydropower Plant and Robotics Club (Pulchowk Campus) are highly acknowledged.
Conflict of Interests Not declared by authors. Appendix: Velocity Profiles for Different Position of Jet
Development of Axial Velocity of Jet for Fully Open Injector
At Needle Tip
101 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
At 5 mm Ahead of Needle Tip At 20 mm Ahead of Needle Tip
At 40 mm Ahead of Needle Tip At 80 mm Ahead of Needle Tip
Development of Axial Velocity of Jet for Quarter Open Injector
At 5 mm Ahead of Needle Tip
102 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
At 5 mm Ahead of Needle Tip At 20 mm Ahead of Needle Tip
At 40 mm Ahead of Needle Tip At 80 mm Ahead of Needle Tip
Development of Axial Velocity of Jet for Half Open Injector
At Needle Tip
103 Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021] Bajracharya et al.
At 5 mm Ahead of Needle Tip At 20 mm Ahead of Needle Tip
At 40 mm Ahead of Needle Tip At 80 mm Ahead of Needle Tip
References ANSYS Inc. (2013). ANSYS Fluent Theory Guide (Release 15.0). http://www.pmt.usp.br/ACADEMIC/martoran/ NotasModelosGrad/ANSYS Fluent Theory Guide 15.pdf Bajracharya, T.R., Acharya, B., Joshi, C. B., Saini, R. P., & Dahlhaug, O. G. (2008). Sand erosion of Pelton turbine nozzles and buckets: A case study of Chilime Hydropower Plant. Wear, 264(3–4), 177–184. https://doi.org/10.1016/J. WEAR.2007.02.021 Bajracharya, T. R. (2007). Efficiency Deterioration in Pelton Turbines due to Sand-Particle-Led Bucket Erosion. Tribhuvan University. Bajracharya, T. R., Shrestha, R., & Timilsina, A. B. (2019). A Methodology for Modelling of Steady State Flow in Pelton Turbine Injectors. Journal of the Institute of Engineering, 15(2), 246–255. https://doi.org/10.3126/jie.v15i2.27674 Bajracharya, T. R., Shrestha, R., & Timilsina, A. B. (2020). Solid Particle Erosion Models and their Application to Predict Wear in Pelton Turbine Injector. Journal of the Institute of Engineering, 15(3), 349–359. https://doi.org/10.3126/ jie.v15i3.32221 Brekke, H., Wu, Y. L., & Cai, B. Y. (2010). Design of Hydraulic Machinery Working in Sand Laden Water. 155–233. https:// doi.org/10.1142/9781848160026_0004 Chhetri, A., Kayastha, R. B., & Shrestha, A. (2016). Assessment of Sediment Load of Langtang River in Rasuwa District, Nepal. Journal of Water Resource and Protection, 08(01), 84–92. https://doi.org/10.4236/jwarp.2016.81007
104 Bajracharya et al. Journal of Engineering Issues and Solutions 1 (1): 88-105 [2021]
Karandikar, D.A. (2015). HVOF Coatings to combat hydro abrasive erosion International Conference On Hydropower Sustainable Development Dehradun 6 th. http://www.ahec.org.in/ICHSD_2015/Presentations/Authors_ presentation_in_PDF/D_A_Karandikar.pdf Felix, D., Albayrak, I., Abgottspon, A., & Boes, R. M. (2016). Hydro-abrasive erosion of hydraulic turbines caused by sediment - a century of research and development. IOP Conference Series: Earth and Environmental Science, 49(12), 122001. https://doi.org/10.1088/1755-1315/49/12/122001 Finnie, I. (1960). Erosion of surfaces by solid particles. Wear, 3(2), 87–103. https://doi.org/10.1016/0043-1648(60)90055-7 Hutchings, I. M. (1981). A model for the erosion of metals by spherical particles at normal incidence. Wear, 70(3), 269–281. https://doi.org/10.1016/0043-1648(81)90347-1 Javaheri, V., Porter, D., & Kuokkala, V.-T. (2018). Slurry erosion of steel – Review of tests, mechanisms and materials. Wear, 408–409, 248–273. https://doi.org/10.1016/J.WEAR.2018.05.010 Morales B, A. M., Pachón, I. F., Loboguerrero U, J., Medina, J. A., & Escobar G, J. A. (2017). Development of a test rig to evaluate abrasive wear on Pelton turbine nozzles. A case study of Chivor Hydropower. Wear, 372–373, 208–215. https://doi.org/10.1016/J.WEAR.2016.11.003 Neopane, H. P., Gunnar Dahlhaug, O., & Cervantes, M. (2011). Sediment Erosion in Hydraulic Turbines. Global Journal of Researches in Engineering Mechanical and Mechanics Engineering, 6, 17–26. https://pdfs.semanticscholar.org/f219/ b5289d91874f8d13090063c34688ebebe978.pdf ?_ga=2.265510923.564185937.1556240431-87428635.1550801047 Neopane, H. P., & Sujakhu, S. (2013). Particle Size Distribution And Mineral Analysis Of Sediments In Nepalese Hydropower Plant: A Case Study Of Jhimruk Hydropower Plant. In Kathmandu University Journal of Science, Engineering and Technology (Vol. 9, Issue I). http://www.ku.edu.np/kusetjournal/vol9_no1/3_Hari_Prasad_Neopane_final.pdf Pandit, H. P. (2005). Nepalese Journal of Engineering. In Nepalese Journal of Engineering (Vol. 1, Issue 1). https://www. nepjol.info/index.php/NJOE/article/view/24 Poudel, L., Thapa, B., Shrestha, B. P., Thapa, B. S., Shrestha, K. P., & Shrestha, N. K. (2012). Computational and experimental study of effects of sediment shape on erosion of hydraulic turbines. IOP Conference Series: Earth and Environmental Science, 15(3), 032054. https://doi.org/10.1088/1755-1315/15/3/032054 Staubli, T., & Hauser, H. P. (2004). Flow visualization - a diagnosis tool for pelton turbines. 1, 1–9. Thapa, B. (2004). Sand Erosion in Hydraulic Machinery [Fakultet for ingeniørvitenskap og teknologi]. https://brage.bibsys. no/xmlui/handle/11250/231204 Wasserman, S. (2021). Engineering.com - Choosing the Right Turbulence Model for Your CFD Simulation. https://www. engineering.com/story/choosing-the-right-turbulence-model-for-your-cfd-simulation Zhang, Z, & Casey, M. (2007). Experimental studies of the jet of a Pelton turbine. Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy, 221(8), 1181–1192. https://doi.org/10.1243/09576509JPE408 Zhang, Z. (2016). Pelton turbines. In Pelton Turbines (1st ed.). Springer International Publishing. https://doi.org/10.1007/978- 3-319-31909-4
105 Journal of Engineering Issues and Solutions
Effects of source digital elevation models in assessment of gross run- off-river hydropower potential: A case study of West Rapti Basin, Nepal
Sunil Bista1, Umesh Singh2,3,*, Nagendra Kayastha3, Bhola NS Ghimire4, Rocky Talchabhadel5 1 Pulchowk Campus, Institute of Engineering (IOE), Tribhuvan University, Lalitpur, Nepal 2 Hydro Lab Pvt. Ltd., Lalitpur, Nepal 3 Geomorphological Society of Nepal (GSN), Kathmandu, Nepal 4 Center for Applied Research and Development (CARD), IOE, Tribhuvan University, Lalitpur, Nepal 5 Texas A &M AgriLife Research, Texas A &M University, EI Paso, Texas, USA *Corresponding email: [email protected] Received: January 27, 2021; Revised: March 20, 2021; Accepted: March 28, 2021
Abstract Advancements in Geographical Information System (GIS), Remote Sensing (RS) technology, hydrologic modeling and availability of wider coverage hydrometeorological data have facilitated the use of GIS and hydrological modelling tools in studies related to hydropower potential. Digital Elevation Model (DEM) is the primary data required for these tools. They have become more accessible and many are freely available. These DEMs have different resolution and their errors vary due to their primary data acquisition techniques and processing methods. However, their effects on the hydropower potential assessment are less investigated. This study evaluates the effects of 6 freely available DEMs: ALOS 12.5 m, SRTM 90 m, SRTM 30 m, ASTER G-DEM version-3 30 m, AW3D 30 m and Cartosat-1 version-3 30 m on the Gross Run-off-River Hydropower Potential (GRHP) assessment, using GIS and hydrological modelling tools. West Rapti River (WRR) basin in Nepal was chosen for the case study. Soil and Water Tool (SWAT) hydrological model, coupled with GIS was used to discretize the WRR basin into several sub-basins/streams. Flow at the inlet and outlet of streams were estimated from the SWAT model whereas the topographic head was extracted from the DEMs. The GRHP of the streams were computed using the estimated stream flow and the topographic head for flows at 40% to 60% Probability of Exceedance (PoE). The total potential of the basin was computed by summing up the potential of all streams. The GRHP of WRR basin for flows at 40% PoE was estimated as 512 MW for ALOS 12.5 m resolution DEM, referred as a base case in this study. The GRHP estimated from the remaining DEMs showed the variation of less than 6% compared to the base case. The topographic head was found to be sensitive with respect to the DEM resolution and the highest variations were observed in the main river channels.
Keywords: Digital Elevation Model (DEM); GIS; hydropower potential; SWAT; West Rapti River Basin
106 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
1. Introduction Nepal is a country with high per capita hydropower potential (Hoes et al., 2017). It receives an average annual precipitation of about 1530 mm spatially varying from below 500 mm to above 5000 mm (Talchabhadel et al., 2018). The total run-off generated is estimated as 225 billion cubic meters annually flowing through more than 6000 rivers, from north to south, into Ganges River in India (WECS, 2005). The rugged hill and mountains cover more than 80 % of the land. The topographic elevation of the country descends from above 8000 m asl in the north to 60 m asl in the south, within a short stretch of about 145 to 241 km (Chaudhary, 2000), providing steep slope to the rivers, favorable for hydropower development. Energy is one of the most important strategic commodities for the socio-economic development of any country (Dhungel, 2016). However, the current hydropower installed capacity in Nepal is only 1,278 MW (NEA, 2020), far below it's potential. The energy sector is still dominated by traditional sources of energy, where biofuels and waste contribute 82 % of the total primary energy supply (Hussain et al., 2019). Lack of natural gas and oil reserves also means that Nepal imports these resources creating economic as well as environmental burden (Gunatilake et al., 2020). So, hydropower is readily seen as a more environmentally friendly and sustainable alternative to these traditional energy sources and natural fuels in Nepal. Besides, hydropower is considered the most advantageous clean source of renewable energy, which is not affected by the fluctuating fuel prices (Singh & Singal, 2017). The Government of Nepal has also put forward ambitious plan for rapid hydropower development within a decade (Bhatt, 2017) to cater the current and future electricity demand. So, a reliable hydropower potential assessment is required for formulating the hydropower development plans and policies. A Gross Run-of-River Hydropower Potential (GRHP) is the theoretical sum of stream flow energy (Arefiev et al., 2015). The studies on assessment of the hydropower potential in Nepal are gradually increasing. Shrestha (1966) first conducted the GRHP study of Nepal, during his doctoral study. The potential of the country was estimated as 83, 500 MW at mean annual flow, based on the technologies and limited hydro- meteorological data available at that time. Only river basins larger than 300 km2 were considered in the study (Shrestha, 2016) due to the coarse resolution of the topography data. The advancement in Geographical Information System (GIS), Remote Sensing (RS) technology, hydrologic modeling and availability of wider coverage hydrometeorological data have enabled the use of GIS and hydrological modelling in hydropower potential related studies. Arfiev et al., (2015) provides a summary of their worldwide application in estimation of hydropower potential. Most of the recent hydropower potential studies in Nepal are also based on these methods. Jha (2010), using the Catchment Area Ratio (CAR) method to estimate river discharge at ungauged locations and the Shuttle Radar Topographic Mission (SRTM) Digital Elevation Model (DEM) resampled at 100 m resolution, estimated the total theoretical ROR hydropower potential of Nepal as 53,836 MW at flow of 40 % Probability of Exceedance (PoE) and 80% efficiency. Bajracharya (2015), using Soil Water Assessment Tool (SWAT) hydrological model and Advanced Space borne Thermal Emission and Reflection Radiometer Global-DEM (ASTER G-DEM) of 30 m resolution, estimated the GRHP of Nepal as 103,341 MW at mean annual flow. The recent study on hydropower potential of Nepal carried out by WECS, (2019) estimated the GRHP as 72,544 MW at 40% PoE, using the Hydrologic Engineering Center-Hydrological Modelling System (HEC-HMS) model and ASTER G-DEM of 30 m resolution. Prajapati (2015), using HEC-HMS and reclassified DEM of 100 m resolution, estimated the ROR hydropower potential of Karnali River basin as 14,150 MW at 40% PoE, 84% efficiency and 10% seasonal outage and riparian release. Aryal et al. (2018) used the SWAT model and SRTM DEM of 90 m resolution to identify the potential sites and estimate the total power potential of the Bagmati River basin at flows of different PoE.
107 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
The other application of GIS in hydropower potential study in Nepal include the rapid spotting of RoR hydropower potential location in Bhote Koshi river basin (Kayastha et al., 2018); identification of the hydropower potential sites and sensitivity analysis of estimated potential under climate change scenarios using RS and Snowmelt Runoff Model (SRM) in Balephi river basin (Kumar et al., 2017); assessment of the optimal distribution of RoR hydropower site, design of a local grid and verification of the resilience of the grid under climate change scenarios in Dudh Koshi river basin (Bocchiola et al., 2020); identification of economically feasible ROR hydropower sites (Magaju et al., 2020). DEM is the basic data in the hydropower potential studies using GIS based approaches with hydrological modelling. The above studies have used different source DEMs. DEM data have become more accessible due to the continuous advancement in GIS and RS technology. Moreover, many are freely available and higher resolution DEMs are also expected to be available in the public domain in the future. These DEMs have different resolution and their errors vary due to their primary data acquisition techniques and processing methods (Mukherjee et al., 2012). In the GRHP studies, the DEMs will mainly influence topographic head of the river reaches. The approach of determining the accuracy of the DEMs based on the elevation of the field control points (Mukherjee et al., 2012) may not be very effective in these studies, mainly because the control points are located in the ridges whereas the hydropower potential has to be evaluated in the river flowing in the valleys. The accuracy is also strongly influenced by the morphology of the terrain (Mukherjee et al., 2012). Despite notable contributions from above hydropower potential related studies, there is a scope to evaluate the effect of different source DEMs on hydropower potential assessment. This study evaluates the effects of six publicly available DEMs on estimation of GRHP, considering only the influence in topographic head. The DEM resolution also affect the drainage area and other topographical parameters like slope etc., thus influencing flow estimation of hydrological model. Higher resolution DEMs are reported to result in higher stream flows (Chaubey et al., 2005; Li & Wong, 2010; Nazari-Sharabian et al., 2020). However, the influence of the DEMs in the hydrological model are not addressed in this study. SWAT hydrological model was used to discretize basin and simulate stream flow at different river reaches. SWAT was selected because it is a physically-based semi-distributed hydrological model available in a public domain. It has been widely used in different watersheds across Nepal to simulate basin hydrology, land-use change, climate change impacts, etc. (Pandey et al., 2019, 2020; Pokhrel, 2018; Talchabhadel et al., 2021). SWAT integrated with GIS have also been used across the various watershed for the identification of hydropower potential sites and estimation of basin potential (Ali et al., 2020; Guiamel & Lee, 2020; Kusre et al., 2010; Pandey et al., 2015; Sammartano et al., 2019). The GRHP of the discretized stream reaches were estimated based on the difference in elevation derived from the DEM and the discharges at different PoE simulated by the SWAT model. The potential is first estimated using ALOS 12.5 m (highest resolution) DEM and compared with the results of the other five DEMs to evaluate their effects on GRHP assessment. We select West Rapti River (WRR) basin as a representative catchment in the mid-hills of Nepal for the case study. Although the current study is limited to a single river basin, the approaches and the findings will be applicable for other river basins across the country.
2. Materials and Methods 2.1 Study Area The WRR basin, with a catchment area of about 6380 km2 is located in the southwestern parts of Nepal. The latitudes and 81°40′10″ to 83°10′55″ longitudes. The river ״basin is bounded within 27°45′10″ to 28° 35′35 originates from the middle mountains of Nepal and descends south from the rugged highlands as shown in Fig. 1. The river is named WRR after the confluence of Madi River with Jhimruk River. Jhimruk and Madi,
108 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] the major tributaries, are both rain-fed rivers; therefore, monsoon rainfall and groundwater discharge are the main contributors to the runoff (Talchabhadel & Sharma, 2014).
Figure 1: Location of a network of hydro-meteorological stations across the West Rapti River (WRR) basin. The shaded area is the topography of the WRR basin. Inset at the top left shows the location of the WRR basin in the country’s map. The length of the main river is 260 km from the origin to the basin outlet. The southern part of the basin, mostly below 500 m above sea level (asl), is dominated by low land, representing about 34 % of the total basin area. The basin is wide at its headwater areas and narrow at the lower part. A diverse climate is observed in the basin due to the variation in elevation ranging from 70 m to 3,609 m asl. The upper part of the basin has a temperate climate, whereas the lower part has a tropical climate (Karki et al., 2016). The temperature along the basin varies from 45°C in the summer in the lower part and falls below 3°C during winter in the upper part of the basin. The basin receives mean annual precipitation of 1577 mm of which more than 80 % occurs during monsoon season (June to September). Since only 1 % of the basin area lies above 3000 m asl, snowmelt contribution to the streamflow can be neglected (see Fig 2.). Therefore, base flow is an important contributor to the river discharge in the non-monsoon season. Jhimruk hydropower plant, with an install capacity of 12.5 MW, is the only operational hydropower project in the basin. Eight RoR hydropower projects, with total estimated installed capacity of about 20 MW, are currently under study (DoED, 2021). Two storage projects, Upper Jhimruk and Madi Khola with an installed capacity of 100 MW and 156 MW, respectively, are also under study. Similarly, Madi-Dang Diversion project and Naumure multipurpose project with installed capacity of 377 MW and 256 MW, respectively, are planned for development.
109 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
2.2 Digital Elevation Model (DEM) Six DEMs of the study area are freely available in the public domain. These DEMs include SRTM (30 m and 90 m resolution)1, ASTER GDEM version-3 30 m resolution2, AW3D 30 m resolution3, Advanced Land Observing Satellite (ALOS) 12.54 m resolution and Cartosat-1 version-3 30 m resolution5. DEM is a crucial input unit for the hydrological models since its resolution affects computation time and model prediction by affecting watershed characteristics such as area, shape, length and slope (Nazari-Sharabian et al., 2020). The delineated watershed and predicted stream network are progressively less accurate as DEM resolution decreases (Chaubey et al., 2005). A large error in the delineated watershed area can yield a large error in model prediction since the runoff generated is directly influenced by the watershed area. Therefore, to minimize the error in streamflow simulation ALOS DEM, freely available DEM of highest resolution, was used for the hydrological modelling. The hypsometric distribution of the basin area is presented in Fig. 2.
Figure 2: Hypsometric division of the study area Most of the basin area, about 35 %, lies within the elevation band of 1000 to 2000 m asl, zone-III. Less than 1 % of the basin area lies above 3000 m asl.
2.3 Hydro-Meteorological Data Daily observed hydro-meteorological data were collected from the Department of Hydrology and Metrology (DHM) of the Government of Nepal. The daily precipitation data from 16 rain gauge stations and temperature (minimum and maximum) data from five climatic stations were used as model input for the period from 1990 to 2009. Location of hydro-meteorological stations in the study area is shown in Fig. 1. Daily discharge data for three hydrologic stations: Nayagaon (upstream), Bagasoti (mid-stream) and Jalkundi (downstream) were only available for the period from 1993 to 2009. The drainage area at these stations is 1960 km², 3841 km², and 5143 km², respectively. Jalkundi hydrologic station is located about 73 km downstream from Bagasoti, whereas Nayagaon hydrologic station is located about 56 km upstream from Bagasoti.
1 https://portal.opentopography.org/datasets 2 https://search.earthdata.nasa.gov/search 3 https://www.eorc.jaxa.jp/ALOS/en/aw3d30/ 4 https://search.asf.alaska.edu/ 5 https://bhuvan-app3.nrsc.gov.in/data/download/
110 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
2.4 Land Use/Cover Map The land cover map of Nepal for the year 2010 at a spatial resolution of 30 m. prepared by the International Center for Integrated Mountain Development (ICIMOD, 2010), was used to represent the landcover map of the study area. Seven land use classes were identified for the study area. The land use map of WRR basin is presented in Fig. 3.
Figure 3: Land use/Land cover map of West Rapti River basin The land use is dominated by forest which accounts for 62 % of the total basin area. The agricultural land covers 32 % of the basin area. Whereas, human settlement occupies only 0.04 % of the basin area.
2.5 Soil Map The Soil and Terrain Database (SOTER) for Nepal at a spatial resolution of 1:1 million (Dijkshoorn & Hunting, 2009), prepared by the International Soil Reference and Information Centre (ISRC), was used to develop the soil map of the study area. The soil map of the study area is presented in Fig. 4. Eleven soil units have been identified in the study area. Eutric Cambisols and Dystric Regosols are dominant soil units in the upper and lower part of the basin, respectively. Eutric cambisols are characterized by loamy texture with a medium rate of water transmission and Dystric Regosols are characterized by sandy loam texture with a high rate of water transmission.
111 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
Figure 4: Soil map of the study area
2.6 Methodology The topographic head and availability of flow are the primary elements required for the hydropower potential estimation. The hydropower potential is calculated using Eq. 1. GRHP is estimated at 100% efficiency (Arefiev et al., 2015). P = x x x H (1)
Where,휌 헀 푄 P power (kW) the density of water (kg/m³) g acceleration due to gravity (m/s2) 휌 Q river discharge (m³/s) H topographic head (m)
The overall methodology adopted in this study is present in Fig. 5. SWAT hydrological model was used to simulate the streamflow and SWAT-CUP was used for sensitivity analysis, calibration and validation. The topographic elevation of the inlet and outlet points were extracted from the ALOS 12.5 DEM, as a base case. A program was developed used to process the GIS data, assess topographic head, developed FDCs and estimate the hydropower potential of stream reaches at different PoEs. Finally, topographic heads were from five more DEMs GRHPs were computed accordingly. The GRHPs were then compared with the base case to study the effects of DEMs on estimation of the GRHP.
2.6.1. Hydrological modeling using SWAT The SWAT is a conceptual, semi-distributed, physically-based hydrologic model developed by the United States Department of Agriculture (USDA). It was developed to assess the impact of management on water supplies, nonpoint source pollution in watersheds and large river basins (Arnold et al., 1998). It can simulate surface runoff, percolation, sediment transport, groundwater and reservoir storage.
112 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
Figure 5: Methodology for the assessment of GRHP SWAT model first discretizes the watershed into the numbers of sub-basins, connected through a stream network. These sub-basins are further divided into Hydrological Response Units (HRUs). HRUs are the smallest units in SWAT model that are comprised of the distinct combination of land use, topography and soil characteristics in a watershed. Runoffs generated from each HRUs within a sub-basin are aggregated and routed to the main outlet of the watershed. SWAT model comprises of two phases: land phase and routing phase. The quality of flow, sediment, pesticide and nutrient loading to the main channel is controlled through the land phase. The movement of water, sediment and nutrient to the watershed outlet is controlled through the routing phase (Neitsch et al., 2011). The land phase of the hydrological cycle in SWAT is simulated based on a water balance equation that accounts for precipitation, runoff, evapotranspiration, permeation and return flow components.
Where,
SWt final water content (mm)
113 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
initial water content (mm) SW0 t time (days)
Rday daily precipitation (mm)
Qsurf daily surface runoff (mm)
Ea daily evapotranspiration (mm)
Wseep the daily permeation (mm)
Qgw daily return flow (mm) SWAT uses the GIS interface for running the model. In this study, we used ArcSWAT2012 as an interface of ArcGIS v. 10.5 to setup the model for WRR basin. A drainage threshold area (DTA) of 3000 ha was defined to generate a river network. The basin was delineated into 97 sub-basins, presented in Fig. 6. Generally, threshold areas of 1000 ha to 10,000 ha have been adopted to delineate the watershed for hydropower potential assessment in Nepal (Bajracharya, 2015; Kayastha et al., 2018; WECS, 2019). The drainage threshold area has no significant effect on streamflow (Jha et al., 2004, Bhatta et al., 2019). Smaller threshold value can represent more spatial variability of elevation for hydropower potential assessment. However, it would increase the computational time significantly and the simulations would not be feasible with the available computational resources for this study. Sub-basins were further divided into 951 HRUs by defining the threshold area of 10 % for each land use, soil class and slope. The threshold value of 5 % to 10 % is commonly used in HRU definition, to eliminate the smaller HRUs and increase the computational efficiency of the model (Masih et al., 2011; Meng et al., 2010; Srinivasan et al., 2010; Starks & Moriasi, 2009). Sixty-five stream reaches with inlet and outlet points were identified in WRR basin. The flow rates were simulated both at the inlet and outlet ends of each stream reaches. Hydropower potential was then estimated at each of these streams. Daily weather data were fed into the model in the form of precipitation data collected from 15 rain gauge stations and minimum and maximum temperature data from 5 climatic stations, as shown in Fig. 6. Soil Conservation Service (SCS) curve number method - which is the function of soil permeability, land use and antecedent soil water condition - was selected to estimate surface runoff (Neitsch et al., 2011). The Hargreaves method - which requires air temperature only - was used to estimate Potential Evapotranspiration (PET). was used to estimate potential evapotranspiration (PET) which requires air temperature only. A variable storage method was adopted to route flow in the channels.
2.6.2. Model calibration and validation Once the model is built and run, calibration and validation were performed, using an independent observed dataset, to check the reliability of the model output. Model calibration is the process of tweaking the parameters within the realistic range for maximizing the objective function by minimizing the variation between simulated and observed data. Validation is the final step, in which the model is re-run using the same parameters and their ranges used in calibration but with different observed data set. In this study, calibration and validation were performed, following the protocol suggested by Abbaspour et al., (2015; 2017). The first step in the model calibration and validation is determining the most sensitive parameters for the watershed. Sensitivity analysis, calibration and validation were performed in SWAT Calibration and Uncertainty Program (SWAT-CUP) using Sequential Uncertainty Fitting (SUFI2) algorithm. SWAT- CUP offers two types of sensitive analysis: local and global. In this study, a global sensitivity analysis was performed to rank the sensitive parameters for the watershed.
114 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
Figure 6: West Rapti basin delineated in SWAT model along with river network and hydro-meteorological stations The model was calibrated and validated at three hydrological stations: Nayagaon, Bagasoti and Jalkundi. The observed discharge data at these stations were split into three distinct periods: warmup, calibration and validation periods, i.e., from 1990-1992, 1993-2002, and 2003-2009, respectively. Twenty-one parameters were selected for sensitivity analysis based on the review of the previous studies conducted in the mountainous region of Nepal (Devkota & Gyawali, 2015; Dhami et al., 2018; Mishra et al., 2018; Pokhrel, 2018; S. Shrestha et al., 2018; Talchabhadel et al., 2021). After sensitivity analysis, the model was calibrated and validated using the most sensitive parameters. The performance of the model was evaluated using the statistical indicators: Nash-Sutcliffe Efficiency (NSE), Percentage Bias (PBIAS) and Coefficient of Determination (R2). Relative Nash-Sutcliff Efficiency (NSE) was also checked to evaluate an over or under prediction of low flows. Details about these indicators can be found in Gupta et al. (1999), Krause et al. (2005) and Moriasi et al. (2007).
2.6.3 Flow duration curve (FDC) A FDC is a graphical representation of the recorded historical variation of stream flows at the monitoring station, such that the percentage of a time-specific flow equaled or exceeded over the historical period (Vogel & Fennessey, 1994). The FDC has been widely used in many hydrological studies related to hydropower engineering, flood control, irrigation, water-quality management (Vogel & Fennessey, 1995); design of RoR power plant (Liucci et al., 2014); hydropower generation, river and reservoir sedimentation, water allocation (Castellarin et al., 2004). The shape of the FDC exhibits hydro-meteorological characteristics of the watershed. It can be developed
115 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al. for daily, weekly, monthly and seasonal at the hydrological station. The lower part of FDCs are highly sensitive to the specific period of record, therefore a median annual FDC for a hypothetical year was preferred (Vogel & Fennessey, 1994). The median annual FDC is derived as a median value of streamflow across the n years of record for each probability of exceedance and is less sensitive to the hydrological extremes. In this study, FDCs were developed at upstream and downstream ends of each stream reach, to determine flows at different PoEs.
2.6.4 Assessment of topographic head The WRR basin was first discretized into a network of sub-basins, river streams and outlet points (shown in Fig. 6.). Each sub-basin is drained by a river stream to an outlet. Both stream and outlet point were represented by unique IDs with the linkage to the sub-basin. A river reach was defined as segment of a river staring from the upstream confluence and ending at the downstream confluence. The upstream confluence point was assumed as headworks and the downstream confluence was assumed as the powerhouse of a theoretical hydropower project (Hall et al., 2004). To estimate the topographic head along the river reach, DEM was overlayed with the river network and sub-basin outlets that were generated during the watershed delineation process. Then raster values were extracted at each sub-basin. Topographic heads for each stream reach within each sub-basin outlets were calculated as a difference in raster values between upstream and downstream sub-basin outlets (Hall et al., 2004). Assessment of the head was started at the main outlet of the basin and progressed towards the upstream till the final outlet. Topographic heads were separately computed for DEMs: SRTM 90 m, SRTM 30 m, ASTER GDEM version-3 30 m, ALOS 12.5 m, AW3D 30 m and Cartosat-1 version-3 30 m.
2.6.5 Estimation of Gross Run-of-River Hydropower Potential (GRHP) The GRHP was calculated for each stream reaches within each sub-basin having an inlet and outlet using the topographic heads and flow rates - simulated by the SWAT model. The total GRHP of the basin was calculated by summing the GRHP of all the stream reaches. The hydropower potential of a stream reach is the sum of potential due to discharge entering the stream and half of the potential due to the discharge entering the stream from the local catchment (Arefiev et al., 2015; Hall et al., 2004).
Where,
G hydropower potential of a stream (kW) K the constant term (equals to 1 for GRHP) flow rate entering the stream reach (m3/s) Qi 3 Qo flow rate leaving the stream reach (m /s) H Elevation difference between inlet and outlet (m)
The gross hydropower potential of WRR basin was calculated as:
Where,
GRHP total gross hydropower potential of the basin (kW)
116 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
K the constant term (equals to 1 for GRHP) gross hydropower potential of ith stream reach (kW) Gi n number of reaches
3. Result and Discussion 3.1 Performance Evaluation of SWAT Model Among the 21 selected parameters, 12 were identified as the most sensitive parameters (p-value <0.05), for the study area, presented in Table 1. ALPHA_BNK, CH_K2, and LAT_TTIME were identified as the most sensitive parameters for WRR basin. The base flow alpha factor for bank storage (ALPHA_BNK) was the most sensitive parameter and reveals that bank storage was a dominant process in the study area. Bank storage contributes flow to the adjacent unsaturated zone during high flow, thereby reducing the peak discharge and maintain the baseflow by releasing storage as the floodplain gradually lowered (Whiting & Pomeranets, 1997). Bank storage supplies the flow to the main channel or reaches within a sub-basin. The effective hydraulic conductivity in the main channel alluvium (CH_K2) affects the channel transmission losses. In SWAT, transmission losses are categorized into bank storage and deep aquifer storage. Bank storage contributes to stream reach as a return flow whereas deep storage contributes to streamflow outside the watershed and is considered as lost from the system (Arnold et al., 1993). Lateral flow travel time (LAT_ TTIME) is the function of hill slope length and saturated hydraulic conductivity. Lag in the release of lateral flow from the soil profile results in a smooth streamflow hydrograph. Table 1: Twelve most sensitive parameters and their calibrated range
Parameter Lower Upper Fitted Rank Description p-value Name bound bound Value Baseflow alpha factor for bank storage 1 ALPHA_BNK 0.00 0 0.7 0.44 (days) Effective hydraulic conductivity in main 2 CH_K2 0.00 0 240 10.20 channel (mm/hr) 3 LAT_TTIME Lateral flow travel time (days) 0.00 0 85 20.61 SCS runoff curve number for moisture 4 CN2 0.00 35 98 varies condition II (-) 5 CH_N2 Manning's "n" value for main channel (m1/3/s) 0.00 0 0.18 0.15 6 SLSUBBSN Average slope length (m) 0.00 10 80 10.87 Saturated hydraulic conductivity for first 7 SOL_K (1) 0.00 0 2000 varies soil (mm/hr) Moisture bulk density for first soil layer 8 SOL_BD (1) 0.00 0.9 2.5 varies (gm/cm3) 9 GW_DELAY Ground water delay time (days) 0.00 240 500 485.75 Threshold water level in shallow aquifer 10 GWQMN 0.00 0 2999 532.44 for base flow (mm) 11 CANMX Maximum canopy storage (mm) 0.01 0.6 66 42.29 12 ALPHA_BF Base flow alpha factor (days) 0.04 0 0.7 0.58
117 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
Table 2: Performance rating of SWAT model Calibration Validation Performance Indicators Nayagaon Bagasoti Jalkundi Nayagaon Bagasoti Jalkundi Nash-Sutcliffe Efficiency (NSE) 0.70 0.61 0.71 0.67 0.64 0.63 Relative NSE rel 0.68 0.75 0.74 0.69 0. 78 0.52 Percentage Bias (PBIAS) 12.5 6.9 -8.1 -1.7 11.7 12.1 Coefficient of Determination (R²) 0.71 0.63 0.71 0.68 0.69 0.70
Figure 7: Observed and simulated daily streamflow hydrograph at three discharge monitoring stations, Nayagaon (upstream), Bagasoti (midstream) and Jalkundi (downstream) for calibration and validation.
118 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
The model’s performance during calibration and validation period at three hydrologic stations is graphically presented in Fig. 7. The simulated daily streamflow at three hydrologic stations was consistent with observed stream flow and daily rainfall. The performance rating of SWAT model is shown in Table 2. The statistical indicators show that the model’s performance is within the acceptable range as suggested by (Moriasi et al., 2007). Nash-Sutcliffe Efficiency (NSE) and Coefficient of Determination (R²) greater than 0.5, Percentage Bias (PBIAS) within ±15 % were observed during calibration and validation period at three hydrological stations indicating good prediction capability of the model. Relative Nash-Sutcliffe Efficiency (NSErel) >0.5 during both calibration and validation period indicates well simulation of low flows at all the hydrologic stations. PBIAS around 12% at upstream, 6% at midstream in calibration and 11% at midstream, 12% at downstream in validation show the model underestimated the daily streamflow at respective hydrological stations, for the both periods. Similarly, negative PBIAS at downstream during calibration and upstream during validation indicates that on average 8% and 2% of the simulated daily discharge data are overestimated. Calibration result is better than validation in terms of NSE. This is because the calibration period includes most of the wet years and calibrated parameters are conditioned on the high flows. Overall, the model exhibits satisfactory performance throughout the simulation period from 1993 to 2009. It is observed that the high flows for the year 2000 at Bagasoti station do not correspond to precipitation pattern observed in the basin and the model heavily underestimated the flow. The authors deem that there is an error in discharge data observed at Bagasoti station during this period.
3.2 Flow Duration Curve The flow duration curve at three hydrologic stations is presented in Fig. 8. The graphical representation of FDC shows that the model underestimated the high flow regimes (PoE <10%) and low flow regimes (PoE >60%) at all hydrological stations. However, the model gradually overestimated the flow at midstream and downstream for PoE 10-30 %. The model slightly overestimated the flow for PoE 30-40 % at midstream and downstream. However, the model well reproduced the flow for PoE 40-60 % at all hydrological stations. This range has been considered for the assessment of hydropower potential of the basin. Run-of-river hydropower project is expected to operate efficiently right across these flow rates for reliable energy generation. In Nepal, RoR hydropower projects are designed at 40% PoE (DoED, 2018).
3.3 Gross Run-off-River Hydropower Potential (GRHP) estimation Ninety-seven stream reaches with a minimum drainage area of 9 km2 were identified within WRR basin, out of which 65 stream reaches with inlet and outlet were analyzed for the assessment of GRHP. The assessment was first carried out for ALOS 12.5 m DEM, which is referred as the base case. GRHP was estimated at the identified locations based on topographic head and FDC at different levels of PoE: Q40, Q45, Q50, Q55 and Q60 resulting in a total power potential of WRR basin of about 512, 363, 291, 247 and 216 MW, respectively, for the base case. The power potentials estimated at different PoE are shown in Table 3. The estimated GRHP at 40% PoE were classified into three categories based on the Department of Electricity Development (DoED, 2018) guidelines, shown in Table 4 and their spatial locations are presented in Fig. 9. Twenty different HPP sites, with a capacity in the range from 10 MW to 50 MW, contribute to about 61 % of total basin potential. The stream reaches were further classified into first order, second order, third order and fourth-order streams (Strahler, 1952). The spatial location of identified hydropower potential sites along with stream reaches is shown in Fig. 9. The attributes of identified streams with estimated potentials are shown in Table 3. The shorter streams of second order with an average bed slope of 2.11 % were characterized as the steeper
119 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al. streams. They consist of 19 hydropower potential sites that contribute to about 13 % of total basin potential for flows at 40 % - 60 % PoE. The steeper stream provides more available heads and results in higher power potential for a given discharge (Kusre et al., 2010). However, the longest stream of fourth-order with a relatively low bed slope of 0.27 % contributes more than 50 % of the total basin potential.
Figure 8: Flow duration curve (daily) at three hydrological stations: Nayagaon (top), Bagasoti (middle) and Jalkundi (bottom). Insert at top right in each plot shows the stretched FDC for 10-60% PoE.
120 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
Table 3: Stream features and estimated GRHP at different level of PoE
Average Average Stream Stream No. of spacing GRHP at different PoE (MW) bed Order length potential between slope (Strahler) (Km) sites two HPP (%) Q40 Q45 Q50 Q55 Q60 sites (km) 1 106.44 1.90 11 9.68 28.46 21.05 16.74 14.13 12.24 2 153.86 2.11 19 8.10 68.26 50.72 40.23 33.07 28.86 3 140.98 1.36 14 10.07 143.83 104.53 85.38 74.08 64.10 4 243.11 0.27 21 11.58 271.3 186.88 148.74 125.56 110.94 Total GRHP 511.85 363.18 291.08 246.83 216.15
Table 4: Classification of estimated hydropower projects at Q40 in West Rapti Basin
Hydropower Projects based on production No. of sites Total Power (MW) % Power capacity (MW) ≤1 4 1.01 0.20 >1 and ≤10 41 196.31 38.35 >10 and ≤50 20 314.52 61.45 Total 65 511.84
Figure 9: Spatial location of hydropower potential sites at Q40 in study area
121 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
This is because a higher-order stream is associated with greater discharge (Costa, 1987), thus contributing to higher potential. The overall spacing between the identified HPP sites ranges from 8.10 km to 11.58 km along the river stream. The HPP sites are closely spaced along the steeper stream because of denser stream network. Only WECS (2019), among the hydropower potential studies carried out in Nepal, reports the potential of WRR basin separately. The WECS study estimated the GRHP of the basin as 595 MW and 745 MW for flow at 40 % PoE, using the flow estimated by empirical method and HEC-HMS hydrological modelling, respectively. These figures are 12 % and 45.5%, respectively, higher than the potential estimated in this study for the same PoE. The WECS study used a DTA of 1000 ha for sub basin discretization. The smaller DTA results in more sub-basins. So, more hydropower potential sites are mainly added in the first order streams which ultimately resulted in increased power potential. The higher difference between the GRHPs estimated for WRR basin, reported in the WECS (2019) study indicates that the GRHP estimation is very sensitive to the method of computing discharges to the river reaches within the basin.
3.4 Effects of Source DEM on Gross ROR Hydropower Potential Five publicly availably DEMs from different sources: ASTER-GDEM version - 3 30 m, AW3D 30 m, Cartosat-I version - 3 30 m, SRTM 30 m and SRTM 90 m were used to study their effects on GRHP assessment. The above nomenclature refers to the name of the DEM followed by its resolution. The basins were delineated with the same DTA of 3000 ha and topographic head were computed for river reaches derived from the DEMs. Stream flow simulated by SWAT model, in the base case, was then used for the estimation of the GRHP. The GRHP was estimated for each DEMs for flows at 40%, 45%, 50%, 55% and 60% PoEs (shown in Table 5). It is observed that SRTM 90 m and the AW3D 30 m DEMs resulted in the highest and the lowest hydropower potential, respectively. The results from SRTM 90 m and AW3D 30 m are about 5 % higher and 3% lower than the base case, respectively. The potential estimated by ASTER G-DEM version - 3 and Cartosat-1 version - 3 30 m are very similar to the base case and vary by less than 1%. Whereas, the potential estimated by SRTM 30 m is about 2% lower than the base case. Table 5: Estimated hydropower potentials using different digital elevation models
Digital Elevation Horizontal Hydropower Potential (MW) Model Resolution (m) Q40 Q45 Q50 Q55 Q60 ALOS 12.5 511.84 363.18 291.08 246.83 216.15 ASTER GDEM 30 514.61 364.87 292.96 248.37 217.63 version - 3 AW3D 30 497.90 353.36 283.49 240.58 210.61
Cartosat-1 version - 3 30 513.08 364.40 292.16 247.86 217.02
SRTM 30 502.53 356.75 286.02 242.63 212.40
SRTM 90 539.28 383.11 307.22 260.75 228.47
122 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
The comparison between the topographic heads of stream reaches computed using different DEMs and the base case is shown in Fig. 10. The SRTM 90 m is most scattered whereas SRTM 30 m is least scattered from the equal value line. The maximum scatter for the 30 m resolution DEMs was observed for river reaches of stream order 4 followed by river reaches of stream order 2 (shown in Fig 11). River reaches of stream order 1 were least scattered. However, for SRTM 90 m DEM, scatteredness of the river reaches increased consistently with the increase in stream order (shown in Fig. 12). On an average the bed slope of river reaches of stream order 2 are the steepest and that of stream order 4 are least steep (Table 3). Topographic heads estimated using low-resolution DEM (SRTM 90 m) are more scattered from the equal value line, thus indicating the sensitivity of topographic head with respect to the DEM resolution.
Figure 10: Comparison of topographic heads assessed using ALOS 12.5 m and other DEMs.
Figure 11: Comparison of difference in topographic head between ASTER G-DEM 30 m and ALOS 12.5 m for different stream orders. Other 30 m resolution DEMs also had similar pattern. The power potential of all stream reaches estimated from different DEMs compared with base case, for flows at 40 % PoE, are shown in Fig. 13. Most of the river reaches have power potential less than 20 MW. Negative heads were obtained at some river reaches due to errors present in DEMs, resulting in negative power potential. Such river reaches were not included while computing the total potential of the basin.
123 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
Figure 12: Comparison of difference in topographic head between SRTM 90 m and ALOS 12.5 m for different stream orders. The trend of power potential is, however, different compared to the trend of the topographic head. This is because of the influence of the discharge also in the power potential. SRTM 30 m is very much close to the equal value line. Whereas, all of the other DEMs are more scattered. However, the total power potential estimated from ASTER G-DEM version - 3 30 m and Cartosat-1 version - 3 30 m are closer to the base case, compared to SRTM 30 m.
Figure 13: Comparison of hydropower potential estimated using ALOS and other digital elevation models.
4. Conclusions An approach based on GIS and hydrological modelling was used to estimate the GRHP of the West Rapti River basin. ALOS DEM, the highest resolution (12.5 m) DEM, available for the study area in the public domain was referred as the base case. The river basin was discretized into 97 sub-basins/ river reaches and the topographic head at the river reaches was assigned using GIS based tools. SWAT, a semi distributed
124 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] hydrological model, was setup to simulate the flow at the inlet and outlets of each river reach. The simulation results showed that the model underestimated the low flows and high flows. However, the flows at 40-60% PoE were well reproduced by the model. Thus, the hydropower potential of the basin was estimated across these flows. The power potentials were estimated at 65 river reaches summing up to a GRHP of the West Rapti River basin as 512 MW for flows at 40% PoE, for the base case. Most of hydropower potential were located in the main channel (i.e., stream order 3 and 4). River reaches with stream order 4 were mild but higher flow contributed in the higher potential. Five DEMs from different sources were used to analyze their effect on the GRHP. The variation of GRHP was estimated less than 6% compared to the base case (i.e., ALOS 12.5 m). SRTM 90 m showed most deviation in the topographic head compared to other DEMs of 30 m resolution. Thus, topographic head was found sensitive with respect to the DEM resolution. The highest deviations are expected in the river reaches located in the lower parts of the basin. Comparison of the results with other- studies, however, showed that the method of computing discharge at the river reaches within the basin will likely have more effect on the GRHP assessment compared to different sources of DEMs (with horizontal resolution ≤90 m). The findings of this study will be helpful to the hydropower developers, water resource planners and policymakers for optimal utilization of water resources in West Rapti River basin. Moreover, the findings, in a general aspect, will also be helpful for the researchers in focusing on the key controlling factors while conducting the Gross RoR Hydropower Potential studies.
Acknowledgments We are thankful to the Department of Hydrology and Meteorology (DHM), Government of Nepal, for providing the hydro-meteorological data. We also would like to thank mechanical department of IOE, Pulchowk Campus for giving access to the computer for model calibration and validation. We would also like to thank the reviewers for their review of the manuscript which helped to improve the paper. Singh's input was partially funded by Energize Nepal Program, Grant Number NPL-12/0032, ENERGIZE NEPAL.
Conflict of Interests Authors declare no conflict of interests.
References Abbaspour, K. C., Rouholahnejad, E., Vaghefi, S., Srinivasan, R., Yang, H., & Kløve, B. (2015). A continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model. Journal of Hydrology, 524, 733–752. https://doi.org/10.1016/j.jhydrol.2015.03.027 Abbaspour, K. C., Vaghefi, S. A., & Srinivasan, R. (2017). A guideline for successful calibration and uncertainty analysis for soil and water assessment: A review of papers from the 2016 international SWAT conference. Water (Switzerland), 10(1). https://doi.org/10.3390/w10010006 Ali, F., Srisuwan, C., Techato, K., Bennui, A., & Suepa, T. (2020). Theoretical Hydrokinetic Power Potential Assessment. Energies, 1–13. Arefiev, N., Badenko, N., Ivanov, T., Kotlyar, S., Nikonova, O., & Oleshko, V. (2015). Hydropower Potential Estimations and Small Hydropower Plants Siting: Analysis of World Experience. Applied Mechanics and Materials, 725–726, 285–292. https://doi.org/10.4028/www.scientific.net/amm.725-726.285 Arefiev, N., Nikonova, O., Badenko, N., Ivanov, T., & Oleshko, V. (2015). Development of automated approaches for hydropowerpotential estimations and prospective hydropower plants siting. Vide. Tehnologija. Resursi - Environment, Technology, Resources, 2, 41–50. https://doi.org/10.17770/etr2015vol2.260
125 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
Arnold, J. G., Srinivasan, R., Muttiah, R. S., & Williams, J. R. (1998). Large area hydrologic modeling and assessment Part i: Model development. Journal of the American Water Resources Association, 34(1), 73–89. https://doi. org/10.1111/j.1752-1688.1998.tb05961.x Arnold, Jeffrey G., Allen, P. M., & Bernhardt, G. (1993). A comprehensive surface-groundwater flow model. Journal of Hydrology, 142(1–4), 47–69. https://doi.org/10.1016/0022-1694(93)90004-S Aryal, A., Magome, J., Pudashine, J. R., & Ishidaira, H. (2018). Identifying the Potential Location of Hydropower Sites and Estimating the Total Energy in Bagmati River Basin. Journal of Japan Society of Civil Engineers, Ser. G (Environmental Research), 74(5), I_315-I_321. https://doi.org/10.2208/jscejer.74.i_315 Bajracharya, I. (2015). Assessment of run-of-river hydropower potential and power supply planning in Nepal using hydro resources. Institut Für Energietechnik Und Thermodynamik E, April, 7–12. http://publik.tuwien.ac.at/files/ PubDat_238319.pdf Bhatt, R. P. (2017). Hydropower Development in Nepal - Climate Change, Impacts and Implications. Renewable Hydropower Technologies. https://doi.org/10.5772/66253 Bocchiola, D., Manara, M., & Mereu, R. (2020). Hydropower potential of run of river schemes in the himalayas under climate change: A case study in the Dudh Koshi basin of Nepal. Water (Switzerland), 12(9). https://doi.org/10.3390/ W12092625 Castellarin, A., Galeati, G., Brandimarte, L., Montanari, A., & Brath, A. (2004). Regional flow-duration curves: Reliability for ungauged basins. Advances in Water Resources, 27(10), 953–965. https://doi.org/10.1016/j.advwatres.2004.08.005 Chaubey, I., Cotter, A. S., Costello, T. A., & Soerens, T. S. (2005). Effect of DEM data resolution on SWAT output uncertainty. Hydrological Processes, 19(3), 621–628. https://doi.org/10.1002/hyp.5607 Chaudhary, R. P. (2000). Forest conservation and environmental management in Nepal: a review. Biodiversity and Conservation, 9(1235–1260). https://doi.org/10.1023/A:1008900216876 Costa, J. E. (1987). Hydraulics and basin morphometry of the largest flash floods in the conterminous under united states. Journal of Hydrology, 93, 313–338. Devkota, L. P., & Gyawali, D. R. (2015). Impacts of climate change on hydrological regime and water resources management of the Koshi River Basin, Nepal. Journal of Hydrology: Regional Studies, 4, 502–515. https://doi.org/10.1016/j. ejrh.2015.06.023 Dhami, B., Himanshu, S. K., Pandey, A., & Gautam, A. K. (2018). Evaluation of the SWAT model for water balance study of a mountainous snowfed river basin of Nepal. Environmental Earth Sciences, 77(1). https://doi.org/10.1007/s12665- 017-7210-8 Dhungel, K. R. (2016). A causal relationship between energy consumption, energy prices and economic growth in Africa. In International Journal of Energy Economics and Policy (Vol. 6, Issue 3, pp. 477–494). Dijkshoorn, K., & Hunting, J. (2009). Soil and Terrian database for Nepal. Report 2009/01. http://www.isric.org/isric/ webdocs/docs/ISRIC_Report_2009_01.pdf DoED. (2018). Guidelines for Study of Hydropower Projects 2018. http://doed.gov.np/storage/listies/December2019/ guidelines-for-study-of-hydropower-projects-2018.pdf DoED. (2021). Hydropower license database. Guiamel, I. A., & Lee, H. S. (2020). Potential hydropower estimation for the Mindanao River Basin in the Philippines based on watershed modelling using the soil and water assessment tool. Energy Reports, 6, 1010–1028. https://doi. org/10.1016/j.egyr.2020.04.025 Gunatilake, H., Wijayatunga, P., & Roland-Holst, D. (2020). Hydropower Development and Economic Growth in Nepal (70). Gupta, H. V., Sorooshian, S., & Yapo, P. O. (1999). Status of Automatic Calibration for Hydrologic Models: Comparison with Multilevel Expert Calibration. Journal of Hydrologic Engineering, 4(2), 135–143. https://doi.org/10.1061/ (ASCE)1084-0699(1999)4:2(135) Hall, D. G., Cherry, S. J., Reeves, K. S., Lee, R. D., Carroll, G. R., Sommers, G. L., & Verdin, K. L. (2004). Water Energy Resources of the United States with Emphasis on Low Head/Low Power Resources. April, 71. Hoes, O. A. C., Meijer, L. J. J., Van Der Ent, R. J., & Van De Giesen, N. C. (2017). Systematic high-resolution assessment of global hydropower potential. PLoS ONE, 12(2), 1–10. https://doi.org/10.1371/journal.pone.0171844 Hussain, A., Sarangi, G. K., Pandit, A., Ishaq, S., Mamnun, N., Ahmad, B., & Jamil, M. K. (2019). Hydropower development in the Hindu Kush Himalayan region: Issues, policies and opportunities. Renewable and Sustainable Energy Reviews, 107(February), 446–461. https://doi.org/10.1016/j.rser.2019.03.010
126 Bista et al. Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021]
ICIMOD. (2010). Land Cover of Nepal. 1–2. https://doi.org/https://doi.org/10.26066/rds.9224. Jha, R. (2010). Total Run-of-River type Hydropower Potential of Nepal. Hydro Nepal: Journal of Water, Energy and Environment, 7(7), 8–13. https://doi.org/10.3126/hn.v7i0.4226 Karki, R., Talchabhadel, R., Aalto, J., & Baidya, S. K. (2016). New climatic classification of Nepal. Theoretical and Applied Climatology, 125(3–4), 799–808. https://doi.org/10.1007/s00704-015-1549-0 Kayastha, N., Singh, U., & Dulal, K. P. (2018). A GIS Approach for Rapid Identification of Run-of-River (RoR) Hydropower Potential Site in Watershed: A case study of Bhote Koshi Watershed, Nepal. Hydro Nepal: Journal of Water, Energy and Environment, 23(23), 48–55. https://doi.org/10.3126/hn.v23i0.20825 Krause, P., Boyle, D. P., & Bäse, F. (2005). Comparison of different efficiency criteria for hydrological model assessment. Advances in Geosciences, 5, 89–97. https://doi.org/10.5194/adgeo-5-89-2005 Kumar, P., Kunwar, S., & Gare, V. (2017). Hydropower Site Investigation and Sensitivity Analysis of Assessed Potential Using Geospatial Inputs. 75, 189–198. https://doi.org/10.1007/978-3-319-55125-8 Kusre, B. C., Baruah, D. C., Bordoloi, P. K., & Patra, S. C. (2010). Assessment of hydropower potential using GIS and hydrological modeling technique in Kopili River basin in Assam (India). Applied Energy, 87(1), 298–309. https://doi. org/10.1016/j.apenergy.2009.07.019 Li, J., & Wong, D. W. S. (2010). Effects of DEM sources on hydrologic applications. Computers, Environment and Urban Systems, 34(3), 251–261. https://doi.org/10.1016/j.compenvurbsys.2009.11.002 Liucci, L., Valigi, D., & Casadei, S. (2014). A new application of flow duration curve (FDC) in designing run-of-river power plants. Water Resources Management, 28(3), 881–895. https://doi.org/10.1007/s11269-014-0523-4 Magaju, D., Cattapan, A., & Franca, M. (2020). Identification of run-of-river hydropower investments in data scarce regions using global data. Energy for Sustainable Development, 58, 30–41. https://doi.org/10.1016/j.esd.2020.07.001 Masih, I., Maskey, S., Uhlenbrook, S., & Smakhtin, V. (2011). Assessing the Impact of Areal Precipitation Input on Streamflow Simulations Using the SWAT Model. Journal of the American Water Resources Association, 47(1), 179–195. https:// doi.org/10.1111/j.1752-1688.2010.00502.x Meng, H., Sexton, A. M., Maddox, M. C., Sood, A., Brown, C. W., Ferraro, R. R., & Murtugudde, R. (2010). Modeling Rappahannock River Basin Using SWAT - Pilot for Chesapeake Bay Watershed. Applied Engineering in Agriculture, 26(5), 795–805. https://doi.org/10.13031/2013.34948 Mishra, Y., Nakamura, T., Babel, M. S., Ninsawat, S., & Ochi, S. (2018). Impact of climate change on water resources of the Bheri River Basin, Nepal. Water (Switzerland), 10(2), 1–21. https://doi.org/10.3390/w10020220 Moriasi, D. N., Arnold, J. G., Liew, M. W. Van, Bingner, R. L., Harmel, R. D., & Veith, T. L. (2007). Model Evaluation Guidelines For Systematic Quantification of Accuracy in Watershed Simulation. American Society of Agricultural and Biological Engineers, 50(3), 885–900. Mukherjee, S., Joshi, P. K., Mukherjee, S., Ghosh, A., Garg, R. D., & Mukhopadhyay, A. (2012). Evaluation of vertical accuracy of open source Digital Elevation Model (DEM). International Journal of Applied Earth Observation and Geoinformation, 21(1), 205–217. https://doi.org/10.1016/j.jag.2012.09.004 Nazari-Sharabian, M., Taheriyoun, M., & Karakouzian, M. (2020). Sensitivity analysis of the DEM resolution and effective parameters of runoff yield in the SWAT model: A case study. Journal of Water Supply: Research and Technology - AQUA, 69(1), 39–54. https://doi.org/10.2166/aqua.2019.044 NEA. (2020). Nepal Electricity Authority Fiscal Year 2019/ 2020-A year in Review. Neitsch, S., Arnold, J., Kiniry, J., & Williams, J. (2011). Soil & Water Assessment Tool Theoretical Documentation Version 2009. Texas Water Resources Institute, 1–647. https://doi.org/10.1016/j.scitotenv.2015.11.063 Pandey, A., Lalrempuia, D., & Jain, S. K. (2015). Assessment of hydropower potentail using spatial technology and SWAT modeling in the Mat River, southern Mizoram, India. Hydrological Sciences Journal, 60(10), 1651–1665. https://doi. org/10.1080/02626667.2014.943669 Pandey, V. P., Dhaubanjar, S., Bharati, L., & Thapa, B. R. (2019). Hydrological response of Chamelia watershed in Mahakali Basin to climate change. Science of the Total Environment, 650(365–383), 1495–1504. https://doi.org/https://doi. org/10.1016/j.scitotenv.2018.09.053 Pandey, V. P., Dhaubanjar, S., & Thapa, B. R. (2020). Spatio-temporal distribution of water availability in Karnali-Mohana Basin, Western Nepal: Hydrological model development using multi-site calibration approach (Part-A). Journal of Hydrology: Regional Studies, 29(100690). https://doi.org/ps://doi.org/10.1016/j.ejrh.2020.100690 Pokhrel, B. K. (2018). Impact of land use change on flow and sediment yields in the Khokana outlet of the Bagmati River,
127 Journal of Engineering Issues and Solutions 1 (1): 106-128 [2021] Bista et al.
Kathmandu, Nepal. Hydrology, 5(2). https://doi.org/10.3390/hydrology5020022 Prajapati, R. N. (2015). Delineation of Run of River Hydropower Potential of Karnali Basin- Nepal Using GIS and HEC- HMS. European Journal of Advances in Engineering and Technology, 2(1)(February 2015), 50–54. Sammartano, V., Liuzzo, L., & Freni, G. (2019). Identification of potential locations for run-of-river hydropower plants using a GIS-based procedure. Energies, 12(18), 1–20. https://doi.org/10.3390/en12183446 Shrestha, Hari M. (2016). Exploitable Potential, Theoretical Potential, Technical Potential, Storage Potential and Impediments to Development of the Potential: The Nepalese Perspective. Hydro Nepal: Journal of Water, Energy and Environment, 19, 1–5. https://doi.org/10.3126/hn.v19i0.15340 Shrestha, Hari Man. (1996). Cadastre of Potential Hydropower Resources in Nepal. In Moscow Power Institute. Moscow Power Institute. Shrestha, S., Shrestha, M., & Shrestha, P. K. (2018). Evaluation of the SWAT model performance for simulating river discharge in the himalayan and tropical basins of Asia. Hydrology Research, 49(3), 846–860. https://doi.org/10.2166/ nh.2017.189 Singh, V. K., & Singal, S. K. (2017). Operation of hydro power plants-a review. Renewable and Sustainable Energy Reviews, 69(August 2016), 610–619. https://doi.org/10.1016/j.rser.2016.11.169 Srinivasan, R., Zhang, X., & Arnold, J. (2010). SWAT ungauged: Hydrological budget and crop yield predictions in the upper Mississippi River basin. Transactions of the ASABE, 53(5), 1533–1546. https://doi.org/10.13031/2013.34903 Starks, P. J., & Moriasi, D. N. (2009). Spatial Resolution Effect of Precipitation Data on SWAT Calibration and Performance: Implications for CEAP. Transactions of the ASABE, 52(4), 1171–1180. https://doi.org/10.13031/2013.27792 Strahler, A. N. (1952). Hypsometric (Area-Altitutde) Analysis of Erosional Topography. Bulletin of the Geological Society of America, 63(1117–1142), 23. https://doi.org/10.1130/0016-7606(1952)63[1117:HAAOET]2.0.CO;2 Talchabhadel, R., Aryal, A., Kawaike, K., Yamanoi, K., Nakagawa, H., Bhatta, B., Karki, S., & Thapa, B. R. (2021). Evaluation of precipitation elasticity using precipitation data from ground and satellite-based estimates and watershed modeling in Western Nepal. Journal of Hydrology: Regional Studies, 33(December). https://doi.org/10.1016/j.ejrh.2020.100768 Talchabhadel, R., Karki, R., Thapa, B. R., Maharjan, M., & Parajuli, B. (2018). Spatio-temporal variability of extreme precipitation in Nepal. International Journal of Climatology, 38(11), 4296–4313. https://doi.org/10.1002/joc.5669 Talchabhadel, R., & Sharma, R. (2014). Real Time Data Analysis of West Rapti River Basin of Nepal. Journal of Geoscience and Environment Protection, 02(05), 1–7. https://doi.org/10.4236/gep.2014.25001 Vogel, R., & Fennessey, N. (1994). Flow-Duration Curves I: New Interpretation and Confidence Intervals. Journal of Water Resoursed Planning and Management, 120(Capítulo 1), 1–9. Vogel, R. M., & Fennessey, N. M. (1995). Flow Duration Curves II: A Review Of Applications In Water Resources Planning. Journal of the American Water Resources Association, 31(6), 1029–1039. https://doi.org/10.1111/j.1752-1688.1995. tb03419.x WECS. (2005). National Water Plan - Nepal. In Water and Energy Commission Secretariat (WECS), Kathmandu. http:// www.moen.gov.np/pdf_files/national_water_plan.pdf WECS. (2019). Assessment of Hydropower Potential of Nepal. http://www.wecs.gov.np/storage/listies/February2021/ final-report-july-2019-on-hydropower-potential.pdf Whiting, P. J., & Pomeranets, M. (1997). A numerical study of bank storage and its contribution to streamflow. Journal of Hydrology, 202(1–4), 121–136. https://doi.org/10.1016/S0022-1694(97)00064-4
128 Journal of Engineering Issues and Solutions
Window to wall ratio and orientation effects on thermal performance of residential building: A case of Butwal Sub-Metropolis
Sanjaya Uprety1, *, Shiva Kafley 1, Barsha Shrestha 1 1 Department of Architecture, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Nepal *Corresponding email: [email protected] Received: January 10, 2021; Revised: March 06, 2021; Accepted: March 28, 2021
Abstract The orientation and glazed surface area used for windows in a building have significant effects on its indoor thermal comfort and overall energy consumption. The increasing use of glazed windows and lack of consideration of orientation in building design have become a major problem in warm and humid regions as windows cover sensitive skin areas for the exchange of energy leading to increased solar gain inside the building. This paper describes the effect of the varied ‘area ratio of glazed window to the wall for different building orientations’ on the thermal performance of the residential building in a warm humid climatic region of Nepal. A typical residential building located in Kalikanagr of Butwal, the fast-urbanizing sub-metropolis of Western Nepal, was selected for the study from 18 houses surveyed using the purposive sampling method. Nine varying values of Window to Wall Ratio (WWR) of glazed façade ranging from 0.1 to 0.9 with a constant increment of 0.1 in north and south façades, and the change in the building orientations were considered for the detailed study. Altogether eighty different test scenarios including base case scenarios were created and annual thermal energy consumption was computed for each test scenario using the Autodesk Ecotect Analysis, 2011. Findings from the study showed that the south orientation is the most appropriate compared to the north-east for all WWR to reduce the building energy consumption and an increase in WWR also results in increased energy consumption. The study concludes the careful considerations of WWR and the south orientation during the designing of building will contribute to efficient energy consumption in residential buildings.
Keywords: Building orientation; energy consumption; residential building; thermal comfort; window to wall ratio.
1. Introduction Buildings and operations, the largest energy-consuming sector in the world, account for the highest share of both global final energy consumption (36%) and energy-related CO2 emission (39%) (IEA, 2019). In Nepal too, total energy consumption in the year 2008/09 was about 9.3 million tonnes of oil equivalent (401 million GJ). The share of residential energy accounts for nearly 87% of the total final energy consumption in Nepal
129 Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021] Uprety et al.
(WECS, 2014). There are growing global and local concerns to reduce the energy consumption and resulting carbon footprint and efforts are being made to develop and deploy sustainable construction technologies, systems, and materials in new and existing buildings to meet the sustainable development goals. As a part of sustainable construction technology, Butera (2005) has highlighted the importance of window glazing as an important technological innovation to achieve indoor thermal comfort1 in enclosed spaces since after the Renaissance period. According to him, it provides a noticeable thermal comfort improvement in the indoor space since glass is capable of trapping solar radiation in the room, and in the winter sunny days, indoor thermal comfort is improved even without a source of heat. This, however, requires a good balance while combining the opaque and transparent elements of the building envelope to achieve the desired daylight, thermal performance, and the building's energy efficiency. Chi et al. (2020) call it a technique (or passive strategy) that aims at controlling and regulating the infiltrated solar radiation and ventilation to create a healthy indoor environment. The good balance, in this sense, is dependent on the Window-to-Wall Ratio (WWR), which is defined in terms of window size expressed as glass ratio of a building façade (Hassan, 2016) and the Building Orientation (BO) that affects sun exposure and resulting thermal acquisition, ventilation, and lighting (Elghamry and Azmy 2017). Several researchers have examined the relationship between the various dimensions of WWR including the use of glazing, shading, and orientation and their effects on the thermal performance of the building. Ghosh and Neogi (2018) in a simulated result on the effect of fenestration’s geometric factors on building energy consumption in a warm and humid climate showed that the increase in WWR in south-facing air-conditioned building cell2; heating, and lighting energy consumption decreased whereas the cooling energy consumption increased. Similarly, eighteen building orientation intervals and eight WWR interval combinations were analyzed using Ecotect 2016 and PHOENICS 2012 simulation software by Chi et al. (2020). The analysis revealed that the best WWRs for N-W 80 and S-E 80 are 0.39 and 0.4 respectively. According to the research performed by Nair et al. (2014) using Design Builder and Energy Plus simulation software, the South orientation and 15% WWR showed the best effect in the case of the composite climate of Rajasthan. Likewise, comparative experimental and simulation studies for thermal performance analysis in residential buildings in Hot-Humid climate conducted by Al-Tamimi and Syed Fadzilb (2010) showed that rooms with high WWR are relatively cool during night time only whereas low WWR performs well results during daytime and night-time. Nearly Zero Energy Building’s most energy-saving WWR design scheme in severe cold areas is that the east WWR ranges from 10% to 15%, the south WWR from 10% to 22.5%, and decrease the north WWR decreases appropriately if the conditions of light and ventilation allow it (Feng et al. 2017). Although energy consumption has been strongly influenced by climatic conditions, structural insulation characteristics, façade configurations, presence of shading devices, the optimal WWR does not seem to vary significantly if the effect of each factor is evaluated individually (Marino et al. 2017). In Nepal, which has a very diverse topography with varying climatic zones, the building design process requires a good understanding of the effects of WWR and BO on the thermal performance of the buildings as the use of glazed windows and their size are sensitive to solar gain and indoor heating. Although several studies were carried out to examine the effect of WWR and BO on the thermal performance of public buildings, not much information is available on such effects in case of residential building types. In this context, this paper aims to analyze the climatic data, find out the best orientation and examine the effect of the WWR and BO on the thermal energy consumption of the residential building in the warm and humid climatic zone mostly prevailing in eastern parts of the Terai (plains) of Nepal. The understanding of the effect of WWR and BO in the thermal energy consumption is expected to have significant practical
1 ASHRE (2010) defines thermal comfort as the “mental state which expresses satisfaction with the thermal environment” 2 The air-conditioned building cell of dimension 5m x 5m x 3m was simulated in Energy Plus simulation software by Ghosh and Neogi (2018)
130 Uprety et al. Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021] implications in achieving better thermal comfort and reduce the running cost of the active system in new construction and retrofitting of the existing buildings. It will help, among others, the design practitioners and policy planners in making design and planning decisions.
2. Study Methods The quantitative methodology was used for the study. A household survey of 18 houses was carried out in Kalianagar of Ward no 10 of Butwal sub-metropolis based on the residential categories, Sub Group A13 (NBC:206, 2015). The houses were selected using the purposive sampling method based on the researcher’s knowledge and experience. A structured questionnaire survey was carried out to get the general idea of the thermal sensation as experienced by the occupants. Out of eighteen surveyed houses, a typical building (latitude 27° 40’ North and longitude 83° 27’ East) representing the general modern residential building typology of Butwal was then selected for the simulation purpose (Fig. 1). “Autodesk Ecotect Analysis, 2011” simulation software4 was used to analyze the selected typical building to evaluate the effect of WRR and BO on energy consumption.
The methodological process further involved the analysis of weather data5 of 10 years (2007-17) obtained from the Department of Hydrology and Meteorology (DHM). The weather file was then edited and created in “WEA” format as input data for Autodesk Ecotect Analysis, 2011. The Ecotect simulation software in which the thermal performance analysis is based on the admittance method as laid out in the CIBSE guide A (CIBSE, 1999), was adopted for its users' friendliness. The different thermal and non-thermal zones were created using field data. The comfort band, calculated by neutrality temperature for the case study area, is then assigned and the selected building was simulated in the mixed-mode in all thermal zones. Input data related to occupant’s activities and building uses were taken from field and the thermal properties of materials from the inbuilt data made available by simulation software itself. The building was modeled and as many as 80 simulated scenarios were developed and interpreted based on the different WWR and orientation to evaluate their effects on thermal performance.
Figure 1: Plan views of the selected typical building.
3 The Sub Group A1in NBC code is defined as general residence which includes any private residential building having sleeping accommodation for up to 40 people and a built up area of not less than 500 square meter 4 Autodesk Ecotect Analysis, 2011” simulation software is widely used to analyse the thermal and energy performance of the building (Al-Tamimi and Syed Fadzilb 2010). 5 Data related to dry bulb temperature, relative humidity, rainfall were collected from Department of Hydrology and Meteorology (DHM). Additional unavailable data like wind speed, wind direction, solar radiations were collected from NASA website. The data from both sources were compared to fill the data gap and finally temperature and relative humidity were used from DHM and solar radiation data from NASA website to construct the building bioclimatic chart
131 Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021] Uprety et al.
3. Results and Discussion The analysis of weather data obtained from DHM revealed that the warmest month is April with a maximum air temperature of 37.03 °C while the coldest month is January with a minimum air temperature of 10.66 °C. Although the relative humidity is high throughout the year; July, August, and September witness higher relative humidity due to the monsoon (Fig. 2). The weather data analysis confirms that the Butwal sub metropolis has a warm and humid climate.
Figure 2: Average monthly air temperature and relative humidity. The climate data including dry bulb temperature, relative humidity, rainfall, solar radiation, wind speed, and directions were edited to create a weather file in “. WEA” for the purpose of weather simulation. The simulated weather data revealed the best and worst orientations. It showed that the best orientation is 187.5° clockwise from the north and the worse orientation is 97.5° clockwise from the north (Fig. 3). The survey result of the case area showed that more than 77% of the total houses surveyed were found to be modern type which is defined as a permanent building having an RCC frame structure. Likewise, the south and north-oriented housing accounted for 28.0 % each. Respondents living in traditional and modern houses surveyed stated that they needed room cooling from April to October whereas the majority of respondents living in modern houses reported that they required room heating in the winter season. A twelve-month electricity unit consumption of the surveyed buildings showed the increment in electricity unit consumption due to the cooling load from the month of March to September with a peak in the month of June. As for the residential indoor thermal satisfaction, the majority of the people living in modern buildings were found to be satisfied despite the inferior thermal comforts in their homes. A typical building (Fig. 4) selected, for the analysis based on the survey result is a three-story, north-oriented with a floor height of 3.35m. The top floor is unoccupied and covered with Corrugated Galvanized Iron (CGI) roofing. The ground floor has timber window frames and timber paneled shutters whereas the first floor has single glazed timber frame windows. Both windows are operable towards the outside in the north and south-
132 Uprety et al. Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021] facing walls. The existing building (base-case) was modeled in ECOTECT simulation software to create sixteen thermal zones (Fig. 5) in such a way that each room on the ground and first floors (Fig. 1) was a zone. The verandas on the first and top floors were considered non-thermal zones as they are not enclosed spaces.
Figure 3: Best and worst orientation in the study area.
The comfort band, calculated by neutrality temperature6 for the case study area, was assigned as 22.5 °C to 27.0 °C for a mixed-mode system7 in all thermal zones. Since none of the residential buildings surveyed used air-conditioning except fans for cooling, the mixed-mode system was chosen as the study intended to compute energy consumption for different case scenarios which otherwise could not be derived by assigning natural ventilation system that gives discomfort hours only. For all zones, the comfort humidity was set as 50% and airspeed as 0.5 m/s for pleasant breeze. Similarly, the air change rate was set at 0.5 ACH for well- sealed condition, and clothing as 0.6 clo. For the base case, the analysis showed that the existing WWR as 0.35 for the ground floor in the north and south-facing wall; 0.27 for north and south-facing first-floor bedrooms; and 0.15 for the south-facing first- floor living room. The data relating to energy consumption patterns, functional activities, and uses were taken from field data (Table 1) and for thermal properties of the materials inbuilt data from the simulation software itself (Table 2).
6 The neutrality temperature was calculated using the formula “Tn equals to 17.6 + 0.31x Tav” where "Tn equals to 17.6 + 0.31x 15.86 = 22.5 °C” for coldest month and “Tn equals to 17.6 + 0.31x 30.41 = 27.0 °C” for warmest months. 7 The mixed-mode system is a combination of air-conditioning and natural ventilation, where the HVAC system shuts down whenever external conditions are within the specified thermostat
133 Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021] Uprety et al.
Figure 4: Two storied building simulated. Figure 5: Thermal zones of simulated building.
Table 1: Field data and zone setting.
Lighting level Zone Zone type Occupants Operation schedule (Lux) Ground bedroom 1 400 Thermal 2 0-24 Ground bedroom 2 400 Thermal 0 18-19 Ground bedroom 3 400 Thermal 2 0-24 Ground bedroom 4 400 Thermal 1 0-24 Ground restroom 200 Thermal 1 7-8 Corridor 200 Thermal 0 - First-floor living kitchen 400 Thermal 3 8-10 First-floor bedroom 1 400 Thermal 0 22-6 First-floor bedroom 2 400 Thermal 2 22-6 First-floor bedroom 3 400 Thermal 1 22-6 First-floor bedroom 4 400 Thermal 3 10-18 First-floor restroom 200 Thermal 1 6-7 First-floor front balcony 200 Non-thermal 0 - First floor back balcony 200 Non-thermal 0 - Staircase 100 Thermal 0 - Top floor 50 Non-thermal 0 -
134 Uprety et al. Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021]
Table 2: Thermal properties of building materials.
Components Materials Thickness (mm) U-Value (W/m2k) Roof CGI 2.6 5.62 130 2.8 Wall Brick Plaster 250 1.9 Door Door core timber 30 3.11 Ceiling RCC with plaster 125 3.25 Floor Stone soling PCC 260 2.78 Timber frame Timber pane 30 3.11 Windows Single glazed 6 5.44
In test scenarios created, the existing WWR and the orientation were changed from the base-case scenario. Considering the different building orientations and combinations of WWR and two evaluation parameters of heating and cooling energy consumption for each case, a total of 80 test scenarios were proposed for investigation. Firstly, a total of eight test case scenarios including base-case were evaluated by rotating building (base-case) in eight different orientations i.e., North (N), North-East (NE), East (E), South-East (SE), South (S), South-West (SW), West (W) and North-West (NW) for existing WWR and resulting heating and cooling energy consumption were obtained. The result showed that annual thermal energy consumption (Fig. 6) decreased by 0.24% for the south orientation and increased by 1.31% for the north-east orientation as compared to the existing north orientation.
Figure 6: Annual energy consumption for existing WWR. Secondly, a constant increment of WWR by 0.1 from 0.1 to 0.9 in eight different building orientations was simulated and 72 test scenarios were compared. For example, WWR was set as 0.1 for all glazed windows on the first floor keeping the timber panel window on the ground floor intact. The building with 0.1 WWR was rotated in a constant increment of 45° clockwise with eight variations in orientation and resulting total heating and cooling energy consumption were obtained from simulation (Fig. 7). Likewise, the effect of WWR and orientation was evaluated by maintaining a constant increment of WWR by 0.1 from 0.2 to 0.9 in eight different orientations. The result showed that the annual thermal load is minimum when WWR is 0.1 and maximum when WWR is 0.9. It also revealed that, compared to base-case annual energy consumption (16.63 MWh), the consumption
135 Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021] Uprety et al. decreased by 1.36% for south orientation and increased by 0.23% for north-east orientation when WWR is 0.1. Likewise, the annual thermal energy consumption increased by 5.24% for the south orientation and 7.16% for the north-east orientation respectively when WWR is 0.9. Overall, the result obtained from 72 test scenarios showed that the annual thermal energy consumption increased linearly with an increase in WWR (Fig. 7).
Figure 7: Annual thermal energy consumption (MWh). The simulated results involving base-case and alternative scenarios revealed that south orientation is best suited and North-East orientation is worst for all WWR in a warm and humid climate, which aligns with the findings from several researchers (Elghamry and Azmy 2017; Nair, Shukla, Shekhar and Jatav 2014). The south as the best orientation for warm and humid climate region is also confirmed by the simulated results of the weather file (figure 3). Likewise, the study also found that increased WWR increases annual thermal load as pointed out in the research by Ghosh and Neogi (2018) and Nair, Shukla, Shekhar, and Jatav (2014). Generally, it is observed that in a warm and humid climate zone, the south orientation is expected to receive more solar radiation increasing the room heating during summer and heat loss during winter requiring a thermal balance to maintain the comfort level throughout the year. However, the result showed the south orientation as best vis-à-vis less significant variations in thermal loads for both summer and winter. The reason for this could be the consideration of effects of WWR and BO alone instead of considering combined effects of climatic conditions, structural insulation characteristics, façade configurations, and presence of shading devices on energy consumption as observed by Marino et al. (2017). This was also found during the thermal sensation survey in the case area as the respondent living in both north-east and south oriented buildings reported hot inside the building during summer.
5. Conclusions In this study, the effects of several WWR and orientation strategies applied to a typical case of a residential building in the warm and humid climate of Butwal were evaluated. The analysis and the obtained results showed that south orientation is best for the construction because when south-oriented, a building can save up to 0.2 % of annual thermal load. When oriented towards the north-east, the annual thermal load is increased by 1.31%. When the WWR and orientations were varied to quantify their effects on the building’s energy consumption, the result confirmed the annual thermal energy consumption is increased linearly with an increase in WWR. The study concludes that the careful considerations of WWR and the south orientation during designing of the building will contribute to efficient energy consumption in residential buildings.
136 Uprety et al. Journal of Engineering Issues and Solutions 1 (1): 129-137 [2021]
Since the study was conducted by selecting only one typical case out of the total surveyed house, the results may not be qualified for the generalization. Hence, more studies are needed to include all possible residential typologies prevailing in the case area.
Conflict of Interests Not declared by authors.
References ANSI/ASHRAE. (2010). ASHRAE Standard 55-2010 "Thermal Environmental Conditions for Human Occupancy". Atlanta: ASHRAE. Bodach, S., Lang, W., & Hamhaber, J. (2014). Climate responsive building design strategies of vernacular architecture in Nepal. Energy and Buildings, 227–242. Butera, F. (2005). Glass architecture: is it sustainable? Passive and Low Energy Cooling for the Built Environment, (pp. 161- 168). Santorini, Greece. Chi, F., Wang, Y., Wang, R., Li, G., & Peng, C. (2020). An investigation of optimal window-to-wall ratio based on changes in building orientations for traditional dwellings. Solar Energy, 64-81. CIBSE. (1999). Environmental design, CIBSE Guide A. London: The Yale Press Ltd. DHM. (2020). Department of Hydrology & Meteorology. Kathmandu, Bagmati, Nepal. Elghamry, R., & Azmy, N. Y. (2017). Buildings orientation and its impact on energy consumption. Al Azhar 14th International Conference (AEIC) on Engineering, Architecture & Technology, (pp. 1-15). Feng, G., Chi, D., Xu, X., Dou, B., Sun, Y., & Fu, Y. (2017). Study on the Influence of Window-wall Ratio on the Energy Consumption of Nearly Zero Energy Buildings. Procedia Engineering, 730–737. Ghosh, A., & Neogi, S. (2018). Effect of fenestration geometrical factors on building energy consumption. Solar Energy, 94- 104. Hassan, A. G. (2016). Parametric Design Optimization For Solar Screens: An Approach For Balancing Thermal And Daylight Performance For Office Buildings In Egypt. Giza: Faculty Of Engineering, Cairo University IEA. 2019., World Energy Statistics and Balances 2019. Marino, C., Nucara, A., & Pietrafesa, M. (2017). Does the window-to-wall ratio have a significant effect on the energy consumption of buildings? A parametric analysis in Italian climate conditions. Journal of Building Engineering, 169– 183. Nair, R. R., Shukla, A., Shekhar, S., & Jatav, A. (2014). Analyzing the effect of building orientation, varied WWR, and building height on solar heat gain and the internal temperature of university buildings located in a composite climate of India. International Journal of Science, Engineering and Technology, 955-973. NBC:206, 2. (2015). Nepal National Building Code. Department of Urban Development and Building Construction. Kathmandu. Tamimi, A., & Fadzilb, S. (2010). Experimental and Simulation Study for Thermal Performance Analysis in Residential Buildings in Hot-Humid Climate (Comparative Study). Science & Technology, 17-25. WECS. (2010). Energy Sector Synopsis Report 2010. Water and Energy Commission Secretariat. Nepal. https://www.wecs. gov.np/ WECS. (2014). Energy consumption situation in Nepal year 2011/2012. Nepal.
137 Journal of Engineering Issues and Solutions
Design and simulation of components of vacuum forming machine using household vacuum cleaner
Navaraj Adhikari1, Nirajan Sharma Timilsina1, Sanskar Gautam1,*, Snehraj Kaphle1, Pratisthit Lal Shrestha1 1Department of Mechanical Engineering, School of Engineering, Kathmandu University, Dhulikhel, Kavre, Nepal *Corresponding email address: [email protected] Received: January 27, 2021; Revised: March 08, 2021; Accepted: March 30, 2021
Abstract Plastic products ranging from toothbrushes to smartphones are an inseparable commodity in daily human life and their impact cannot be underestimated. This paper aims to design and simulate the vacuum forming process using readily available materials in context of Nepal. Vacuum forming process is a thermoforming process where the heated plastic sheet derives the shape of the mold through the application of vacuum and is used to make packaging products and other household products. Simulations were done to find out the optimum distance between the plastic sheet and the heater, arrangement of the wire in the heater, load bearing capacity of the design and the flow of vacuum in the arrangement. Nichrome wire coiled as heater coil is used as the heating material and laid in a spiral path with the plastic sheet 35mm below provided the best heating results and 1800W vacuum cleaner provided the necessary pressure of 85-90kPa and velocities of 100- 115m/s while the steel posts provided adequate strength.
Keywords: Coiled heater, mold, vacuum forming process, household vacuum cleaners
1 Introduction Plastic is material consisting of any of a wide range of synthetic or semi-synthetic organic compounds that are malleable and so can be molded into solid objects. Plastics have been in use in almost every aspect of our lives. Plastics due to their low cost, ease of manufacture, versatility, and imperviousness to water, are used in as many products of different areas, ranging from paper clips to spacecraft. They are specially used in area of packaging and home piping. (Science History Institute, n.d.). Amongst the various methods of processing plastics, vacuum forming is one of the most common methods for making plastic products. The process involves heating a plastic sheet until soft and then draping it over a mold. A vacuum is applied sucking the sheet onto the mounds. Using prototypes for the mounds makes
138 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] it economically feasible to produce low quantities of large parts and to operate medium size production runs. The sheet is then removed from the mold. As low pressures are used, the vacuum forming process is a low-cost process. (Awari Mahesh Prakash, 2016). Low vacuum pressures available through vacuum pumps/ cleaners, heating through ceramic heating wires and easy availability of plastic sheets make vacuum forming process suitable for low cost operations. (Sumit G. Wankhade, 2018) With the significant advantage of low cost initial setup and operation, the paper aims to design a vacuum forming machine and simulate the various components for best specification and configuration.
2 Methodology This study follows a methodological flow chart as shown in Fig. 1.
Figure 1: Study design
2.1 Working Principle The vacuum forming process is explained as follows and depicted in Fig. 2. a. The vacuum forming tool or mold is loaded in to the vacuum forming machine and warmed b. The plastic material, in sheet form, is loaded on to the vacuum forming machines material clamp. c. Heaters, located above the sheet, then heat the sheet of vacuum forming material until it softens. d. An automatic levelling device then supports the softened sheet of vacuum forming material with air. e. The mold is either raised to meet the bottom surface of the sheet of material or the clamp is lowered and a vacuum of air is applied in order to draw the sheet over the shape of the vacuum forming tool (mold). The plastic is cooled with air to set hard. The vacuum forming can be removed by hand or with the use of air
139 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al. to carry out any necessary secondary operations e.g., trimming (by hand). (Toolcraft Plastics, n.d.)
Figure 2: Vacuum forming process : 1-sheet of plastic is heated, 2-heated plastic takes shape of mold, 3- plastic sheet is removed (PVC Plastic Sheets, n.d.)
2.2 Components The major important components in this process are:
• Clamp: The clamp holds the plastic sheet to be formed. It can be made from wood. (Gringery, 1999) • Heaters: The heaters will heat the plastic sheet to temperatures ranging from 80°C to 140°C through radiation. (Gringery, 1999) • Mold: The mold provides the shape of the product. Molds can be made from aluminum, wood, fiberglass etc. The mold has holes so that the air can be sucked out of the mold when the material forms creating vacuum. (Gringery, 1999) • The Vacuum System: Vacuum pumps are used to suck the air out of mold. The vacuum pumps should be able to main a pressure under 1 bar, approximately -0.83 bar. (Gringery, 1999) • Plastic Sheet: The material of the plastic sheet is to be determined so as to heat the sheet accordingly to its forming temperature. (Gringery, 1999) • Trimming Tools: After the sheet is formed and released, it needs to be trimmed to remove the unnecessary parts of the sheet and get the required product Depending on the thickness of the sheet, snippers to chisels may be used. (Gringery, 1999)
2.3 3D Modelling The heating chamber consists of nichrome wire in loops that will heat the plastic sheet. The clamps will hold the plastic sheet and will allow for the movement of plastic sheets. The vacuum table consists of holes which will create the necessary suction pressure for the vacuum to be applied. The vacuum to be created is to be applied through the vacuum input hole as shown in Fig. 3.
140 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
Figure 3: CAD model 2.4 Selection of Components The major aspects that determine the quality and efficiency of the process are: i. The selection of heating source ii. The vacuum created iii. The time of heating for the plastic to be used
2.4.1 Selection of heating source For automatic vacuum forming system in industries for mass production, quartz heaters with pyrometer are used as it has less thermal mass compared to other heaters. (Science History Institute, n.d.)But for low cost initial setup, ceramic heater can be taken as a better option. Ceramic heaters consist of coiled resistance wire (such as Nichrome, Kanthal, and Constantan) elements set in molded china clay or ceramic. Due to following properties of ceramic heater, it is a better option compared to other heaters; 1. Slightly more thermal mass compared to quartz heater but has comparatively less thermal mass than other heaters. 2. Available in round, square or rectangular shapes, they can be flat (for maximum proximity) or curved (to provide a parabolic reflector which radiates more effectively). 3. They radiate long wavelength heat which is readily absorbed by thermoplastics.
2.4.2 Selection of vacuum source The vacuum to be created depends upon the thickness of plastic sheet. The atmospheric pressure is 101700 Pa. (Metrological Forecasting Division, 2020) Generally, the vacuum generated by vacuum cleaner of 1400 Watts to 1800 Watts is about 98,205 Pa to 88,046 Pa. (Gringery, 1999). The aim of vacuum (P) is to create the drag force (F) through the hole of area A in the vacuum table. The drawing force of vacuum is given by, F=P×A (1) For ‘n’ no of holes, the vacuum pressure is given by, F=n ×P ×A (2) The vacuum cleaner to be used will be 1800-Watt drum vacuum cleaner having suction power of 350 W.
141 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
Table 1: Plastic and thier properties
Forming S Plastic Type Properties Temp. Colors Price Applications Availability N (°C) • hard, rigid, good impact strength and Acrylonitrile luggage, weather resistance 1. Butadiene 105 L U sanitary parts, R • available in different Styrene (ABS) electrical parts texture and finishes • good formability Copolyster • Easily forming, good 2. (PETG/ impact strength 120-200 L E Medical; food U VIVAK) • transparent
• Hard, rigid, high Light diffusers, Polycarbonate impact resistance, 3. 150 A E signs, machine Y (PC/ Lexan) • good formability, guards • translucent
• High forming temperature Food Polypropylene • good impact 4. 105-180 L I containers, Y (PP) strength toys, tanks • Sheet sag is inevitable Caravan parts, • Successful forming vehicular Polyethylene 5. • Good impact 140-196 A I parts, Y (PE) strength enclosures, housing • Strong & tough thermoplastics Packaging, Polyvinyl • Transparent in thin machine 6. chlorine gauge 110-140 A I Y guards, car (PVC) • Good impact tyres strength, fire resistive • Homopolymer Food • Transparent packaging, • Rigid brittle, poor medical Polystyrene UV resistance 7. 100 A I applications, Y (PS/HIPS) • Low impact strength electronics, but high tensile cosmetic, cups, strength (35-55 container MPa) A-Any, E-Expensive, I-Inexpensive, L-Limited, R-Rare, U-Unclear, Y-Yes
142 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
2.4.3 Selection of plastic sheet PVC is selected as the material of choice for the purposes of this study because of its availability in sheets, inexpensiveness and relatively easily attainable forming temperature of 110°C. Properties of various types of plastics are summarized in Table 1 and material selection for various components in Table 2. Table 2: Material selection of components
S.N. Component Material Description 1400-watt drum vacuum cleaner 1 Vacuum Source Vacuum Cleaner capable of 350 watts suction pressure Heat Source 1400 watts 2 Nichrome wire coil Length of wire: 10 m Thickness: 1 mm 3 Plastic Sheet PVC sheet Size: 270 mm × 210 mm Heating Chamber & Vacuum Aluminum Plate / Holes in vacuum table to be made 3 Table Ply board through drilling 4 Pillar Square steel hollow tube 5 Clamp Mild Steel Plate
2.5 Wire Calculations Nichrome wire was selected as the heating material in this project owing to its thermal properties and availability.
2.5.1 Wire dimension calculations The heater coil is to be formed from a nichrome wire. For purposes of the project simulation, the dimensions of nichrome wire are assumed as:
Diameter of heater coil D=20 mm Diameter of heater coil wire d=1.016mm=18 AWG Total length of heater coil wire l=10m≈32ft
Length in one turn l1= π(D-d)=19π mm
No of turns
2.5.2 Power generated by nichrome wire The wire is to be coiled up and laid in a path and let the wire maintain a temperature of over 500°C using normal AC current.
So, resistance of wire, R=13.5Ω If the temperature is to be maintained at 538°C (Hotwireformcutterinfo, n.d.), then current, I=10.1 A
143 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
From Ohm’s law, operating voltage is
V=I ×R=10.1 ×13.5 =136.35 V From Joule’s law, power output is P=I ×V (3) (3) =136.35×10.1 =1377.22 W
2.6 Heat Transfer Calculations Vacuum forming is a thermoforming process where the plastic sheet is heated to its forming temperature and then processed over the mold by vacuum. Heat is transferred from the heating source to the plastic sheet. Heat is transferred from the wire usually through radiation. The three modes of transfer of heat are conduction, convection and radiation. The numerical analysis is done to find out the medium of heat transferred. The setup is as shown in Fig. 4.
Figure 4: Set-up in CAD model The coil is made from nichrome and is heated to temperatures around 538°C with the PVC plastic sheet placed 2cm below as shown in Fig. 5. Both the coil and plastic sheet are in the dimensions of A4 (297mm x 210mm). The atmospheric temperature is assumed to be a 25°C. The coil is assumed as a screen or a horizontal surface with lower side hot because the coil heats the plastic sheet placed at its bottom and the upper part of the coil is placed on an insulated rack.
144 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
Figure 5: Schematic diagram showing arrangements of Nichrome wire and plastic sheet
2.6.1 Heat transferred by convection The lower surface of the coil gets heated to temperatures of around 538°C. This causes the air in contact with the surface to flow leading to heat transfer by convection. Since, no fans are used, it is natural convection. Also, the coil is assumed as horizontal surface having hot bottom side with the dimensions of A4. For horizontal surface facing downwards, the characteristic length is
(4)
Surface temperature is and the atmospheric temperature is Film temperature is
(5)
At film temperature for dry air properties values are (C.P. Kothandaraman, 2007)
Dynamic Viscosity Prandtl Number Thermal Conductivity
Coefficient for thermal expansion of air
The Rayleigh number is
(6)
Nusselt’s number is
(7)
Coefficient of thermal convection is
145 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
(8)
From the schematic diagram in Fig. 5, the heat transfer can be analyzed by steady state analysis (please refer Fig. 6) as the temperature difference remains constant.
Figure 6: Steady state analysis for convection Area is A=.297×.210=0.06237 m2
Length of air film, ,la=20 mm=0.02 m
-3 Length of plastic sheet, lp=1mm=10 m Thermal conductivity of air at 25°C, ka=0.02634 W/mK Thermal conductivity of PVC, kp=0.19 W/mK Coefficient of convection, h=5.71 W/m2 K Total resistance is
Rtotal=Rconvection+Rair,conduction+Rplastic.conduction (9)
Rtotal
The =15.07total heat ℃/W transfer by steady state is
146 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
(10)
2.6.2 Heat transfer by radiation The setup is as shown in the schematic diagram in figure4 The various values are as follows
From above calculations it can be seen that radiation is the main mode of heat transfer rather than convection. Hence, the effectiveness of the heating setup can be calculated as
(12)
3 Radiation Simulation The radiation simulation was performed for the vacuum forming machine for the following purposes:
147 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
• to understand the spread of heat in the plastic sheet due to different wiring path configurations • to find out the optimum distance between the coil and the plastic sheet Different wire path configurations affect the temperature distribution in the plastic sheet. The variation of distance affects the amount of heat to be absorbed by the plastic sheet and also, the variation helps to control the heat and thus minimize the risk of melting of plastic sheet.
3.1 Simulation for Wire Path Configuration 3.1.1 Setup Nichrome wire, which is the heating element in the machine delivers 900W to the plastic sheet placed parallel 2cm below. Due to computing issues, the whole solenoid structure of the wire is not used for the simulation, rather a wire of the same diameter and the path structure were used. The simulation tested spiral path, double spiral path and longitudinal path structures of a 12 AWG (2.05 mm) nichrome wire. The setup is as in Fig. 7.
Figure 7: Set-up for wire configuration 3.1.2 Results The simulation was done in Solidworks Flow Simulation to find out the spread of heat in the plastic sheet due to different wire paths. From the setup as mentioned above following results (Fig. 8, Table 3) were found: As the plastic sheet, placed on clamps, gets heated, the sheet expands and drags downwards until it melts. Obviously, due to gravity, the centre most part of the sheet would drag down the most. At both spiral and longitudinal paths, the maximum temperature was above 200 at the centres with longitudinal having a larger area above this temperature. So, both wire path configurations are very risky to be used in heating the plastic sheet due to the risk of melting. However, in the double spiral wire configuration,℃ the heat is unevenly divided and is maximum at the centres of each quadrant of the plastic sheet which poses the risk in melting at these sites. So, considering the risk of melting, the spiral wire configuration is concluded to be used as it offers a minimum area for melting of plastic sheets. 3.2 Simulation for Distance Between Coil and Plastic Sheet 3.2.1 Setup In this simulation the nichrome heating coil is made into a spring of diameter 2cm and free length of 5cm with pitch of 5mm. So there are 10 turns in the setup coil. The diameter of wire of spring is 1.016mm i.e., 18AWG as in other previous simulation and this wire is heated to 538°C or 811 K. Following this, the plastic sheet and the coils were arranged for different vertical distances. The simulation was done for 2cm, 5cm, 8cm and 10cm distances. Likewise, the material of wire is nichrome and the plastic sheet is PVC and the
148 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] default values of emissivity were given as in solid works. Emissivity for nichrome and PVC was 0.98 and 0.91 respectively. The simulation was done in Solidworks Flow Simulation.
Figure 8: Simulation result for spiral coil (top left), double spiral coil (top right) and longitudinal coil (bottom) Table 3: Simulation results for wire path configuration
Coil path Max Temp (°C) Min Temp (°C) Heat distribution Maximum at center and gradual Spiral 252 122 decrease to the outer sides Maximum at the centre of each Double spiral 150 39 quadrants and heat is distributed as kidney shaped Maximum at center at an larger area Longitudinal 252 108 than spiral coil
149 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
3.2.2 Result The simulation results are summarized in Table 4 and visualized in Fig. 9 and Fig. 10. Table 4: Simulation results for distance between coil and sheet
Coil-sheet Max Temp Forming temperature spread Remarks distance (°C) (regions above 110°C) 2cm 190 Above forming temperature 50 mm along the path of wire In forming temperature 3.5 cm 124 30mm along the path of wire range(110-140 °C) 5cm 103 Below forming temp - 8cm 30 Below forming temp - 10cm 30 Below forming temp -
Figure 9: Simulation results for 2cm (top), 5cm(top right), 5 cm(bottom left) and 10cm(bottom right)
From above simulations following can be concluded:
• 2cm-3.5cm is the best distance between the coil and plastic sheet to reach the forming temperature • Distances above 3.5cm result in heating of plastic sheet below its forming temperature • Considering the width of temperature distribution in plastic sheet, 3.5 cm results in a lower area of temperature above 110°C while 2cm results in larger area, so 3.5 cm distance can lower the risk of melting of PVC. • Since, PVC melts at 150°C, heat transferred by 2cm distance resulted in temperatures of 195°C leading to maximum chances of melting of plastic sheet. However, the maximum temperature offered by 3.5 cm is 124°C, so 3.5 cm distance is the safest option to prevent melting of plastic sheet. • Hence, it can be concluded that 3.5cm is the best vertical distance between the plastic sheet and the coil.
150 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
Figure 10: Temperature distribution in 2cm (left) and 3.5cm (right)
4 Fluid Flow Simulation The fluid simulation was performed for vacuum forming machine for the following purposes: • to understand the velocity distribution from several input to single output • to find out pressure distribution Different values of velocity and pressure of a vacuum for suction has different behavior for final product and also, these wide range of values helps to control the suction pressure to get best version of the final product.
4.1 Calculations Considering the vacuum cleaner works at full efficiency (i.e., there are no losses), the fluid flow can be analyzed using Bernoulli’s equation.
Figure 11: Schematic diagram and CAD diagram
151 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
In the schematic diagram (Fig. 11), I, II and III represent the regions of setup at the inlet of the vacuum hose, inside the vacuum chamber and at the holes respectively with P, v and z representing the pressure, velocity and the height at the concerned regions respectively Corresponding areas in respective regions as per CAD design. Known parameters for calculation and areas of different regions are shown in Table 5 and Table 6. Table 5: Known parameters for calculation
Table 6: Areas of different regions
If in the setup, a vacuum cleaner with suction flow rate of 1.9m3/min (Total 1400 Watt Vacuum Cleaner TVC14301, n.d.)is used, which is connected to the vacuum hose inlet i.e., region I, then the velocity at the inlet is
Since no losses are assumed, we can use the continuity equation i.e., the mass flow rate remains constant. However, air gets compressed at the holes and the density decreases. Considering the fluid flow in vacuum cleaners, the effect of density of air is assumed to be negligible. Hence the analysis is done using constant volumetric flow rate. Since, there is only one inlet region and one outlet region, from continuity equation, we have, (13)
152 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
In regions I and II, applying Bernoulli’s equation, we get
(14)
Since,z1=z2, i.e., both are at same height
(15)
In regions II and III, applying Bernoulli’s equation, we get
(16)
Using (15) in above we get
(17)
Applying values from table 1, we get
Applying values of velocities from above and density of air as 1.23 kg/m3, we get
So, the pressure difference due to the velocities is just 0.72kPa. Also, there is a sudden expansion of area from region III to II and the corresponding head loss is
Also, there is sudden contraction of area from region II to I and the corresponding head loss is
153 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
For A1/A2 = 0.04 , K = 0.45 (Losses Due to Sudden Contraction, 2021), hence,
So, the total pressure difference is
4.2 Simulation For the setup of fluid simulation, vacuum section is divided into simple elements that can be used as discrete local approximation, generally known as mesh as shown in Fig. 12. These influences the accuracy, convergence and speed of simulation. For this section, there are 73,927 nodes and 387,314 elements.
Figure 12: Meshing of CAD model
4.3 Results In the vacuum section, the maximum velocity of 111.7 m/s was recorded at the tip of the hole. The velocity variation can be seen unevenly distributed from input to output. Similarly, the maximum pressure of 86,150 Pa was recorded ar the wall. The pressure variation can be seen as shown in figure 34. Also, the minimum pressure was 85,040 Pa recorded at the entry holes. The pressure diffrence is 1.1 kPa. The pressure deviations found theorotically was 1.3 kPa. The results are depicted in Fig. 13.
5 Structural Simulation
5.1 Calculation for Steel Pillar Here, Modulus of elasticity of steel, E=2.14×1011 2
Cross sectional N/m area, A=40×20 mm2 =800 mm^2 Assume, Applied load, P =10 N
154 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
Figure 13: Fluid flow simulation results for velocity(above) and pressure(below) We know that stress,
Again,
So, the total deformations are
155 Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021] Adhikari et al.
5.2 Results of Simulation of Vertical Frame From above theoretical calculations and ANSYS simulation analysis, the equivalent stress is 0.0125 N/mm2 N/mm2 and 0.0121 N/mm2 respectively. Both values are nearly equal, so the simulated structure resists the force and it is feasible and applicable.
Figure 14 Equivalent stress simulation Figure 15 Total deformation simulation
6 Conclusions From the various simulations the following can be concluded: • The heating element-nichrome wire coil, should be placed in a spiral path to achieve the better heat distribution and minimum melting risk in comparison to longitudinal and double spiral path configurations • For heater coil diameter 20mm, the optimum distance between the nichrome wire and plastic sheet was found to be 35mm • The structural steel pillar was able to supports the clamp and the heater. • 1400 watt household vacuum cleaner was able to provide velocity of 100-115 m/s at the holes and vacuum pressure of 85-90 kPa which is suitable for vacuum forming.. The above results are only based on simulation and need to be validated by testing.
Acknowledgements The authors would like to thank Kathmandu University for carrying out this research and Mr. Pratisthit Lal Shrestha for supervising this research.
Conflict of Interests Not declared by authors.
References Awari Mahesh Prakash, P. S. (2016). Design & Development of Vaccum Forming Machine & Die. International Journal on
156 Adhikari et al. Journal of Engineering Issues and Solutions 1 (1): 138-157 [2021]
Theoretical and Applied Research in Mechanical Engineering (IJTARME), 5(1). C.P. Kothandaraman, S. S. (2007). Heat and Mass Transfer Databook. New Delhi: New Age International (P) Ltd, Publishers. Gringery, V. (1999). Secrets of Building a Vacuum Forming Machine. David J. Gringey Publishing. Hotwireformcutterinfo. (n.d.). Calculations Nichrome. Retrieved January 17, 2021, from http://hotwirefoamcutterinfo.com/_ NiChromeData.html Losses Due to Sudden Contraction. (2021, March 5). Retrieved from NPTEL: https://nptel.ac.in/content/storage2/ courses/112104118/lecture-14/14-7_losses_sudden_contract.htm Metrological Forecasting Division. (2020, March 14). (Department of Hydrology and Metrology, Government of Nepal) Retrieved March 14, 2020, from http://www.mfd.gov.np/city?id=31 PVC Plastic Sheets. (n.d.). (India Mart) Retrieved March 20, 2020, from https://www.indiamart.com/proddetail/5mm-pvc- sheet-14943147130.html Science History Institute. (n.d.). The History and Future of Plastics. (Science History Institute) Retrieved November 11, 2019, from https://www.sciencehistory.org/the-history-and-future-of-plastics Sumit G. Wankhade, N. L. (2018). Design and Fabrication of Small Format Vacuum Forming Machine. International Journal for Research in Applied Science & Engineering Technology (IJRASET), 2045-2048. Toolcraft Plastics. (n.d.). Vacuum Forming Process-Explanation, Schematic View & Video. Retrieved November 11, 2019, from https://www.toolcraft.co.uk/vacuum-forming/advice/help-vacuum-forming-process.htm Total 1400 Watt Vacuum Cleaner TVC14301. (n.d.). (Hardwarepasal.com) Retrieved January 22, 2020, from https://www. hardwarepasal.com/product/total-1400-watt-vacuum-cleaner-tvc14301
157 Journal of Engineering Issues and Solutions
Development of rainfall – runoff model for extreme storm events in the Bagmati River Basin, Nepal
Nirajan Devkota1,*, Narendra Man Shakya2 1 Department of Civil Engineering, Kathford International College of Engineering and Management, (Affiliated to Tribhuvan University), Lalitpur, Nepal 2 Department of Civil Engineering, Pulchowk Campus, Institute of Engineering, Tribhuvan University, Lalitpur. Nepal *Corresponding Email: [email protected] Received: January 29, 2021; Revised: March 29, 2021; Accepted: March 31, 2021
Abstract This study is based on the Bagmati river basin that flows along with the capital city, Kathmandu which is a small and topographically steep basin. Major flood occurring in 1993 and 2002 as stated in the report of DWIDP shows that the basin is subjected to water-induced disaster in monsoon season affecting people and property. This study focuses on the development of a rainfall-runoff model for Bagmati basin in HEC-HMS using the Synthetic Unit Hydrograph (SUH) with Khokana as the outlet. The coefficients for SUH like Lag time coefficient (Ct), peak discharge coefficient (Cp), unit hydrograph widths at 50% and 75% of peak and base time were determined calibrating the Synder’s equation where Ct varies from 0.244 to 1.016 and Cp varies from 0.439 to 0.410. The rainfall-runoff model in HEC-HMS has been calibrated from daily data of 1992-2013 and validated from hourly data for July 2011, August 2012, and July 2013. Furthermore, the model has been tested to compare the discharge for various return periods with the observed ones which are in close agreement. The determination of Peak Maximum Flood (PMF) using the calculated Peak Maximum Precipitation (PMP) is also another application of the model which can be used to design various hydraulic structures. Thus the values of coefficients, Ct and Cp can be used to construct unit hydrograph for the basin. Moreover, the satisfactory performance of the model during calibration and validation proves the applicability of the model in flood forecasting and early warning.
Keywords: Bagmati basin, event-model, HEC-HMS, unit hydrograph
1. Introduction While majority of the basins worldwide are ungauged (Hrachowitz et al., 2013), hydrological behavior in ungauged basins are poorly known. It has implications on estimating runoff, water availability, flood events, and inundation depth/extent. A rainfall-runoff model usually produces the runoff hydrograph as a response to a rainfall hyetograph as input (Beven, 2012). The actual shape and timing of the response hydrograph for a particular watershed is a function of many physiographic, land use, and climatic variables (Chow, 1988).
158 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021]
Several efforts are made over the years to overcome the prediction and forecasting challenges in ungauged basins. They include but not limited to developing a wide variety of data acquisition techniques: as improving rainfall-runoff measurements using real-time monitoring, weather radars, and satellites (Li & Shao, 2010) and enhance the capability of hydrological modeling by using geographic information system (GIS) (Jain et al., 2004). Though those efforts have resulted in a set of new tools and methods, putting the approaches of prediction in ungauged basins to practice is still a challenge (Efstratiadis et al., 2014). A variety of simulation models, ranging from simple empirical models (e.g. transfer function, data-driven, regression, etc.) to complex physically-based distributed models are available for generating a watershed response in both gauged and ungauged basins. Simulation using physically–based models are generally better compared to lumped and empirical models (Hughes, 2010), as they can account for the system heterogeneity and simulate the hydrological process in a watershed (Cibin et al., 2014). However, in the case of ungauged locations, a synthetic UH is a simple and effective method of the rainfall-runoff simulation (Fedorova et al., 2018). The UH can then be used as a transfer function to transfer rainfall into runoff using appropriate hydrological models (Saghafian, 2006). When there is no observed long-term discharge, prediction of runoff is more challenging to practical applications such as design of drainage infrastructure, flood forecasting, and for watershed management tasks such as water allocation and climate impact analysis. In the case of ungauged river basin, one of the options is to develop Synthetic UHs for runoff prediction, flood forecasting as well as early warning system (Sherman, 1932). To develop UH to a catchment, detailed information about the rainfall and the resulting flood hydrograph is needed. However, such information would be available only at a few locations, and in a majority of catchments especially those which are at remote locations, the data would normally be very limited. In order to construct UHs for such areas, empirical equations of regional validity which relate the hydrograph characteristics to the basin characteristics are to be available. Unit hydrographs derived from such relationships are known as synthetic – unit hydrographs (Bhunya, 2011). Some of the physical characteristics of the watershed to develop hydrograph includes peak flow rate (Qp), time to peak (tp), time base (tb), and width of unit hydrograph at 0.5Qp and 0.75Qp (i.e., W0.5 and W0.75), respectively. In addition, simultaneous adjustments are required for the area under the synthetic UH to be unity. There are three types of synthetic UHs. One among them relates hydrograph characteristics (peak flow rate, base time, etc.) to watershed characteristics (Snyder UH), while another one is based on a dimensionless unit hydrograph (Soil Conservation Service), and last one is based on the model of watershed storage (Clark Unit hydrograph). However, these relations are quite empirical and as such cannot be expected to be universally applicable. Their applications, in general, should be restricted to the region in which they have derived Hoffmeister et. al. (1977) developed a synthetic UH for an un-gauged basin in New Zealand and Sudhakar et al. (2015) tested three different methods viz. Snyder method, Common’s dimensionless method, SCS dimensionless hydrograph for India. The study concluded that Snyder UH method gives the best results as compared to later ones. Due to lack of hourly measured rainfall and runoff long term data our work also applies the Snyder UH. It requires only basin characteristics as input. The basin characteristics of the watershed are extracted from the HCE – GeoHMS model, a data preparation tool for HEC – HMS (Hydraulic Engineering Center – Hydrologic Modelling System) model. For the city basin like Baghmati, the density of the observation network is spare but issues of flooding and inundation prevail due to extreme rainfall, topography, blockage of natural drainage, and poor planning/ design/construction practices. Such issues are more frequent in highly urbanized Kathmandu Valley, the capital city of Nepal. In such a case, we can develop a synthetic hydrograph by calibrating the continuous rainfall data and validating with different rainfall events to make the model applicable for predicting
159 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya flooding/runoff at several ungauged locations in the urban area. The objectives of this study are: i) to develop a synthetic UH for an ungauged basin; ii) to develop the rainfall-runoff model using HEC-HMS and iii) evaluate the applicability of the model for estimating flood of different return periods as well as probable maximum flood using the developed model.
2. Materials and Methods 2.1 Study Area The Kathmandu Valley, located in the upper part of the Bagmati River Basin is chosen as a system or basin boundary. It originates at Shivpuri Mountain (2679m) and draining the Kathmandu Valley, the river flows through the Middle Mountains and Siwalik range before entering into Terai (see Fig.1). The basin area above Khokana is 612 km2 and the elevation of the basin varies from 1260 m to 2679 meters above the mean sea level (masl). The average length of the river up to Khokana is 35 km and the average gradient is 0.0025. There are several tributaries of different orders originating from the middle mountains that feed the Bagmati River. These are Manahara, Dhobi Khola, Bishnumati, Balkhu Khola, Hanumante Khola, and Nakkhu Khola. The basin always faces the problems of flash floods and inundation during the rainy season, which cause severe human and property losses. The real-time rainfall and runoff observations are therefore essentially required in the basin to conduct flood and inundation predictions. This study focused on tributaries of the upper urban reach of the Bagmati basin and developed a rainfall-runoff model and analyzed/ interpreted the simulated results.
Figure 1: Location and associated details of the Kathmandu Valley watershed in the uppermost part of the Bagmati River Basin
160 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021]
2.2 Development of Synthetic UH As shown in the overall methodological framework (see Fig.2), Synthetic UH methods were utilized to determine runoff hydrograph for ungauged sites. The physical parameters of sub-basins (i.e., area of sub- basin, length of the river, centroidal length, the slope of the river) were extracted from HEC – GeoHMS and processed further in MS –Excel program for generation of synthetic UH. The synthetic UH of Snyder is based on relationships between the characteristics of a standard UH and basin morphology. Snyder formulates an equation for the effective rainfall duration (Tr), the peak direct runoff rate (qp), and the basin lag time (Tl). From these relationships, the following five characteristics of a required UH for a given effective rainfall duration was calculated (Chow, 1988): the peak discharge (qp), the basin lag (Tl), the base time (Tb), and the widths, W (in time units) of the UH at 50 and 75 percent of the peak discharge. In this particular study, the model parameters we seek to estimate are Snyder’s equations coefficient, namely, coefficient of slope (Ct) and coefficient of peak (Cp). The method calibrates the Snyder coefficients (i.e., Ct and Cp) such that direct runoff must be 1cm or within 10% relative error and the observed hourly maximum discharge and calculate peak discharge within 10% difference. The calculated peak discharge is obtained by multiplication of ordinate of Synthetic UH and point rainfall depth. In addition, this method determines the peak discharge, lag time, and time to peak by using characteristic features of the watershed. Also, the lag time is compared with the time of concentration (Tc) such that Tl range from 50 to 75 percentage of Tc. It is generally taken as 60% of Tc (USACE, 2005). The Tc is calculated from the Kripich formula. Snyder's equations are described as follows:
• Basin Lag (tp): tp = 0.75Ct (L × Lc)0.3; where L is Length of stream, and Lc is centroidal length of stream. • Peak Discharge (Qp): Qp = ; where Cp is the Coefficient of the peak, A is Basin area and Tp is time to peak. • Base period (Tb): Tb =5 (Tp + D/2); where Tp is time to peak and D is unit duration. • Unit Duration (Tr): Tr = Tp /5.5 • Correction for actual Duration: Tl1 =Tp + (D – Tr)/4 • Width at 50% and 75% of peak discharge (W50 and W75): W50 = 5.9 / (qpl)1.08 & W75 = 3.4 / (qpl)1.08
2.3. Hydrological Model Development 2.3.1 Selection of a hydrological model The HEC HMS model was chosen for developing the hydrological model as it has the capability of simulating the rainfall-runoff process for dendritic watersheds in space and time (Oleyiblo & Li, 2010). Since the old version of HEC – HMS being lump model, 4.2.1 version of HEC – HMS model is used as a semi-distributed model. This model is used widely across the globe to simulate runoff to a wide variety of watersheds. The semi-distributed models evaluate basin response by partially representing spatial heterogeneity by dividing the basin into several sub-basins, depending upon the resolution of available input data (HMS, 2000). HEC – HMS model has been used successfully in different parts of the world, including Nepal, for catchment modeling. Some examples include determining hydropower potential (Prajapati, 2015), climate change impact assessment (Babel et al., 2014), and development of the rainfall-runoff model (Khadka & Bhaukajee, 2018).
161 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya
Figure 2: Methodological framework adopted in this study
2.3.2 HEC – HMS model set – up Basically, HEC-HMS requires four model components: Basin model, meteorological model, control specifications, and input data (time series, paired data, and gridded data). The Basin model describes different elements of the hydrological system (subdivision, reaches, junction, sources, sinks reservoir, and diversion) and their connectivity that represent the movement of water through the drainage system (Guide & Manual, 2008). As present in Fig.2 the HEC – GeoHMS extension in ArcGIS was selected to derive topography-related inputs to HEC-HMS. It usages the spatial analyst extension in ArcGIS to develop hydrologic parameters as input data for HEC – HMS (Vizina & Hradilek, 2012). Analyzing the digital terrain information, HEC – GeoHMS transforms the drainage paths and watershed boundaries into a hydrologic data structure that represents the watershed response to precipitation. The spatial analyst extension was used for terrain pre- processing based on digital elevation model (DEM) and stream data. A terrain model was used as an input to derive eight additional data sets that collectively describe the drainage patterns of watersheds and allows for a stream of sub-basin delineations. The first five data sets in grid representation for the flow direction are DEM reconditioning, fill sink, flow accumulation, stream definition, and stream segmentation. The next two data sets are watershed polygons and the drainage line processing. The last one is aggregated watershed. Outputs after terrain preprocessing serve as a spatial database for the study. After terrain processing, the HMS project is generated for the study area. The stream gauging station
162 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] at Khokana (550.05) is considered as the control point for the project generation. Then the sub-basins automatically generated during terrain processing are merged and added at major river junctions. The study area is divided into 13 sub-basins as indicated in Fig. 3. The spatial-temporal precipitation distribution was calculated by the gauge weight method. The Thiessen polygon technique was used to determine the gauge weights, the area is shown in Table 1 and the following input data were used for the meteorological module: daily precipitation, daily temperature, elevation, and long-term mean monthly actual potential evapotranspiration. And finally, the HMS basin model was used to generate HEC – HMS model including a background map (see Fig. 3).
Figure 3: HEC - HMS model set-up HEC – HMS uses a separate model to represent each component of the runoff process. For every sub- basin and reach all the hydrological parameters have been initially estimated and simulated the models for observed boundary conditions to compute out either the watershed runoff hydrograph or a channel outflow hydrograph. Then the computed hydrograph is compared with the observed hydrograph to evaluate how well the model “fits” the real hydrological system. The model parameters were adjusted until satisfactory results are obtained. While setting up the HEC-HMS model, the following methods were selected for various hydrological processes. Loss Method: After the occurrence of precipitation, the loss method controls the partitioning between intercepted water, infiltrated, and the water that leaves the catchment as direct runoff. Water that survives a loss method leaves the catchment as quick flow. The loss method used in this study is the Deficit and Constant, which is a quasi-continuous model of precipitation loss where initial loss can recover after a prolonged period of no rainfall and is most suitable for continuous simulation (Majidi & Shahedi, 2012). The parameters of
163 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya this method are constant rate, initial deficit, and maximum deficit. Table 1: Thiessen area of rain gauge stations
Thiessen Area Rain gauge station Thiessen Weightage sq. km 1015 45 0.074 1022 48 0.079 1029 53 0.087 1030 30 0.049 1035 40 0.066 1039 73 0.120 1043 41 0.067 1052 24 0.039 1059 61 0.100 1060 48 0.079 1071 59 0.097 1073 32 0.052 1074 44 0.072 1075 12 0.020
Direct runoff method: The transformation method controls the channel surface runoff concentration time. Water concentration is recorded in a hydrograph, thus transformation methods attempt to build the right hydrograph using catchment characteristics (Halwatura & Najim, 2013). In this study, the Snyders user-specified UH method was selected. Objective functions of percent error in peak and volume were used to determine the best fit between observed and simulated hydrographs. The objective function of percent error in peak only shows the well-fitting of peak discharge of simulated and observed hydrographs. The objective function of percent error in volume calculations is based on just volumes of observed and simulated hydrographs. Base flow method: Base flow is influenced by groundwater and is closely related to watershed characteristics. The recession model has been used often to explain the drainage from natural storage in a watershed (Linsley et al., 1982). It defines the relationship of Qt, the base flow at any time t, to an initial value as: Qt = Qokt Error! Bookmark not defined. Where Qo = initial base flow and k = an exponential decay constant. In the HMS model, the variables for the base flow method by the recession are initial discharge, ratio to peak, and recession constant. Reach routing: A channel or reach s an element with one or more inflow and only one outflow. Inflow comes from other elements in the basin model. The routing models included in HEC – HMS program are the fundamental equations of open channel flow (the momentum equation and the continuity equation). Together the two equations are known as the St. Venant equations or the dynamic wave equations. The momentum equation accounts for forces that act on a body of water in an open channel. In simple terms, it equates the sum of gravitational force, pressure force, and friction force to the product of fluid mass and
164 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] acceleration. In one dimension, the equation is written as: Sf = So - dydx- Vgdvdx- 1gdvdt To solve this equation we have used the Muskingum - Cunge method (Barry & Bajracharya, 1995).
2.3.3 Model calibration and validation Automatic calibration in conjunction with manual calibration was used to determine a practical range of the model parameter values preserving the hydrograph shape and minimum error in volume. Nelder and Mead optimization method (Nelder & Mead, 1965) were used as automatic calibration algorithm, which aims to minimize a specific objective function, such as the sum of the absolute error, the sum of the squared error, percent error in peak and peak weighted root mean square error (Guide & Manual, 2008). This study selected the sum of squared error objective function for automatic calibration. Daily river discharge data for the period of 1992 - 2013 at Khokana (550.05) was used for model calibration. And the model was validated with storm event by real-time hourly data in the years 2011, 2012, and 2013. The objectives of the model calibration and validation were to match simulated volumes, peaks, and timing of hydrographs with the observed ones.
2.4 Rainfall Frequency Analysis For evaluating the applicability of UH, various scenarios were evaluated using rainfalls of different return periods. Rainfall frequency in the study area was analyzed based on rainfall data at 15 stations spread across the basin. Hydrognomon software was used for frequency analysis (Kozanis et al., 2010). With the best fit statistical method 5, 10, 25, 50, 100, 500,1000,10,000 year return period rainfall data have been calculated taking input as daily time series data. Extreme Value (EV) 1 – Max (Gumbel), Log Pearson – III, Extreme value method, and Lognormal distribution were used to determine extreme values. From Kolmogorov – Smirnov Test in Hydrognomn software, EV1 – Max (Gumbel) gave a good performance (98.6%) than other methods. Therefore, results from the Gumbel method were finally adopted.
2.5 Probable Maximum Precipitation and Flood Estimation According to WMO (2009), probable maximum precipitation (PMP) can be defined as the greatest depth of precipitation for a given duration meteorologically possible for a given size storm area a particular time of year, with no allowance made for long – term climate trends. Several techniques of PMP estimation are available based on the availability of rainfall data, catchment size and location, and meteorological conditions responsible for extreme rainfalls. A simplified statistical method developed by Hershfield (1961), recommended by the WMO (2009), and used widely across the globe is adopted in this study to estimated PMP. As per this method, PMP is estimated as; PMP = Xn + K × Sn
Where Xn̅ is the mean of maximum daily annual rainfall sample, Sn is standard deviation, and K is a factor depending on Xn and estimated using the following equation: ̅ The PMP thus estimated̅ was subject to adjustments of mean to maximum rainfall, standard deviation (SD) to maximum rainfall, and mean and SD to the length of a data record as per the guidelines in WMO (2009a). In this case, the adjustment factor for mean was 1.01, adjustment factor of SD was 1.06, and adjustment for area reduction factor was 1.1.
165 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya
The probable maximum flood (PMF) value, as well as hydrograph, was estimated by feeding PMP as rainfall input to the event model. In addition, floods of different year return periods were also estimated by feeding rainfall of different return periods.
2.6 Data and Sources Both geo-spatial and time-series data were used in this study as elaborated in Table 2. Table 2: Data and sources used in this study
Data set Data Resolution Data description/ Properties Data Source Unit type (Time Frame) Spatial EarthExplorer 30m × 30m Terrain (m) Digital Elevation Model (DEM) Grids (usgs.gov) grids Spatial Stream network and its physical Generated from Stream (m) vectors properties (e.g. Length, gradient) DEM Department of 15 stations Daily observed precipitation Time - Hydrology and (1992 - 2013) Precipitation (mm) Series Meteorology 5 stations Hourly observed precipitation (DHM), Nepal (2011 - 2013) http://www. Monthly evapotranspiration data fao.org/nr/ 3 Stations Evapotranspiration Time - downloaded from CLIMWAT 2.0 water/infores_ (Mean (mm per day) Series tools databases_ monthly ) climwat.html Daily Observed discharge and 1 Station River Discharge (m3 Time - instantaneous maximum Discharge (1992 - 2013) DHM per sec) Series 1 Station Hourly observed discharge (2011 - 2013) 3. Results and Discussion 3.1 Unit Hydrograph for the Study Area The calibrated Ct and Cp depend on the physical parameters of the watershed and are also defined as regional coefficients. After satisfying the required criteria, the value of Ct and Cp were calibrated for each sub-basin. The physical parameters extracted from HEC – GeoHMS for each sub-basins and calibrated Ct and Cp for each sub-basin are shown in Table 3. The calibrated value of Ct ranges from 0.244 to 1.016 and Cp ranges from 0.410 to 0.439. According to Subramanya (2017) and Acanal (2021) the coefficient Ct and Cp in the range of 0.3 to 6.0 and 0.31 to 0.93, respectively. In our research, the sub-basins having greater slope responds to smaller Ct value and in the case of small slope, the Ct value is larger than others. In the case of Hanumante river (sub-basin W1040), is the largest basin among all. But due to the longest flow path is largest for Manahara river (Sub – basin W710), the basin lag time is greater for Manahara river with comparison to Hanumante river. But due to largest basin, the peak discharge is greater for Hanumante river. The basin lag, time to peak, base time and peak discharge varies across the sub basins. This depends on the regional coefficients Ct and Cp and physical parameters of the watershed. The sub-basin having longest flow path and mild slope result in longer time to generate peak flow. UH for selected sub-basins representing
166 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] major tributaries are shown for sample only (see Fig. 4). All these graphs have been used to routing the direct runoff from each sub-basin. Table 3: Calibrated coefficient and summary parameters of Snyder Unit Hydrograph Name of L Lc Area Qpeak Tl Tp Tb Slope Ct Cp W50% W75% Sub-Basin km km km2 m3/s hr hr hr W710 29.04 16.73 73.7 0.04 0.4 0.42 41.47 1.9 2.56 11.51 3.98 2.27 W720 16.48 8.02 26.67 0.06 0.31 0.43 26.1 1.02 1.72 6.82 2.19 1.25 W900 18.23 10.49 44.28 0.06 0.3 0.43 41.01 1.09 1.79 7.19 2.33 1.33 W930 9.88 3.93 14.74 0.005 0.79 0.42 8.81 1.78 2.45 10.87 3.73 2.13 W1040 26.8 10.64 184.01 0.03 0.46 0.42 104.92 1.87 2.54 11.36 3.93 2.24 W1150 12.3 5.09 31.39 0.002 1.02 0.41 12.94 2.64 3.27 15.42 5.57 3.18 W1240 5.96 2.61 5.85 0.01 0.6 0.43 5.7 1.03 1.73 6.85 2.2 1.26 W1290 12.47 3.88 41.95 0.01 0.81 0.42 23.09 1.94 2.61 11.74 4.08 2.32 W1300 15.02 5.95 65.44 0.09 0.27 0.44 80.15 0.77 1.49 5.51 1.72 0.98 W1340 17.31 8.47 31.03 0.03 0.39 0.43 24.63 1.3 1.99 8.31 2.75 1.57 W1350 11.72 5.99 37.03 0.09 0.24 0.44 51.56 0.66 1.38 4.88 1.5 0.85 W1390 8.41 3.77 7.32 0.02 0.47 0.43 7.31 0.99 1.7 6.69 2.14 1.22 W1400 20.67 12.67 49.04 0.06 0.3 0.43 41.88 1.2 1.89 7.76 2.54 1.45
3.2 HEC – HMS Model Performance The model performance in calibration and validation was evaluated by visual inspection of the calculated and observed hydrographs and Nash – Sutcliffe Efficiency (Nash & Sutcliffe, 1970). The continuous data from 1992 to 2013 was used for calibration (see Fig. 5). Results indicate that the simulated hydrograph is comparable with the observed one, responds well to rainfall, and it can reproduce the overall hydrological pattern. The NashSutcliffe Efficiency in calibration is obtained as 0.753. Due care was given during calibration to ensure peak discharge and baseflows are reasonably reproduced. The model therefore can be considered as acceptable for further application. The developed UH and other calibrated parameters were fed with the model as input along with real-time data. The model was then applied to simulate the flood hydrograph at the outlet of Khokana. Available real- time hourly hydro-meteorological data for the storm event of July 2011, August 2012, and July 2013 were used for validation. From Fig. 6, it can be easily observed that the simulated discharge follows the similar trend of recorded discharge and also observed the Nash – Sutcliffe Efficiency was 0.90, 0.637, and 0.731 for 2011, 2012, and 2013 respectively which were in acceptable range as per Moriasi et al. (2007). Here, in the case of 2011, the rainfall was started from 9:00 AM July 30, and maximum precipitation recorded as 10.06 mm at 5:00 AM Jul 31. It seems 20 hours of regular rainfall with small fluctuations generate the peak flow of 461.6 m3/sec at 10:00 AM July 31. And the model also simulates the rainfall and gives the peak discharge of 480.4 m3/sec at the same time. However in the case of the event of 2012, the rainfall starts from 5:00 PM Aug 2, but the intense rainfall of 27.7 mm within 3 hours was recorded on 5:00 AM Aug 3. In this event, the peak discharge of 188.7 m3/sec was recorded at 3 Aug 11:00 AM and the model gives the peak value of 185 m3/sec at 7:00 AM 2 Aug. This is because of short duration instance rainfall. Also in the case of 2013 the rainfall starting from 5:00 PM and peaks after 12 hours up to 6.66 mm/hour. This precipitation gives rise to 175.7 m3/sec of flow at 9:00 AM but the model pretends the flow of 182.0 m3/sec at 8:00 AM.
167 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya
a)
Figure 4: Unit hydrograph for selected sub-basins. a) For sub-basin W1040 which represents the Hanumante river. b) For sub-basin W710 which represents the Manahara river. c) For sub-basin W1400 which represents the Nakhu river. d) For sub-basin W1350 which represents the upper part of Bagmati river.
Figure 5: Comparison of observed versus simulated daily stream flows for the continuous-time period at Khokana outlet (index (530.05)) from 1992 to 2013.
168 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021]
Figure 6: Validation of the model with the event of Jul 2011, Aug 2012, and Jul 2013 Model parameters deficit and constant, recession, simple canopy, and simple surface were calibrated. The calibrated model parameters for each sub-basin are presented in Table 4. For the model parameters, highly sensitive parameters were ratio to peak and recession constant they influenced for base flow.
169 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya
Table 4: Calibrated model parameters for HEC-HMS model of Bagmati River Basin
Deficit and Constant Recession Simple Canopy Simple Surface Ratio Parameters Constant Initial Maximum Initial Recession Initial Max Initial Max to Rate Deficit Deficit Discharge Constant Storage Storage Storage Storage Peak Sub basins mm/hr mm mm m3/sec % mm % mm W1150 0.346 30.00 150.00 4.090 0.364 0.233 5.188 9.287 5.00 9.170 W1040 0.235 50.34 248.64 4.599 0.650 0.348 5.000 10.294 5.00 5.387 W1240 0.707 50.00 350.00 4.743 0.347 0.367 5.126 10.250 5.00 12.874 W1290 0.691 49.00 250.00 5.850 0.453 0.378 4.999 5.210 5.00 13.233 W1300 0.624 40.00 300.00 4.820 0.348 0.529 5.651 5.699 5.00 8.591 W1340 0.653 50.00 250.00 4.080 0.567 0.237 7.387 5.875 5.00 5.943 W1350 0.356 45.00 185.00 2.809 0.233 0.344 4.107 10.107 5.00 12.310 W1390 0.797 40.00 220.00 5.078 0.656 0.237 4.651 5.634 5.00 10.118 W1400 0.394 56.00 250.00 4.800 0.632 0.235 4.360 19.888 5.00 15.770 W710 0.651 45.00 200.00 4.800 0.584 0.346 5.000 4.445 5.00 14.771 W720 0.679 50.00 350.00 5.080 0.348 0.454 4.900 7.950 5.00 12.090 W900 0.893 45.00 195.00 5.853 0.636 0.310 5.105 15.825 5.00 15.825 W930 0.663 60.00 210.00 5.825 0.651 0.364 3.049 9.521 5.00 9.800
3.3 Application of the Model for Estimating Extreme Events We have chosen the date of 23 July 2002 as the storm event to simulate extreme events because of the maximum flood recorded on that day from 1993 to 2013. The event value of rainfall, simulated discharge from the model, and estimated discharge using frequency analysis are present in Table 5. The maximum daily discharge of this flood event at the Khokana was 814 m3/s. The flood was generated by 5 – days of continuous rainfall starting from 19th through 23rd July 2002. The total cumulative rainfall amount for the 5 – day’s period is 276.4 mm, with a maximum of 175.6 mm on 23rd Jul 2002. The rainfall of earlier days would have generated favorable moisture conditions and then an increase in the base flow for that flood. The event model was then applied for estimating design floods as well as PMF at the outlet of the basin. Floods of different return periods were estimated using the deterministic approach, that is, by feeding the calibrated event model with rainfall of different return periods and compared with estimates from the probabilistic approach (i.e., flood frequency analysis) (see Table 5). It indicates the estimated design floods using the determinist approach are compared with those estimated using the probabilistic approach. As estimates from the probabilistic approach are slightly higher, they are recommended as design floods for the dry port. The floods of 50, 100, and 500 year return periods are estimated as 957.5 m3/s, 1,080.4 m3/s, and 1,364.3 m3/s, respectively. The estimated PMF is 3629.70 m3/s and the PMF hydrograph is shown (see Fig. 7).
4. Conclusions In this study, HEC – HMS model was used for input daily rainfall and compare with daily observed discharge from 1992 to 2013 in calibration at the gauging station of Khokana (530.05). And for validation, hourly rainfall
170 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] was compared with hourly observed real-time discharge for the event of 2011, 2012, and 2013. For the study, the spatially averaged precipitation in each sub-basin is derived using the Thiessen polygon method in ArcGIS. In this research work, the Bagmati Basin was divided into thirteen sub-basins, and unit hydrograph was developed for each sub-basin wherein the discharge of the outlet i.e. Khokana (550.05) was only taken. The UH for each sub-basin was developed by using Snyder equations. And the Snyder equations use the physical parameters of the basins. The dimension of UH was fixed so that the area of UH belongs to unity. After the UH development, the flood discharge at the basin outlet was then estimated by combining the sub-basins, using flood routing procedures. For calibration of the model parameters, Trial and error is performed to compare the precipitation and observed discharge. Based on this, by calibrating Snyder’s equations, the coefficients required for the development of synthetic unit hydrographs, Lag time Co-efficient (Ct), Peak discharge coefficient (Cp), unit hydrograph widths at 50% and 75% of the peak and base time were determined. The lag time coefficient (Ct) for the watersheds ranges from 0.244 – 1.016. The peak discharge coefficients (Cp) of the unit hydrographs of the watershed range from 0.439 – 0.410, these values are recommended to construct UH of the Bagmati basin. All the HEC HMS model parameters are calibrated with the entering of calibrated Snyder UH. The results obtained are satisfactory and acceptable. During the model parameter calibration, the ratio to peak and recession constant module for the contribution of the base flow was highly sensitive. The applicability of the model is also ensured by extreme rainfall of different return periods. The simulated discharge was compared with the calculated discharge of different return periods, which gives a reasonable response. And again the calculated PMP of the basin was processed in the model and determine the PMF for the basin. The one-day PMP for the basin is 612.91 mm using Hersfield’s method. The peak outflow at the outlet of the study area is 3629.70 m3/sec. which is 1.94 times higher than the storm of 10,000 year return period. Table 5: Simulated discharge and predicted discharge
Simulated Discharge from Difference in Return period Input Rainfall discharge from frequency analysis Discharge HEC – HMS using observed data Year mm m3/sec m3/sec 5 113.1 542.6 534.8 -1.4% 10 134.05 670.9 666.8 -0.6% 25 160.4 832.5 833.7 0.1% 50 180.0 938.3 957.5 2.0% 100 199.5 1072.3 1080.4 0.7% 500 244.4 1349.3 1364.3 1.1% 1000 263.7 1468.6 1486.4 1.2% 10000 327.9 1865.4 1891.7 1.4% PMP 612.9 3629.70 (PMF)
As specified in the flood control and management manual by WECS, bridges, cross-drainage structures, river training structures, and other hydraulic structures are designed for 100 years to return periods’ flood (WECS, 2019). However important projects like dry port may require flood greater than 100 years return period. Construction of dry port has been proposed in Chovar which lies little upstream of the outlet of our study basin and design of such project requires probable maximum flood. In this study, PMF has been estimated for the probable maximum precipitation for the basin which can be used in designing the river training structure upstream of Chovar and similar applications can be done for other river basins with
171 Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021] Devkota and Shakya similar projects. Furthermore, the satisfactory performance of the model proves its applicability in flood forecasting. The extreme rainfall events can be efficiently simulated to obtain the flood. Moreover, simulating the model for hourly rainfall can be used to fix the warning and danger level for the river. Hence this model can be applied in flood forecasting and early warning.
PMP and PMF 4000 0 3500 100 3000 200 2500 300 (mm) (m3/se c ) 2000 PMP PMF 400 g e r a in fall
1500 500 a R
Dis c h 1000 600 500 700 0 800 17-Jul-02 19-Jul-02 21-Jul-02 23-Jul-02 25-Jul-02 27-Jul-02 29-Jul-02 Date
Figure 7: Estimated Probable maximum flood at the basin outlet
Conflict of Interests Not declared by authors.
References Acanal, N. (2021). Snyder-gamma synthetic unit hydrograph. Arabian Journal of Geosciences, 14(4), 271. https://doi. org/10.1007/s12517-021-06531-7 Babel, M. S., Bhusal, S. P., Wahid, S. M., & Agarwal, A. (2014). Climate change and water resources in the Bagmati River Basin, Nepal. Theoretical and Applied Climatology. https://doi.org/10.1007/s00704-013-0910-4 Barry, D. A., & Bajracharya, K. (1995). On the Muskingum-Cunge flood routing method. Environment International. https:// doi.org/10.1016/0160-4120(95)00046-N Beven, K. (2012). Rainfall-Runoff Modelling. In Rainfall-Runoff Modelling. https://doi.org/10.1002/9781119951001 Bhunya, P. K. (2011). Synthetic Unit Hydrograph Methods: A Critical Review. The Open Hydrology Journal. https://doi. org/10.2174/1874378101105010001 Chow, V. T. (1988). Applied Hydrology. McGraw-Hill Series in water resources and environmental engineering. Cibin, R., Athira, P., Sudheer, K. P., & Chaubey, I. (2014). Application of distributed hydrological models for predictions in ungauged basins: A method to quantify predictive uncertainty. Hydrological Processes. https://doi.org/10.1002/ hyp.9721 Efstratiadis, A., Koussis, A. D., Koutsoyiannis, D., & Mamassis, N. (2014). Flood design recipes vs. reality: Can predictions for ungauged basins be trusted? Natural Hazards and Earth System Sciences. https://doi.org/10.5194/nhess-14-1417-2014 Fedorova, D., Kovár, P., Gregar, J., Jelínková, A., & Novotná, J. (2018). The use of snyder synthetic hydrograph for simulation of overland flow in small ungauged and gauged catchments. Soil and Water Research. https://doi.org/10.17221/237/2017-SWR Guide, U., & Manual, T. R. (2008). Technical Reference Manual. Texas Instruments. Halwatura, D., & Najim, M. M. M. (2013). Application of the HEC-HMS model for runoff simulation in a tropical catchment.
172 Devkota and Shakya Journal of Engineering Issues and Solutions 1 (1): 158-173 [2021]
Environmental Modelling and Software. https://doi.org/10.1016/j.envsoft.2013.03.006 Hershfield, D. M. (1961). Rainfall Frequency Atlas of the United States, for Durations from 30 Minutes to 24 Hours and Return Periods from 1 to 100 Years. Technical Paper No. 40. HMS, H. (2000). Hydrologic Modeling System HEC-HMS Technical Reference Manual. US Army Corps of Engineers. https:// doi.org/CDP-74B Hoffmeister, G., & Weisman, R. N. (1977). Accuracy of synthetic hydrographs derived from representative basins la précision des hydrogrammes synthétiques dérivés des bassins représentatifs. Hydrological Sciences Bulletin. https://doi. org/10.1080/02626667709491719 Hrachowitz, M., Savenije, H. H. G., Blöschl, G., McDonnell, J. J., Sivapalan, M., Pomeroy, J. W., Arheimer, B., Blume, T., Clark, M. P., Ehret, U., Fenicia, F., Freer, J. E., Gelfan, A., Gupta, H. V., Hughes, D. A., Hut, R. W., Montanari, A., Pande, S., Tetzlaff, D., … Cudennec, C. (2013). A decade of Predictions in Ungauged Basins (PUB)-a review. In Hydrological Sciences Journal. https://doi.org/10.1080/02626667.2013.803183 Jain, M. K., Kothyari, U. C., & Ranga Raju, K. G. (2004). A GIS based distributed rainfall-runoff model. Journal of Hydrology. https://doi.org/10.1016/j.jhydrol.2004.04.024 Khadka, J., & Bhaukajee, J. (2018). Rainfall-Runoff Simulation and Modelling Using HEC-HMS and HEC-RAS Models : Case Studies from Nepal and Sweden. Lund University. Kozanis, S., Christoforides, A., & Efstratiadis, A. (2010). Scientific Documentation of Hydrognomon Software (Version 4). Development of Database and Software Application in a Web Platform for the National Database and Meterological Information. ITIA Research Team, National Technical University of Athens Available from: Http://? Www.? Itia.? Ntua.? Gr/? Getfile. Li, M., & Shao, Q. (2010). An improved statistical approach to merge satellite rainfall estimates and raingauge data. Journal of Hydrology. https://doi.org/10.1016/j.jhydrol.2010.01.023 Linsley, R. K., Kohler, M. A., & Paulhus, J. L. H. (1982). Hydrology for engineers. Third edition. Majidi, a, & Shahedi, K. (2012). Simulation of Rainfall-Runoff Process Using Green-Ampt Method and HEC-HMS Model (Case Study: Abnama Watershed, Iran). International Journal of Hydraulic …. https://doi.org/10.5923/j.ijhe.20120101.02 Moriasi, D. N., Arnold, J. G., Liew, M. W. Van, Bingner, R. L., Harmel, R. D., & Veith, T. L. (2007). M e g s q a w s. 50(3), 885–900. Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual models part I - A discussion of principles. Journal of Hydrology. https://doi.org/10.1016/0022-1694(70)90255-6 Nelder, J. A., & Mead, R. (1965). A Simplex Method for Function Minimization. The Computer Journal. https://doi.org/10.1093/ comjnl/7.4.308 Oleyiblo, J. O., & Li, Z. J. (2010). Application of HEC-HMS for flood forecasting in Misai and Wan’an catchments in China. Water Science and Engineering. https://doi.org/10.3882/j.issn.1674-2370.2010.01.002 Prajapati, R. N. (2015). Delineation of Run of River Hydropower Potential of Karnali Basin- Nepal Using GIS and HEC- HMS. European Journal of Advances in Engineering and Technology. Saghafian, B. (2006). Nonlinear transformation of unit hydrograph. Journal of Hydrology. https://doi.org/10.1016/j. jhydrol.2006.04.026 Sherman, L. K. (1932). Streamflow from rainfall by the unit graph method. Engineering News Record, 108. Subramanya, K. (2017). About the Author. In Reservoir Engineering in Modern Oilfields (3rd ed.). Tata McGray - Hill Publishing Company, New Delhi. https://doi.org/10.1002/9781119284628.about Sudhakar, B., Anupam, K., & Akshay, O. (2015). Snyder Unit Hydrograph and GIS for Estimation of Flood for Un-Gauged Snyder Unit Hydrograph and GIS for Estimation of Flood for Un-Gauged. December. https://doi.org/10.4172/2157-7587.1000195 USACE. (2005). Technical standard for water-table monitoring of potential wetland sites. Wetlands Regulatory Assistance Program. Vizina, Š., & Hradilek, V. (2012). Comparison of semi-distributed and distributed approaches of the hydrological rainfall-runoff modelling and determination of morphological and morphometric characteristics of watersheds. 12th International Multidisciplinary Scientific GeoConference and EXPO - Modern Management of Mine Producing, Geology and Environmental Protection, SGEM 2012. WECS, G. (2019). Flood Control and Management Manual. Government of Nepal. World Meteorological Organization. (2009). 22. Manual for Estimation of Probable Maximum Precipitation, Operational Hydrology Report 1, 2nd edition, Publication 332 (Issue 1045).
173 Journal of Engineering Issues and Solutions
A study on FTTH implementation and migration in Nepal
Naba Raj Khatiwoda1, Babu R Dawadi2,* 1Wireline and Customer Service Directorate, Operation and Maintenance Department, Nepal Telecom 2Department of Electronics and Computer Engineering, Institute of Engineering, Tribhuvan University, Nepal *Corresponding email: [email protected] Received: January 10, 2021; Revised: March 17, 2021; Accepted: April 02, 2021
Abstract The increasing demand of high speed data results into extensive enhancement on different telecommunication technologies through wireline and wireless technologies. Optical Fiber technology is being popular for fixed broadband technologies and for backhaul network data for network convergence and media device interaction. Fiber to the home (FTTH) is gaining momentum of deployments in many countries all around the world. Passive optical network (PON) utilizes point to multipoint (P2MP) topology and is becoming suitable, cost effective, and promising solutions as compared to existing copper based telecommunication infrastructure. PON architecture is cheaper than other architectures due to dynamic bandwidth allocation and common resources that can be used by different subscribers and especially for home subscribers. This paper presents a study on the effective deployment of PON based FTTH network at Nepal by referring the deployment scenario of Nepal Telecom (NT), while this network design, deployment, and implementation provides a lesson learn for cost effective deployment of such network to other stakeholders of developing countries having similar territory and implementation challenges.
Keywords: Broadband; FTTH, network deployment; PON; telecommunication.
1. Introduction Information and communication technology (ICT) enables us to exchange information instantly from any part of the world. However, there are several challenges for a developing country like Nepal to have robust and high capacity telecommunication services for its people due to poor telecommunication infrastructure, where less than three percentage (2.43%) people of total population have access of landline infrastructure (NTA, 2021). There is limited availability of broadband services in most of the areas of the country. Mobile communication is available to almost all the citizens, but only the voice communication service is available. Mobile broadband (3G, LTE/4G) is being expanded rapidly in the country, but it is not enough for all citizens and for all parts of the country.
174 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
In Nepal, there is well planned fixed telephone network in limited area only and most parts of the country have poor telecommunication infrastructure. But the government owned organization such as Nepal Telecom (NT) and other private internet service providers (ISPs) are working massively to expand FTTH service and building FTTH infrastructure. About 6,00,600 subscribers are using copper based telephone lines and still time division multiplexing (TDM) switches are in operation (NTA, 2021). NT is the service provider for landline phone in Nepal. It is working aggressively to replace the TDM or legacy switches and copper based networks by the next generation network (NGN) IP switches and optical fiber network basically FTTH network. It is a great challenge to expand telecommunication facilities in all parts of the country, because there is difficult and adverse territorial structure such as mountains, hills, lack of physical infrastructure such as roads, electricity, and lack of proper planned residential structure (Dawadi et al., 2019). Fixed landline is limited to urban and some parts of the sub-urban area and district headquarters with limited places in rural areas. Digital divide is a prevalent in the country. In this paper, design and implementation of FTTH technology is discussed and cost effective implementation of FTTH and migration of FTTH from copper based network is analyzed and different models for different territories are discussed and recommended. This design and implementation method to deploy FTTH network would provide a lesson learn for the cost effective deployment of such network in the developing countries. The FTTH network architecture can be modified in different capacities with slight modification of Fiber distribution cabinet (FDC) and Fiber access point (FAP), and can be used at different sites depending on the population distributions. The rest of the paper is organized as follows. In section 2, the background of FTTH technology and its benefits with related work on network deployment are discussed. Section 3 presents methodology of PON based FTTH deployment considering network deployment of NT. Section 4 presents the results and analysis of implementation including capital and operational expenditure (CAPEX and OPEX) calculation of network deployment. Section 5 provides discussion with possible recommendations, while section 6 concludes the paper.
2. Background and Related Work FTTH is an advanced technology transmitting signal in the form of light through the Fiber cable medium to provide an ultra-high speed, reliable communication, high quality voice, data, and video transmission. The transmission of voice, data, and video through the same medium and at the same time is called triple play. Besides having higher investment in the initial stage, optical Fibers are effective and suitable substitutes of existing copper cable based telephony and cable TV networks. Optical Fiber is replacing copper cables due to its various advantages. It provides higher bandwidth (Tera bit per second - Tbps vs. 10Gbps in copper wires) and longer distance transmission (around 50 km vs. up to 5 km in copper wires). It is lighter in weight (weighs around 10 times less than copper cables) and has longer lifetime and much less attenuation (Khatimi et al., 2019). Optical Fiber is used for transmitting a volume of data at a higher speed with lower cost. Optical Fibers are costly at the initial stage of deployment, but their durability minimizes the overall cost in a long run. Optical Fibers are recently widely used in various types of networks such as backhaul network, computer network e.g. wide area networks(WANs), metropolitan area networks (MANs) as well as local area networks(LANs) with Ethernet-optical interface standards and in access network (‘the last mile and ‘will complete all-optical-network-evolution’) as well. Optical Fibers are also used in data transmission in various fields like surgery, automobile industry, space, military applications, decorations, and lighting (Babani et al., 2014). The copper based telecommunication network is not capable to cope up with the huge demand of data
175 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi transmission required for the reliable data service. The evolution of advanced telecommunication technologies and the rapid increase in data usage in different fields resulted into a huge demand of data speed. Online classes, e-learning, telemedicine, online video conference, online business, online payment, online shopping, and live streaming need reliable and high speed Internet service. FTTH is the only complete solution for such type of applications. The combination of Fiber and copper network as a hybrid technology is also possible. NT is also using these hybrid combinations using IP switches and copper based access network. But this hybrid combination is also not suitable due to limitations and problems of copper cable as compared to optical Fiber networks. In FTTH, the Fiber is directly connected from an exchange or switching office to the subscriber’s home. Generally, the exchange offices are called central offices (CO) and the Fibers from a CO are connected to different subscriber’s premises. The connection mechanisms and approaches are different based on the FTTH network design. FTTH network design is carried out differently based on the types of subscriber’s premises, i.e. the size of building, number of residents in the building, number of building in a community, etc. Thus, general term of FTTx is frequently used, where the term ‘x’ refers to a node, curb, home, premise, or business (Sahu et al., 2019). Fig. 1 shows the general architecture of FTTH. The major components of the architecture are briefly summarized here. Optical line terminal (OLT): OLT is located in CO and it is the end point of internet service providers. It controls the bidirectional data flow from subscribers’ end to backbone transmission network and vice versa. OLT takes voice, data, and video from backbone transmission network and broadcast to all subscribers, while there is flow of information in the downstream direction. When upstream information flow takes place, OLT accepts the traffic from subscribers and forwards to the backbone transmission network.
Figure 1: FTTH general architecture
176 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
Fiber to the business (FTTB): FTTB is designed for enterprises subscribers or business houses. The Fiber is directly connected from OLT to the active devices located inside the enterprises and they have access dedicated resources such as bandwidth. Fiber to the node (FTTN): An active switching equipment i.e. node such as digital subscriber line access multiplexer (DSLAM) is placed near the subscribers’ premises. Copper access network can also be used. FTTN is suitable at a distance of about 1.5 Km. Fiber to the premises (FTTP): FTTP is designed for larger number of subscribers in a single building or premises. Multi dwelling unit (MDU) are placed at the premises and the network is connected to individual subscribers through the MDU. The premise can be a single family house or group of single family house within a large building (Rigby, 2014). Cloud: This is the network structure that connects all the network elements through Internet protocol (IP) and multi-protocol level switching (MPLS). It connects the server related to voice, data and Video. DSLAM: DSLAM is a switching equipment that aggregates traffic from different digital subscriber line (DSL) and aggregates the traffic using multiplexing techniques and sends to backbone transmission network. De-multiplexing is carried out for the downstream traffic. Splitter Cabinet: it is a cabinet containing optical splitter inside it. Optical splitters are passive devices which divides the optical signal into numerous optical signals. The split ratios are 1:2, 1:4, 1:8, and 1:16 and so on. The splitters make the different subscribers use the same optical resource from OLT to splitter (Lokhande & Singh, 2017).
2.1 FTTH Network Architecture Optical Fiber network can be deployed in two ways: Active Optical Network (AON) and Passive Optical Network (PON).
2.1.1 Active optical network AON architecture as the name suggests uses the active or powered equipment such as routers or switches. It uses electrical switching equipment for data routing. It has dedicated optical Fiber terminated to the subscriber’s premises and the subscribers get the dedicated Fibers and bandwidth. The AON network structure is point-to-point (PTP) and can provide data transmission at a distance of around 100 Km (Abdel et al., 2018). This network is costly and has frequent power issues due to active equipment, but generally considers high quality of service (QOS) network due to dedicated network and suitable for corporate users. It is also considered as a high security network. Fig. 2 shows the network architecture for AON system. The different components of the AON architecture are: Ethernet switches: These are the active devices such as router or switches that have capable of switching and forwarding the incoming and outgoing traffic to the desired destination. Optical Network Terminal (ONT): ONT is an active device used at or inside the subscriber’s premises. ONT de-multiplexes the incoming signal (downstream) into voice, data, and video, then sends to appropriate destination i.e. voice to telephone set, data to computer or mobile, and video to TV set. Similarly ONT aggregates upstream traffic from triple play devices and forwards to CO (Mata, 2014).
177 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi
Figure 2: AON architecture
2.1.2 Passive optical network PON uses passive equipment or non-powered equipment from CO to outside of customers’ premises. The passive devices are basically PON splitters. PON architecture has point to multipoint (P2MP) structure and the Fibers and bandwidth are shared by the different subscribers. It does not use active components such as amplifiers, repeaters, and shaping circuits. Due to the high bandwidth of optical Fiber, the shared resources are not of much interest but it makes sensitive for quality service when there is high traffic and much more active subscribers at a time. A single Fiber is connected from central office to splitter and the splitter splits the Fiber and connected to different subscribers, thus sharing the Fiber from CO to a splitter by the connected subscribers. Fig. 3 shows the PON based system architecture. In the central office, optical line terminator (OLT) is located and optical Fiber is connected from OLT to an optical splitter through optical distribution frame (ODF). As discussed earlier, the OLT generally performs bandwidth allocation and data routing. It manages bidirectional traffic from OLT to customer’s premises and vice versa. Optical splitters are the passive splitters that split an optical signal equally into several low power signals. One of them is selected on the basis of the communication service requirements. Splitters are capable of multiplexing and de-multiplexing the optical signals from and to the connected ONTs. Splitters are capable of transmitting the traffic in both direction i.e. both upstream and downstream. An optical splitter or PON splitter is used inside the subscriber’s premise or it can be also used outside of the premises. PON splitter is also used inside FDC. PON splitter can be used in different stages. If they are used in first stage, they are generally called as L1 or S1 splitter and in second stage, as L2 or S2 splitters. L2 or S2 splitters are used inside a FAP. Basically FDC and FAP are passive optical splitters. Splitters are connected between ONT and OLT and there may be more than one stage of splitter connections. Basically, there is optical network unit (ONU) connected at customer’s premises and it communicates other ONTs, but ONU and ONTS are considered the same device. The ONT has the interfaces to connect triple play devices such as RJ-11 interfaces for telephone set, LAN interfaces for data and video. The PON network is suitable for the distance at around 20km for quality data transmission and it uses wavelength range of 1310 nm for upstream at speed of 1.25Gbps, 1490nm at 2.5Gbps for downstream, and 1550nm for video transmission (Horvath et al., 2020).
178 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
Figure 3: PON architecture
2.1.3 AON vs. PON As every system has both benefits and shortcomings, AON and PON systems have the same. On the basis of some key parameters, for example, signal distribution, equipment, cost and coverage distance, here is comparison between these two systems. Table 1 shows the difference between these two systems (Larsen et al., 2010; Mahloo et al., 2015). Table 1: Comparison between AON and PON Parameters AON PON Coverage distance 100Km 20Km Bandwidth Dedicated, same as bandwidth of OLT port Sharing Signal distribution Unique signal Same signal to all Cost Expensive due to dedicated resources Cheaper System failure and Complex due to active components if there is Simpler due to passive devices maintenance frequent power failure Complex as there are shared Network modification Easier due to dedicated resources resources
In AON networks, subscribers can use dedicated Fiber optic and bandwidth, while in PON networks, there is sharing of optical Fiber and bandwidth by using PON splitter. Optical Fibers from OLT are shared using L1 splitter and further can be shared by using L2 splitter before reaching to customers premises as well. Due to the dedicated resources, tracing of problems would be easier in AON network, while it is somewhat difficult in PON network. Since, there are powered or active equipment in AON network, there may be more issues of power if somewhere fails the power. In PON network, no such issues take place due to the passive Fiber optic except in two end terminals. AON devices are costlier than PON and AON devices that need frequent
179 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi maintenance than PON devices due to the necessity of power. In AON system, network can be extended up to 90 to 100km from OLT i.e. central office, while in PON system it is up to 20km only. If there are subscribers closer to the central source of the data, the deployment is easier (Larsen et al., 2010). Apart from these reasons, AON networks are industry standards, more reliable than PON networks, easier to add new devices, and are suitable for corporate customers. Due to the shared bandwidth and resources causing slow in busy hours but are cheaper than AON network and suitable for household customers.
2.1.4 PON standardization PON network is basically designed for shared usage, due to this, it is cheaper. It uses multiplexing/de- multiplexing techniques to transmit many user’s data through the same Fiber. The multiplexing mechanism is based on time division multiple access and time division multiplexing. The Institute of electrical and electronics engineers (IEEE) and the study groups of ITU’s telecommunication standardization sector (ITU-T) have developed the different series of PON standards. The full services access network (FSAN) group has developed technical specifications that have been the basis for ITU-T standards. Table 2 summarizes the different PON standards (ITU-T Q2 / Study Group 15, 2018; Muciaccia et al., 2014). Table 2: PON standards
Technology (Standards) Features It is the first complete standard of PON, enhanced version of APON based BPON(G.983.1/G.983.5) on the ATM protocol with dynamic bandwidth allocation and protection functions. BPON (G.983.1/G.983.5) is standardized in 2005 and has bit rate of 155/622 Mbps (upstream/downstream). Adopts P2MP structure and PON transmission. It is the most effective communication method to realize the “three networks in one” and “last EPON (IEEE 802.3ah) mile” due to its cost effective deployment. It has symmetric bandwidth of 1.25 Gbps. Provides for 10/1Gbps (downstream/upstream) bitrates is the most 10G-EPON (IEEE 802.3av) common Standards. GPON supports multiuser through PON splitter and has high bandwidth, GPON (ITU-T G.984) but it is more expensive and complex. It has the bit rate of 1.25/2.5 Gbps. (upstream/downstream) XG-PON standard introduced in 2010 provides bit rate of 2.5-10/10Gbps) XG-PON (ITU-T G.987) with highest number of end-users (64-128 Gbps (upstream/downstream) and longer reach (20-60KM). Introduced in 2013, provides bit rate of greater than 10/40Gbps NG-PON2 (ITU-T G.989.1) (upstream/downstream) with highest number of end-users (>128) and longer reach (40-60KM). High speed PON (G.Hsp.x) Provides high speed of 50Gbps. It is under development.
2.1.5 Telecommunication service scenario Total population of Nepal is 29,876,531 (NTA, 2021), the telecommunication penetration in Nepal is 130.34%, among them 127.9% are of mobile users. It shows that the number of mobile users is more than the total population of the country. It is due to the fact that more than one mobile phone lines are subscribed by a person but not all citizens are using the mobile. There is no or limited network coverage at far rural part of
180 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] the country. As discussed earlier, the share of basic telephony is about 2.43%. In most part of the country, the fixed telephone or landline network availability is limited (NTA, 2021). Fig. 4 shows the broadband share in Nepal for different technologies. Broadband penetration is mainly from mobile network, it is around 74.09%, remaining are from wired and wireless fixed broadband.
Figure 4: Broadband share in Nepal (source: NTA) Due to the lack of fixed telephony network infrastructure, the fixed broadband network in Nepal is limited to only about 22.46% of the Nepal’s total Population. Fixed broadband technologies include ADSL, FTTH, and lease lines (including copper, optical, and connected by microwave links). Though only 22.46% of Nepal’s total Population are having access of fixed broadband, FTTH has the larger share of 86.64 % and it is growing day by day (NTA, 2021). It is due to the optical Fiber deployment by NT and other private ISPs. Fig. 5 shows the share of fixed broadband services in Nepal.
Figure 5: Fixed broadband in Nepal (source: NTA) NT is migrating copper based ADSL system to optical Fiber broadband technology. Project work for FTTH infrastructure is under deployment. Still, NT has dominant number of ADSL subscribers and they are
181 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi getting lower due to migration process. Fig. 6 shows the proportion of ADSL, migrated FTTH subscribers from copper cable network, and new FTTH subscribers in NT. The figure is based on the data of NT’s ADSL, FTTH subscribers as of august 2020.
Figure 6: Fixed broadband users in NT (source: NT) Since, only NT is providing the fixed landline service in Nepal, it has been planning to replace the legacy system with the FTTH service. Fig. 7 shows that 43.47% of the existing total subscribers are having TDM legacy switch and NT is working to replace the legacy switch with NGN IP/IMS switches very soon.
Figure 7: TDM switch vs. IP-based switch (source: NT)
2.2 Related Works Some related works for the implementation of PON based FTTH architecture are very fruitful and carried out in Nepal and other countries. For reliable optical Fiber high speed data communication, there should be robust telecommunication infrastructure available for the service. Because the existing legacy system cannot support high bandwidth data transmission. There is a pressure for network service providers to upgrade the network infrastructure to meet their demands. The high speed data transmission capability of optical Fiber is instrumental for the world to realize the concept of smart city, smart payment, smart transportation, Internet of things, and smart of everything. A cost effective design and implementation of GPON based network is presented for Baghdad/Al-Gehad city. Wavelength division multiplexing (WDM) based FTTH solution results in minimization of CAPEX and allows for flexibility and adaptability (Kadhim & Hussain, 2013a). A GPON based architecture is proposed by the authors for Kosovo (Caka & Hulaj, 2011). The concrete possibilities for practical realization of FTTH network are analyzed based on the territorial structure of Kosovo. A model network for ten houses were used for analysis.
182 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
Innovative ideas are developed in Europe for infrastructure sharing. In France, rules for duct sharing are defined and agreed in 2008 and different operators shared the common duct infrastructure of France telecom. Tools for installing optical Fiber in domestic water pipe is developed and used in Germany (FTTH Council Europe, 2014). FTTH can be a milestone for efficient and fast health care system in rural areas. Hypothetical structuring equation models have been proposed and the test has been carried out in India (Bag et al., 2019). It shows the implementation of FTTH systems results in enhancing the efficiencies of emergency health care system. Authors of (Khabzli et al., 2019) proposed a PON-based FTTH network architecture for Pekanbaru Citraland housing in Indonesia. The simulated results show that the link budget value has met the IEEE 802.3 standard. In this paper, we present a design and implementation process for PON based FTTH for Nepal considering network deployment of NT. In Nepal, the project for optical Fiber project for backhaul data transmission is expanding all over the country. NT and other private telecommunication service providers are building Fiber infrastructure in the Terai region and expanding towards hilly region rapidly. Fig. 8 shows the copper network and optical FTTH network reachability status by NT. Copper network is expanded to all district headquarters but there is no uniform expansion to other places. NTA is also building optical Fiber transmission network expansion utilizing the rural telecommunications development fund (RTDF) (NTA, 2018).
Figure 8: Copper and optical network reachability by NT (map source: ncthakur.itgo.com) There are challenges to deploy optical Fiber, aerial as well as underground (UG) and all-dielectric self- supporting (ADSS). ADSS is the aerial optical Fiber used for transmitting data that are used on electric utility power lines or high voltage transmission lines (Efficiency, 2015). Tensile element is provided with nonmetallic reinforcement and it does not need messenger wire. There should be a close coordination among various types of physical infrastructure providing agencies such as road department, drinking water supply department, and electrical authority. The coordination process is happening frequently, but lack of proper communication results in breakdown of optical Fiber, while construction and maintenance work of road, electricity, and drinking water. So, complete synchronization is yet to be realized and there should be proper planning to expand the Fiber through bridges, overhead bridges, and poles.
183 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi
3. Methodology A study of design and implementation of PON based FTTH for 500 subscribers capacity is carried out in NT. Its capital expenditure (CAPEX) as well as operational expenditure (OPEX) is estimated and possible payback period is also calculated. Let us discuss on first design aspect of PON FTTH network. Data collections and analysis have been carried out from the field survey. The location and capacity of OLT, ODF, the length and types of feeder cables, the FDC capacity, the length and types of aerial or distribution cable and types of FAPs are calculated based on precisely collected survey data. The location of FDC and FAPs are fixed accurately with proper survey data. The residential density that may include the number of building, capacity of building, the number flats etc. the possibility of expansion of residential area also considered precisely. The structure of road, highways, drinking water, and pole structure of electricity, where it is to be shared are analyzed and coordinated properly. The proper route of feeder, aerial cable according to the residential structure are identified. The geographic information system (GIS) application becomes the effective tools to solve the problems with field survey data. The network ring also designed correctly for redundancy purposes in case of main link fails. Generally, OLT and ODF sites are in CO. Fig. 9 shows the summary of design steps of FTTH network using bottom up approach i.e. from subscriber residents to CO.
Figure 9: Design steps of PON based FTTH network
184 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
Fig. 10 shows FSA, CSA and DSA service are boundaries in design aspect (Lokhande & Singh, 2017).
Figure 10: Service area boundaries Feeder service area (FSA) design: FSA is planned in such a way that it covers the maximum home/customer to meet the key performance indicator (KPI) and design from OLT to first stage designer. Customer service area (CSA) design: Planning the CSA/S2 splitter cluster requires building information, open plots, under construction building information, and land base data, which contains building, roads, streets parks, playground etc. Design from 2nd level splitter onwards covers 10-40 subscribers. Distribution service area (DSA) design: Fiber route planned between L1 splitter to L2 splitter is called as distribution route. Distribution sections connect neighboring buildings from L1 splitter locations in its designated cluster. Feeder and distribution Fiber: Feeder Fiber is used to connect ODF and FDC. After completing the site for OLT, feeder laying design is carried out and possible locations of FDC and FAP are designed accurately from field survey data by using GIS tools. Generally, 12/24/48/96/144 core Fiber cable is used. It is underground (UG) Fiber cable and designed accordingly. Distribution Fibers are aerial Fiber of 12/24/48 core that are used to connect FDC and FAP depending upon choice of different their capacities. Drop Fiber (DF) and ONT: Drop Fiber cable is used between FAPs and ONT. It can be one- or two-core Fiber. The length of DF is calculated according to the distance between FAP and subscriber’s home. The FAP is placed in a location such that the total length of DF can be minimized. ONTs are installed inside the home premises of the users. DF with built-in connector is used and connector can be made according to the choice of the required length of the DF. Generally, the length of DF is up to 50m and if there is longer distance i.e. greater than fifty meters, outdoor terminal box can also be used for signal repetition and generation purposes. Fig. 11 shows the FTTH architecture for 512 capacity FDC of ODN Network. In this network, L1 splitter or FDC has 1:8 split ratio resulting 64 Fiber access points (FAPs) from 8 cores of Fiber. Each FAP or L2 splitter has 1:8 split ratio further thus resulting 512 connection port for subscribers. For copper based telephone migration, exchanges up to the capacity of 2000 to 5000 subscribers and such type of FDC is suitable for cities such as Itahari, Mahendranagar, and Surkhet. The network architecture for ODN shown in Fig. 11 can be modified and applied for FDCs with the capacities of 256 and 1024 subscribers. The major differences are the number of feeder cable, split ratio of L1 and L2 splitter. Table 3 shows the comparisons
185 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi of the three types of ODN architecture. Fig. 12 summarizes the implementation and deployment steps of PON based FTTH network.
Figure 11: ODN architecture for FDC capacity of 512 Civil works: Trenching, digging and basement of FDC, laying of duct for Fibers are carried out. In NT, civil department coordinates the design and implementation of civil related works. The major challenges for this type of work is to establish coordination with the related other offices such as road department, urban development department, electricity authority, and other related departments (Nyarko‐Boateng et al., 2020). Table 3: Comparison of different ODN architecture
Parameters FDC-512 FDC-256 FDC-1024 Number of OLT Ports 8 4 16 Number of L1 Splitter 8 4 16 L1 Split ratio 1:8 1:16 1:4 Number of L2 Splitter 64 64 64 L2 Split ratio 1:8 1:4 1:16
4. Implementation, Results, and Analysis CAPEX is related to the cost associated with the infrastructure such as physical assets, building, network equipment, and software (Schneir & Xiong, 2014). OPEX refers to the cost associated with day to day activities or business for running and maintaining the system continuously and preventing from failures. In this paper, we consider cost estimation for ODN basically from OLT to ONT and analyze for best optimization combination of CAPEX and OPEX. CAPEX includes mainly three types of costs, these are infrastructure setup cost, system network equipment cost, and customer’s premises equipment cost. The infrastructure setup cost includes the cost of ODF, Fiber (e.g., feeder, aerial, splicing cost, and drop Fiber), POS splitter (e.g., FDC and FAPs), in-house infrastructure in the subscribers premises, and related other components. The cost also includes the civil work cost like trenching, digging, and ducting for underground Fiber laying. System network equipment cost refers to the cost associated with network equipment for routing and switching. Switches, routers, OLT, and related cards and accessories, racks, power equipment such as UPS,
186 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] air conditioning, and rectifiers. Customer premises equipment (CPE) cost includes the cost related to ONT (e.g., CPEs, telephone interfaces, LAN cables and interfaces, radio interface, and power backup for CPEs. (Rokkas et al., 2012; Wang, 2017).
Figure 12: Implementation and deployment steps of PON based FTTH network. OPEX is the cost incurred by business in its day to day business for smooth operation and maintenance of the FTTH system. OPEX includes wages and salaries of employees, travelling expenses, preventive and corrective maintenance cost. The major OPEX for FTTH operation and maintenance is the service provisioning cost. This is about provisioning of subscribers’ services such as addition, removal and editing service profile of the subscribers. Service profile includes details of subscribers such as name, address, documents, types of service package etc. Service provisioning is carried out by the network management system (NMS). It includes Fiber management (such as splicing and patch cord management), and fault management such as failure detection and recovery through system, card and other components replacement and repair. Energy consumption cost covers the power consumption by active equipment and air-conditions. Maintenance cost includes the maintenance of physical infrastructure, preventive and corrective maintenance cost. Floor space
187 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi management cost includes mainly the cost associated with space leasing for setting up equipment (Kadhim & Hussain, 2013b; Kulkarni et al., 2008).
4.1 Network Life Cycle of a Typical FTTH The cost breakdown can be well depicted by examining the following project life cycle of typical projects, which will be fruitful for cost breakdown. A typical network lifecycle consists of the following five stages:
• Planning phase • Implementation phase • Migration phase • System power up and running phase • Dismantle phase Planning phase: It includes analysis of demographic as well as geographical factors which affects costs for deployment stage. Survey of entire network, calculation of number and location of network equipment, estimation of CAPEX and OPEX, developing business model, choice of proper technology, and choice of proper work order are to be finalized in planning phase. Implementation phase: In this phase, the project is deployed. For FTTH, all civil related works e.g. trenching, digging, and lying of ducting can be carried out. The installation of network equipment, laying down of Fibers, installation and testing of PON splitter, and testing of all equipment should be carried out in this phase. Larger amount of cost and resources are needed in this phase. Migration phase: After the deployment phase is over, migration phase should be carried out. Database collection of all subscribers, their service status are collected, rearranged and updated according to the new service. If there is a mismatch, then proper migration cannot be performed. The cost is related to activity based cost. It includes basically customer’s equipment cost such as cost of CPEs and accessories and administrative and CPE installation cost while migrating. System power up and running phase: This phase includes mainly operation and maintenance. Proper preventive and corrective maintenance procedure are analyzed and carried out. The main cost includes OPEX cost like equipment maintenance, splicing and patching (for operations and maintenance), parts repaired and replacement costs. Marketing and promotions, and continuous fixed costs are also included in this phase. Dismantle phase: In this Phase, all the existing subscribers are migrated into new service and all the old network equipment will be dismantled (Casier et al., 2008).
4.2 CAPEX and OPEX calculation for ODN The CAPEX calculation presented in this paper is only for an ODN network of an urban area having 500 subscribers. The other costs like backbone transmission and core network are not included. The CAPEX includes OLT and core equipment, UG feeder Fiber, aerial distribution Fiber, FDC, FAP, drop Fiber, customer premises equipment (CPE), implementation cost and civil work cost. By calculating all the costs, the CAPEX per subscriber is come out to be NPR 9937.17(≈9938). This is calculated for specific time period and it may vary with time due to the changes in the prices of Fibers and labors and other equipment. The calculation is based on ODN architecture for 512 subscribers. The estimated values are assumed based on the current
188 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] market values and considering previous different quotations and tender. Table 4 shows the values of each parameters. The calculation is carried out using the following formula.
Cost per subscriber for OLT and core equipment (1024 Capacity) Total cost for OLT and core equipment (1024-capacity) 650000.00 = = = 634.77 Number of subscriber 1024
The total quantity is specified for UG feeder Fiber, aerial distribution Fiber, FDC, and FAP. Implementation cost is just enough for 512 subscribers.
Total cost for UG feeder fiber 300000.00 Cost/subscriber for UG Fiber = = 585.94 Number of subscriber = 512 Each subscriber need one CPU, so the unit cost is kept same. Total cost =∑(Total cost for each parameter)+ Overhead cost (15% assumed) Table 4: CAPEX calculation for ODN
Parameters Quantity Total cost(NPR) Cost per subscriber (NPR) OLT and core equipment (1024 1 Set 650000.00 634.77 capacity) UG feeder Fiber 1000 meters 300000.00 585.94 Aerial distribution Fiber 4000 meters 280000.00 546.88 FDC 1 set 25000.00 48.83 FAP 64 set 96000.00 187.50 Drop Fiber 1 set 600.00 600.00 CPE 1set 2500.00 2500.00 Implementation cost 275000.00 537.11 Calculated total Civil cost 1000.00 1000.00 cost Installation and marketing 2000.00 2000.00 Total cost 8641.02 Overhead cost(15% assumed) 1296.15 Total investment (CAPEX) 9937.17
Fig. 13 shows the different cost and their contributions in percentage. The CAPEX would increase in suburban and rural area due to the transportation cost and labor cost. CAPEX is estimated for urban, sub- urban, and rural area. Fig. 14 shows the CAPEX for urban, sub-urban, and rural areas. CAPEX increases in rural area due to heavily increase in transportation cost. Similarly, OPEX includes operation and maintenance cost, customer care cost, International bandwidth for voice, IPTV, data and other VAS services, and interest of investment. Analyzing and calculating OPEX, the OPEX per subscriber per month is estimated to be NPR 532.81. The calculation is carried out based on current international bandwidth price, operation and maintenance cost as 10% of total revenue, and considering interest of investment as 10%. Income is assumed based on the current average income of
189 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi voice, data, and TV subscription rate. Fig. 15 shows the different contribution share for OPEX per month in percentage.
Figure13: Investment share in percentage Internet revenue per month = NPR 800.00 IPTV revenue per month = NPR 200.00 Voice revenue per month = NPR 250.00 Total revenue per month = NPR 800.00 + NPR 200.00 +NPR 250.00 = NPR 1250.00
Figure14: CAPEX for different terrain
Expenses are assumed as, International bandwidth (20% of total consumption) per month = 0.2* NPR800 =NPR160.00 IPTV per month (70%) = 0.7 x 200.00 = NPR 140.00
190 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
Voice (10%) per month = 0.1 x 250 = NPR 25.00 Operation, maintenance, and customer care cost per month = 10% of total revenue = 0.1 x 1250 = NPR 125.00 9937.17 Interest of investment per month = 10% of total investment = 0.1x = NPR 82.81 12 Total investment per month = NPR160.00 + NPR 140.00 + NPR 25.00 + NPR 125.00 + NPR 82.81 =NPR 532.81
Figure 15: OPEX contributions in percentage Payback period: Analyzing CAPEX, OPEX, and revenue calculations, the payback period for urban area is about 20 months. It is considered that the network utilization ratio is 0.5. Fig. 16 shows payback period for different geography. The payback period is longer (38 months) for rural areas due to high transportation, carrying, and labor cost. In some area, transportation is to be carried out by helicopters, aero planes.
Figure 16: Payback period for different geography 4.3 Implementation of GIS Tool for Optical Network Planning A GIS tool is very fruitful to design FTTH network due to its availability, much more functional abilities, and
191 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi openness of software (Matrood et al., 2014). GIS software tools are very fruitful for effective design of FDCs, FAP, feeder Fiber and aerial Fiber, and for estimation of drop Fibers. Fig. 17 shows network design for both feeder Fiber and FDC, FAPs, and aerial Fiber. For the accurate Fiber network planning for both aerial and feeder Fiber, GIS tools is very effective. Fig. 18 shows a sample figure of network diagram drawn using GIS tool (CableCad) for aerial cable and FDC location and a feeder. Table 5: Power loss measurement from OLT to L1 splitter (source: NT)
Port No. Signal Power loss OLT to ODF(dBm) At input of L1 splitter (FDC no. 1)(dBm) 1 3.3 1.1 2 4.4 0.46 3 2.41 0.6 4 2.94 1.49 5 2.36 0.5 6 3 1.96 7 3.12 1.54 8 3 1.17 Average Power loss = 3.06625 dBm Average Power loss = 1.1025 dBm
Table 6 shows the measured values of output of L1 splitter, input and output of a particular FAP or L2 splitter of 1:8 split ratio. The average value of power loss is -8.72375 dBm and -18.23 dBm respectively. So the values are satisfactory for network deployment. The value of loss is -8.88dBm at input of L2 splitter. Similarly, the maximum value of loss is -18.3dBm and minimum value of loss is -18.16dBm at output of L2 splitter. The above values suggest that the network deployment is satisfactory. Table 6: Power loss measurement from L1 splitter to L2 splitter (source: NT)
Port Output of L1 splitter Input of L2 (FAP1) L2 output (FAP1)(dBm) No. (FDC No. 1)(dBm) (dBm) 1 -8.88 -8.88 -18.2 2 -8.69 -18.22 3 -8.7 -18.25 4 -8.69 -18.28 5 -8.77 -18.18 6 -8.72 -18.16 7 -8.65 -18.25 8 -8.69 -18.3 Average power Loss = -8.72375 Average power Loss = -18.23 dBm dBm
4.4 Loss Measurement The practical optical Fiber loss of up to -22dBm for the last mile or end devices from ODF is considered to be sufficient. The power loss from OLT to ODF is 4dBm. Table 5 shows the measured values from OLT to ODF. The average value is 3.0663 dBm. Similarly, the power at the input of a sampled FDC or L1 splitter of
192 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021]
1:8 split ratio is also measured and the average value turn out to be 1.1025 dBm. This measurement is taken of feeder cable up to the length of 3km the distance between FDC and FAP is up to 1km. Optical power meter (OPM) is used for loss measurement, optical time-domain reflectometer (OTDR) is used for finding Fiber cable’s fault, and splicing machines are used to splice the optical Fiber whenever there is requirements to ensure quality of optical Fiber.
Figure 17: FDC, aerial fiber and FAP design map by GIS tools-CableCad (source: NT)
193 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi
Figure 18: Feeder fiber design layout by GIS tools (source: NT)
5. Discussion and Recommendation In this paper, the PON architectures are discussed and implementation in the context of Nepal is analyzed. The different architectures are suitable in different terrain structure in Nepal. It is considered to have reliable optical Fiber backhaul link across all parts of the country. Microwave links should be updated to higher capacities. Based on the CAPEX and OPEX discussion, the following different architecture is recommended in different terrain.
• For corporate customer, AON is suitable, PON architecture can also be used for that purpose too. • For densely populated area, PON having capacity of 1024 and higher is suggested that reduces CAPEX and OPEX. • For sub-urban area, PON architecture with capacity of 512 is recommended. • For rural area, where there is dispersed inhabitation around 200 to 400 population in small market. For sub-urban area PON architecture with capacity of 256 is suitable. For remote area, where there may not be Fiber transmission network, there should be improvement of microwave transmission network and PON having capacity of 64 or 32 is recommended, but it is too costly for implementation. RTDF fund can be utilized for such scenario as Government of Nepal has a policy to use the fund for the rural areas.
6. Conclusions In this paper, the FTTH implementation scenario in the context of Nepal is presented. Cost effective implementation approach is also discussed and some architectural approaches are recommended that are suitable based on demographic condition of Nepal. Backhaul transmission network with optical link is
194 Khatiwoda and Dawadi Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] mandatory for better quality service of FTTH network. Based on the territory, the discussed architectures can be modified and designed in future. High capacity FDCs are proposed for high density area to replace copper network. Microwave link should be upgraded to avoid bottleneck of data traffic. Integrated planning is necessary for building telecommunication infrastructure. While planning for road infrastructure, planning for electric poles, ADSS, underground cable, Fiber laying on the side of bridges is to be carried out jointly. The discussed design and implementation methodology would be a lesson learn to other stakeholders of the developing nations like Nepal.
Acknowledgments We are grateful to Nepal Telecom for providing related facts and figures for analysis and visualization of the work carried out. We appreciate necessary feedbacks and suggestions provided from the planning, project, and other departments of NT.
Conflict of Interests Authors declare no conflict of interests.
References Abdel, E., Soleit, A., & Obour, A.-. (2018). Proceedings of the 11 th ICEENG Conference , 3 - 5 April , 2018 Kobry El-Kobbah , Cairo , Egypt 11 th International Conference on Electrical Engineering ICEENG 2018 A New Adaptive Chaotic Key Generator. April. Babani, S., Bature, A. A., Faruk, M. I., & Dankadai, N. K. (2014). Comparative Study Between Fiber Optic and Copper in Communication Link. International Journal of Technical Research and Applications, 2(2), 59–63. Bag, M., Giri, A., Gangopadhyay, S., & Majumder, J. (2019). Implementation of fiber to the home (FTTH)-based technologies: Opening a new dimension for health care industry in India. International Journal of Engineering and Advanced Technology, 8(6), 5004–5009. https://doi.org/10.35940/ijeat.F9091.088619 Caka, N., & Hulaj, A. (2011). Optimization of FTTH network in Kosovo through the implementation of GPON architecture and analysis of the cost of the implementation. International Journal of Communications, 5(4). Casier, K., Verbrugge, S., Meersman, R., Colle, D., Pickavet, M., & Demeester, P. (2008). A clear and balanced view on FTTH deployment costs. Journal of the Institute of Telecommunications Professionals, 2(3), 27–30. Dawadi, B. R., Rawat, D. B., Joshi, S. R., & Keitsch, M. M. (2019). Towards energy efficiency and green network infrastructure deployment in Nepal using software defined IPv6 network paradigm. Electronic Journal of Information Systems in Developing Countries. https://doi.org/10.1002/isd2.12114 Efficiency, E. (2015). ITU-T Rec. L.102. Optical Fibre Cables for Aerial Application. Itu. FTTH Council Europe. (2014). White Paper: Innovative FTTH Deployment Technologies. 1–28. Horvath, T., Munster, P., Oujezsky, V., & Bao, N. H. (2020). Passive optical networks progress: A tutorial. Electronics (Switzerland), 9(7), 1–31. https://doi.org/10.3390/electronics9071081 ITU-T Q2 / Study Group 15. (2018). ITU-T PON standards - progress and recent activities. Ofc 2018 © Osa 2018, 18. Kadhim, D. J., & Hussain, N. A.-R. (2013a). Design and Implementation of a Practical FTTH Network. International Journal of Computer Applications, 72(12), 50–56. https://doi.org/10.5120/12550-9273 Kadhim, D. J., & Hussain, N. A. (2013b). Link and Cost Optimization of FTTH Network Implementation through GPON Technology. Communications and Network, 05(03), 438–443. https://doi.org/10.4236/cn.2013.53b2081 Khabzli, W., Diono, M., Hasri, E., & Noptin, P. (2019). Design of Gigabit Ethernet Passive Optical Network ( GEPON ) Based Fiber To The Home ( FTTH ) Network in Pekanbaru Citraland Housing. 14(1), 130–136. Khatimi, H., Setya Wijaya, E., Rizky Baskara, A., & Sari, Y. (2019). Performance Comparison Between Copper Cables and Fiber Optic in Data Transfer on Banjarmasin Weather Temperature Conditions. MATEC Web of Conferences, 280, 05022. https://doi.org/10.1051/matecconf/201928005022 Kulkarni, S., Polonsky, B., & El-Sayed, M. (2008). FTTH network economics: Key parameters impacting technology decisions. NETWORKS 2008 - 13th International Telecommunications Network Strategy and Planning Symposium, 3–10. https://doi.
195 Journal of Engineering Issues and Solutions 1 (1): 174-196 [2021] Khatiwoda and Dawadi
org/10.1109/NETWKS.2008.4763668 Larsen, C. P., Gavler., A., & Wang, K. (2010). Comparison of active and passive optical access networks. 2010 9th Conference of Telecommunication, Media and Internet, CTTE 2010, June 2010. https://doi.org/10.1109/CTTE.2010.5557694 Lokhande, M., & Singh, A. (2017). Design and Implementation of FTTH. International Research Journal of Engineering and Technology, 10, 2395–56. Mahloo, M., Chen, J., & Wosinska, L. (2015). PON versus AON: which is the best solution to offload core network by peer-to- peer traffic localization. Optical Switching and Networking, 15, 1–9. Mata, E. C. da. (2014). Design of a Network Access Based on FTTH for Sao Tome and Principe (Vol. 54). Matrood, Z. M., George, L. E., & Mahmood, F. H. (2014). A Simple GIS Based Method for Designing Fiber-Network. International Journal of Engineering and Innovative Technology (IJEIT), 4(2), 49–57. Muciaccia, T., Gargano, F., & Passaro, V. (2014). Passive optical access networks: state of the art and future evolution. Photonics, 1(4), 323–346. NTA. (2018). http://nta.gov.np/wp-content/uploads/2018/01/ToR-of-Consultant-for-Monitoring-and-Evaluation-Broadband. docx. NTA. (2021). Nepal Telecommunications Authority MIS Report. In Nepal Telecommunication Authority (Vol. 194, Issue 144). Nyarko‐Boateng, O., Xedagbui, F. E. B., Adekoya, A. F., & Weyori, B. A. (2020). Fiber optic deployment challenges and their management in a developing country: A tutorial and case study in Ghana. Engineering Reports, 2(2), 1–16. https://doi. org/10.1002/eng2.12121 Rigby, P. (2014). FTTH Handbook. FTTH Council Europe, 5, 1–161. Rokkas, T., Neokosmidis, I., Katsianis, D., & Varoutas, D. (2012). Cost analysis of WDM and TDM fiber-to-the-home (FTTH) networks: A system-of-systems approach. IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews, 42(6), 1842–1853. https://doi.org/10.1109/TSMCC.2012.2227999 Sahu, N., Kher, S., Saxena, M., & Sahu, A. (2019). Introduction of Fiber To The Home Technology. March. https://doi.org/10.13140/ RG.2.2.23634.20169 Schneir, J., & Xiong, Y. (2014). Cost analysis of network sharing in FTTH/PONs. IEEE Communications Magazine, 52(8), 126–134. https://doi.org/10.1109/MCOM.2014.6871680 Wang, K. (2017). Migration Towards Next Generation Optical Access and Transport Networks.
196 JOURNAL OF ENGINEERING ISSUES AND SOLUTIONS
Hariharbhawan, Pulchowk, Lalitpur Email : [email protected] Phone : 01-5010251 ISSN 27951529 Web : https://neanepal.org.np/journal