Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Micro grid for All India Institute of Medical Sciences, Madurai

Micro grid for All India Institute of Medical Sciences, Madurai Keywords: new healthcare facilities; load assessment; microgrid proposal; optimization Received: 15 December 2020; Accepted: 10 March 2021 © The Author(s) 2021. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (http:// creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 255 These are not just energy challenges for healthcare Introduction systems in India; power shortages in a number of coun- Hospitals and healthcare facilities are the foremost tries are still a barrier to effective health services. Energy energy-intensive consumers of the commercial sector issues for hospitals vary in low- and high-income coun- and, as vital components of the delivery system of med- tries. The availability of sustainable energy for common ical services, they are liable for a considerable portion medical needs is the key challenge for health facilities in of total commercial energy consumption in India [ , 1 2]. low-income countries, especially those in sub-Saharan For effective functionality, perpetual operations and Africa [9]. Many healthcare facilities in these countries are state-of-the-art medical equipment, most hospitals must facing a shortage of electricity for critical services, such as have a continuous power supply. Globally, they are under lighting, heating and medical devices. This would lead to tumultuous pressure to achieve more while controlling decreased diagnostic capabilities and treatment facilities, costs and limiting waste, and to offer enhanced patient shortened working hours during the day and shortages consideration [3 , 4]. of committed healthcare professionals due to job dissat- In India, many healthcare facilities are located near elec- isfaction [10]. On the other hand, high-income countries trical grid networks as the primary power-supply resource. are focusing on improving productivity and the practice of Also, power failures and supply interruptions during high- renewable energy to minimize energy use, reduce running demand periods have been problematic in grid-connected costs and environmental impacts [11–15]. cities and regions [5]. In one major event, the biggest grid In the face of rising energy costs, rising demand, the failure in history, over 30 and 31 July 2012 in the northern escalating role of power-dependent technology in med- part of India, massive load demand, poor supply manage- ical care and rising numbers of aggressive storms all over ment and interruptions in transmission services led to a the world, the issue of resilience keeps several healthcare recurrent power-grid breakdown that affected 650 million managers up at all times [16–18]. Thus, the acquisition and people across the country, home to half of India’s popula- appropriate use of electrical energy is a pivotal step for tion. By March 2019, the Indian government had granted any intricate structure that needs to reach the ideal energy USD 2.5 billion to provide electricity to all households [6]; level. Particularly for healthcare facilities, there should be access to affordable electricity remains a major concern for a high level of concern about energy supply and mainten- Indian health services. Even after the remedial measures, ance, as they are the most vital facilities for continuing people were unable to ignore the historic grid breakdown primary care and emergency care. Since recent incidents and blamed it for the series of major longevity instances have led to power outages, there is no bias that puts hos- around the country, most of which were triggered by the pitals and their patients at serious risk. Hospitals are ex- breakdown of life-sustaining equipment such as ventila- ploring innovative energy models and highly developed, tors and incubators. reliable and cost-effective technologies such as microgrids Twenty patients died along with eight children, and to address these critical energy concerns [19]. many others had a tormenting experience when the While monitoring and managing energy supply and de- Kurnool Government General Hospital went into dark- mand, a microgrid will incorporate and regulate a broad ness for 12 hours after the power supply had been shut range of energy sources [20–23]. As a result, in an emer - down from 7  p.m. on Thursday, 22 June 2017 to the fol- gency situation, the microgrid will be able to supply elec- lowing morning [7]. More horrible still, no senior spe- tricity to the hospital whenever the adjacent grid breaks cialist, not even the Resident Medical Officer, was in the down. Whereas the utility is not bound to control black- emergency clinic throughout the blackout. Of the 20 pa- outs, the microgrid works 24 hours a day and throughout tients reported to have passed on, 4 were ladies. A  se- the year [24–26]. The microgrid is constantly on the job, un- nior specialist admitted that power disturbance caused like the backup generators that need to be kept under con- the vast majority of deaths. As per source, power was trol to ensure they can function in a crisis. In addition, if switched off when police outpost personnel snarled elec- not required for backup purposes, the microgrid allows the trical connections to high-tension wires on clinic prem- healthcare facility to achieve its cost and sustainability ob- ises to draw power illegally. jectives through demand–response, peak shaving and load On 24 July 2016, 21 patients died from power outages at management, and excess-energy sales to the grid other the state-run Gandhi Hospital, the most prominent 1200- than the use of on-site generators and energy-storage re- bed hospital in the city of Hyderabad [8]. A few specialists sources when electricity rates are high on the grid [27, 28]. said that power had initially stumbled at around 3  p.m. Microgrid technology is very effective, as can be seen and then continued to do so on a regular basis. In spite of from the installation of California’s leading large-scale the fact that there were four generators on standby, the microgrid for healthcare services. In 2018, at the Kaiser emergency clinic guaranteed that they had peak power Permanente facility in Richmond, California, the California generation and could not be used to identify the explan- Energy Commission issued a USD 4.78 million grant to ation for the stumbling after electrical cables were later Charge Bliss, Inc. to develop, build and run the foremost cut off. Deaths occurred in specialty clinics, as well as in renewable-energy-based microgrid for a healthcare hos- the Surgical Intensive Care Unit and the Emergency Ward. pital in California [29]. The hospital was the only public Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 256 | Clean Energy, 2021, Vol. 5, No. 2 hospital serving western Contra Costa County to provide based on the smart-grid concept of providing two-way the local population with critical care. In addition, the area power and communication facilities to enhance the reli- was affected by a greater degree of environmental con- ability of the power supply at all times. Of course, due to tamination and subsequent health consequences. A  new smart-grid technology, the operating and maintenance microgrid controller was also developed for the project (O&M) cost of AIIMS is slightly higher than that of con- that will isolate the hospital’s life-safety emergency-power ventional microgrid systems. division, including emergency lighting and exit signs, and provide emergency power services. The microgrid resulted in an annual output of 395 000 kWh of renewable energy, 1 System modelling a 4-hour reduction in demand of 220 kW from the battery AIIMS Madurai will have a hospital capacity of 750 beds and a reduction in emissions of 739 metric tons of CO /yr. with an interim departmental distribution of beds as The Ministry of Health & Family Welfare, Government shown in Table 1. of India, approved a project for the development of a new All India Institute of Medical Sciences (AIIMS) in Madurai, Tamil Nadu, with a budget of USD 178 mil- 1.1 Site location lion in 2018 [30]. AIIMS includes the establishment of hospitals, trauma-centre facilities, medical colleges, Thoppur village, Madurai-625008, is the proposed location residential complexes and related facilities/services, for the new AIIMS. The village of Thoppur is located in the mostly based on AIIMS, New Delhi, and six proposed south of Madurai, Tamil Nadu State, India. It is located 20 new AIIMS. The aim is to establish the new AIIMS as a km from Madurai South Sub-district Headquarters and nationally significant educational institution to provide 20 km from Madurai District Headquarters. As per the quality medical education, as well as tertiary healthcare Global Positioning System, its coordinates are 9° 52.3’ N facilities for the people of Madurai and the surrounding (latitude) and 78° 1.5’ E (longitude), respectively. Fig. 1 il- area. The main purpose of this paper is to provide a re- lustrates the site of the planned Thoppur AIIMS, Madurai. liable, economical and efficient power supply based on The annual average air temperature over a 30-year period a microgrid for the new AIIMS Madurai. It is the right (January 1984 to December 2013) was found to be 27.34°C time to propose the inclusion of the microgrid, because and the annual average wind speed at 50 m above the sur - the project is in the tendering stage. The greatest sig- face of Earth over a 30-year period was 4.97 m/s, according nificance of this suggestion is that AIIMS New Delhi (es- to NASA’s Prediction of the Worldwide Energy Resource tablished in 1956) ranked first in India by the National (POWER) Database [35]. The solar radiation and the clear - Institutional Ranking Framework [31], ranked first in ness index of the proposed site as per the US National South Asia and 231st in the Life Sciences and Medicine Renewable Energy Laboratory database [36] are shown in category worldwide by QS WUR [32], recognized and Fig. 2 and the annual average irradiance value is 5.61 kWh/ introduced the value of the microgrid only in 2017 [33, m /day. 34]. The novelty of the proposed microgrid is that it was suggested at the beginning of the approval stage of the Table 1: Department-wise bed distribution for the proposed AIIMS Madurai itself during the preparation of the in- AIIMS healthcare facility in Madurai frastructure. Recently, the existing AIIMS in India began planning renewable-energy systems with PV systems Departments Beds only a decade or a few decades after its inception. a) Speciality Departments 360 Having PV alone is unreliable and, in countries like India (Surgical & Allied Specialities, Medicine & Allied with four seasons, operating efficiency is <14%, resulting Specialities, and Obstetrics & Gynaecology) in a higher unit cost of electricity generation. Moreover, b) Super Speciality Departments 215 because PV systems need vast land, the construction of (Cardiology, Cardio-thoracic Vascular solar farms without public support is not straightfor - Surgery-CTVS, Gastroenterology, Surgical ward. In light of all these issues, a wind turbine (WT), Gastroenterology, Nephrology, Urology, as another renewable-energy system, has been included Neurology, Neuro-surgery, Paediatric Surgery, in this project, as the wind profile in the proposed area Burns & Plastic Surgery, Medical oncology, Surgical Oncology, Radiation Oncology, is favourable. At this site, a 1-kW WT with a height of 30 Endocrinology and Pulmonary Medicine) m will produce an energy output of 1750 kWh/yr with c) Other facilities 175 an attractive payback period of <7  years. The proposed (Intensive Care Units (ICUs) & Critical Care, microgrid is such that >50% of its maximum demand Trauma, AYUSH Facilities, PMR Department and could be met by both PV and wind turbines, thus redu- Paid Beds) cing dependence on the utility grid, apart from environ- Total number of beds 750 mental credits. In addition, the proposed microgrid is Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 257 Fig. 1: The approved and inaugurated location for AIIMS in Thoppur Madurai. This figure was used with permission from [47]. Fig. 2: The solar radiation and the clearness index of the proposed site as per the US National Renewable Energy Laboratory database [36] devices with lower energy consumption are chosen. LED 1.2 Demand assessment lighting fixtures with built-in harmonic-suppression sys- One of the key tasks for the energy management of the tems are considered for lighting in all areas and buildings new system is the assessment of demand. Energy con- in accordance with the National Building Code (NBC) 2016 sumption depends on activity; climate and changes in oc- [37], the Energy Conservation Building Code (ECBC) [38] and cupancy, at different times of the day; days of the week; the Indian Standard Code. The demand assessment has weather conditions; and seasons. India has four seasons been estimated on the basis of the covered area of various and this is properly considered in the assessment of de- buildings/blocks as per NBC 2016 considering the lighting mand. Besides, the hospital loads are classified as critical load as 13 W per square metre and power load as 55 W per and non-critical loads, depending on their importance, and square metre minimum. A 10% load was considered for fu- this type of classification has been done during demand ture expansion. Table 2 shows the different loads and the assessment in this project to give priority for availability respective locations and the quantity of the total amount duration. A load diversity varying from 0.5 to 1.0 has been of the connected load requirement in addition to the total considered. All medical devices, such as ventilators, anaes- demand for kVA at a power factor of 0.9. thesia machines, blood and infusion heaters, MRIs, x-rays and computed tomography scan (CTS), etc., require steady and good-quality electricity supplies for their operation 1.3 Utility requirement and are considered to be critical loads, while hospital loads, Tamil Nadu Generation and Distribution Corporation such as heating, ventilating, and air conditioning (HVAC) Limited, TANGEDCOUP [39] under state government con- systems and water heaters, etc., are grouped together as trol has a generation capacity of 18 747.28 MW consisting non-critical loads. Modern and sophisticated medical Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 258 | Clean Energy, 2021, Vol. 5, No. 2 Table 2: Category-wise (power, lighting and residential loads, and total load) demand assessment of AIIMS Madurai a) Power loads Load and location kW Load and location kW OPD & Diagnostic Block 550 Night shelter 135 Medical equipment 905 Auditorium 98 Hospital Block (720 beds) 2320 Fire station 87 Ayush Hospital (30 beds) 115 Dining hall 75 Medical & Nursing College 1040 Kitchen equipment 44 HVAC 1925 Director’s residence and servant quarters 40 Lifts (56 nos) 710 Guest house 38 Plumbing and pump load 560 Mortuary 32 Apartment blocks 885 Waste-management block 32 PG hostel blocks 802 Laundry 32 Working nurses’ hostel 420 Shopping centre 22 Boys’ and girls’ hostels 340 Cafeteria 14 Total load in kW 11 221 Load in kVA at 0.90 power factor 12 468 b) Lighting loads Load and location kW Load and location kW OPD & Diagnostic Block 139 Fire station 7 Hospital Block (720 beds) 624 Dining hall 19 Ayush Hospital (30 beds) 30 Kitchen 24 Medical & Nursing College 267 Director’s residence and servant quarters 10 Plumbing and pumping house 7 Guest house 11 Apartment blocks 230 Mortuary 11 PG hostel blocks 207 Waste-management block 15 Working nurses’ hostel 111 Laundry 11 Boys’ and girls’ hostels 85 Shopping centre 11 Sports-field lighting 93 External lighting 11 Night shelter 37 Cafeteria 7 Auditorium 81 Total load in kW 2048 Load in kVA at 0.90 power factor 2276 c) Lighting and power loads of residential blocks Residential loads Lighting and power loads in kW 1131 Lighting and power loads in kVA at 0.90 power factor 1257 d) Total load in kW and kVA Total connected loads Lighting and power connected loads in kW 14 400 Lighting and power loads in kVA at 0.90 power factor 16 001 1.4 Backup power supply of TANGEDCO state-owned ventures, Central Generating In the event of power failure on the part of the utility, Stations shares and Private Power Purchase. In addition, diesel-generator (DG) sets are recommended for backup. the state has projects of up to 1 047 961 MW in renewable- The main concern is that the storage of diesel fuel should energy sources such as wind, solar, biomass and cogener - be there at all times. The main fuel-storage tanks must be ation. It will supply electricity to meet the electrical load tracked and refilled when they run down. Automated con- requirements of AIIMS, Madurai, from its nearby 110/11-kV trol systems are available to inform the generator operator substation located in Kappalur, Madurai, at a distance of <2 when the level of the tank goes below the specified level. km from AIIMS. AIIMS Madurai will have its own 20-MVA The capacity of the backup power supply has been fixed distribution substation with 2  × 10 MVA, 11/0.415-KV dis- on the basis of the maximum demand. For the connected tribution transformers. The utility has a commercial tariff load of 14  400 kW, taking into account the average div-er of USD 0.085/kWh and USD 5/kW. The buy-back tariff for sity factor of 0.8, the maximum demand is 11 520 kW. As renewable-energy supplies is USD 0.062/kWh, so that con- a result, a backup power supply of 12 000 kVA for DGs has sumers can sell their excess energy from renewable energy been recommended. It is advisable to use a large number to the grid. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 259 of generators rather than a single large-capacity generator and Nursing College, the Ayush Block, the auditorium and in terms of reliability, operating efficiency and economics. the Computer Centre. It is recommended to use a 4-hour, The most significant locations to be provided with 100% 1-MWh capacity Li-ion battery [40]. It offers outstanding backup power are: functionality for microgrid implementation. Depending on the size of the microgrid, the Li-ion battery can be used as an energy-storage system due to its extended size range. OPD & Diagnostic Block, Hospital Block (720 Full backup Furthermore, its falling price and the recovery of perform- beds), Ayush Hospital (30 beds), Medical power & Nursing College, HVAC loads, plumbing supply ance and lifetime enhance the applicability of this Li-ion and pump loads, all chillers and all AHUs, battery. fire station, mortuary, laundry, auditorium, guest house, dining, shopping complex and director’s residence 1.6 Renewable-energy options Renewable energy is an attractive choice worth exploring, For providing a 100% backup supply, four 2000-kVA (4  × depending on availability, accessibility and cost, policies 2000-kVA) generators are suitable and the strategic loca- and incentives, and pricing and regulations for electri- tion for their placement is near the Hospital Block. city. A variety of renewable options are worth considering: Other facilities and loads are provided with backup solar, wind and biomass resources. This initiative currently power of four 1000-kVA (4 × 1000-kVA) DGs, including: lifts targets both solar and wind energy systems. (56 total); apartment blocks; PG hostel blocks; working At the location of AIIMS Thoppur, Madurai (latitude: 9° nurses’ hostel; boys’ and girls’ hostels; night shelter; 52.3’ N and longitude: 78° 1.5’ E), the annual average solar sports-field lighting; waste-management block; cafeteria; radiation is found to be 5.61 kWh/m/day. Fig. 2 shows that, residential blocks; lighting and fan loads; water and fire- regardless of the four seasons, solar radiation is available fighting pumps; street lighting; and emergency services. throughout the year. The average annual radiation lasts for The location for the DGs is near the sports field. All DGs 5 months (42%) per year and the maximum radiation takes are of an outdoor type with hospital-type silencers and place for a period of 3 months (25%). acoustic enclosures. The total capacity of the DG backup The universal method for determining the electricity power is therefore 12 000 kVA. generated by a photovoltaic system [41] is as follows: (2) E = A ∗ r ∗ H ∗ PR kWh where E represents the energy produced (kWh), A repre - 1.5 Energy storage sents the total area of the solar panel (m ), r represents the Energy storage is quickly moving forward and becoming solar-panel yield or efficiency (%), H represents the annual a key player in the future of microgrids. In the context of average solar radiation on tilted panels without shading the hybrid renewable-energy concept for the healthcare and PR represents the performance ratio, coefficient for microgrid, the integration of wind and solar energy be- losses (between 0.5 and 0.9, default value = 0.75). comes inevitable. As both wind and solar have variable The site is lucrative so that, with a payback period of outputs, storage technologies have an immense potential <7  years, a 1-m PV panel will generate ~2050 kWh/yr. In to smooth the supply of energy from these sources. This view of the yield and the rate of return, a 4-MW PV system storage option helps operators to dynamically balance en- with a capacity of ~35% of AIIMS Madurai’s maximum de- ergy on the grid by alternatively injecting and consuming mand has been proposed. While fixing the capacity, the excess electricity. Without a balance of supply and de- annual solar variation has been duly considered and the mand, the stability of microgrids becomes questionable. backup power supply has been fixed with a diversity factor Various energy-storage technologies contribute to stability of 0.8. by operating at different stages of the microgrid from gen- The AIIMS site has an average annual wind speed of eration to end use by customers. Battery storage is prefer - 4.97 m/s at an anemometer height of 30 m. It lasts for 4380 able in hospitals due to portability, flexibility and economic hours per annum, as seen in Fig. 3. The power output of a concerns. The capacity of a storage plant is given as the WT is given by the expression [42, 43]: following expression [40]: (3) C =(E ∗ AD) / (η ∗ η ∗ DOD) Wh(1) P = 1/ ∗ ρ ∗ A ∗ v ∗ C Watts Wh L inv bat T 2 p where P represents the power output (W), ρ represents the where C represents the battery capacity (Wh), rep E - Wh L T air density, 1.22 kg/m, A represents the turbine swept area resents the average energy load per day, AD represents (m ), v represents the wind speed (m/s) and C represents is the number of autonomous days of the battery, ɳ inv p the performance coefficient of the WT. represents the efficiency of the inverter, ɳ represents bat At this site, a 1-kW WT with a height of 30 m will the efficiency of the battery and DOD represents the produce 1750 kWh/yr of energy. Higher WT-generator battery-discharge depth. capacity with higher hub height is an efficient and eco- Battery backup is provided for operation theatres, essen- nomical option for better yield. An attractive payback tial loads and medical equipment of the Hospital, Medical Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 260 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 3: Annual average wind speed at the AIIMS Thoppur Madurai site. This figure was used with permission from [47]. period of <7 years is feasible with 900-kW capacity WTs. Java-based graphical user interface that supports the Two such WTs (2 × 900 kW) are proposed for this AIIMS decision-making function of surviving urban develop- project. Including PV solar and wind power, the contri- ment. The objective of this instrument is to simulate and bution of renewable energy to the maximum demand of optimize the flow of building-related resources (energy, AIIMS Madurai will be 50.35%. water and waste) and their inter-relationships, as well as Since weather (solar irradiance, wind) varies from year to investigate their dependence on urban climates. to year, a slightly higher capacity to cope with the rise and fall of generation due to year-to-year wind and solar irradi- ance was considered with a diversity factor of 0.8 when 2.3  EAM designing the power-generation capacity of the microgrid. EAM is used to evaluate the economic viability of microgrids [46]. It is capable of optimizing the capacity of microgrids through the proper choice of equipment 2 Optimization platform selection, power rating, capital and running costs, and As the AIIMS Microgrid infrastructure has been finalized, lifetime equipment. There is provision for a comparison the next step is to evaluate its performance and whether of the optimized cost of the microgrid against the energy or not it can meet the energy requirements of AIIMS cost of the utility. Madurai without sacrificing its reliability. To do so, an on- line platform becomes essential to achieve the objective economically. Computational approaches are usually used 2.4 HOMER Grid to model, test and forecast energy efficiency in the de- Based on software developed by the National Laboratory sign of microgrid projects. There are a number of software for Renewable Energy (NREL) [47], HOMER Grid is designed packages available for the optimization mission and it is for the economic and engineering assessment of grid- wise to select the most flexible one. connected and off-grid energy systems. The key capability of HOMER Grid is to simulate the efficiency of any specific configuration of the energy system. However, on specified 2.1  DER-CAM systems, the programme is also adequately capable of run- DER-CAM was built by Lawrence Berkeley National Lab, ning economic optimization and sensitivity/uncertainty Berkeley, California, USA [44]. It is a decision-support device analysis. It is important to note that the optimization is for decentralized power systems. The important objective performed on parameters that the designer has control of DER-CAM is to conduct a techno-economic comparison over. Sensitivity analysis is based on factors that are sub- of distinct on-site electricity-generation technologies or ject to uncertainty or shift that are beyond the control of microgrids and to optimize the costs of on-site electricity the designer, like wind speed and the price of fossil fuel. generation using linear programming techniques. HOMER Grid input data include customer load profiles for electrical and thermal energy, any resources and fuel used by the electrical and thermal power-generation sys- 2.2  CitySim tems, energy-system components, electrical and thermal Built by the Laboratory of Solar Energy and Physics load curves with a resolution of ≤1 minute, technical ef- Buildings (LESO-PB), Switzerland [45], CitySim has a ficiency, O&M costs, emission limitations and sensitivity Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 261 parameters. HOMER Grid outputs the results of the assess- For AIIMS Madurai, demand is estimated based on data ment and analysis in the graph format and in the compre- from other AIIMS with the same bed ability and similar hensive data reports. location status. Taking into account the number of beds; Based on the extended facilities, the accuracy of the re- the number of medical and nursing students; the likely sults and the ease of availability, the HOMER Grid platform number of outpatients every day; the number of phys- was preferred for the design of the microgrid for AIIMS icians, nurses and support staff; the number of hostel Thoppur, Madurai. inmates; and the number of campus residents, the head count is ~6200 people. The hourly load for the approximate head counts was reached on the basis of the same refer - 3 AIIMS Madurai Microgrid ence hospitals, as in Table 3. Unlike standard commercial loads, for 9–18 hours a day, HOMER Grid, the identified optimization platform, performs the hourly load stays at its peak for 10 hours. The min- simulation calculations on an hourly basis throughout imum load occurs during the early morning and late night, 8760 hours of the year [48]. Based on the energy balances, when there will be no outpatients and most loads will the estimations compare the thermal and electrical en- be working at less than critical loads; lifts; and medical ergy both supplied and required by the system. Based on equipment. the energy balances, the estimations compare the thermal The HOMER Grid database for electricity consump- and electrical energy both supplied and required by the tion provided the seasonal and yearly load profiles for system. HOMER Grid decides on the optimal configuration AIIMS Madurai Microgrid as in Figs 4 and 5 . The most of the system as well as analyses of the cost of the system noteworthy month-to-month load utilization would be in on the basis of these estimates. There was a significant August, equal to 11 017.9 kW, and the load factor is found assumption that the energy efficiency of the storage bat- to be 0.43. tery remains constant over the life of the battery, while, depending on the battery power and the ambient tempera- ture, it varies, among other factors. Furthermore, the nom- 3.2 Microgrid architecture inal voltage, the power curve, the lifetime curve and the Fig. 6 depicts the proposed architecture for the microgrid minimum charge state of the battery are considered con- with its primary-load requirement. It includes the utility stant. An interest rate of 6% per year and a 25-year project grid as the main electricity supply with a simple tariff, a lifespan are assumed in the computation. backup DG, battery energy-storage system (BESS), two WTs each of 900-kW capacity, three PV solar systems (2  × 1000 3.1 Daily, seasonal and yearly load profiles kW + 1 × 2000 kW) with a total capacity of 4 MW and a con- The actual daily-load demand is uncertain but predictable verter. The WTs and the PV solar systems have their own because AIIMS Madurai is in the development process. independent converters. Apart from that, being a critical Table 3: Estimated daily-load profile of AIIMS Madurai Hour and load in kW Hour 0 1 2 3 4 5 6 7 8 9 10 11 Load, kW 2910 2910 2915 2915 2915 3015 3230 4314 5380 6111 6111 6111 Hour 12 13 14 15 16 17 18 19 20 21 22 23 Load, kW 6111 6111 6111 6111 6111 6111 6111 5901 5190 3950 2935 2910 Fig. 4: The seasonal load profile of AIIMS Madurai Microgrid. This figure was used with permission from [47]. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 262 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 5: The yearly load profile of AIIMS Madurai Microgrid. This figur e was used with permission from [47]. has the flexibility and suitability to simulate future design configurations in the name of the options: Reference: Only the utility grid Option 1: Utility grid, generator backup and WT with bat- tery storage system (Utility, DG, WT, BESS); Option 2: Utility grid, generator backup and PV solar with battery storage system (Utility, DG, PV, BESS); Option 3: All—utility grid, generator backup, WT, PV solar with battery storage system (Utility, DG, WT, PV, BESS). The output performances of all optimization options are compared against the reference on the basis of (i) annu- alized cost; (ii) electricity production, consumption and sales; (iii) economic; (iv) annual utility bill; (v) fuel con- sumption; and (vi) emission release. 4 Results and discussions 4.1 Reference option: only with the utility grid The primary input is the primary-load data from Table 3 and the power-grid tariff as USD 0.085/kWh for electricity and USD 5/kW for demand in order to perform this opti- mization. The energy buy-back tariff is USD 0.062/kWh for excess-energy sales. Output optimization yields: Fig. 6: The proposed architecture of the AIIMS Microgrid with its primary- load requirement. This figure was used with permission from [47]. Total electricity production (kWh): 41 062 500 Total net present cost (NPC) (USD): 52 957 160.61 healthcare facility, to highlight the role of a converter in LCOE (USD/kWh): 0.09976 the microgrid, a separate converter has been included as a Consumption charge (USD): 3 490 312.50 reliability concern. Demand charge (USD): 606 155.71 Apart from the utility tariff, the appropriate initial cap- CO emission (kg/yr): 52 957 160 ital cost, replacement cost and operation and maintenance Carbon monoxide (kg/yr): 0.0 costs of the DG, renewable-energy sources, storage battery Unburned hydrocarbons (kg/yr): 0.0 and the converter are given in Table 4, which are the inputs Particulate matter (kg/yr): 0.0 for the optimization. SO emission (kg/yr): 112 511 NO (kg/yr): 55 024 3.3 Optimization options The optimization indicates the monthly need for grid electricity to satisfy the load demand of the microgrid, as Microgrid optimization has been carried out with different shown in Fig. 7, which incurred an annual USD 4 096 468.21 combinations (or so-called options here) of integrating re- utility bill. newable and backup generation with BESSs. HOMER Grid Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 263 Table 4: Optimization input data for DG, WT, PV, BESS and converter Component costs Component Initial capital Replacement O&M Diesel generator (DG) (Auto size Genset., 12 000 kW) USD 750/kW USD 500/kW USD 0.01/ operating hour Wind turbine (WT), 900 kW (Enercon E-44, 55-m hub height) USD 749 700 USD 749 700 USD 36 000/yr WT, 900 kW (1) (Enercon E-44, 55-m hub height) USD 749 700 USD 749 700 USD 36 000/yr PV panel, 1 MW (ABB PVS 800–1000) USD 625 000 USD 625 000 USD 25 000/yr PV panel, 1 MW (1) (ABB PVS 800–1000) USD 625 000 USD 625 000 USD 25 000/yr PV panel, 2 MW (ABB PVS 980–2000) USD 1250 000 USD 1 250 000 USD 50 000/yr BESS (battery) storage, 7030 Ah (generic 4-hour 1-MW Li-ion, USD 500 000 USD 500 000 USD 5000/yr 90% roundtrip efficiency) Converter (generic system converter, inverter 2880 kW and USD 300/kW USD 300/kW – rectifier 2880 kW) Utility tariff Energy: USD 0.085/kWh Demand: USD 5/kW Energy buy-back: USD 0.062/kWh Fig. 7: The monthly utility-grid electricity requirement to meet the load demand of AIIMS Madurai 4.2 Option 1: Utility, DG, WT, BESS Fig. 8 depicts the microgrid architecture for Option 1 con- sisting of the 12  000-kW DG, two WTs each of 900-kW capacity, a generic 4-hour 1-MW Li-ion kW battery and the 2880-kW generic system converter. All are integrated with the utility-grid supply system; their relevant data are obtained from the HOMER Grid database and the load and utility-grid data are already available in the optimization platform. The outcomes of the Option 1 optimization are listed in Tables 5–12 with the reference-option comparison. According to the findings, the DG set had a total fuel consumption of 38  054  L for 32 hours and contributed 137 392 kWh/yr of electricity, with an average electrical output of 4294 kW. Its contribution to the overall output of electricity under Option 1 is just 0.335%. This is be- cause of the major contribution from the wind turbines and the BESS even during demand-management situ- ations to reduce the emission contribution by the DG set. The generated power output per year indicating the minimum (3000 kW), mean and the maximum (7695 kW) values of the DG are shown in Fig. 9. The specific fuel consumption works out to be 0.277  L/kWh and a mar - Fig. 8: The microgrid architecture for Option 1 consisting of Utility, DG, WT, BESS with converter.This figur e was used with permission ginal generation cost of USD 0.236/kWh at a mean elec- from [47]. trical efficiency of 36.7%. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 264 | Clean Energy, 2021, Vol. 5, No. 2 Table 5: Optimization of Options 1, 2 and 3—comparison of annualized costs Option 1 (Utility , Option 2 (Utility , Option 3 (Utility, DG, Category DG, WT, BESS) DG, PV, BESS) WT, PV, BESS) Capital cost (USD) 812 174.56 996 169.44 1 108 159.48 Replacement cost (USD) 36 976.92 45 225.08 80 506.98 Operation and maintenance cost (USD) 2 578 060.98 2 210 797.12 1 998 367.91 Fuel cost (USD) 38 053.99 43 163.71 43 973.54 Salvage (USD) –126 094.08 –111 543.33 –131 692.56 Total cost (USD) 3 339 172.37 3 183 812.01 3 099 315.34 Table 6: Optimization of Options 1, 2 and 3—comparison of electricity production, consumption and sales Option 1 (Utility, DG, WT, Option 2 (Utility, DG, PV, Option 3 (Utility, DG, WT, Category BESS) BESS) PV, BESS) Electricity production kWh/yr % kWh/yr % kWh/yr % DG 137 392 0.335 150 094 0.362 152 034 0.366 WT 900 1 575 747 3.84 -- -- 1 575 747 3.84 WT 900 (1) 1 575 747 3.84 -- -- 1 575 747 3.84 PV 1000 -- -- 1 801 544 4.34 1 801 544 4.34 PV 1000 (1) -- -- 1 801 544 4.34 1 801 544 4.34 PV 2000 -- -- 3 610 387 8.68 3 610 387 8.68 Grid purchases 37 774 048 91.985 34 142 167 82.28 30 973 516 74.594 Total 41 062 934 100 41 505 737 100 41 490 520 100 Electricity consumption and sales AC primary load 41 062 500 100 41 062 500 100 41 062 500 100 Grid sales 434 0.001 1462 0.0036 5471 0.0133 Total 41 062 934 100 41 063.962 100 41 067 971 100 Excess energy 0.0 0.0 58 159 0.14 27 885 0.0672 Renewable energy RE fraction 7.68% 17.36% 25.04% Max. RE penetration 119% 175% 212% Table 7: Optimization of Options 1, 2 and 3—economic comparisons Option 1 (Utility, DG , Option 2 (Utility , Option 3 (Utility , Category WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Present worth (USD) 1 086 663 3 095 088 4 187 418 Annual worth (USD/yr) 84 058 239 419 323 915 Return on investment (%) 9.5 9.9 9.8 Internal rate of return (%) 13.0 13.6 13.6 Simple payback (yr) 6.96 6.71 6.75 Discounted payback (yr) 9.22 8.79 8.85 Total NPC (USD) 43 167 210 41 158 780 40 066 450 LCOE (USD/kWh) 0.08132 0.07753 0.07547 Operating cost (USD) 2 526 998 2 187 642 1 991 156 Fig. 10 presents the monthly average renewable-energy requirement, which has been sold to the utility. Fig. 11 por - power generation. Both WTs contribute 3 151 494 kWh an- trays the monthly electricity production of the microgrid. nually, which is equal to 7.68% of 41  062  934 kWh/yr, the With respect to the reference option, there is a reduc- total electricity generated by the Option 1 microgrid con- tion in carbon-dioxide emission by 1 978 346 kg/yr, sulphur figuration. They operated for 8109 hours/yr with a LCOE dioxide by 8766  kg/yr and nitrogen oxides by 3817  kg/yr. of USD 0.0648/kWh. The annual electricity purchase from The Option 1 has a present worth of USD 1  086  663, an the utility grid is 37  774  048 kWh (91.985%). There is an annual worth of USD 84  058/yr, a return on investment excess of 434 kWh after meeting the microgrid energy of 9.5% and an internal rate of return of 13.0% with a Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 265 Table 8: Optimization of Options 1, 2 and 3—annual utility-bill comparisons Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Consumption charge (USD) 3 210 767.20 2 901 993.58 2 632 409.66 Demand charge (USD) 544 523.51 451 593.26 436 507.97 Total (USD) 3 755 290.71 3 353 586.84 3 068 917.63 Table 9: Optimization of Options 1, 2 and 3—savings in annual utility bill against reference model (utility and DG) Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Consumption charge (USD) 279 545.30 588 318.92 857 902.84 Demand charge (USD) 61 632.20 154 562.45 169 647.74 Total (USD) 341 177.50 742 881.37 1 027 550.58 Table 10. Optimization of Options 1, 2 and 3—fuel-consumption comparisons Option 1 (Utility , Option 2 (Utility , Option 3 (Utility, DG, Category DG, WT, BESS) DG, PV, BESS) WT, PV, BESS) Total fuel consumed (L) 38 054 43 164 43 974 Specific fuel consumption (L/kWh) 0.277 0.288 0.289 Mean electrical efficiency (%) 36.7 35.3 35.1 Marginal generation cost (USD/kWh) 0.236 0.236 0.236 Hours of operation 32 44 46 Mean electrical output (kW) 4294 3411 3305 Table 11: Optimization of Options 1, 2 and 3—emission comparisons Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Carbon dioxide (kg/yr) 23 972 809 21 690 836 19 690 368 Sulphur dioxide (kg/yr) 103 745 93 826 85 149 Nitrogen oxides (kg/yr) 51 207 46 420 42 186 Table 12: Optimization of Options 1, 2 and 3—reduction (–)/increase (+) in emissions compared with the reference option Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Carbon dioxide (kg/yr) –1 978 346 –4 260 664 –6 261 132 Sulphur dioxide (kg/yr) –8766 –18 685 –27 362 Nitrogen oxides (kg/yr) –3817 –8604 –12 838 simple payback period of 6.96 years. The total NPC is USD Option 1.  There are two 1-MW PV systems and a single 43 167 210, the LCOE is USD 0.08132/kWh and the operating 2-MW PV systems along with the DG and the BESS with cost is USD 2 526 998. There are savings in utility bills on system converter. All are interconnected with the utility electricity-consumption charges USD –279 545.30 and de- main supply as in Fig. 12. As before, the data for the newly mand charges USD –61 632.20 against the reference option added PV systems are obtained from the HOMER Grid besides demand-reduction revenue of USD 279  681. As a database. The renewable-energy-capacity addition is more whole, the economic metrics of Option 1 are attractive than double in this option (4 MW in Option 2 versus 1.8 when compared with those of the reference option. MW in Option 1). Tables 5–12 portray the outcomes of the Option 2 opti- mization with a comparison against Options 1 and 3. The 4.3 Option 2: Utility, DG, PV, BESS total electricity production by this microgrid Option 2 is The significant difference between Options 1 and 2 is the 41  505  737 kWh/yr, of which the renewable contribution addition of PV systems in Option 2 in the place of WTs in is 7 213 475 kWh/yr (17.37%), whereas the DG has a minor Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 266 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 9: The DG-generated power output per year indicating its minimum (3000 kW), mean and maximum (7695 kW) values Fig. 10: The monthly average power output of both wind turbines Fig. 11: The monthly electricity production of the Option 1 microgrid for a 12-month period Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 267 contribution of 0.362%. The PV systems operate for 4376 has been reduced by 3 631.881 kWh (10.64%), which means hours/yr and the LCOE is USD 0.0406/kWh. The utility pur - that the dependence on the utility grid has been decreased. chase of energy is 34  142  167 kWh as the major share of Fig. 13 presents the annual primary-load variations, the 82.28%. In Option 2, the electricity purchase from the utility grid purchases and the grid-demand limit. Fig. 14 depicts the demand–response coordination of all the elements of Option 2. The demand incentive is set to USD 35/kW. The load peak is 8000 kW and the demand-reduction revenue works out to be USD 279 681. There is an excess of electricity production that has been sold to the grid. Option 2 has a return on invest- ment of 9.9% and an internal rate of return of 13.6% with a simple payback period of 6.71  years. The LCOE is USD 0.077532/kWh, whereas, in Option 1, it was USD 0.08132/ kWh. There are total savings on utility bills for electricity charges of UDS 742  881.37 against the reference option and, with reference to Option 1, the savings are higher by USD 401  703.87. There are considerable reductions on emissions in Option 2 with respect to the reference op- tion such as carbon-dioxide reduction by 4 260 664 kg/yr, sulphur-dioxide-emission reduction by 18  685  kg/yr and nitrogen-oxides reduction by 8604 kg/yr. These reductions are much higher than those of Option 1, too, because of the higher amount of renewable penetration. 4.4 Option 3: Utility, DG, WT, PV, BESS Option 3 is the renewable-energy-enhanced option con- sisting of both WTs and solar PV systems. The renewable- energy total capacity is 5.8 MW (consisting of 1.8 MW WT and 4 MW PV), which is >52% of the daily maximum Fig. 12: The microgrid architecture for Option 2 consisting of Utility, demand. All the data are already available in the opti- DG, PV, BESS with converter.This figure was used with permission from [47]. mization platform with the support of the HOMER Grid Fig. 13: Annual primary load, grid purchases and grid-demand limit in Option 2 Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 268 | Clean Energy, 2021, Vol. 5, No. 2 database. The demand-management option has been pro- generation is found to be 41  490  520 kWh, which is the vided as before. The microgrid architecture of this option highest generation among all the three options. While is shown in Fig. 6. The comprehensive outcome results are comparing the annual utility energy cost among all the presented in Tables 5–12. options, the reference option incurs the highest utility In Fig. 15, the monthly total energy production by all cost because of no renewables contributed. Among the the resources in Option 3 is given. The annual energy renewable-integrated Options 1–3, Option 3 has the lowest Fig. 14: The demand–response events in Option 2 Fig. 15: The monthly energy production in Option 3 with Utility, DG, WT, PV, BESS Fig. 16: The annual utility energy cost including the consumption and demand charges Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 269 utility energy cost, as seen in Fig. 16. In percentage meas- is no renewable-energy source. It is a good sign that the ures, the reference option is 100, Option 1 is 91.985, Option grid dependency has been reduced by >25% because of the 2 is 82.25 and Option 3 is the lowest, equal to 74.574. 25.04% energy production by the renewables (both the WT Apart from the electricity purchase from the utility, the and PV) and a small contribution of 0.366% by the DG. renewable-energy contribution has a significant impact on The major renewable-energy contribution resulted in the total energy requirement to meet the microgrid energy environmental friendliness especially in healthcare en-vir demand. Fig. 17 portrays the renewable-energy contribu- onments by a reduction in CO by 6 261 132 kg/yr, SO by 2 2 tions by all options except the reference case, since there 27 362 kg/yr and NO by 12 838 kg/yr against the reference option. Among all the options, these are the highest reduc- tion quantities in emissions. The generic 4-hour 1-MW Li-ion BESS has an autonomy of 0.899 hours, a storage wear cost of USD 0.025/kWh and a usable normal capacity of 4216 kWh. It has a lifetime throughput of 4  452  292 kWh and an annual throughput of 296 819 kWh/yr. The energy-content daily profile of the BESS is shown in Fig. 18. Both the DG and BESS coordinate with the renewable sources (WT and PV) in the demand- management events of the microgrid. Whenever there is no renewable-energy generation, the roles played by the DG and BESS are significant. On economic metrics, Option 3 has a return on invest- ment of 9.8%, an internal rate of return of 13.6% and a pay- back period of 6.75 years. Though all the three options have payback periods of <7 years and internal rates of return of ~13.6%, the LCOE is attractive for Option 3 at USD 0.07547/ kWh. Besides, the operating cost is much lower in Option 3 and its annualized cost is the lowest among all the options in spite of the addition of 5.8 MW of renewable-energy Fig. 17: The annual electricity contribution by all the renewable sources in MWh from each option systems. Fig. 18: The daily energy-content profile of the BESS for a 12-month period Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 270 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 19: Utility-grid power outage for 4 hours 4.5 Utility-grid outage the region, the microgrid has been designed to maintain supply reliability even if the utility power outage is up to As the main purpose of the incorporation of the microgrid the maximum sanctioned demand of the microgrid. is to improve the power-supply resiliency, a power-outage case is considered based on the existing state utility grid’s operating conditions. During certain days, there would be a 4.6 Lessons learned and scope for future reduction in consumer demand such as 60%, which means energy fields that the consumer could make use of only 40% of the sanc- The proposed microgrid would be the first attempt at tioned maximum demand. Another prevailing situation is healthcare facilities in India since its first day of operation a 100% planned power cut or power-supply outage for cer - to ensure the availability of electricity. Renewable energy tain hours of a day, such as 13 hours to 17 hours. Under such as solar and wind have been considered for economic that situation, all consumers in that region are in darkness and environmental advantages. Biomass, apart from these unless backup power-supply facilities are available for two, may be the next option in the near future, depending them. The 100% planned outage is considered here as the on the production of waste. From a reliable point of view, unexpected power outage due to the grid’s technical func- more types of energy sources would be beneficial. In add- tional issues for the simulation purpose. ition to battery backup, fuel cells could be preferred for A 100% power outage of the utility grid for 4 hours backup power supply. Although bilateral power and com- during a daytime period has been simulated with the munication facilities are introduced in the name of smart- microgrid functioning (as in Option 2: Utility, DG, PV, BESS) grid technology, it is only the initial stage and, depending without the WTs because of low wind below the minimum on the state-of-the-art technological changes, it needs to in turbine speed. The 100% power outage occurred at 13.00 be modernized. hours. During that period, the maximum demand of AIIMS In this optimization platform, demand management was 7000 kW. The microgrid responded and all the loads has been duly considered without many grid failures (only were served by the battery backup system. Being daytime, one example has been provided). In future work, various there was a PV generation of ~4000 kW; hence, the DG magnitudes of grid failures, scheduled power cuts and the was not triggered. The supply interruption continued for energy harnessed from the hospital non-hazardous waste another hour while the AIIMS demand was ~5500 kW. As shall be considered. there was a reduction in PV generation to 3500 kW, the DG has been activated and the loads were met by the backup supply. Likewise, for the continued power outage until 5 Conclusion 17.00 hour, the generator and the PV generation main- tained the continued power supply through the BESS (Fig. This paper proposed a 20-MVA microgrid for the newly 19). Considering the prevailing power-outage situations of approved 750-bed AIIMS healthcare facility in Madurai. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 271 [4] Buonomano A, Calise F, Ferruzzi G, et al. Dynamic energy per - AIIMS hospitals already in operation, some of which formance analysis: case study for energy efficiency retrofits of had been established several decades before with no hospital buildings. Energy, 2014, 78:555–572. microgrid resources, have now begun to incorporate re- [5] Allcott  H, Collard-Wexler  A, O’Connell  S. How do electricity newable energy due to energy shortages and grid fail- shortages affect industry? Evidence from India. American ures. The authors of this paper considered that it will be Economic Review, 2014, 106:1–45. economical and environmentally feasible to introduce a [6] The Times of India PTI. 24-hour power to all by March 2019. microgrid with a utility-grid interconnection option as https://timesofindia.indiatimes.com/india/24-hour-power- to-all-by-march-2019-govt/article-show/62284777.cms (15 the electricity supply at the planning stage itself to en- December 2020, date last accessed). sure service continuity and reliability. A detailed load as- [7] The New Indian Express, Madhusudhakar K. 20 die during sessment has been carried out considering the number 12-hr power cut at Kurnool Hospital in Andhra Pradesh. of beds, the medical and nursing students, doctors, 2017. https://www.newindianexpress.com/states/andhra- nurses and supporting staff, outpatients and visitors, pradesh/2017/jun/23/20-die-during-12-hr-power-cut-at- critical and non-critical electrical loads, hostel facilities kurnool-hospital-in-andhra-pradesh-1619975.html.%20 (15 and resident staff quarters, etc. The demand forecast re- December 2020, date last accessed). [8] The Times of India, TNN. 21 die in Hyderabad govt Hospital, sulted in 16 000 kVA as the connected load capacity at a staff blame power cut. 2016. https://timesofindia.indiatimes. 0.90 power factor. Allowing space for future expansion, com/india/21-die-in-hyderabad-govt-hospital-staff-blame- 20 MVA has been finalized as the capacity of the pro- power-cut/article-show/53359874.cms (15 December 2020, posed microgrid. date last accessed). Within 2-km distance, the state utility grid is available for [9] Adair-Rohani H, Zukor K, Bonjour S,et  al. Limited electricity interconnection as the primary supply. For accommodating access in health facilities of sub-Saharan Africa: a systematic renewable energy, the locally available natural resources review of data on electricity access, sources, and reliability. Glob Health Sci Pract, 2013;1:249–261. are considered. The wind speed and solar radiation are fa- [10] World Health Organization and World Bank. Access to vourable; hence, 4 MW of PV panels and 1.8 MW of WTs are modern energy services for health facilities in resource- proposed for the microgrid with DG and BESS backup. For constrained settings: a review of status, significance, chal- optimization purposes, the HOMER Grid platform has been lenges and measurement. 2015. https://apps.who.int/ used. Various options of renewable capacities were tested iris/handle/10665/156847 (15 December 2020, date last and, as a result, a combined 5.8-MW capacity of wind and accessed). PV solar generation with 12 MW DG and 1 MW BESS were [11] Bonnema E, Studer D, Parker A, et  al. Large hospital 50% en- ergy savings: technical support document. 2010. https://apps. finalized. On economic metrics, a return on investment who.int/iris/handle/10665/156847 (15 December 2020, date of 9.8%, an internal rate of return of 13.6% and a payback last accessed). period of 6.75  years were achieved besides an attractive [12] Carpenter D, Hoppszallern S. Advancing efficiency: 2011 hos- LCOE as USD 0.07547/kWh. Avoided annual emissions of pital energy management survey. Health Facility Management, 6 261 132 kg of carbon dioxide, 27 362 kg of sulphur dioxide 2011, 24:15–22. and 12  838  kg of nitrogen oxides have been achieved as [13] Kantola M, Saari A. Renewable vs. traditional energy manage- compared to the mere utility-grid power supply. ment solutions—a Finnish hospital facility case. Renewable Energy, 2013, 57:539–545. [14] Modern Healthcare. Hospital shift from fossil fuel to renew- able energy sources. 2014. http://www.modernhealthcare. Acknowledgements com/article/20141108/MAGAZINE/311089981 (15 December C.P. did 60% of the work. P.N. did 40% of the work. 2020, date last accessed). [15] European Commission. Towards zero carbon hospitals with renewable energy systems. 2013. https://ec.europa.eu/energy/ intelligent/projects/en/projects/res-hospitals (15 December Conflict of Interest 2020, date last accessed). None declared. [16] Schneider Electric. A report on energy efficient hospitals— visiting the realities. https://www.cii.in/webcms/Upload/ CII%20Report%20on%20Energy%20Efficient%20Hospitals%20 References %E2%80%93%20visiting%20the%20realities.pdf (15 December [1] CADDET Energy Efficiency Maxi Brochure 5. Saving energy 2020, date last accessed). with energy efficiency in hospitals. 1999. https://www.scribd. [17] Lawrence Berkeley National Laboratory. High performance com/document/176818872/Energy-Efficiency-in-Hospitals- healthcare buildings: a roadmap to improved energy efficiency. Maxi-Brochure-5-CADDET (9 December 2020, date last 2009. https://indoor.lbl.gov/publications/high-performance- accessed). healthcare-buildings (15 December 2020, date last accessed). [2] Sathwik  R, Sayali  S, Vaishnavi  D. Bringing energy efficiency [18] García-Sanz-Calcedo J, Al-Kassir A, Yusaf T. Economic and en- for hospital building through the conservative and preventive vironmental impact of energy saving in healthcare buildings. measures. International Journal of Innovative Technology and Applied Sciences, 2018; 8, 440. Exploring Engineering (IJITEE), 2019, 8:1081219. [19] Schneider Electric. Microgrids: achieving reliable power for [3] Wang  T, Li  X, Liao  PC, et  al. Building energy efficiency for our most critical facilities. 2019. https://microgridknowledge. public hospitals and healthcare facilities in China: barriers com/healthcare-microgrids-reliable-power/ (15 December and drivers. Energy, 2016, 103:588–597. 2020, date last accessed). Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 272 | Clean Energy, 2021, Vol. 5, No. 2 [20] Microgrid Knowledge Editors. Microgrids in hospitals economictimes.indiatimes.com/news/renewable/aiims-to- minimize threats of electrical outages. 2020. https:// go-green-with-solar-power-plant-reduce-power-bill-by-50- microgridknowledge.com/microgrids-in-hospitals/ (15 per-cent/57784132 (15 December 2020, date last accessed). December 2020, date last accessed). [34] The Asian Age. AIIMS to go green, to set up solar power plant. [21] Alessandro  B, Matteo  M, Gabriele  C, et  al. Microgrid design 2017. https://www.asianage.com/metros/delhi/230317/aiims- and operation for sensible loads: Lacor hospital case study in to-go-green-to-set-up-solar-power-plant.html (15 December Uganda. Sustainable Energy Technologies and Assessments, 2019, 2020, date last accessed). 36: 100535. Doi: 10.1016/j.seta.2019.100535. [35] NASA’s Open Data Portal. Prediction of Worldwide Energy [22] Alexis L, Miguelde SM, Alberto GM, et  al. Sustainable microgrids Resources (POWER). https://data.nasa.gov/Earth-Science/ with energy storage as a means to increase power resilience Prediction-Of-Worldwide-Energy-Resources-POWER-/wn3p- in critical facilities: an application to a hospital. International qsan (15 December 2020, date last accessed). Journal of Electrical Power & Energy Systems, 2020, 119: 105865. [36] NREL, US Department of Energy. National renewable energy [23] Arun Kumar V, Verma A. Optimization and implementation of laboratory database. https://www.nrel.gov/research/data- hybrid energy sources in remote and grid active microgrids-a tools.html (15 December 2020, date last accessed). case study for Indian Scenario 2016. In: IEEE 7th Power [37] National Building Code of India 2016. Standards & Codes. India International Conference (PIICON), Bikaner, India, 25–27 https://bis.gov.in/index.php/standards/technical-department/ November 2016, 1–6. national-building-code/ (15 December 2020, date last accessed). [24] Arun K, Ashu V, Rajbans T. Optimal techno-economic sizing of [38] Energy Conservation Building Codes. Energy Conservation a multi-generation microgrid system with reduced depend- Building Codes 2017. https://beeindia.gov.in/sites/default/ ency on grid for critical health-care, educational and indus- files/BEE_ECBC%202017.pdf (15 December 2020, date last trial facilities. Energy, 2020, 208, 118248. accessed). [25] Margaret  AW, Rajini  V. Cost benefit and technical analysis [39] TANGEDCOUP. Generation data. https://www.tangedco.gov. of rural electrification alternatives in southern India using in/generation.html (15 December 2020, date last accessed). HOMER. Renewable and Sustainable Energy Reviews, 2016, [40] Alberto B, Idoia SM, Pablo S, et al. Lithium-ion batteries as dis- 62:236–246. tributed energy storage systems for microgrids. In: Academic [26] Asian Development Bank. Handbook on Microgrids for Power Press. Distributed Energy Resources in Microgrids. Amsterdam: Quality and Connectivity. 2020. https://www.adb.org/docu- Elsevier Science, 2019, 143–183. ments/handbook-microgrids-power-quality-connectivity (15 [41] PV Software. Simulation and design of solar systems. https:// December 2020, date last accessed). photovoltaic-software.com/ (15 December 2020, date last [27] Xu  D, Long  Y. The impact of government subsidy on renew- accessed). able microgrid investment considering double externalities. [42] Palanichamy C, Sundar Babu N, Nadarajan C. Privatizing and Sustainability, 2019, 11:3168. restructuring Indian power sector: an overview. Journal of the [28] Darrell Proctor. Micro generation with macro possibilities. Institution of Engineers, 1999, 80:23–30. 2020. https://www.powermag.com/micro-generation-with- [43] Palanichamy C, Sundar Babu N, Nadarajan C. Renewable en- macro-possibilities/ (15 December 2020, date last accessed). ergy investment opportunities in Mauritius—an investor’s [29] California Energy Commission. A novel, renewable energy perspective. Renewable Energy, 2004, 29:703–716. microgrid for a California healthcare facility. April 2019. [44] Lawrence Berkeley National Laboratory (LBNL). DER-CAM https://ww2.energy.ca.gov/2019publications/CEC-500-2019- User Manual 2015. https://gridintegration.lbl.gov/der-cam (15 034/CEC-500-2019-034.pdf (15 December 2020, date last December 2020, date last accessed). accessed). [45] Robinson  D, Haldi  F, Kämpf  J, et  al. CitySim: comprehen- [30] Ministry of Health & Family Welfare Government of India. No. sive micro-simulation of resource flows for sustainable MoH & FW /Z-28016/168/2018-PMSSY-III/ RFP /1/ 2019. http:// urban planning. In: Proceedings of the 11th International IBPSA pmssy-mohfw.nic.in/files/tender/Reply%20to%20the%20 Conference, Glasgow, Scotland, UK, 2009, 1083–1090. Queries%20in%20Pre-Bid%20Meeting.pdf (15 December 2020, [46] Mendes  G, Ioakimidis  C, Ferrão  P. On the planning and ana- date last accessed). lysis of integrated community energy systems: a review and [31] Ministry of Education. National Institutional Ranking Framework survey of available tools. Renewable and Sustainable Energy 2020. https://www.nirfindia.org/2019/MEDICALRanking.html Reviews, 2011, 15:4836–4854. (15 December 2020, date last accessed). [47] Homer Energy. HOMER Energy—the microgrid software. [32] All India Institute of Medical Sciences, New Delhi. Top uni- https://www.homerenergy.com/products/software.html (15 versities 2014. https://www.topuniversities.com/subject- December 2020, date last accessed). rankings/2020 (15 December 2020, date last accessed). [48] Lambert T, Gilman P, Lilienthal Micr P. o power System Modelling [33] The Economic Times. AIIMS to go green with solar power with HOMER: In Integration of Alternative Sources of Energy. plant, reduce power bill by 50 per cent. 2017. https://energy. Hoboken, NJ: John Wiley & Sons, 2006. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Clean Energy Oxford University Press

Micro grid for All India Institute of Medical Sciences, Madurai

Clean Energy , Volume 5 (2) – Jun 1, 2021

Loading next page...
 
/lp/oxford-university-press/micro-grid-for-all-india-institute-of-medical-sciences-madurai-KiSQgKYtpq
Publisher
Oxford University Press
Copyright
Copyright © 2021 National Institute of Clean-and-Low-Carbon Energy
ISSN
2515-4230
eISSN
2515-396X
DOI
10.1093/ce/zkab007
Publisher site
See Article on Publisher Site

Abstract

Keywords: new healthcare facilities; load assessment; microgrid proposal; optimization Received: 15 December 2020; Accepted: 10 March 2021 © The Author(s) 2021. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (http:// creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 255 These are not just energy challenges for healthcare Introduction systems in India; power shortages in a number of coun- Hospitals and healthcare facilities are the foremost tries are still a barrier to effective health services. Energy energy-intensive consumers of the commercial sector issues for hospitals vary in low- and high-income coun- and, as vital components of the delivery system of med- tries. The availability of sustainable energy for common ical services, they are liable for a considerable portion medical needs is the key challenge for health facilities in of total commercial energy consumption in India [ , 1 2]. low-income countries, especially those in sub-Saharan For effective functionality, perpetual operations and Africa [9]. Many healthcare facilities in these countries are state-of-the-art medical equipment, most hospitals must facing a shortage of electricity for critical services, such as have a continuous power supply. Globally, they are under lighting, heating and medical devices. This would lead to tumultuous pressure to achieve more while controlling decreased diagnostic capabilities and treatment facilities, costs and limiting waste, and to offer enhanced patient shortened working hours during the day and shortages consideration [3 , 4]. of committed healthcare professionals due to job dissat- In India, many healthcare facilities are located near elec- isfaction [10]. On the other hand, high-income countries trical grid networks as the primary power-supply resource. are focusing on improving productivity and the practice of Also, power failures and supply interruptions during high- renewable energy to minimize energy use, reduce running demand periods have been problematic in grid-connected costs and environmental impacts [11–15]. cities and regions [5]. In one major event, the biggest grid In the face of rising energy costs, rising demand, the failure in history, over 30 and 31 July 2012 in the northern escalating role of power-dependent technology in med- part of India, massive load demand, poor supply manage- ical care and rising numbers of aggressive storms all over ment and interruptions in transmission services led to a the world, the issue of resilience keeps several healthcare recurrent power-grid breakdown that affected 650 million managers up at all times [16–18]. Thus, the acquisition and people across the country, home to half of India’s popula- appropriate use of electrical energy is a pivotal step for tion. By March 2019, the Indian government had granted any intricate structure that needs to reach the ideal energy USD 2.5 billion to provide electricity to all households [6]; level. Particularly for healthcare facilities, there should be access to affordable electricity remains a major concern for a high level of concern about energy supply and mainten- Indian health services. Even after the remedial measures, ance, as they are the most vital facilities for continuing people were unable to ignore the historic grid breakdown primary care and emergency care. Since recent incidents and blamed it for the series of major longevity instances have led to power outages, there is no bias that puts hos- around the country, most of which were triggered by the pitals and their patients at serious risk. Hospitals are ex- breakdown of life-sustaining equipment such as ventila- ploring innovative energy models and highly developed, tors and incubators. reliable and cost-effective technologies such as microgrids Twenty patients died along with eight children, and to address these critical energy concerns [19]. many others had a tormenting experience when the While monitoring and managing energy supply and de- Kurnool Government General Hospital went into dark- mand, a microgrid will incorporate and regulate a broad ness for 12 hours after the power supply had been shut range of energy sources [20–23]. As a result, in an emer - down from 7  p.m. on Thursday, 22 June 2017 to the fol- gency situation, the microgrid will be able to supply elec- lowing morning [7]. More horrible still, no senior spe- tricity to the hospital whenever the adjacent grid breaks cialist, not even the Resident Medical Officer, was in the down. Whereas the utility is not bound to control black- emergency clinic throughout the blackout. Of the 20 pa- outs, the microgrid works 24 hours a day and throughout tients reported to have passed on, 4 were ladies. A  se- the year [24–26]. The microgrid is constantly on the job, un- nior specialist admitted that power disturbance caused like the backup generators that need to be kept under con- the vast majority of deaths. As per source, power was trol to ensure they can function in a crisis. In addition, if switched off when police outpost personnel snarled elec- not required for backup purposes, the microgrid allows the trical connections to high-tension wires on clinic prem- healthcare facility to achieve its cost and sustainability ob- ises to draw power illegally. jectives through demand–response, peak shaving and load On 24 July 2016, 21 patients died from power outages at management, and excess-energy sales to the grid other the state-run Gandhi Hospital, the most prominent 1200- than the use of on-site generators and energy-storage re- bed hospital in the city of Hyderabad [8]. A few specialists sources when electricity rates are high on the grid [27, 28]. said that power had initially stumbled at around 3  p.m. Microgrid technology is very effective, as can be seen and then continued to do so on a regular basis. In spite of from the installation of California’s leading large-scale the fact that there were four generators on standby, the microgrid for healthcare services. In 2018, at the Kaiser emergency clinic guaranteed that they had peak power Permanente facility in Richmond, California, the California generation and could not be used to identify the explan- Energy Commission issued a USD 4.78 million grant to ation for the stumbling after electrical cables were later Charge Bliss, Inc. to develop, build and run the foremost cut off. Deaths occurred in specialty clinics, as well as in renewable-energy-based microgrid for a healthcare hos- the Surgical Intensive Care Unit and the Emergency Ward. pital in California [29]. The hospital was the only public Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 256 | Clean Energy, 2021, Vol. 5, No. 2 hospital serving western Contra Costa County to provide based on the smart-grid concept of providing two-way the local population with critical care. In addition, the area power and communication facilities to enhance the reli- was affected by a greater degree of environmental con- ability of the power supply at all times. Of course, due to tamination and subsequent health consequences. A  new smart-grid technology, the operating and maintenance microgrid controller was also developed for the project (O&M) cost of AIIMS is slightly higher than that of con- that will isolate the hospital’s life-safety emergency-power ventional microgrid systems. division, including emergency lighting and exit signs, and provide emergency power services. The microgrid resulted in an annual output of 395 000 kWh of renewable energy, 1 System modelling a 4-hour reduction in demand of 220 kW from the battery AIIMS Madurai will have a hospital capacity of 750 beds and a reduction in emissions of 739 metric tons of CO /yr. with an interim departmental distribution of beds as The Ministry of Health & Family Welfare, Government shown in Table 1. of India, approved a project for the development of a new All India Institute of Medical Sciences (AIIMS) in Madurai, Tamil Nadu, with a budget of USD 178 mil- 1.1 Site location lion in 2018 [30]. AIIMS includes the establishment of hospitals, trauma-centre facilities, medical colleges, Thoppur village, Madurai-625008, is the proposed location residential complexes and related facilities/services, for the new AIIMS. The village of Thoppur is located in the mostly based on AIIMS, New Delhi, and six proposed south of Madurai, Tamil Nadu State, India. It is located 20 new AIIMS. The aim is to establish the new AIIMS as a km from Madurai South Sub-district Headquarters and nationally significant educational institution to provide 20 km from Madurai District Headquarters. As per the quality medical education, as well as tertiary healthcare Global Positioning System, its coordinates are 9° 52.3’ N facilities for the people of Madurai and the surrounding (latitude) and 78° 1.5’ E (longitude), respectively. Fig. 1 il- area. The main purpose of this paper is to provide a re- lustrates the site of the planned Thoppur AIIMS, Madurai. liable, economical and efficient power supply based on The annual average air temperature over a 30-year period a microgrid for the new AIIMS Madurai. It is the right (January 1984 to December 2013) was found to be 27.34°C time to propose the inclusion of the microgrid, because and the annual average wind speed at 50 m above the sur - the project is in the tendering stage. The greatest sig- face of Earth over a 30-year period was 4.97 m/s, according nificance of this suggestion is that AIIMS New Delhi (es- to NASA’s Prediction of the Worldwide Energy Resource tablished in 1956) ranked first in India by the National (POWER) Database [35]. The solar radiation and the clear - Institutional Ranking Framework [31], ranked first in ness index of the proposed site as per the US National South Asia and 231st in the Life Sciences and Medicine Renewable Energy Laboratory database [36] are shown in category worldwide by QS WUR [32], recognized and Fig. 2 and the annual average irradiance value is 5.61 kWh/ introduced the value of the microgrid only in 2017 [33, m /day. 34]. The novelty of the proposed microgrid is that it was suggested at the beginning of the approval stage of the Table 1: Department-wise bed distribution for the proposed AIIMS Madurai itself during the preparation of the in- AIIMS healthcare facility in Madurai frastructure. Recently, the existing AIIMS in India began planning renewable-energy systems with PV systems Departments Beds only a decade or a few decades after its inception. a) Speciality Departments 360 Having PV alone is unreliable and, in countries like India (Surgical & Allied Specialities, Medicine & Allied with four seasons, operating efficiency is <14%, resulting Specialities, and Obstetrics & Gynaecology) in a higher unit cost of electricity generation. Moreover, b) Super Speciality Departments 215 because PV systems need vast land, the construction of (Cardiology, Cardio-thoracic Vascular solar farms without public support is not straightfor - Surgery-CTVS, Gastroenterology, Surgical ward. In light of all these issues, a wind turbine (WT), Gastroenterology, Nephrology, Urology, as another renewable-energy system, has been included Neurology, Neuro-surgery, Paediatric Surgery, in this project, as the wind profile in the proposed area Burns & Plastic Surgery, Medical oncology, Surgical Oncology, Radiation Oncology, is favourable. At this site, a 1-kW WT with a height of 30 Endocrinology and Pulmonary Medicine) m will produce an energy output of 1750 kWh/yr with c) Other facilities 175 an attractive payback period of <7  years. The proposed (Intensive Care Units (ICUs) & Critical Care, microgrid is such that >50% of its maximum demand Trauma, AYUSH Facilities, PMR Department and could be met by both PV and wind turbines, thus redu- Paid Beds) cing dependence on the utility grid, apart from environ- Total number of beds 750 mental credits. In addition, the proposed microgrid is Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 257 Fig. 1: The approved and inaugurated location for AIIMS in Thoppur Madurai. This figure was used with permission from [47]. Fig. 2: The solar radiation and the clearness index of the proposed site as per the US National Renewable Energy Laboratory database [36] devices with lower energy consumption are chosen. LED 1.2 Demand assessment lighting fixtures with built-in harmonic-suppression sys- One of the key tasks for the energy management of the tems are considered for lighting in all areas and buildings new system is the assessment of demand. Energy con- in accordance with the National Building Code (NBC) 2016 sumption depends on activity; climate and changes in oc- [37], the Energy Conservation Building Code (ECBC) [38] and cupancy, at different times of the day; days of the week; the Indian Standard Code. The demand assessment has weather conditions; and seasons. India has four seasons been estimated on the basis of the covered area of various and this is properly considered in the assessment of de- buildings/blocks as per NBC 2016 considering the lighting mand. Besides, the hospital loads are classified as critical load as 13 W per square metre and power load as 55 W per and non-critical loads, depending on their importance, and square metre minimum. A 10% load was considered for fu- this type of classification has been done during demand ture expansion. Table 2 shows the different loads and the assessment in this project to give priority for availability respective locations and the quantity of the total amount duration. A load diversity varying from 0.5 to 1.0 has been of the connected load requirement in addition to the total considered. All medical devices, such as ventilators, anaes- demand for kVA at a power factor of 0.9. thesia machines, blood and infusion heaters, MRIs, x-rays and computed tomography scan (CTS), etc., require steady and good-quality electricity supplies for their operation 1.3 Utility requirement and are considered to be critical loads, while hospital loads, Tamil Nadu Generation and Distribution Corporation such as heating, ventilating, and air conditioning (HVAC) Limited, TANGEDCOUP [39] under state government con- systems and water heaters, etc., are grouped together as trol has a generation capacity of 18 747.28 MW consisting non-critical loads. Modern and sophisticated medical Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 258 | Clean Energy, 2021, Vol. 5, No. 2 Table 2: Category-wise (power, lighting and residential loads, and total load) demand assessment of AIIMS Madurai a) Power loads Load and location kW Load and location kW OPD & Diagnostic Block 550 Night shelter 135 Medical equipment 905 Auditorium 98 Hospital Block (720 beds) 2320 Fire station 87 Ayush Hospital (30 beds) 115 Dining hall 75 Medical & Nursing College 1040 Kitchen equipment 44 HVAC 1925 Director’s residence and servant quarters 40 Lifts (56 nos) 710 Guest house 38 Plumbing and pump load 560 Mortuary 32 Apartment blocks 885 Waste-management block 32 PG hostel blocks 802 Laundry 32 Working nurses’ hostel 420 Shopping centre 22 Boys’ and girls’ hostels 340 Cafeteria 14 Total load in kW 11 221 Load in kVA at 0.90 power factor 12 468 b) Lighting loads Load and location kW Load and location kW OPD & Diagnostic Block 139 Fire station 7 Hospital Block (720 beds) 624 Dining hall 19 Ayush Hospital (30 beds) 30 Kitchen 24 Medical & Nursing College 267 Director’s residence and servant quarters 10 Plumbing and pumping house 7 Guest house 11 Apartment blocks 230 Mortuary 11 PG hostel blocks 207 Waste-management block 15 Working nurses’ hostel 111 Laundry 11 Boys’ and girls’ hostels 85 Shopping centre 11 Sports-field lighting 93 External lighting 11 Night shelter 37 Cafeteria 7 Auditorium 81 Total load in kW 2048 Load in kVA at 0.90 power factor 2276 c) Lighting and power loads of residential blocks Residential loads Lighting and power loads in kW 1131 Lighting and power loads in kVA at 0.90 power factor 1257 d) Total load in kW and kVA Total connected loads Lighting and power connected loads in kW 14 400 Lighting and power loads in kVA at 0.90 power factor 16 001 1.4 Backup power supply of TANGEDCO state-owned ventures, Central Generating In the event of power failure on the part of the utility, Stations shares and Private Power Purchase. In addition, diesel-generator (DG) sets are recommended for backup. the state has projects of up to 1 047 961 MW in renewable- The main concern is that the storage of diesel fuel should energy sources such as wind, solar, biomass and cogener - be there at all times. The main fuel-storage tanks must be ation. It will supply electricity to meet the electrical load tracked and refilled when they run down. Automated con- requirements of AIIMS, Madurai, from its nearby 110/11-kV trol systems are available to inform the generator operator substation located in Kappalur, Madurai, at a distance of <2 when the level of the tank goes below the specified level. km from AIIMS. AIIMS Madurai will have its own 20-MVA The capacity of the backup power supply has been fixed distribution substation with 2  × 10 MVA, 11/0.415-KV dis- on the basis of the maximum demand. For the connected tribution transformers. The utility has a commercial tariff load of 14  400 kW, taking into account the average div-er of USD 0.085/kWh and USD 5/kW. The buy-back tariff for sity factor of 0.8, the maximum demand is 11 520 kW. As renewable-energy supplies is USD 0.062/kWh, so that con- a result, a backup power supply of 12 000 kVA for DGs has sumers can sell their excess energy from renewable energy been recommended. It is advisable to use a large number to the grid. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 259 of generators rather than a single large-capacity generator and Nursing College, the Ayush Block, the auditorium and in terms of reliability, operating efficiency and economics. the Computer Centre. It is recommended to use a 4-hour, The most significant locations to be provided with 100% 1-MWh capacity Li-ion battery [40]. It offers outstanding backup power are: functionality for microgrid implementation. Depending on the size of the microgrid, the Li-ion battery can be used as an energy-storage system due to its extended size range. OPD & Diagnostic Block, Hospital Block (720 Full backup Furthermore, its falling price and the recovery of perform- beds), Ayush Hospital (30 beds), Medical power & Nursing College, HVAC loads, plumbing supply ance and lifetime enhance the applicability of this Li-ion and pump loads, all chillers and all AHUs, battery. fire station, mortuary, laundry, auditorium, guest house, dining, shopping complex and director’s residence 1.6 Renewable-energy options Renewable energy is an attractive choice worth exploring, For providing a 100% backup supply, four 2000-kVA (4  × depending on availability, accessibility and cost, policies 2000-kVA) generators are suitable and the strategic loca- and incentives, and pricing and regulations for electri- tion for their placement is near the Hospital Block. city. A variety of renewable options are worth considering: Other facilities and loads are provided with backup solar, wind and biomass resources. This initiative currently power of four 1000-kVA (4 × 1000-kVA) DGs, including: lifts targets both solar and wind energy systems. (56 total); apartment blocks; PG hostel blocks; working At the location of AIIMS Thoppur, Madurai (latitude: 9° nurses’ hostel; boys’ and girls’ hostels; night shelter; 52.3’ N and longitude: 78° 1.5’ E), the annual average solar sports-field lighting; waste-management block; cafeteria; radiation is found to be 5.61 kWh/m/day. Fig. 2 shows that, residential blocks; lighting and fan loads; water and fire- regardless of the four seasons, solar radiation is available fighting pumps; street lighting; and emergency services. throughout the year. The average annual radiation lasts for The location for the DGs is near the sports field. All DGs 5 months (42%) per year and the maximum radiation takes are of an outdoor type with hospital-type silencers and place for a period of 3 months (25%). acoustic enclosures. The total capacity of the DG backup The universal method for determining the electricity power is therefore 12 000 kVA. generated by a photovoltaic system [41] is as follows: (2) E = A ∗ r ∗ H ∗ PR kWh where E represents the energy produced (kWh), A repre - 1.5 Energy storage sents the total area of the solar panel (m ), r represents the Energy storage is quickly moving forward and becoming solar-panel yield or efficiency (%), H represents the annual a key player in the future of microgrids. In the context of average solar radiation on tilted panels without shading the hybrid renewable-energy concept for the healthcare and PR represents the performance ratio, coefficient for microgrid, the integration of wind and solar energy be- losses (between 0.5 and 0.9, default value = 0.75). comes inevitable. As both wind and solar have variable The site is lucrative so that, with a payback period of outputs, storage technologies have an immense potential <7  years, a 1-m PV panel will generate ~2050 kWh/yr. In to smooth the supply of energy from these sources. This view of the yield and the rate of return, a 4-MW PV system storage option helps operators to dynamically balance en- with a capacity of ~35% of AIIMS Madurai’s maximum de- ergy on the grid by alternatively injecting and consuming mand has been proposed. While fixing the capacity, the excess electricity. Without a balance of supply and de- annual solar variation has been duly considered and the mand, the stability of microgrids becomes questionable. backup power supply has been fixed with a diversity factor Various energy-storage technologies contribute to stability of 0.8. by operating at different stages of the microgrid from gen- The AIIMS site has an average annual wind speed of eration to end use by customers. Battery storage is prefer - 4.97 m/s at an anemometer height of 30 m. It lasts for 4380 able in hospitals due to portability, flexibility and economic hours per annum, as seen in Fig. 3. The power output of a concerns. The capacity of a storage plant is given as the WT is given by the expression [42, 43]: following expression [40]: (3) C =(E ∗ AD) / (η ∗ η ∗ DOD) Wh(1) P = 1/ ∗ ρ ∗ A ∗ v ∗ C Watts Wh L inv bat T 2 p where P represents the power output (W), ρ represents the where C represents the battery capacity (Wh), rep E - Wh L T air density, 1.22 kg/m, A represents the turbine swept area resents the average energy load per day, AD represents (m ), v represents the wind speed (m/s) and C represents is the number of autonomous days of the battery, ɳ inv p the performance coefficient of the WT. represents the efficiency of the inverter, ɳ represents bat At this site, a 1-kW WT with a height of 30 m will the efficiency of the battery and DOD represents the produce 1750 kWh/yr of energy. Higher WT-generator battery-discharge depth. capacity with higher hub height is an efficient and eco- Battery backup is provided for operation theatres, essen- nomical option for better yield. An attractive payback tial loads and medical equipment of the Hospital, Medical Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 260 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 3: Annual average wind speed at the AIIMS Thoppur Madurai site. This figure was used with permission from [47]. period of <7 years is feasible with 900-kW capacity WTs. Java-based graphical user interface that supports the Two such WTs (2 × 900 kW) are proposed for this AIIMS decision-making function of surviving urban develop- project. Including PV solar and wind power, the contri- ment. The objective of this instrument is to simulate and bution of renewable energy to the maximum demand of optimize the flow of building-related resources (energy, AIIMS Madurai will be 50.35%. water and waste) and their inter-relationships, as well as Since weather (solar irradiance, wind) varies from year to investigate their dependence on urban climates. to year, a slightly higher capacity to cope with the rise and fall of generation due to year-to-year wind and solar irradi- ance was considered with a diversity factor of 0.8 when 2.3  EAM designing the power-generation capacity of the microgrid. EAM is used to evaluate the economic viability of microgrids [46]. It is capable of optimizing the capacity of microgrids through the proper choice of equipment 2 Optimization platform selection, power rating, capital and running costs, and As the AIIMS Microgrid infrastructure has been finalized, lifetime equipment. There is provision for a comparison the next step is to evaluate its performance and whether of the optimized cost of the microgrid against the energy or not it can meet the energy requirements of AIIMS cost of the utility. Madurai without sacrificing its reliability. To do so, an on- line platform becomes essential to achieve the objective economically. Computational approaches are usually used 2.4 HOMER Grid to model, test and forecast energy efficiency in the de- Based on software developed by the National Laboratory sign of microgrid projects. There are a number of software for Renewable Energy (NREL) [47], HOMER Grid is designed packages available for the optimization mission and it is for the economic and engineering assessment of grid- wise to select the most flexible one. connected and off-grid energy systems. The key capability of HOMER Grid is to simulate the efficiency of any specific configuration of the energy system. However, on specified 2.1  DER-CAM systems, the programme is also adequately capable of run- DER-CAM was built by Lawrence Berkeley National Lab, ning economic optimization and sensitivity/uncertainty Berkeley, California, USA [44]. It is a decision-support device analysis. It is important to note that the optimization is for decentralized power systems. The important objective performed on parameters that the designer has control of DER-CAM is to conduct a techno-economic comparison over. Sensitivity analysis is based on factors that are sub- of distinct on-site electricity-generation technologies or ject to uncertainty or shift that are beyond the control of microgrids and to optimize the costs of on-site electricity the designer, like wind speed and the price of fossil fuel. generation using linear programming techniques. HOMER Grid input data include customer load profiles for electrical and thermal energy, any resources and fuel used by the electrical and thermal power-generation sys- 2.2  CitySim tems, energy-system components, electrical and thermal Built by the Laboratory of Solar Energy and Physics load curves with a resolution of ≤1 minute, technical ef- Buildings (LESO-PB), Switzerland [45], CitySim has a ficiency, O&M costs, emission limitations and sensitivity Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 261 parameters. HOMER Grid outputs the results of the assess- For AIIMS Madurai, demand is estimated based on data ment and analysis in the graph format and in the compre- from other AIIMS with the same bed ability and similar hensive data reports. location status. Taking into account the number of beds; Based on the extended facilities, the accuracy of the re- the number of medical and nursing students; the likely sults and the ease of availability, the HOMER Grid platform number of outpatients every day; the number of phys- was preferred for the design of the microgrid for AIIMS icians, nurses and support staff; the number of hostel Thoppur, Madurai. inmates; and the number of campus residents, the head count is ~6200 people. The hourly load for the approximate head counts was reached on the basis of the same refer - 3 AIIMS Madurai Microgrid ence hospitals, as in Table 3. Unlike standard commercial loads, for 9–18 hours a day, HOMER Grid, the identified optimization platform, performs the hourly load stays at its peak for 10 hours. The min- simulation calculations on an hourly basis throughout imum load occurs during the early morning and late night, 8760 hours of the year [48]. Based on the energy balances, when there will be no outpatients and most loads will the estimations compare the thermal and electrical en- be working at less than critical loads; lifts; and medical ergy both supplied and required by the system. Based on equipment. the energy balances, the estimations compare the thermal The HOMER Grid database for electricity consump- and electrical energy both supplied and required by the tion provided the seasonal and yearly load profiles for system. HOMER Grid decides on the optimal configuration AIIMS Madurai Microgrid as in Figs 4 and 5 . The most of the system as well as analyses of the cost of the system noteworthy month-to-month load utilization would be in on the basis of these estimates. There was a significant August, equal to 11 017.9 kW, and the load factor is found assumption that the energy efficiency of the storage bat- to be 0.43. tery remains constant over the life of the battery, while, depending on the battery power and the ambient tempera- ture, it varies, among other factors. Furthermore, the nom- 3.2 Microgrid architecture inal voltage, the power curve, the lifetime curve and the Fig. 6 depicts the proposed architecture for the microgrid minimum charge state of the battery are considered con- with its primary-load requirement. It includes the utility stant. An interest rate of 6% per year and a 25-year project grid as the main electricity supply with a simple tariff, a lifespan are assumed in the computation. backup DG, battery energy-storage system (BESS), two WTs each of 900-kW capacity, three PV solar systems (2  × 1000 3.1 Daily, seasonal and yearly load profiles kW + 1 × 2000 kW) with a total capacity of 4 MW and a con- The actual daily-load demand is uncertain but predictable verter. The WTs and the PV solar systems have their own because AIIMS Madurai is in the development process. independent converters. Apart from that, being a critical Table 3: Estimated daily-load profile of AIIMS Madurai Hour and load in kW Hour 0 1 2 3 4 5 6 7 8 9 10 11 Load, kW 2910 2910 2915 2915 2915 3015 3230 4314 5380 6111 6111 6111 Hour 12 13 14 15 16 17 18 19 20 21 22 23 Load, kW 6111 6111 6111 6111 6111 6111 6111 5901 5190 3950 2935 2910 Fig. 4: The seasonal load profile of AIIMS Madurai Microgrid. This figure was used with permission from [47]. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 262 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 5: The yearly load profile of AIIMS Madurai Microgrid. This figur e was used with permission from [47]. has the flexibility and suitability to simulate future design configurations in the name of the options: Reference: Only the utility grid Option 1: Utility grid, generator backup and WT with bat- tery storage system (Utility, DG, WT, BESS); Option 2: Utility grid, generator backup and PV solar with battery storage system (Utility, DG, PV, BESS); Option 3: All—utility grid, generator backup, WT, PV solar with battery storage system (Utility, DG, WT, PV, BESS). The output performances of all optimization options are compared against the reference on the basis of (i) annu- alized cost; (ii) electricity production, consumption and sales; (iii) economic; (iv) annual utility bill; (v) fuel con- sumption; and (vi) emission release. 4 Results and discussions 4.1 Reference option: only with the utility grid The primary input is the primary-load data from Table 3 and the power-grid tariff as USD 0.085/kWh for electricity and USD 5/kW for demand in order to perform this opti- mization. The energy buy-back tariff is USD 0.062/kWh for excess-energy sales. Output optimization yields: Fig. 6: The proposed architecture of the AIIMS Microgrid with its primary- load requirement. This figure was used with permission from [47]. Total electricity production (kWh): 41 062 500 Total net present cost (NPC) (USD): 52 957 160.61 healthcare facility, to highlight the role of a converter in LCOE (USD/kWh): 0.09976 the microgrid, a separate converter has been included as a Consumption charge (USD): 3 490 312.50 reliability concern. Demand charge (USD): 606 155.71 Apart from the utility tariff, the appropriate initial cap- CO emission (kg/yr): 52 957 160 ital cost, replacement cost and operation and maintenance Carbon monoxide (kg/yr): 0.0 costs of the DG, renewable-energy sources, storage battery Unburned hydrocarbons (kg/yr): 0.0 and the converter are given in Table 4, which are the inputs Particulate matter (kg/yr): 0.0 for the optimization. SO emission (kg/yr): 112 511 NO (kg/yr): 55 024 3.3 Optimization options The optimization indicates the monthly need for grid electricity to satisfy the load demand of the microgrid, as Microgrid optimization has been carried out with different shown in Fig. 7, which incurred an annual USD 4 096 468.21 combinations (or so-called options here) of integrating re- utility bill. newable and backup generation with BESSs. HOMER Grid Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 263 Table 4: Optimization input data for DG, WT, PV, BESS and converter Component costs Component Initial capital Replacement O&M Diesel generator (DG) (Auto size Genset., 12 000 kW) USD 750/kW USD 500/kW USD 0.01/ operating hour Wind turbine (WT), 900 kW (Enercon E-44, 55-m hub height) USD 749 700 USD 749 700 USD 36 000/yr WT, 900 kW (1) (Enercon E-44, 55-m hub height) USD 749 700 USD 749 700 USD 36 000/yr PV panel, 1 MW (ABB PVS 800–1000) USD 625 000 USD 625 000 USD 25 000/yr PV panel, 1 MW (1) (ABB PVS 800–1000) USD 625 000 USD 625 000 USD 25 000/yr PV panel, 2 MW (ABB PVS 980–2000) USD 1250 000 USD 1 250 000 USD 50 000/yr BESS (battery) storage, 7030 Ah (generic 4-hour 1-MW Li-ion, USD 500 000 USD 500 000 USD 5000/yr 90% roundtrip efficiency) Converter (generic system converter, inverter 2880 kW and USD 300/kW USD 300/kW – rectifier 2880 kW) Utility tariff Energy: USD 0.085/kWh Demand: USD 5/kW Energy buy-back: USD 0.062/kWh Fig. 7: The monthly utility-grid electricity requirement to meet the load demand of AIIMS Madurai 4.2 Option 1: Utility, DG, WT, BESS Fig. 8 depicts the microgrid architecture for Option 1 con- sisting of the 12  000-kW DG, two WTs each of 900-kW capacity, a generic 4-hour 1-MW Li-ion kW battery and the 2880-kW generic system converter. All are integrated with the utility-grid supply system; their relevant data are obtained from the HOMER Grid database and the load and utility-grid data are already available in the optimization platform. The outcomes of the Option 1 optimization are listed in Tables 5–12 with the reference-option comparison. According to the findings, the DG set had a total fuel consumption of 38  054  L for 32 hours and contributed 137 392 kWh/yr of electricity, with an average electrical output of 4294 kW. Its contribution to the overall output of electricity under Option 1 is just 0.335%. This is be- cause of the major contribution from the wind turbines and the BESS even during demand-management situ- ations to reduce the emission contribution by the DG set. The generated power output per year indicating the minimum (3000 kW), mean and the maximum (7695 kW) values of the DG are shown in Fig. 9. The specific fuel consumption works out to be 0.277  L/kWh and a mar - Fig. 8: The microgrid architecture for Option 1 consisting of Utility, DG, WT, BESS with converter.This figur e was used with permission ginal generation cost of USD 0.236/kWh at a mean elec- from [47]. trical efficiency of 36.7%. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 264 | Clean Energy, 2021, Vol. 5, No. 2 Table 5: Optimization of Options 1, 2 and 3—comparison of annualized costs Option 1 (Utility , Option 2 (Utility , Option 3 (Utility, DG, Category DG, WT, BESS) DG, PV, BESS) WT, PV, BESS) Capital cost (USD) 812 174.56 996 169.44 1 108 159.48 Replacement cost (USD) 36 976.92 45 225.08 80 506.98 Operation and maintenance cost (USD) 2 578 060.98 2 210 797.12 1 998 367.91 Fuel cost (USD) 38 053.99 43 163.71 43 973.54 Salvage (USD) –126 094.08 –111 543.33 –131 692.56 Total cost (USD) 3 339 172.37 3 183 812.01 3 099 315.34 Table 6: Optimization of Options 1, 2 and 3—comparison of electricity production, consumption and sales Option 1 (Utility, DG, WT, Option 2 (Utility, DG, PV, Option 3 (Utility, DG, WT, Category BESS) BESS) PV, BESS) Electricity production kWh/yr % kWh/yr % kWh/yr % DG 137 392 0.335 150 094 0.362 152 034 0.366 WT 900 1 575 747 3.84 -- -- 1 575 747 3.84 WT 900 (1) 1 575 747 3.84 -- -- 1 575 747 3.84 PV 1000 -- -- 1 801 544 4.34 1 801 544 4.34 PV 1000 (1) -- -- 1 801 544 4.34 1 801 544 4.34 PV 2000 -- -- 3 610 387 8.68 3 610 387 8.68 Grid purchases 37 774 048 91.985 34 142 167 82.28 30 973 516 74.594 Total 41 062 934 100 41 505 737 100 41 490 520 100 Electricity consumption and sales AC primary load 41 062 500 100 41 062 500 100 41 062 500 100 Grid sales 434 0.001 1462 0.0036 5471 0.0133 Total 41 062 934 100 41 063.962 100 41 067 971 100 Excess energy 0.0 0.0 58 159 0.14 27 885 0.0672 Renewable energy RE fraction 7.68% 17.36% 25.04% Max. RE penetration 119% 175% 212% Table 7: Optimization of Options 1, 2 and 3—economic comparisons Option 1 (Utility, DG , Option 2 (Utility , Option 3 (Utility , Category WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Present worth (USD) 1 086 663 3 095 088 4 187 418 Annual worth (USD/yr) 84 058 239 419 323 915 Return on investment (%) 9.5 9.9 9.8 Internal rate of return (%) 13.0 13.6 13.6 Simple payback (yr) 6.96 6.71 6.75 Discounted payback (yr) 9.22 8.79 8.85 Total NPC (USD) 43 167 210 41 158 780 40 066 450 LCOE (USD/kWh) 0.08132 0.07753 0.07547 Operating cost (USD) 2 526 998 2 187 642 1 991 156 Fig. 10 presents the monthly average renewable-energy requirement, which has been sold to the utility. Fig. 11 por - power generation. Both WTs contribute 3 151 494 kWh an- trays the monthly electricity production of the microgrid. nually, which is equal to 7.68% of 41  062  934 kWh/yr, the With respect to the reference option, there is a reduc- total electricity generated by the Option 1 microgrid con- tion in carbon-dioxide emission by 1 978 346 kg/yr, sulphur figuration. They operated for 8109 hours/yr with a LCOE dioxide by 8766  kg/yr and nitrogen oxides by 3817  kg/yr. of USD 0.0648/kWh. The annual electricity purchase from The Option 1 has a present worth of USD 1  086  663, an the utility grid is 37  774  048 kWh (91.985%). There is an annual worth of USD 84  058/yr, a return on investment excess of 434 kWh after meeting the microgrid energy of 9.5% and an internal rate of return of 13.0% with a Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 265 Table 8: Optimization of Options 1, 2 and 3—annual utility-bill comparisons Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Consumption charge (USD) 3 210 767.20 2 901 993.58 2 632 409.66 Demand charge (USD) 544 523.51 451 593.26 436 507.97 Total (USD) 3 755 290.71 3 353 586.84 3 068 917.63 Table 9: Optimization of Options 1, 2 and 3—savings in annual utility bill against reference model (utility and DG) Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Consumption charge (USD) 279 545.30 588 318.92 857 902.84 Demand charge (USD) 61 632.20 154 562.45 169 647.74 Total (USD) 341 177.50 742 881.37 1 027 550.58 Table 10. Optimization of Options 1, 2 and 3—fuel-consumption comparisons Option 1 (Utility , Option 2 (Utility , Option 3 (Utility, DG, Category DG, WT, BESS) DG, PV, BESS) WT, PV, BESS) Total fuel consumed (L) 38 054 43 164 43 974 Specific fuel consumption (L/kWh) 0.277 0.288 0.289 Mean electrical efficiency (%) 36.7 35.3 35.1 Marginal generation cost (USD/kWh) 0.236 0.236 0.236 Hours of operation 32 44 46 Mean electrical output (kW) 4294 3411 3305 Table 11: Optimization of Options 1, 2 and 3—emission comparisons Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Carbon dioxide (kg/yr) 23 972 809 21 690 836 19 690 368 Sulphur dioxide (kg/yr) 103 745 93 826 85 149 Nitrogen oxides (kg/yr) 51 207 46 420 42 186 Table 12: Optimization of Options 1, 2 and 3—reduction (–)/increase (+) in emissions compared with the reference option Option 1 (Utility , Option 2 (Utility , Option 3 (Utility , Category DG, WT, BESS) DG, PV, BESS) DG, WT, PV, BESS) Carbon dioxide (kg/yr) –1 978 346 –4 260 664 –6 261 132 Sulphur dioxide (kg/yr) –8766 –18 685 –27 362 Nitrogen oxides (kg/yr) –3817 –8604 –12 838 simple payback period of 6.96 years. The total NPC is USD Option 1.  There are two 1-MW PV systems and a single 43 167 210, the LCOE is USD 0.08132/kWh and the operating 2-MW PV systems along with the DG and the BESS with cost is USD 2 526 998. There are savings in utility bills on system converter. All are interconnected with the utility electricity-consumption charges USD –279 545.30 and de- main supply as in Fig. 12. As before, the data for the newly mand charges USD –61 632.20 against the reference option added PV systems are obtained from the HOMER Grid besides demand-reduction revenue of USD 279  681. As a database. The renewable-energy-capacity addition is more whole, the economic metrics of Option 1 are attractive than double in this option (4 MW in Option 2 versus 1.8 when compared with those of the reference option. MW in Option 1). Tables 5–12 portray the outcomes of the Option 2 opti- mization with a comparison against Options 1 and 3. The 4.3 Option 2: Utility, DG, PV, BESS total electricity production by this microgrid Option 2 is The significant difference between Options 1 and 2 is the 41  505  737 kWh/yr, of which the renewable contribution addition of PV systems in Option 2 in the place of WTs in is 7 213 475 kWh/yr (17.37%), whereas the DG has a minor Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 266 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 9: The DG-generated power output per year indicating its minimum (3000 kW), mean and maximum (7695 kW) values Fig. 10: The monthly average power output of both wind turbines Fig. 11: The monthly electricity production of the Option 1 microgrid for a 12-month period Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 267 contribution of 0.362%. The PV systems operate for 4376 has been reduced by 3 631.881 kWh (10.64%), which means hours/yr and the LCOE is USD 0.0406/kWh. The utility pur - that the dependence on the utility grid has been decreased. chase of energy is 34  142  167 kWh as the major share of Fig. 13 presents the annual primary-load variations, the 82.28%. In Option 2, the electricity purchase from the utility grid purchases and the grid-demand limit. Fig. 14 depicts the demand–response coordination of all the elements of Option 2. The demand incentive is set to USD 35/kW. The load peak is 8000 kW and the demand-reduction revenue works out to be USD 279 681. There is an excess of electricity production that has been sold to the grid. Option 2 has a return on invest- ment of 9.9% and an internal rate of return of 13.6% with a simple payback period of 6.71  years. The LCOE is USD 0.077532/kWh, whereas, in Option 1, it was USD 0.08132/ kWh. There are total savings on utility bills for electricity charges of UDS 742  881.37 against the reference option and, with reference to Option 1, the savings are higher by USD 401  703.87. There are considerable reductions on emissions in Option 2 with respect to the reference op- tion such as carbon-dioxide reduction by 4 260 664 kg/yr, sulphur-dioxide-emission reduction by 18  685  kg/yr and nitrogen-oxides reduction by 8604 kg/yr. These reductions are much higher than those of Option 1, too, because of the higher amount of renewable penetration. 4.4 Option 3: Utility, DG, WT, PV, BESS Option 3 is the renewable-energy-enhanced option con- sisting of both WTs and solar PV systems. The renewable- energy total capacity is 5.8 MW (consisting of 1.8 MW WT and 4 MW PV), which is >52% of the daily maximum Fig. 12: The microgrid architecture for Option 2 consisting of Utility, demand. All the data are already available in the opti- DG, PV, BESS with converter.This figure was used with permission from [47]. mization platform with the support of the HOMER Grid Fig. 13: Annual primary load, grid purchases and grid-demand limit in Option 2 Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 268 | Clean Energy, 2021, Vol. 5, No. 2 database. The demand-management option has been pro- generation is found to be 41  490  520 kWh, which is the vided as before. The microgrid architecture of this option highest generation among all the three options. While is shown in Fig. 6. The comprehensive outcome results are comparing the annual utility energy cost among all the presented in Tables 5–12. options, the reference option incurs the highest utility In Fig. 15, the monthly total energy production by all cost because of no renewables contributed. Among the the resources in Option 3 is given. The annual energy renewable-integrated Options 1–3, Option 3 has the lowest Fig. 14: The demand–response events in Option 2 Fig. 15: The monthly energy production in Option 3 with Utility, DG, WT, PV, BESS Fig. 16: The annual utility energy cost including the consumption and demand charges Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 269 utility energy cost, as seen in Fig. 16. In percentage meas- is no renewable-energy source. It is a good sign that the ures, the reference option is 100, Option 1 is 91.985, Option grid dependency has been reduced by >25% because of the 2 is 82.25 and Option 3 is the lowest, equal to 74.574. 25.04% energy production by the renewables (both the WT Apart from the electricity purchase from the utility, the and PV) and a small contribution of 0.366% by the DG. renewable-energy contribution has a significant impact on The major renewable-energy contribution resulted in the total energy requirement to meet the microgrid energy environmental friendliness especially in healthcare en-vir demand. Fig. 17 portrays the renewable-energy contribu- onments by a reduction in CO by 6 261 132 kg/yr, SO by 2 2 tions by all options except the reference case, since there 27 362 kg/yr and NO by 12 838 kg/yr against the reference option. Among all the options, these are the highest reduc- tion quantities in emissions. The generic 4-hour 1-MW Li-ion BESS has an autonomy of 0.899 hours, a storage wear cost of USD 0.025/kWh and a usable normal capacity of 4216 kWh. It has a lifetime throughput of 4  452  292 kWh and an annual throughput of 296 819 kWh/yr. The energy-content daily profile of the BESS is shown in Fig. 18. Both the DG and BESS coordinate with the renewable sources (WT and PV) in the demand- management events of the microgrid. Whenever there is no renewable-energy generation, the roles played by the DG and BESS are significant. On economic metrics, Option 3 has a return on invest- ment of 9.8%, an internal rate of return of 13.6% and a pay- back period of 6.75 years. Though all the three options have payback periods of <7 years and internal rates of return of ~13.6%, the LCOE is attractive for Option 3 at USD 0.07547/ kWh. Besides, the operating cost is much lower in Option 3 and its annualized cost is the lowest among all the options in spite of the addition of 5.8 MW of renewable-energy Fig. 17: The annual electricity contribution by all the renewable sources in MWh from each option systems. Fig. 18: The daily energy-content profile of the BESS for a 12-month period Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 270 | Clean Energy, 2021, Vol. 5, No. 2 Fig. 19: Utility-grid power outage for 4 hours 4.5 Utility-grid outage the region, the microgrid has been designed to maintain supply reliability even if the utility power outage is up to As the main purpose of the incorporation of the microgrid the maximum sanctioned demand of the microgrid. is to improve the power-supply resiliency, a power-outage case is considered based on the existing state utility grid’s operating conditions. During certain days, there would be a 4.6 Lessons learned and scope for future reduction in consumer demand such as 60%, which means energy fields that the consumer could make use of only 40% of the sanc- The proposed microgrid would be the first attempt at tioned maximum demand. Another prevailing situation is healthcare facilities in India since its first day of operation a 100% planned power cut or power-supply outage for cer - to ensure the availability of electricity. Renewable energy tain hours of a day, such as 13 hours to 17 hours. Under such as solar and wind have been considered for economic that situation, all consumers in that region are in darkness and environmental advantages. Biomass, apart from these unless backup power-supply facilities are available for two, may be the next option in the near future, depending them. The 100% planned outage is considered here as the on the production of waste. From a reliable point of view, unexpected power outage due to the grid’s technical func- more types of energy sources would be beneficial. In add- tional issues for the simulation purpose. ition to battery backup, fuel cells could be preferred for A 100% power outage of the utility grid for 4 hours backup power supply. Although bilateral power and com- during a daytime period has been simulated with the munication facilities are introduced in the name of smart- microgrid functioning (as in Option 2: Utility, DG, PV, BESS) grid technology, it is only the initial stage and, depending without the WTs because of low wind below the minimum on the state-of-the-art technological changes, it needs to in turbine speed. The 100% power outage occurred at 13.00 be modernized. hours. During that period, the maximum demand of AIIMS In this optimization platform, demand management was 7000 kW. The microgrid responded and all the loads has been duly considered without many grid failures (only were served by the battery backup system. Being daytime, one example has been provided). In future work, various there was a PV generation of ~4000 kW; hence, the DG magnitudes of grid failures, scheduled power cuts and the was not triggered. The supply interruption continued for energy harnessed from the hospital non-hazardous waste another hour while the AIIMS demand was ~5500 kW. As shall be considered. there was a reduction in PV generation to 3500 kW, the DG has been activated and the loads were met by the backup supply. Likewise, for the continued power outage until 5 Conclusion 17.00 hour, the generator and the PV generation main- tained the continued power supply through the BESS (Fig. This paper proposed a 20-MVA microgrid for the newly 19). Considering the prevailing power-outage situations of approved 750-bed AIIMS healthcare facility in Madurai. Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 Palanichamy and Naveen | 271 [4] Buonomano A, Calise F, Ferruzzi G, et al. Dynamic energy per - AIIMS hospitals already in operation, some of which formance analysis: case study for energy efficiency retrofits of had been established several decades before with no hospital buildings. Energy, 2014, 78:555–572. microgrid resources, have now begun to incorporate re- [5] Allcott  H, Collard-Wexler  A, O’Connell  S. How do electricity newable energy due to energy shortages and grid fail- shortages affect industry? Evidence from India. American ures. The authors of this paper considered that it will be Economic Review, 2014, 106:1–45. economical and environmentally feasible to introduce a [6] The Times of India PTI. 24-hour power to all by March 2019. microgrid with a utility-grid interconnection option as https://timesofindia.indiatimes.com/india/24-hour-power- to-all-by-march-2019-govt/article-show/62284777.cms (15 the electricity supply at the planning stage itself to en- December 2020, date last accessed). sure service continuity and reliability. A detailed load as- [7] The New Indian Express, Madhusudhakar K. 20 die during sessment has been carried out considering the number 12-hr power cut at Kurnool Hospital in Andhra Pradesh. of beds, the medical and nursing students, doctors, 2017. https://www.newindianexpress.com/states/andhra- nurses and supporting staff, outpatients and visitors, pradesh/2017/jun/23/20-die-during-12-hr-power-cut-at- critical and non-critical electrical loads, hostel facilities kurnool-hospital-in-andhra-pradesh-1619975.html.%20 (15 and resident staff quarters, etc. The demand forecast re- December 2020, date last accessed). [8] The Times of India, TNN. 21 die in Hyderabad govt Hospital, sulted in 16 000 kVA as the connected load capacity at a staff blame power cut. 2016. https://timesofindia.indiatimes. 0.90 power factor. Allowing space for future expansion, com/india/21-die-in-hyderabad-govt-hospital-staff-blame- 20 MVA has been finalized as the capacity of the pro- power-cut/article-show/53359874.cms (15 December 2020, posed microgrid. date last accessed). Within 2-km distance, the state utility grid is available for [9] Adair-Rohani H, Zukor K, Bonjour S,et  al. Limited electricity interconnection as the primary supply. For accommodating access in health facilities of sub-Saharan Africa: a systematic renewable energy, the locally available natural resources review of data on electricity access, sources, and reliability. Glob Health Sci Pract, 2013;1:249–261. are considered. The wind speed and solar radiation are fa- [10] World Health Organization and World Bank. Access to vourable; hence, 4 MW of PV panels and 1.8 MW of WTs are modern energy services for health facilities in resource- proposed for the microgrid with DG and BESS backup. For constrained settings: a review of status, significance, chal- optimization purposes, the HOMER Grid platform has been lenges and measurement. 2015. https://apps.who.int/ used. Various options of renewable capacities were tested iris/handle/10665/156847 (15 December 2020, date last and, as a result, a combined 5.8-MW capacity of wind and accessed). PV solar generation with 12 MW DG and 1 MW BESS were [11] Bonnema E, Studer D, Parker A, et  al. Large hospital 50% en- ergy savings: technical support document. 2010. https://apps. finalized. On economic metrics, a return on investment who.int/iris/handle/10665/156847 (15 December 2020, date of 9.8%, an internal rate of return of 13.6% and a payback last accessed). period of 6.75  years were achieved besides an attractive [12] Carpenter D, Hoppszallern S. Advancing efficiency: 2011 hos- LCOE as USD 0.07547/kWh. Avoided annual emissions of pital energy management survey. Health Facility Management, 6 261 132 kg of carbon dioxide, 27 362 kg of sulphur dioxide 2011, 24:15–22. and 12  838  kg of nitrogen oxides have been achieved as [13] Kantola M, Saari A. Renewable vs. traditional energy manage- compared to the mere utility-grid power supply. ment solutions—a Finnish hospital facility case. Renewable Energy, 2013, 57:539–545. [14] Modern Healthcare. Hospital shift from fossil fuel to renew- able energy sources. 2014. http://www.modernhealthcare. Acknowledgements com/article/20141108/MAGAZINE/311089981 (15 December C.P. did 60% of the work. P.N. did 40% of the work. 2020, date last accessed). [15] European Commission. Towards zero carbon hospitals with renewable energy systems. 2013. https://ec.europa.eu/energy/ intelligent/projects/en/projects/res-hospitals (15 December Conflict of Interest 2020, date last accessed). None declared. [16] Schneider Electric. A report on energy efficient hospitals— visiting the realities. https://www.cii.in/webcms/Upload/ CII%20Report%20on%20Energy%20Efficient%20Hospitals%20 References %E2%80%93%20visiting%20the%20realities.pdf (15 December [1] CADDET Energy Efficiency Maxi Brochure 5. Saving energy 2020, date last accessed). with energy efficiency in hospitals. 1999. https://www.scribd. [17] Lawrence Berkeley National Laboratory. High performance com/document/176818872/Energy-Efficiency-in-Hospitals- healthcare buildings: a roadmap to improved energy efficiency. Maxi-Brochure-5-CADDET (9 December 2020, date last 2009. https://indoor.lbl.gov/publications/high-performance- accessed). healthcare-buildings (15 December 2020, date last accessed). [2] Sathwik  R, Sayali  S, Vaishnavi  D. Bringing energy efficiency [18] García-Sanz-Calcedo J, Al-Kassir A, Yusaf T. Economic and en- for hospital building through the conservative and preventive vironmental impact of energy saving in healthcare buildings. measures. International Journal of Innovative Technology and Applied Sciences, 2018; 8, 440. Exploring Engineering (IJITEE), 2019, 8:1081219. [19] Schneider Electric. Microgrids: achieving reliable power for [3] Wang  T, Li  X, Liao  PC, et  al. Building energy efficiency for our most critical facilities. 2019. https://microgridknowledge. public hospitals and healthcare facilities in China: barriers com/healthcare-microgrids-reliable-power/ (15 December and drivers. Energy, 2016, 103:588–597. 2020, date last accessed). Downloaded from https://academic.oup.com/ce/article/5/2/254/6278351 by DeepDyve user on 25 May 2021 272 | Clean Energy, 2021, Vol. 5, No. 2 [20] Microgrid Knowledge Editors. Microgrids in hospitals economictimes.indiatimes.com/news/renewable/aiims-to- minimize threats of electrical outages. 2020. https:// go-green-with-solar-power-plant-reduce-power-bill-by-50- microgridknowledge.com/microgrids-in-hospitals/ (15 per-cent/57784132 (15 December 2020, date last accessed). December 2020, date last accessed). [34] The Asian Age. AIIMS to go green, to set up solar power plant. [21] Alessandro  B, Matteo  M, Gabriele  C, et  al. Microgrid design 2017. https://www.asianage.com/metros/delhi/230317/aiims- and operation for sensible loads: Lacor hospital case study in to-go-green-to-set-up-solar-power-plant.html (15 December Uganda. Sustainable Energy Technologies and Assessments, 2019, 2020, date last accessed). 36: 100535. Doi: 10.1016/j.seta.2019.100535. [35] NASA’s Open Data Portal. Prediction of Worldwide Energy [22] Alexis L, Miguelde SM, Alberto GM, et  al. Sustainable microgrids Resources (POWER). https://data.nasa.gov/Earth-Science/ with energy storage as a means to increase power resilience Prediction-Of-Worldwide-Energy-Resources-POWER-/wn3p- in critical facilities: an application to a hospital. International qsan (15 December 2020, date last accessed). Journal of Electrical Power & Energy Systems, 2020, 119: 105865. [36] NREL, US Department of Energy. National renewable energy [23] Arun Kumar V, Verma A. Optimization and implementation of laboratory database. https://www.nrel.gov/research/data- hybrid energy sources in remote and grid active microgrids-a tools.html (15 December 2020, date last accessed). case study for Indian Scenario 2016. In: IEEE 7th Power [37] National Building Code of India 2016. Standards & Codes. India International Conference (PIICON), Bikaner, India, 25–27 https://bis.gov.in/index.php/standards/technical-department/ November 2016, 1–6. national-building-code/ (15 December 2020, date last accessed). [24] Arun K, Ashu V, Rajbans T. Optimal techno-economic sizing of [38] Energy Conservation Building Codes. Energy Conservation a multi-generation microgrid system with reduced depend- Building Codes 2017. https://beeindia.gov.in/sites/default/ ency on grid for critical health-care, educational and indus- files/BEE_ECBC%202017.pdf (15 December 2020, date last trial facilities. Energy, 2020, 208, 118248. accessed). [25] Margaret  AW, Rajini  V. Cost benefit and technical analysis [39] TANGEDCOUP. Generation data. https://www.tangedco.gov. of rural electrification alternatives in southern India using in/generation.html (15 December 2020, date last accessed). HOMER. Renewable and Sustainable Energy Reviews, 2016, [40] Alberto B, Idoia SM, Pablo S, et al. Lithium-ion batteries as dis- 62:236–246. tributed energy storage systems for microgrids. In: Academic [26] Asian Development Bank. Handbook on Microgrids for Power Press. Distributed Energy Resources in Microgrids. Amsterdam: Quality and Connectivity. 2020. https://www.adb.org/docu- Elsevier Science, 2019, 143–183. ments/handbook-microgrids-power-quality-connectivity (15 [41] PV Software. Simulation and design of solar systems. https:// December 2020, date last accessed). photovoltaic-software.com/ (15 December 2020, date last [27] Xu  D, Long  Y. The impact of government subsidy on renew- accessed). able microgrid investment considering double externalities. [42] Palanichamy C, Sundar Babu N, Nadarajan C. Privatizing and Sustainability, 2019, 11:3168. restructuring Indian power sector: an overview. Journal of the [28] Darrell Proctor. Micro generation with macro possibilities. Institution of Engineers, 1999, 80:23–30. 2020. https://www.powermag.com/micro-generation-with- [43] Palanichamy C, Sundar Babu N, Nadarajan C. Renewable en- macro-possibilities/ (15 December 2020, date last accessed). ergy investment opportunities in Mauritius—an investor’s [29] California Energy Commission. A novel, renewable energy perspective. Renewable Energy, 2004, 29:703–716. microgrid for a California healthcare facility. April 2019. [44] Lawrence Berkeley National Laboratory (LBNL). DER-CAM https://ww2.energy.ca.gov/2019publications/CEC-500-2019- User Manual 2015. https://gridintegration.lbl.gov/der-cam (15 034/CEC-500-2019-034.pdf (15 December 2020, date last December 2020, date last accessed). accessed). [45] Robinson  D, Haldi  F, Kämpf  J, et  al. CitySim: comprehen- [30] Ministry of Health & Family Welfare Government of India. No. sive micro-simulation of resource flows for sustainable MoH & FW /Z-28016/168/2018-PMSSY-III/ RFP /1/ 2019. http:// urban planning. In: Proceedings of the 11th International IBPSA pmssy-mohfw.nic.in/files/tender/Reply%20to%20the%20 Conference, Glasgow, Scotland, UK, 2009, 1083–1090. Queries%20in%20Pre-Bid%20Meeting.pdf (15 December 2020, [46] Mendes  G, Ioakimidis  C, Ferrão  P. On the planning and ana- date last accessed). lysis of integrated community energy systems: a review and [31] Ministry of Education. National Institutional Ranking Framework survey of available tools. Renewable and Sustainable Energy 2020. https://www.nirfindia.org/2019/MEDICALRanking.html Reviews, 2011, 15:4836–4854. (15 December 2020, date last accessed). [47] Homer Energy. HOMER Energy—the microgrid software. [32] All India Institute of Medical Sciences, New Delhi. Top uni- https://www.homerenergy.com/products/software.html (15 versities 2014. https://www.topuniversities.com/subject- December 2020, date last accessed). rankings/2020 (15 December 2020, date last accessed). [48] Lambert T, Gilman P, Lilienthal Micr P. o power System Modelling [33] The Economic Times. AIIMS to go green with solar power with HOMER: In Integration of Alternative Sources of Energy. plant, reduce power bill by 50 per cent. 2017. https://energy. Hoboken, NJ: John Wiley & Sons, 2006.

Journal

Clean EnergyOxford University Press

Published: Jun 1, 2021

References