CET 97 DOI: 10.3303/CET2297016 Paper Received: 31 May 2022; Revised: 10 August 2022; Accepted: 1 September 2022 Please cite this article as: Zulkafli N.I., Sukri M.F., Tahir M.M., Muhajir A., Hanak D.P., 2022, Optimal Operation of Chillers Plant in Academic Building by using Linear Programming Approach, Chemical Engineering Transactions, 97, 91-96 DOI:10.3303/CET2297016 CHEMICAL ENGINEERING TRANSACTIONS VOL. 97, 2022 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Jeng Shiun Lim, Nor Alafiza Yunus, Jiří Jaromír Klemeš Copyright © 2022, AIDIC Servizi S.r.l. ISBN 978-88-95608-96-9; ISSN 2283-9216 Optimal Operation of Chillers Plant in Academic Building by using Linear Programming Approach Nur Izyan Zulkaflia,*, Mohamad Firdaus Sukria, Musthafah Mohd Tahira, Asjufri Muhajira, Dawid Piotr Hanakb aCentre for Advanced Research on Energy (CARe), Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia bSchool of Water, Energy and Environment, Cranfield University, Bedfordshire MK43 0AL, United Kingdom nurizyan@utem.edu.my The operation of chillers plant in the HVAC system is not ideally efficient in major academic buildings because the chillers are operated without accounting for the cooling load demand of the air-conditioning area of the building. The current operation of the chillers plant may fall under two cooling effects, namely excessive cooling effect that will reduce the efficiency of the chillers and inadequate cooling effect that may create an unpleasant environment inside the air-conditioning area. This is because the cooling load demand of the air-conditioning area of the building is unknown and is not numerically measured. The main aim of this study is to find the optimal operation of chillers plant in HVAC system by formulating an optimization framework with the main goal of minimizing energy consumption and electricity costs. The formulation of the optimization framework for chillers plant operation is modelled as Linear Programming (LP) to establish real representation of the chillers plant. The energy consumption profile, cooling load demand, cooling capacity and COP can be obtained by using historical data analysis. The optimization framework is modelled in GAMS v38.2.1 and solved by CPLEX solver to obtain the optimum input for the chillers plant in the HVAC system. The minimum total power consumption can be achieved by optimally coordinating the operation of chillers, cooling towers and AHUs while maintaining room temperature. From the cost comparison analysis between the current and optimal chillers plant, considerable cost reduction is expected to be less than 5 % if the chillers plant is operating efficiently or greater than 30 % if the chillers plant is not functioning effectively. Therefore, this study is beneficial to the administration of academic building to find a strategic action plan to promote cost optimization in the operation of the chillers plant. 1. Introduction Buildings are one of the primary electricity consumers that is accounting for up to 52 % of total electricity consumption, in comparison to 45 % by the industrial sector and 3 % by the other sector in 2017 (Energy Commission, 2017). The heating, ventilation, and air conditioning system (HVAC) contributes to a major fraction of electricity consumption in commercial buildings. A central chillers network that is commonly used in the HVAC system for providing sufficient cooling air for the buildings is the largest electricity consumer in the building. Therefore, energy management for the chillers network has become essential to investigate potential strategies for reducing power demand and improving the energy efficiency of the chillers. The main operational performance of the chiller can be monitored by measuring the coefficient of performance (COP). COP is usually not uniform throughout the operation period due to differences in the sequencing control strategy of multiple chillers (Liu et al., 2017) and fluctuation in cooling load demands (Sala-Cardoso et al., 2019) that contribute to significant changes in power consumption profile. The performance of the chillers can be further improved by considering cooling load demands in the optimization approach to find the opportunity and the flexibility of chiller operation to reduce electrical power utilization (Shao et al., 2019). It is hard to predict the cooling load demands of the building because of inaccurate information at the design stage and limited operational data to predict system performance (Huang et al., 2018). The cooling load demands may vary in a 91 large range depending on weather conditions and the number of occupants in the building. Since the operation of chiller plant is strongly dependent on the dynamic change of cooling load demand of the building, the current practice in the academic building by using the conventional scheduling method of manual control may not be accurate enough in commanding the HVAC to operate as efficiently as possible. Optimization techniques can be adopted to find the best operation strategy for HVAC, especially for chillers which reflect the dynamic change of cooling load, energy efficiency, and energy cost without scarifying human thermal comfort (Varbanov et al. 2021). The majority literatures, neglected the impact of power change of the chiller system and instead concentrated primarily on the air-loop heat exchange from thermal zones (Wang et al., 2022). In this paper, integrated chiller-AHU plant, which consists of a cooling tower system, a chiller system, and air handling units are considered with time-varying air mixed temperature in the optimisation framework. The contribution of this study is the proposed optimization model is expected to represent the actual operation of the chillers plant by including important equations (e.g., energy consumption, cooling load, cooling tower effectiveness). The second contribution is the percentage of cost saving between the current and optimal chiller plant is known to further investigate its process flexibility and availability for reducing power demand. 2. Problem statement The main focus of the optimization model for the operation of the chillers network is to find the best operating condition of the chillers plant considering the detailed operation of chillers, cooling tower, and air handling units for an academic building. The cooling supply and demand of the building are considered for the chiller units to operate based on the predicted cooling demand by the building space. A linear programming (LP) model is presented for the optimisation problem considered in this study. The optimisation problem of chillers network is formally defined according to the following characteristics: (i) a given time horizon of eight hours is divided into an equal length of an hour; (ii) a number of two chillers with the given minimum and maximum chiller’s temperature (i.e., return and supply temperature of chilled water); (iii) a set of compressors in the chillers network; (iv) a number of thirteen AHUs with minimum and maximum air temperature. (i.e., return and supply air temperature); (v) a set of pumps to transfer the flow of chilled water to AHU and condenser water to the cooling tower, and (vi) a fixed tariff electricity price is introduced. The main optimal results that need to be made by the optimization model for every time are: (i) the amount of cooling load of the chiller; (ii) the chilled water return and supply temperature for chillers; (iii) the hot air and cooled air supply temperature for AHU; (iv) the air mass flow rate for the cooling tower; (v) the cooling requirement of the building; (vi) the power consumption for each compressor in the chiller system; (vii) the heat rejection from the cooling tower. The main optimal results are achieved by formulating the overall optimization model of the chiller plant to minimise the electricity costs. 3. Optimisation framework The operation of chiller network is modelled as Linear Programming (LP) model. The main part of the optimisation framework consists of: (i) constraints related to the chiller system, (ii) constraints related to the cooling tower, (iii) constraints related to AHU, and the (iv) objective function. The parameters and variables of the mathematical model are written as small letters for parameters and capital letters for variables. 3.1 Equations for chiller system The chiller system consists of the evaporator, condenser, compressor and expansion valve. Eq(1) models the cooling capacity of the evaporator (𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒), and Eq(2) describes the heat transfer for the condenser (𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐). Eq(3) defines the power consumption of a chiller (𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑒𝑒𝑖𝑖) that is closely related to the temperature difference of the evaporator based on study by Deng et al. (2015). Coefficients 𝛽𝛽(𝑖𝑖,𝑡𝑡) and 𝛾𝛾(𝑖𝑖,𝑡𝑡) are obtained by regression analysis of actual measured chiller data (Theng et al., 2021). 𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 = �̇�𝑚𝑖𝑖 𝑐𝑐ℎ𝑤𝑤𝑐𝑐𝑤𝑤�𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑖𝑖 − 𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑠𝑠 � ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (1) 𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 + 𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑒𝑒𝑖𝑖 ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (2) 𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑒𝑒𝑖𝑖 = 𝛽𝛽(𝑖𝑖,𝑡𝑡)�𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑖𝑖 − 𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑠𝑠 � + 𝛾𝛾(𝑖𝑖,𝑡𝑡) ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (3) The chilled water return temperature (𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑖𝑖 ) is modeled according to Eq(4) that are based on partial load ratio (𝑃𝑃𝑃𝑃𝑃𝑃(𝑖𝑖,𝑡𝑡)). PLR is the ratio of cooling capacity of the evaporator with the maximum cooling capacity based on Eq(5). 92 𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑖𝑖 = 𝑇𝑇(𝑖𝑖,𝑡𝑡) 𝑠𝑠 𝑃𝑃𝑃𝑃𝑃𝑃(𝑖𝑖,𝑡𝑡)𝑞𝑞𝑐𝑐(𝑖𝑖) 𝑚𝑚𝑒𝑒𝑚𝑚 �̇�𝑚𝑖𝑖 𝑐𝑐ℎ𝑤𝑤𝑐𝑐𝑤𝑤 ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (4) 𝑃𝑃𝑃𝑃𝑃𝑃(𝑖𝑖,𝑡𝑡) = 𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑞𝑞𝑐𝑐(𝑖𝑖) 𝑚𝑚𝑒𝑒𝑚𝑚 ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (5) Eq(1) to (5) are developed according to energy model of established commercial chillers by (Thangavelu et al., 2017). 𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 = 𝛿𝛿𝑖𝑖𝑞𝑞𝑖𝑖 𝑐𝑐ℎ𝑤𝑤ℎ𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 𝜂𝜂𝑖𝑖 𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (6) 𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑐𝑐𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 = 𝛿𝛿𝑖𝑖𝑞𝑞𝑖𝑖 𝑐𝑐𝑐𝑐𝑤𝑤ℎ𝑐𝑐𝑐𝑐𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 𝜂𝜂𝑖𝑖 𝑐𝑐𝑐𝑐𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 ∀𝑖𝑖 ∈ 𝐼𝐼, 𝑡𝑡 ∈ 𝑇𝑇 (7) Eq(6) and (7) are the calculation for power of chilled water pump (𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒) and condenser water pump (𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒) . Volume flowrate, pump head and efficiency are taken from energy audit report. 3.2 Equations for cooling tower Cooling tower is an important unit in chiller operation to continuously reject heat from cooling water cycle in the condenser by using ambient air through forced convection. The relation between the cooling water and air cycle in the cooling tower may affect the efficiency of the chiller. Eq(8) and Eq(9) is adopted from the model by (Liu et al., 2017). The performance of cooling tower (𝑄𝑄(𝑗𝑗,𝑡𝑡) 𝑐𝑐𝑡𝑡 ) is measured based on cooling tower effectiveness (𝜀𝜀𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖). �𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑖𝑖∈𝐼𝐼 = �̇�𝑚(𝑗𝑗,𝑡𝑡) 𝑒𝑒𝑖𝑖𝑖𝑖 �ℎ𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖,𝑐𝑐𝑐𝑐𝑡𝑡 − ℎ𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖,𝑖𝑖𝑐𝑐� ∀𝑗𝑗 ∈ 𝐽𝐽, 𝑡𝑡 ∈ 𝑇𝑇 (8) 𝑄𝑄(𝑗𝑗,𝑡𝑡) 𝑐𝑐𝑡𝑡 = 𝜀𝜀𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖�̇�𝑚(𝑗𝑗,𝑡𝑡) 𝑒𝑒𝑖𝑖𝑖𝑖 �ℎ𝑗𝑗 𝑠𝑠𝑒𝑒𝑡𝑡,𝑖𝑖𝑐𝑐 − ℎ𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖,𝑖𝑖𝑐𝑐� ∀𝑗𝑗 ∈ 𝐽𝐽, 𝑡𝑡 ∈ 𝑇𝑇 (9) 𝜀𝜀𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖 = �ℎ𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖,𝑐𝑐𝑐𝑐𝑡𝑡 − ℎ𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖,𝑖𝑖𝑐𝑐� �ℎ𝑗𝑗 𝑠𝑠𝑒𝑒𝑡𝑡,𝑖𝑖𝑐𝑐 − ℎ𝑗𝑗 𝑒𝑒𝑖𝑖𝑖𝑖,𝑖𝑖𝑐𝑐� ∀𝑗𝑗 ∈ 𝐽𝐽 3.3 Equations for air handling unit The purpose of AHU is to exchange heat between chilled water from the evaporator and the mixed air in the building space. The total heat rejected by the building (𝑄𝑄(𝑘𝑘,𝑡𝑡) 𝑏𝑏𝑐𝑐𝑖𝑖𝑖𝑖𝑐𝑐𝑖𝑖𝑐𝑐𝑏𝑏) composed of two types of heat that is known as sensible heat (𝑆𝑆𝑆𝑆(𝑘𝑘,𝑡𝑡)) and latent heat (𝑃𝑃𝑆𝑆(𝑘𝑘,𝑡𝑡)) as shown in Eq(10) and Eq(11). 𝑆𝑆𝑆𝑆(𝑘𝑘,𝑡𝑡) is the heat transfer equation between mixed air temperature and supply air temperature. 𝑃𝑃𝑆𝑆(𝑘𝑘,𝑡𝑡) is the amount of energy released during phase transition according to mixed air and supply air moisture contents. 𝑄𝑄(𝑘𝑘,𝑡𝑡) 𝑏𝑏𝑐𝑐𝑖𝑖𝑖𝑖𝑐𝑐𝑖𝑖𝑐𝑐𝑏𝑏 = ��𝑆𝑆𝑆𝑆(𝑘𝑘,𝑡𝑡) + 𝑃𝑃𝑆𝑆(𝑘𝑘,𝑡𝑡)� ∀𝑘𝑘 ∈ 𝐾𝐾, 𝑡𝑡 ∈ 𝑇𝑇 (10) 𝑆𝑆𝑆𝑆(𝑘𝑘,𝑡𝑡) = �̇�𝑚(𝑘𝑘,𝑡𝑡) 𝑠𝑠𝑐𝑐𝑒𝑒𝑒𝑒𝑖𝑖𝑠𝑠𝑐𝑐𝑘𝑘 𝑆𝑆𝑆𝑆 �𝑇𝑇(𝑘𝑘,𝑡𝑡) 𝑚𝑚𝑖𝑖𝑚𝑚 − 𝑇𝑇(𝑘𝑘,𝑡𝑡) 𝑠𝑠𝑐𝑐𝑒𝑒𝑒𝑒𝑖𝑖𝑠𝑠� ∀𝑘𝑘 ∈ 𝐾𝐾, 𝑡𝑡 ∈ 𝑇𝑇 (11) 𝑃𝑃𝑆𝑆(𝑘𝑘,𝑡𝑡) = �̇�𝑚(𝑘𝑘,𝑡𝑡) 𝑠𝑠𝑐𝑐𝑒𝑒𝑒𝑒𝑖𝑖𝑠𝑠𝑐𝑐𝑘𝑘 𝐿𝐿𝑆𝑆 �𝑀𝑀(𝑘𝑘,𝑡𝑡) 𝑚𝑚𝑖𝑖𝑚𝑚 − 𝑀𝑀(𝑘𝑘,𝑡𝑡) 𝑠𝑠𝑐𝑐𝑒𝑒𝑒𝑒𝑖𝑖𝑠𝑠� ∀𝑘𝑘 ∈ 𝐾𝐾, 𝑡𝑡 ∈ 𝑇𝑇 Eq(12) describes the mixed air temperature (𝑇𝑇(𝑘𝑘,𝑡𝑡) 𝑚𝑚𝑖𝑖𝑚𝑚) with the rise or fall degree of the surrounding air temperature (𝑡𝑡𝑚𝑚𝑖𝑖𝑡𝑡𝑡𝑡𝑐𝑐𝑐𝑐𝑡𝑡𝑠𝑠𝑖𝑖𝑐𝑐𝑒𝑒) during normal weather conditions. The purpose of this equation is to formulate the mixed air temperature inside the building space with the changes in surrounding air temperature. 𝑇𝑇(𝑘𝑘,𝑡𝑡) 𝑚𝑚𝑖𝑖𝑚𝑚 = 𝑡𝑡𝑚𝑚𝑖𝑖𝑡𝑡𝑘𝑘 + 𝑡𝑡𝑚𝑚𝑖𝑖𝑡𝑡𝑡𝑡𝑐𝑐𝑐𝑐𝑡𝑡𝑠𝑠𝑖𝑖𝑐𝑐𝑒𝑒 ∀𝑘𝑘 ∈ 𝐾𝐾, 𝑡𝑡 ∈ 𝑇𝑇: 𝑡𝑡 = 1 (12) 𝑇𝑇(𝑘𝑘,𝑡𝑡) 𝑚𝑚𝑖𝑖𝑚𝑚 = 𝑇𝑇(𝑘𝑘,𝑡𝑡−1) 𝑚𝑚𝑖𝑖𝑚𝑚 + 𝑡𝑡𝑚𝑚𝑖𝑖𝑡𝑡𝑡𝑡𝑐𝑐𝑐𝑐𝑡𝑡𝑠𝑠𝑖𝑖𝑐𝑐𝑒𝑒 ∀𝑘𝑘 ∈ 𝐾𝐾, 𝑡𝑡 ∈ 𝑇𝑇: 𝑡𝑡 > 1 93 The linking relation between the cooling capacity of the evaporator (i.e., cooling supply) and heat rejected by the building (i.e., cooling demand) is shown in Eq(13). Penalty variables (𝑃𝑃𝑡𝑡) with high penalty costs are imposed in this equation to avoid unsatisfied cooling capacity generated from the chiller plant. �𝑄𝑄(𝑖𝑖,𝑡𝑡) 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑖𝑖∈𝐼𝐼 = � 𝑄𝑄(𝑘𝑘,𝑡𝑡) 𝑏𝑏𝑐𝑐𝑖𝑖𝑖𝑖𝑐𝑐𝑖𝑖𝑐𝑐𝑏𝑏 𝑘𝑘∈𝐾𝐾 + 𝑃𝑃𝑡𝑡 ∀𝑡𝑡 ∈ 𝑇𝑇 (13) 3.4 Objective function The objective function for the optimization modelling is minimising the total electricity cost which composes of the power consumption for chiller, chilled water pump and condenser water pump. The fixed electricity tariff for commercial buildings according to tariff of TNB Berhad is used as cost coefficient (𝜙𝜙𝑡𝑡𝑒𝑒𝑖𝑖𝑒𝑒𝑐𝑐) in Eq(14). 𝑚𝑚𝑖𝑖𝑚𝑚� � 𝜙𝜙𝑡𝑡𝑒𝑒𝑖𝑖𝑒𝑒𝑐𝑐 �𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑒𝑒𝑖𝑖 + 𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 + 𝑊𝑊(𝑖𝑖,𝑡𝑡) 𝑐𝑐𝑐𝑐𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒� + 𝜙𝜙𝑡𝑡 𝑒𝑒𝑒𝑒𝑐𝑐𝑒𝑒𝑖𝑖𝑡𝑡𝑠𝑠𝑃𝑃𝑡𝑡 𝑖𝑖∈𝐼𝐼,𝑡𝑡∈𝑇𝑇 � (14) 4. Optimal operation of chiller plant – A case study of an academic building In this part, the description of a case study is presented. The HVAC system consists of two chillers (i1 and i2) and two cooling towers (j1 and j2) that are operating simultaneously throughout the duration of air-conditioning time of 10 h from 8:00 AM until 5:00 PM (t0 - t10). The building is a 4-storey building with a total number of 17 air handling units (k1 to k17). 4.1 Description of a case study The schematic diagram of the chillers network in HVAC system is shown in Figure 1. The chillers network involves two water cycles. The first water cycle is the condenser water. The condenser water from the cooling tower enters the condenser in the chiller to absorb heat from the chilled water. The condenser water leaves the chiller at a higher temperature and is then pushed by the pump to the cooling tower where the air is forced by the fan inside to reject heat to the environment. The condenser water then re-enters the chiller at a cooler temperature, completing the cycle. The second water cycle is the chilled water. The chilled water from the evaporator in the chiller enters the pump to distribute the chilled water around the building where it enters the air handling units (AHU). The chilled water exchanges the heat with the hot air that is collected by AHU within the building. The cold air is then released by the AHU throughout the building. Finally, this chilled water returns to the chiller at a higher temperature to complete the cycle. Figure 1: Schematic diagram of chillers network in HVAC system 94 Table 1 displays the list of input parameters for the case study. The input parameters are obtained from a recent energy audit report for an academic building in a public university in Malaysia. Table 1: Input parameters for the case study Symbol Value Description Symbol Value Description �̇�𝑚𝑖𝑖 𝑐𝑐ℎ𝑤𝑤 29 kg/s Chilled water mass flow rate for chiller i1 �̇�𝑚(𝑘𝑘,𝑡𝑡) 𝑠𝑠𝑐𝑐𝑒𝑒𝑒𝑒𝑖𝑖𝑠𝑠 7–32 kg/s supply air flow rate from AHU to room �̇�𝑚𝑖𝑖 𝑐𝑐ℎ𝑤𝑤 30 kg/s Chilled water mass flow rate for chiller i2 �̇�𝑚𝑖𝑖 𝑐𝑐𝑐𝑐𝑤𝑤 23 kg/s Condenser water mass flow rate for chiller i1 𝑞𝑞𝑐𝑐(𝑖𝑖) 𝑚𝑚𝑒𝑒𝑚𝑚 536 kJ Maximum cooling capacity for chiller i1 �̇�𝑚𝑖𝑖 𝑐𝑐𝑐𝑐𝑤𝑤 60 kg/s Condenser water mass flow rate for chiller i1 𝑞𝑞𝑐𝑐(𝑖𝑖) 𝑚𝑚𝑒𝑒𝑚𝑚 580 kJ Maximum cooling capacity for chiller i2 𝜂𝜂𝑖𝑖 𝑐𝑐𝑤𝑤𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 0.57 Chilled water pump efficiency 𝑐𝑐𝑘𝑘 𝑆𝑆𝑆𝑆 1.21 Sensible heat coefficient 𝜂𝜂𝑖𝑖 𝑐𝑐𝑐𝑐𝑒𝑒𝑐𝑐𝑚𝑚𝑒𝑒 0.77 Condenser pump efficiency 𝑐𝑐𝑘𝑘 𝐿𝐿𝑆𝑆 3.00 Latent heat coefficient 𝜙𝜙𝑡𝑡𝑒𝑒𝑖𝑖𝑒𝑒𝑐𝑐 USD 0.083 /kWh Electricity price 4.2 Results of the case study This case study has been solved by using the Eq(1) until (14) considering input parameters in Table 1. The optimisation problem has been solved under standard configuration and a zero optimality gap in GAMS version 38.2 by using CPLEX solver in Intel(R) Core (TM) i7. The optimal results that are obtained are analyzed and discussed below. Figure 2a shows the normalised total heat requirement of the building spaces per time. The normalized total cooling requirement is calculated by finding the ratio of cooling requirement to the maximum cooling requirement for each time. The maximum cooling requirement is at time t6 because the total sensible heat is the highest at this time due to large air temperature difference between supply air and mixed air temperature. The temperature difference in t6 was in the range of 1.11 to 1.3 times higher than that of the other time point. Figure 2b displays normalized percentage of cooling load and power input for each chiller. The maximum cooling requirement represents maximum cooling load as shown at time t6. Although cooling load at this time is maximum, the power input especially for chiller i2 is at minimum level in comparison to other time. The reason for this is because the coefficient (beta and gamma) for power input is lower at this time point. Figure 2: (a) Cooling requirement of building spaces; (b) Comparison between power input and cooling load Table 2 shows the optimal operating condition for chiller plant specifically for time t6. The coefficient of performance (COP) for chiller i2 is better than chiller i1 because the ratio between the cooling load and power input for chiller i2 is higher than the ratio for chiller i1. According to ASHRAE – Chiller Plant Efficiency chart (Hartman, 2001) on chiller plant performance indicator, the average COP for chiller plant is 4.87 which indicates a high-efficiency optimised chiller plant. Table 2: Optimal operating condition for chiller plant Chiller Chilled water supply temperature (°C) Chilled water return temperature (°C) Power Input (kW) Cooling Load (kW) COP Condition of chiller plant i1 7.80 11.4 93.54 438.95 4.69 Optimal i1 7.30 11.80 112.64 465.20 4.13 Current i2 7.78 11.7 97.73 493.25 5.04 Optimal i2 7.70 12.10 113.50 516.43 4.55 Current 98.5 % 99.0 % 99.5 % 100.0 % t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 time 70 % 80 % 90 % 100 % t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 time Cooling Load i1 Cooling Load i2 Power Input i1 Power Input i2 (b)(a) 95 Figure 3 shows the total electricity cost and electricity cost components comparison of the current chiller plant and the optimal result. The percentage of cost saving is around 23 %. This shows that the current electricity cost that is based on energy audit can be potentially minimised if the operating condition follows the optimized result. The percentage cost reduction is closer to 30 % for not functioning effectively. Therefore, the current chiller plant is not considered as highly efficient due to the big gap between the current and optimal chiller plant. Figure 3: Comparison of (a) total electricity cost for optimal result and energy audit report (percentage); (b) electricity cost components for current and optimal chiller plant 5. Conclusions In this study, the optimal operation of the chiller plant in an academic building has been presented by accounting for the operation of chillers, cooling towers and air handling units. From the results, it is observed that the percentage of optimal cost is lower than the current cost by about 23 %. This demonstrates a direct result of percent energy improvement that can be made by the building owner to achieve low electricity cost and enhanced energy efficiency through optimal energy consumption of the chiller plant. Therefore, this could result in obtaining high energy star rating to achieve sustainable energy practice for an academic building. Acknowledgments The authors are gratefully recognised Universiti Teknikal Malaysia Melaka (UTeM) for providing financial support under short term research grant with grant number, PJP/2021/FKM/S01809. References Deng K., Sun Y., Li S., Lu Y., Brouwer J., Mehta P.G., Chakraborty A., 2015, Model predictive control of central chiller plant with thermal energy storage via dynamic programming and mixed-integer linear programming, IEEE Transactions on Automation Science and Engineering, 12(2), 565–579. Energy Commission, 2017, Performance and Statistical Information on Electricity Supply Industry in Malaysia 2017, Energy Commission Malaysia, Putrajaya, Malaysia. Hartman T.B., 2001, All-variable speed centrifugal chiller plants, Ashrae Journal, 43, 43–53. Huang P., Huang G., Augenbroe, G., Li S., 2018, Optimal configuration of multiple-chiller plants under cooling load uncertainty for different climate effects and building types, Energy and Buildings, 158, 684–697. Liu Z., Tan H., Luo D., Yu G., Li J., Li Z., 2017, Optimal chiller sequencing control in an office building considering the variation of chiller maximum cooling capacity, Energy and Buildings, 140, 430–442. Sala-Cardoso E., Delgado-Prieto M., Kampouropoulos K., Romeral L., 2019, Predictive chiller operation: A data- driven loading and scheduling approach, Energy and Buildings, 208. Shao Z., Gholamalizadeh E., Boghosian A., Askarian B., Liu Z., 2019, The chiller’s electricity consumption simulation by considering the demand response program in power system, Applied Thermal Engineering, 149, 1114–1124. Thangavelu S.R., Myat A., Khambadkone, A., 2017, Energy optimization methodology of multi-chiller plant in commercial buildings, Energy, 123, 64–76. Theng L.M., Tan J., Liew P.Y., Tan L.S., 2021, A review of manufacturing sustainability assessment tool selection criteria: A quantitative score-rating system versus process-data sustainability assessment, Chemical Engineering Transactions, 89, 523-528. Varbanov P.S., Klemeš J.J., Chin H.H., 2021, Minimisation of the resource demands and environmental footprints – the need for smart symbiosis networks, Chemical Engineering Transactions, 89, 421-426. Wang W., Zhou Q., Pan C., Cao F., 2022, Energy-efficient operation of a complete chiller-air handing unit system via model predictive control, Applied Thermal Engineering, 201, 117809. 0 % 20 % 40 % 60 % 80 % 100 % Optimal chiller plant Current chiller plant (a) 0 % 20 % 40 % 60 % 80 % 100 % current i1 optimal i1 current i2 optimal i2 Chiller Chilled Water Pump Condenser Water Pump (b) 96 016.pdf Optimal Operation of Chillers Plant in Academic Building by using Linear Programming Approach