Building occupancy behavior is becoming increasingly important and playing a significant role in the modeling of building energy performance, intelligent operations of building systems, and design of the future building system. Image-based, threshold and mechanical, motion sensing, and radio-based sensing are commonly used for occupancy sensing in buildings. However, those sensing technologies suffer from high cost, inevitable sensor errors, and scalability issues. Thus, they are not widely implemented in buildings. With the vast development of information technologies in the era of the internet-of-things (IoT), occupant sensing and data acquisition are not limited to traditional approaches. Social media could provide new near-real-time data sources that might contain occupancy information in space and in time. Utilizing public APIs provided by social media services, it is possible to attain the geographic information through either geo-tagged posts from Twitter or Facebook check-in messages. These datasets explicitly indicate the occupant presence and could be used to estimate the occupancy. However, it is well known that most social media users probably are not willing to disclose their location information. To increase the volume of the social media datasets, this paper provides a methodology to detect the implicitly geo-tagged posts that hold valuable occupancy information through a text classification and semantic analysis. The implicitly geo-tagged posts refer to the posts that could be inferred for the human occupancy information, but the user does not add the location to the posts. In this framework, first, the data acquisition tool is utilized to obtain the Tweets containing the textual location information such as a museum, a restaurant, etc. Then, training and evaluation datasets are pre-selected respectively with labels regarding whether the posts contain relevant occupancy information. The Word2Vec, a word embedding algorithm, is used to convert the text to the vectors as the input to the classification model, alongside with some Tweet-content-based features. In addition, several popular classification machine learning techniques are adopted to train the model. The preliminary results show that the proposed methodology could detect the implicitly geo-tagged posts from social media with relatively high classification performance. We also presented a preliminary result of a typical occupancy schedule to be used by the building energy models based on social media data. Results from this preliminary study will lay the foundation of the occupancy sensing through the social media data mining, which aims to provide another data source for occupancy sensing in buildings.
In this paper we present a methodology to map individual occupants' thermal preference votes and indoor environmental variables into personalized preference models. Our modeling approach includes a new Bayesian classification and inference algorithm that incorporates hidden parameters and informative priors to account for the uncertainty associated with variables that are noisy or difficult to measure (unobserved) in real buildings (for example, the metabolic rate, air speed and occupants’ clothing level). To demonstrate our approach, we conducted an experimental study in private offices by considering thermal comfort delivery conditions that are representative of typical office buildings. Personalized preference models were developed with the training dataset and the developed algorithms were used in a detailed validation process. The proposed model showed better prediction performance compared to previous methods. Towards realization of preference-based control systems, this study also addresses practical limitations associated with controlling model complexity and data efficiency as well as using effective model evaluation metrics to train reliable personalized preference models in the real world.
According to the adaptive approach for thermal comfort, such as described for occupant controlled, naturally conditioned spaces in ASHRAE Standard 55-2013, occupants may find thermal comfort in a wider temperature range compared to that specified through heat balance model. Our research is investigating the energy savings potential through applying aspects of the adaptive approach to air conditioned buildings and the development of an optimization scheme for thermal comfort with peak energy cooling demand. This paper presents the results of an ongoing study investigating the impact of temporarily increased cooling setpoint temperature during simulated demand response events on building occupants at a university campus in the southeastern U.S., and the occupant's adaptive behaviors for reducing thermal discomfort. More than 1500 data points were collected and analyzed to date in classrooms, dining spaces and lobby areas, with additional data collection conducted during the 2017 cooling season. This paper focuses on the results from classrooms. To ensure that the results are applicable to real-world situations, testing was conducted in the subject's everyday environment and normal routines with the subjects being unaware that testing was ongoing up until the time of their survey questions. The surveys recorded human factors like gender differences and time of exposure to the thermal environment, any adaptive behaviors taken by the occupant's, and environmental conditions such as ambient temperature and zone temperatures, relative humidity, and CO2 levels. As would be expected, variations in people's perceptions existed, but the results showed that increasing the cooling setpoint temperature up to 77 degrees F (25 degrees C) did not generally result in thermal discomfort among occupants. In some cases, occupant thermal satisfaction actually improved for those preferring a warmer room, since the setpoint temperature in these zones is typically around 70-72 degrees F or 21-22 degrees C. Increasing zone temperatures to more than 77 degrees F (25 degrees C) resulted in the average actual mean vote greater than +0.5 as measured by the ASHRAE seven-point thermal comfort scale. Women generally perceived their thermal environment cooler than men in the same environment and with similar clothing level (CLO value). The outdoor ambient temperature also affected the occupant's thermal comfort vote, presumably through adaptive behaviors such as adjustment to their clothing type and level or possibly their overall expectations, although there is no conclusive evidence for changing expectations so far from this study.
Life cycle assessment of data centers has indicated that the main environmental impacts stem from the embodied impacts, energy efficiency and source energy mix of the facility mechanical and electrical systems and equipment and IT equipment. Metrics to describe the in-use efficiency of the power and cooling infrastructure are well adopted, however the other areas often go unmonitored, in part due to lack of data, metrics and difficulty of measurement. Without a holistic view of impact, burden shift may occur, i.e. where action is taken to reduce the environmental impact of one area which moves or increases the impact to another area. Life cycle assessment considers a range of environmental impacts grouped into areas of protection: human health and climate change, ecosystem quality and resources. In this paper a metric comprising separate measures for each main area of impact is described, identifying key variables and analysing their range so that their effect on environmental impact may be understood and used as a basis for decision making. The operational component includes the influence of energy and water consumption, PUE, WUE and renewable energy usage. The embodied component considers variables such as utilization, equipment lifecycle, materials and disposal in mechanical and electrical systems and plant and IT equipment. These are related to the IT output via a productivity proxy. The paper includes preliminary analysis of the application of the metric
This study presents the development of personalized models of occupant satisfaction with the visual environment in private perimeter offices. A set of experiments was designed and conducted to collect comparative preference data from four human test-subjects. A probit model structure, with assumed latent satisfaction utility model in the form of multivariate Gaussian function, was adopted for developing the preference model. Four distinctive satisfaction and preference models were trained with a sequential Monte Carlo algorithm using experimental data. The posterior estimations of model parameters, visual preference probabilities and inferred satisfaction utility functions were investigated and compared, with results reflecting the different characteristics of the subjects. The developed visual satisfaction utility function was designed for use in personalized control, where occupants could balance their own satisfaction expectation and energy considerations.
This paper presents a preliminary analysis of the impact of a gas-fired absorption heat pump (GAHP) on heating and cooling energy consumption of a library building in Ontario, Canada. A single effect gas-fired absorption heat pump (GAHP) that uses ammonia-water as the working pair has been installed at the library building. Energy consumption data from utility bills from 2012-2014 were collected. Linear regression analysis was conducted in PRISM using outdoor dry bulb temperature. The building characteristics in terms of energy consumption per heating degree day and cooling degree day were used along with the heating and cooling capacity curves of the GAHP to determine annual heating and cooling energy consumptions and demands. The annual energy consumptions for heating and cooling of the GAHP were compared with conventional cooling equipment having seasonal energy efficiency ratio (SEER) of 14 and heating equipment having annual fuel utilization efficiency (AFUE) of 76%. Results show an average annual reduction of 11.1% in energy consumption, 22.7% reduction in energy cost and 9.7% reduction in greenhouse gas (GHG) emissions for the three years when GAHP is used for both heating and cooling. Preliminary analysis shows that GAHPs have great potential for heating applications in Canada and these could operate with a net reduction of greenhouse gas emissions when used for both heating and cooling.
Combination space and water heating systems have historically offered end users and installation contractors numerous benefits over conventional, standalone equipment - typically a gas furnace or electric heat pump paired with a standard gas or electric storage water heater. Combination systems, typically driven by a gas-fired potable (tankless water heater) or non-potable (boiler) water heating system, offer benefits including reduced equipment costs with one "thermal engine", reduced installation costs through requiring only one vent/gas line/condensate drain, and when deployed effectively, they can yield consistent high operating efficiency and reduced cost. Concerning high operating efficiency, these "combi" systems often cost-effectively improve the operating efficiency of domestic hot water (DHW) production to a condensing efficiency (>90%), where the economics of a standalone DHW system at a similar efficiency level are often difficult. For the majority of U.S. homes that are heated by natural gas, particularly in colder climates of the Pacific Northwest, Midwest and Northeast, the authors have demonstrated a gas heat pump-driven residential combi system, through laboratory and field testing, a step up in combi system efficiency. The gas heat pump (GHP) component is based on a low-cost single-effect absorption cycle and, through prior testing, has demonstrated an operating efficiency of a projected AFUE of 140% for Climate Region IV as defined by ANSI Z21.10.4. The GHP has a nominal output of 80 kBtu/hr at 47 degrees F and is capable of 4:1 modulation, to load follow when necessary. Developed to provide space and water heating to a residence, integrating with a forced-air heating distribution via a hydronic air coil and heating an indirect storage tank for DHW, the GHP-driven combi system is shown to meet a home's space and water heating loads simultaneously, through simulated- use testing in a laboratory and through a field demonstration over 12 months at a residence in Tennessee. The authors outline the performance of the GHP-driven combi system as a function of loading, operating conditions (e.g. ambient temperatures), and system control strategies. Additionally, the impact of system design considerations for component sizing and control, such as the indirect storage tank size, and system performance during defrost events are explored.
Many HFC refrigerants have a high global warming potential (GWP), and therefore have the potential to contribute to climate change. Increasing concern due to the presence of greenhouse gases, including high GWP refrigerants, in the atmosphere and their effect on global climate is leading to international policies aimed at phasing down the use of refrigerants with high global warming potentials. The HVAC&R industry is actively developing new environmentally friendly alternatives to replace the high GWP fluids currently in use. R-134a has been widely used in medium pressure refrigerant products, including medium and large capacity chillers, as the replacement for R-12. However, because of R-134a's high GWP(AR5) of 1300, the US EPA, through its Significant New Alternatives Program (SNAP) program, has placed a ban on the use of R-134a in chillers beginning in 2014. R-516A, also known as ARM-42, has been proposed as an alternative to R-134a. R-516A is an azeotropic blend of mostly R-1234yf (77.5%wt) with some R-134a (8.5%wt) and R- 152a (14%wt) that has thermodynamic characteristics very similar to R-134a. R-516A has a significantly lower GWP(AR5) of 130. However, it is mildly flammable, receiving an ASHRAE Standard 34 classification of A2L. Thermodynamic calculations predict that the capacity should be 1.7% lower and the efficiency should be 2.1% lower for R-516A compared to R-134a when operating at conditions typical of the AHRI Standard 550/590 rating point. This paper describes the thermodynamic characteristics of R-516A and reports the results of tests that were performed on a 105 RT (370 kW) air cooled screw chiller with R-134a and R-516A. Performance was measured over a range of ambient air temperatures from 65 degrees F to 125 degrees F (18 degrees C to 52 degrees C). At the AHRI Standard 550/590 rating conditions, measured capacity when running the unit with R-516A was about 2% lower and EER was reduced by 3.5% to 4.0% compared to R-134a. Given these results, R-516A can be considered a design compatible alternative to R-134a in water chiller products.
In current study, the thermal and hydraulic performance of a novel bare tube heat exchanger prototype, manufactured by using stainless steel tubing with outer diameter of 0.8 mm, is experimentally investigated under dry and wet conditions. The inlet air dry bulb temperature is 26.7 degrees C; the inlet air relative humidities are 35% (dry), 50% (wet) and 70% (wet) respectively; the inlet air frontal velocity are 3, 6 and 9 m/s respectively. The inlet water temperature is 12 degrees C and the water mass flow rates are 20, 35 and 50 g/s respectively. The effects of inlet air humidity, air flow rate, water flow rate and orientation (vertical and horizontal) of heat exchanger were discussed. The test results show that the total heat transfer capacity increases as inlet air humidity increases, the air flow rate increases and water flow rate increases. The change of latent heat is affected by inlet air humidity, water flow rate and heat exchanger geometry. Airside pressure drop increases as air flow rate increases and inlet air humidity increases. Water flow rate increase has negligible influence when the surface is total wet or total dry and leads to pressure drop increase when surface is partical wet. Power law correlations of Chilton-Colburn heat transfer factor j, mass transfer factor j(m) and friction factor f were developed for Reynolds number range of 200 similar to 600. Maximum deviations are +/- 15%, +/- 10% and +/- 10%, respectively.
A whole new class of refrigerant chemistry has been developed as a reaction to new regulatory actions to restrict and lower the direct global warming potential (GWP) impact of many hydrofluorocarbons (HFC) or fluorocarbon (or F-gas) refrigerants. These new refrigerants can be referred to as alkenes, or unsaturated hydrocarbons which means they contain two or more carbon atoms linked by a double bond. Today it is common for these refrigerants to be referred to as hydrofluoroolefins (HFOs), olefins or unsaturated HFCs and as a result of the double bond, can react quickly in the atmosphere which results in short atmospheric lives and thus results in low global warming impacts.It is confusing to many that a refrigerant with high atmospheric reactivity can have low reactivity in heating ventilation air conditioning and refrigeration (HVACR) equipment. These new refrigerants have been evaluated in laboratory testing under accelerated temperature simulated HVACR equipment conditions and have shown a variety of chemical reactions, but in general have shown acceptable stability. This paper will provide an overview the atmospheric chemistry reactions of these new olefin refrigerants, which atmospheric lives are measured in days, and compare and contrast them to very limited chemical reactions for these new olefin refrigerants in accelerated life laboratory testing for HVACR equipment.
Ground source heat pumps (GSHPs) can provide an efficient way of heating and cooling buildings due to their high operating efficiencies. The implementation of these systems in urban environments could have further benefits. In such locations the ground source heat is potentially more accessible via alternative sources such as through underground railways (URs). This paper investigates to what extent the heat in the soil surrounding an UR tunnel could enhance the operation of urban GSHPs installations. To address this, a numerical investigation was set out which included a parametric study considering a number of geometrical options of the systems. The results showed that heat extraction rates of GSHPs installed near UR tunnels can be significantly improved by up to similar to 43%.
The Canadian High Arctic Research Station (CHARS), located in Ikaluktutiak (Cambridge Bay), Nunavut, is a world-class, interdisciplinary facility that anchors a strong research presence in Canada's north. This paper describes the resilient and sustainable approaches to facility design and operations that were utilized to ensure superior performance in the harsh environment of the Canadian Arctic.Canada's Arctic presents significant challenges to the design, construction, and operation of resilient and sustainable buildings, including a very cold and dry climate, long supply chains, periods of 24-hour light and darkness, and energy and water costs that are 5-10 and 20-30 times more expensive, respectively, than in southern Canada. As such, the design of the facility emphasized optimization of energy performance, water use reduction, and indoor environmental quality. Energy simulations were conducted early and throughout the design process in order to optimize system sizing and emergency power requirements.The simulation results during the design process indicate an estimated 70% reduction in annual energy use compared to a reference building. This paper discusses some of the approaches utilized to improve resilience and sustainability, including high-efficiency heat recovery to counteract laboratory ventilation requirements, energy-efficient lighting and HVAC equipment, adiabatic humidification, and increased envelope insulation. Modeling of renewable energy generators for net-zero energy operation was also conducted; this paper outlines some of the opportunities and limitations of this approach in a small dieselelectric microgrid.A robust energy monitoring system was implemented for building commissioning, simulation model calibration, on-going energy management, and assessment of performance gaps. In order to verify building performance, significant commissioning efforts were undertaken to test the response of various systems in normal and emergency modes. The monitoring system will also support the client in their efforts to use CHARS as a `living laboratory' for testing of sustainable technologies appropriate for the Arctic regions of the world.
Perfectly unidirectional (laminar) flows are rare. Such flows can be attempted either by supplying the air through HEPA filters covering the entire ceiling or by placing restraining panels around the array of supply (laminar) diffusers. In most other cases the flows are "semi unidirectional" - unidirectional only in the core of the supply air stream and recirculating/mixing in the other locations. This study analyzes behavior of such semi unidirectional flows under isothermal and non-isothermal conditions. A CFD model of a virtual test chamber is developed to study the effect of supply airflow rates/discharge velocity on the airflow patterns, velocity, and temperature distribution. A total of four supply airflow rates ranging from 200 cfm (94.4 lps) to 800 cfm (377.6 lps) are analyzed. Non-isothermal analyses are performed for two different locations of sensible heat sources - one with the heat source on a table and the other with heat sources near the ceiling. This study indicates that under isothermal conditions the discharge velocity (flow rate) has little impact on the flow behavior of the supply air jet. For the conditions analyzed, the velocity of the isothermal downward air jet decreases (and becomes non-unidirectional) after traveling about 60 percent of the distance from the discharge. However, in the case of non-isothermal conditions the discharge velocity can significantly affect the directionality of the flow. When a heat source was placed on the table and when the cooling capacity of the supply air jet was four times the cooling load, the air jet could overcome the upward buoyant force of the hot air. When heat sources were placed near the ceiling, the hot air surrounding the heat sources entrained into the downward moving jet causing acceleration in the centerline velocity, which showed to decrease with increasing airflow rate. Interestingly such acceleration was lowest for the lowest supply airflow rate. This study further indicates that for non-isothermal conditions Archimedes Number can be a good initial indicator for estimating the directionality of the supply air jet. This number should be smaller than 0.2 for the cold air jet to travel further without substantially losing its unidirectional behavior.
This paper presents an innovative approach to procuring operating room HVAC systems: the operating room will be designed, built, and then tested for sterile performance. Much of the innovation lies in the final test, which is designed to assure that the built operating room achieves sterility prior to the owner's acceptance. Such performance tests of operating rooms are not uncommon in Europe; however, this delivery approach is the first of its kind in the United States.In the current market, several vendors offer integrated operating room ceiling solutions, adapting the design and construction technology from the clean room industry. This owner solicited proposals, interviewed vendors, and selected two vendors to provide operating room ceilings on future projects. The resulting agreements between the owner and the vendors includes a test of the operating room to meet ISO 14644 Class 6 and microbial contamination not to exceed 10 CFU/m3. These tests shall be conducted before the close-out phase of the project after the operating room is cleaned and sterilized per the applicable protocols.This paper also describes the process of selecting the performance criteria. Because there is no specific clinical or medical evidence pointing to a certain level of cleanliness, the owner selected ISO Class 6 based on a comparisons of international standards, known studies of U.S. operating room performance, and coordination with internal stakeholders.
Large, tall, multi-zone spaces often utilize mid-level mixing systems to condition the occupied zones. Efficiency of such mid-level mixing systems in maintaining desired temperature and humidity within the occupied zones has implications not only on occupant thermal comfort and indoor air quality; but on fan energy consumption. Optimization of airflow rate requirements for such systems thus becomes essential and is typically performed using Computational Fluid Dynamics (CFD). Traditional CFD modeling techniques for airflow rate optimization involve simulating several design iterations, analyzing results from individual analysis and eventually making an analytically driven design decision. It is to be noted though that it is often impossible to fully identify the effects that changes to airflow rates in one zone have on temperatures both within the zone it is directly serving as well as other zones within the space using such traditional solution methods. This paper summarizes the optimization of airflow rates to maintain a desired set-point temperature throughout the occupied zone of a large multi-zone space; by means of implementation of a Proportional-Integral (PI) control loop within the ANSYS Fluent commercial CFD code. The paper describes a steady state case study simulating a space consisting of multiple zones each served by separate variable air volume (VAV) boxes and terminal devices. Internal heat gains are applied as volumetric heat sources to multiple zones. Solar heat gain is computed using the Solar Load Model within ANSYS Fluent incorporating a 3D ray tracing methodology. Conduction though the envelope is calculated within the CFD model. These heat sources result in non-uniform heat gain in multiple zones. PI controls were then applied for each of the zones independently within a single CFD study. The supply airflow rates in each zone is set as a control variable, while the temperature within the zone it serves is applied as a measured process value, which is compared against the set-point temperature - the desired process value - and is used in turn by the PI feedback loop to control the airflow rate. This paper presents the temperature distributions and airflow rates of the resulting optimized design and its improvements over the baseline design for this particular case study.
Many building standards, guidelines and codes focus on the structure, envelope, and energy use of the building, but do not detail or explore the environment through the lens of building occupants and their experience of the interior environment. These documents help owners and designers meet the requirements for quality design and dictate a certain standard of care, but fall short with respect to the perceived acoustical quality and function of the space. Current research is exploring opportunities for human centric design, how we can improve occupant experience and support physiological and psychological wellbeing, and propose metrics for standards and guidelines to optimize acoustical conditions in the built environment.
This paper investigates a modeling strategy, intended for model-based control, for an air-based thermal energy storage device used for peak-load shedding. The device, denominated an Electric Thermal Storage (ETS) heating system, is designed to "store" electric energy as heat during hours when energy costs are lower and demand charges are not incurred. It may be used in commercial, institutional or multi-unit residential buildings. The ETS core consists of high-density ceramic bricks that may be heated to very high temperatures (up to 900 degrees C [1652 degrees F]) via electric heating elements. ETS devices are available in two different models, which may be connected respectively to air- or water-based HVAC systems; this study focuses solely on air-based applications. ETS systems are particularly useful in regions where space heating is predominantly electricity-based, and where outdoor temperatures are very low; these conditions can result in high electric peaks due to high heating demand. The storage capacity of the commercial ETS devices ranges between 320 and 960 kWh [1,091,885 and 3,275,656 BTU/h], which is large enough to replace backup duct heaters (or water tank heaters) powered by fossil fuels, thus limiting harmful emissions while also reducing peak demand.Although ETS devices have proven their usefulness to reduce peak loads, there are still some issues regarding their operation. Currently, ETS has basic control logic which involves knowledge of current power demand and outdoor temperature. This control logic can cause the system to store more energy than required during the shoulder season and thus increase the electricity bill at the same time. While previous research has addressed the development of detailed physics-based models for these types of ETS systems, this particular work aims to develop and compare simple (yet robust and generalized) control-oriented models for the ETS. The main purpose of these control-oriented models is the rapid simulation of the ETS device in order to assess load management strategies and help in decision-making. The investigated modeling approach is gray-box modeling (in which physical models are combined with measured data to complete the model). These control-oriented models are intended to be used, along with knowledge of future conditions (such as electricity pricing, occupancy, weather forecasts etc.), to plan ahead the operation strategies within the Building Automation System to better regulate electrical loads while maintaining comfort.
We propose a drip-feeding cooling system that efficiently cools high thermal density servers using multiprocessor implementation and General-purpose computing on graphics processing units (GPGPU) used for high-performance computing such as machine learning. We evaluated this system through computational fluid dynamics (CFD) simulation and an experiment. In the CFD simulation, two types of fluorine inert liquid and three types of silicone oil were used as refrigerant. In the experiment, one type of fluorinated inert liquid and one type of silicone oil were used. As a result, this system could efficiently cool heat generation by about 16 kW per rack at a power usage efficiency of 1.04 or less and reduce a floor load to 500 kg/m(2) (102 lbs/ft(2)) or less.
Economic and environmental benefits of combined cooling, heating and power (CCHP) systems can be enhanced with integration of solar photovoltaic (PV) arrays by virtue of recent cost reductions of PV arrays. Such hybrid PV and CCHP systems could be potentially profitable in California (CA) that has one of the most expensive electric rates in the U.S and also provides incentive programs favorable for distributed and renewable energy systems. The objective of this study is to investigate economic feasibility of implementing the hybrid PV and CCHP system with a grid connection for a representative large office building in CA.The payback period of such combined system is estimated based on a conventional separated production system where electricity is imported from the grid and heat is produced from a boiler. For the economic analysis, actual electric and gas rates in CA and the market price for each unit of the hybrid systems are used. In this study, the modified following electrical load ( MFEL) strategy is applied to operate hybrid PV and CCHP system. The MFEL can determine the required power output of the power generation unit (PGU) accounting for the electricity requirement for the electric chiller that dynamically varies depending on the electrical, cooling and heating loads of the building while avoiding excess electricity production. In the studied hybrid system with the MFEL, 28% and 66.5% of the annual electrical load of the building is met by PV and PGU, respectively. Only 5.5% of the electrical demand is purchased from the grid. Due to reduced electrical dependency on the grid mostly during the peak time in summer, significant amount of the annual cost saving as much as $965,000 is achieved and finally the payback period of the hybrid PV and CCHP system is 5.9 years. This payback period can be reduced when considering additional incentive programs provided from CA. Therefore, implementing hybrid PV and CCHP systems in CA is cost- competitive to the conventional systems and economically feasible.
Cooling with ice thermal storage can be the most cost-effective, reliable system approach to cooling different types of buildings. The ice thermal storage can reduce energy costs by shifting the cooling cost from on-peak to off-peak periods. The paper proposes an integrated approach to size and operate the ice thermal storage with different loads and weather conditions. The variables such as chiller and ice storage sizes and charging and discharge times during different seasonal conditions are considered. Different building types, locations, utility rate structure are investigated to evaluate the advantages of optimal ice thermal design and operation. The study considers the effect of the ice thermal storage on the chiller performance and the associated energy cost and demonstrates the cost saving achieved from optimal ice storage design. Two load estimation techniques are evaluated, including linear regression model and artificial neural networks. The results show a significant cost energy saving can be obtained by optimal ice storage design through the proposed approach.
