IIIT Hyderabad Publications |
|||||||||
|
Distributed Task Queues for parametric EnergyPlus simulations on the webAuthor: Shanmukh Dontu 201402105 Date: 2022-01-11 Report no: IIIT/TH/2022/2 Advisor:Vishal Garg AbstractOver the past few decades, there has been substantial growth in electricity demand of urban buildings across the globe due to rapid urbanisation and usage of energy-intensive systems. Buildings account for over 1/3rd of global energy usage and nearly 40% of energy-related carbon emissions. Due to this fact, energy-efficient buildings with appropriate design standards are adopted by the architects. Performing modelling and simulation using software such as EnergyPlus to assess the scope of energy consumption of a building requires expertise. Web-based parametric building energy simulation tools can help architects perform their desired studies without having much software expertise. Traditional web-based tools use a synchronous request-response method. It is utilized for smaller Input/Output CPU interactions between the web client and the web server. It blocks the client until the task assigned is completed by the server. EnergyPlus is CPU-intensive and multiple simulation requests on the web server can become a bottleneck to the synchronous approach. The parallel processing methods have been explored by the researchers to address this issue but the development of scalable and fault-tolerant systems have not been discussed so far. In this regard, two domain orientated web applications have been developed using asynchronous Distributed Task Queues (DTQ) in this work. The first application is a simple roof savings calculator where the benefits of a cool roof, sloped roof and green roof are evaluated over a conventional roof using EnergyPlus software. There are several independent cool roof and green roof calculators available online which are either look-up table or synchronous simulation engine based. They are limited to certain weather data and can block multiple users accessing the application simultaneously. The roof savings tool developed in this work uses DTQs which brings everything at the same place by providing an interface to compare the cool roof, sloped roof and the green roof over a conventional roof on every location provided by the EnergyPlus website which are over two thousand weather locations. It also includes a parametric simulation option which can let users decide the thickness of the roof based on the energy savings potential. This is the only tool available online for world-wide usage having such diverse simulating options and is meant to handle peak calculator usage. In the second application, the DTQ technique is implemented considering the notable complexity of early-stage design optimisation of a building. During the early stages of building design, architects tend to evaluate various strategies to simulate building energy consumption. There can be several numbers of building parameters to be evaluated such as passive roof strategies, window glazing material, the orientation of the building, window overhang depth, window-to-wall ratio, etc. In a situation having five-building parameters with ten design alternatives for each parameter, the web server needs to evaluate a hundred thousand simulation models with building energy simulation software. With each simulation model taking one minute to process, it may take about seventy days on a single-core system to complete the group of simulations. Using asynchronous DTQ technique, computational cores can be scaled based on the simulation requirements. DTQs does real-time processing while supporting task scheduling and CPU resource management as well. Early Design Optimisation tools (eDOT) for Air-Conditioned buildings have already existed in the past but the development of fault-tolerant systems has not been mentioned yet. DTQ software like Celery provides scalable and reliable features. Even if any of the systems go offline, the rest of the systems can still communicate with the Queue and perform tasks of the disconnected systems. There is an increasing concern of energy efficiency in buildings due to high dependency on mechanical systems for cooling and heating. In this work, the eDOT tool is enhanced by incorporating a Mixed-Mode simulation option. This hybrid approach for space conditioning uses natural ventilation and utilizes mechanical systems only when required to reduce energy consumption without compromising on occupant comfort. Energy Management System (EMS) algorithms are developed for Mixed Mode building models in EnergyPlus to control cooling loads, heating loads and natural ventilation based on certain conditions which include zone air relative humidity, outside relative humidity, zone air temperature, outside air temperature and cooling/heating set points. Parametric simulations for annual building energy consumption are carried out with the eDOT tool for New Delhi and San Francisco weather data. Design strategies with lowest energy profiles for AC and Mixed-mode buildings are examined and compared for both the cities. The cut-off Energy per Intensity (EPI) for identifying strategies is considered to be bottom 15% of the EPI range as more than 70% of the solutions are obtained in that domain. The impact of various design parameters on each other is also demonstrated. The design strategies are mapped using parallel axis graphs and energy distribution of the case study is generated with histograms. Potential energy savings of 10% and 20% has been observed with Mixed Mode buildings for the New Delhi and San Francisco weather data respectively while comparing with AC buildings. Hence, the concept of Distributed Task Queues in a simple and robust web-based application for handling multiple EnergyPlus simulations has been demonstrated in this work. Both the tools support multiple users accessing the application at the same time without compromising on the web user experience. These tools provide a platform for designers, architects, engineers, consultants, researchers, students, house owners or any organization to study the energy efficiency of buildings considering various parameters in all the climatic zones of the world. Full thesis: pdf Centre for IT in Building Science |
||||||||
Copyright © 2009 - IIIT Hyderabad. All Rights Reserved. |