Creating virtual indoor monitoring sensors to enhance LCA inventory for monitoring energy and well-being in buildings during use phase

Mustafa, Mohamed 2024. Creating virtual indoor monitoring sensors to enhance LCA inventory for monitoring energy and well-being in buildings during use phase. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF (Thesis) - Accepted Post-Print Version Download (8MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (237kB)

Abstract

The buildings’ operational phase is considered the longest phase where buildings contribute the most to their overall environmental impact. In this context, Life Cycle Assessment (LCA) helps to quantify both the background and foreground of the environmental impact of the buildings’ components and performances. Different LCA dimensions were introduced to tackle this issue, aiming to quantify the environmental impact of the buildings’ components and performances. In this setting, the LCA impact significantly depends on the life cycle inventory components. However critical issues evolve around these components identified to affect the overall LCA impact. One major component is the indoor sensors used to provide input data to the inventory with the scope of energy optimisation while maintaining optimum indoor conditions. The relevant research has shown considerable progress along this path, yet with notable shortcomings. These include a lack of understanding of the significance of the trade-off between the characteristics of the integrated technologies and their optimisation efficiency. This gap was further highlighted, particularly in the adoption of indoor sensors considering their environmental impact weightings against their optimisation outputs. The primary aim of this research is to create and integrate virtual indoor monitoring sensors into LCA inventory to optimize energy consumption and well-being performance during buildings’ use phase. Accordingly, this Thesis presents a holistic approach to virtualise indoor monitoring sensors while providing credible measurements for energy optimisation purposes. Informed by current research, the methodology was demonstrated in a step-by-step approach to answer the research questions. Mainly, different simulation engines were used for different purposes. For instance, the Computational Fluid Dynamic (CFD) simulation and thermal imaging were used to optimise the physical sensors’ positions to guarantee high-accuracy measurements at a later stage of the virtualisation. The methodology also traced and analysed multiple dynamic and static indoor boundary conditions of influence on the sensors measurements. This approach has significantly helped in understanding the sensing measurements behavior iii under different conditions, which provided more certainty around the virtual measurements. The virtual sensors resulted in a decrease of CO2 emissions by 11.698.76 ton of CO2 per each physical sensing unit. Accordingly, the developed solution contributed to phasing out the physical sensors and therefore, their associated embodied carbon. Thus, this finding is considered significant in eliminating the associated carbon from a pivotal LCA inventory component. Furthermore, the resulting high-accuracy measurements of the developed virtual sensors also factored in maintaining the occupants’ well-being conditions and the associated energy consumption. The founded equations in virtualising indoor sensors resulted from the extensive CFD, and EnergyPlus simulations and data analysis. These simulations were used to define variables governing the indoor sensors’ measurements were practical representations of different indoor and outdoor influencing factors that control indoor measurement behaviours. The achieved virtual sensors’ high-accuracy measurements were validated by using physical sensors in the case study zone. Furthermore, the comparison to the widely used Machine Learning (ML) models indicated higher accuracy of this Thesis’ framework. In summary, the key findings of this Thesis open a new path to virtual indoor sensors research. The identification of the indoor environment parameters of interest to energy and well-being performances narrows the scope of the assessment to a building’s casespecific. As a result, this finding has effectively helped in reducing the number of needed sensors. Furthermore, the findings emphasised the need to optimise sensors’ locations for higher accuracy measurements. The software simulations used to identify optimum sensors’ locations have also contributed to finding a relationship between different location measurements, which was used to reduce the number of sensors. Additionally, the total virtualisation equation found to virtualise indoor temperature represents another significant contribution to the indoor virtual sensors’ research field. Overall, the successfully achieved new level of highly accurate virtual sensors’ measurements of temperature, pressure, CO2 levels, and humidity can be counted as a significant step in narrowing the physical components of the LCA inventory.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Engineering
Uncontrolled Keywords:	Indoor monitoring sensors; Virtual sensors; Embodied carbon; BIM; Machine learning prediction; LCA inventory
Date of First Compliant Deposit:	8 November 2024
Last Modified:	08 Nov 2024 16:40
URI:	https://orca.cardiff.ac.uk/id/eprint/173784

Actions (repository staff only)

Edit Item