Analysis of Interactions between Raw Material and Energy Demands for Data Centers
Fernando Andres Peñaherrera Vaca
The amount of professional data centers in Germany is continuously increasing. This results in a growing demand for ICT components, such as servers and storage units. This, in addition to their short service times and high intensity of critical raw material content, results in a higher demand for valuable raw materials and primary energy for their production. The evaluation of the energy and material demands comes with challenges regarding poor data quality, a focus on operational phases, and lack of proper information on End-of-Life strategies. Recovery of valuable materials from high grade electronic components is considered strategic because of their high concentration of critical metals but at the same time represents a knowledge gap.The goal of this dissertation is to analyze the raw material and primary energy demands for the cradle-to-grave lifecycle of professional data centers and its equipment, outside of the operational phase, and to analyze possible resource savings that come from different End-Of-Life strategies. The interdependencies of these resource demands within several scenarios is to be analyzed. To achieve this, Life Cycle Assessment of different data center systems and their components is conducted, with consideration of the infrastructure and inventorial composition of such facilities.This requires gathering and updating data on raw material content and on devices composition, preparation of models for analysis, and evaluation of results through appropriate Life Cycle Impact Assessment Methods that reflect the use of material and energy resources, as current indicators consider mostly the operational phase, and omit the importance of different materials to the economy. Different scenarios for technologies applied for End-of-Life treatment and recycling are evaluated. Due to the number of models required and the volume of input and output data, automation of model building, calculations, and of results evaluation is required. An assessment of input data quality allows evaluating the results quality and validity, which serves to provide an overview of the data improvements. The information system used to calculate resource consumption is evaluated within the scope of this research. This provides a validated methodological basis for holistic resource consumption evaluation.
März / 2024
phdthesis
OFFIS Autoren
