52 / 2021-08-31 00:09:28

Uncertainty Analysis Using Monte Carlo of Selective Artificial Ground Freezing Systems for the Safety of Deep Underground Mine

selective artificial ground freezing,Monte Carlo analysis

Underground water flooding is a major threat to the safety of wet underground mines. As such, the construction of the Cigar Lake Uranium Mine (Saskatchewan, Canada) underwent two major flooding incidents which severely damaged the mine infrastructure and delayed the production by eight years. Eventually, the flooding issues were resolved by installing an artificial ground freezing (AGF) system that froze all the underground water in the vicinity of the orebody as shown in Figure 1. The AGF process is accomplished by continuously circulating a coolant from refrigeration plants in the ground surface to the ore deposits located at a depth of more than 400 meters underground. Such large AGF systems demand intensive energy supply which come at significant costs. Accordingly, the area in between the orebody and the ground-surface are insulated by an air chamber since ground freezing is not required. The insulated area is often referred to as the passive zone in selective artificial ground freezing (S-AGF) systems.



Most of studies on S-AGF systems in the literature analyze the ground freezing process in the active zone [1,2]. Nevertheless, few studies examined the energy consumption of S-AGF systems. Zueter et al. [3] derived the first fully conjugate CFD model for S-AGF systems and scaled an optimum air gap thickness in the passive zone that results in minimum energy consumption in the passive zone. Zueter et al. [4] then developed a reduced-order model that reduces the computational cost of field-scale simulations of ideal S-AGF systems, shown in Figure 1(a), by more than 99% as compared to CFD models. Afterwards, Zueter et al. [5] extended the reduced-order model to capture the effect of freeze-pipe eccentricity as shown in Figure 1(b). It was found that eccentric freeze-pipes cause higher heat dissipation in the passive zone by 20-200%, depending on many uncertain parameters, such the ground porosity, ice nucleation temperature, pipe design, and coolant thermal energy.



Evidently, there is a high level of uncertainty in S-AGF applications. In this study, we aim to conduct uncertainty analysis using Monte Carlo simulations to investigate the effect of various design, geological and operating parameters on the energy consumption of S-AGF plants as well as the phase-transition-front expansion in the passive zone. A reduced-order mathematical model of S-AGF systems that addresses the effect of freeze-pipe eccentricity developed in our previous work [5] will be employed due to its high computational speed and reliable data. The results of the study will assist engineers and practitioners in the AGF industry at quantifying and estimating the effect of different parameters in the S-AGF process. Consequently, the design and operation of S-AGF systems can be tuned according to the most dominant parameters. Ultimately, the study aims to ensure the safety of underground mines by determining the most significant parameters that hinder the S-AGF process and then adapting mitigation strategies, while minimizing the operational costs and carbon emissions of the refrigeration plants.