Conference PaperPDF Available

A combined ventilation and thermodynamic model for dry surfaces to predict and optimize the NHEA cooling and heating capacity for Creighton mine

Authors:
  • Microsoft (Vancouver)

Abstract and Figures

Historically, Creighton Mine has operated without the use of mechanical refrigeration for more than 100 years. The Natural Heat Exchange Area (NHEA) of Creighton Mine is part of an old block caving area divided into 4 blocks of trenches that were used to extract the ore. Currently, it is used year round to stabilize the outside temperature before the airflow enters the fresh air system of the mine. Initially, the first blocks (5 & 6) were brought into operation to condition the air for the ventilation system in the early 1960s; after the successful use of these initial blocks to cool and heat the air, number 2 Block was added in 1982. The number 1 Block was then added to the NHEA system in 2000. The four available blocks are now in operation and at 100% capacity. Today Creighton Mine is reaching 8000 ft (2438 m) in depth without requiring the use of a mechanical refrigeration system, as it is being presently managed. This paper presents the combined ventilation and heat exchange model, for dry surfaces, calibrated with the historical data, developed to predict the heat exchange that will occur in the NHEA. Having generated the functions and parameters to estimate the heat exchange through the NHEA, the system was then optimized for use of the cooling and heating capacity and compared with the historical data to show potential improvements that could be achieved under different operating strategies than those used in the past.
Content may be subject to copyright.
13th United States/North American Mine Ventilation Symposium, 2010 – Hardcastle & McKinnon (Eds.)
A combined ventilation and thermodynamic model for dry
surfaces to predict and optimize the NHEA cooling and heating
capacity for Creighton mine
E. Acuña, A. Tousignant & N. Bilal
MIRARCO, Laurentian University, Sudbury, Ontario, Canada
L. Fava & D. Goforth
Mathematics and Computer Science, Laurentian University, Sudbury, Ontario, Canada
D.O’Connor & C. Allen
Vale Inco Limited, Sudbury, Ontario, Canada
ABSTRACT: Historically, Creighton Mine has operated without the use of mechanical refrigeration for more than
100 years. The Natural Heat Exchange Area (NHEA) of Creighton Mine is part of an old block caving area
divided into 4 blocks of trenches that were used to extract the ore. Currently, it is used year round to stabilize the
outside temperature before the airflow enters the fresh air system of the mine. Initially, the first blocks (5 & 6)
were brought into operation to condition the air for the ventilation system in the early 1960s; after the successful
use of these initial blocks to cool and heat the air, number 2 Block was added in 1982. The number 1 Block was
then added to the NHEA system in 2000. The four available blocks are now in operation and at 100% capacity.
Today Creighton Mine is reaching 8000 ft (2438 m) in depth without requiring the use of a mechanical
refrigeration system, as it is being presently managed. This paper presents the combined ventilation and heat
exchange model, for dry surfaces, calibrated with the historical data, developed to predict the heat exchange that
will occur in the NHEA. Having generated the functions and parameters to estimate the heat exchange through the
NHEA, the system was then optimized for use of the cooling and heating capacity and compared with the
historical data to show potential improvements that could be achieved under different operating strategies than
those used in the past.
1 Introduction
Creighton Mine uses the Natural Heat Exchange Area (or
System, NHEA) as its primary intake for the ventilation
system of the mine and also as its primary air conditioning
system. Unlike regular mechanical refrigeration systems,
this one is based on the natural heat exchange generated
between the rock mass and the airflow that goes through
the different trenches of the NHEA and then into the main
ventilation system that provides fresh air for the mine. The
current NHEA provides clean cooling and heating capacity
to Creighton Mine and avoids the costs of heating and
mechanical refrigeration.
The benefit of better management of the NHEA for
Creighton Mine is to delay the installation of mechanical
refrigeration and hence, to delay the capital and
operational expenditures of such a mechanical
refrigeration system, thereby reducing the production cost
for the mine.
2 Background
The surface temperature of the mine can vary from 30°C
during summer to -40°C during winter. Because of this
variance in the outside temperature the rock mass of the
NHEA is warmed up during summer and the heating
capacity used during winter. Likewise, the rock mass is
cooled during winter and the cooling capacity used during
summer. This enables the full usage of a natural cooling
and heating system year round.
As Creighton Mine develops deeper, the temperature in
the lowest levels of the mine is approaching the values at
which the mine is considering the installation of
mechanical refrigeration. In the past, mechanical
refrigeration was never required by the mine because the
NHEA system, manually operated, provided enough
cooling and heating to keep the mine in production.
Previous studies, have shown that the cooling capacity of
the NHEA as it is currently being operated will no longer
be sufficient taking into account the planned underground
expansion of the mine.
As the current cooling and heating system is being
manually operated and the number of potential
293
© 2010, MIRARCO - Mining Innovation
configurations for each period of operations is very large,
Creighton Mine management would like to explore
alternatives to make better use of the NHEA. The possible
delay of a mechanical refrigeration installation would defer
capital costs and reduce operating costs while guaranteeing
safe and healthy operational conditions for the miners. In
this study the alternative considered was to model the heat
exchange process and then try to optimize it in order to
maximize the cooling capacity that can be obtained from
the NHEA based on the historical data provided by
Creighton Mine.
2.1 NHEA description
The NHEA consists of a caved area of the mine which is
divided into distinct sets of Blocks (1, 2, 5 and 6) that are
associated with main ore passes and a number of
associated tramways/trenches per block; each equipped
with a ventilation door at the end of the trench that controls
the airflow to the fresh air system of the mine. Blocks 1, 2,
5 and 6 have 17, 37, 18 and 24 trenches respectively, with
each trench reporting a certain amount of airflow at a
certain temperature when they are open. The NHEA has 96
trenches in total, which have to be manually operated in
order to control the full area. Considering only the
possibility of the door being open or closed the potential
combinations for each day of operation, with one state of
the doors to be kept all day, are 296 = 7.92E+28. The
number of combinations clearly makes it difficult to decide
which doors are to be kept open and which have to be
closed. Additionally a minimum number of doors have to
be open per block in order to provide the airflow required
by the mine without generating a large pressure drop for
the main and booster fans. Currently, the ventilation doors
are being opened or closed manually using a predefined
protocol. The protocol is followed when and where
possible due to safety conditions in particular trenches,
operating conditions of the doors and available man power.
When a door is open, surface air is drawn through the
rock mass contained in the open pit by fans located
underground into the mine ventilation system. The air is
collected from all blocks on 800 Level after it passes
through the rock mass where the heat exchange
phenomenon occurs. Figure 1 presents a general schematic
of the airflow path to get into Creighton fresh air system.
The temperature data collected and used to build year
round profiles for each trench are from dry bulb sensors
installed in the side of the door of each trench of the
NHEA. The airflow volumes of the trenches are a function
of the number of doors open in the NHEA, and measured
manually at this time. The airflow volumes measured at
the bottom of each block are fairly consistent all year
round; however, depending on how many doors are left
open on each block, the airflow volume per trench can
change drastically.
Figure 2 presents a schematic of the NHEA, an old
caving area, and the trenches that constitute each block.
From the schematic the four block structures can clearly be
identified, where Block 1 is the first from left with Blocks
2, 5 and 6 in succession to the right.
2.2 Heat exchange
Heat transfer is mainly achieved through three mechanisms
of heat exchange: conduction, convection and radiation. In
the particular case of underground mine ventilation, some
Figure 1 Airflow path schematic.
294
Figure 2 NHEA schematic.
procedures have already been developed to model the heat
exchange between a rock surface and the airflow in an
opening (McPherson, 1993). These approaches assume a
constant rock temperature equal to the virgin rock
temperature (VRT) and not a varying rock temperature, as
is the case in the NHEA. Also, unique conditions of the
heat exchange between the fragmented rock mass of the
NHEA and the airflow passing through are not included in
the function presented by McPherson (1993) as the
procedure was not developed for that purpose. The initial
methodology, for this project, developed for heat exchange
between a rock surface and the airflow in an opening was
modified to fit a curve to each trench temperature profile
and will be explained in Section 4.
To the best of our knowledge, little work has been done
to model the heat exchange of airflow passing through
fragmented rock mass for underground environments. The
NHEA is a unique system that has been delivering good
results for several decades but no mathematical modelling
has been done to quantify how the heat exchange happens
in each trench to fully understand the system and for
prediction. Johnson (2006) attempted to predict the thermo
and psychrometric properties of the intake air passing
through rock mass for a hypothetical case study similar to
the NHEA but smaller in size. Because the rock mass was
modeled as a whole system and not as an integrated set of
individual trenches, Johnson’s (2006) work cannot be
applied when generating operational solutions for the
trenches of the NHEA.
2.3 Genetic algorithms (GA)
A genetic algorithm (GA) is a metaheuristic method based
on Darwin’s theory of the evolution of species. A genetic
algorithm is an iterative search technique for solving linear
and nonlinear problems. The technique does not guarantee
optimality, but can be used to find good feasible solutions
in a reasonable amount of time. A genetic algorithm
explores the solution space, generating solutions,
evaluating their fitness or value, and ‘learning’ how to
build better solutions. They have shown their capability to
optimize main ventilation systems as shown by Acuña et al
(2009), Lowndes et al (2005) and Lowndes & Yang
(2004), considering fans and regulators, but have not been
applied to ventilation systems considering heat exchange
optimization. A GA has an advantage over other
optimization techniques when a simulation must be
generated in order to calculate the fitness or value of the
solution. In this particular case the proposed solution is
simulated with the ventilation model and the
thermodynamic model to generate the value of the
proposed solution.
3 Optimization Methodology
The NHEA solution can be formulated as a binary
optimization problem that has to be solved for each period
of operation. Initially the length of operating periods was
narrowed down to monthly, then later weekly and lastly
daily. The binary decision corresponds to deciding if the
state of the door is open or closed, subject to the following
constraints: a minimum number of doors that have to be
open per block, operational constraints on the doors of the
trenches and the heat exchange capacity of the NHEA. The
objective function or value function is to minimize the
temperature of the incoming air for the mine. In order to
perform this optimization three models were developed: a
ventilation model, a heat exchange model and an
optimization model.
3.1 Ventilation model
The intake system for Creighton Mine consists of the
following: air from surface is drawn through the rock mass
295
then passes through trenches and is then collected on the
800 Level of the NHEA where it is distributed into three
parallel fresh air systems. Given this structure a simplified
ventilation model was developed first assuming that all
trenches can be considered as parallel airways that go from
surface to the 800 level and then to the fresh air systems.
To estimate the resistance of these parallel airways a
ventilation audit was performed in the NHEA and
resistance values were calculated based on the square law
for turbulent airflow. Additionally, fixed quantity airflows
were set for the bottom of each block (1, 2, 5 and 6) to
model and simulate the average airflow that each block
usually collects. Then for each combination of the states
(open/closed) of the control doors in the trenches an
airflow volume was calculated for each trench.
3.2 Heat exchange model (dry surfaces – data)
Referring to the particular case of the NHEA, it was
assumed that the heat exchange process can be divided in
two main states: when the door is open or when the door is
closed. In the case where the door is closed, very little
airflow (leakage) is assumed to be passing through the
trench. In those conditions, it is expected that the
predominant heat exchange mechanism is the conduction
from surface to the door of the trench. When the door is
open and airflow can be calculated at the door of the
trench, it is assumed that convection is the predominant
heat exchange mechanism. Using these assumptions to
predict the temperature at the door of each trench for each
period, a combined function of convection and conduction
was developed to describe standard heat exchange
occurring at the NHEA and to fit the historical data. For
the approach presented in this study, only convection and
conduction were modeled as part of the heat exchange
process. Radiation was disregarded because it was
considered as not significant in terms of the heat exchange
impact (McPherson, 1993).
The procedure developed by Pierre Mousset-Jones,
included in McPherson (1993) to model the convection
heat exchange between rock mass and air was not intended
to be used under the conditions of this project. In order to
use the concept for the NHEA study, the procedure was
slightly modified to be able to properly fit the historical
data. Also, the standard formula for heat exchange through
conduction, included in Incropera et al (2006), was
implemented in conjunction with the convection function
to fit the historical data.
Due to the limitation of the historical data (only dry
bulb temperatures were available), the heat exchange
model developed is only based on dry bulb temperatures.
The wet bulb, relative humidity and all additional
equations related to psychrometrics were not considered in
these heat exchange models. It can be argued that this will
not create an accurate model; but the objective was to
generate the best approximation possible for a curve that
fits the available dry bulb temperature profile of the
trenches. In simpler terms, this corresponds to the ‘crystal
ball’ that will learn from the historical data to predict the
heat exchange in the NHEA for the future.
3.3 Optimization model
The optimization model consists of the ventilation and heat
exchange model integrated with a genetic algorithm to
decide if each door should remain open or closed for each
period in order to minimize the temperature of the air
entering the fresh air system of the mine. The chromosome
or solution representation is a sequence of 0s and 1s, for
each trench, where a 0 represents a closed door of a trench
and 1 an open door. This decision process is run for each
period (monthly, weekly and daily) and the best result
obtained is kept. Each block is run independently to obtain
better solutions. Because the airflow is fixed at the bottom
of each block, it makes no difference if they are all run
together or independently. As mentioned before the
constraints are the minimum number of doors that have to
be open for each block: 6, 13, 6 and 8 for Blocks 1, 2, 5
and 6 respectively. This corresponds to a third of the
available doors for each block.
The objective or value function considered in this study
is the weighted average of the airflows and the
temperatures contributed by each trench. The optimization
was run for each period (monthly, weekly and daily) but
the overall goal was to minimize the temperature year
round for the entire mine. For this reason the GA was
given two different objectives: the first objective was to
achieve a target temperature set for each block; once the
target temperature was achieved, the second objective of
the GA became maintaining the temperature of the rock
mass as cool as possible for the next period.
4 Application to Creighton Mine Historical Data
Creighton Mine had three years of data stored in its system
including the dry bulb temperature and the state of the
door for each trench on a daily basis from February 2006
until November 2008. Although the data was not 100%
complete, the most complete portions of the data were
used. The first portion between February 2006 and
February 2007 was used to calibrate the heat exchange
model. The second portion between August 2007 and
December 2007 was used to test the heat exchange model
on its capacity to predict the heat exchange and the
temperatures of the NHEA. The surface or intake
temperature of the air was assumed to be the temperatures
provided for the Sudbury area by Environment Canada.
4.1 Heat exchange model results
The model developed considered convection and
conduction for each trench as two parallel processes
happening simultaneously, and was calibrated and tested
with the available data. The objective of the heat exchange
model was to minimize the error between the sensor
reading and the estimation calculated by the model during
the calibration period. The error of the calibration was
calculated and is presented in Table 1. Then based on the
296
results obtained for the calibration of the model the
temperatures for each trench were calculated for the period
between August 2007 and December 2007 on a monthly,
weekly and daily basis in order to estimate the error of the
prediction. The prediction error for each block is presented
in Tables 2, 3, 4 and 5.
Table 1 Trench calibration error (°C)
Period Type Maximum Minimum Average
Monthly 2.29 0.04 0.7
Weekly 2.06 0.08 0.86
Daily 2.24 0.09 0.91
From Table 1 it can be observed that as the aggregation
of the data decreases going from monthly to daily, the
error of the calibration increases as expected. What is
interesting to observe is that the error for the daily
calibration is under 1°C. This means that on average the
temperature of the sensor of the trenches could be
estimated up to a year ahead with an average error of 1°C.
When looking at each particular trench it can also be
observed that the maximum average error is 2.29°C and
the minimum is almost zero.
One of the initial assumptions was that the weighted
average of the airflows and temperatures of the trenches of
a block will deliver a good approximation of the
temperature of the block. Because the main objective was
to predict the temperature of the trenches and not the
temperature of the block, two errors were calculated for
the prediction error: the first one reflects how well the
weighted temperature of the trenches was approximated
(C2) and the second one measures how well the block
temperature was approximated (C3). Each block is
equipped with an additional sensor at the bottom of the
block that allows the comparison.
Table 2 Trench prediction error (°C) for Block 1
Period
Type
Range of
Temperatures Curve Fit
Error
Test
Error
Monthly -0.20°C to
9.30°C
C2 0.5 0.77
C3 0.34 0.42
Weekly -0.53°C to
9.44°C
C2 1 0.87
C3 0.4 0.41
Daily -0.56°C to
9.44°C
C2 0.52 0.91
C3 0.48 0.26
Table 3 Trench prediction error (°C) for Block 2
Period
Type
Range of
Temperatures Curve Fit
Error
Test
Error
Monthly 1.15°C to
6.65°C
C2 0.81 1.26
C3 0.86 0.48
Weekly 1.11°C to
6.78°C
C2 1.06 1.27
C3 0.97 0.66
Daily 1.09°C to
7.03°C
C2 0.83 3.37
C3 1.02 1.07
From Tables 2, 3, 4 and 5 it can be observed that
Block 1 delivered the best results with all the errors from
the different periods (monthly/weekly and daily), under 1
°C. These results were not as good for Blocks 2, 5 and 6
because Block 1 had the best quality data among the four
blocks. Block 2 delivered good results for almost all cases
with Blocks 5 and 6 performing the poorest.
Table 4 Trench prediction error (°C) for Block 5
Period
Type
Range of
Temperatures Curve Fit
Error
Test
Error
Monthly -4.68°C to
8.44°C
C2 1.79 4.13
C3 1.84 3.1
Weekly -5.51°C to
8.89°C
C2 1.84 4.14
C3 1.98 3.3
Daily -5.62°C to
8.89°C
C2 1.85 3.95
C3 1.79 2.97
Table 5 Trench prediction error (°C) for Block 6
Period
Type
Range of
Temperatures Curve Fit
Error
Test
Error
Monthly 5.00°C to
6.67°C
C2 2.22 2.08
C3 2.6 1.15
Weekly 4.49°C to
6.67°C
C2 2.1 1.64
C3 2.11 0.74
Daily 4.17°C to
6.67°C
C2 2.08 1.42
C3 2.19 0.74
Table 6 presents the solving time, required by the
computer to generate the calibration of the parameters for
each trench for each period (monthly/weekly/daily). The
results presented in this table can drastically change
depending on the CPU capacities of the computer as well
as the size of the data to be calibrated. The relationship is
near linear. Based on runtime of current results, only the
monthly and the weekly approach can be implemented on
a regular basis. The daily approach will need some
additional work to reduce the solving time to make it
practical for daily basis.
Table 6 Solving time for heat exchange model
Period Block 1 All Blocks
Monthly 25.5 min 2.4 hours
Weekly 2.3 hours 12.8 hours
Daily 15 hours 3.5 days
4.2 Optimization model results
Once the heat exchange model was built for each block
and the errors that each trench were producing provided a
measure of the accuracy of the prediction, the genetic
algorithm optimization model was integrated with the
ventilation and the heat exchange model and run for each
period (monthly/weekly and daily) from February 2006
until November 2008. The main idea was to observe the
performance of the optimization model against the results
obtained in the testing interval between August 2007 and
December 2007. The data of the testing interval was not
previously used in calibrating the models.
Two different settings were tested: in the first case the
temperature of the trenches was allowed to vary
297
unconstrained and in the second case the temperature was
constrained to a minimum of 0°C. Keeping the temperature
above 0°C reduces the formation of ice which seals off the
trench so it is no longer available for contributing to
condition the air.
Table 7 Unconstrained monthly optimization
improvement (°C)
Block Delta
°C
Error
Trench
Error
Block Improvement
1 1.91 1.15 0.42 0.76
2 4.15 1.59 0.48 2.56
5 5.37 2.95 3.1 2.27
6 2.53 1.79 1.15 0.74
Table 8 Constrained monthly optimization improvement
(°C)
Block Delta
°C
Error
Trench
Error
Block Improvement
1 1.79 1.15 0.42 0.64
2 0.92 1.59 0.48 0
5 1.85 2.95 3.1 0
6 2.52 1.79 1.15 0.73
Table 9 Unconstrained weekly optimization improvement
(°C)
Block Delta
°C
Error
Trench
Error
Block Improvement
1 2.09 1.15 0.41 0.94
2 4.02 1.56 0.66 2.46
5 6.29 2.89 3.3 2.99
6 2.38 1.54 0.74 0.84
Table 10 Constrained weekly optimization improvement
(°C)
Block Delta
°C
Error
Trench
Error
Block Improvement
1 2.33 1.15 0.41 1.18
2 0 1.56 0.66 0
5 2.69 2.89 3.3 0
6 2.38 1.54 0.74 0.84
For each block the conservative approximation of the
improvement predicted by the optimization model was
calculated as the difference between the nominal
improvement and the largest error, either C2 or C3. The
delta value of the tables is calculated as the difference
between the highest temperature registered during the
testing interval for the block and the highest temperature
calculated by the optimization model for the same interval
of time. The results obtained are presented in Tables 7, 8,
9, 10, 11 and 12 for each period.
The “unconstrained” results in Tables 7, 9 and 11 can
be used to define an upper bound on the benefits that can
be achieved by a different operation of the NHEA; and the
constrained results define a lower bound. Table 13 presents
these upper and lower boundaries of the improvements for
the full NHEA. This calculation was done again using the
weighted average of the airflows and the temperatures
obtained on each block.
Table 11 Unconstrained daily optimization improvement
(°C)
Block Delta
°C
Error
Trench
Error
Block Improvement
1 1.82 1.01 0.26 0.81
2 4.15 1.42 1.07 2.73
5 5.87 2.3 2.97 2.9
6 2.98 1.34 0.74 1.64
Table 12 Constrained daily optimization improvement (°C)
Block Delta
°C
Error
Trench
Error
Block Improvement
1 1.97 1.01 0.26 0.96
2 0 1.42 1.07 0
5 4.3 2.3 2.97 1.33
6 2.98 1.34 0.74 1.64
Table 13 Optimization improvement to NHEA (°C)
Period Type Lower Bound Upper Bound
Monthly 0.31 1.7
Weekly 0.49 1.83
Daily 0.72 1.99
As can be observed in Table 13 the daily approach
offers the largest potential improvement ranging from
0.72°C up to 1.99°C. The reason is that the optimization
model micromanages the trenches on a daily basis and thus
obtains the largest improvement. However in operational
terms this is the most difficult result to obtain because the
NHEA is not automated and it requires having someone
ready to change the state of the doors every day.
4.3 Justification of the simplified ventilation model
Because of the simplicity of the ventilation model that
considered parallel airways and a fixed quantity at the
bottom of each block, a specific ventilation model was
built into the code to solve the airflows and a VnetPC2007
model (Vale Inco Standard) was developed to confirm the
results. When both models were run the results were
exactly the same but the solving time of the specific
ventilation model was dramatically faster than the
VnetPC2007 model because it didn’t have to solve the full
298
model, only the NHEA portion and because it didn’t have
to use the Hardy Cross iterative method. Table 14 presents
the solving times for both models for each period
(monthly/weekly/daily).
Table 14 Solving time for NHEA optimization
Period Type All blocks
Parallel Airways VnetPC
Monthly 4 min 2 days
Weekly 4 min 2 days
Daily 4 min 2 days
Although the results are encouraging for the parallel
airways model it is limited to this simple configuration. If
a more complex ventilation model is built to better
represent the airflow through the NHEA, then
VnetPC2007 will be the required ventilation solver to use.
5 Conclusions and further work
Both the ventilation and the heat exchange model
developed were able to successfully generate a first
approximation to estimate the temperature of the fresh air
system (NHEA) of Creighton Mine under different
operation strategies. Additionally, an optimization model
was integrated with the ventilation and heat exchange
models to optimize the use of the NHEA year round based
on the historical data provided. The results provided by the
optimization exercise demonstrated the potential to
increase the benefit obtained from the NHEA depending
on how it is managed. Controlling the ventilation doors on
a daily basis yielded the most favorable results (lower
temperatures). Results of this study are positive and
encouraging but more work is required in order to improve
the running time of the models and the accuracy of the
approximations, to improve on the validation of the models
in the field. In addition more experiments will be required
to better estimate how much of the upper boundary for
improvement can actually be achieved.
Acknowledgements
The authors would like to express their gratitude to Vale
Inco’s staff and management for their cooperation in this
work and their permission to present the findings. The
genetic algorithm was developed using GAlib genetic
algorithm package, written by Matthew Wall at the
Massachusetts Institute of Technology. The authors would
like to thank MVS Engineering for their involvement in
the project and David Winsett for his cooperation
integrating VnetPC solver to the models developed in this
study.
References
Acuña, E. Hardcastle, S.G., Maynard, R., Fava, L., Hall, S. &
Dunn, P. (2009), The application of genetic algorithms to
multiple period ventilation systems for multi-level mine
operations. In Proceedings of the Orebody Modelling and
Strategic Mine Planning Symposium, pp 265-270.
Incropera, F., DeWitt, D., Bergman, T. & Lavine, A. (2006),
Fundamentals of heat and mass transfer, 6th edition, John
Wiley & Sons.
Johnson, O.S. (2006), Thermo- and psychrometric properties of
intake air passing through fragmented strata. In Proc. 11th
U.S./North American Mine Ventilation Symposium, pp 29-
34, Taylor & Francis Group Plc.
Lowndes, I.S., Fogarty, T., & Yang, Z.Y. (2005), The application
of genetic algorithms to optimize the performance of a mine
ventilation network: the influence of coding method and
population size. In Soft Computing, Vol. 9, pp. 493-506.
Lowndes, I.S. & Yang, Z.Y. (2004), The application of GA
optimization method to the design of practical ventilation
systems for multi-level metal mine operations. In Mining
Technology, Vol. 113, pp 45-58.
McPherson, M.J. (1993), Subsurface Ventilation and
Environmental Engineering, London: Chapman & Hall.
299
300
... Simpler analytical models have also been developed to find better ways to operate these systems Such is the case of the work by Acuña et al. (2010) and Saeidi, Romero, Fava, and Allen (2017), which, using different simplifications, explored the improvement of the thermal storage process through, respectively, operation optimization or design changes. These studies base their lighter computational effort with respect to CFD on the use of a reduced number of nodes where the balance equations are defined. ...
Conference Paper
Most underground mines require heating or cooling systems to meet health and safety standards in the working areas. In regions where daily or seasonal fluctuations of environmental conditions exist, these systems tend to be over-sized and operate at low average load factors, which increase both investment and operating costs. One possible approach to reduce the temperature variations of the fresh air supplied to the mine, and so reducing the investment in large heating or cooling systems, is to use rock as a thermal regenerator or thermal storage system. Packed rock beds act as thermal inertial units, damping the atmospheric temperature oscillations to provide a more constant fresh air temperature. However, although regenerators exist in some mines, and packed bed rocks have been studied in detail in solar thermal power applications, there is a lack of tools to design new and sufficiently large storage systems of this kind. This work fills the existing gap by presenting a methodology to design and evaluate the impact on the ventilation systems of these regenerators. The method is validated using existing rock bed systems found in the literature and illustrates possible designs for small (daily attemperation) and large size (seasonal attemperation) regenerators in both hot and cold climates.
... Simpler analytical models have also been developed to find better ways to operate these systems Such is the case of the work by Acuña et al. (2010) and Saeidi, Romero, Fava, and Allen (2017), which, using different simplifications, explored the improvement of the thermal storage process through, respectively, operation optimization or design changes. These studies base their lighter computational effort with respect to CFD on the use of a reduced number of nodes where the balance equations are defined. ...
Article
Most underground mines require heating or cooling systems to meet health and safety standards in the working areas. In regions where daily or seasonal fluctuations of environmental conditions exist, these systems tend to be over-sized and operate at low average load factors, which increase both investment and operating costs. One possible approach to reduce the temperature variations of the fresh air supplied to the mine, and so reducing the investment in large heating or cooling systems, is to use rock as a thermal regenerator or thermal storage system. Packed rock beds act as thermal inertial units, damping the atmospheric temperature oscillations to provide a more constant fresh air temperature. However, although regenerators exist in some mines, and packed bed rocks have been studied in detail in solar thermal power applications, there is a lack of tools to design new and sufficiently large storage systems of this kind. This work fills the existing gap by presenting a methodology to design and evaluate the impact on the ventilation systems of these regenerators. The method is validated using existing rock bed systems found in the literature and illustrates possible designs for small (daily attemperation) and large size (seasonal attemperation) regenerators in both hot and cold climates.
... The Natural Heat Exchange Area (NHEA) at Creighton Mine is an old sublevel cave area that is used to condition the incoming air from surface that goes into the three main fresh air systems of the mine. As presented by [1] the area is divided into trenches which are considered independent or parallel entrances for the airflow from surface. The air at ambient surface temperature passes through the fragmented rock mass sitting on the top of each trench where the heat exchange occurs. ...
Conference Paper
Full-text available
Creighton Mine has more than 100 years of production history and is continuing to develop deeper with the plan to reach 10,000 feet. The deeper levels of the mine have already reached 8000 ft level without the use of a mechanical refrigeration system. This has been achieved through the use of the Natural Heat Exchange Area (NHEA) implemented in the early 1960's. The NHEA is a remaining sublevel caving area mainly divided into 4 blocks with slusher trenches that were used to extract the ore. Today, it serves the purpose of conditioning the outside air temperature before being fed into the fresh air system of the mine. As Creighton mine progresses deeper, the temperature on the lowest levels will approach the limit of Vale's guidelines for working in heat. This paper presents the study developed to analyze the possibility of implementing modifications to the gathering area of the NHEA. The objective was to improve the distribution of the cooled air to the deeper levels and outline the impact that this change can have in terms of the additional levels that could be operated underground without the use of mechanical refrigeration. This study supports the efforts currently under development for energy conservation and to reduce the capital and operational costs of Creighton Mine while extending the expected life of the mine.
Conference Paper
Full-text available
Historically, the methodology for ventilation optimisation has been based on case studies and expert knowledge. It has yet to significantly benefit from optimisation techniques used and proven in other fields of engineering. The results from this non-mathematical approach are demonstrably suboptimal in terms of energy consumption, because of the focus on supplying the peak demand over a period of time. While mine planners generate production schedules to extract the ore efficiently, they cannot be sure if the ventilation support required for the scheduled activities will be in place at the necessary time. This generally forces a deviation from the mining schedule, leading to a lower Net Present Value (NPV) resulting from development and production losses. This paper outlines the architecture of a multiple period mine ventilation tool to optimise the ventilation infrastructure and its operation for multilevel mines, to be later used by the planning team or to be integrated with a scheduling tool. The objective of the proposed approach is to improve the overall NPV of schedules by accounting for ventilation requirements. The improvement in the NPV can be realised in two main areas: first, allowing the schedules to be executed as planned, reducing (or avoiding) development and production losses, and, second, reducing ventilation costs. This paper presents a comparison between single period and multiple period approaches to improving mine ventilation systems. An application to a real size problem is presented and solved with both approaches, and then compared in terms of the solving time and the values obtained through each approach.
Article
This paper presents the results of a series of recent studies to investigate the improved optimal design and planning of the main ventilation systems of complex multi-level mine networks using a genetic algorithm (GA) routine. It is concluded that the method may be successfully used to investigate the benefits offered by alternative ventilation configurations and to determine the most practical and cost-effective ventilation system during the various mine planning stages (short-, medium- and long-term). In particular, the optimum number, location and duties of booster fans, and size and location of regulators may be evaluated and selected. The objective of the final ventilation design is to identify the ventilation system that operates in both a safe and energy-efficient manner. The method has been refined to include the selection of practical fan units that match the optimal fan operating duties determined by the GA routine. The paper also reviews general optimisation strategies that may be employed to enhance and maintain the performance of mine ventilation systems.
Article
This paper presents an application of genetic algorithms (GAs) to the solution of a real-world optimisation problem. The proposed GA method investigates the optimisation of a mine ventilation system to minimise the operational fan power costs by the determination of the most effective combination of the fan operational duties and locations. The paper examines the influence that both the encoding method and the population size have on the performance of the GA. The relative performance of the GA produced by the use of two different encoding methods (a binary and a hybrid code) and various solution population sizes is assessed by performing a two way ANOVA analysis. It is concluded that the genetic algorithm approach offers both an effective and efficient optimisation method in the selection and evaluation of the cost-effective solutions in the planning and operation of mine ventilation systems.