University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Colorado School of Mines
|
Although the results were site-specific, it clearly shows the relation between longer face
lengths and higher cumulative methane emissions at the tailgate corner.
The recent trend of extending face length can change this ratio. A study by Diamond and
Garcia (1999), Krog et al. (2006) and Schatzel et al. (2006) all suggest the trend of mining longer
longwall faces to increase productivity will also result in higher methane accumulations at the
tailgate corner. Schatzel et al. (2006) and Krog et al. (2006) produced empirical models to
estimate methane emissions in longwall face for mine operating in the Pittsburgh coalbed. This
estimation is based on a three-day ventilation survey study on a 300 m wide longwall face. This
data was then used to predict methane emissions on longer longwall faces, as shown in Figure
2.39.
Figure 2.39: Prediction curve for methane emissions in the Pittsburgh coalbed (Schatzel et al.,
2006, public domain)
A notable result from this study is that the cumulative methane increases seem to follow a
linear regression trend, with approximately a 25% increase in methane emissions for every 61 m
of increase in longwall face width. Krause (2015) described a method to estimate the amount of
methane emitting to the working face based on the shearer’s cutting cycle time.
45
|
Colorado School of Mines
|
(2.2)
𝐿𝑠𝑚𝑒𝛾𝑧𝑀𝑜𝜂𝑠
𝑠𝑐
𝑉𝐶𝐻4
= 100𝑡
Where:
= predicted volume of methane emitted to the longwall working per 1 min during the cycle
𝑠𝑐
o 𝑉f𝐶 𝐻c4utting with a shearer with time t, m3CH 4/min
= Length of the longwall, m
= Height of the mined longwall, m
𝐿𝑠
= Coal density, mg/min3
𝑚𝑒
= Shearer web depth, m
𝛾
= Initial methane content of the cut seam in the longwall, m3CH /min
𝑧 4
= Degassing degree of the mined seam, %
𝑀𝑜
= Duration of cutting cycle, min
𝜂𝑠
The relationship between the degassing degree of the mined seam can be estimated based on
𝑡
the methane content, which can be seen in Figure 2.40.
Figure 2.40: Degassing degree of the mined seam while being cut with a shearer, based on its
methane content (Krause, 2009, used with permission)
It is important to note this experiment was performed in U- and Y-type ventilation systems,
and the results may not be applicable for bleeder ventilation systems due to differences in
pressure and airflow distribution across the face. Higher methane outgassing can be expected
46
|
Colorado School of Mines
|
with bleeder ventilation due to the higher pressure drop between the in-situ methane source and
the longwall face.
To estimate a coal’s methane, an operator can perform direct measurement tests by vertical
drilling from the surface or using in-seam horizontal drilling during operations. If direct testing is
not feasible, gas content can be roughly estimated based on the depth and rank of the coal.
However, this estimation usually results in over prediction of the methane content (Kissell,
2006). Figure 2.41 shows methane content estimations for different ranked coals at various
depths based on hydrostatic head assumption.
Figure 2.41: Indirect method to estimate methane content based on depth for different coal ranks
(Kissell, 2006, public domain)
2.4.2 Methane inflow from the gob area
In a longwall operation, methane may come from three main locations: the seam being
mined, lower coal seams and the upper seam (the rider seam). Lidin (1961), Thakur (1981),
Winter (1975) and Gunther and Belin (1967) have all conducted studies to determine the extent
of gas emissions from surrounding gas sources to the working seam, as shown in Figure 2.42.
47
|
Colorado School of Mines
|
Figure 2.42: Extent of gas emissions space within the gob (Kissell, 2006, public domain)
Based on the graph, the working seam remained the primary source of gas emissions found
at the working mine level. Additional gas inflow can come from the surrounding coal seams,
with the rider seam having more influence on gas emissions compared to the seam below the
mine floor. Range of influence within the surrounding coal seam can vary from 160 m to 300 m
above the working seam and 40 m to 80 m below it.
A study by Krause (2015) identified the relationship between longwall face length, seam
inclination and the extent of coal degassing from the surrounding strata. The results of this study,
as shown in Figure 2.43, suggest that for the typical face length of 300 m to 400 m, the extent of
methane inflow from the surrounding strata can be estimated to be between 200 m and 270 m
from above and 60 m and 100 m from below the extracted coal seam.
48
|
Colorado School of Mines
|
Figure 2.43: Degassing range of coal depending on the longwall length and inclination (Krause,
2015, after Krause & Lukowicz, 2000, used with permission)
Assuming there is enough information on the gas contents of surrounding coal seam layers,
the results of these studies (Kissel, 2006; Krause, 2015) can be used to estimate the amount of
methane produced in the gob area as a longwall is mined.
Methane inflow from the gob is generally not a concern in bleeder ventilation systems, as
the bleeder fan pulls incoming gas toward the back of the panel. However, mine operators must
be aware of factors that can push the methane-air gas mixtures from the gob back to the face,
such as immediate roof caving condition and changes in tailgate ventilation conditions. These
phenomena were demonstrated in the UBB mine disaster in 2010 and modeled using the CFD
approach by Brune and Sapko (2012) and Juganda et al. (2017). Other factors, such as sudden
drops in barometric pressure, can also result in methane gas migrating from the gob into the
surrounding bleeder entries (Lolon et al, 2017).
Methane Monitoring Practices in U.S. Longwall Coal Mines
2.5.1 Monitoring location
Monitoring working conditions is important for underground operations, particularly coal
mines where methane emissions and ignition always pose a threat. The number of sensors, sensor
placement and concentration limits all play a vital role in successful methane monitoring
practices.
49
|
Colorado School of Mines
|
Currently, point-based monitoring practices are used to assess ventilation conditions at the
longwall face. Assessing an explosion hazard based on a point-type instrument reading, however,
can be misleading, as it is highly dependent on the location being measured. MSHA requires
all methane concentration testing be made at least 0.3 m (12 inches) from the roof, face, ribs and
floor (30 CFR §75.351), primarily because methane often enters the mine workings as a
localized source at a high concentration. Figure 2.44 illustrates how methane gas enters mine
workings through cracks or fissures in the coal, roof or floor rock, and is diluted by the moving
air stream.
Figure 2.44: Illustration of methane being diluted in a moving air stream (Kissell, 2006, public
domain)
This illustration shows how measuring locations can have significant impact on methane
readings. One can expect to receive high readings if measuring close to the source, especially
since there is less air flowing in the boundary layer to dilute it. As methane is lighter than air,
emissions from the roof can form stratification layers with higher concentrations near the roof.
Several studies have determined the minimum required air velocity to prevent methane layering.
Raine (1960) found that an air velocity of 0.51 m/s (100 ft/min) measured at the roof is sufficient
to prevent layering. Further study by Bakke and Leach (1962) suggested the 0.51 m/s
(100 ft/min) requirement was only applicable for horizontal entries, and that a higher air velocity
is required for inclined entries. A follow-up laboratory study by Bakke and Leach (1962) found
the minimum air velocity is dependent upon the methane release rate; they thus proposed an
equation for estimating the minimum required air velocity to prevent layering, called a “layering
number” that is expressed as follows:
50
|
Colorado School of Mines
|
(2.3)
𝑈
𝐿 = 3 𝑉
Where L is the layering number (dimen3si7o.n√ le𝑊ss), U is the air velocity (feet per minute), V is
the methane release rate (cubic feet per minute), and W is the entry width (feet).
According to Blake and Leach, gas mixing by turbulence begins at layering numbers larger
than 2, and a minimum layering number of 5 is required for adequate dilution. These numbers
are based on testing to determine required minimum air velocity to dilute the methane layer
released from a single point at the roof. The main issue with this approach is that the methane
release rate is not known until a proper ventilation survey is conducted. Figure 2.45 shows the
experimental results for different methane source locations. A more recent study by McPherson
(2002) estimated a minimum air velocity of 0.4 m/sec or 80 ft/min, in agreement with previous
study results.
Figure 2.45: Methane layering test results for different source locations (Taylor et al., 2010,
public domain)
Past studies performed on ventilation patterns and methane emissions at longwall faces have
suggested the majority of face methane stems from coal breakage by the shearer (Cecala et al.,
1985a, 1989; Denk and Wirth, 1991). Two methane monitors are usually installed at the longwall
face, one mounted at the shearer and the other near the tailgate side of the face. The shearer
methane concentration reading is generally higher than that at the tailgate. An experiment
51
|
Colorado School of Mines
|
conducted by Kissell and Cecala, et al. (2006) showed that during the tail-to-head pass, methane
concentrations at the shearer would exceed 1.0% several times, but no methane concentration
above 1.0% was recorded by the tailgate methane monitor. This suggests that the methane
concentration recorded by the tailgate monitor may not be a good representation of face
conditions.
The shearer is a primary ignition source for longwall operations, making the placement
location of the methane monitor on the shearer body critical. Unlike portable handheld detectors,
where a peak emission can be easily missed because of infrequent reading intervals, machine-
mounted monitors are expected to operate continuously and must be able to identify emission
peaks and automatically de-energize electrical equipment when levels exceed prescribed limits.
Cecala et al. (1993) conducted a full-scale laboratory study to determine the best methane
monitoring location on the shearer and the effect of turbulence caused by water sprays on
methane readings, as shown in Figure 2.46.
Figure 2.46: A diagram of experimental methane monitoring locations (Cecala et al., 1993,
public domain)
52
|
Colorado School of Mines
|
In this experiment, the shearer is set up to simulate a tail-to-head cutting sequence. A single
fan was used to deliver 6.6 m3/s (14,000 cfm) of fresh air at 1.3 m/s (250 ft/min) velocity, while
7.5 L/min of methane gas at a 93% concentration was continuously released from eight points in
a simulated sump coal block at 207 kPa (30 PSIG). Pencil point water sprays that delivered
60 L/min (16 gpm) of water at 760 kPa (110 PSI) were used in this experiment and were directed
toward the head side drum. For gas sampling, four, 2 L/min (0.5 gpm) constant flow pumps were
used to draw air samples of 8 L/min (2 gpm) from each location through 8 m (26 ft) of 10 mm
(3/8-in) semi-rigid tubing. Table 2.4 shows a comparison of the recorded methane concentrations
at these different monitoring locations, with and without the water sprays.
Table 2.4: Recorded methane concentrations at different monitoring locations, with and without
the use of water sprays (Remake from Cecala et al., 1993)
Test #1 Test #2
CH Concentration CH Concentration
4 4
Water On Water Off Water On
Location Location
ppm SD ppm SD ppm SD
1 36.5 11.3 45.5 14.2 4 77.7 31.4
2 27.5 9.3 27.1 8.2 4A 12.6 4.4
3 36.2 9.6 29.0 7.6 6 60.1 16.2
4 82.9 25.2 43,5 10.8 6A 22.3 8.1
5 65.3 17.4 48.2 9.3 8 72.8 9.8
6 74.8 19.0 46.5 13.1 8A 26.8 7.5
7 84.7 10.6 42.0 7.0 10 70.8 5.7
8 82.9 13.0 38.0 7.3 10A 40.7 3.7
9 80.6 11.7 40.0 6.3 11 8.7 1.5
10 76.7 7.6 42.0 5.1 12 33.8 9.4
53
|
Colorado School of Mines
|
A notable result from this experiment was that, compared to the no-spray case, use of this
type of water spray generated substantial turbulence, which resulted in higher methane
concentrations at the 10 gas sampling locations. The result of the second test also showed the
importance of sensor placement, as all four sampling locations (4A, 6A, 8A and 10A) located
slightly further back from the rest of the locations only managed to pick up about half of the
concentrations recorded by their counterparts (4, 6, 8 and 10).
However, since these four monitor locations were farther from the coal face, they were less
prone to damage or covered with coal dust and soaked by water sprays. It was considered to be
the better monitoring location choice, even though it returned a much lower methane
concentration reading compared to the other 10 locations. Thus, out of the 14 tested sampling
locations, Cecala et al. (1993) suggested the optimum location for a single shearer-mounted
methane sensor to be the top face side of the shearer body, at least 1.8 m (6 feet) downstream of
the head side cowl. The results of this study are still currently being used as guidelines for
shearer-mounted sensor placement.
A more recent CFD modeling study by Wang et al. (2017) on methane flow characteristics
on a longwall face with a progressively sealed ventilation system outlined that readings obtained
from a shearer monitor can be approximately two to three times lower than concentrations
around the TG drum. While this study is based on a progressively sealed ventilation system
commonly used in Australia, these results can be used as a general reference when analyzing
methane distribution in bleeder ventilation systems, especially for the headgate and middle
section of the longwall face where both systems show similar flow patterns.
Figure 2.47 illustrates a cross-section view of methane distribution inside the face area at
four different locations. The results also show methane concentration contrasts between the coal
face and where the methane sensor is typically located can vary significantly along the longwall
face. This case emphasizes the need to re-evaluate sensor placement and effectiveness when
assessing the explosion hazard at the longwall face.
54
|
Colorado School of Mines
|
Figure 2.47: Methane concentration distribution across the face at 2 m above the floor (Wang et
al., 2017, used with permission)
2.5.2 Types of methane monitors
MSHA categorized methane monitoring practices into two categories, methane monitors and
methane detectors, based on their use and placement [30 CFR §75.342]. Methane monitors
should be permanently mounted to provide continuous readings near the face. This sensor must
be located as close to the face as practicable, but at least 0.3 m (12 in.) from the roof, face, ribs
and floor [30 CFR §75.342(a)(3) and 30 CFR §75.323(a)]. It should also be equipped with a
warning signal to alert workers when methane concentrations reach 1.0% [30 CFR
§75.342(b)(1)] and 2.0% methane by volume [30 CFR §75.342(c)(1)], and an electrical relay to
cut power to the mining machine, in this case the longwall shearer, when methane concentrations
reach 2.0% [30 CFR §75.342(c)(1)]. As discussed earlier, under certain conditions MSHA may
grant exceptions to the 1.0% and 1.5% statutory limits, raising them to 1.5% and 2.0%,
respectively. A study by National Institute for Occupational Safety and Health (NIOSH) (Taylor
et al., 2010) found that a good monitor placement should be six to eight feet from the face and on
the return air side where concentrations are usually highest, and to reduce the risk of the sensor
head being damaged from falling rock and water sprays.
55
|
Colorado School of Mines
|
Methane detectors are generally small, portable, battery powered and designed to be worn on
the worker’s clothing. They are used to make periodic gas checks at each active mining face
[30 CFR §75.362)(d)(1)(iii)] and at outby locations at a distance of at least 30 cm (12 inches)
from the roof, face, ribs and floor [30 CFR §75.323(a)].
2.5.3 Types of sensors, calibration and response time
In general, there are two types of sensor heads used in underground coal operations: the
combustible catalytic bead sensor and the infrared sensor. The catalytic sensor includes a very
fine platinum wire contained within an alumina bead coated with a catalyst material, typically
platinum or palladium. This sensor operates by heating up the active bead until it reaches a
temperature sufficient to promote combustion of the methane-oxygen mixtures on the catalyst
surface. The generated heat increases the resistance of the wire inside the bead, this change is
detected by a Wheatstone bridge circuit to determine the methane concentration. The other
inactive bead in the Wheatstone bridge, which is not treated with the catalyst material, is used to
compensate for changes in temperature, pressure and humidity (Taylor et al., 2010). Figure 2.48
shows the components of a catalytic bead sensor.
Figure 2.48: Schematic of catalytic bead sensor components (Taylor et al., 2010, public domain)
A major limitation with this sensor type is that it is only accurate for methane concentrations
below 8%, and also requires an oxygen concentration above 10% (Taylor et al., 2010). Above
8%, instruments with catalytic heat of combustion sensors are no longer reliable.
56
|
Colorado School of Mines
|
Most methane monitors used in coal mines use catalytic heat of combustion sensors, but some
infrared monitors and detectors have been approved by MSHA for underground use. Figure 2.49
shows the components of an infrared sensor.
Figure 2.49: Schematic of infrared sensor components (Taylor et al., 2010, public domain)
Infrared sensors utilize the property of methane gas that absorb infrared light at certain
wavelengths (e.g., 1.33, 1.66, 3.3 and 7.6 µm; Taylor et al., 2010). Optical absorption filters are
used to limit the infrared light, narrowing the wave band absorbed by the measured gas. The
detector will then measure the intensity of the filtered light, which varies inversely with the
methane concentration.
Unlike the catalytic bead type, infrared sensors can accurately measure methane
concentrations without the minimum oxygen requirement, and in a concentration range of up to
100% methane. However, the sensor is highly sensitive to the filter used in the infrared
instrument. Without a proper filter, wavelengths of the transmitted light absorbed by ethane and
higher hydrocarbons can produce an exaggerated infrared detector response. This type of sensor
is suitable for restricted areas where methane concentrations are expected to exceed the operating
limit of the catalytic combustion type.
57
|
Colorado School of Mines
|
MSHA regulations require that all methane monitors and detectors be calibrated at least
once every 31 days [30 CFR §75.342(a)(4)] according to procedures specified by the monitor
manufacturer. The calibration process involves flooding the sensor head with calibration gas
(usually 2.5% methane by volume) and recording the value read by the instrument. Generally, a
methanometer that measures concentrations within 0.1% of the “zero” gas and within 0.2% of
the calibration gas, with a visual display that stabilizes within two minutes after the application
of calibration gas, is considered to be properly calibrated (Taylor et al., 2010). Other useful
reference values for calibration are 40% response time (1% concentration) and 80% response time
(2% concentration), which correspond to the warning device concentration limits prescribed in the
regulations (30 CFR §75.342(b) and (c)). To be used in underground operations, the sensor head
must also be covered with a dust cap to protect the sensor elements or sensor chamber from dust
or water (30 CFR §27.22).
Based on a study by Kissell (2006), the most important factors that determine a machine-
mounted monitor response time are the sample transport time, which is defined as the time a
volume of methane released from the coal face takes to reach the monitor sensor head, and the
monitor response time, the time it takes for the monitor to respond to the methane after it reaches
the monitor sensor head. Currently, there are no MSHA written response time criteria for
methanometers mounted on mining machines.
NIOSH has conducted extensive studies on the performance of methane sensors typically
used in the mining industry and have identified factors affecting their performance and response
time. Results show the sensor response time is highly affected by the gas flow rate (increasing
flow rate through the sensor head will reduce response time) and the dust cap design (how far the
calibration gas has to travel internally to reach the sensor head). To study the sensors’ response
time, Taylor et al. (2010) performed tests with a calibration cup typically used in underground
mining operations, as shown in Figure 2.50, and a test box in a laboratory environment, which is
shown in Figure 2.51.
58
|
Colorado School of Mines
|
The tests were conducted with five different sensors, three of the catalytic type and two of
the infrared type, and all MSHA-approved sensors commonly used in the industry. Based on the
test results, the catalytic bead-type sensor has a response time of approximately 30 to 40 seconds
when tested with calibration cup, and approximately 25 seconds when tested with the test box
and attached dust cap. The response time further decreased to within a range of 4 to 14 seconds
when the sensors were tested without the dust cap.
The response time for infrared sensors ranged from 10 to 33 seconds when tested in the test
box with dust cap attached, and reduced to three seconds without dust cap. The important thing
to consider is that in active mining applications, the calibration cup is more practical for use but
is not as accurate as the test box, and the dust cap should not be removed. In addition, the test
results with the test box are a better representation as they do not have additional restrictions
placed by the calibration cup that requires airflow to travel a greater distance to reach the sensor
head, as previously shown in Figure 2.50. The results of these tests suggest that the dust cap
design is more important in determining response time than the type of sensor.
Figure 2.52: Example of infrared sensor heads with dust caps (Taylor et al., 2010, public
domain)
For comparison, wearable methane detectors typically use catalytic heat of combustion
sensors that measure methane levels up to 2.5% by volume. Methane detectors usually have
faster response times than machine-mounted methane monitors, with 90% response times varied
from 8 to 20 seconds. However, person-wearable devices have limited filtering capabilities for
removing airborne water and dust compared to the machine-mounted methanometer (Taylor et
al., 2010).
60
|
Colorado School of Mines
|
CHAPTER 3
CFD MODEL DEVELOPMENT AND VALIDATION
CFD models of longwall bleeder ventilation systems were developed to analyze airflow
patterns, methane gas distribution and explosion-prone areas. The longwall face working area,
especially when the shearer is cutting coal at the tailgate corner areas, are the focus of this
research as it is where the ignition is most likely to occur, as previously referenced in the
background study. These two areas were chosen as the critical locations where the explosive
mixture is most likely to form because of high turbulence and a lack of fresh air to dilute
incoming methane gas.
To ensure the simulation results closely resemble real mine conditions, the longwall face
model includes the operational components typically found in a longwall operation: a shearer,
rotating shearer drums, a stageloader, face conveyor, shield supports, face curtain, gob plate and
the headgate and tailgate drives, along with rotating shearer drums to simulate cutting action.
The detail on the modeled longwall component is presented on the model geometry section. the
Different shearer locations and ventilation scenarios are tested, and the results is presented in the
following section of this report. The resulting models were used to analyze the effectiveness of
point-based methane monitoring currently practiced in the industry and provide
recommendations to improve the reliability of the current methane monitoring practice to detect
explosion hazard at the longwall face.
Another important aspect of this research is the longwall gob flow model, which governs
leakage across the face. Extensive mine ventilation survey measurements (Peng and Chiang,
1986; Thakur, 2006; Krog et al., 2006; Schatzel et al. 2006 & 2012), , a physical scaled model
(Gangrade et al., 2017 & 2019) have demonstrated that gob characteristics such as permeability
and porosity distribution, along with roof caving conditions, will govern the airflow and leakage
patterns in the working face. However, gob characteristics are site-specific and can vary greatly
depending on overburden conditions.
Several studies have been done by other researchers (Esterhuizen & Karacan, 2007; Marts et
al., 2014; Wedding, 2014) to determine the permeability and porosity of longwall gob. Although
the range of gob permeability and porosity distribution can vary greatly depending on the
overburden caving characteristics at the sites tested, results of these studies show similar trends
61
|
Colorado School of Mines
|
regarding gob flow patterns, with the highest permeability and porosity behind the face shields
and around the gob edges. The gobs become significantly less permeable towards the center.
Since gob characteristic development and validation is not part of this study, different gob caving
characteristics, porosity and permeability distributions found by other researchers will be used as
a part of parametric study to ensure that the general trend of the simulation results holds true. It
should be noted that, due to varying characteristics at mines, the resulting models are not
intended to give an exact prediction, but rather offer recommendations based on the trends
observed in the simulation results.
Some preliminary work has been conducted to establish modeling approaches and
parameters that would support the development of a longwall face model that represents the
ventilation conditions at a real operating mine. These studies consist of testing methane gas
inflow method, wall roughness adjustments, testing the modeling approach for rotating shearer
drums, and walkaway and gob curtain setup.
Preliminary CFD Studies
3.1.1 Effect of gob characteristics and immediate roof caving condition
Previous research and case studies done by other researchers have indicated that factors such
as immediate roof caving conditions, gob permeability, and tailgate ventilation set-up can
significantly impact the airflow pattern across the longwall face area. Bleeder ventilation relies
on leakage from the face area to push explosive methane-air mixtures away from the active roof
caving area immediately behind the longwall face, which can extend up to 45 m behind the
shields (Thakur, 2006). Unfortunately, this face leakage is not controllable by the mine operator,
since it is highly dependent on gob permeability and immediate roof caving (Krog et al., 2014;
Gangrade et al., 2019). Figure 3.1 shows a schematic of a longwall face with uncaved, immediate
roof directly behind the shields.
A preliminary CFD study has been done to analyze impact of caving conditions behind the
shields on face and tailgate ventilation, along with the phenomenon of methane in- and
outgassing from the gob to the face due to roof falls in the tailgate inby the face.
62
|
Colorado School of Mines
|
Figure 3.1: Cross-section view of a longwall face and immediate roof area behind the shields
The geometry and caving conditions used for this study were based on a previous study by
Juganda (2017) to analyze specific ventilation conditions at the Upper Big Branch mine just prior
to the 2010 explosion. The modeled panel is 1 km long, 300 m wide and the coal seam is 2.2 m
high. The gateroad pillar dimensions are 24.4 m by 18.3 m. Gateroad dimensions are 6 m by
2.2 m, consisting of three headgate entries, a belt entry and three tailgate entries. The gob and
fracture zone heights are 10 m and 7 m, respectively. The shearer is cutting out the tailgate
corner of the longwall face. In this test, the shearer drums were modeled as simple, cylindrical
drums.
Two immediate roof caving scenarios were considered in this study. Good caving conditions
represent a case where the immediate roof fully caves immediately behind the shields while poor
caving conditions represent a case where the roof directly behind the shields hangs up and does
not fully cave, resulting in a void behind the shields that extends up to 15 m inby the face on the
tailgate side. Figure 3.2 shows the ventilation network of the modeled longwall panel. It should
be noted that investigators observed a similar, poor caving pattern after the Upper Big Branch
Mine explosion (Phillips, 2012).
The longwall face is supported by 176 shields. Each shield is 7 m long, 1.8 m wide and
2.2 m high. The exceptions are the first and last three shields, which have longer canopies to
accommodate the headgate and tailgate drive. On the back of each shield, there is a ~0.27 m2
opening that allows air to exit and enter the face. On the headgate side, a ventilation curtain
extends from the rib of the chain pillar to about shield no. 6.
63
|
Colorado School of Mines
|
Figure 3.2: Plan view of a CFD model for a study with immediate roof caving conditions
The gob characteristics used in this study were based on a model developed by Marts et al.
(2014). Gob porosity ranged from 14% in the gob center to 40% near the face, while gob
permeability ranged from 2x10-8 to 7x10-7 m2. The gob fringes at the headgate and tailgate sides
of the gob were also modeled as porous media with uniform porosity of 50% and a permeability
value of 2x10-7 m2.
A bleeder ventilation system was used, delivering ~47 m3/s of fresh air to the two headgate
entries. 9.4 m3/s was assumed to leak through the headgate curtains, and 4.7 m3/s of air was
returned through the belt entry, resulting in ~33 m3/s of air delivered to the longwall face. Each
of the three tailgate entries supplies 4.7 m3/s of fresh air. Figure 3.3 shows the results of the CFD
simulations of airflow distribution across the face for the two roof caving conditions.
Figure 3.3: Comparison of airflow velocity contours for the two caving conditions
64
|
Colorado School of Mines
|
At the headgate side (HG), the models show that after entering the longwall face, some of
the air begins to leak through the shield gaps as it travels along the face because of the bleeder
fan-generated pressure drop at the back of the panel. The poor caving conditions leave more
open space behind the shields, providing a pathway for the leaked air and reducing the face
quantity. Figure 3.4 shows the comparison of airflow quantities inside the longwall face for the
two immediate roof caving conditions.
Figure 3.4: Comparison of airflow leakage rates for the two caving conditions
Airflow distribution across the face reveals minimum leakage between shields no. 1 and 6
due to the face ventilation curtain. However, significant leakage occurred starting at shield no. 7
through the last shield. Under poor caving conditions, the mine may not meet the minimum
requirement of 14 m3/s (30,000 cfm) of face air per 30 CFR §75.325, which could cause
insufficient face ventilation and an explosion risk around the tailgate corner area. The extent of
air leakage is highly dependent on caving conditions behind the shields, the pressure drop across
the face and the ventilation system in use. In progressively sealed gobs, the fresh air that leaks
into the gob at the headgate corner and along the face will be pulled back into the face through
the shield gaps as it approaches the tailgate. In bleeder systems, the air leaking into the gob does
not return to the face, but rather flows directly toward the bleeder fan at the back.
65
|
Colorado School of Mines
|
The CFD modeling results also indicate major air leakage near the headgate corner. A
portion of this leaked fresh air flows through the high permeability area along the gob’s headgate
side, then enters back into the headgate entry and mixes with fresh air that passed though the
headgate curtain. The remainder of leaked fresh air toward the tailgate flows either along the gob
void behind the shields of the face, or through the gob directly toward the back of the panel.
Depending on the immediate roof caving conditions, most supplied air may leak through the
shield gaps and flow behind the shields parallel to the face, leaving only a small quantity of fresh
air to flow inside the last couple of shields where the shearer is located.
In addition to the caving conditions, compaction of the gob in this area also has a significant
impact on the leakage occurring near the headgate corner area. Figure 3.5 shows the airflow
leakage at the headgate corner area, while Figure 3.6 and Figure 3.7 show the airflow
distributions across the longwall face for the two different immediate roof caving scenarios and
three gob permeabilities.
The results show that gob permeability impacts are more significant in the fully caved
immediate roof condition compared to the poorly caved condition. While the immediate roof
caving condition is a factor that cannot be controlled by the mine operator, it is important to
understand the impact it can have on longwall face and tailgate ventilation.
Figure 3.5: A comparison of airflow leakage at the headgate corner in good caving (left) and
poor caving (right) immediate roof
66
|
Colorado School of Mines
|
forming directly behind the shields. This is a known safety issue with bleeder ventilated gobs.
Since active caving occurs in the area behind the shields, there is a hazard of rock-on-rock or
rock-on-steel frictional ignition. The gob has much higher porosity and permeability around its
fringes compared to the gob center (Esterhuizen & Karacan, 2007; Marts et al., 2014; Wedding,
2014), and most face-to-gob leakage occurs near the headgate and tailgate corners of the
longwall face. Figure 3.8 and Figure 3.9 show results of CFD modeling of methane
concentrations inside the gob for different gob permeabilities. The base gob permeability value,
used in Figure 3.8 is based on a study by Marts et al. (2014), while Figure 3.9 shows the result
for a scenario of higher gob resistance, but for the same amount of methane inflow into the gob.
For better visualization of explosive gas mixtures, only gas mixtures within the explosive range
of approximately 5.5% to 14% CH by volume are shown. It is assumed that the oxygen
4
concentration is above the minimum required for explosive conditions.
Figure 3.8: Volume rendering of methane mole fraction inside the gob, shown from overview
(top) and cross-section views (bottom) for gob permeability 6.9 x 10-6 m2 (edge) to 2.0 x 10-7 m2
(center)
68
|
Colorado School of Mines
|
Figure 3.9: Volume rendering of methane mole fraction inside the gob, shown from overview
(top) and cross-section views (bottom) for gob permeability 6.9 x 10-7 m2 (edge) to 2.0 x 10-8 m2
(center)
In Figure 3.8 fresh air that leaks from the face dilutes the methane concentration inside the
gob below the explosive range, immediately behind the shields. In and Figure 3.9, due to higher
gob resistance, the same amount of supplied fresh air only manages to push the explosive
mixtures away from the face in the headgate and tailgate corner sides of the gob, while allowing
the formation of explosive gas mixtures directly behind the shields in the center of the face. This
poses an explosion and fire risk for miners working in the face, as the immediate roof caves
following the face advance. Other factors, such as gob ventilation borehole failures (Brune and
Saki, 2017), ventilation control failures (Brune and Sapko, 2012; Juganda et al., 2017) or sudden
drops of barometric pressure (Lolon et al., 2016) can all potentially lead to methane outgassing
from the gob to the surrounding crosscut or longwall face.
69
|
Colorado School of Mines
|
3.1.2 Modeling methane gas emanating from the coal face
In this study, only methane released from the uncut coal face is modeled. It is assumed that
the coal seam supplies methane at the boundary in an evenly distributed manner from an
“infinitely” large reservoir . A simple geometry, shown in Figure 3.10, was used to test different
modeling methods for simulating methane emanating from the coal face. This model only
contains open, rectangular mine workings, without any obstructions. The tailgate side is modeled
slightly longer compared to the headgate side to allow the flow to re-attach before reaching the
outlet boundary. In this test, a velocity inlet boundary was assigned to the inlet and set to deliver
18 m3/s of fresh air. The outlet gauge pressure is set to 0 Pa gauge pressure. The coal face
supplies a total of 0.36 m3/s of pure methane resulting in typical methane concentrations found in
operating longwall mines. A wall roughness constant of 1, and a roughness height of 3 cm were
assigned to all wall surfaces based on trials with the model.
Figure 3.10: Model geometry for a coal face methane inflow test
70
|
Colorado School of Mines
|
Two sets of settings were tested in this simple model:
• A method to introduce methane gas into the model. Modeling options included a
pressure boundary, a velocity inlet boundary and using a source term applied to the
porous medium that represents the methane reservoir behind the coal face.
• A method to introduce the methane gas into the longwall face: either the entire coal
face releases methane or the methane flows from multiple, discrete openings.
Two modeling methods were tested to simulate methane emanating from the coal face:
• Method 1: A porous medium represents a methane reservoir behind the coal face. The
entire methane release surface is in direct contact with the mine wall.
• Method 2: A porous medium represents a methane reservoir behind the coal face.
Methane flows into the mine entry through a discrete connection.
Table 3.1 summarize the scenarios for modeling the methane gas emanating from coal face
Table 3.1: Tested scenarios for modeling the methane gas emanating from coal face
Method Interface method Methane inlet boundary type
Pressure
1 Direct contact with porous medium Velocity
Source term
Pressure
2 Discrete contact with porous medium Velocity
Source term
Method 1 results
Figure 3.11 shows the geometry used to simulate Method 1. The methane reservoir behind
the coal face is modeled as a porous medium with viscous resistance value of 1x 1012 /m2.
71
|
Colorado School of Mines
|
The models show that using either a pressure or a velocity inlet boundary pushes the bulk
face air flow towards the back of the entry. Both cases also significantly over predict the amount
of methane entering the model. For the source term case, the airflow distribution along the face is
more realistic as it compares well to the base case without methane. Unlike the pressure and
velocity inlet boundary condition, the source term input is assigned to the porous medium cell
with a preferable flow direction, which minimizes shifting the bulk airflow in the longwall face
due to the artificial pressure drop created across the porous medium. Based on these trials, only
the source term input is considered viable and is used for further modeling.
Method 2 results
Figure 3.14 shows the geometry used to simulate Method 2. The methane reservoir behind
the coal face is modeled as a porous medium with a viscous resistance value of 1x 1012 /m2.
Figure 3.14: Model geometry for coal face methane inflow - Method 2
74
|
Colorado School of Mines
|
Similar to Method 1, the pressure inlet boundary, velocity inlet boundary and source term
input were tested with this method. For the velocity inlet boundary condition, the velocity value
assigned is set to supply a total of 0.36 m3/s of pure methane. The pressure value assigned to the
pressure inlet is based on the resulting pressure produced from the velocity inlet case. The
methane inlet direction specification method is set normal to boundary in both cases. For the
source term approach, the methane gas input of 0.0131 kg/m3.s, is calculated to produce
0.36 m3/s based on the 18.1 m3 volume of porous medium region that represents the methane
reservoir behind the coal face.
Figure 3.15 and Figure 3.16 show the velocity and methane concentration contour plot
comparison between the model with and without methane source. The contour plot shows a plan
view at 1.8 m above the mine floor.
The velocity contour plot comparison shows that utilizing Method 2 does not significantly
alter the bulk flow case, similar to when Method 1 with source term input. Based on these
results, the three tested inlet input are viable for use with the discrete connections method, as
they produced similar flow pattern and methane distribution.
Figure 3.15: Contour plot showing a comparison of velocity distribution for different methane
inlet sources - Method 2
75
|
Colorado School of Mines
|
Figure 3.18: Vector plot showing a comparison of methane gas leaving the porous zone
Comparing the results, the discrete connection case predicts higher methane accumulations
near the roof of the entry compared to the direct surface contact case. In the discrete connections
case, the vector plot comparison shows methane entering the entry at much higher velocities. The
upward flow direction is due to buoyancy and occurs in both cases. A higher methane velocity
can be expected when the discrete connections are modeled with smaller openings and larger
spacing between the discrete connections. The required pressure to push methane through the
discrete openings will also increase, which will produce unrealistic pressures at the boundary and
result in poor model convergence. In conclusion, the combination of Method 1 with a source
term input produces a reasonable result in terms of airflow and methane distribution, while
offering several advantages:
• Easy adjustment of methane input quantity
• Simplified assumptions for the number and size of openings representing fractures in
the coal, and minimal local effects of methane emanating from the coal face
• Simplified meshing for the discrete source connections
77
|
Colorado School of Mines
|
3.1.3 Wall roughness adjustment
Including obstructions represented by shields, the armored face conveyor, the longwall
shearer, the tailgate and headgate drives and the crusher in the headgate entry provides better
accuracy of gas distribution, turbulent flows and pressure drops across the longwall face. Still,
these obstructions did not produce sufficient pressure drops so additional adjustments on wall
roughness for the face and mine entries were required to reflect pressure drops and airflow
resistances typically found in underground coal mines.
Several studies have been done by researchers to determine the friction factor values
typically found in underground coal mines. Table 3.2 and Table 3.3 outlines estimated Atkinson
friction factors for various coal mine airway conditions.
Table 3.2: Friction Factors for Different Airway Types (Kharkar et al., 1974)
Value of k, kg/m3 (x10-10 lbf-min2/ft4)
Straight Curved
Type of Slightly Moderately Slightly Moderately
Clean Clean
Airway Obstructed Obstructed Obstructed Obstructed
Smooth 0.0046 0.0052 0.0063 0.0058 0.0072 0.0080
lined (25) (28) (34) (31) (39) (43)
Unlined
0.0080 0.0091 0.0113 0.0115 0.0126 0.0137
(rock-
(43) (49) (61) (62) (68) (74)
bolted)
0.0124 0.0139 0.0152 0.0158 0.0161 0.0167
Timbered
(67) (75) (82) (85) (87) (90)
80
|
Colorado School of Mines
|
Table 3.4: Standardized Friction Factors for Coal Mine Airways (Prosser and Wallace, 2002)
Intake Drift Return Drift Belt Drift Cribbed Drift
Average Value 0.0075 (40.6) 0.0087 (47.0) 0.0106 (57.0) 0.0678 (365.5)
Maximum Value 0.0115 (61.9) 0.0113 (61.1) 0.0176 (94.7) 0.1441 (776.6)
Minimum Value 0.0048 (26.0) 0.0057 (30.5) 0.0046 (24.3) 0.0452 (243.7)
Std. Deviation 0.0022 (11.8) 0.0018 (9.5) 0.0064 (34.3) 0.0252 (135.7)
Number of
23 15 5 7
Measurements
The mean friction factor for return entries is generally higher than that of intake entries. This
result is to be expected, considering intake entries are better maintained than return entries. The
friction factor for the cribbed drift appears to be significantly higher than other entries, and that
seems to vary based on cribbing dimensions and set-up.
A parametric study was done with FLUENT to determine the equivalent wall roughness
parameter that represents the conditions of an active mine. The simulation set-up for this study is
presented in Figure 3.23. The geometry has the same opening area with the typical mine entry
opening used in the U.S. longwall mine, while the length is set to allow the flow to fully
developed before reaching the outlet.
Figure 3.23: Simulation set-up for an equivalent wall roughness study
82
|
Colorado School of Mines
|
The results, presented in Figure 3.25, show that an additional roughness adjustment is
required to achieve the more realistic airway roughness condition found in typical underground
coal mines.
Figure 3.25: Equivalent Atkinson’s friction factor as a function of wall roughness height
Without any adjustments and a roughness height of 0 m, the airway friction factor closely
resembles a smooth airway. Based on the values listed in literature, a wall roughness height of
0.15 m to 0.50 m represents a reasonable value for the moderately to highly obstructed coal mine
entries presented in Table 3.2 and Table 3.3. The longwall model presented in this study utilizes
a 3 cm thick wall roughness with roughness constant of 1 for the longwall face equipment wall
boundary, and 20 cm thick wall roughness for the main and bleeder entries. These wall
roughnesses lead to airflow resistances that are similar to those measured in typical underground
coal mines. The wall roughness in the face is an order of magnitude lower than what is used in
the bleeder entries due to the inclusion of explicit obstructions, and the wall roughness is needed
to only compensate smaller obstructions such as cables, chains, hoses, etc. that have not been
explicitly modeled. The additional wall roughness is necessary to be assigned for the mine
entries as the model does not account for the roughness created by the roof support and the
surrounding coal that represent the ribs, roof, and floor of the mine entries.
84
|
Colorado School of Mines
|
Figure 3.29: Closeup view of a velocity vector plot showing comparison of flow around the
drums, both for stationary and a 30 RPM rotating drum
The results show considerable differences when the drum is modeled as a cylinder or when
the drum is modeled as stationary, spiral drum. Depending on the study’s focus, these
assumptions may significantly affect results. For example, modeling the shearer drums as non-
rotating, cylindrical drums may be acceptable when conducting a study on gas mixing inside the
gob area, though this simplification is not appropriate when studying gas mixing near the coal
face, especially when methane emanates from the face.
Generally, there are two modeling methods available to simulate rotating objects in CFD:
rotating surface and rotating fluid zones. In a rotating surface, an angular velocity is assigned to
the surface so it can rotate relative to adjacent cells. In the case of rotating fluid, the angular
velocity is assigned to a fluid body instead of a surface.
Although the rotating fluid body is typically easier to use for simple geometries, there is a
fundamental issue with this approach; to achieve a realistic result, the modeler must make a good
assumption on the thickness of the rotating fluid body, and this thickness will vary depending on
complexity of the geometry and surrounding turbulence conditions. In the case of rotating
87
|
Colorado School of Mines
|
The results show that by modeling them as rotating fluid bodies with certain thicknesses
may artificially create a boundary layer trapping any flow inside these fluid zones. By assigning
the surrounding drums and picks as rotating surfaces, a more realistic situation can be modeled
where the assigned angular velocity only applies to the drums and picks, instead to the
surrounding fluid zone. Thus, the rotating surface method is more suitable to model a rotating
spiral drum with multiple picks.
The next test is to determine which method is more appropriate for modeling the parts of the
drums that are engaged in cutting coal. This test uses a simplified longwall face model that
includes part of the longwall face area along with a simplified shearer. The modeled shearer
drums still retain the important features found in real shearer drums, such as the spiral shape and
individual cutting picks. Figure 3.32 shows the geometry of the simplified longwall face and
shearer drum model.
Figure 3.32: Geometry for the rotating spiral drums study, showing model overview (left) and
the shearer drum model (right)
89
|
Colorado School of Mines
|
Two models are considered in this test:
• Model 1: It is assumed that when the drums cut the coal, only the top half of the picks are
in contact with the wall that represents the coal face, leaving an approximate 5 cm gap
between the wall and the shearer drum’s main cylinder. This is shown in Figure 3.33, top.
• Model 2: It is assumed that when the drums cut the coal, all of the picks are in contact
with the coal face, leaving no gap between the face and the shearer drums. This means
that areas of the shearer drums directly in contact with the coal face do not need to be
modeled. This is shown in Figure 3.33, bottom.
In the CFD model, the shearer is in the middle of the longwall face and is cutting toward the
headgate. Both headgate and tailgate shearer drums rotate at 30 rpm in the direction shown in
Figure 3.33, with the airflow near the drums initially set to 0.1 m/s. This air speed minimizes the
turbulence caused by the face airflow and allows better visualization of the effect of the drum
rotation.
Figure 3.33: Geometry comparison for two rotating spiral drums modeling approaches
Figure 3.34 compares the results of the two modeling approaches, looking from a plan view
at 1.5 m above floor, while Figure 3.35 shows the comparison from front view.
90
|
Colorado School of Mines
|
The results show that leaving a gap between the drum’s main cylinder and the wall creates a
larger turbulence region around the drums, which results in more air from the main flow stream
being drawn closer to the coal face. This is shown in Figure 3.34, compared to the case where
only parts of the drums are modeled and rotating. Higher turbulence effect can be observed near
the coal face between the rotating headgate and tailgate drums, as shown in Figure 3.35.
The second scenario represents the case of a shearer in the middle of the longwall face and
cutting toward the tailgate, i.e. in the direction of air flow. All other parameters remain
unchanged. Figure 3.36 shows the results comparison of the two modeling approaches, from a
plan view at 1.5 m above the floor, while Figure 3.37 shows the comparison for two front views.
As in the previous case, leaving a gap between the drums and the wall creates a larger
turbulence region around the drums. However, since the headgate drum is rotating against the
main flow in this case, less airflow is being drawn into the area between the shearer body and the
face. Considering parts of the drums engaged in cutting coal would not have their gaps
completely filled with broken coal, modeling half of the picks’ length engaged inside the uncut
coal face should produce a simulation result closer to real mining conditions. This full shearer
drums rotating modeling approach is used throughout the rest of this study.
Figure 3.36: Comparison of airflow distribution between the two drums rotating methods from a
plan view 1.5 m above floor, with shearer cutting toward tailgate
92
|
Colorado School of Mines
|
This modeling approach has been tested with a smaller model that only include a single
headgate entry, a full 300 m longwall face, a tailgate entry and gob area behind the shields that
extends up to 6 m deep into the gob and 3 m high, as shown in Figure 3.45.
Figure 3.45: Geometry showing a reduced longwall face model
A 6-m-wide gob fringe is included in the partial model to simulate interaction between the
flow inside the longwall face and the gob and allow the airflow inside the gob to re-enter the
longwall face depending on the simulated ventilation scenario. Based on study by Thakur (2006),
this gob region is within the 45 m active roof caving area immediately behind the longwall face.
This gob depth also allows simplification of the gob permeability and porosity. Looking at the
results of gob permeability studies by Esterhuizen and Karacan (2007), Wedding (2014) and
Marts et al. (2014), the permeability and porosity value in this gob region is approximately
uniform across the face, with an exception on the headgate corner and tailgate corner area, where
there are increase in permeability due to the caving characteristic. In this study, uniform
permeability and porosity value is assigned to the 6 m deep gob, as the main purpose of the gob
model is to simulate reasonable airflow leakage across the face. Different gob permeability
100
|
Colorado School of Mines
|
Figure 3.47: Velocity contour plot showing results of profile boundary conditions extraction for
reduced model from plan view 1.5 m above floor
In this test, the mine utilizes a bleeder ventilation system with a back return on the tailgate
side. The tailgate entry inby the longwall face is kept opened to allow the face air to pass to the
bleeder entry through the crosscut directly inby the face. The model is set to supply 48.1 m3/s of
fresh through the two headgate entries. A total of 10.4 m3/s are assumed to leak through the two
headgate curtains (shown as ‘C’ in Figure 3.47) and 4.7 m3/s of air is returned through the belt
entry, resulting in 33 m3/s of air delivered to the face. For reference, bleeder ventilation system
in U.S. longwalls typically delivers a total of 33 - 47.1 m3/s of fresh air to longwall face. Inby the
headgate, a ventilation curtain extends from the rib of the chain pillar until shield number 3 to
direct the supplied fresh air into the face. Each tailgate entry and the two headgate side bleeder
entries are set to supply 4.5 m3/s of air. The air quantities at the back of the bleeder are controlled
with a series of stoppings and bleeder regulators. Regulator R3 (Figure 3.41) is set to allow 4.5
m3/s of air to pass through, while R4 and R5 are closed to force air from the headgate entries to
sweep the back end of the longwall panel and move the potential explosive gas mixtures towards
the center of the gob. R6 is kept fully open.
102
|
Colorado School of Mines
|
The graphs in Figure 3.48 show airflow distribution inside the longwall face area for the two
cases. The x-axis represents the shield number, with the first shield located on the headgate
corner of the longwall face. The blue-colored line represents the case with a full bleeder model,
while the red-colored line represents the case when only the longwall face and small part of the
gob is modeled.
Figure 3.48: Comparison of airflow quantity inside the longwall face between a full panel
bleeder model and a reduced longwall face model utilizing profile boundary conditions
The resulting comparison of velocity distribution at the longwall face and amount of air
leaking across the face show the ability of the reduced model to maintain accuracy while limiting
the model to only include areas of interest. The highest reported discrepancy is 2.6% at shield
number 140, which is within an acceptable range. In this research, the reduced model geometry
that only include the longwall face area and small section of the gob is used. Every time the
simulated inlet conditions changes, or change in tailgate ventilation setup, the full model is re-
run and the profile data are extracted for the boundary condition for the reduced model.
Meshing Setup
Table 3.5 shows the setup for the base mesh.
103
|
Colorado School of Mines
|
Table 3.5: Mesh setup
Mesh parameter Base Mesh Note
Size function Curvature
Min cell size (cm) 3
Max cell size (cm) 30
Max face size (cm) 10 30 cm for gob
Additional adjustment Applicable to
Body sizing (cm) 10 Porous medium representing uncut coal face
Face sizing (cm) 3 Shearer drums
Coal face; Top of Shearer body; Top of TG
Face sizing (cm) 10
drive; Shield roof
Face sizing (cm) 5 Coal face in contact with shearer drums
Total number of cells 31.5 Mil
The longwall face model is divided into four bodies, consisting of:
• The main longwall face without the shearer;
• Part of the longwall face where the shearer is located;
• The uncut coal face, modeled as porous medium; and
• The gob behind the shields, modeled as porous medium
The accuracy of the numerical modeling solution is highly dependent on cell quality.
ANSYS identified the major parameters that can be used to assess overall mesh quality,
including overall cell skewness, orthogonal quality and aspect ratio. ANSYS (2014) provided the
following definition for cell properties:
104
|
Colorado School of Mines
|
• Skewness is defined as the difference between the shape of the cell and the shape of an
equilateral cell of equivalent volume. Table 3.6 shows guidelines used to assess mesh
quality based on the maximum cell skewness values.
Table 3.6: Cell Skewness Guidelines (ANSYS, 2014)
Excellent Very good Good Acceptable Bad Unacceptable
0 – 0.24 0.25 – 0.49 0.50 – 0.79 0.80 – 0.94 0.95 – 0.97 0.98 – 1.00
• The aspect ratio is a measure of cell stretching; the recommended value for this property is
below 5 for flow located away from the walls, noting an exception for quadrilateral,
hexahedral and wedge cells inside the boundary layer. In general, the maximum aspect ratio
should be kept below 35 for the stability of the energy solution.
• Orthogonal quality for cells is computed using the vector from the cell centroid to each of its
faces, the corresponding face area vector and the vector from the cell centroid to the
centroids of each adjacent cell. Table 3.7 shows guidelines used to assess mesh quality based
on the maximum cell skewness values.
Table 3.7: Cell Orthogonal Quality Guidelines (ANSYS, 2014)
Unacceptable Bad Acceptable Good Very good Excellent
0 – 0.001 0.001 – 0.14 0.15 – 0.19 0.20 – 0.69 0.70 – 0.95 0.95 – 1.00
In order to produce good overall mesh quality for the model, the longwall panel model was
separated into four parts, as shown in Figure 3.49, and then joined together in the ANSYS
FLUENT using non-conformal mesh method.
105
|
Colorado School of Mines
|
Table 3.9: FLUENT Solver Settings
General Settings
ANSYS Fluent
Pressure Based
Solver Type
Absolute
Velocity Formulation
Steady State
Time
9.81 m/s2
Gravity
3.4.2 FLUENT model settings
Table 3.10 summarizes the FLUENT model settings used in this study. The realizable k- ε
turbulence model provides improved predictions for flows involving rotation, boundary layers
under strong adverse pressure gradients, separation, and recirculation. It also captures the mean
flow of the complex geometries and predicts reattachment length with good accuracy. This
turbulence model is suitable to model airflow inside the longwall face area, where the flow is
fully turbulent, and many flow separation and reattachment occurs due to the obstructions by the
longwall equipment.
In CFD modeling, it is important that the mesh accurately predicts the velocity gradients
across boundary layers. For turbulent flow modeling, the first cell from the wall should be within
the thin viscous sub-layer. This is difficult to achieve for complex flows in complicated
geometries as it would require a rather fine mesh near all walls. This would significantly increase
the computational time. A wall function allows the use of a larger mesh near the walls, while still
giving a good prediction of the velocity gradient across the boundary layer. The standard wall
function is suitable for cases when the first cell cannot be placed within the viscous sub-layer and
where this cell lies in the log-layer region. This wall function provided reasonably accurate
predictions for the majority of high Reynolds number, wall-bounded flows (ANSYS, 2017). This
wall function also allows wall roughness adjustments. Such adjustments are necessary to achieve
airway resistances typically found in mine entries, and to produce realistic pressure drops across
the longwall face.
116
|
Colorado School of Mines
|
The porous media model is applied to the simulated uncut coal face and gob using the
superficial velocity formulation that follows Darcy’s flow. The inertial resistance term for the
porous media is set to zero, and only the viscous resistance is assigned to the porous zone. The
porous medium model does not consider the particle size distribution, physical porosity and
tortuosity. In order to estimate the pressure drop across the porous zone, ANSYS Fluent treat the
porous media model as an added momentum sink in the governing momentum equations, which
resulted in superficial velocity inside the porous medium.
Table 3.10: FLUENT Model Settings
Transport Models
ANSYS Fluent
On
Energy
Realizable k– ε with Standard Wall Function
Viscous
Zone: Gob, Uncut coal face
Porous Media Model Porous Formulation: Superficial Velocity
Species Transport
Diffusion Energy Source
Species
Mixture Material: Methane - Air
101,325 Pa
Operating Pressure
288.16 K
Operating Temperature
3.4.3 FLUENT materials settings
FLUENT material settings for gas species are shown in Table 3.11. The methane-air mixture
formulation includes five species; species transport was further simplified by removing oxygen,
carbon dioxide, water vapor and nitrogen from the equation, leaving only two species, methane
and air, to be modeled. The use of these two species reduced the computational time and model
complexity because the solver only required solving the methane species transport equation
instead of multiple species.
117
|
Colorado School of Mines
|
Table 3.11: FLUENT Materials Settings
Materials - Settings
ANSYS Fluent
Methane, Air
Gas Species
Incompressible-ideal-gas
Compressibility
Mixing-law
Specific Heat
Thermal Conductivity 0.0454 W/m-K
Methane-air mixtures
Dynamic Viscosity- 1.72 x 10-5 kg/m-s
Methane-air mixtures
Kinetic theory
Mass Diffusivity
The decision to use incompressible flow is based on the expected pressure drop across the
face of 75-150 Pa, depending on the face leakage rate, and a reported airflow velocity lower than
10 m/s inside the the longwall face area, far below Mach 0.3 (~100 m/s) typically used as a
threshold to treat the flow as compressible.
3.4.4 FLUENT discretization and solution settings
The FLUENT solution method settings are given in Table 3.12. Semi-Implicit Method for
Pressure-Linkage Equations (SIMPLE) is used for the pressure-velocity coupling method. In the
SIMPLE algorithm, the pressure gradient term is calculated using the pressure distribution from
the initial or previous iteration, and the velocity field is approximated by solving the momentum
equation. Thus, the discretized momentum equation and pressure correction equation are solved
implicitly, while the velocity correction is solved explicitly. The least squares cell based gradient
scheme was used to solve spatial discretization. The remaining momentum, turbulence, energy
and species transport equations were set as a second-order scheme for better accuracy.
118
|
Colorado School of Mines
|
values typically found in operating mines and met the regulatory requirements prescribed in 30
CFR. Therefore, the resulting airflow distribution within the model should give a good
representation of real longwall mine ventilation conditions.
Figure 3.70: Boundaries location for reduced bleeder model
Table 3.13 shows the model boundaries. Both shearer drums are modeled as rotating walls at
30 RPM, as shown in Section 3.1.4. A wall roughness of 3 cm with a roughness constant of 1 is
assigned to all wall boundaries, except for the shearer drums. The wall roughness for both
shearer drums are set to 0.5 cm with roughness constant of 1.
Table 3.13: FLUENT model zone type and boundaries
Average Gauge
Code Boundary Boundary Type Airflow Quantity (m3/s)
Pressure Value (Pa)
A HG entry Inlet- pressure profile 41 188
B TG entry Inlet- velocity profile 4.5 106
Varies depending on the
C Gob Outlet- pressure profile 127
tested scenario
Varies depending on the
D TG bleeder entry Outlet- pressure profile 100
tested scenario
Zone Zone type Viscous resistance (/m2) Porosity
Uncut coal face Porous medium 1 x 108 5%
Gob Porous medium 1.45 x 105 40%
120
|
Colorado School of Mines
|
A sensitivity study varied airflow quantity and gob resistance. Results are listed in
Appendices B and C.
Model Convergence
For a steady-state simulation, good convergence is achieved when all discrete conservation
equations (momentum, energy, etc.) are conformed in all cells to a specified tolerance, and there
are no longer any significant changes in the solution with further iterations. To determine
convergence, FLUENT recommends checking the following:
• Discrete conservation equations are solved to a specific tolerance
• Overall mass, momentum, energy and scalar balances
• A decrease in residuals for all equations, except energy, to at least 1x10-3, or three order
of magnitude
• An energy residual decrease to at least 1x10-6,or three order of magnitude
• Monitor relevant key variables
Following this guideline, the convergence criteria used for all models in this study are as
follows:
• Continuity: 1x10-4
• Momentum: 1x10-3
• Turbulence parameter, Kappa and Epsilon: 1x10-3
• Species: 1x10-5
• Energy: 1x10-9
Several surface monitors were set at various locations in the model, including the
monitoring of the airflow quantity passing longwall shields 1, 40 and 150 (s-1, s-40, s-150 in
Figure 3.71), and the methane concentration leaving the model through the tailgate bleeder entry
(out-ch4 in Figure 3.71)
Figure 3.71 shows the convergence result for the reduced longwall face model, while Figure
3.72 shows the model mass flow balance. Fore reference, the initial residuals for each parameter
is shown below, rounded to one significant figure:
121
|
Colorado School of Mines
|
The values highlighted by the red box show the surface monitor to further check model
convergence. The residuals appear steady and reach the target residual value for each parameter.
There is also no longer significant change in the monitored key variables with further iterations
and the overall mass balance is achieved, thus model convergence is reached.
Mesh Independence
A mesh independence study was also conducted to further verify the result. This process
involved refining the mesh and analyzing the percent change in a variable of interest to an
acceptable value. The same model setup and convergence criteria were used in this test. Table
3.14 shows the two mesh parameters used for the mesh independence study.
Table 3.14: Mesh parameter for mesh independence study
Mesh parameter Base Mesh Refined Mesh Note
Size function Curvature Curvature
Min cell size (cm) 3 1.5
Max cell size (cm) 30 15
Max face size (cm) 10 10 30 cm for gob
Additional adjustment Applicable to
Porous medium representing uncut
Body sizing (cm) 10 10
coal face
Face sizing (cm) 3 2 Shearer drums
Coal face; Top of Shearer body;
Face sizing (cm) 10 5
Top of TG drive; Shields roof
Coal face in contact with shearer
Face sizing (cm) 5 5
drums
Total number of cells 31.5 Million 49 Million
Lines 1 and 2 in Figure 3.73 are used to verify mesh independence in the bulk flow region.
Line 1 is located 1.5 m above the armored face conveyor area and 1 m away from the coal face,
while line 2 is located 3 m away from the coal face and 1.5 m above the floor. Line 2 represents
the shearer operator walking area. Figure 3.73 shows the locations of these two lines, while
123
|
Colorado School of Mines
|
The comparison using volume-weighted averages at the shearer drums show an insignificant
difference between the two mesh: below 2%. The comparison using two lines on the shearer
body show the same trend, with approximately 2% difference in the predicted methane
concentration between the two mesh. Based on these test results, the base mesh with 31.5 million
cells can be considered acceptable for use in this research.
Model Validation
Model and result validations are necessary to confirm the predicted results represent the
conditions observed in a real longwall operation. The longwall equipment geometries, such as
the shields and shearer, were based on information gathered from a well-known longwall
equipment manufacturer and further simplified to account for the meshing and computational
time constraints. Although some of the details of the equipment was simplified for meshing and
modeling the important physical features that would potentially impact the flow was left intact.
The dimensions of the mine entries, chain pillar and panel size were based on dimensions in
typical U.S. longwall operation. Operational conditions, such as the amount of fresh air supplied
to the longwall face area, stopping locations and regulators settings were based on conditions
typically found in US mine longwall operations.
The longwall ventilation conditions simulated in this study were not based on any specific
mine; the purpose of this research was to find general patterns in varying factors affecting
ventilation conditions in the longwall face in order to make results presented in this study
applicable for general longwall bleeder ventilation conditions.
Ventilation conditions vary by operation. However, assuming the mines use the same
ventilation system, some general trends can be observed, which include:
- Continuous leakage of fresh air from the face to the gob, and the higher accumulation of
methane as the supplied air travels from the headgate to tailgate side of the face;
- Higher leakage around the headgate and tailgate corners of the face due to the high
porosity and permeability around the edge of the gob; and
- Methane accumulation seems to follow linear regression based on ventilation surveys
done in several longwall operations.
132
|
Colorado School of Mines
|
In order to validate the CFD model, several parameters can be compared with the results
obtained from the ventilation surveys or scaled physical models conducted by other researchers.
Note that further validation, such as the effect of shearer cowls and machine-mounted sensor
placement, will be further discussed in forthcoming chapters.
3.7.1 Air leakage patterns across the longwall face
The leakage pattern across the face highly depends on the characteristic of the gob and
immediate roof caving conditions, and can vary significantly for each mine. The gob parameters
used in this research were based on previous research done by Marts (Marts et al., 2014a, 2014b
and 2015) validated using subsidence data obtained from cooperating longwall coal mines in the
U.S. Different gob permeability values were also tested as a part of sensitivity study, and are
presented in the appendix C. Figure 3.86 shows the amount of airflow inside the longwall face,
as it leaks from the headgate to the tailgate side of the face for the case when the shearer is
cutting the tailgate corner of the face. Shield number one is located at the headgate corner. This
result is obtained using the reduced model that only include the full longwall face and small
section of the gob.
Figure 3.86: Air quantity distribution along a longwall face, based on the reduced model
133
|
Colorado School of Mines
|
There was 41 m3/s of fresh air supplied at the headgate side. By the time the air reached the
tailgate side, only around 23 m3/s (56% of the supplied fresh air from headgate side of the face)
remained inside the face due to leakage from the face to the gob. The remaining amount of
airflow at the tailgate corner area still meets the minimum statutory requirement of 14.2 m3/s
specified in 30 CFR §75.325. The leakage is within the expected range based on studies by
Krickovic and Findlay (1971), Peng and Chiang (1986) and Thakur (2006). Based Thakur
(2006), for a longwall operation with a face width of 300 m, about 70% of supplied fresh air may
leak into the gob by the time it reaches the tailgate corner.
A ventilation study by Peng and Chiang (1986) estimated 20% to 40% leakage for a good
immediate roof caving case and around 60% leakage for a poorly caved case. However, this
survey involved shorter longwall face widths, around 200 m based on the reported number of
longwall shields used. This is compared to modern longwall of 300 m or wider. A study using
tracer gas by Krog (Krog et al., 2014) on a 300 m longwall face reported that only about half the
airflow reaching the tailgate bleeder entry travels from inside the longwall face area, while the
rest travels through the gob area behind the shields.
Figure 3.87: Estimated face air leakage based on longwall face width, m (Thakur, 2006)
134
|
Colorado School of Mines
|
Figure 3.90: Physical longwall model test results for 13.7 m gob void spaces behind shields
scenario (Gangrade et al., 2019). Note that “bleederless” refers to a progressively sealed gob.
Ventilation surveys done at a real longwall operation or physical model usually does not
cover the entire cross-section of the face due to safety and its impracticality. They are mostly
focused on the operator walkway area and area on top of the armored face conveyor, where the
bulk flows are located. Thus, the reported air quantity is usually based on the average velocity
measurements done in these areas and multiplied by what the surveyor considers to be the
effective cross-section area. This can lead to an exaggeration of the resulting airflow quantity. In
comparison, the reported airflow quantity presented in this study covers the entire cross-section
of the longwall face, including areas with slower airflow movement, such as the flow around the
shields’ hydraulic jacks.
3.7.2 Methane emission trend across the face
Figure 3.91 shows the cumulative methane concentration across the longwall face for the
case when the shearer is cutting at the tailgate corner of the face. The methane concentration
reported here was obtained from the CFD model by multiplying the amount of airflow passing
through the listed shield number and multiplying it by the weighted average of the methane mole
fraction for the respective shield cross-section area.
136
|
Colorado School of Mines
|
CHAPTER 4
EFFECTIVE METHANE MONITORING LOCATIONS
Effect of Shearer Location on Airflow and Gas Distribution along the Face
The next part of the study is to simulate the effect of shearer location on the airflow and gas
distribution inside the face in four shearer cutting scenarios. This test is used to study the
correlation between the shearer location and the reported methane concentration by the sensors
mounted on the shearer body and tailgate drive.
For all scenarios, the following model setup is used:
• The model is based on a bleeder ventilation system with a back return
• The modeled longwall face length is 300 m, in addition to the 6 m wide entry that represent
the headgate and tailgate entries on both side of the face. Resulting in 312 m total length.
• Pressure or velocity profiles from the previously solved case of longwall bleeder model is
assigned to the inlet and outlet boundary condition
• The headgate inlet supplies 41 m3/s of fresh air, while the tailgate entry is also set as an inlet
and supplies 4.7 m3/s of fresh air
• Gob extending up to 6 m behind the shields is included in the model to simulate the effect of
leakage from the face
• A 20-cm-thick porous medium is used to represent the active coal face and produces
0.12 m3/s of methane gas, uniformly distributed across the longwall face.
• A straight ventilation curtain is used at the headgate corner and extends from the coal rib
until shield 3
• The headgate and tailgate drums are set to rotate at 30 rpm in their respective direction,
along with the use of shearer cowls
Figure 4.1 shows the geometry and location of the shearer for the four scenarios:
• Scenario 1: Shearer located 4 m from the headgate corner and cutting toward the headgate,
location shown by ‘A’ in Figure 4.1.
• Scenario 2: Shearer in the middle of the longwall face, cutting toward the tailgate, location
shown by ‘B’ in Figure 4.1.
139
|
Colorado School of Mines
|
• Scenario 3: Shearer located 10 m from the tailgate corner and cutting toward the tailgate,
location shown by ‘C’ in Figure 4.1.
• Scenario 4: Shearer cutting out the tailgate corner of the longwall face, location shown by
‘D’ in Figure 4.1.
Figure 4.1: Longwall face geometry showing four different shearer cutting scenarios
The following section will show in detail the airflow and methane distribution around the
shearer for these scenarios.
4.1.1 Scenario 1: Shearer cutting close to the headgate corner
In this scenario, the shearer is assumed to be located close to the headgate corner and cutting
toward the headgate, with the shearer headgate drum located 4 m away from the headgate corner.
Figure 4.2 and Figure 4.3 show the volume rendering of velocity and methane concentration for
scenario 1 from a plan view, while Figure 4.4 shows an enhanced view of the methane
concentration around the shearer from the front. The accumulation of methane on the headgate
side of the shearer is due to recirculation and entrainment of methane from the nearby cut coal
face. Note that the contour range for the methane concentration plot is set to 2% maximum for
better visualization and comparison with all scenarios presented in this study.
140
|
Colorado School of Mines
|
The results show the use of ventilation curtains helps direct the flow into the face. In this
scenario, methane accumulation can be observed around the both shearer drums, though it does
not reach the explosive range based on steady-state simulations. This does not rule out the
possibility of an explosive range occurring during a transient case.
The simulation results identify several poorly ventilated areas in the face. These include the
areas between the shearer drums and the cowls and that between the shearer drums and the uncut
coal face. Poorly ventilated areas around the shearer drum are a concern, since this is where
methane can accumulate and form explosive mixtures.
4.1.2 Scenario 2: Shearer in the middle of the longwall face
Two cases are tested in this scenario: the shearer cutting toward the headgate, and the
shearer cutting toward the tailgate. Figure 4.5 and Figure 4.6 show the volume renderings of
velocity and methane concentration for scenario 2 from a plan view, while Figure 4.7 shows the
close-up view of the methane concentration around the shearer from the front.
Figure 4.5: Plan view showing volume rendering of velocity for scenario 2
Figure 4.6: Plan view showing volume rendering of methane concentration for scenario 2
142
|
Colorado School of Mines
|
Figure 4.7: Close-up view showing volume rendering of methane concentration around the
shearer for scenario 2
Based on the results, the shearer drum that cut the roof component of the coal face seems to
be poorly ventilated, which resulted in higher methane accumulations between the drum and the
uncut coal face (compared to the drum that cut the bottom component of the coal face). This is to
be expected, considering methane will accumulate at the roof due to buoyancy and the use of
drum cowls also helps to trap methane in this area.
4.1.3 Scenario 3: Shearer cutting close to the tailgate corner
In this scenario, the shearer’s assumed location is close to the tailgate corner and is cutting
toward the tailgate, with the shearer tailgate drum located 10 m away from the tailgate corner.
Figure 4.8 and Figure 4.9 show the volume rendering of velocity and methane concentrations for
scenario 3 from a plan view, while Figure 4.10 shows the close-up view of the methane
concentration around the shearer from the front.
Figure 4.8: Plan view showing volume rendering of velocity for scenario 3
143
|
Colorado School of Mines
|
By the time the tailgate drum completes the pass and part of the drum is exposed to the
tailgate entry, the fresh air supplied from the tailgate entry outby the face helps to dilute the
methane around the tailgate drum. This is shown by a higher methane concentration around the
headgate drum compared to the tailgate drum. This is one of the main advantages of using a
bleeder system with tailgate back return.
4.1.5 Shearer-mounted and tailgate drive sensor performance
Comparing the four cutting scenarios, the main concern is the location(s) of the highest
methane accumulation when the shearer is cutting close to the tailgate corner, but before the
tailgate drum being exposed to the tailgate entry, as represented in scenario 3 in this study. To
verify this, two lines that represent the typical locations of methane sensors (shown in Figure
4.14) are used to compare the recorded methane concentrations for the four scenarios.
Figure 4.14: Geometry showing methane concentration sampling locations
The two sampling locations are atop the tailgate drive and attached to the shearer body. The
line that represents the typical sensor placement atop the tailgate drive is located 301.5 m away
from the headgate corner and span across the entire 5 m length of the tailgate drive body. This
line only moved forward toward the coal face as the shields advance. The line that represents the
shearer-mounted sensor changes depending on the shearer location along the face. Figure 4.15
146
|
Colorado School of Mines
|
Based on the results, the sensor located on the tailgate drive predicted approximately the
same methane concentration values when the shearer is located far from the tailgate corner,
which in this case was when it was located close to the headgate corner and at the middle of the
face. In addition, there seems to be a trend of higher methane concentrations predicted as the
sensor moved toward the tail-end side of the tailgate drive, and the highest predicted peak value
was about twice that of the lowest predicted values. This is a significant increase, considering
both ends of this tested line are only five meters apart.
When the shearer approached the tailgate corner, the predicted value at the tailgate drive
sensor started to deviate from the previous two cases. The shearer body diverted the incoming
flow toward the back of the shields, as shown in Figure 4.11, allowing the tailgate drive sensor to
pick up a higher methane concentration.
Conversely, the shearer-mounted sensor predicted much lower methane concentration values
compared to the sensor located at the tailgate drive for scenario 1. This is to be expected, since
the tailgate drive sensor is located in the area with the highest cumulative methane concentration
along the face, as the shearer is cutting in the area with the lowest cumulative methane
concentration near the headgate corner. However, when the shearer was cutting close to the
tailgate corner, the shearer-mounted sensor started to report higher concentration values than the
sensor located on the tailgate drive.
These results show the importance of sensor placement and understanding methane
distribution along the face for different cutting scenarios. The tailgate drive sensor is useful when
the shearer is cutting away from the tailgate corner, as it records the cumulative methane across
the face. It does not, however, represent the ventilation conditions around the shearer, which are
a main ignition source. In comparison, the shearer-mounted sensor provides a more accurate
representation of the ventilation conditions around the shearer. Looking at the results, it is
difficult to determine direct correlation between the shearer sensor and tailgate drive sensor
readings that can be used to set up a reliable ignition prevention system.
Evaluation of Current Industry Practices
The next element of the study analyzed the effectiveness of sensor placement in predicting
ventilation conditions directly at the coal face. Current industry practices rely on the use of point-
based methane monitoring typically installed atop the tailgate drive and mounted on the shearer
148
|
Colorado School of Mines
|
body, in addition to periodic methane readings performed by miners working at the longwall
face.
The major issue with this practice is that there is no known correlation between the
measured methane concentrations at these locations and the main area of concern, directly at the
coal face. Out of the listed monitoring locations, the closest one to the coal face is the shearer-
mounted sensor, located at least 1.5 m away from the face and at a height typically around half of
the mining height. Considering methane tends to accumulate on the mine roof due to buoyancy,
these monitoring locations may not be a good representation of the ventilation condition around
the shearer drums. The effectiveness is being tested for these sensor placements to assess the
ignition hazard at around the shearer drums using the ventilation scenario, when the shearer is
cutting close to the tailgate corner.
Two cases are simulated:
• Case 1: The ‘warning’ state. This case represents the condition when one or both of the
sensors report methane concentrations between 1%-2% CH , which would have
4
triggered the warning sign, but is not enough to trigger the shearer automatic shut-off
function.
• Case 2: The ‘shut-off’ state. This case represents the condition when one or both of the
sensors report barely passing 2% methane concentration, which would trigger the
shearer automatic shut-off function.
In this test, the methane gas concentration by volume is represented using iso-surface with
different colors, e.g. light blue for 1%, green for 2%, orange for 4% and red for 5.5%. The 1%
methane concentration represents the prescribed limit when the sensor starts giving a warning
sign to the operator. At a 2% concentration, it has reached the limit when the automatic shut-off
function should trigger and de-energize the shearer. The 4% and 5.5% concentrations represent
the transition to being explosive and within explosive ranges, respectively.
4.2.1 Case 1: Warning state
In this case, the amount of methane introduced into the model from the coal face is increased
from 0.12 m3/s to 0.32 m3/s. Figure 4.17, Figure 4.18 and Figure 4.19 show the methane
distribution around the shearer for case 1 from different viewpoints, while Figure 4.20 shows the
149
|
Colorado School of Mines
|
Looking at the iso-surfaces plot, about one-third of the tailgate drum already operates in an
explosive gas mixture zone and thus can potentially result in a face ignition. Based on the sensor
readings, the shearer-mounted sensor should already have given a warning to the operator, but
will not trigger the automatic shut-off function.
Conversely, the tailgate drive sensor is still reporting a value below 1% methane. The
sudden methane concentration drops predicted at the tail-end side of the tailgate drive sensor is
due to this location is already exposed to the tailgate entry that supplied fresh air from outby the
face. The methane concentration plot also shows a noticeable difference in methane readings
between the head side and the tail side of the shearer body, with the tail side of the shearer body
reporting a higher methane concentration. These results agree with the sensor placement
recommendations based on a study by Cecala (1993).
4.2.2 Case 2: Shearer shut-off state
In this case, the amount of methane introduced into the model from the coal face is increased
from 0.12 m3/s to 0.40 m3/s. Figure 4.21, Figure 4.22 and Figure 4.23 show methane distribution
around the shearer for case 2 from a different viewpoint, while Figure 4.24 shows the methane
concentration predicted by the shearer-mounted and tailgate drive sensor (shown in yellow
lines), respectively.
Figure 4.21: Iso-surfaces plot showing methane distribution around the shearer drum from the
front view, for shearer shut-off state
152
|
Colorado School of Mines
|
Figure 4.24: Predicted methane concentration by the sensor placed on the shearer body and
tailgate drive, for shearer shut-off state
Looking at the methane concentration around the shearer, the tailgate drum is currently
operating in transition to an explosive zone, with about half of the drum covered by the explosive
gas mixtures iso-surface. This scenario will likely lead to face ignition. Based on the sensor
readings, the shearer-mounted sensor can potentially shut off the shearer as soon as it measures a
methane concentration of 2% or higher. In comparison, the sensor located on the tailgate drive
still predicted a value below 1% concentration.
As pointed out in the previous case, sensor placement plays an important role in preventing
face ignition. In this scenario, the shearer-mounted sensor will trigger the shearer shut-off
function if it is installed at the tail side of the shearer body. Based on these two cases, only the
sensor mounted on the shearer seems to be useful in preventing possible face ignition.
4.2.3 Effects of shearer cowls on gas mixtures around shearer drums
Studies done by Cecala (1993) suggest shearer cowls can result in gas accumulation during
cutting. To test this, two shearer drums models were tested with and without the cowls.
154
|
Colorado School of Mines
|
A scenario where the shearer is located at mid-face and cutting toward the tailgate is used to
analyze the effect of the cowls on ventilation conditions and formation of explosive gas zones
near the drums. Figure 4.25 shows the comparison of a velocity contour plot 0.5 m from the coal
face for the two models, while Figure 4.26 shows comparison of methane concentrations around
the shearer drums. For better visualization, only methane concentrations higher than 2% around
the shearer are shown.
The comparison of velocity contour plots shows the obstruction of flow by the cowls and the
resulting poorly ventilated area between the drums and the cowls. Without the cowls, the fresh
air still managed to help dilute the area between the drums and the face, as shown by a lower
overall methane concentration around the drums. However, in the case of the drums utilizing
cowls, the results clearly show an accumulation of methane between the drums and the cowls.
Based on these results, the use of shearer cowls can increase the likelihood of face ignition by
trapping incoming methane from the face between the uncut coal and the cowls.
Figure 4.25: Comparison of cowl vs. no-cowl shearer drums, showing velocity contour plot at
0.5 m from the coal face
155
|
Colorado School of Mines
|
Figure 4.30: Effect of a gob plate on airflow distribution around the tailgate corner area, showing
velocity streamlines, overview
Gob plates that cover the full mining height direct the flow toward the opening in front of
the shields and closer to the coal face. A major issue with this ventilation condition is the use of a
gob plate can significantly affect the performance of the methane sensor typically installed atop
the tailgate drive. To test this, the two lines representing the potential location of the methane
sensors shown in Figure 4.28 are used to compare the recorded methane concentrations for the
two cases. The scenario used for this test is the same one used in the shearer shut-off state, when
one sensor picked up a 2% methane concentration and shut off the shearer. Figure 4.31 and
Figure 4.32 show a comparison of methane distribution around the tailgate corner, while Figure
4.33 and Figure 4.34 show the comparison of the predicted methane concentrations by the sensor
located on the tailgate drum and shearer body, respectively. The blue and green iso-surface
represent the boundary where the methane gas reaches the 1% and 2% concentrations,
respectively, while the red iso-surface represents the 5% concentration boundary. Note that, in
this model, the tailgate drive is 1.6 m high, or about half the mining height.
158
|
Colorado School of Mines
|
In bleeder ventilation systems with a back return, and when no gob plate is used, the
methane at the tailgate corner is being diluted and pushed toward the back of the shields by the
fresh air coming from the tailgate entry (brown streamlines). This set-up helps ventilate the
tailgate corner of the face by continuously diluting methane with supplied fresh air from the
tailgate entry.
The use of a gob plate channels the flow toward the opening in front of the face and prevents
fresh air forming in the tailgate entry to dilute the tailgate corner area. At the same time, the gob
plate blocked some of the airflow coming from the headgate side of the face and induced more
leakage into the gob, which explains the higher methane concentration readings by both the
shearer-mounted and tailgate drive sensors, compared to the case where no gob plate was used.
There is also a noticeable increase in methane accumulation next to the tailgate drum, shown by
a larger red iso-surface area in the tailgate with a gob plate.
Overall, gob plate use is not recommended for bleeder ventilation systems that utilize
tailgate back return. Preventing fresh air from entering the tailgate to dilute methane at the
tailgate corner area may increase face ignition risk during the shearer tailgate cut-out sequence.
4.2.5 Effect of changes in tailgate ventilation set-up on sensor performance
In this test, two tailgate ventilation scenarios, presented in Figure 4.35: Effect of a tailgate
ventilation set-up on the airflow distribution at the tailgate corner, showing velocity streamlines
were tested. The first scenario represents a normal tailgate with back return ventilation
conditions, with the airflow from the longwall face mixing with fresh air from the tailgate entry
and flows toward the bleeder entry through the open crosscut in the tailgate inby the face. This
tailgate set-up allows face air to sweep and ventilate the gob’s tailgate corner.
The second scenario represents the case when the previously supported tailgate entry
collapsed and the flow from the face is directed into the next crosscut outby the face. Some
longwall operators purposely direct the flow outby the face as part of their normal ventilation
plan, while others, such as in the UBB mine explosion in 2010, were forced to make adjustments
to direct flow outby the face due to roof failures that blocked the previously supported bleeder
entry inby the face.
161
|
Colorado School of Mines
|
Figure 4.39: Effect of a tailgate ventilation set-up on tailgate drive sensor performance
Looking at the iso-surfaces’ graphs, the scenario when the flow directed outby the face
shows a higher methane concentration overall, and directly at the coal face due to the shifting of
the bulk flow toward the front side of the shields. This is also reflected by the concentration of
methane predicted by the shearer sensor, which show overall higher values for the roof fall
scenario. Since this is still based on a bleeder ventilation system, some of the airflow will still
flow inby the face through the gob and collapse in the tailgate entry. The redirection of flow
outby the face reduces the amount of fresh air available to dilute methane around the tailgate
drive, resulting in higher methane concentrations predicted by the tailgate drive sensor.
These results show the importance of utilizing bleeder ventilation with a back return. In the
case where flow is directed outby the face, the recirculation zone (as shown in Figure 4.35) can
potentially trap methane around the tailgate corner area and can pose an ignition risk during the
tailgate cut-out sequence.
164
|
Colorado School of Mines
|
CHAPTER 5
METHANE EXPLOSIVE ZONE PROXIMITY DETECTION
Detection Method
In order to set up a system that can reliably assess the ventilation condition around the
shearer drums, the proposed detection method will rely on methane readings and the correlation
between multiple sensors in relation to the shearer location along the face. The number of
sensors and placement are based on the following considerations:
• Placement in a reasonably safe location;
• A correlation shown based on shearer location that can be used to set up a methane
explosion preventive action;
• Provides readings that represent the ventilation condition around the shearer drum,
thus preventing premature shearer shut-off; and
• Reasonably spaced out to ensure the sensors can trigger the automatic shearer shut-
off function in time, while considering the shearer cutting speed and sensor response
time
The proposed new monitoring practice is to install sensors on the roof of the longwall shield,
30 cm (1 ft) away from the front tip, as shown in Figure 5.1 This measurement location is still
within the prescribed regulation of having the measurement at least 30 cm (12 inches) away from
the methane source, in this case the coal face. Figure 5.2 shows the longwall face model with red
lines to evaluate the projected methane reading when the proposed shield tip sensors are installed
on every shield.
165
|
Colorado School of Mines
|
Using the same scenarios and two cases tested in the previous sections (section 4.1 and 4.2),
Figure 5.3 shows the projected methane concentrations predicted by multiple sensors for the
scenario where the shearer is located mid-face and 10 m away from the tailgate corner, cutting
toward the tailgate (section 4.2 case 2).
Figure 5.3: Projected methane concentration reading by the shield sensors
In this scenario, the shield sensor located 296 m away from the headgate already predicted
methane concentrations within explosive ranges, while the shearer-mounted and tailgate drive
sensors predicted methane concentrations around 2% and 0.9%, respectively. It is also important
to note the shield tip sensors located 150 m from the headgate corner already predicted
concentrations exceeding the 2% shut-off limit and would have shut off the shearer if the same
regulatory limit is used for the shield tip sensor.
The methane concentrations predicted by multiple sensors on the shield tip show a linear
trend correlation, with noticeable peak concentrations predicted directly on the return side of the
tailgate drum. This peak concentration is expected to move along with the shearer and with a
concentration value that better represents the average concentration around the tailgate drum.
167
|
Colorado School of Mines
|
This peak concentration is used to set up a monitoring system to slow down or stop the shearer
before the shearer reaches the area where methane is already within an explosive range.
From a reliability perspective, installing a sensor on every single shield is preferable to
provide better correlation between sensors and to mitigate the impact of having a single or
multiple failure of sensors. However, the modern longwall face typically consists of more than
150 shields, depending on the face width, which can lead to sensor maintenance and calibration
issues in real practice. A proposed number of sensors should be reasonable from an operational
perspective, while still maintaining the ability to fulfill the role of providing methane reading
correlation between sensors.
To determine a suitable number of sensors, three sensor placement cases are tested. The
scenario for this test is when the shearer located in the middle of the longwall face and cutting
toward the tailgate. Figure 5.4 shows the geometry of longwall face with a sensor installed on the
roof of the shield at every 5 shields. Note that the sensors shown in the figures are for
visualization purpose and not physically model. The methane concentration values predicted by
these sensors are obtained using a probe function on a point where these sensors are assumed to
be located.
Figure 5.4: Geometry of longwall face showing shield sensor placement for sensor spacing test
Figure 5.5, Figure 5.6 and Figure 5.7 show the predicted methane concentration by these
multiple sensors for the case when a sensor is installed every 5 shields, 10 shields, and 20 shields
168
|
Colorado School of Mines
|
Figure 5.7: Methane concentration predicted by the shield sensors for sensor placement every 20
shields
Comparing the results, the case when a sensor is installed every 10 shields still managed to
capture the same profile as the case with a sensor installed every 5 shields, with multiple sensors
showing deviation from the linear trend that indicate the shearer location. However, when the
sensor is installed every 20 shields, the shearer location only indicated by the peak concentration
predicted by a single sensor. Considering the distance of 20 shields is 40 m apart for this case,
this is not a preferable setup, as it increases the likelihood of the sensor failing to pick up the
peak methane concentration around the shearer due to the large distance between sensors. This is
demonstrated in the case when the sensor placement began at shield 10, instead of shield 20,
shown in Figure 5.8.
Figure 5.8: Methane concentration predicted by the shield sensors for sensor placement every 20
shields at different sensor starting placement
170
|
Colorado School of Mines
|
In this case, the next sensor on the tail side of the shearer barely picks up the peak
concentration in front the shearer. These sensors placement reduce the reliability of the proposed
method to shut off the shearer prior to entering the explosive mixtures zone. Therefore, installing
a sensor at least every 10 shields is considered to be the minimum number of sensors required to
ensure reliability of this method. Increasing the number of sensors increase the method reliability
in detecting the peak methane concentration, but at the same time increasing the risk of
premature shearer shut off due to sensor malfunction or not properly calibrated. It can also put an
operational constraint due to the requirement to calibrate the sensors.
To further test the viability of this proposed methane sensor placement, 6 shearer cutting
scenarios, shown in Table 5.1, are used.
Table 5.1: Shearer cutting scenarios for shield sensor test
Scenario Case Shearer center location
1 Shield 75
Shearer cutting toward tailgate 2 Shield 115
3 Shield 143
4 Shield 140
Shearer cutting toward headgate 5 Shield 118
6 Shield 78
Sensor Performance: Shearer Cutting Toward the Tailgate
During the head to tail cut, the tailgate drum posed an ignition risk when cutting the roof of
the coal seam. In this scenario, the shearer started from area with the lowest methane
accumulation at the headgate corner and continues to move into area with the highest methane
concentration at the tailgate corner, increasing the risk of methane explosion as the shearer
getting close to the tailgate corner. To evaluate the proposed sensor location performance, three
shearer locations are tested and represent scenario when the shearer is cutting toward the tailgate
corner. In these tests, it is assumed that a sensor is installed on the tip of the shield at 5 shields
spacing. The location of the shearer-mounted sensor and tailgate drive sensor are based on
approximate location shown in the diagram of UBB sensor location (Page et al., 2011). The
amount of methane introduced into the model from the coal face is set to 0.40 m3/s for all cases.
The rest of the boundary condition remains the same, as listed in section 3.4.5.
171
|
Colorado School of Mines
|
The iso-surfaces plot show that the area next to the tailgate drum already reaching 4% CH
4
concentration, with small volume of explosive mixtures start forming around the drum. Looking
at the predicted methane concentration by multiple sensors, the shield tip sensors show linear
concentration increase across the face with notable sudden change in concentration predicted by
shield 70 and 80. The sudden drop of concentration predicted at shield 70 is due to the shield has
not been fully advanced, resulting in larger distance between the coal face and the sensor. The
sensor at shield 80 is located 3 m away from the tailgate drum and started to pick up the methane
buildup as the shearer tailgate drum approaching this sensor. In this case, the sensor at shield 80
already predicted around 4.0% CH , while the shearer-mounted and tailgate drive sensors still
4
predicted approximately 1.2% and 0.6% CH respectively. In this case, the shearer-mounted
4,
sensor should start giving a warning to the operator, as it exceeds the 1% CH limit prescribed in
4
the regulation.
5.2.2 Case 2: Shearer at shield 115
Figure 5.11 and Figure 5.12 show the methane concentration around the shearer and the
predicted methane concentration by the multiple sensors located on the tip of the shields when
the shearer is located between shield 112 and 118, and cutting toward the tailgate.
Figure 5.11: Iso-surfaces plot showing methane distribution around the shearer when the shearer
is located at shield 115 and cutting toward the tailgate
173
|
Colorado School of Mines
|
Figure 5.12: Methane concentration readings for multiple methane sensor placement when the
shearer is located at shield 115 and cutting toward the tailgate
As the shearer move closer to the tailgate corner, the methane concentration around the
drums increase, as shown by larger area covered by the 4% and 5.5% methane iso-surfaces, and
already posed an ignition hazard. This is also shown by the peak methane concentration by the
sensor at shield 120. This sensor is located 3 m away from the tailgate drum and predicted
methane concentration around 4.9%. In comparison, the shearer-mounted and tailgate drive
sensors predicted approximately 1.8% and 0.6% CH respectively. At this point, the shearer-
4,
mounted sensors should still trigger the warning sign, but not enough to trigger the automatic
shut-off function on the shearer as it still below the 2% limit.
The iso-surface plots result also show that the placement of the shearer-mounted sensor is
highly sensitive and could lead to failure to shut off the shearer. For the same condition, a sensor
placed on the front tail side of the shearer body would have triggered the automatic shearer shut-
off function.
5.2.3 Case 3: Shearer at shield 143
Figure 5.13 and Figure 5.14 show the methane concentration around the shearer and the
predicted methane concentration by the multiple sensors located on the tip of the shields when
the shearer is located between shield 140 and 146, and cutting toward the tailgate.
174
|
Colorado School of Mines
|
peak concentration is only 4.0%, which is lower than in the previous case (4.9%) when the
shearer is cutting between shield 112 and 188. This is due to the shield 150 sensor is still located
7 m away from the tailgate drum in this case. The peak concentration will increase as the shearer
moves closer to the sensor. In comparison, the shearer-mounted and tailgate drive sensors
predicted approximately 2.1% and 0.9% CH respectively. At this point, the shearer-mounted
4,
sensor should already trigger the automatic shut-off function on the shearer as it reaches the 2%
CH limit. Note that even in this condition, the sensor located on the tailgate drive has not trigger
4
the warning sign.
5.2.4 Explosion hazard assessment during Headgate – to – Tailgate pass
To better visualize and analyze these sensors performance, the comparison of the predicted
methane concentration by the sensor located at the shields tip for different shearer location as it
cut toward the tailgate is plotted in a single graph. Figure 5.15 shows the predicted methane
concentration by the shield sensors installed on every 10 shields, while Figure 5.16 shows the
projected peak methane concentration predicted by these sensors as the shearer moved toward
the tailgate. Note that the shearer location listed in the graph legend refer to the center location of
the shearer. For example, ‘shearer at S-75’ refer to the case when the shearer is located between
shield 72 and 78.
Figure 5.15: Predicted methane concentration by the multiple shield sensors as the shearer
cutting toward the tailgate
176
|
Colorado School of Mines
|
Figure 5.16: Projected methane concentration reading by multiple shield sensors as the shearer
cutting toward the tailgate
The results show that the peak concentration moved along with the shearer and the peak
concentration also increase as it gets closer to the tailgate corner. This shows the viability of the
proposed multi-sensors monitoring system to be used to assess the hazard around the shearer as it
moves toward the tailgate.
Table 5.2 shows the comparison between methane concentration around the shearer drums
and the reading by the multiple sensors. The average CH concentration around the drum covers
4
the area between the uncut coal face and cowls, and extend half meter from the front coal face, as
shown in Figure 5.17. The shield sensor reading is based on the peak methane value predicted by
the nearest sensor located on the tail-side of the shearer, while the projected shield sensor reading
represent the peak concentration that will be predicted by the sensor immediately after the
shearer pass through this sensor.
177
|
Colorado School of Mines
|
Figure 5.17: Longwall face geometry showing hazard assessment with multiple sensors
Table 5.2: Comparison of methane reading by multiple sensors during head-to-tail-pass
Ave CH Ave CH Shield
4 4 Shield Shearer TG drive
Shearer conc. at HG conc. at TG sensor
sensor sensor sensor
location drum drum projected
2.1% 4.4% 4.0% 4.7% 1.2% 0.6%
Shield 75
2.9% 5.4% 4.9% 5.8% 1.8% 0.6%
Shield 115
3.2% 5.2% 4.0% 5.5% 2.1% 0.9%
Shield 143
Based on the results, the shearer location does not seem to significantly affect the tailgate
drive sensor reading until the shearer is cutting close to the tailgate corner and divert the flow
toward the back of the shields. In comparison, the shearer-mounted sensor reading give better
indication of the ventilation condition around the shearer, as shown by the increase in the CH
4
concentration reading as the shearer moves closer to the tailgate corner. Even then, the methane
reading barely passing the 2% shearer shut-off limit and may not trigger the shut-off function
considering it is still within the 10% sensor accuracy ranges. The predicted concentration is
approximately two to three times lower of the peak concentration predicted by the shield tip
sensors, and about three to four times lower than average concentration around the tailgate drum.
In comparison, the projected shield sensors reading seems to be a good representative of the
condition around the tailgate shearer drum.
178
|
Colorado School of Mines
|
The decision to shut off the shearer can be made by comparing the predicted peak
concentration with the methane concentration around the shearer drums. Because the model does
not consider the effect of water sprays attached to the shearer drum, a proper assessment should
be based on the size of the explosive mixtures around the drums and the ventilation condition
upfront of the drums. For example, the small high methane accumulation located between the
drums and the uncut front coal face, shown to have reached explosive ranges, may have lower
concentration in real mining condition due to the additional turbulence created by the water
sprays and chunk of coal cut from the face. By looking at the ventilation condition along the
drums cutting path, preventive action can be made by shutting-off the shearer prior to cutting
through area with large volume of explosive gas mixtures.
Since the shield sensor located closer to the coal face compared to the shearer-mounted and
tailgate drive sensors, the predicted concentration will be higher than the shearer-mounted sensor
and better represent concentration around the drums. However, this will require a new regulatory
limit to be used for the shield tip sensors in order to not prematurely shut off the shearer.
Additional regulatory limit should be setup to automatically slow down the shearer cutting speed
when the sensor reads methane concentration above certain value until lower methane
concentration is predicted by the nearby sensors. Considering typical shearer cutting speed of
15 m/min, and the sensor delay around 30 to 40 seconds for catalytic type, or 8 to 20 seconds for
infrared type (Taylor et al., 2010) before reporting the reading results, slowing down the shearer
will allow the next sensor to shut off the shearer when necessary, while at the same time
reducing the rate of methane outgassing from the coal face by slowing down the rate of freshly
opened coal surface. For more reliability, the system should be setup to give warning, slow down
the shearer, or shut off the shearer when high methane concentration readings are reported
consecutively by two sensors located next to each other.
Sensors Performance: Shearer Cutting Toward the Headgate
In this scenario, the shearer starts from area with the highest methane accumulation at the
tailgate corner and continue to move into area with lowest methane concentration at the headgate
corner. Considering this, methane explosion most likely to occur when the shearer is still cutting
close to the tailgate corner. The explosion risk reduce as the shearer moves away from the
tailgate corner toward the fresh air at the headgate. During the tail to head cut, the headgate drum
179
|
Colorado School of Mines
|
Looking at the iso-surface plot, there is no peak methane concentration that will be picked-
up by the shield sensor on the head side of the shearer. This is due to the headgate drum cutting
against the bulk flow that continuously dilute the area upstream the headgate drum and prevent
methane buildup. However, the impact of air diversion due to the blockage by the shearer body
allow methane to build up around the tailgate drum. In this case, both the headgate and tailgate
drums are poorly ventilated, as shown by both drums covered by the 4% CH iso-surfaces. The
4
shield tip sensor located 13 m away on the tail side of the shearer, at shield 150, already
predicted methane concentration around 3.7%, while the sensor located 2 m away, at shield 135,
on the head side of the shearer predicted a 3.1% CH . In comparison, the shearer-mounted and
4
tailgate drive sensors predicted approximately 2.2% and 1.1% CH , respectively. At this point,
4
the shearer-mounted sensor should trigger the automatic shut-off function on the shearer as it
reaches the 2% CH limit.
4
5.3.2 Case 5: Shearer at shield 118
Figure 5.20 and Figure 5.21 show the methane concentration around the shearer and the
predicted methane concentration by the multiple sensors located on the tip of the shields when
the shearer is located between shield 115 and 121, and cutting toward the headgate.
Figure 5.20: Iso-surfaces plot showing methane distribution around the shearer when the shearer
is located at shield 118 and cutting toward the headgate
181
|
Colorado School of Mines
|
Figure 5.21: Methane concentration readings for multiple methane sensors placement when the
shearer is located at shield 118 and cutting toward the headgate
Looking at the iso-surface plot, the methane concentration around the drums decrease as the
shearer moves away from the tailgate corner, as shown by the reduction in the area around the
drums being covered by the 4% and 5.5% CH iso-surfaces. In this case, the shield tip sensor
4
located 7 m away on the tail side of the shearer, at shield 125, predicted methane concentration
around 3.6%, while the sensor located behind the headgate drum, at shield 115, predicted a
2.9% CH . In comparison, the shearer-mounted and tailgate drive sensors predicted
4
approximately 2.0% and 0.9% CH respectively. At this point, the shearer-mounted sensor is
4,
within the 2% CH shearer shut-off limit.
4
5.3.3 Case 6: Shearer at shield 78
Figure 5.22 and Figure 5.23 show the methane concentration around the shearer and the
predicted methane concentration by the multiple sensors located on the tip of the shields when
the shearer is located between shield 75 and 81, and cutting toward the headgate.
182
|
Colorado School of Mines
|
approximately 1.3% and 0.7% CH respectively. At this point, the shearer-mounted sensor
4,
should still give a warning sign as it passes the 1% CH limit.
4
5.3.4 Explosion hazard assessment during Tailgate – to – Headgate pass
To better visualize and analyze these sensors performance, the comparison of the predicted
methane concentration by the sensor located at the shields tip for different shearer locations as it
cut toward the headgate is plotted in a single graph. Figure 5.24 shows the predicted methane
concentration by the shield sensors installed on every 10 shields, while Figure 5.25 shows the
projected peak methane concentration predicted by these sensors as the shearer moved toward
the tailgate. Note that the shearer location listed in the graph legend refer to the center location of
the shearer. For example, ‘shearer at S-78’ refer to the case when the shearer is located between
shield 75 and 81.
Figure 5.24: Predicted methane concentration by the multiple shield sensors as the shearer
cutting toward the headgate
184
|
Colorado School of Mines
|
Figure 5.25: Projected methane concentration reading by the multiple shield sensors as the
shearer cutting toward the headgate
Table 5.3 shows the comparison between methane concentration around the shearer drums
and the reading by the multiple sensors. The shield sensor reading is based on the peak methane
value predicted by the nearest sensor located on the tail-side of the shearer, while the projected
shield sensor reading represent the peak concentration that will be predicted by the sensor
immediately after the shearer pass through this sensor and the shield has advanced.
Table 5.3: Comparison of methane reading by multiple sensors during tail-to-head-pass
Ave CH Ave CH Shield
Shearer 4 4 Shield Shearer TG drive
conc. at HG conc. at TG sensors
location sensors sensor sensor
drum drum projected
Shield 140 4.3% 4.5% 3.7% 3.8% 2.2% 1.1%
Shield 118 4.2% 4.4% 3.6% 3.7% 2.0% 0.9%
Shield 78 3.2% 3.6% 2.8% 2.8% 1.3% 0.7%
Looking at the results, the reading by proposed monitoring location at the roof of the shields
does not shows a distinct peak methane reading during the tail to head pass. Even though higher
185
|
Colorado School of Mines
|
methane concentration is shown around headgate drum, the shield sensor located on the tail side
of the shearer is still showing significantly higher methane reading than the one on the head side
of the shearer. This still makes the tail side sensor as the primary sensor to detect explosion
hazard around the drums, similar to the case when the shearer is cutting toward the tailgate.
Based on the predicted methane concentration, the shield sensor directly on the tail side of the
shearer reports value approximately 1% lower than the average methane concentration around
the drums.
Similar to the head-to-tail pass, the tailgate drive sensor is not effective in assessing the
ignition risk around the drums, with methane reading barely reaching the 1% warning limit. In
comparison, the sensor mounted on the shearer body seems to be more effective in preventing
ignition hazard during the tail to head cutting sequence, as the shearer-mounted sensor is
effectively closer to the coal face compared to H-T pass, due to the main flow start from smaller
opening on the headgate side into larger opening on the tailgate side. However, the predicted
methane concentration is still about two to three times lower than average concentration around
the drums.
CFD Modeling of Methane Explosions in the Longwall Face
The CFD ventilation model can be further used to quantify the methane explosion hazard by
integrating it with a combustion model to simulate 3D methane gas explosions. The results
presented in this section are part of a collaboration work with Strebinger (2019) as a deliverable
for the NIOSH funded project titled “Combustion Modeling for Fire and Explosion Prevention in
Longwall Gobs”. Refer to Strebinger (2019) for background on the 3D CFD combustion
modeling part of the results presented in this section.
Methane combustion modeling is computationally intensive and has more restrictions on
mesh size and quality than modeling fluid flow. For example, laminar methane-air flames have a
flame thickness on the order of 1 mm and a quenching distance approximately between 2-3 mm
for a stoichiometric flame (9.5% methane by volume) at 300 K and 101 kPa (Barnett & Hibbard,
1959; Andrews & Bradley, 1972). For combustion modeling, the mesh size must be on the order
of millimeters in order to fully resolve the propagation of the flame front. However, the fluid
flow boundary layers and other key fluid flow features in the full-scale ventilation model are
much larger than the flame thickness. Thus, the base mesh for the fluid flow can be larger than
186
|
Colorado School of Mines
|
what is required to fully resolve the flame reaction front. Therefore, mesh allocation will be
important to ensure model accuracy with acceptable computation times.
To resolve the flame front propagation, the model uses three levels of mesh adaption on the
temperature gradient every 10 μs. This method has proved useful when simulating methane
combustion in both small and large domains (Strebinger et al., 2019). The mesh is refined in
critical areas where the ignition is simulated. Once the flame expands and travels towards the
model boundaries, the model is expanded into the adjacent zones and data interpolated to allow
flame propagation into the adjacent zones. This method can be extended to simulate explosions
propagating through the entire longwall mine.
Figure 5.26 shows a plan view of air flow velocity inside the longwall face, along with a
close-up, isometric view of velocity contours around the shearer drums in the tailgate corner area
used for this test.
Figure 5.26: Volume rendering of velocity inside longwall face from plan view (top) and
velocity contour plot showing close-up view of flow around shearer drums
187
|
Colorado School of Mines
|
For the methane inflow source, this test only considers methane emanating from the uncut
coal face around the shearer location. Figure 5.27 shows a volume rendering of the methane
mole fraction around the shearer drums for this ventilation scenario. The rendering is limited to
the area near the shearer drums for better visualization.
Figure 5.27: Volume rendering of methane mole fraction around shearer drums and ignition
location at HG drum.
The ventilation scenario depicted in Figure 5.26 represents the case where a roof fall blocks
the tailgate inby the face, forcing the return air outby towards the nearest open crosscut. This
leaves insufficient fresh air to dilute methane in the tailgate corner, resulting in the formation of
an explosive mixtures of methane and air near the face and around the shearer drums. Figure
5.27 shows methane accumulations between the headgate drum and the uncut coal as well as
between the tailgate drum and the cowl while the shearer is cutting towards the tailgate.
For the first scenario, it is assumed an ignition occurs at the coal face while the headgate
drum is cutting the coal face. The ignition location is 8 cm away from the headgate drum in the
horizontal direction, 1.27 m above the floor, and 2 cm away from the coal face. This ignition
location was chosen to represent a likely face ignition scenario in real mining operation during
the tailgate cut-out. In the model, the shearer drums rotate with a fully developed flow. It is
assumed that the drums rotation at 30 RPM is relatively slow compared to the pressure wave
generated during the simulated methane ignition and combustion event; thus, the drums rotation
are switched off and the drums are treated as stationary. Considering the time scale of the
188
|
Colorado School of Mines
|
explosion, on the order of milliseconds, the continuous movement of the shearer and rotation of
the drums should not have a significant impact on the pressure waves generated from the
methane ignition.
To model the resulting combustion event, the following settings are used in ANSYS Fluent
v. 18.2:
• 3-D, compressible flow
• Viscous-Standard k-ω model (low Reynolds number and shear flow corrections)
• Energy Equation
• Species Transport (volumetric reactions, finite rate chemistry, laminar flame speed theory)
• First order in time and space
• Time Step = 10 μs
• Methane-air 2 step mechanism
• Pressure-velocity coupling
• Adaptive meshing (3 levels) on the temperature gradient every time step
The model was initialized with a 524 cm3 spherical body of stoichiometric methane-air (9.5%
methane by volume) at the headgate drum, as shown in Figure 5.27. The spark is initiated using
the ANSYS Fluent Spark Model (v18.2) with the following setup:
• Ignition energy, E = 60 mJ
ign
• Ignition energy duration = 2 ms
• Initial kernel radius = 2 cm
• Laminar kernel expansion
Figure 5.28 shows a volume rendering of the explosion overpressure and initial explosion
temperatures after ignition near the headgate drum. The flame front itself can be visualized by
the contour of temperature at the adiabatic flame temperature which is approximately 2,200 K
for methane at standard temperature, 298 K, and standard pressure, 101.325 kPa. Figure 5.28
shows that the wave front from the explosion overpressure is expanding much more quickly than
the flame front. Additionally, Figure 5.28 shows that the expanding pressure wave increases both
the pressure and temperatures of the unburned gases inside the face as noted by the preheat zone
on. This is important because increased preheating in the unburned gases can increase
189
|
Colorado School of Mines
|
CHAPTER 6
CONCLUSIONS AND RECOMMENDATIONS
This study shows the importance of understanding the airflow and gas mixtures distribution
across the longwall face in order to design an effective methane monitoring system. Current U.S.
regulations only specify that a methane sensor should be mounted on all cutting equipment and
other fixed locations, but does not specify the number of sensors or the optimum locations for
each placement. Extensive CFD modeling studies show that, when using these measurement
points, the current 1% and 2% methane concentration limit does not provide a good indication
whether explosive gas mixtures presents around the shearer drums. Likewise, the current
regulatory requirement and industry practice of maintaining a minimum amount of airflow at the
tailgate corner in combination with methane reading from two single point-based sensors
installed on the shearer body and tailgate drive is not adequate to warn of and prevent methane
ignition hazards at the face. In addition, using the same percent CH limits for both the shearer-
4
mounted and TG drive sensor can lead to either premature or late shearer shut-off, as the
effectiveness of point-based sensors is highly dependent on the sensor placement in relation to
the potential ignition source, which in this case the shearer drums.
The sensor installed on the tailgate drive is not reliable in detecting ignition hazards around
the shearer, as its readings do not relate to the shearer cutting position. The use of a TG drive
sensor to assess the ventilation condition in the tailgate corner area is also highly dependent on
the tailgate ventilation setup and the use of gob plate. Simulation results show that the use of a
gob plate or directing the flow outby the face can make tailgate sensor placement ineffective for
detecting ignition hazards directly at the coal face.
Original Contributions and Impact of the Research
Although several studies have been aimed at identifying airflow patterns and gas distribution
inside the longwall face, little work has been done in evaluating the effectiveness of the current
industry practice of relying on point-based methane readings for methane explosion prevention
in the active longwall face area.
194
|
Colorado School of Mines
|
• Using advanced CFD modeling techniques, this research has shown that the current, point-
type methane monitoring practice is insufficient to detect face ignition hazard at the
longwall face.
• This study presents and propose a new methane monitoring method to assess the ignition
risk at the longwall face on a conceptual level. These results can be used as a guideline to
help develop a more reliable explosion preventive system at the longwall face.
The UBB mine explosion in 2010 shows that explosive gas mixtures can formed around the
shearer drums without being detected by the methane sensors mounted on the shearer body and
tailgate drive. The CFD modeling approach presented in this research can be used to evaluate the
effectiveness of different sensor placement and assess the ventilation condition around the
shearer drums for different ventilation scenarios. By integrating the combustion model into the
ventilation model, the potential damage if face ignition occurs can be assessed. This information
is useful for evaluating the current explosion prevention and mitigation strategies, including
ventilation control placement, and development of explosion barriers.
Recommendation
Based on the simulation results, several recommendations can be made to improve the
ventilation condition and methane monitoring practice at the longwall face:
For longwall operation that utilize bleeder ventilation, the tailgate back return setup can help
reduce the face ignition risk during the tailgate corner cut-out. The fresh air supplied from the
tailgate entry outby the face can help to dilute and prevent methane accumulation at the tailgate
corner area. Shifting the flow inby the face also helps the methane sensor that is typically
installed on the tailgate drive to pick-up any sudden increase of methane around this area.
Looking at the methane distribution around the shearer, the shearer-mounted sensor should
be installed on the tail-side of the shearer body, and as close as possible to the coal face. Mine
operators and equipment manufacturers should consider installing additional sensors as close as
possible to the front and on the tail side of the shearer body.
The use of shearer cowls can increase the face ignition risk by blocking the incoming fresh
air to ventilate the shearer drums and trapping the incoming methane from the coal face between
the cowls and the drums, especially during the head-to-tail cutting sequence.
195
|
Colorado School of Mines
|
The use of a tailgate gob plate that covers the entire mining height may further reduce the
effectiveness of the tailgate drive sensor as the gob plate channels the airflow towards the front
of the shields.
Longwall mine operators should implement a more reliable methane monitoring system by
utilizing multiple sensors along the face. Knowing the correlation between methane readings at
the shield tips and actual explosion hazard established in this dissertation, regulators must
consider setting different methane concentration limits for machine mounted and fixed location
sensors depending on the sensor placement and set standard guidelines for the sensor placement.
By utilizing CFD modeling, this research has demonstrated that there is a more reliable
monitoring method to assess the ignition risk at the longwall face that can be used to complement
the current monitoring practice. The proposed methane explosion warning method relies on
multiple sensors reading that capture more accurately the peak methane concentration. Sensor
readings can be used to slow down or shut down the shearer prior to entering an ignition prone
area. The proposed multi-sensor warning system provides a more accurate representation of
potential explosive methane concentrations around the shearer drums compared to the current
monitoring practice that relies on two individual sensors placed on the shearer body and tailgate
drive. The multi-sensor system can be set up to give a warning, slow down the shearer, or shut
off the shearer when high methane concentrations, e.g. 3.5% +/-0.5% reading, are reported by
two consecutive shield tip sensors.
Recommendations for Future Work
The results presented in this research are still at the conceptual level and required further
study to assess the practicality of this method to be implemented in the actual mining
environment.
The following are not part of the scope of this research and should be considered to be
implemented in future research:
196
|
Colorado School of Mines
|
• Feasibility of the recommended sensor placement at the shield tips from a mechanical and
electronic technical perspective
• Modeling water sprays mounted on the shearer drums behind the picks
• Modeling physical rotation of the shearer drums cutting the coal face as transient model
• Transient model with moving shields and shearer that capture the real time methane
monitoring mechanism, considering time lag and sampling time, among others.
• More detailed modeling on the methane gas emanating from the coal face and inclusion of
other methane sources, including the face conveyor, crusher, roof and floor strata and the
gob.
• ANSYS Fluent can be coupled with ANSYS Mechanical for structural analysis. This can be
used to assess the potential damage of different mine explosion scenarios, which can lead to
improvements in mine design and ventilation layout, as well as structural design of
ventilation control structures and mine seals.
• Adaptive mine ventilation design and explosion hazard visualization through real-time
methane monitoring with data analytics. This can be achieved by using CFD model to
establish correlation of the measured methane concentration by the arrangement of sensors
located in multiple locations throughout the face and utilizing machine learning to setup a
system that can predict the existence of explosive mixtures ahead of the shearer and
automatically shut-off the shearer before it reach the explosion prone area.
197
|
Colorado School of Mines
|
APPENDIX A
EFFECT OF DRUMS ROTATING SPEED ON METHANE DISTRIBUTION AROUND
SHEARER DRUMS
Modern longwall shearer drums typically operate within 30 RPM to 50 RPM (Komatsu,
2018; Caterpillar, 2019). The ventilation case used in this test is when the shearer is located
between shield 140 and 146, cutting toward the tailgate. All three cases have the same amount of
fresh air (41 m3/s) supplied from the headgate side of the longwall face, and the same amount of
methane source (0.4 m3/s, pure methane) supplied from the uncut coal face. Figure A1 and
Figure A2 show comparison of methane distribution around the shearer drums for different
drums rotating speed.
The results show the importance of modeling the drums with rotating wall boundary
condition. The two cases with rotating shearer drums at 30 RPM and 50 RPM show higher
methane accumulation around the shearer drums compared to the case with stationary drums.
Higher shearer drums rotational speed induces higher turbulence around the drums, pulling more
methane from the surrounding area, resulting in higher methane accumulation around the drums,
as shown by comparing the 30 RPM and 50 RPM drums cases. However, there no significant
change in the overall results between modeling the drums rotating at 30 RPM or 50 RPM.
Figure A1: Plan view showing volume rendering of methane distribution around shearer drum
for different drums rotating speed
208
|
Colorado School of Mines
|
APPENDIX B
IMPACT OF SUPPLIED FACE AIR QUANTITY ON METHANE DISTRIBUTION
AROUND SHEARER
In this test, different supplied airflow quantity from headgate are compared, 41 m3/s and 33
m3/s, while the rest of the model setup remain the same. The gob is set to have 40% porosity and
viscous resistance value of 1.45x105 /m2. The shearer is located between shield 140 and 146,
cutting toward the tailgate. A total of 0.4 m3/s methane gas is supplied into the model through the
uncut coal face. Figure B1, Figure B2, and Figure B3 show comparison of methane
concentration for the two cases from different view point, while Figure B4 shows the comparison
of the reported methane concentration by the shearer mounted sensor and tailgate drive sensor
for the two cases.
Decreasing the amount of supplied fresh air from the headgate side of the face reduce the
amount of available fresh air that manage to reach the tailgate side of the face. This lack of fresh
air resulted in overall higher methane concentration inside the face, as can be seen by comparing
the methane iso-surfaces plots for both cases. The case with the 33 m3/s fresh air inlet shows that
the shearer tailgate drum is now covered by methane concentration higher than 5.5%.
This is also reflected in the reported methane concentration by the shearer and tailgate drive
sensors. The two cases show the same methane concentration trend, with overall higher methane
reported by the case with 33 m3/s supplied fresh air. In the second case, the shearer mounted
sensor should already trigger the automatic shut-off function as it reported methane
concentration higher than 2%, and the tailgate drive sensor should already give warning sign as it
reported methane concentration higher than 1%.
210
|
Colorado School of Mines
|
APPENDIX C
IMPACT OF GOB RESISTANCE ON METHANE DISTRIBUTION AROUND SHEARER
This section shows the results of parametric study for different gob resistances. In this test,
the gob resistance refers to the porous medium viscous resistance (/m2) used to represent the gob
directly behind the shields, up to 5 deeps into the gob. Three gob resistance values were tested in
this study: 1.5x105 /m2, 1.5x106 /m2, and 1.5x107 /m2 . For reference, the gob resistance value of
1.5x105 /m2 , based on study by Marts et al. (2015) was used as a base case for all the modeling
results presented in this report. The ventilation scenario used in this test is when the shearer is
located between shield 140 and 146, cutting toward the tailgate. All three cases have the same
amount of fresh air (41 m3/s) supplied from the headgate side of the longwall face, and the same
amount of methane source (0.4 m3/s, pure methane) supplied from the uncut coal face. Figure
C1 and Figure C2 show the comparison of the methane distribution around the shearer drums
from front view and plan view respectively.
Figure C1: Iso-surfaces plot showing methane distribution around the shearer drum from the
front view, for different gob resistance
213
|
Colorado School of Mines
|
ABSTRACT
A substantial amount of base and precious metals are produced globally using
open pit mining methods combined with heap leach processing of ores. Results from
mine planning and metallurgical testwork taken in conjunction with project capital costs,
operating costs, metal prices, project risk and corporate objectives are used to develop
heap leach parameters during project mine valuation and detailed engineering. Once in
operation, these parameters are then applied for the duration of the project unless leach
extraction results do not meet projections.
Selective placement of ores on a heap provides a technique whereby knowledge
of the ore characteristics and available leach capabilities, taken in conjunction with heap
design parameters and the mine plan, allow ores to be sequentially mined and placed in a
configuration on a heap so as to improve, if not optimize, leach extraction. Ores may be
selectively placed within a heap lift section by characteristic such that leach parameters
applied to the placed ore columns are targeted to produce better leach extraction results.
Ores may be segregated, averaged or placed according to some quantifiable characteristic
goal so as to achieve the desired lift section configuration.
A review of the literature revealed that no work to date had been performed on
sequential selective placement of ore as received from an open pit. Multiple random and
selective ore placement scenarios, each using 240 to 243 ore blocks per lift section, were
modeled to evaluate and compare leach extraction from each. Design criteria were
prepared to reflect typical mine and heap leach operating parameters. Characteristics for
over 250 ore blocks located on a mining bench within a Nevada gold deposit were
obtained for use in the model. The specific ore characteristics that were considered are
gold grade (toz Au/ton) and cyanide consumption (lbs CN/ton ore) of the ore when put
under leach.
iii
|
Colorado School of Mines
|
Selective placement scenarios and lift section configurations included segregating
ore blocks by their respective gold grades and averaging ore blocks by their respective
cyanide consumptions. Ore blocks segregated by grade were divided into either two
volumes of 120 blocks each or three volumes of 81 blocks each, composing one-half or
one-third of a lift section footprint, respectively.
Ore blocks selectively placed by cyanide consumption were averaged in one case
across one-third of a lift section row and in the other case across an entire lift section row.
The lift section configured with selectively placed ore blocks exhibited a 2.1% increase in
leach extraction in comparison to the same ore blocks configured randomly and leached
with an equal volume of lixiviant. Ultimately, overall leach extraction was improved
3.3% when the higher-grade lift section constructed was leached with an additional
lixiviant volume of 10%. Averaging the cyanide consumption characteristic through
selective placement yielded increases in the average grades of many ore columns while
substantially reducing the number of ore columns with inordinately high cyanide
consumption, a beneficial affect.
Selective ore placement may be used to improve leach extraction and decrease the affects
of deleterious constituents. Mine plans may be impacted to provide the best sequence of
ore blocks to a heap. Improved ore characterization and advances in leach technology
applied in conjunction with selective ore placement can improve process metallurgy and
project economics.
|
Colorado School of Mines
|
CHAPTER 1
INTRODUCTION
Exploitation of rocks and minerals from the earth has provided mankind with all
material items not derived from animals and vegetables. Extraction of specific rocks or
minerals from locations where deposits were naturally concentrated minimized the effort
required to obtain these resources. People needing these materials would congregate at
these locations and exploit the resources. These areas where rocks and minerals were
gathered became known as “mines” and those individuals that gathered rocks and
minerals became known as “miners”. The rocks and minerals gathered became known as
“ore” and the rocks and minerals that were of no value yet needed to be moved so as to
extract the ore became known as “waste”. These developments are well summarized in
the classic publication by Georg Bauer (pen name Georgius Agricola) titled “De Re
Metallica” in 1556.
Open pit mining coupled with heap leach extraction is a major and common
pairing of mining and processing techniques for the production of copper and gold
(Bartlett, R.W., 1997). Mined ore is moved and placed on a heap for lixiviant application
and metal extraction. Material below ore grade is transported to waste. Higher and/or
lower grade ores are occasionally treated on separate heaps, but this is an exception.
Very rarely are ores segregated on a primary heap. A schematic diagram of a typical
heap leach operations is presented in Figure 1.1 - Schematic of a Typical Heap Leach
Operation.
1
|
Colorado School of Mines
|
Evaporation Rain and Snow Ore to Heap
From Mine
Barren Solution Barren Solution
(Recycle) Application
Heap
(Metal Extraction)
Heap Leach Pad Liner System
Reagent Water Make-Up
Pregnant Solution
Make-Up From Well
Draindown
Pump Pump
Process Plant Pregnant Solution
(Metal Recovery) Pond
Figure 1.1 - Schematic of a Typical Heap Leach Operation
Heap leach facilities are typically constructed when low-grade ores are
encountered within a deposit wherein the metal contained is not economically viable to
extract otherwise. Heap leach technologies have evolved to handle randomly placed ores
that are commingled on a leach pad in one volumetric unit before leaching. The concept
developed and investigated herein considers benefiting from the selective placement of
ores by their distinguishable characteristics subsequently followed by targeted leaching of
the individual ore volumes generated during placement.
Ores have many characteristics that may enhance or be deleterious to
metallurgical leach extraction processes (Malhotra, D., 2006). Characteristics most
important to heap leach operations are evaluated during metallurgical testing and
development as mine valuation progresses though the feasibility and development stages.
Ore grade; however, is commonly the only characteristic that is closely followed
throughout mine valuation and operation.
2
|
Colorado School of Mines
|
Modem day best practice heap leach operations developed beginning
approximately 40 years ago, several decades before mine planning software and use of
the global positioning system (GPS) became an industry standard (Johnson, P.H., 1977).
Mines today commonly use GPS in conjunction with computerized mine planning /
scheduling / dispatching and database software to precisely locate both the mineral
reserve in the ground and the mining equipment in the pit (Zoschke, L.T., 2000). These
systems are used to track, manipulate and store information about ore characteristics and
the movement of material once mined. A photograph of a typical open-pit mine / heap
leach operations is presented in Figure 1.2- Girilambone Copper Company Heap
Leaching Operation, NSW.
The author contends that although being embraced and used on the most basic
level at virtually all mines, these advanced mine software and survey systems have not
been employed to their fullest extent. These systems have been extensively used to
improve mine planning, dispatch and fleet operations; however, there has been no
concerted effort to use these technologies to better plan the placement of ores on a heap.
There is an opportunity to substantially enhance metallurgical extraction by
applying these advanced mine software and survey systems to designate locations on a
heap where ore should be delivered while considering its metallurgical attributes. Leach
conditions may be improved and better controlled for each specific ore resulting in a
more efficient overall heap leach process through which more metal can be extracted
from the deposit. Understanding and committing to the benefits of selective placement
on a heap may also require that the mine alter, to some extent, the sequence of ore
excavation from the pit so as to deliver an improved, if not optimized, ore block sequence
for placement on the heap. Ultimately, more could be economically mined and leached
due to improved efficiencies resulting in decreased costs per unit of production.
3
|
Colorado School of Mines
|
Given a deposit deemed most economically amenable to exploitation by
employing heap leach processes, optimization beyond the use of current techniques
begins with more thorough evaluation of fundamental characteristics exhibited by ores.
Results from such evaluations would be used to influence, if not direct, the mining
sequence and ore placement location on the leach pad during construction of the lifts.
Ores would be segregated on the pad during lift construction by their relevant
metallurgical characteristics so as to allow these segregated ores to be treated with
improved leach extraction process parameters designed specifically for that ore.
Observation and consideration of ore characteristics in addition to grade should be
developed whereby enhancements to the leach extraction process may be achieved
through the knowledge of these characteristics and the use of this knowledge during mine
scheduling and heap placement.
The scope of this dissertation is limited to placement of run-of-mine ore from an
open pit onto a leach pad. Five comparative scenarios that focus on two ore
characteristics associated with the construction of one complete leachable lift section are
presented as examples. Models were developed to simulate leaching for the two ore
characteristics and were compared for both random ore placement and selective ore
placement scenarios on metallurgical and economic bases. Although a substantially
greater effort would be required to evaluate more ore characteristics, leach extraction
process parameters and multiple lifts for an entire heap leach operation, the work
presented explicitly demonstrates the concept and application to an extent that will allow
additional work to be performed in this area of study.
5
|
Colorado School of Mines
|
CHAPTER 2
LITERATURE SURVEY, HISTORICAL BACKGROUND, ENABLING
TECHNOLOGIES AND ADVANCED SELECTIVE
PLACEMENT OPPORTUNITIES
Three areas related to heap leaching and selective placement were reviewed in the
literature. Presented first is background pertaining to open pit mining and the heap leach
process. Second, mine information systems and global positioning systems as applied to
mine fleet operations are reviewed, both of which are enabling selective ore placement
technologies. Finally, literature applicable to the enhancement of selective ore placement
and the benefits there from in the areas of improved ore characterization, improved
selective placement decision making and improved leach conditions for selectively
placed ores is reviewed third and last.
Literature from several sources was surveyed including U.S. patents, U.S.
government documents, mining and mineral processing texts, papers presented at
industry symposiums and meetings, mining collage class notes, mine equipment vendor
information and recollection of discussions with mine industry experts. Additional
information from equipment vendors, engineering firms and other researchers has been
extracted from the world-wide-web and is referenced as such.
The library databases searched and the number of references found in each
database using the keywords “heap leach" are listed in Table 2.1 - Databases Searched
for Keywords “Heap Leach” and Number of References Found", below.
7
|
Colorado School of Mines
|
Table 2.1 - Databases Searched for Keywords “Heap Leach” and
Number of References Found
Database Number of references to keywords “heap leach”
U.S. Patents 116
Compendex 22
Metadex 136
Dissertations Advanced Search 13
Geobase 39
Georef 69
GeoScience World 85
SciFinder 225
The literature and information identified by the author as pertinent to the topic of
selective ore placement was reviewed for its applicability and is presented herein. Only
one document specifically discusses the concept of selective ore placement as a method
of enhancing metallurgical recovery from heap leach systems (Hurfst, U., 1989). It is
presented in the section on selective ore placement and decision process improvement.
No other literature was uncovered that directly relates to the proposed concept.
2.1 Open Pit Mining
Open pit mining is the method of choice when ores are within reach for extraction
from the earth’s surface. Several texts dedicated specifically to this type of mining
include Open Pit Mine Planning and Design, 1998, Hustrulid and Kuchta; Surface
Mining 2nd Edition, 1990, Kennedy ed., Open Pit Mine Planning and Design, 1979,
Hustrulid ed., and Surface Mining, 1968, Pfleider, E. P. ed. Additional references can be
8
|
Colorado School of Mines
|
found in the SME Mining Engineering Handbook, 2nd edition, 1992, Hartman ed. and the
SME Mining Engineering Handbook, 1973, Cummins A. B and Given I. A. eds.
Peck, J.P., 2000, acknowledges open pit operations as described above but
continues in more detail about the importance of mine fleet monitoring, communications
technologies and information management; these issues being requisite for selective ore
placement.
When considering selective placement techniques, it is imperative to ensure that
mine equipment selected for a mine fleet can operate within the physical parameters of
the proposed open pit and heap (Hustrulid, W., 1984). Information regarding operational
capacity, tipped bed dump height, turning radius and ground bearing pressure of haulage
trucks used to move excavated ore onto a heap leach pad may all be found in the
equipment manuals provided by the equipment manufactures. Caterpillar equipment was
considered in the preparation of this dissertation, the pertinent information being located
in the Caterpillar Performance Handbook, Volume 37, 2006. The internet addresses for
various equipment manufacturers that publish this information on the web can be found
in Table 2.2 - Mine Equipment Companies with Internet Addresses.
Table 2.2 - Mine Equipment Companies with Internet Addresses
Caterpillar - http://www.cat.com/.
Komatsu - http://www.komatsu.com/.
Euclid-Hitachi http://www.euclid-hitachi.com/. and
Terex - http://www.terex.com/
Liebherr - http://www.liebherr.com/lh/en/
Open pit haulage has been extensively examined in the literature due to its
significant contribution to mine costs (Sweigard, R. J., 1992, Ramani, R. V., 1990,
9
|
Colorado School of Mines
|
Crawford III, J. T., 1979). Ore is delivered to the pad after excavation either by truck or
conveyor depending on technical and economic preferences. Ore is stacked by the
dumping of ores received from the mine on top of ores having already been placed on the
heap earlier in the mining sequence. This is called stacking or lift building (Muhtadi,
O.A., 1988). Although detailed investigations to reduce haulage costs persist in the
literature, the author has found no investigations that technically or economically
consider selective final placement of ore on a heap. The author submits that no additional
time is transpired nor distance transversed to selectively place ore. The same amount of
material must be hauled over the same distance; therefore, selective ore placement only
influences the placement locations of sequentially delivered ore to the lift section.
Current mine cut sequencing techniques are discussed in the literature with regard
to geomechanics and operations (Hustrulid and Kuchta, 1998, p. 285); however, little if
any attention is given toward optimization of the block removal sequence with respect to
final placement on a heap (Miller, G., 1998). Mine fleet dynamics and dispatch is
discussed in the literature with reference to blend-quality and grade but only to the extent
that such blended end products are affected by the combined production from various
locations within a pit (Yingling, J. C., p. 801). Specific techniques for transporting and
stacking ores, as related to mine production scheduling, also consider that it is difficult to
maintain ore permeability when the stacking procedure has trucks continually rolling
over the placed ore material. This is commonly countered by plug dumping followed by
leveling with a dozer (Bernard, G.M., 1993).
2.2 Heap Leaching
Heap leaching combines geochemical, geotechnical and hydrometallurgical
techniques by which raw ores are statically treated on open-air impermeable leach pads to
10
|
Colorado School of Mines
|
extract the contained metals from ores that would otherwise not be economic to process
(Bartlett, R.W., 1997, Marsden, J.O., et. al., 1993). A detailed review of the heap leach
method still in practice today may be found in U.S. Patent 4,017,309 (Johnson, P., 1977).
Hustrulid and Kuchta, 1998, Malouf, E.E., 1990, Van Zyl, D.J.A., 1988 describe
the design, construction and operations of open pit mines and heap leach facilities
including aeration of a heap, size and depth of a heap, composition of ore and rock going
to a heap, deleterious constituents included with ore and the size of material placed.
Gold heap leaching as practiced today was first suggested U.S. Bureau of Mines
in 1967 and was adopted for commercial application in the early 1970’s (Thorstad, L.E.,
1987, Eisele, J.A. et. al. 1984). Before this time there had been dump leaching of copper
and uranium materials using sulfuric acid solutions as a lixiviant; however, the concise
methods used to uniformly stack material, apply solution and recover the extracted metals
were far less than optimal (Taylor, J.H. and Whelan, P.P., 1942).
With the advent of formal heap leach design and operation it became necessary to
develop ancillary industrial processes, including laboratory tests and procedures to
investigate ore characteristics, which could support metallurgical benefrciation from the
various ores. The U.S. Bureau of Mines (Heinen, H.J., McClelland, G.E. and Lindstrom,
R.E., 1979; McClelland, G.E., Pool, D.L. and Eisele, J.A., 1983; McClelland, G.E. and
Eisele, J.A., 1981) developed many of these tests. Additional work also continued in the
private sector (McClelland, G.E., 1988; McClelland, G.E. and van Zyl, D.J.A., 1988)..
Recent changes in mine management philosophy incorporate thinking which
embraces heap leaching as the preferred metal extraction process unless economics
(based on ore grade, metal price, metal recovery and project costs) dictate the need for
greater liberation and processing as achieved by mill/leach or mill/float operations.
Phelps Dodge Morenci operations, one of the largest open pit mines in the world,
exemplify this change in philosophy with their recently implemented “mine-to-leach”
operations.
11
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.