University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Virginia Tech
|
Repeating the methodology described in the previous sections, results were also obtained
for options 2, 4, and 6 – using the Spendrup 305-183-880 fan as an adjustable pitch fan
only, a fixed pitch fan with VFD, and an adjustable pitch fan with VFD, respectively.
These results are shown in Tables 3.5 through 3.7 below. For option 6, a plot of
efficiency capabilities, similar to that of the fan in Figure 4.7, is shown in Figure 4.10.
Note that the zone of highest efficiency (greater than 84%) is larger in this plot,
permitting more operating points to fall within this region.
Table 4.5 – Results for option 2, adjustable pitch Spendrup 305-183-880 fan
Fan Speed, Pitch Quantity, Head, in. Air Power, Brake Power, HP Calculated
Year rpm Setting kcfm WG HP (Vnet) (from fan curves) Efficiency
1 880 1 350 3.44 190 391 55%
2 880 1 350 3.44 190 391 55%
3 880 1 348 3.67 202 397 55%
4 880 1 347 3.88 212 403 55%
5 880 1 345 4.09 223 411 57%
6 880 1 344 4.29 232 416 57%
7 880 1 342 4.59 248 421 60%
8 880 1 340 4.91 263 430 60%
9 880 1 338 5.23 278 441 63%
10 880 2 428 7.14 481 638 75%
11 880 2 425 7.44 498 654 77%
12 880 2 420 7.93 524 677 79%
13 880 3 498 10.03 788 958 83%
14 880 3 491 10.56 817 991 84%
15 880 3 482 11.19 851 1023 84%
16 880 4 541 13.50 1150 1395 84%
45
|
Virginia Tech
|
Table 4.6 – Results for option 4, fixed pitch Spendrup 305-183-880 fan with VFD
Fan Speed, Pitch Quantity, Head, in. Air Power, Brake Power, HP Calculated
Year rpm Setting kcfm WG HP (Vnet) (from fan curves) Efficiency
1 340 4 220 1.78 62 76 82%
2 340 4 220 1.78 62 76 82%
3 355 4 229 1.97 71 87 82%
4 380 4 240 2.27 85 109 78%
5 405 4 256 2.69 108 132 82%
6 430 4 270 3.08 131 159 82%
7 460 4 286 3.58 161 197 82%
8 490 4 304 4.10 196 238 82%
9 525 4 324 4.75 243 295 83%
10 565 4 347 5.56 304 369 82%
11 600 4 368 6.29 365 442 83%
12 640 4 392 7.19 444 537 83%
13 685 4 418 8.28 546 661 83%
14 730 4 445 9.42 661 801 83%
15 780 4 475 10.79 808 978 83%
16 835 4 508 12.40 992 1201 83%
Table 4.7 – Results for option 6, adjustable pitch Spendrup 305-183-880 fan with VFD
Fan Speed, Fan Pitch Quantity, Head, in. Air Power, Brake Power, HP Calculated
Year rpm Setting kcfm WG HP (Vnet) (from fan laws) Efficiency
1 340 4 220 1.78 62 75 83%
2 340 4 220 1.78 62 75 83%
3 400 3 229 2.00 72 86 84%
4 425 3 241 2.31 88 105 83%
5 450 3 254 2.65 106 126 84%
6 480 3 270 3.07 131 154 85%
7 515 3 287 3.61 164 194 84%
8 550 3 306 4.16 201 237 85%
9 585 3 324 4.76 243 288 85%
10 625 3 346 5.49 299 351 85%
11 665 3 367 6.25 361 425 85%
12 710 3 391 7.18 443 519 85%
13 760 3 418 8.26 544 638 85%
14 810 3 445 9.40 659 773 85%
15 865 3 475 10.76 806 943 85%
16 835 4 508 12.40 992 1201 83%
46
|
Virginia Tech
|
For each iteration of the simulation, the speed, slip, and torque are calculated based on
the results of the previous cycle, up to the time limit of 60 seconds (1200 iterations). The
relevant results are the torque produced by the motor and the motor efficiency. Torque, τ,
is calculated by Equation 15, and efficiency, η , by Equation 16 which is based on the
m
calculated power in and power out. Power out, P , is found by Equation 14 using the
out
calculated values of torque and speed. Power in, P , is a function of input voltage and
in
required current, which is also calculated in the model.
3×V 2×(r /s)
τ= TH 2 (15)
ω×(R +r /s)2 +(X + X )2
TH 2 TH 2
Where, V = Thévenin voltage
TH
X = Thévenin reactance
TH
R = Thévenin resistance
TH
ω = rotational speed
r = rotor resistance
2
s = slip
P
η = out (16)
m P
in
Based on these equations, the motor was modeled over varying speeds and loads, and the
slip, mechanical speed, efficiency, and torque were recorded at steady state operation.
Table 4.8 contains a summary of the data.
50
|
Virginia Tech
|
Table 4.9 – Fan and motor efficiency comparison
Fan Only (1) Fan with VFD (2) Difference (2)-(1)
Total Total Total
Fan Eff, Motor Eff, Power, Fan Eff, Motor Eff, Power, Fan Eff, Motor Eff, Power,
Year % % HP % % HP % % HP
1 38 87 283 79 68 115 40 -19 -168
2 38 87 283 79 68 123 40 -19 -161
3 39 87 289 74 73 135 35 -14 -155
4 40 88 303 76 76 158 35 -12 -144
5 57 90 376 76 79 184 19 -11 -192
6 58 90 393 77 81 215 19 -9 -178
7 60 90 403 76 84 255 16 -6 -148
8 64 90 406 77 86 308 13 -4 -99
9 70 92 548 76 88 369 6 -4 -179
10 71 92 566 74 89 454 3 -3 -111
11 72 92 571 77 90 532 5 -2 -39
12 78 93 759 77 91 648 -1 -2 -111
13 80 93 773 75 92 787 -5 -1 14
14 80 94 1030 78 93 945 -3 -1 -85
15 77 94 1401 77 93 1148 0 -1 -253
16 77 93 1430 77 93 1420 1 0 -10
54
|
Virginia Tech
|
4.6 Summary and Discussion of VFD Effects on Fan Operation
With the modeling complete, the next stage of the research was to analyze the use of
VFDs and try to explain any trends or effects of such operation. A short list of each
option modeled is contained in Table 4.10. Table 4.11 contains a summary of the
updated power requirements for each modeled option at all operating points. These
values serve as the basis for more accurate comparison between the options. The power
input values were calculated based on the air power required and losses from the VFD,
motor, and fan (Equation 17).
1 1 1
P = P × × × (17)
in a η η η
d m f
Where, P = power input required
in
P = air power
a
η = efficiency of the drive, d, motor, m, and fan, f
Table 4.10 – Summary of scenarios modeled
Option Scenario
1 Fan 1, adjustable pitch
2 Fan 2, adjustable pitch
3 Fan 1, with VFD
4 Fan 2, with VFD
5 Fan 1, adjustable pitch and VFD
6 Fan 2, adjustable pitch and VFD
Fan 1: Spendrup 274-165-880
Fan 2: Spendrup 305-183-880
55
|
Virginia Tech
|
Table 4.11 – Total input power requirements, HP
Option
Year 1 2 3 4 5 6
1 279 381 116 115 114 114
2 278 381 116 115 114 114
3 283 404 135 126 136 126
4 297 424 158 150 158 145
5 369 430 184 175 184 167
6 385 446 215 203 212 197
7 395 453 255 242 246 239
8 398 480 308 286 302 285
9 537 483 369 345 355 337
10 554 690 450 428 437 402
11 560 695 532 507 526 482
12 744 712 648 603 617 582
13 758 1,014 782 741 755 707
14 1,009 1,038 945 888 917 849
15 1,373 1,082 1,148 1,073 1,148 1,024
16 1,401 1,466 1,420 1,318 1,420 1,318
Overall, operating with a VFD in options 3-6 reduces the input power requirements.
Referring to Table 4.9, comparing the efficiencies of Spendrup 274-165-880 fan at
adjustable pitch only (option one) and variable speed only (option three), there is a
definite advantage to using a variable speed drive. When combined with the
corresponding loss of motor efficiency the advantage is slightly diminished. These
differences become more apparent when the options are compared against a single
standard, in this case, operation with option one The power differences compared to
option one for each year are shown in Table 4.12. Negative values represent a power
savings over option one.
56
|
Virginia Tech
|
Table 4.12 – Power difference for each option vs. option one
Option
Year 1 2 3 4 5 6
1 0 102 -163 -164 -165 -165
2 0 103 -162 -162 -164 -163
3 0 121 -149 -157 -147 -158
4 0 128 -138 -146 -138 -151
5 0 61 -185 -194 -185 -202
6 0 62 -170 -182 -173 -188
7 0 58 -140 -153 -149 -156
8 0 82 -91 -112 -96 -114
9 0 -54 -168 -192 -182 -200
10 0 136 -104 -127 -117 -152
11 0 136 -28 -53 -34 -78
12 0 -32 -96 -141 -126 -162
13 0 256 25 -16 -3 -50
14 0 29 -65 -121 -92 -161
15 0 -291 -225 -300 -225 -349
16 0 65 19 -84 19 -83
As expected, with the larger fan in option two, the power required is higher, except at
three operating points. The lower power at these points can be explained by the increased
efficiency of the larger fan and low performance of the small fan. At these points, the
smaller fan has a pitch setting that does not quite meet the ventilation requirements,
resulting in operation at the next higher pitch setting which then results in a higher than
required head and quantity. It is at these points that the larger fan can operate at a distinct
pitch setting without over-ventilating. If both fans were operated at an infinite range of
pitch settings, the smaller fan would always require lower power than the larger one.
Using the larger fan with a VFD though, in options four and six, results in lower power
requirements than using the smaller fan conventionally. The higher efficiency that the
fan is capable of over a wider range of operating conditions is the reason for this trend.
Although the power consumption is lower, the determination of whether this is an
economically feasible option is discussed in Chapter 4.
During certain years the power values do not fit the expected trend. These instances
occur where power requirements are greater for fans operating with a VFD than without.
The case of year 13, option three versus option one, is one such example. The fan
without a VFD is capable of meeting the airflow requirements (428 kcfm at 8.38 in. w.g.)
at a speed of 880 rpm with an efficiency of 80%. When this fan is operated at variable
speeds, the operating point, 414 kcfm at 8.10 in. w.g., is met at a fan speed of 715 rpm
with 75% fan efficiency. Although this results in a lower air power requirement (529 hp
57
|
Virginia Tech
|
vs. 563 hp), accounting for the decrease of fan efficiency results in a greater brake power.
This occurs for a combination of the following two reasons. The method used to
determine optimum fan speed relied on meeting the optimum operating conditions as
established in the ventilation model, and as such, did not account for fan performance.
Essentially the process consisted of minimizing the air power requirements without
regard for optimizing fan operation. Although this effect is present in all years of
operation, it is not noticeable because the increased fan efficiency and reduced power
requirements compensate for this small difference.
A second cause for the decrease in performance is that the variable speed fan cannot
operate at exactly the same point as the fixed speed fan because the precision of speed
adjustment was limited to 5 rpm. In addition, there was a 1% loss of motor efficiency
and a 2% loss of efficiency through the drive that resulted in greater power requirements.
These effects are more apparent in this case because the original fan (option one) was
already operating at an efficient point without a VFD. This is more evident in year 16,
where the fan operates at approximately the same head, quantity, and efficiency with and
without a VFD. There is a 19 hp increase in total power when a VFD is used, which is
primarily caused by the 2% loss of efficiency through the drive.
The results at years 13 and 16 do highlight one of the issues regarding variable speed
fans, specifically, how accurately can or should the fan be adjusted. A few operating
points were re-modeled with speeds accurate to 1 rpm, and as such, the optimum fan
speed was found to be slightly lower than when an accuracy of 5 rpm was used. The
lower speed decreases the power requirements a small amount as well. For this research,
as in an actual mining environment, certain assumptions as estimates were used. There is
variability in how the regulators and injectors are operated, as well as what the precise
ventilation requirements at the face are. It did not seem practical to use a fan speed
accurate to within 1 rpm, when a minor change in an injector or regulator in the mine
model, or a requirement change of a few hundred cfm of airflow at the face or belt could
change the required fan speed by more than few rpm. The accuracy of 5 rpm was
deemed sufficient to illustrate the benefits of reduced speed.
The general trend seen for all options using a VFD is that if the fan is sized for the most
extreme operating conditions, during lower ventilation requirements there is more benefit
of using variable speed than when then fan is operating nearer its designed operating
range. This was the initial assumption that is now supported by the modeling. Power
58
|
Virginia Tech
|
1,000,000
800,000
600,000
400,000
200,000
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
(200,000)
(400,000)
(600,000)
(800,000)
Year
sralloD
5002
Option 2
Option 3
Option 4
Option 5
Option 6
Figure 5.1 – Cumulative cash flow – savings over option 1
To properly compare the options economically, this situation is a case of looking at
mutually exclusive alternatives for service producing projects. Because these values
represent costs, and the initial capital investment yields a decrease in operating costs, the
option with the minimum cost is desired (Stermole and Stermole 2000).
Using the net cost analysis, the net present cost of each option is calculated to the nearest
$1000, at minimum rates of return (ROR) of 10, 20, 30, and 40%, shown in Table 5.3.
With a rate of return under 10%, option 6 has the minimum cost, and is therefore the best
option, but at 20-30% rates of return, option 5 has the minimum cost. If the minimum
rate of return is 40%, it is better to do nothing and keep option 1.
Table 5.3 – Net present cost for each option, $
Option
Interest Rate 1 2 3 4 5 6
10% 1,807,000 2,140,000 1,556,000 1,565,000 1,534,000 1,531,000
20% 997,000 1,250,000 901,000 949,000 892,000 935,000
30% 680,000 888,000 664,000 726,000 659,000 719,000
40% 530,000 710,000 561,000 628,000 558,000 625,000
The rate of return for each option can also be directly calculated, as shown in Table 5.4.
Since only options 3-6 yield results better than option 1, these differences are used to
62
|
Virginia Tech
|
determine an incremental rate of return. The minimum attractive rate of return for
comparison is 30%. Although using the larger fan with the VFD results in cost savings
over option 1, the minimum attractive rate of return is not met, therefore, these options
are not economically viable.
Table 5.4 – Rate of return for each option
Option 3 4 5 6
ROR 33.0% 24.4% 33.7% 25.4%
With a decrease in capital costs on fixed pitch fans as little as 4%, option 3 becomes the
best option at a ROR of 35%. Table 5.5 shows the resulting ROR based on a 6%
reduction in price for fixed pitch fans. If fixed pitch fans can be manufactured for less
than adjustable pitch fans, using adjustable pitch fans with a VFD is a potentially less
economical option.
Table 5.5 – ROR for lower cost fixed pitch fans
Option 3 4 5 6
ROR 34.8% 25.7% 33.7% 25.4%
5.2 Additional Considerations
There are other benefits of using a VFD, aside from the yearly energy savings. There is
no drawback to installing a fan that is suitably sized for the future or is even oversized, at
the beginning of the mine life, since it is only operated at the required conditions. This
can actually lead to avoiding the situation of needing a fan replacement when pressure
and head are increased at a later point of the mine’s life. At a cost of increased power
consumption, the fan can even be operated at higher than rated speeds if necessary. The
effects of such action on the motor and fan have not been evaluated, but the drive has the
capability. In mines with multiple fans, it should be possible to use VFDs to balance
each fan’s performance such that the entire mine’s ventilation is optimized.
In any mine, having a fan that is capable of a smooth start from 0 rpm to operating speed
reduces the inrush current normally associated with powering up a large motor. As
described by Sartain and McDaniel of a P&M Coal Mining Co. operation in Alabama,
this results in minimizing the size of the standby generator that is required to provide
80% backup capability. In this example, P&M was able to reduce the size of their
63
|
Virginia Tech
|
CHAPTER 6: SUMMARY AND CONCLUSIONS
Primary mine ventilation fans provide airflow and pressure to properly circulate air to the
working faces in the mine. The most commonly used primary mine fan is the axial flow
type. This type of fan traditionally operates at a fixed speed based on motor design and
power supplied to the fan. Over the life of the mine, ventilation requirements change
with changing mine conditions and layouts. Usually, there is a time when the fan is
underutilized, is operated at a higher than necessary point, or operated at its rated
capacity. In response to the changing conditions of the mine, it is common practice to
control airflow with either dampers or by changing the pitch of the fan blades. However,
this method of fan operation results in inefficiencies during the life of the mine. Most
often, a mine will operate a fan at a higher airflow than necessary when low ventilation
requirements exist, simply because it is more efficient than operating the fan at lower
conditions. In all but a narrow range of airflows and pressures, the fan does not operate
at its optimum efficiency.
A solution to this problem is the use of a variable frequency drive with the fan motor.
This drive allows the fan to operate at reduced speeds and power. The fan speed, and
consequently the operating point, can be adjusted to maintain a high fan efficiency while
still meeting the ventilating requirements of the mine throughout the mine life. There are
many advantages to this approach of controlling airflow. If a VFD is used to control fan
speed, energy savings can be realized by operating a fan at a point that meets the
minimum ventilation requirements without wasting energy. The use of VFDs also
provides flexibility in mine operations by potentially allowing the fan to be adjusted as
needed or “on the fly” during mining.
Technical requirements of ventilation, VFDs, motors, and fans were reviewed to
determine the feasibility of installing and using VFDs for mine fan applications. A
simple coal mine was designed and modeled to perform a ventilation analysis that
incorporates variable speed fans. It was then used to develop and illustrate a method for
determining the appropriate reduced speed fan settings to achieve optimum operating
conditions. This research illustrates some of the possible benefits of using a variable
frequency drives in mine fan applications. Although using VFDs decreases motor
efficiency at slower speeds, the significant increase in fan efficiency more than
compensates for this loss.
65
|
Virginia Tech
|
Economic considerations were outlined for the example mine, but due to the numerous
factors required that are specific to an individual mining operation, exact economic
implications could not be made. The general trend seen with the mine that was modeled
indicates that there are definite energy savings possible by using a VFD. The magnitude
of these savings combined with the capital investment required will determine whether
using VFDs is a viable alternative to conventional mine ventilation. This research shows
the potential of significant energy savings, and that the application of VFDs warrants
further consideration in ventilation planning.
6.1 Future Research
The effects of VFDs on more complicated mines and mines with multiple fans warrants
further investigation. Since each individual fan can be operated with a VFD, it should be
possible to optimize the entire ventilation system to a relatively precise state. Areas of
future research on this subject also include methods of controlling the fan speeds
dynamically. If the conditions at certain locations in the mine are monitored for air
quantity and quality, it might be possible to adjust the fan speed as needed to meet
ventilation requirements.
66
|
Virginia Tech
|
Table B.1 – Year 16 branch properties
Resistance Calculated
Resistance per Length Length Parallel Resistance Branch Surface
From To Fqi Type (P.U.) (R/1000ft) (ft) Factor (P.U.) Code State Description
10 100 R/L 0.0264 800 1 0.02112 Intake Intake Intake Shaft
30 300 R/L 0.3033 2816 1 0.85409 Neutral Intake Belt Slope (Intake)
200 20 F R/L 0.0013 800 1 0.00104 Return Exhaust Exhaust Shaft
100 104 R/L 0.0266 1000 3 0.00296 Intake Neither
104 106 R/L 0.0266 1000 3 0.00296 Intake Neither
106 110 R/L 0.0266 1000 3 0.00296 Intake Neither
110 112 R/L 0.0266 1000 3 0.00296 Intake Neither
112 116 R/L 0.0266 1000 3 0.00296 Intake Neither
116 118 R/L 0.0266 1000 3 0.00296 Intake Neither
118 122 R/L 0.0266 1000 3 0.00296 Intake Neither
122 124 R/L 0.0266 1000 3 0.00296 Intake Neither
124 128 R/L 0.0266 1000 3 0.00296 Intake Neither
128 130 R/L 0.0266 1000 3 0.00296 Intake Neither
130 134 R/L 0.0266 1000 3 0.00296 Intake Neither
134 136 R/L 0.0266 1000 3 0.00296 Intake Neither
136 140 R/L 0.0266 1000 3 0.00296 Intake Neither
140 142 R/L 0.0266 1000 3 0.00296 Intake Neither
142 146 R/L 0.0266 1000 3 0.00296 Intake Neither
146 150 R/L 0.0266 700 3 0.00207 Intake Neither
150 152 R/L 0.0266 500 3 0.00148 Intake Neither
152 154 R/L 0.0266 300 3 0.00089 Intake Neither
154 158 R/L 0.0266 1000 3 0.00296 Intake Neither
158 160 R/L 0.0266 100 3 0.0003 Intake Neither
160 162 Q R/L 0.0266 500 3 0.00148 Intake Neither Face 1
162 260 R/L 0.0301 250 1 0.00752 Return Neither
162 460 R/L 0.0301 250 1 0.00752 Return Neither
300 304 R/L 0.0428 1000 1 0.0428 Neutral Neither
304 306 R/L 0.0428 1000 1 0.0428 Neutral Neither
306 310 R/L 0.0428 1000 1 0.0428 Neutral Neither
310 312 R/L 0.0428 1000 1 0.0428 Neutral Neither
312 316 R/L 0.0428 1000 1 0.0428 Neutral Neither
316 318 R/L 0.0428 1000 1 0.0428 Neutral Neither
318 322 R/L 0.0428 1000 1 0.0428 Neutral Neither
322 324 R/L 0.0428 1000 1 0.0428 Neutral Neither
324 328 R/L 0.0428 1000 1 0.0428 Neutral Neither
328 330 R/L 0.0428 1000 1 0.0428 Neutral Neither
330 334 R/L 0.0428 1000 1 0.0428 Neutral Neither
334 336 R/L 0.0428 1000 1 0.0428 Neutral Neither
336 340 R/L 0.0428 1000 1 0.0428 Neutral Neither
340 342 R/L 0.0428 1000 1 0.0428 Neutral Neither
342 346 R/L 0.0428 1000 1 0.0428 Neutral Neither
346 350 R/L 0.0428 800 1 0.03424 Neutral Neither
350 352 R/L 0.0428 500 1 0.0214 Neutral Neither
352 354 R/L 0.0428 200 1 0.00856 Neutral Neither
354 358 R/L 0.0428 1000 1 0.0428 Neutral Neither
358 360 R/L 0.0428 100 1 0.00428 Neutral Neither
360 260 Q R/L 0.0428 200 1 0.00856 Neutral Neither Belt 1
200 202 R/L 0.0301 1000 2 0.00752 Return Neither
202 204 R/L 0.0301 0 2 0 Return Neither
204 206 R/L 0.0301 1000 2 0.00752 Return Neither
206 208 R/L 0.0301 1000 2 0.00752 Return Neither
208 210 R/L 0.0301 0 2 0 Return Neither
210 212 R/L 0.0301 1000 2 0.00752 Return Neither
212 214 R/L 0.0301 1000 2 0.00752 Return Neither
214 216 R/L 0.0301 0 2 0 Return Neither
216 218 R/L 0.0301 1000 2 0.00752 Return Neither
218 220 R/L 0.0301 1000 2 0.00752 Return Neither
90
|
Virginia Tech
|
Table B.1 cont. – Year 16 Branch Properties
Resistance Calculated
Resistance per Length Length Parallel Resistance Branch Surface
From To Fqi Type (P.U.) (R/1000ft) (ft) Factor (P.U.) Code State Description
220 222 R/L 0.0301 0 2 0 Return Neither
222 224 R/L 0.0301 1000 2 0.00752 Return Neither
224 226 R/L 0.0301 1000 2 0.00752 Return Neither
226 228 R/L 0.0301 0 2 0 Return Neither
228 230 R/L 0.0301 1000 2 0.00752 Return Neither
230 232 R/L 0.0301 1000 2 0.00752 Return Neither
232 234 R/L 0.0301 0 2 0 Return Neither
234 236 R/L 0.0301 1000 2 0.00752 Return Neither
236 238 R/L 0.0301 1000 2 0.00752 Return Neither
238 240 R/L 0.0301 0 2 0 Return Neither
240 242 R/L 0.0301 1000 2 0.00752 Return Neither
242 244 R/L 0.0301 1000 2 0.00752 Return Neither
244 246 R/L 0.0301 0 2 0 Return Neither
246 250 R/L 0.0301 500 2 0.00376 Return Neither
250 252 R/L 0.0301 500 2 0.00376 Return Neither
252 254 R/L 0.0301 500 2 0.00376 Return Neither
254 256 R/L 0.0301 1000 2 0.00752 Return Neither From Face 1
256 258 R/L 0.0301 500 2 0.00376 Return Neither
258 260 R/L 0.0301 100 2 0.00075 Return Neither
400 402 R/L 0.0301 1000 2 0.00752 Return Neither
402 404 R/L 0.0301 0 2 0 Return Neither
404 406 R/L 0.0301 1000 2 0.00752 Return Neither
406 408 R/L 0.0301 1000 2 0.00752 Return Neither
408 410 R/L 0.0301 0 2 0 Return Neither
410 412 R/L 0.0301 1000 2 0.00752 Return Neither
412 414 R/L 0.0301 1000 2 0.00752 Return Neither
414 416 R/L 0.0301 0 2 0 Return Neither
416 418 R/L 0.0301 1000 2 0.00752 Return Neither
418 420 R/L 0.0301 1000 2 0.00752 Return Neither
420 422 R/L 0.0301 0 2 0 Return Neither
422 424 R/L 0.0301 1000 2 0.00752 Return Neither
424 426 R/L 0.0301 1000 2 0.00752 Return Neither
426 428 R/L 0.0301 0 2 0 Return Neither
428 430 R/L 0.0301 1000 2 0.00752 Return Neither
430 432 R/L 0.0301 1000 2 0.00752 Return Neither
432 434 R/L 0.0301 0 2 0 Return Neither
434 436 R/L 0.0301 1000 2 0.00752 Return Neither
436 438 R/L 0.0301 1000 2 0.00752 Return Neither
438 440 R/L 0.0301 0 2 0 Return Neither
440 442 R/L 0.0301 1000 2 0.00752 Return Neither
442 444 R/L 0.0301 1000 2 0.00752 Return Neither
444 446 R/L 0.0301 0 2 0 Return Neither
446 450 R/L 0.0301 500 2 0.00376 Return Neither
450 452 R/L 0.0301 500 2 0.00376 Return Neither
452 454 R/L 0.0301 500 2 0.00376 Return Neither
454 456 R/L 0.0301 1000 2 0.00752 Return Neither From Face 1
456 458 R/L 0.0301 500 2 0.00376 Return Neither
458 460 R/L 0.0301 100 2 0.00075 Return Neither
200 400 R/L 0.0301 700 1 0.02107 Return Neither
202 402 R/L 0.0301 700 1 0.02107 Return Neither
208 408 R/L 0.0301 700 1 0.02107 Return Neither
214 414 R/L 0.0301 700 1 0.02107 Return Neither
220 420 R/L 0.0301 700 1 0.02107 Return Neither
226 426 R/L 0.0301 700 1 0.02107 Return Neither
91
|
Virginia Tech
|
Table B.1 cont. – Year 16 Branch Properties
Resistance Calculated
Resistance per Length Length Parallel Resistance Branch Surface
From To Fqi Type (P.U.) (R/1000ft) (ft) Factor (P.U.) Code State Description
232 432 R/L 0.0301 700 1 0.02107 Return Neither
238 438 R/L 0.0301 700 1 0.02107 Return Neither
244 444 R/L 0.0301 700 1 0.02107 Return Neither
256 456 R/L 0.0301 700 1 0.02107 Return Neither
204 304 R 22.5 1 22.5 Default Neither
304 104 R 22.5 1 22.5 Default Neither
104 404 R 22.5 1 22.5 Default Neither
254 354 R 22.5 1 22.5 Default Neither
354 154 R 22.5 1 22.5 Default Neither
154 454 R 22.5 1 22.5 Default Neither
258 358 R 22.5 1 22.5 Default Neither
358 158 R 22.5 1 22.5 Default Neither
158 458 R 22.5 1 22.5 Default Neither
206 306 R 22.5 1 22.5 Default Neither
306 106 R 22.5 1 22.5 Default Neither
106 406 R 22.5 1 22.5 Default Neither
210 310 R 22.5 1 22.5 Default Neither
310 110 R 22.5 1 22.5 Default Neither
110 410 R 22.5 1 22.5 Default Neither
212 312 R 22.5 1 22.5 Default Neither
312 112 R 22.5 1 22.5 Default Neither
112 412 R 22.5 1 22.5 Default Neither
216 316 R 22.5 1 22.5 Default Neither
316 116 R 22.5 1 22.5 Default Neither
116 416 R 22.5 1 22.5 Default Neither
218 318 R 22.5 1 22.5 Default Neither
318 118 R 22.5 1 22.5 Default Neither
118 418 R 22.5 1 22.5 Default Neither
222 322 R 22.5 1 22.5 Default Neither
322 122 R 22.5 1 22.5 Default Neither
122 422 R 22.5 1 22.5 Default Neither
224 324 R 22.5 1 22.5 Default Neither
324 124 R 22.5 1 22.5 Default Neither
124 424 R 22.5 1 22.5 Default Neither
228 328 R 22.5 1 22.5 Default Neither
328 128 R 22.5 1 22.5 Default Neither
128 428 R 22.5 1 22.5 Default Neither
230 330 R 22.5 1 22.5 Default Neither
330 130 R 22.5 1 22.5 Default Neither
130 430 R 22.5 1 22.5 Default Neither
234 334 R 22.5 1 22.5 Default Neither
334 134 R 22.5 1 22.5 Default Neither
134 434 R 22.5 1 22.5 Default Neither
236 336 R 22.5 1 22.5 Default Neither
336 136 R 22.5 1 22.5 Default Neither
136 436 R 22.5 1 22.5 Default Neither
240 340 R 22.5 1 22.5 Default Neither
340 140 R 22.5 1 22.5 Default Neither
140 440 R 22.5 1 22.5 Default Neither
242 342 R 22.5 1 22.5 Default Neither
342 142 R 22.5 1 22.5 Default Neither
142 442 R 22.5 1 22.5 Default Neither
246 346 R 22.5 1 22.5 Default Neither
346 146 R 22.5 1 22.5 Default Neither
92
|
Virginia Tech
|
Table B.1 cont. – Year 16 Branch Properties
Resistance Calculated
Resistance per Length Length Parallel Resistance Branch Surface
From To Fqi Type (P.U.) (R/1000ft) (ft) Factor (P.U.) Code State Description
146 446 R 22.5 1 22.5 Default Neither
150 500 R/L 0.0266 1300 2 0.00864 Intake Neither
500 501 R/L 0.0266 500 2 0.00332 Intake Neither
501 502 R/L 0.0266 500 2 0.00332 Intake Neither
502 504 R/L 0.0266 1000 2 0.00665 Intake Neither
504 506 R/L 0.0266 1000 2 0.00665 Intake Neither
506 508 R/L 0.0266 1000 2 0.00665 Intake Neither
508 510 R/L 0.0266 1000 2 0.00665 Intake Neither
510 512 R/L 0.0266 1000 2 0.00665 Intake Neither
512 514 R/L 0.0266 800 2 0.00532 Intake Neither
514 518 R/L 0.0266 100 2 0.00066 Intake Neither North Face
518 616 R/L 0.0301 250 1 0.00752 Return Neither
518 636 R/L 0.0301 250 1 0.00752 Return Neither
350 700 R/L 0.0428 1200 1 0.05136 Neutral Neither
700 701 R/L 0.0428 500 1 0.0214 Neutral Neither
701 702 R/L 0.0428 500 1 0.0214 Neutral Neither
702 704 R/L 0.0428 1000 1 0.0428 Neutral Neither
704 706 R/L 0.0428 1000 1 0.0428 Neutral Neither
706 708 R/L 0.0428 1000 1 0.0428 Neutral Neither
708 710 R/L 0.0428 1000 1 0.0428 Neutral Neither
710 712 R/L 0.0428 1000 1 0.0428 Neutral Neither
712 714 R/L 0.0428 800 1 0.03424 Neutral Neither
714 716 R/L 0.0428 0 1 0 Neutral Neither
716 636 Q R/L 0.0428 250 1 0.0107 Neutral Neither North Belt
600 250 Q R/L 0.0301 1000 1 0.0301 Return Neither
600 602 R/L 0.0301 1000 1 0.0301 Return Neither From N Face
602 604 R/L 0.0301 1000 1 0.0301 Return Neither
604 606 R/L 0.0301 1000 1 0.0301 Return Neither
606 608 R/L 0.0301 1000 1 0.0301 Return Neither
608 610 R/L 0.0301 1000 1 0.0301 Return Neither
610 612 R/L 0.0301 1000 1 0.0301 Return Neither
612 614 R/L 0.0301 800 1 0.02408 Return Neither
614 616 R/L 0.0301 0 1 0 Return Neither
620 252 Q R/L 0.0301 1000 1 0.0301 Return Neither
620 622 R/L 0.0301 1000 1 0.0301 Return Neither From N Face
622 624 R/L 0.0301 1000 1 0.0301 Return Neither
624 626 R/L 0.0301 1000 1 0.0301 Return Neither
626 628 R/L 0.0301 1000 1 0.0301 Return Neither
628 630 R/L 0.0301 1000 1 0.0301 Return Neither
630 632 R/L 0.0301 1000 1 0.0301 Return Neither
632 634 R/L 0.0301 800 1 0.02408 Return Neither
634 636 R/L 0.0301 0 1 0 Return Neither
600 500 R 22.5 1 22.5 Default Neither
500 700 R 22.5 1 22.5 Default Neither
700 620 R 22.5 1 22.5 Default Neither
602 502 R 22.5 1 22.5 Default Neither
502 702 R 22.5 1 22.5 Default Neither
702 622 R 22.5 1 22.5 Default Neither
604 504 R 22.5 1 22.5 Default Neither
504 704 R 22.5 1 22.5 Default Neither
704 624 R 22.5 1 22.5 Default Neither
606 506 R 22.5 1 22.5 Default Neither
506 706 R 22.5 1 22.5 Default Neither
706 626 R 22.5 1 22.5 Default Neither
93
|
Virginia Tech
|
Table B.1 cont. – Year 16 Branch Properties
Resistance Calculated
Resistance per Length Length Parallel Resistance Branch Surface
From To Fqi Type (P.U.) (R/1000ft) (ft) Factor (P.U.) Code State Description
608 508 R 22.5 1 22.5 Default Neither
508 708 R 22.5 1 22.5 Default Neither
708 628 R 22.5 1 22.5 Default Neither
610 510 R 22.5 1 22.5 Default Neither
510 710 R 22.5 1 22.5 Default Neither
710 630 R 22.5 1 22.5 Default Neither
612 512 R 22.5 1 22.5 Default Neither
512 712 R 22.5 1 22.5 Default Neither
712 632 R 22.5 1 22.5 Default Neither
614 514 R 35.2 1 35.2 Default Neither
514 714 R 35.2 1 35.2 Default Neither
714 634 R 35.2 1 35.2 Default Neither
152 550 R/L 0.0266 1000 2 0.00665 Intake Neither
550 551 R/L 0.0266 500 2 0.00332 Intake Neither
551 552 R/L 0.0266 500 2 0.00332 Intake Neither
552 554 R/L 0.0266 1000 2 0.00665 Intake Neither
554 556 R/L 0.0266 1000 2 0.00665 Intake Neither
556 558 R/L 0.0266 1000 2 0.00665 Intake Neither
558 560 R/L 0.0266 1000 2 0.00665 Intake Neither
560 562 R/L 0.0266 1000 2 0.00665 Intake Neither
562 564 R/L 0.0266 800 2 0.00532 Intake Neither
564 568 R/L 0.0266 100 2 0.00066 Intake Neither South Face
568 666 R/L 0.0301 250 1 0.00752 Return Neither
568 686 R/L 0.0301 250 1 0.00752 Return Neither
352 750 R/L 0.0428 1000 1 0.0428 Neutral Neither
750 751 R/L 0.0428 500 1 0.0214 Neutral Neither
751 752 R/L 0.0428 500 1 0.0214 Neutral Neither
752 754 R/L 0.0428 1000 1 0.0428 Neutral Neither
754 756 R/L 0.0428 1000 1 0.0428 Neutral Neither
756 758 R/L 0.0428 1000 1 0.0428 Neutral Neither
758 760 R/L 0.0428 1000 1 0.0428 Neutral Neither
760 762 R/L 0.0428 1000 1 0.0428 Neutral Neither
762 764 R/L 0.0428 800 1 0.03424 Neutral Neither
764 766 R/L 0.0428 0 1 0 Neutral Neither
766 666 Q R/L 0.0301 250 1 0.00752 Return Neither South Belt
650 450 Q R/L 0.0301 1000 1 0.0301 Return Neither From S Face
650 652 R/L 0.0301 1000 1 0.0301 Return Neither
652 654 R/L 0.0301 1000 1 0.0301 Return Neither
654 656 R/L 0.0301 1000 1 0.0301 Return Neither
656 658 R/L 0.0301 1000 1 0.0301 Return Neither
658 660 R/L 0.0301 1000 1 0.0301 Return Neither
660 662 R/L 0.0301 1000 1 0.0301 Return Neither
662 664 R/L 0.0301 800 1 0.02408 Return Neither
664 666 R/L 0.0301 0 1 0 Return Neither
670 450 Q R/L 0.0301 1000 1 0.0301 Return Neither From S Face
670 672 R/L 0.0301 1000 1 0.0301 Return Neither
672 674 R/L 0.0301 1000 1 0.0301 Return Neither
674 676 R/L 0.0301 1000 1 0.0301 Return Neither
676 678 R/L 0.0301 1000 1 0.0301 Return Neither
678 680 R/L 0.0301 1000 1 0.0301 Return Neither
680 682 R/L 0.0301 1000 1 0.0301 Return Neither
682 684 R/L 0.0301 800 1 0.02408 Return Neither
684 686 R/L 0.0301 0 1 0 Return Neither
650 750 R 22.5 1 22.5 Default Neither
94
|
Virginia Tech
|
Table B.1 cont. – Year 16 Branch Properties
Resistance Calculated
Resistance per Length Length Parallel Resistance Branch Surface
From To Fqi Type (P.U.) (R/1000ft) (ft) Factor (P.U.) Code State Description
750 550 R 22.5 1 22.5 Default Neither
550 670 R 22.5 1 22.5 Default Neither
652 752 R 22.5 1 22.5 Default Neither
752 552 R 22.5 1 22.5 Default Neither
552 672 R 22.5 1 22.5 Default Neither
654 754 R 22.5 1 22.5 Default Neither
754 554 R 22.5 1 22.5 Default Neither
554 674 R 22.5 1 22.5 Default Neither
656 756 R 22.5 1 22.5 Default Neither
756 556 R 22.5 1 22.5 Default Neither
556 676 R 22.5 1 22.5 Default Neither
658 758 R 22.5 1 22.5 Default Neither
758 558 R 22.5 1 22.5 Default Neither
558 678 R 22.5 1 22.5 Default Neither
660 760 R 22.5 1 22.5 Default Neither
760 560 R 22.5 1 22.5 Default Neither
560 680 R 22.5 1 22.5 Default Neither
662 762 R 22.5 1 22.5 Default Neither
762 562 R 22.5 1 22.5 Default Neither
562 682 R 22.5 1 22.5 Default Neither
664 764 R 35.2 1 35.2 Default Neither
764 564 R 35.2 1 35.2 Default Neither
564 684 R 35.2 1 35.2 Default Neither
95
|
Virginia Tech
|
Figure C.1 is the first page in the Excel workbook. Clicking the ‘Start’ button runs a
Visual Basic for Applications (VBA) macro that opens the form in Figure C.2.
Click the 'Start' button to run the speed calculator
Start
Notes: Solver must be installed in Excel, and marcos enabled.
- To enable macros, select 'Security...' from the Tools-->Macro menu
- To install solver, go to the Tools-->Add-ins menu and check the box next to "Solver Add-in"
- If you receive a "Compile Error: Can't find project or library" message, view the help worksheet tab
below for instructions.
Figure C.1 – Excel Start Page
Figure C.2 – Program’s primary screen
This screen lists the steps necessary to solve reduced speed fan curves, numbered 1 to 4.
Fan curve data and operating point data are required inputs. Clicking these buttons opens
the forms to input this information, Figures C.3 and C.4, respectively. The program
allows for up to six fan curves to be inputted, and up to 20 operating points. Fan curve
data required includes digitized points of both the fan curve and power curve.
97
|
Virginia Tech
|
Figure D.5 – Chart of convergence station #053 channel 1 extension monitoring,
20-inch scale……………………………………………………………………..63
Figure D.6 – Chart of convergence station #053 channel 1 extension monitoring,
concentrated scale………………………………………………………………..63
Figure D.7 – Chart of convergence station #053 channel 3 extension monitoring,
20-inch scale……………………………………………………………………..64
Figure D.8 – Chart of convergence station #053 channel 3 extension monitoring,
concentrated scale………………………………………………………………..64
Figure D.9 – Chart of convergence station #053 channel 5 extension monitoring,
20-inch scale……………………………………………………………………..65
Figure D.10 – Chart of convergence station #053 channel 5 extension monitoring,
concentrated scale………………………………………………………………..65
Figure D.11 – Chart of convergence station #053 channel 7 extension monitoring,
20-inch scale………………..……………………………………………………66
Figure D.12 – Chart of convergence station #053 channel 7 extension monitoring,
concentrated scale………………………………………………………………..66
Figure D.13 – Map of convergence station #040’s extensometer locations……………………..67
Figure D.14 – Chart of convergence station #040 channel 1 extension monitoring,
20-inch scale……………………………………………………………………..68
Figure D.15 – Chart of convergence station #040 channel 1 extension monitoring,
concentrated scale………………………………………………………………..68
Figure D.16 – Chart of convergence station #040 channel 3 extension monitoring,
20-inch scale……………………………………………………………………..69
Figure D.17 – Chart of convergence station #040 channel 3 extension monitoring,
concentrated scale………………………………………………………………..69
Figure D.18 – Chart of convergence station #040 channel 5 extension monitoring,
20-inch scale……………………………………………………………………..70
Figure D.19 – Chart of convergence station #040 channel 5 extension monitoring,
concentrated scale………………………………………………………………..70
Figure D.20 – Chart of convergence station #040 channel 7 extension monitoring,
vii
|
Virginia Tech
|
20-inch scale……………………………………………………………………..71
Figure D.21 – Chart of convergence station #040 channel 7 extension monitoring,
concentrated scale………………………………………………………………..71
Figure D.22 – Map of convergence station #077’s extensometer locations...…………………...72
Figure D.23 – Chart of convergence station #077 channel 1 extension monitoring,
20-inch scale……………………………………………………………………..73
Figure D.24 – Chart of convergence station #077 channel 1 extension monitoring,
concentrated scale………………………………………………………………..73
Figure D.25 – Chart of convergence station #077 channel 3 extension monitoring,
20-inch scale……………………………………………………………………..74
Figure D.26 – Chart of convergence station #077 channel 3 extension monitoring,
concentrated scale………………………………………………………………..74
Figure D.27 – Chart of convergence station #077 channel 5 extension monitoring,
20-inch scale……………………………………………………………………..75
Figure D.28 – Chart of convergence station #077 channel 5 extension monitoring,
concentrated scale………………………………………………………………..75
Figure D.29 – Chart of convergence station #077 channel 7 extension monitoring,
20-inch scale……………………………………………………………………..76
Figure D.30 – Chart of convergence station #077 channel 7 extension monitoring,
concentrated scale……………………………………………………………..…76
Figure D.31 – Map of convergence station #078’s extensometer locations...…………………...77
Figure D.32 – Chart of convergence station #078 channel 1 extension monitoring,
20-inch scale……………………………………………………………………..78
Figure D.33 – Chart of convergence station #078 channel 1 extension monitoring,
concentrated scale………………………………………………………………..78
Figure D.34 – Chart of convergence station #078 channel 3 extension monitoring,
20-inch scale……………………………………………………………………..79
Figure D.35 – Chart of convergence station #078 channel 3 extension monitoring,
concentrated scale………………………………………………………………..79
Figure D.36 – Chart of convergence station #078 channel 5 extension monitoring,
viii
|
Virginia Tech
|
20-inch scale……………………………………………………………………..80
Figure D.37 – Chart of convergence station #078 channel 5 extension monitoring,
concentrated scale………………………………………………………………..80
Figure D.38 – Chart of convergence station #078 channel 7 extension monitoring,
20-inch scale……………………………………………………………………..81
Figure D.39 – Chart of convergence station #078 channel 7 extension monitoring,
concentrated scale (high)………………………………………………………...81
Figure D.40 – Chart of convergence station #078 channel 7 extension monitoring,
concentrated scale (low)…………………………………………………………82
Figure D.41 – Map of convergence station #080’s extensometer locations……………………..83
Figure D.42 – Chart of convergence station #080 channel 1 extension monitoring,
20-inch scale……………………………………………………………………..84
Figure D.43 – Chart of convergence station #080 channel 1 extension monitoring,
concentrated scale………………………………………………………………..84
Figure D.44 – Chart of convergence station #080 channel 3 extension monitoring,
20-inch scale……………………………………………………………………..85
Figure D.45 – Chart of convergence station #080 channel 3 extension monitoring,
concentrated scale…………………………………………………………..……85
Figure D.46 – Chart of convergence station #080 channel 5 extension monitoring,
20-inch scale……………………………………………………………………..86
Figure D.47 – Chart of convergence station #080 channel 5 extension monitoring,
concentrated scale………………………………………………………………..86
Figure D.48 – Chart of convergence station #080 channel 7 extension monitoring,
20-inch scale……………………………………………………………………..87
Figure D.49 – Chart of convergence station #080 channel 7 extension monitoring,
concentrated scale………………………………………………………………..87
Figure D.50 – Map of convergence station #074’s extensometer locations……………………..88
Figure D.51 – Chart of convergence station #074 channel 1 extension monitoring,
20-inch scale……………………………………………………………………..89
Figure D.52 – Chart of convergence station #074 channel 1 extension monitoring,
ix
|
Virginia Tech
|
Chapter 1: Introduction
Convergence in mining is best described as the deformation in an entry or at the face in which
the distance between the roof and floor decreases. In underground coal operations, deformation
almost exclusively refers to convergence. Excessive convergence may be caused by a
combination of heavy pressure from overburden and insufficient support within the excavation.
This causes movement in the surrounding rock, which can lead to a multitude of problems at a
mining operation. Excessive convergence will reduce the headroom in a mine, making it
difficult to move machinery and carry out mining operations [1]. Movement in the rock can also
increase the density of the fracture network, which can lead to a variety of roof fall hazards that
could slow down operations and threaten the safety of the miner.
In order to prevent excessive convergence, a roof control plan is implemented at every
underground mine with the intent of stabilizing the mine environment. The roof control plan
incorporates directives for the development of a support system, from the type of support being
used to the procedures that need to be implemented with every advancement in the excavation.
Plans are designed specifically for the characteristics of the ribs, roof, coal, and overburden in
the immediate area and will vary between mines [2].
Despite the proactive nature of implementing a roof control plan, the arranged maintenance may
prove ineffective for unforeseen reasons. It remains critical that entries and faces be monitored
continuously to observe any changes in the mining environment. Although common practice
among the industry is to use a conventional tape measure to detect deformation, this approach
does not offer the same level of sensitivity or ability to continuously monitor as a properly
implemented extensometer.
The focus of the past few decades has been placed prominently on mine safety. With the passing
of the MINER Act in 2006, great emphasis was been placed on the ability of the operators to
wirelessly locate and communicate with miners underground [3]. Nine years later, the
development of these communication systems has greatly increased the ability to transfer
information immediately to the surface from underground and vice-versa.
Considering the development of other technologies in the industry, it is inexcusable that
instrumentation has fallen behind the curve. Proper implementation of monitoring instruments
1
|
Virginia Tech
|
can reduce the likelihood of disasters through the immediate detection of hazardous conditions.
With access to wireless technologies developed as a result of the MINER Act, there is currently a
need to develop devices that can transfer data to operators working both underground and on the
surface in real time.
The goal of this research is to design and construct a contact extensometer that measures
displacement in the roof. The device will incorporate a potentiometer, a device that transfers
mechanical energy into an electric pulse of proportional strength [4]. The particular type of
transducer we are interested in simulating is the string transducer, which is capable of measuring
a change in linear position. The string potentiometer is typically composed of a rope and a
spring-loaded spool attached to a sensor that will emit a signal as the spool rotates [4].
The philosophy behind string transducer installation is quite simple. The first step requires that
the transducer be mounted in a fixed position with the wire rope attached to a moveable object.
As movement occurs, extension of the wire rope rotates a sensor that is capable of producing an
electrical output signal proportional to the wire rope extension and velocity. The tension in the
rope is controlled by an internal torsion spring that will retract the rope when there is no force
extending the wire rope [5].
The particular advancement in this technology is the development of a borehole extensometer
capable of continuous monitoring of the coal mine roof for extension. The extensometer setup
should be durable, self-supporting in the roof, and capable of continuous data collection and
wireless data communication. The extensometer should be designed for installation in the
MSHA test hole, defined in 30 CFR 75.204 [6].
2
|
Virginia Tech
|
Chapter 2: Literature Review
Rock Conditions in Underground Coal Mines
Underground coal mines are designed to maximize the recovery of a coal seam while ensuring
the safety of miners. The desire to increase recovery led to development of longwall mining,
which incorporates new design challenges to the engineer. As the push to increase recovery
continues, design engineers must also develop systems that manage potential hazards as a means
of preserving the mine environment. Ground instability and the redistribution of stresses due to
mining activity present the greatest challenge to the engineer. A thorough understanding of these
conditions is essential to maximize the potential of mine design.
The causes of ground instability in underground coal mines can be divided into five different
categories. The consequence of understanding these five areas is the information necessary to
devise a proactive and preventative plan to reduce the number of hazards in the mine
environment.
The first of these five categories is the geological factors associated with the mining environment
and its immediate surroundings. In the roof, causes of instability include fracture density, joints,
strata content, local faults, and local lineaments to name a few. In the floor, soft underclay can
contribute to unstable ground and floor heave. In coal seams, rib rolls, washouts, cleats, and
joints are the main causes of ground instability [7].
The second category of issues that could lead to ground instability is the geotechnical properties
of the rock. These properties represent the characteristic nature of the rock, serving as the main
features associated with rock mechanics. Often, the geotechnical properties of rock are
determined in a laboratory setting as opposed to the underground environment. The list of
geotechnical properties includes the unconfined compressive strength of the rock, the shear
strength of the rock, the modulus of elasticity, and the Poisson’s ratio [7].
The third group that contributes to ground instability is the hydrogeologic factors in the mining
environment. Hydrogeologic factors describe both the mine’s access or unavailability to water.
A list of potential hydrogeologic factors include the location and size of local aquifers, thickness
3
|
Virginia Tech
|
of impermeable zones, the hydraulic pressure, porosity, and permeability of the rock, and the
chemical composition of the water [7]. One study of eastern Kentucky coals found that out of
250 roof falls that occurred across five different room-and-pillar coal mines, water was present at
78.0% of the roof fall locations, the most common symptom of roof fall in the investigation [8].
In-situ stresses are the fourth category of causes of ground instability. In-situ stresses are the
pressures associated with any rock location underground. Knowledge of the magnitude and
directions of in-situ stresses is vital to mine design engineering [1]. Regional and local lateral
stresses are the primary factors when determining in-situ stresses. The release of lateral stresses
due to excavation or erosion can also have an impact on ground instability [7].
Mining processes are the fifth and final category associated with causing ground instability. One
example of this category is a mine designed to recover material under bodies of water, which
will indict a variety of stress changes and potential flooding as mining progresses. Another
example is mining that occurs in both over- and under-lying seams, where support systems and
mining activity need to be considered and scheduled according to ground control issues [7].
These factors need to be identified in the design process and accounted for in the long-term
plans.
Convergence
Convergence is the reduction of an entry height due to stress redistributions resulting from
various mining activities [9]. It is best described as the deformation in an entry or at the face in
which the distance between the roof and floor decreases. In underground coal operations,
deformation almost exclusively refers to convergence [1].
Excessive convergence is caused by a combination of heavy pressure from overburden and
insufficient support within the excavation. This can lead to a multitude of problems at a mining
operation. Excessive convergence will reduce the headroom in a mine, making it difficult to
move machinery and carry out mining operations [1].
In order to prevent excessive convergence, it is critical that entries and faces be monitored
continuously to observe any changes in the mine environment. It may seem obvious that this can
4
|
Virginia Tech
|
be achieved with conventional tape measures and rulers, but these devices do not offer the same
level of sensitivity or constant ability to monitor as an extensometer [1].
In longwall mining, the amount of convergence in an area of the mine is directly related to its
proximity to the active longwall panel. The side abutment is the area influenced by the stress on
either side of the longwall panel and can be defined by the following equation:
√ (1)
where W is the width of the side abutment (in feet) on either side of the active panel, and h is the
s
overburden depth (in feet) above the panel [9]. The pressure associated with the tailgate (gob
side) of the longwall panel is typically two-to-three times greater than the pressure associated
with the headgate side. The lowest zone of pressure is dispersed from the middle third of the
longwall panel and is projected towards the longwall face, in an area considered the front
abutment [10].
With regard to location, the greatest amount of convergence on the headgate side occurs right
where the panel rib is located. As the distance from the longwall panel on the headgate side
increases, the amount of convergence decreases. As the longwall face approaches a particular
location, the amount of convergence increases at an accelerating rate [9].
The tailgate side of the longwall panel experiences a different pattern of convergence than the
headgate side. Overall, the tailgate entry has more convergence than the headgate entry.
Convergence in the tailgate actually increases as the distance between the active panel and the
location of interest increases. This difference in convergence distribution can be attributed to
stress dispersion attributed to an area in the mine that is exhibiting its first longwall pass versus
its second longwall pass. In the first instance of a longwall pass, the magnitude of stress in a
particular area is only anticipated to increase by 40%-350%; on the second pass, the stress likely
to increase by 160%-1000% of the original stress [9]. Stress redistribution is the greatest
contributing factor to convergence in a longwall coal mine.
5
|
Virginia Tech
|
Roof Fall
Roof fall can be defined as “the natural or spontaneous fall of rock from the mine roof and ribs”
[1]. Roof falls and rock failures are inhibiting to mining operations, as they lead to downtime in
production and present and an immediate and dangerous threat to employees [8]. Coal mines are
inherently more susceptible to roof fall than other types of underground mines because of the
nature of the coal seam and the mining conditions. This susceptibility has led to identifying the
characteristics of many different types of roof falls that occur in coal mines. Certain geologic
conditions, such as paleochannels, pinch-outs, and slickenslides, have an immense effect on the
stability of a roof in a coal mine [11].
A paleochannel rock failure features weak shale strata in the roof that is unsupported by a newer
sandstone layer. When a paleochannel is mined underneath, the shale loses its primary means of
support and collapses into the mine entry. Paleochannels are exceedingly dangerous because of
their size. Fortunately, paleochannels can often be detected in the exploration process by
inferring the border between sandstone and shale and laying that that boundary over the mine
map. The most difficult paleochannels to identify are the ones that are less than 10 meters in
length. When a paleochannel is identified, the roof can be strapped and bolted to prevent an
imminent roof fall. Any paleochannel collapses that are between 9.3 and 15.5 meters in size are
called “washouts” [12].
A pinch-out occurs in the roof when there is a sudden termination in the roof strata. Pinch-outs
are difficult to predict before roof fall occurs because the location where the strata terminates is
rarely made available through exploration data. When a pinch-out is identified, additional
bolting is required to support the weakened beam [12].
Slickenslides, also known as “slips,” are the most common of all roof failures. Slickenslides
develop when there is movement in the roof strata between shale and claystone layers, forming
curved edges in the roof. These edges tend to be very smooth and develop with a high variability
of dip, with the median angle of dip being 30 . Roof falls in areas where slickenslides can be
controlled with additional bolting and strapping, effectively reducing the area of coverage for
each piece of support and increasing the number of supports to provide better hold [12].
6
|
Virginia Tech
|
Joints are near vertical planar failures that occur not only in the roof, but most locations
underground. In horizontal coal seams, it is unlikely to see failure solely due to jointing in the
roof because the rock continues to receive adequate support. However, if major faulting has
occurred near the coal seam, joint is likely to contribute to local rock failures [B].
The Roof Rating Index (RRI) was developed as a means of evaluating the roof and its
characteristics to increase awareness of areas susceptible to roof fall. With the RRI,
characteristics attributed to causing roof falls (roof content, structures, fractures, etc.) are
compared against the features of the underground location (entry/crosscut, three- or four-way
intersection) to determine the likelihood, size, and type of roof fall faulting that will occur [11].
With this system, mine operators can take extra precaution in areas where roof fall is likely to
occur in a proactive effort to prevent rock failure.
Despite monitoring records of geologic features, it has been ascertained that the major
contributors to roof fall may not be as obvious as anticipated. A study of eastern Kentucky coal
mines investigated 250 roof fall accidents across five different room-and-pillar mines. Most
locations where roof fall occurred had been rated as “good” or better in earlier surveys of the
roof [8].
The investigation took a variety factors into consideration. The results of the study showed that
the presence of water in the roof (78.0%) was a better indicator of potential rock failure than any
other factor, including the evidence of cracks from the mine entry (75.2%). The results also
showed that a vast majority of roof falls took place in locations over 30 meters from the nearest
active mining face (70.4%) and occurred less than 30 weeks after the excavation was made
(71.3%). Over half (60.8%) of the areas that exhibited roof fall were rated as having “good” or
better roof conditions in a prior examination. Oddly, rock failure in the ribs and floor were not
great indicators of rock failure in the roof, as rib sloughing occurred in less than half of the roof
fall locations (48.8%) and floor heaving in approximately one-out-of-eight (12.0%) locations.
This study concluded that geologic factors that cannot be easily measured or evaluated from the
mine entry contribute the most to roof fall in coal mines [8].
7
|
Virginia Tech
|
Ground Control
Ground control refers to study of rock masses that undergo a change in equilibrium [1]. In
mining, this definition can be applied to both underground and surface operations. Removal of
large quantities of material from an area will have and effect on the equilibrium of adjacent rock
masses, permitting the in-place rock to expand in the direction of the free face. It is essential to
understand the mechanics of this interaction in order to safely plan and design mines.
Ground control can also define the means in which a mine chooses to support to its surrounding
environment during advancement. The objective of ground control in mining is to optimize the
support systems in order to insure rock stability and control failure [1]. Prevention of ground
control issues is critical to the health of the mine, as rock failures contribute to downtime and
interrupt normal operations. Rock failures also present a hazard to the employee, as roof falls
and other similar stability issues contribute injuries and, in some cases, fatalities [8].
The purpose of developing a roof control plan is to ensure the preservation of the mining
environment as it was designed. Within the plan are designations for supports and procedures to
be implemented, specifically selected to accommodate the properties of the ribs, roof, coal, and
overburden parameters in the mine [2]. There are three primary objectives that require attention
when assessing the validity of any ground control plan. These three goals are reinforcement,
retaining, and holding, as each goal provides a different mode of stabilization within the rock. It
is important to assess the features of the rock mass before determining which method(s) will be
used [13].
Reinforcement is the ground control method that increases the strength of the rock and reduces
the natural tendency of the rock to lose strength due to fracture. A reinforcing support system
tightens the components of the rock mass in order to minimize the inconsistencies and potential
points of failure [13]. In essence, a reinforcing system will artificially increase the cohesion of
the rock, assuring that the rock mass will remain in one piece, even as the individual rock layers
begin to break.
Retaining is the process of supporting broken rock in the immediate location of the rock failure
[13]. In many situations, this may be required in order to improve the safety of the mine
environment. However, ground control systems are able to take advantage of the broken rock by
8
|
Virginia Tech
|
using it to prevent future failures and provide resistance to rock bursts. At a minimum, a
retaining support system will reduce the severity of a rock burst by providing an immediate
shield of resistance to the energy released composed primarily of previously broken rock.
Holding is the process by which unstable rock is attached or mounted to stable ground. Ideally,
this means that the rock surrounding the mine workings will behave as if it were a singular rock
mass [13]. The stresses that act on the in-situ rock will determine the methodology used
pertaining to the holding process.
Often times, the method of ground control selected for a particular mine is based on two sets of
standards. The first, and more paramount, set of standards taken into consideration is safety
[13]. Often, these standards are set forth by a governing body (MSHA, OSHA, etc.), but the
leaders of the industrial sector may choose to improve upon and go beyond these guidelines to
better ensure the safety of their employees. However, the rule of thumb remains that if the
design is unsafe, it is not worth the investment.
The second set of standards that must be looked into is the cost to operate and maintain the
support system in both active and inactive sections of the mine. The main expenses associated
with ground control are affiliated with materials, personnel, and maintenance. The most
profitable ground control systems are the ones that can minimize the expenditures in each of
these three categories without inhibiting the production plans [13].
In underground coal mines, MSHA sets the standards of practice for roof control plans and
manages the plan approval process. All roof control plans need to be approved by the MSHA
District Manager. The MSHA District Manager is responsible for retaining the geological
information of the area in which he/she is assigned so that appropriate assessments of submitted
roof control plans can be made [2].
MPAS, the Mine Plan Approval System, is the program that monitors the status of roof control
plans at the District Office. When any plans or revisions to plans are sent to out to be approved,
the document is placed in MPAS. Once in place, the office has 45 days to complete its
evaluation of the plan or must provide documented reasoning as to why the plan could not be
complete in 45 days. During the evaluation process, the District Office may request additional
information regarding the roof control plan, in which case the operator must reply or face
9
|
Virginia Tech
|
disapproval of the submitted plan. MPAS tracks all requests, updates, approvals, and evaluations
for each mine using the mine’s ID number [2].
When the evaluation is complete, the mine operator will receive a written document from the
District Manager. In the case that the plan is approved, the document will verify the plan’s
approval and the plan can be implemented. If the plan is not approved, the operator will receive
feedback from the District Manager identifying any problems or inaccuracies in the plan,
possible solutions to any issues, in addition to a timeframe in which the revisions can be
resubmitted [2].
The MSHA handbook Roof Control Plan Approval and Review Procedures identifies the
requirements of the roof control plan for each component that needs to be submitted. The
essential pieces of information for each portion of the document are cited, often referencing the
laws established 30 CFR 75 as a justification for each item listed. The handbook also provides
the operator with information regarding the submittal of mine specific requests, instructions on
using software developed by NIOSH to design mine features, and information on the protection
of miners from rock failures [2].
The roof support section of the handbook establishes that a record of all equipment and the
parameters of each type of equipment needs to be submitted alongside drawings that
demonstrates how the supports will be installed for both underground intersections and mine
entries. Bolting patterns must include specific locations and types of bolts that are being
installed. Tensioned roof bolts are required to be included in the bolting plan and are identified
as a permanent support in the roof control plan. Additional information regarding the use of
supplementary supports is also provided [2].
When using tensioned roof bolts, there is a requirement that test holes must be drilled in the same
vicinity. 30 CFR 75.221 (a) (10) states “[w]hen mechanically anchored tension roof bolts are
used, the roof control plan should include intervals at which test holes will be drilled,” [6].
These test holes should be placed at intervals that take into account the roof strata and the depth
of the test hole [2]. Test holes are required to be “at least 12 inches above the anchorage horizon
of the mechanically tensioned bolts being used,” as stated by 30 CFR 75.204 (f) (2) [6].
Information extracted from the test holes provides the information necessary to evaluate the
10
|
Virginia Tech
|
effect of the tension bolts on the roof strata and identify whether a different type of support is
necessary [2].
Beyond the scope of the aforementioned, it is necessary to collect data reflecting the
characteristics of the mining environment in order to begin constructing ideas on what ground
control methods may be applicable for the mining environment. This data is often relayed to the
engineering staff by way of underground instrumentation: technology designed to reveal the
stress, strain, or load on a particular location. Common practice suggests testing various points
within the mine to gain a grasp of the larger picture and better interpret the data. This
fundamental practice is likely to yield mixed results, requiring that the person analyzing the data
to be an expert on the topic. A good sample of data points will permit proper interpolation of the
data, providing speculative data for the remaining portions of the mine [13].
Mining engineers will often rely on results yielded by underground instrumentation in order to
design appropriate ground control plans for the given conditions. Inaccurate assessment of the
in-situ rock may lead to improper or inadequate support, so the need for proper instrumentation
to identify the surrounding stresses is critical [13]. The consequences of misguided designs can
be disastrous and, in some cases, fatal. The proper evaluation of instrumentation data should be
at the forefront of any design.
Ground Displacement Monitoring Technologies
An extensometer is a device that measures displacement. In underground mining, there are
extensometers that monitor convergence and borehole deformation; the focus of this project is on
the former [14]. There are two general categories in which all extensometers will fall under:
contact and non-contact.
A contact extensometer measures displacement by monitoring the physical changes in the
environment by contacting the extents of the area of interest from a fixed location. The most
common contact extensometers are typically mounted at two locations, identifying any
displacement between the two locations.
One type of contact extensometer used in the mining industry is the tube extensometer. The tube
extensometer provides two applications in the mining industry. The first function of a tube
extensometer is to measure the convergence of roof and floor in a mine as a function of time.
11
|
Virginia Tech
|
This process indicates the stability (or instability) of the development and indicates if any
additional support will be needed. The second application of a tube extensometer is to measure
the separation in roof layers by placing levels at several horizons at on the roof, resulting in
another measure of roof stability [14].
The tube extensometer features a telescoping tube design with a dial indicator and two contact
seats, anchored to the roof and the floor to measure the change distance. The tube extensometer
also features the master bar, which is made of same material as the tube of the extensometer.
The master bar acts as a reference standard and can be used to gauge any error that may exist.
The tube extensometer also has a thermometer to regularly assess any errors that need to be
corrected based on a change in temperature [14].
Another contact extensometer is the tape extensometer. The tape extensometer is a portable
device that uses hand-held measurements. One of the advantages of the tape extensometer is that
it can measure distances in both the horizontal and vertical direction. Measurements are taken
from preexisting base points with precision and speed. The stability of the mine can be assessed
by evaluating the system’s ability to repeat results between two reference points for consecutive
rounds of measurements [14].
The tape extensometer is constructed of steel engineering tape with punched holes and a
compression spring that controls the tension of the tape. The alignments of the anchor points are
not critical because the hook fittings at either end of the device are not affected by the orientation
of the anchor points. The tape extensometer often has a resolution of 0.001 in. (0.025 mm) [14].
Current designs often have digital reading capabilities and can endure harsh environments. The
tape extensometer is easy to use but requires the skill of an experienced employee during
operation [15].
A third type of contact extensometer used is the magnet extensometer. The magnet extensometer
is used to measure heave and settlement in excavations. The magnet extensometer indicates the
depth at which settlement has occurred as well as the total amount of settlement [14]. This
instrument is easy to handle and is very accurate, being able to repeat measurements with an
error of +/- 3 mm (0.1 in.) [15].
12
|
Virginia Tech
|
Implementation of the magnet extensometer is simple. Magnets line the exterior walls of an
access pipe in the roof of an underground excavation. As the ground shifts, the magnets shift as
well. A magnetic probe attached to a graduated cable is then placed in the access pipe to identify
the present location and orientation of each magnet as a means of monitoring any shift that has
occurred between readings. If the magnets do not move over an extended period of time, it can
be assumed the ground is settled. In cases where the heave is expected to exceed 3%,
telescoping sections can be installed to provide more fortitude to the access pipe [15].
Non-contact extensometers often use surveying equipment or lasers to indicate the location of
two or more points from a fixed location. Lasers can be used if placed in a secure, fixed location
with the intent of measuring displacement as it happens, yielding a rate of displacement.
Surveying equipment can be used to regularly monitor the location of multiple points of interest
to determine if any displacement has occurred based on a fixed-point. Surface mounted
extensometers are often used to measure the absolute displacement between various locations by
using surveying equipment. Surveying systems are often viewed as large-scale extensometers
[14]. The surveying equipment in a surface mounted extensometer can be part of an automated
system or can be operated by a technician. Data collected will be used in triangulation analysis
to evaluate the relative locations of each point of interest and detect any displacement that may
have occurred since the previous set of measurements. In the surface setting, most evidence of
extension and slope instability is a result of the development of tension cracks. A regular survey
of preexisting points of interest will can be used to monitor the effects of these fractures over
time [14].
Although this application of surveying equipment is typically used in surface monitoring, it is
not unrealistic to apply this technique to monitor convergence in underground operations.
Similar to the tape extensometer, the surveying equipment can monitor the distance between two
reference points, with one located on both the roof and the floor. Consecutive measurements
will indicate is any displacement has occurred with very good accuracy (often +/- 0.005 feet).
Another type of non-contact extensometer is the laser extensometer that uses laser interferometry
to record information. A laser extensometer operates by examining the surface of the specimen,
generating a virtual image of the specimen. Consecutive tests from the same fixed point will
13
|
Virginia Tech
|
generate several models of the surface over a specific period of time. When these virtual images
of the specimen are arranged chronologically, any deformation that occurs can be detected [16].
When monitoring for convergence, a laser extensometer can be used to analyze the roof and
indicate the current position of all points of interest. Analysis of consecutive virtual models over
a known period of time would provide information to where (if any) deformation is occurring as
well as provide a rate at which deformation occurs. Laser extensometers are particularly good
for very quick testing of high volume areas, making them particularly beneficial for use in
convergence studies [16].
The model piece of equipment that can be associated with analyzing roof conditions via the
MSHA required test hole is the Golder Associates Remote Reading Telltale System. The
Telltale provides information on rock deformation above a mine entry with a multi-color rod that
moves to indicate that deformation has occurred. The Telltale can be hard- wired in succession
to an interrogation unit that has the capability of distributing the information collected by up to
400 Telltales to a computer on the surface. The convenience of the Telltale’s operation has led
to its worldwide use in the mining industry [17].
Wireless Communications Underground
In 2006, the Mine Improvement and New Emergency Response Act (or MINER Act) was signed
into law as an amendment to the Federal Mine Safety and Health Act of 1977. The new law
required that all underground coal mines install wireless two-way communication and tracking
systems by 2011. At the time, no such communication system was readily available or
permissible for operation in an underground coal mine [3]. The MINER Act inspired an
industry-wide rededication to wireless technology implementation in underground mines.
Wireless communication networks have been associated with the mining industry for nearly a
century. There are three different mediums in which wireless communications can be distributed
in an underground mine setting: through-the-earth (TTE), through-the-wire (TTW), and through-
the-air (TTA) [18].
The first of these three mediums that was used was through-the-earth communications.
Beginning in the 1920s, communicators were developed to pick up radio signals underground.
These devices were so well-established that the U.S. Bureau of Mines commercially offered
14
|
Virginia Tech
|
carrier-current radios and TTE signaling by the 1940s. Due to the limitations of TTE,
particularly the retrieval rate of data and the bulkiness of the equipment, it was effectively
terminated in the mining industry by 1950. In recent years, a renewed interest in TTE
technology has increased, almost exclusively since the MINER Act was passed in 2006. The rise
in interest can be attributed to the reliability of TTE systems in the event of a mine disaster
where other wireless systems would fail. Currently, TTE is one of the most commonly used
methods for tracking individuals in underground coal mines [18].
Through-the-wire communication (although the name might suggest otherwise) is a wireless
system that came to fruition in the 1950s. The concept of TTW technology was developed when
it was discovered that conductors, such as electrical wires and metal pipes, could propagate low
frequency transmissions throughout an underground mine. Although this was not completely
understood by scientists at the time, this discovery led to the creation of the leaky feeder systems
used today. TTW systems are considered hybrid or semi-wireless systems since it is partially
wired and partially wireless [18].
Through-the-air communications have been a major component in underground mine
communications since the early 2000s. TTA is capable of providing numerous benefits in
underground mines, including two-way communications, remote controlled equipment and
sensing, and ability to track miners and equipment. WLAN networks, created through the use of
off-the-shelf wireless units, are the most reliable in the underground setting. Using RFID
technology in combination WLAN components has been the most effective means of tracking
equipment underground [18].
There are times when a mine will choose to implement a combination of these three types of
networks to increase the range of the signals. One study, performed in an underground coal mine
in India, used TTE radio technologies to reach TTW leaky feeder networks underground, which
propagated the signal deep into the extents of the mine. Neither system was capable of reaching
that depth alone, so integration provided the means access [19].
Monitoring systems in underground coal are critical to the mine’s performance, as they relay
information regarding the conditions of the mine environment. Traditionally, these systems have
been hard-wired, which is difficult when you consider the extent and size of an underground
mine [20]. In addition, the commitment of mines to wireless communication systems through
15
|
Virginia Tech
|
the requirement of the MINER Act should provide the necessary infrastructure to complete the
task of wireless monitoring systems.
The ideal method of data collection for the underground environment has been designed. The
monitoring system would be permitted to work under one of two settings. The first setting is an
“urgent event setting,” which would place the monitoring equipment in a state of high alert and
produce data in real-time. The second setting described is as “long term periodic monitoring,”
where regularly scheduled data collections occur [20].
The sensors, described as “nodes” in the wireless network, would have three states of operation:
sleep, update, and awake. In the sleep mode, sensors run on low power and do not collect data in
an effort to save energy in the unit. The update mode is activated during periodic monitoring
sessions, permitting the recovery of data in a short interval, before returning to the sleep mode.
The awake mode is activated when the urgent event setting is initiated, and makes the sensor
begins to take readings in real-time, presenting the most current available information to the user
[20].
16
|
Virginia Tech
|
Chapter 3: Development of a Wireless Borehole
Extensometer for Monitoring Convergence in
Underground Mines
Design Considerations
The intention of this exercise is to create an extensometer that will be able to detect convergence
in a coal mine by measuring the extension in the roof. In order to develop the design, it is
important to be able to quantify the aspects of the environment in which the device will be
installed.
The first component that needs to be quantified is the MSHA-required test holes that are
introduced in the roof. These holes have a one-inch diameter and must extend at least 12 inches
beyond the permanent support. In most cases, the permanent support being referred to are
torque-tension bolts, which are typically six feet long. This establishes that the tests holes are
required to be seven feet in depth by law. Mines can also choose to evaluate secondary supports
(cable bolts) via the test holes. In cases where this occurs, the same 12-inch addition will apply
to the depth of the hole. Assuming the average cable bolt is 10 feet in length, it can be
anticipated that average depth of test holes in these mines are 11 feet.
Asserting that the average depth of a test hole will be 7-11 feet provides insight on what kind of
measurement tool can be used. Reasonably, it would be difficult to imagine more that 10% of
the measurable roof strata to experience extension before breaking with respect to the original
anchorage depth. Although this is a severely rough estimate, it provides some insight to the
anticipated amount of movement that the device needs to be capable of identifying. On the short
end, the maximum amount of extension that should be able to be measured is 8.4 inches; on the
long end, maximum anticipated extension 13.2 inches. These figures will be identified as the
approximate range of extensometer being developed. In addition, an anchor capable of
withstanding movement within the roof strata and a method of installation that sets the anchor at
the top of the test hole needs to be designed for the extensometer.
The mine environment must also be considered when developing the extensometer.
Underground mines tend to be one of the most hazardous environments on the planet, with large
17
|
Virginia Tech
|
equipment, personnel, and raw materials constantly moving in constrained areas. In addition, the
ground structures that surround the mine are continuously experience changes in equilibrium as
mining activity advances, causing minor shifts in the ground due to the change in stress.
When considering the mine environment in the design process, a few things become evident.
The most exterior portion of the extensometer needs to be exceedingly durable such that it can
protect the measurement equipment from the everyday hazards of the mine environment.
Secondly, the extensometer apparatus should be relatively small to avoid interference with
regular mining operations. Another trait the extensometer should have is the capability of
supporting itself in the roof.
Data collection and wireless transmission are two design components that are closely related.
Data should be collected continuously and at a predetermined rate. This feature is desired
because it would provide a timeframe of geotechnical events monitored by the extensometer,
allowing mine operators to attribute the deformation to a particular event (blasting on the
surface, nearby longwall activity, etc.) or is the result of natural equilibrium restoration, which
would provide the opportunity to evaluate the roof independent of outside activity. It is also
desired that the equipment used to collect data has the ability to transmit the information
wirelessly. This would reduce the amount of electrical components added to the device and
ensure that there is nothing lost in the transmission from data collection to wireless
communication.
The cost of producing the extensometer is also critical to the design process. Ideally, the
extensometer would be inexpensive to produce on economic principles alone. However, given
the area of the mine roof the extensometer is capable of evaluating, the ideal design would have a
low cost of production so that more extensometers can be inserted in relatively close proximity
to give a more comprehensive evaluation of the roof. The ability to produce an inexpensive
device is essential to creating the best representation of the convergence in the mine.
For summary, the objectives of the design can be broken down into five components:
1) The extensometer must give a proper evaluation of roof strata extension for
convergence measurements.
2) The extensometer must retain the ability to continuous monitor the immediate
environment for convergence.
18
|
Virginia Tech
|
3) The extensometer must be able to transmit information wirelessly throughout the
mine.
4) The extensometer should have a casing and anchorage system suitable for the mine
environment and capable of supporting itself to the roof.
5) The extensometer must be inexpensive to fabricate and install.
Extensometer Development
The tasks involved in the development of the extensometer were numerous and consisted of
many failed trials before a final design was established. In this section, a review of the device
components and the information that led to final design will be reviewed.
Measuring Device: Potentiometer
To create an electrical signal capable of wireless transmission, a measurement tool must be used
that can create an electrical signal when triggered by the data-logging component. To meet this
requirement, a potentiometer has been incorporated into the extensometer design. A
potentiometer is a device that transfers mechanical energy into an electric pulse of proportional
strength [4]. The potentiometer that has been designated for this design is the Bourns 3590S-2-
102L Wirewound Potentiometer, as seen in Figure 3.1 below. The potentiometer chosen has 10-
turn rotating arm and a 1,000 ohm (1 k ) resistance. This potentiometer was chosen because it
features the same resistance and number of turns as the JX-PA extensometer discussed in
Chapter 3.
Figure 3.1 - Bourns 3590S-2-102L Wirewound Potentiometer
For this extensometer, the goal is to create a motion that is normally attributed to string
transducers. String transducers are typically composed of a rope and a spring-loaded spool
19
|
Virginia Tech
|
attached to a sensor that will change resistance as the spool rotates, which causes a change in the
signal emitted [4]. In order to induce tension in the extensometer and allow the potentiometer to
recoil, a constant force spring will be attached to the arm. The constant force spring is wrapped
around the arm of the potentiometer via a set of rubber grommets, ensuring the spring will not
slip as it is pulled. On the outer edge of the constant force spring is a hole that can be attached to
a string that will provide a connection to the anchor at the top of the test hole. An example of a
constant force spring can be seen in Figure 3.2.
Figure 3.2 - Constant force spring
Four different styles of constant force springs were chosen to be analyzed based on their inner
diameter, extended length, and price. After careful inspection, two of the springs, the Gardner
GCF-04-15 and the Gardner GCF-04-20, were deemed acceptable for use in the extensometer
design. The Gardner 04-10 was considered unreliable as it would twist and turn before recoiling.
The Gardner 04-25 was determined to be too wide and too difficult to manipulate by hand to
give any further consideration in design process.
Table 3.1 displays the characteristics of both the Gardner 04-15 and Gardner 04-20 for
comparison [22]. From the initial inspection, it is evident that the Gardner 04-15 spring is the
more sensitive of the two springs, meaning that it could provide more accurate data than a larger
spring such as the Gardner 4-20. On the other hand, the Gardner 4-20 is longer, thicker, and
wider than its counterpart, indicating it would be more durable and dependable addition to the
design that would also increase the range of the extensometer.
20
|
Virginia Tech
|
Figure 3.3 - 3-D printed extensometer base
The lid portion of the extensometer casing went through three generation of design. The crucial
element that would determine the success of design was the mechanism designed to hold the
extensometer in place. The original design features two thick-walled semi-cylinders, each with a
radius of 0.5”, with a path for a bolt to securely enter and push the two semi-cylinders apart and
create a wedge. Between the two semi-cylinders is a slot that went through the bottom of the lid
that would permit the constant force spring to emerge and extend up and into the test hole. In
each of the four corners is a hole designed to permit an attachment screw to go through to attach
to the base of the extensometer casing. Unfortunately, the thick-walled semi-cylinders proved
too stiff to be moved and would not grip the inside the borehole.
The second generation of the lid then featured two thin-walled semi-cylinders, each with an outer
diameter of 1.15” and an inner diameter of 1”, erected 1” above the surface. This design was
chosen because of the limited, but available, flexibility of the plastic in mind. This design
permitted the plastic to bend inwards to be inserted into the test hole. When released, the semi-
cylinders would try to retain their original position, providing a means of support the entire
extensometer casing. The drawback to this design was when the constant force spring was
introduced. As the spring emerged through the slot that was placed between the thin-walled
cylinders, it would begin to twist and go off-center, creating a movement susceptible to error
22
|
Virginia Tech
|
Figure 3.6 - Torsion spring as a borehole anchor
Data-Logging and Wireless Communication Equipment: MIDAS
For data logging and wireless communications, an outside instrument was used to complete the
project. The Golder Associates Miniature Data Acquisition System (MIDAS) both collects data
and transmits it wireless, thus fulfilling the remaining requirements of the design objectives.
The MIDAS was developed by NIOSH as a way to collect data in underground coal mines and is
capable of using wired or wireless data transmission. Serving as a self-powered, intrinsically
safe (electrically self-contained) data-collector, the MIDAS datalogger, as seen in Figure 3.7,
was designed with the intention of updating older equipment being used in U.S. coal mines. The
datalogger is capable of monitoring up to eight instruments at a single time. Wireless data
collection is feasible through the MIDAS user interface, as seen in Figure 3.8, and stored in a
MicroSD card, located in the back panel of the user interface [21].
For this study, the borehole extensometer was hard-wired into the MIDAS datalogger. Data was
logged at varying rates that were highly dependent on the experiment being performed. Data
was collected via the MIDAS user interface and transferred to a computer spreadsheet through
the MicroSD card.
25
|
Virginia Tech
|
soldering the wires to the proper terminal on the potentiometer. Solder the wires outside of the
extensometer casing such that you do not melt the plastic of the casing.
The next step is to attach the spring to the potentiometer. Depending on the spring chosen will
determine this setup. For the GCF-04-15, the connection to the rotating arm of the potentiometer
uses two grommets that have an inner diameter of 0.25” and an outer diameter of 0.5625”. This
set of grommets attached firmly to the potentiometer arm and the GCF-04-15 wraps tightly
around the outside of the grommets, as seen in Figure 3.9, so there is no need for additional
adhesive.
Figure 3.9 - Grommets around the potentiometer arm
To set up an extensometer with the GCF-04-20, an extra step must be taken. The grommets that
can be used for this constant force spring are larger, with an inner diameter of 0.375” and an
outer diameter of 0.625”. Since the inner diameter of these grommets is too large to
independently wrap around the arm of the potentiometer, electric tape is wrapped around the arm
of the potentiometer until it is suitable for the grommet to be wrapped tightly around it.
When the grommets are secure in either case, test the setup by placing the constant force spring
around the grommets so it will unwind in a clockwise direction. Turn the potentiometer arm as
far counterclockwise as possible. Look to ensure that the grommets do not slip and that the
potentiometer arm turns at a consistent rate as the spring is being pulled. If there is slippage,
additional tape may become necessary to connect the constant force spring and potentiometer. If
the setup can be verified as competent, the potentiometer may be placed in the designated area of
27
|
Virginia Tech
|
Figure 3.11 - Completed wireless borehole extensometer
Installation and Data Collection
In order to install the extensometer, some required and auxiliary equipment will be necessary.
The required equipment includes 0.5” PVC piping, cut to an appropriate length, and electrical
wiring. PVC was selected because of its durability and relatively light weight. The length of the
PVC piping will be dependent on the test hole depth. The reason 0.5” PVC pipe is used is to
ensure that the pipe will not get lodged into test holes if there is bending in the test hole. The
PVC may be cut into more than one piece to ease the burden of moving this equipment around
the mine and reconnected as it is inserted into the borehole. The electrical wiring will be used to
connect the potentiometer inside of the extensometer to the MIDAS datalogger.
In the event that the test hole in the roof is beyond reach, a ladder will be necessary for the
installation process. In addition to the extensometer, it is recommended that the MIDAS
datalogger is also attached to the roof in an effort to remove the datalogger from moving
equipment and other types of intervention. The MIDAS can be effectively attached to the roof
mesh via rope or hooks. Wire mesh cutters may be appropriate if roof mesh is present to cut a
29
|
Virginia Tech
|
hole big enough to install the extensometer as far into the test hole as possible. A hammer can be
used to hit the bottom of the PVC pipe when the anchor is being installed.
The installation procedure begins outside of the mine where the extensometers are connected to
the MIDAS datalogger. Up to four extensometers can be connected to the MIDAS datalogger
simultaneously. In order to attach the two components, an electrical wire must be soldered at
each end to the following corresponding parts: ground (1, black), output (2, red), and input (3,
white). When the electrical wiring is set up for all devices, reattach the lid of the MIDAS
datalogger to ensure that the device retains its permissible standards. When the lid is replaced, it
is important to activate the datalogger using the MIDAS user interface. At this time, the user can
name the datalogger, set the time on the datalogger, set the rate of data collection, and turn data
collection on, ensuring that the device is operating correctly by checking to see if a green light is
blinking. The string should be cut to the anticipated test hole length and attached to the constant
force spring at one end and torsion spring anchor at the other. The wiring for each set of
dataloggers and extensometers should be neatly wound and bound before travel underground to
ensure no intertwining of wire occurs.
Once underground, the first part of the installation process is to examine test holes in the area
with the PVC pipe. In the best-case scenario, these test holes would be examined shortly after
the test hole is drilled to ensure that little-to-no manipulation has occurred. A desired test hole
will retain the same depth as indicated in the mines roof control plan for anchor installation. If
there is any mesh wiring present, it may be necessary to cut a hole in the mesh to provide enough
access for the extensometer during installation.
If the test hole is found suitable, the test hole installation can begin. Placing the torsion spring
coil-side up into the test hole, use the PVC pipe to push the spring up to the top of the test hole.
As the constant force spring emerges from the extensometer, hold the extensometer against the
roof; this will ensure that the constant force spring will coil back inside extensometer during
installation. When the anchor has reach the top of the test hole, slowly insert the extensometer
into the hole, paying attention to the constant force spring reentering the device. When the two
half-cylinders reach the roof, squeeze the top portion of each half-cylinder together until it can be
inserted. Put pressure on the bottom of the extensometer until it has been properly secured in the
roof. Check the immediate area to ensure there are no threats to the extensometer or the
30
|
Virginia Tech
|
electrical wiring. When all applicable extensometers are installed, secure the MIDAS datalogger
to a safe location in the immediate area, preferably to roof or rib mesh.
To collect data, the miner must have the MIDAS user interface and be in line-of-sight and within
200 feet of the MIDAS datalogger of interest. When using the interface, the data collection
process is activated from the user interface and stored on a MicroSD flash memory card that is
installed within the user interface. Data collected will reflect all the information that has been
stored since the datalogger was activated or since the last time the datalogger had been reset. If
data has been collected, the datalogger will be placed in a “paused” mode and the blinking light
will turn red. If the user would like to continue collecting data, it is necessary to place the
datalogger back into “run” mode. Information stored on the MicroSD card can be easily
uploaded to a user-friendly spreadsheet developed specifically for the MIDAS.
Calibration
The purpose of calibrating the extensometer is to understand the rate at which each respective
spring will change the output of the potentiometer. When dealing with raw data, it is important
to understand this rate of change since the null reading at the beginning of tests can vary based
on how the extensometer is set up. Both setups will undergo the same controlled evaluation, and
each set of data produced will be given a quadratic equation of best fit using least-squares
regression. The quadratic equation was picked to represent the line of best fit because the ever-
changing outer diameter of the constant force spring ensures that the rate of change in the
potentiometer outputs will not be linear. The y-intercept of the equation of best fit for each
extensometer setting will be set to zero to represent that no change in displacement has occurred.
To calibrate the extensometer, an experiment was run to simulate a test hole and the possible
extension that may occur in a mine setting. To perform this experiment, a replica test hole had to
be constructed. For the length of the original test hole, two pieces of 5’ long, 1.25” PVC pipe
were employed. This size pipe was selected because it features a 1” inner diameter, the same
diameter as the MSHA-required test holes and did not require any change to the original
extensometer design.
31
|
Virginia Tech
|
To simulate extension in the roof, a portion of 1.25” PVC pipe was cut into six segments, each
segment cut to approximately 1” in length. Each segment had a slit cut into the side to provide a
means of permitting the string attached to the anchor to slide through. A picture of the six
segments is available in Figure 3.12.
Figure 3.12 - Six 1” segments made from 1.25” PVC pipe
In order to run the experiment, both 5’ segments of the PVC pipe were lined up end-to-end on
the floor. Two extensometers, one with each type of spring setup, were installed to a MIDAS
datalogger and with anchors attached. One at a time, an extensometer was installed, with the
anchor crossing the intersection between the two pipes and continuing into the second pipe. The
data logger was turned on to collect data once every five seconds.
After about 30 seconds, the pipes were pulled apart and a 1” segment was installed in-between
the two long pieces of PVC. After another 30 seconds, the pipes were pulled apart and another
segment was added. This process was repeated until all the segments were used. When the first
extensometer had completed the experiment, the string was cut and both the extensometer and
the anchor were removed from the PVC pipes.
This same process was repeated with the second extensometer. In both cases, the segments were
used in the exact same order and orientation to ensure that both extensometer setups experienced
that same simulated extension. Every precaution was made to ensure that the experimental
apparatus remained identical.
32
|
Virginia Tech
|
installation where the PVC is removed and the extensometer replaces it in the test hole was more
difficult with the larger diameter PVC pipe. The issues experienced with the 0.75” diameter
PVC pipe led to the use of 0.50” diameter PVC pipe in future installations.
The last item that needs to be mentioned was the use of the GCF-04-10 constant force spring in
the extensometer design. Originally, the GCF-04-10 was the most sensitive spring considered
for the extensometer and was favored over other springs because it was the most likely to detect
a small change in the ground environment. However, when introduced to the mining
environment and the envisioned installation process, the GCF-04-10 setup did not work. In
addition to twisting uncontrollably as the anchor was being installed, it was not strong enough to
maintain its grip around the potentiometer arm, ejecting from the extensometer to a fair distance
up the test hole. The complications involved in the installation of the extensometer with the
GCF-04-10 spring eliminated it from future design consideration.
The second attempt to install equipment at the aforementioned mine took place on November
20th, 2014. At this time, two borehole extensometers were successfully installed in the #2
headgate and the 10th crosscut. At the time of the installation, the longwall was currently
running at a location between the 15th and 14th crosscut.
The intention of installing equipment at this location was to gauge the amount of deformation
that could be attributed to the side abutment stresses. Permission was not granted to install
equipment in the entry immediately adjacent to the longwall panel, where the highest amount of
deformation would typically occur on the headgate side of the active panel. In addition, a
stopping was located halfway through the 10th crosscut at the time of installation, providing a
limited number of locations to install the extensometers.
One of the extensometers installed was set up with a GCF-04-15 spring. This extensometer was
connected to the MIDAS datalogger in channel 5. This extensometer was installed
approximately halfway between the headgate entry and the track entry in the 10th crosscut. The
anchor used was a 3-D printed stopper. The only issue was that the initial failure of two 3-D
printed anchors. The arms broke as they had in the previous installation attempt. The third
anchor was installed with no issue.
37
|
Virginia Tech
|
Field Test - Roof Extension Readings
3.5
3
)
.n
i(2.5
g
n Channel 1 -
id
a e 2 GCF-04-20
R
n Channel 5 -
o
is1.5 GCF-04-15
n
e
t
x
E
f 1
o
o
R
0.5
0
Date
Figure 3.20 - Roof extension readings of field test installation
In order to properly evaluate the data set, it is important to identify the outlier. Figure 3.21 gives
a clear insight as to the data obtained between December 7th, 2014 at 12:00PM and December
9th, 2014 at 12:00AM. After careful consideration, it has been determined that the spike in the
data set was due to an electrical malfunction in the MIDAS datalogger. Although it was
speculated, the raw data figure between the smaller magnitude spike and the larger magnitude
spike is not exactly double (3852 to a range of 7821-8200), but it is close. The reason for this
error is unclear, but has occurred in previous studies using the MIDAS datalogger.
The outliers were removed from the data set. Updated roof extension charts are presented with
regards to raw data set and the actual length of extension are present in Figures 3.22 and 3.23.
42
|
Virginia Tech
|
segment addition. Another potential source of error in the calibration test is the true length of
each of the 1” segments. Although they were cut to the 1” specification, the edges were not truly
straight and may have attributed to a small deviation in the readings.
Another potential source of error is the constant force spring. Although the spring moves fairly
linearly for the first few inches it is extended outside of the extensometer, it will begin to curl is
the extensometer is experiencing excessing extension. This would rarely be a contributor to
extreme error, but could affect the overall accuracy of readings at great lengths. Fortunately, the
borehole acts a secondary slot extension, forcing the constant force spring to remain primarily
straight in the test hole.
Discussion of Results and Conclusions
The results for the Channel 1 – GCF-04-20 extension data indicates that little to no extension of
the roof occurred between the anchor point at the top of the test hole and the extensometer. The
range of the raw data corresponds to a displacement of less than 0.003”, a negligible amount. If
no extension or movement occurred in the roof at this location, the extensometer data is correct.
The results of the Channel 5 – GCF-04-15 extension data indicate similar results to that of the
Channel 1 – GCF-04-20. A low range of extension in the roof only correlated to a displacement
reading of about 0.003”.
This low range of displacement is likely due to the placement of the extensometers. Had the
extensometers been located nearer to the longwall panel on the headgate side or experienced a
second longwall-pass with the adjacent panel, there would be a greater chance that significant
displacement would have occurred. However, the limited availability of suitable test holes and
the permissions the mine allowed for testing prevented this from happening.
No information regarding the condition of the site where the installation took place has been
provided since installation that could assert a conclusion as to if any equipment error occurred.
To conclude, a wireless borehole extensometer was developed and implemented in a lab setting
as well as in an underground coal mine. The extensometer was designed to mimic the traits of a
string transducer and was implemented with a wireless data component. The exterior casing of
the extensometer is adequate for the mine environment and acts as a self-supporting system that
45
|
Virginia Tech
|
allows the extensometer to hold itself onto the roof at the location of a MSHA required test hole.
The 3-D printer permitted several iterations of the extensometer casing to be produced in a
timely fashion, where it could then be evaluated for its feasibility regarding the design
objectives. A cheap, effective anchor was discovered to replace a less competent alternative.
The anchor, when in place at the top of the test hole, acted as a second means of support for the
extensometer to hold it in place against the roof.
The results of the lab study prove that the borehole extensometer that was developed can
adequately measure extension in a borehole environment. When extension in the roof occurs, the
extensometer is capable of transmitting that information through a potentiometer and to a data-
logging device that relays information regarding the magnitude of displacement.
In the field, the extensometer was successfully installed, but returned data that could not verify
whether the device was recording data properly. Further field testing is necessary to ensure the
device is appropriate for the roof extension monitoring.
46
|
Virginia Tech
|
Chapter 4: Future Work
If work is to be continued in this area of research, more underground installations need to occur
in areas where convergence occurs at a premium. Installing the extensometer at these locations
alongside a redundant and commercially available extensometer will increase the likelihood that
measurements will be taken and can be verified with other equipment. The great uncertainty in
this study is the actual impact the extensometer can have in the mine environment. Components
of the extensometer design can only be improved on once they are met with failure, so having
more opportunities to install would inevitably develop the extensometer further.
Another point of emphasis that needs to be improved is the wireless communication system.
Currently, information can only be recorded using the MIDAS user interface when in line of
sight of the datalogger. The objective of future work should be to provide the extensometer the
ability to communicate data in real-time to the surface. Improving this wireless feature in the
extensometer would place this equipment on the cutting edge of mining technology. The likely
course of action to pursue this goal is to install a WLAN wireless device and set it up to the
extensometer; readings will be taken and can be communicated to a leaky feeder system, and
provide information to the surface.
A design feature that can be improved is the feature of tension for the potentiometer. While the
constant force spring performs adequately, the goal for this objective would be to improve upon
the ease of use in the system. A spring-loaded potentiometer could serve as the source of
tension, although a manufactured one is typically ten times the cost of a potentiometer that is not
spring-loaded.
If the constant force spring is chosen to continue in future iterations of the design, the grommets
should be replaced with a part that is specifically fabricated to connect the spring and the
potentiometer arm. This will reduce the likelihood of error due to the component parts of the
device and ensure no slippage occurs between the two parts.
The permissibility of the device also needs to be evaluated. The parts of the device should be
secure to one another and permit no airway with access to the outside environment. This may
47
|
Virginia Tech
|
22785 4521 4521 29192 6274
22784 4520 4520 29257 6339
25701 7437 7437 29320 6402
26252 7988 7988 29383 6465
26184 7920 7920 29467 6549
26183 7919 7919 29474 6556
26112 7848 7848 29534 6616
29493 11229 11229 29594 6676
29286 11022 11022 29613 6695
29286 11022 11022 29644 6726
29417 11153 11153 29715 6797
32273 14009 14009 29732 6814
32274 14010 14010 29703 6785
32273 14009 14009 29734 6816
32273 14009 14009 29734 6816
35852 17588 17588 29771 6853
35852 17588 17588 29789 6871
35853 17589 17589 29693 6775
35853 17589 17589 29778 6860
20801 2537 29741 6823
18484 220 29729 6811
18199 -65 29930 7012
18200 -64 23674 756
18200 -64 22923 5
18200 -64 22919 1
18200 -64 22918 0 0
18200 -64 22918 0 0
18201 -63 22918 0 0
18201 -63 22918 0 0
18201 -63 22918 0 0
18201 -63 26230 3312 3312
18201 -63 26194 3276 3276
18202 -62 26177 3259 3259
18201 -63 26177 3259 3259
18201 -63 26178 3260 3260
18201 -63 26177 3259 3259
18201 -63 26126 3208 3208
18202 -62 26126 3208 3208
18201 -63 29913 6995 6995
18201 -63 29870 6952 6952
18201 -63 29284 6366 6366
18201 -63 29624 6706 6706
18201 -63 29607 6689 6689
56
|
Virginia Tech
|
For data logging and wireless communications, an outside instrument was used to complete the
project. The Golder Associates Miniature Data Acquisition System (MIDAS) both collects data
and distributes it wireless, thus fulfilling the remaining requirements of the design objectives.
The MIDAS was developed by NIOSH as a way to collect data in underground coal mines and is
capable of using wired or wireless data transmission. Serving as a self-powered, intrinsically
safe data-collector, the MIDAS datalogger, as seen in Figure D.2, was designed with the
intention of updating older equipment being used in U.S. coal mines. The datalogger is capable
of monitoring up to eight instruments at a single time [21]. Wireless data collection is feasible
through the MIDAS user interface, as seen in Figure D.3, and stored in a MicroSD card, located
in the back panel of the user interface.
For this study, the borehole extensometer was hard-wired into the MIDAS datalogger. Data was
logged at varying rates that were highly dependent on the experiment being performed. Data
was collected via the MIDAS user interface and transferred to a computer spreadsheet through
the MicroSD card.
Figure D.2: MIDAS datalogger
59
|
Virginia Tech
|
Figure D.3: MIDAS user interface
Installation Procedure
The first portion of the installation is connecting the JX-PA to the MIDAS datalogger. Up to
four extensometers can be attached to each MIDAS datalogger. It is essential to provide an
adequate amount of additional wire to each extensometer to ensure the extensometers can reach
the desired underground support locations.
The next step is to attach the extensometer to the standing support. Ideally, the extensometer
will be placed at a location very close to the roof or the floor in a space in which it is unlikely to
be tampered. For timber supports, the easiest method of attaching the extensometer is to take a
screw, place it in one of the anchorage holes of the extensometer, and screw it in to the wood.
For cement-filled supports, the support will need to have a hole punched in the side before the
screw can be inserted; it is recommended that a handheld pick-ax or a designed hole-puncher be
used to complete this task. A second screw can be attached in the opposite screw hole of the
extensometer if possible to add extra anchorage (this may not be feasible in cement-filled
supports).
On the opposite end of the support, a screw is installed directly above or below the extensometer,
depending on which location the extensometer was installed. Attach a string that is
approximately the length equal to the distance between screw and the extensometer to the hook
60
|
Virginia Tech
|
of the extensometer. Expelling the cord from the extensometer, wrap the string around the
screw, such that the extensometer cord is fully extended to its 20-inch maximum length. Knot
the string around the screw, ensuring that the string is secure while maintaining the 20-inch
extension.
Repeat the previous steps with all of the extensometers. When all are installed, ensure that the
electrical wires and the MIDAS datalogger secure and out of harm’s way and away from areas of
heavy traffic. Interference with the electrical wires or the MIDAS datalogger may cause
inaccurate results or a loss of transmission from the extensometer to the datalogger.
Installation and Data Collection
The mine where the installation of convergence stations occured is in northern West Virginia. It
is mine a high-grade bituminous coal in the Lower Kittanning seam. The average cutting height
in the mine is 96” and the average depth of cover is 600 feet. The roof in the mine is considered
very good in most areas. To date, the mine has completed two longwall panels and is currently
mining a third panel.
Several convergence stations were installed and monitored throughout the course of one year.
The location and feedback of each monitoring station has been compiled in this section. The
convergence stations can be identified by the number associated with its respective datalogger.
The graphs display the length of string extension of each extensometer attached to that specific
extensometer. Each extensometer data set will produce two graphs: one on a 0-20 inch scale,
and another concentrated on the range of the data representing any deformation that may have
occurred. Any outliers due to transportation or faulty signal transmission will be available on the
first chart and explained if necessary. In theory, as the extension of the extensometer decreases,
the amount of convergence in the entry is increasing as the artificial support deforms.
The first convergence station installed incorporated the #053 MIDAS datalogger. This
convergence station began recording information on November 19th, 2013 and was mounted in
roof supports that were located in the bleeders at the beginning of the first longwall panel. A
map of location of the convergence station can be found in Figure 3.4. The extensometers
connected into channels 1 and 3 were installed in cement-filled supports. The extensometers
connected into channels 5 and 7 were attached to Link-N-Lock timber supports. A data point
61
|
Virginia Tech
|
Logger 77 - Channel 7
20
)
.n 18
i( 16
n
o 14
is
n 12
e
t 10
x
E 8
r
e 6
t
e m 4
o 2
s
n 0
e
t
x
E
Date
Figure D.29 – Chart of convergence station #077 channel 7 extension monitoring, 20-inch scale
Logger 77 - Channel 7
20
)
.n
i( 19.9
n
o
is 19.8
n
e
t 19.7
x
E
r e 19.6
t
e
m 19.5
o
s
n 19.4
e
t
x
E
Date
Figure D.30 – Chart of convergence station #077 channel 7 extension monitoring, concentrated scale
Convergence station #078 was installed and began collecting data on April 11th, 2014. This
convergence station was mounted in roof supports that were located in the headgate of the first
longwall panel in and around crosscut 17. A map of location of the convergence station can be
found in Figure D.31. All the extensometers were connected into cement-filled supports. A data
point was collected at this site once every 12 hours. The most recent date of data collection was
October 30th, 2014. Charts regarding the extension of the #078 extensometers can be found in
Figures D.32 through D.40. The significant portion of channel 7 is shown over two graphs.
76
|
Virginia Tech
|
abandon this type of support for financial and maintenance-related reasons when it became
apparent that the overall strength of the columns was unnecessary.
Installation of the convergence stations also had a wide range of parameters that favored use of
timbers over cement-filled supports for convergence monitoring. In the timbers, screws could
easily penetrate the wood with no additional effort, simplifying the task and permitting the
installation of a second anchor screw for support in many cases. In the cement-filled supports,
punching a hole in the exterior shell required a lot more effort and was often met with a great
deal of trial-and-error until a suitable penetration point could be established. This often altered
the final placement location of the extensometer, making the process less ideal. In cases where
the cement-filled supports had wooden boards placed between the column and the roof, the
extensometer was connected to the boards rather than the cement column.
The MIDAS datalogger would sometimes misrepresent the amount of deformation in the
standing support. This can be attributed to an electrical error within the device. These outliers
happened in multiple data loggers and across multiple channels. Although the error would often
correct itself, these electrical communication errors contribute to holes within the data set.
Overall, the recommendation for the use of the convergence station is to install in areas where
some form of timber supports is used. The timbers flex and deform more easily alongside the
mine environment, providing a good representation of convergence in the area.
93
|
Virginia Tech
|
A detailed justification for the selection of a novel mine tracer gas and
development of protocols for GC-ECD analysis of SPME sampling in static
and turbulent conditions for assessment of underground mine ventilation
systems
Susanne Whitney Underwood
Abstract
Tracer gas surveys are a powerful means of assessing air quantity in underground mine
ventilation circuits. The execution of a tracer gas style ventilation survey allows for the direct
measurement of air quantity in locations where this information is otherwise unattainable. Such
instances include inaccessible regions of the mine or locations of irregular flow. However, this
method of completing a mine ventilation survey is an underused tool in the industry. This is
largely due to the amount of training required to analyze survey results. As well, the survey is
relatively slow because of the time required to perform analysis of results and the time required
to allow for the total elution of tracer compounds from the ventilation circuit before subsequent
tracer releases can be made. These limitations can be mitigated with the development of a
protocol for a novel tracer gas which can be readily implemented with existing technology.
Enhanced tracer gas techniques will significantly improve the flexibility of ventilation surveys.
The most powerful means to improve tracer gas techniques applied to mine ventilation surveys is
to alter existing protocols into a method that can be readily applied where tracer surveys already
take place.
One effective method of enhancing existing tracer gas survey protocols is to simply add a
second tracer gas that can be detected on a gas chromatograph – electron capture detector (GC-
ECD) using the same method as with the existing industry standard tracer, sulfur hexafluoride
(SF ). Novel tracer gases that have been successfully implemented in the past called for complex
6
analysis methods requiring special equipment, or were designed for inactive workings.
ii
|
Virginia Tech
|
Experimentation with perfluoromethylcyclohexane (PMCH) and SF allowed for ideal
6
chromatographic results. PMCH is a favorable selection for a novel tracer to work in tandem
with SF due to its chemical stability, similar physical properties and detection limits to SF , and
6 6
its ability to be applied and integrated into an existing system. Additionally, PMCH has been
successfully utilized in other large-scale tracer gas studies.
Introduction of a novel tracer gas will make great strides in improving the versatility of
underground tracer gas ventilation surveys, but further improvement to the tracer gas technique
can be made in simplifying individual steps. One such step which would benefit from
improvement is in sampling. Solid phase microextraction (SPME) is a sampling method that is
designed for rapid sampling at low concentrations which provides precise results with minimal
training. A SPME extracting phase ideal for trace analysis of mine gases was selected and a GC-
ECD protocol was established. The protocol for fiber selection and method optimization when
performing trace analysis with SPME is described in detail in this thesis. Furthermore, the
impact of sampling with SPME under varying turbulent conditions is explored, and the ability of
SPME to sample multiple trace analytes simultaneously is observed.
iii
|
Virginia Tech
|
1 Introduction
1.1 Preface
This thesis contains three primary documents which have been submitted for scholarly
review or are prepared for submittal. Chapter 3, Selection of a novel mine tracer gas for
assessment of ventilation systems in underground mines, was submitted to the 2012 SME peer
review conference proceedings and was presented at said conference in Seattle, WA. Chapter 4,
Gas chromatograph- electron capture detector method for solid phase microextraction sampling
of mine gases, has been submitted to the Journal of Chromatography A and is currently under
review. Chapter 5, Evaluation of solid phase microextraction as a sampling method in turbulent
conditions and for the simultaneous sampling of multiple trace analytes, is ready for submission
to the Journal of Occupational and Environmental Hygiene. Additionally, a literature review
introducing tracer gas analysis and its application to mine ventilation surveys, perfluorocarbon
tracers, and solid phase microextraction is included.
1.2 Background
Mine ventilation surveys executed via tracer gas surveys are a direct means of
determining air quantity in an underground ventilation circuit. The application of a tracer gas
survey allows for the measurement of air quantity in locations of a mine with irregular flow or
where the cross-sectional area cannot be measured to accommodate a more traditional mine
ventilation survey, as well as locations which are inaccessible to personnel [1]. Unfortunately,
tracer gas surveys are an underused tool in the mining industry, however, due to the large amount
of training required to analyze the results as well as the process being relatively slow. While a
tracer gas survey is a powerful tool, the time necessary to gather all pertinent information can be
daunting due to the time required to transport the gas samples from the collection site to the
location where they can be analyzed, as well as the time necessary for the applied tracer to
completely elute from the ventilation circuit before another tracer release can take place.
1
|
Virginia Tech
|
1.3 Objectives
While the scope of this research project is large, two primary objectives are addressed in
this thesis. The objectives addressed in this thesis are designed to meet the objectives of the
research project, but also to enhance existing tracer gas survey protocols as applied in the mining
industry. The research objectives listed below are discussed in the following sections.
Objective 1: Improve existing methodologies for tracer gas surveys by making them more
flexible and to reduce the time required to execute a comprehensive survey
Objective 2: Investigate sampling methods which will simplify tracer gas surveys on-site
that can be easily integrated into existing protocols
1.3.1 Objective 1
The major factor hindering tracer gas surveys in underground mines is the existence of
only one industry standard tracer gas: sulfur hexafluoride (SF ) [2]. SF is used as the industry
6 6
standard because it is a non-toxic compound which is chemically and thermally stable. It exists
as a gas with a minute atmospheric presence of 3 molar parts per trillion (ppt) in the ambient
atmosphere [3]. This tracer gas is easily detected down to a few parts per trillion when
performing analysis with a gas chromatograph with an electron capture detector (GC-ECD) [2].
Limitations arise with the use of only one tracer gas because only parameters under investigation
can be observed, eliminating the ability to monitor baseline conditions. Additionally, the
application of only one tracer reduces the number of data points which can be collected. Most
significantly, the execution of a comprehensive ventilation survey requires that the circuit be
observed in segments with sufficient time allowed for the total elution of the tracer gas from the
mine before subsequent segments can be observed. These limitations can be mitigated with the
development of a protocol for a novel tracer gas which can be used in tandem with SF .
6
2
|
Virginia Tech
|
This thesis will discuss the selection of a tracer gas novel to the mining industry. The
protocol for adjusting existing analysis methods employing SF will be discussed. This thesis
6
will also emphasize the non-toxic nature of the selected tracer compound such that it may be
released in active workings of underground mines. Additionally, the response of the selected
tracer will be discussed in relation to the response of SF to analysis systems.
6
1.3.2 Objective 2
When tracer gas surveys are executed at underground mines, samples are often shipped to
off-site laboratories for analysis, eliminating the need for miner training in gas chromatography
analytical methods. This practice leaves the major need for miner training in the realm of
sampling. Traditional sampling mechanisms in underground mines include glass syringe bottles
with tight fitting rubber stoppers [4], disposable plastic syringes and Tedlar bags [2], and
evacuated containers [5]. The performance of these sampling mechanisms was previously
evaluated for the broader research project [6], but the application of solid phase microextraction
(SPME) was reevaluated in this thesis.
SPME sampling was reevaluated due to the limited training required to apply the
sampling mechanism on-site and the robust, precise results consistently obtained with its
application. This thesis will detail the selection process of the appropriate SPME fiber for mine
ventilation surveys, as well as a protocol for developing an optimized method for sample
analysis with a GC-ECD. Furthermore, the impact of sampling with SPME under turbulent
conditions and sampling multiple mine tracer gases simultaneously will be studied.
3
|
Virginia Tech
|
2 Literature Review
2.1 Introduction to Trace Analysis
Gas chromatography (GC) is the premiere technique employed to achieve separation and
analysis of volatile compounds. GC is used to analyze both organic and inorganic compounds of
any phase including gases, liquids, and solids. GC allows for rapid analysis which is
simultaneously efficient and sensitive. Methods used in GC analysis allow for quantitative
analysis of small samples while achieving high accuracy at a relatively low cost. The primary
limitation of GC analysis is limited to the characteristics of the sample. Analytes must be
volatile and must not be thermally labile. Large samples of pure compounds and samples that
require intense preparation are also a limiting factor. Finally, GC analysis alone cannot be used
to identify or confirm the identity of a peak. A secondary method such as GC with mass
spectrometer is necessary for confirmation of peak identity [7].
The instrumentation of a GC unit includes five primary components: sample inlet,
column, oven, detector, and data system. When a sample is injected into a GC unit, it is injected
at the sample inlet. A carrier gas constantly flows through the sample inlet through the column,
and finally to the detector. The carrier gas serves as a vehicle for transporting the sample
through the system. The flow rate of the carrier gas is controlled to ensure reproducible retention
times. Upon departure from the sample inlet, the sample flows with the carrier gas into the
column, which is enclosed in the oven. The oven maintains the temperature of the column
allowing for isothermal analysis (constant temperature) or analysis using temperature
programming. Columns in a GC system can either be packed or capillary. If a capillary column
is used, it may be either a wall coated open tubular (WCOT), porous layer open tubular (PLOT),
or a support-coated open tubular (SCOT) column. Each type of column contains a mobile and
4
|
Virginia Tech
|
stationary phase. While the sample is inside the column, the individual components within the
sample partition between the gas and liquid phases creating separation based upon the relative
solubility of the components in the liquid phase and the sample relative vapor pressures.
Separation allows for the components of the sample to pass through the detector individually for
quantification. [7].
The selection of the detector for a GC unit defines the types of analytes which may be
observed in the unit. While many specialized detectors exist for GC units, the most popular
detectors include the flame ionization detector (FID), the thermal conductivity detector (TCD),
and the electron capture detector (ECD) [7]. The FID is the most widely used GC detector. The
FID employs an oxy-hydrogen flame to burn the sample to produce ions. The ions are collected
and an ion current is produced. The FID allows for sensitive analysis for organic compounds
only and is a stable analysis method with excellent linearity [8]. The TCD is a differential
detector that employs filaments incorporated into a Wheatstone Bridge circuit that measures the
thermal conductivity of the analyte in the carrier gas in comparison with the thermal conductivity
of the carrier gas alone. The TCD is universal with moderate sensitivity and linearity [9]. The
ECD is a highly sensitive, highly selective detector with only fair linearity that is employed for
compounds that capture electrons [10]. The ECD uses a 63Ni source that releases radioactive β-
particles. The β- particles then collide with the make-up gas, nitrogen. The results of collision
include the production of free electrons with a high standing current. When an electronegative
analyte elutes from the column and enters the detector, it captures free electrons. The ECD
measures the difference in the standing current and the reduced current when the analyte captures
free electrons [7].
5
|
Virginia Tech
|
2.2 Overview of Trace Analysis Applied to Mine Ventilation Surveys
Ventilation controls exist in underground mines to provide a sufficient quantity and
quality of air to locations of the mine where personnel travel and work. The air supplied to
personnel should be safe to breath and provide reasonable comfort [1]. The ability to measure
air quantity is important to be able to determine the adequacy of existing ventilation systems in
underground mining operations. Additionally, the knowledge of air quantity will aid in rescue
and recovery efforts in a post mine-disaster situation. As such, the measurement of air quantity
through mine ventilation surveys occurs in all underground coal mines. Air quantity is typically
determined using Equation 2.1. Q is the quantity of air measured in ft3/min, A is the cross-
sectional area at the point of measurement recorded in ft2, and V is the air velocity in units of
ft/min. When air quantity is calculated, the cross-sectional area and the air velocity must be
measured. Common methods of determining air velocity include the use of smoke tubes, vane
anemometers, velometers, and pitot tubes [11]. The cross-sectional area is typically measured
directly.
Equation 2.1
Determination of air quantity using direct measurement (i.e., via Equation 2.1) is
common because it is a rapid method with a low cost. However, in situations where direct
measurements of air velocity and/or cross-sectional area may not be observed, other methods of
determining air quantity must be applied. One such method of quantifying airflow is the tracer
gas technique. In addition to being applicable to locations inaccessible to personnel, due to
either location or safety in the case of an emergency, and locations of irregular cross-sectional
area, the tracer gas technique for ventilation surveys may be used in areas of the mine where
flow is irregular [11]. Tracer gas surveys may also be employed in locations of excessively
6
|
Virginia Tech
|
turbulent flow, such as a main shaft [1]. Furthermore, tracer gas surveys are favorable in
situations where there is any of the following conditions exist: there is recirculation of return air
into intake air, there is leakage from adjacent mines, there is lost air, or there are unknown transit
flow times through stoped areas [12].
The tracer gas technique allows for the direct measurement of the air quantity by
controlling the injection rate of tracers into the observed ventilation circuit, which provides a
means of measuring air flow [1]. Following the release of the tracer gas into the mine
atmosphere, the gas is recaptured and analyzed in a gas chromatograph (GC) to determine the air
quantity. Tracer gases that have been used to varying degrees of success in mine surveys include
hydrogen, nitrous oxide, carbon dioxide, ozone, radioactive krypton 85, and sulfur hexafluoride
(SF ) [1].
6
2.2.1 Performance of a Mine Ventilation Tracer Gas Survey
Tracer gas surveys can be accomplished through two different methods. The first method
is a continuous release of tracer which allows for a steady state to develop downstream of the
release point and is best applied to high velocity airways. Once the tracer gas has been
sufficiently mixed into the mine air downstream, samples are collected so that the concentration
of tracer present can be observed [1]. Analysis of concentration of the tracer is achieved through
GC analysis. The air quantity is then determined using Equation 2.2. Again, Q is the quantity of
air measured in ft3/min, Q is the feed rate of the tracer measured in ft3/min, and C is the
g
concentration of the tracer gas in the sample measured in ft3/ft3 [11].
Equation 2.2
The other method of conducting a tracer gas survey is better applied to portions of a
ventilation circuit where airflow is slow. This method requires the release of a known mass of
7
|
Virginia Tech
|
the tracer gas into the system. Following the release, the presence of the tracer gas is observed
downstream forming a peak with the passage of time [1]. The resultant air quantity is a function
of the various concentrations of tracer in samples over time, and can be calculated by Equation
2.3. The variables Q and Q have been previously defined; C is the average of the measured
g av
concentrations in units of ft3/ft3, and τ is the total time for which tracer(s) were measured in
T
minutes [11].
Equation 2.3
2.2.2 Disadvantages of Tracer Gas Mine Ventilation Surveys
The tracer gas technique of performing ventilation surveys is an advantageous tool, but
the technology is lacking due to the large amount of time required to complete a survey. The
time limitation is largely due to the fact that SF is the only tracer compound applicable in the
6
mining industry as a standard [2]. SF is a favorable tracer gas because it is non-toxic, odorless,
6
colorless, and both chemically and thermally stable. Additionally, SF is a manmade product
6
which is already present at approximately 3 ppt in the ambient atmosphere [3]. SF is the mining
6
industry standard tracer gas for ventilation surveys for the aforementioned reasons, as well as the
fact that only small volumes of the gas are required to trace large ventilation systems. Also, SF
6
may be measured accurately down to a few parts per trillion when utilizing a GC-ECD system
[2].
In addition to the time requirements of tracer gas surveys, another significant limitation is
the fact that, at present, these surveys are only performed with a single tracer. This means that
only the parameter under investigation can be monitored, and precludes monitoring of baseline
ventilation conditions. Additionally, the use of one tracer reduces the amount of data points
which could be collected to allow for a more comprehensive ventilation survey. The most
8
|
Virginia Tech
|
detrimental impact of having only one trace compound is the time that must be allowed for a full
ventilation survey to take place. When a ventilation survey calls for multiple injects of SF , the
6
survey must be divided into different portions for each injection to allow sufficient time for the
previously injected tracer to elute completely from the ventilation survey. The application of
additional tracers is expected to help resolve the problems associated with single-gas surveys [2].
2.2.3 Previously Studied Tracer Gases for Mining Applications
In one particular study, a selection of Freon gases, or chlorofluorocarbon gases, were
tested to be used in tandem with SF with the successful implementation of one as a tracer
6
compound to be used in tandem with SF [2]. Freon gases were tested as tracers because they
6
satisfy the general tracer requirements: non-toxic, non-corrosive, non-radioactive, chemically
stable, thermally stable, odorless, colorless, readily obtainable, inexpensive, easily stored and
transported, of gaseous form at mine temperatures and pressures, and not naturally occurring in
the environment. Additionally, Freon gases are compatible with the existing sampling systems
for SF , and they are able to be measured in a method similar to that of SF . Freon gases known
6 6
to have been previously tested include Freon-12, Freon-13, Freon-13B1, Freon-22, Freon-23, and
Freon-116 [2]. Of the Freon gases tested, only two could be measured with sensitivity similar to
that of SF : Freon-12 and Freon-13B1. The method used to analyze the Freon gases in tandem
6
with SF used a Perkin-Elmer Sigma model 2000 gas chromatograph paired with an electron
6
capture detector. The method requires the application of two separate columns to be used in
tandem in a parallel system incorporating a switching valve located in the oven of the gas
chromatograph which allows each column to be placed in turn on stream with the detector. The
system requires a molecular sieve 5A column for SF and Freon-13B1 and a Porapak Type P
6
column for Freon-12. The Freon gases were sampled using plastic syringes and tedlar bags.
9
|
Virginia Tech
|
Samples were collected and stored for 24 hours before analysis took place to simulate real world
delays during analysis. Tedlar bags proved to be a reliable means of sample collection, but
plastic syringe proved to be susceptible to sample loss, particularly during their initial use.
When the plastic syringes were used for a second time, sample loss was reduced, which indicates
that initial loss was probably due to adsorption of the trace compounds onto the internal part of
the plastic syringe. While sample loss was reduced for plastic syringes, the amount of loss was
still at an unacceptable level for Freon-13B1, reducing sampling options to only sampling via
tedlar bags. The results of the experimentation with Freon gases as mine tracers was successful;
however, background levels of Freon gases must be measured before analysis due to their
widespread application as a refrigerant and propellant in aerosol cans [2].
Another tracer compound for which a methodology has been produced is 133Xe [13]. The
radioactive isotope may serve as a trace compound because it is detectable at a concentration of
4.5x10-10 µg/g. 133Xe has a sufficiently short half-life to encourage a specific activity
characteristic amenable to detection at low concentrations, but, at 5.25 days, the half-life is not
so short that it will decay within the timeframe of the testing period. The decay of the tracer
during testing would result in a difficult testing method that would not allow for reproducibility.
133Xe is a noble gas. As such, 133Xe is both chemically and physiologically inert. As a
radioactive element, 133Xe decays by beta emission, but it also releases gamma-rays and X-rays.
The use of 133Xe as a trace compound requires the use of the pulse-dilution tracer method
utilizing total-count measuring for trace analysis [13].
As with conventional tracer gas technique ventilation surveys, the measurement of cross-
sectional area is not required to determine air flow; however, the utilization of the total-count
method for measurement of the tracer means that radioactivity is measured to indicate tracer
10
|
Virginia Tech
|
concentration. Trace analysis is completed using a radioactive detector system which contains a
large-area proportional counter, amplifier, high voltage generator, and datalogger. The
radioactive detector system is sufficiently sensitive to detect 133Xe because it detects both low-
energy gamma-ray emitters and β emission. Radioactive trace analysis is employed in the field,
and has successfully been used in model and abandoned mines. Application of the method
requires baseline measurements both before and after the release and capture of the radioactive
tracer, which is released in a pulse-dilution method. The duplication of baseline measurements
ensures that the radioactive material has eluted from the mine [13].
Tracer gas surveys have also been applied to the evaluation of auxiliary ventilation
systems, which can be heavily influenced by leakage in tubing. Formerly these airflows and
pressure losses were measured using pitot tubes and pressure gauges, a method which is time
consuming and can be imprecise when compared with the tracer gas method. The application of
SF as a tracer gas allows for the survey to be completed rapidly with only two to three persons
6
required to complete the survey [14].
Tracer gas surveys can be used to determine the cause of anomalies within a ventilation
circuit. For instance, SF has been successfully used in a tracer gas survey to determine not only
6
the presence of air movement within the sealed regions of a mine, but also the direction in which
the air was moving [15]. This study ascertained the knowledge that air was leaking from old
workings of another seam, and that atmospheric pressure changes have a significant impact of
not only seal leakage, but also on the concentration of gas samples gathered at seal sample pipes.
Additionally, SF was the tracer gas employed to determine air velocity in the gob of a coal mine
6
to determine if the caved in regions of the mine were properly ventilated. Furthermore, SF was
6
11
|
Virginia Tech
|
used to determine that reverse leakage from an intake airway was cause by the pressure pulses
instigated by a traveling train in the adjacent airway [15].
2.3 Introduction to Perfluorocarbon Tracers
The term perfluorocarbon tracer, or PFT, defines a group of volatile organic compounds
[16] which can be individually quantified during trace analysis. The compounds are fully
fluorinated hydrocarbons. The overwhelming presence of fluorine in PFTs qualifies this group
of compounds as ideal tracers because they have properties such as: low atmospheric background
levels, an atmospheric lifetime on the order of 3,000 years, inert qualities, and the ability to be
detected at trace levels in the parts per trillion range with a GC-ECD [17]. PFTs are sensitive to
detection due to their reaction with free electrons; they have great stability where their
counterparts would degrade due to thermal or environmental conditions,; and they tend to be
non-toxic [18]. PFTs are particularly sensitive to detection via a GC-ECD because the high
number of fluorine atoms present paired with the structure of the molecules causes PFTs to have
high electron affinities on the order of three eV [19]. The stability of PFTs is a direct result of
the strong fluorine-carbon bond which forms the basis of their chemical structure [19].
The discussed characteristics of PFTs make them ideal tracers for a variety of
applications including underground leak detection, atmospheric transport and diffusion studies,
and oil and gas reservoir studies [19]. PFCs containing 1 to 4 carbon atoms are gases, 5 to 19
carbon atoms are liquids, and 20 to 24 atoms are solids [20]. The boiling points for PFCs
existing as liquids range between 44.5 and 125.2⁰C [21]. PFCs are also particularly non-polar
[22]. Their non-polar state reduces their vulnerability to loss by adsorption. The susceptibility
of PFTs to sample loss due to high volatility is a problem when using PFTs as liquid tracers.
12
|
Virginia Tech
|
2.3.1 Non-toxicity of PFTs
Due to their non-toxic nature, perfluorocarbon compounds (PFCs) have been used in a
variety of medical applications. One well documented application of PFCs in the medical field is
the application of a perfluorochemical liquid as a surfactant in subjects with respiratory ailments.
For instance, with PFCs existing as high density compounds with low surface tension, liquid
PFCs are used in partial liquid ventilation in premature infants with severe respiratory distress
syndrome because they act as a surfactant, physically opening the alveoli [23]. The research
stemmed from the successful application of the partial liquid ventilation treatment with
perfluorochemical liquid in premature lambs ailed by respiratory distress syndrome [24]. The
perfluorochemical liquid has also been used to treat congenital diaphragmatic hernia in lambs
[25]. Piglets were used to model the impact of treating adults affected with respiratory distress
syndrome with perfluorocarbon associated gas exchange [26]. Canine lung injuries were also
successfully treated with partial liquid breathing via perfluorochemical liquid [27]. The injected
PFCs leave the respiratory system via exhaled air, but trace amounts were found to remain in the
subjects of the various studies. The residual PFCs present in all subjects had no adverse effects
due to the non-toxic nature of PFCs.
The application of perfluorocarbon liquids in vitreoretinal surgeries, or any surgical
process treating the eye, macula, or vitreous fluid, has been extensively evaluated with the
discovery of no adverse effects associated with the use of these compounds [28]. Due to their
non-toxic nature and their ability to carry oxygen, the ability of PFCs to serve as red blood cell
substitutes has been investigated [29]. PFCs have also served as synthetic oxygen-carrying
solutions in anesthetized patients during surgery without impacting the body in any manner
outside of increasing the mixed venous oxygen tension in the patient [30]. PFCs have also been
13
|
Virginia Tech
|
used as a medium for contrast enhancement in animals and people [31]. Beyond their medical
applications, perfluorocarbon resins are used in the lubricating and softening of human and
animal skin due to their inert nature [32]. Due to their non-toxic nature, PFCs are the preferred
tracers used to tag illicit drugs and money which come into direct contact with people [33].
2.3.2 Methods of Analysis
The analysis of PFCs is a subject which has acquired a great deal of attention due to the
application of the compounds to long term atmospheric tracing experiments. The
chromatographic methods of analysis have focused on the column type, the polymeric extracting
phase, and detector selections. While PFCs have low polarity in common, they also have
multiple degrees of polarity and varying physical properties presents a conundrum when
attempting to apply one general methods to the gamut of compounds [20]. Early efforts to
separate PFCs were limited to success with only the lighter of the compounds due to the
existence of only packed columns. Middle weight PFCs with four to nine carbon atoms were
successfully separated in a packed column with graphitized carbon black serving as the solid
support coated with a moderately polar stationary phase [34]. It is likely that the greatest success
when using packed columns was found with the use of graphitized carbon black as the solid
support because it is completely nonpolar, is able to be heated to high temperatures
accommodating a wider variety of boiling points, and it has small, nonporous particles allowing
for stable suspensions and compact layers [35]. Later capillary columns were successfully used
to achieve ultrasensitive and selective separation of various PFCs [36]. Further success was
found with the commercial production of PLOT columns due to their affinity for fixed gases,
light hydrocarbons and volatile solvents [37]. PLOT columns allow for the separation of PFCs
whose boiling points are very close to one another because their application increases resolution
14
|
Virginia Tech
|
by removing liquid phase as a factor influencing the chromatographic results, and in turn
eliminating column bleed [38]. Alumina chloride (Al O /Na SO ) has proved to be the best
2 3 2 4
stationary phase used in PLOT columns to optimize sensitivity when performing analysis of
PFTs [39].
Several detectors have been used during analysis of PFCs with varying degrees of
success. The TCD has been used to successfully quantify PFCs containing between four and
nine carbon atoms [34]. Some PFCs, perfluoroalkanes specifically, have been successfully
detected with the FID [40] with relatively low sensitivity. This low response was attributed to
the dilution instigated by the carrier gas as the analytes pass over the detector [41]. Sensitivity
when performing analysis of PFCs with an FID can be increased with the application of a
hydrogen rich flame [42]. The most documented success achieved during the separation of PFCs
has used the ECD. Early studies showed that the sensitivity of the ECD to varying PFCs ranged
from low to high [43]. Since then ECD sensitivity has improved to the low parts per trillion
range [44]. Femtogram levels of detection are achievable with electron-capture negative ion
chemical ionization mass spectrometry [36], but it is not necessary with the achievable
sensitivity of the ECD.
2.4 Applications of Perfluoromethylcyclohexane in Trace Analysis
Perfluoromethylcyclohexane (PMCH) has a molecular weight of 350 g/mol and the
following chemical formula: C F . The chemical structure of PMCH is displayed in Figure 2-1
7 14
[45]. Having seven carbon atoms, PMCH exists as a liquid at room temperature. PMCH has a
boiling point of 76⁰C [46], but, like all PFCs, it is highly volatile. The density of PMCH is 1.8
g/cm3 [47]. PMCH is non-toxic, but can be irritating to the eyes or skin when direct contact
15
|
Virginia Tech
|
occurs and can be irritating to the respiratory system when inhaled [48]. Symptoms are
alleviated with a thorough rinsing with water.
Figure 2-1: Chemical Structure of PMCH
2.4.1 Use as a Gaseous Tracer
PMCH in its standard liquid form is often used as a tracer liquid in oil and gas reservoirs,
but is most commonly applied as an atmospheric tracer where it is vaporized and dispersed in its
gaseous. The movement to understand the impact of humans on climate and the environment on
regional and global scales has allowed for the method of releasing PFCs including PMCH to
develop. PMCH was the tracer gas released in a neighborhood to understand the processes
dictating flow and pollutant dispersion on this scale [49]. On a larger scale, PMCH was mixed
with a stream of nitrogen gas which was then heated and enabled the release of PMCH gas from
designated points in Dayton, Ohio and Sudbury, Ontario. This allowed for the simulation of the
transport and diffusion of pollutants sampling 1000 km from the release points in the Cross-
Appalachian Tracer Experiment [50]. The Across North America Experiment was comprised of
66 tracer releases from two locations that were sampled up to 3000 km away However, due to
the frequency of releases, individual tracer plumes were not able to be separated [51]. PMCH
was selected as one of the tracer gases to be used in the European Tracer Experiment to
16
|
Virginia Tech
|
determine atmospheric long-range transport models to assess the threat of point source releases
across Europe. Tracer gases released from France were tracked at 168 ground stations across 20
countries [52].
Beyond long range transport and diffusion modeling PMCH is used as a tracer gas to
measure air infiltration and ventilation in homes and commercial buildings and to pinpoint leak
sources in a variety of systems. Methodologies employing PMCH as a tracer gas to measure the
magnitude and point sources of air infiltration in home ventilation systems have been extensively
developed [44]. Methodologies have also been developed for more complex ventilation circuits
employing two HVAC systems. With regard to leak detection, PMCH gas can be used in tandem
with other PFTs to locate and size leaks at depth within a variety of soils with resolutions of
fractions of an inch [53]. PMCH has been successfully applied to leak detection applications in
subsurface barriers [54] and buried natural gas pipelines [53]. PMCH can also be used as a
tracer gas to improve leak testing in components of operating systems of the electric utility
industry, as well as rapid screening of the seal of packaged foods [55].
2.4.2 Use in Tandem with SF
6
PMCH and SF are known to be good tracer gases with a great deal of existing literature
6
to allow for methodologies for new applications to be readily developed. Like PMCH, SF has
6
been discovered to be a powerful atmospheric tracer [3] and has been successfully employed as
an atmospheric tracer gas in Perth to develop an urban dispersion model [56]. Also like PMCH,
SF has been used as a tracer gas to measure air infiltration rates in homes [57]. However, the
6
analytical cost of using the two tracer gases in tandem presents significant challenges. Initial
success for the simultaneous analysis of both gases was met with a two column GC system as
applied in oil and gas reservoirs [58]. More recently, an alumina chloride PLOT column has
17
|
Virginia Tech
|
been considered, which allows for the simultaneous analysis of both PMCH and SF6 in one
simple GC-ECD system [6].
2.5 Solid Phase Microextraction
Solid Phase Microextraction (SPME) is a solvent-free sampling mechanism capable of
integrating sample extraction, concentration and introduction of the sample into a GC in a simple
method which can be applied on-site [59]. Designed to be a sampling device for complex liquid
and gaseous matrices, many of the original applications of SPME were in water analysis. SPME
was developed to alleviate the disadvantages of the existing solid phase extraction (SPE)
technologies. SPE is a means of liquid-liquid extraction in which analyte is absorbed onto a solid
support and then desorbed by means of thermal desorption or applications of a solvent. The
disadvantages of SPE include necessary modifications to the GC injector port when desorbing
analytes via thermal desorption, significant increases in analysis time, high variation, and loss of
analyte to the plastic SPE cartridges [60]. The resultant SPME product is an inexpensive, easy to
use sampling device that has all of the advantages of SPE technologies, but few of the pitfalls.
2.5.1 Theory of SPME
SPME is a process based on equilibration; when collecting a sample with SPME, a small
amount of extracting phase coats a fiber. The extracting phase collects sample for a
predetermined amount of time, beginning as soon as the fiber contacts the matrix. When
sufficient time is allowed, the extracting phase will reach concentration equilibrium with the
matrix being sampled; once concentration equilibrium occurs, the extracting phase will no longer
collect sample. This process can be described as a multiphase equilibration process where the
phases considered include the extracting phase and a homogeneous matrix. The mass of analyte
collected in field situations where the matrix being sampled has a very large volume is expressed
18
|
Virginia Tech
|
in Equation 2.4. The mass of analyte collected is defined as n, K is the fiber coating-sample
fs
matrix distribution constant, V is the fiber coating volume, and C is the initial concentration of
f 0
a given analyte in the sample. The volume of sample collected by the SPME fiber is directly
proportional to the concentration of analyte in the matrix without any dependence on sample
volume [61].
Equation 2.4
2.5.2 Function of SPME
SPME works in two primary phases. Solute absorption from matrix onto the fiber occurs
during the first phase. A SPME fiber traditionally consists of a small amount of solid extracting
phase coated onto a small fused silica rod. Essentially, the SPME fiber functions much like an
inverted segment of a capillary column. SPME fibers can be exposed directly to headspace or
solution [62].
Transfer of analytes from the SPME fiber into the GC system occurs during the second
phase. SPME may be applied to GC or liquid chromatography (LC). Additionally, SPME is a
capable sampling method to achieve robust results when using a portable GC on-site [63] in
addition to its ability to be analyzed on a traditional GC system with either manual or automated
injections. Analytes that have adsorbed onto the fiber are thermally desorbed into the GC unit at
the sample inlet. The use of SPME requires that the sample inlet be fitted with a septa designed
for high heat applications, and a glass liner with an inner diameter of a specified inner diameter
[64].
2.5.3 Fiber Type
SPME fibers are contained within an apparatus for use, but the fiber itself consists of a
needle, plunger and fiber core. Three types of cores exist for SPME fibers. The first and
19
|
Virginia Tech
|
original core is the fused silica core. The fused silica core is made of standard optical quartz, and
is favorable for its endurance of high heat and general durability. Fused silica, however, is
fragile. Due to the fragility of fused silica, Stableflex fibers were created. Stableflex fibers are
fused silica with a plastic coating. The advantage of Stableflex fibers is their flexibility. The
third core available is a metal core. Metal fibers are inert because they do not contain iron. They
are favorable because better bonding with the fiber coating is achieved compared to the bonding
of coating with fused silica and Stableflex fibers. Additionally, metal fibers do not require glue
to attach the fiber into the assembly, and there are no extraneous peaks with metal fibers. Metal
fibers, however, are more expensive than the traditional fused silica and Stableflex fibers [64].
2.5.4 Fiber Coating
Common extracting phases applied to the fibers include carboxen (CAR),
polydimethylsiloxane (PDMS), divinylbenzene (DVB), polyacrylate, polyethylene glycol (PEG),
and combinations of the aforementioned phases. When selecting an extracting phase for a SPME
fiber, the adsorptive qualities, polarity, porosity, and coating thickness should be considered.
The adsorptive qualities are largely dependent on whether the selected extracting phase is a
liquid coating such as PDMS or polyacrylate or a solid coating like CAR [65]. Liquid coatings
extract analyte via absorption such that the analyte partitions through the extracting phase while
solid coatings extract via adsorption where the analyte collects on the surface of the extracting
phase [66]. Adsorbent fibers physically trap or chemically bond analytes to the fiber. Adsorbent
fibers are porous materials with high surface areas. The porous qualities of adsorbent fibers
make them better for trace analysis. Additionally, increasing porosity causes the analytes to be
retained more tightly. Varying pore size increase analyte selectivity. Fibers can be micro-,
meso-, and macro-porous [64]. The limited capacity of adsorbent fibers causes analytes to
20
|
Virginia Tech
|
compete for sites. Absorbent fibers, on the other hand, achieve extraction by partitioning.
Analytes partition into the liquid phase such that the thickness of the coating is of high
importance. A coat of higher thickness increases sample capacity, but it also increases extraction
time and desorption time. The elimination of competition amongst analytes for sites on the fiber
allows for fibers capable of high capacity. Finally, polarity of the coating impacts the analyte
selectivity. More polar phases allow for better recovery, but have a shorter life.
2.5.5 Sample Extraction and Recovery
Sample extraction with SPME is achieved by exposing the fiber to the sample. Care
should be used when the fiber is exposed to ensure that it does not come into direct contact with
anything but the sample due to its fragile nature. If the sample is contained, the first step in
sample extraction is to pierce the sample septum with the fiber secured within the apparatus.
Then the fiber should be exposed for the appropriate amount of time. Once the sampling time
has elapsed, the fiber should be withdrawn back inside the apparatus. Only once the fiber is
retracted should the needle be withdrawn from the sample container.
When employing SPME for sample analysis, sample extraction is affected directly by
time. In order to achieve reproducibility, the fiber should be exposed to the sample for the same
amount of time during each use. Every fiber reaches a point of equilibrium in which the fiber is
fully saturated. This time can be measured by creating an uptake or equilibrium curve which
plots the resultant peak areas as processed through a GC against the time during which the fiber
was exposed to the sample matrix [67]. The time of equilibrium is influenced by kinetic effects
[68].
Some such kinetic effects include stirring and agitation, which reduces the equilibrium
time, and may increase the sample temperature. When stirring/agitation of the sample is
21
|
Virginia Tech
|
consistent, the precision for all analyses increases. While unnecessary for volatile compounds,
compounds with high distribution constants require agitation. Sample extraction may also be
improved with alterations to the salt and pH of the sample. Salt increases analyte uptake, but is
not helpful when collecting large, nonpolar analytes. Reducing the pH for acidic compounds and
increasing the pH for basic compounds will improve sample extraction. Finally, direct versus
headspace sampling impacts extraction [64].
Sample recovery is performed in the same fashion as sample extraction. The SPME
apparatus pierces the GC inlet septum. Following the penetration of the septum, the fiber is
exposed. The analyte desorbs from the phase coating the fiber while it is exposed in the sample
inlet of the GC. Following the completion of analysis, the fiber is retracted back into the SPME
apparatus, at which time the apparatus may be removed from the GC unit. Sample recovery is
directly impacted by fiber selection, sample modifications, extraction time, desorption
conditions, inlet design, and column selection [69]. The application of SPME to sampling during
the course of underground mine ventilation surveys may introduce a robust on-site sampling
mechanism which could simplify survey procedures, encouraging an increase in the use of the
technique.
22
|
Virginia Tech
|
3 Selection of a Novel Mine Tracer Gas for Assessment of Ventilation Systems
in Underground Mines
3.1 Abstract
The tracer gas method of conducting ventilation surveys is a means of determining air
quantity in underground mines with knowledge of the concentration of the tracer in the
atmosphere and the flow rate at which the tracer is being released. The technique is useful in
inaccessible areas of underground mines and areas where traditional point measurement of
velocity is not practical. Sulfur hexafluoride (SF ) is the industry standard tracer [2] used in
6
underground mines because it is safe, stable, and not naturally occurring in the mine
environment. The implementation of a second tracer will increase the versatility of the tracer gas
technique allowing for simultaneous releases for the study of interrelated ventilation circuits, and
for conducting multiple experiments in less time. This paper will detail the selection of
perfluoromethylcyclohexane (PMCH) as a novel tracer with sensitivity, physical properties, and
analysis characteristics similar to SF . Methods for the release of liquid PMCH into underground
6
mines will be recommended, and the vulnerabilities of sample loss due to condensation of vapor
PMCH will be explored. Finally, the benefits and implications of using PMCH as a second tracer
will be discussed.
3.2 Introduction
The measurement of air quantity through mine ventilation surveys is a common
occurrence in underground mining operations. The knowledge of air quantity in the mine is
significant in the determination of the adequacy of existing ventilation systems, and in providing
information after a mine disaster [11].
23
|
Virginia Tech
|
While the calculation of air quantity from the direct measurement of air velocity and
cross-sectional area is most commonly used due to the ease with which these surveys may be
conducted, the low cost of the equipment, and the small amount of time it demands, it is not
always a feasible means to determine air quantity. Places where traditional point measurement
of velocity is not practical include locations in the mine which are inaccessible to personnel,
locations where the cross-sectional area cannot be measured through simple measurement or
photographic methods, and locations that are not deemed safe in the case of an emergency. Also,
equipment used for point measurement is not always appropriate for very low velocities, such as
leakage through stoppings. The tracer gas technique of determining the air quantity is a viable
method of conducting ventilation surveys where other methods are not applicable. Additionally,
the tracer gas technique is advantageous in areas where air flow is irregular [11].
The tracer gas technique allows for the direct measurement of the air quantity. The tracer
gas technique requires the release of a tracer element into the mine atmosphere, which can later
be analyzed. There are two general methods of conducting the tracer survey. One method calls
for the continuous release of the tracer into the airway allowing equilibrium to be reached.
Samples are collected and processed by means of a gas chromatograph (GC) which, combined
with the knowledge of the method, produces a peak area that indicates the concentration of the
tracer in the collected sample. The air quantity is then determined using Equation 3.1.
The quantity of air, Q, measured in cubic feet per minute (ft3/min), Q is the feed rate of the
g
tracer measured in ft3/min, and C is the concentration of the tracer gas. This concentration is
typically unit less, but measurements may be classified on a volumetric basis so that samples
may be measured in units of cubic feet per cubic feet (ft3/ft3). The tracer gas technique may also
be implemented by releasing a known mass of tracer into the airway and collecting samples until
24
|
Virginia Tech
|
the tracer has passed through the point of collection. The resultant air quantity is a function of
the various concentrations of tracer in samples over time, and the final relationship defined in
Equation 3.2. The variables Q and Q are measured in ft3/min, C is the average of the
g av
measured concentrations, a unitless number that may be classified volumetrically in units of
ft3/ft3, and τ is the total time for which trace elements were measured in minutes.
T
Equation 3.1
Equation 3.2
While the tracer gas technique adds variability to the available means of determining air
quantity, ventilation surveys are still limited in inaccessible regions because the tracer gas
technique is a slow method. Not only is the speed of the survey limited by the amount of time
required by the gas chromatograph to process samples, but also by the need to allow background
levels of the tracer element to be removed from the ventilation system. The industry standard for
mine tracer gas applications is sulfur hexafluoride (SF ). The addition of a second, novel tracer
6
element will increase the versatility of tracer surveys in that multiple airways/leakages could be
traced simultaneously, eliminating the problem of the long wait time between runs with the use
of only one tracer element. Use in a mine setting requires that the tracer element be easily
detected and analyzed, absent from the mine air, not absorbed by chemical or physical means in
the air way, and non-reactive, toxic, corrosive, nor explosive [11]. Additional desirable
properties of tracer elements include that they are readily attainable, easily transported, and
inexpensive. To be used in tandem with SF the tracer must exhibit high sensitivity (similar to
6
that of SF ) to obtain desirable chromatographic results. Perfluorocarbon tracers (PFTs),
6
specifically perfluoromethylcyclohexane (PMCH), meet the demands required of a novel tracer.
25
|
Virginia Tech
|
3.3 Background
The knowledge that PMCH is a suitable mine tracer warrants further investigation of the
chemical. While PMCH is a liquid at room temperature, it is highly volatile with a boiling point
of 76⁰C [70]. As such, PMCH vaporizes at a relatively low temperature, and will remain in the
vapor state through cooler temperatures. PMCH has a molecular weight of 350 g/mol, which is
favorable because it is much heavier than SF (molecular weight of 146 g/mol). The difference
6
in molecular weights will allow for the easy separation of peaks during GC analysis. The
chemical formula of PMCH is C F , and it contains no unsaturated bonds as seen in Figure 3-1.
7 14
The fully saturated bonds cause PMCH to be extremely stable both chemically and physically.
As such, PMCH is also biologically inert so that, even if ingested or inhaled, it is not harmful to
humans [71]. In addition to being safe for people at trace levels, PMCH is safe to use in the
environment. The fully saturated fluorine atoms do not react with the ozone ensuring that PFTs
do not affect the ozone layer [71]. Also, the ambient background of all PFTs is approximately
0.03 parts per trillion making any single PFT an ideal trace element [71].
Figure 3-1: Fully Fluorinated PMCH Structure [5]
PMCH has been used extensively as a tracer element in industries outside of mining.
One such study used PMCH to study the dispersion mechanisms in the atmosphere of an urban
26
|
Virginia Tech
|
environment. On three separate occasions PMCH was released into a neighborhood setting in a
vapor state, and samples were collected anywhere from one to nine kilometers from the site of
release, with measured concentrations ranging from 0.003-2.16 µg/m3 [49]. In this case PMCH
was successfully incorporated as a tracer gas in an open environment/neighborhood setting, and
was traced at very low concentrations a great distance from the source. PMCH was also used as
a tracer in much the same fashion in the European Tracer Experiment, in tandem with
perfluoromethylcyclobutane (PMCB) [72].
Further application of PMCH as a tracer element includes its use in oil and gas reservoirs.
As within the ambient atmosphere, the background of PFTs is also very low in underground
settings so as not to interfere with tracer experimentation. In one case, PMCH is used in tandem
with SF to study flow patterns between injection and production wells. PFTs are favorable for
6
these purposes due to their high detectability at low concentrations and chemical inertness, as
well as their successful use in atmospheric transport studies and marine studies. Results from
this reservoir study measured PMCH tracers in the parts per billion range [58].
Another instance in which PMCH was used as a tracer gas element was in home
ventilation systems. In order to determine where in the ventilation system heated or cooled air is
lost, PMCH was introduced to directly measure air infiltration in homes under actual living
conditions. The study was designed to be long term to best determine where air infiltration
occurred, and the study neither suffered sample loss of the volatile PMCH vapor to condensation,
nor were there any side effects on residents in the home. Additionally, adsorption losses of the
PMCH tracer gas were negligible. [44].
Based on the three discussed uses of PMCH as a tracer gas, it is evident that a human
population can safely be exposed to vapor PMCH without fear of harmful side effects.
27
|
Virginia Tech
|
Additionally, PMCH has been used as a successful tracer in both open environments, as well as
enclosed ventilation systems, indicating negligible sample loss to absorption of the tracer to
exterior elements, as well as limited condensation of the tracer. Furthermore, PMCH is proven
to be readily detectable at very low concentrations and great distances from its source. In one
case the minimum detection limits of various PFTs were evaluated. An Al O /Na SO column
2 3 2 4
was successfully implemented in experimentation. The limit of detection of PMCH was
determined to be 50 pg [39]. When considering the volume of air that flows through
underground mines, it is apparent that the determined detection limits fall far below the ppm
range routinely used in mine ventilation survey evaluations.
The use of multiple tracers in mine ventilation applications is, by no means, a novel
concept. Gases such as nitrous oxide and helium have been used as mine tracers with little
success due to their low detectability [2]. A gas with low detectability would be difficult to
manage in an underground mine due to the vast volume that would be required to be released in
order to be detectable in tandem with SF . Freon gases including Freon-13B1 and Freon-12 were
6
also selected as acceptable tracers to be used with SF due to their properties making them
6
suitable tracers for underground applications, their compatibility with existing sampling systems,
and due to the fact that they are able to be measured in the same or a similar method as SF [2].
6
The resultant method used a two column system GC with an electron capture detector (ECD).
The results of the experimentation with Freon gases as mine tracers was successful; however,
background levels of Freon gases must be measured before analysis due to their application as a
refrigerant and propellant in aerosol cans [2].
Further experimentation with mine tracers applied the use of 133Xe, a radioactive isotope
of xenon gas. This isotope of xenon gas is a suitable radioactive tracer because it is chemically
28
|
Virginia Tech
|
and physiologically inert, and its half-life of 5.25 days is sufficiently long that it will not decay
before the completion of the testing and that it will be sufficiently detectable. The radioactive
tracer was released in abandoned mines. Background measurements were observed before and
after the release/capture of the radioactive tracer to ensure that pooling and absorption of the
tracer did not occur. The radioactive tracer was released using a pulse-dilution method, and was
detected using a radioactive detector designed for the study [13].
3.4 Application
A primary consideration in the selection of a novel tracer to use in underground mining
applications is the sensitivity of the new element relative to that of SF . PMCH is known to be an
6
exceedingly sensitive tracer. As discussed, it has been successfully used as a tracer element in
both gaseous and liquid states to high levels of success. SF and PMCH have been successfully
6
used in tandem in oil and gas reservoirs. In one case SF and PFTs including PMCH as well as
6
PMCB, perfluoromethylcyclopentane (PMCP), perfluorodimethylcyclohexane (PDMCH), and
perfluorotoluene were successfully used simultaneously in analysis. Each of the aforementioned
PFTs is a fully saturated compound. A GC with an ECD was used to detect the tracer elements,
utilizing two columns in tandem [58]. The successful use of PMCH and SF in tandem in oil
6
reservoirs indicates that PMCH is a promising option for a novel tracer in mine survey
applications.
In order to successfully use a two tracer system in mining applications, it is important to
have a simple analysis protocol to minimize analysis time, maintenance, and training. As such, a
two column system is not suitable.
An ECD is used to detect the concentration of tracer in the air samples. The ECD uses a
63Ni source that releases radioactive β- particles. The β- particles then collide with the make-up
29
|
Virginia Tech
|
gas consisting of nitrogen and helium. The results of collision include the production of free
electrons and a high standing current. When an electronegative analyte elutes from the column
and enters the detector, it captures free electrons. The ECD measures the difference in the
standing current and the reduced current when the analyte captures free electrons. The
mechanics of the ECD makes it a highly selective detector. [7].
As demonstrated with other tested tracers, including carbon tetrafluoride (CF ) and
4
carbon octafluoropropane (C F ), difficulties were encountered in selecting a tracer with
3 8
sufficient sensitivity to be able to be used simultaneously with SF , as well as separated from
6
oxygen in air. Both Freon gases CF and C F were experimented with using SBP-1 Sulfur, ZB-
4 3 8
624, TG Bond Q+, and TG Bond Q columns. These two Freon gases were selected due to the
previous success of Freon gas application as mine ventilation tracers [2], as well the fact that
they do not pose known health risks to people, have molecular weights sufficiently different
from SF to indicate that separation is possible, and they are both readily available in compressed
6
gas cylinders. Additionally, the presence of fluorine compounds in their chemical makeup
indicate that they should be easily detectable using an ECD. In order to be selected as the novel
tracer, it was important that one or both of the Freon gases be successfully applied in tandem
with SF . To satisfy this requirement, the selected novel tracer must be as sensitive as SF on the
6 6
same method. A method of detection using one column on a GC-ECD is desirable, and the
Freon gas must present itself as a Gaussian shaped peak that is clearly separated from both air
and SF on chromatograms.
6
While testing the selected Freon gases on the SBP-1 Sulfur column, sufficient separation
between peaks could not be achieved, nor were the peak areas of the Freon gases comparable to
30
|
Virginia Tech
|
those of SF . The resultant average retention times and peak areas of the pertinent gases for the
6
two experimental methods are listed in Tables 3-1 and 3-2, respectively.
Table 3-1: Average Retention Time of Gases on SBP-1 Sulfur Column
Column Average Retention Time (min)
Temperature O SF CF C F
2 6 4 3 8
40⁰C 1.10 1.22 1.10 1.20
20⁰C 1.08 1.19 1.10 1.24
Table 3-2: Average Peak Area of Gases on SBP-1 Sulfur Column
Column Average Peak Area
Temperature SF CF C F
6 4 3 8
40⁰C 8.91x107 1.43x104 1.44x104
20⁰C 7.69x107 1.56x104 1.30x104
While testing on the ZB-624 column, similar problems were encountered as with the
SBP-1 Sulfur column. Additionally, CF was not detected using the ZB-624 column. The
4
resultant retention times and peak areas of the pertinent gases for the three experimental methods
are listed in Tables 3-3 and 3-4, respectively.
Table 3-3: Average Retention Time of Gases on ZB-624 Column
Column Average Retention Time (min)
Temperature O SF CF C F
2 6 4 3 8
40⁰C 2.14 2.22 --- 2.23
20⁰C 2.14 2.22 --- 2.27
0⁰C 2.14 2.24 --- 2.17
Table 3-4: Average Peak Area of Gases on ZB-624 Column
Column Average Peak Area
Temperature SF CF C F
6 4 3 8
40⁰C 5.93x107 2.07x103 1.02x104
20⁰C 6.12x107 --- 8.63x103
0⁰C 6.57x107 --- 2.20x105
31
|
Virginia Tech
|
Testing on the TG Bond Q+ column produced similar results as with the ZB-624 column.
Difficulties were encountered with the production of peaks from CF . The resultant retention
4
times and peak areas of the pertinent gases for the three experimental methods are listed in
Tables 3-5 and 3-6, respectively.
Table 3-5: Average Retention Time of Gases on TG Bond Q+ Column
Column Average Retention Time (min)
Temperature O SF CF C F
2 6 4 3 8
40⁰C 1.94 2.22 2.12 3.49
20⁰C 1.96 5.22 --- 5.22
0⁰C 1.96 9.65 --- 9.84
Table 3-6: Average Peak Area of Gases on TG Bond Q+ Column
Column Average Peak Area
Temperature SF CF C F
6 4 3 8
40⁰C 5.93x107 2.79x103 1.15x105
20⁰C 1.091x108 --- 6.15x104
0⁰C 1.100x106 --- 4.13x104
Testing using the TG Bond Q column resulted in adequate separation of C F from
3 8
oxygen, but not from SF . CF was not able to be analyzed on the selected column as it
6 4
produced no peaks, compared to the limited peaks produced using the ZB-624 and TG Bond Q+
columns. The resultant retention times and peak areas of the pertinent gases for the five
experimental methods are listed in Tables 3-7 and 3-8, respectively.
Table 3-7: Average Retention Time of Gases on TG Bond Q Column
Column Average Retention Time (min)
Temperature O SF C F
2 6 3 8
60⁰C 2.30 3.10 3,09
50⁰C 2.33 3.42 3.41
40⁰C 2.24 3.88 3.88
20⁰C 2.27 5.74 5.70
0⁰C 2.27 10.6 ---
32
|
Virginia Tech
|
Table 3-10: Average Peak Area of Gases on HP-AL/S Column
Column Average Peak Area
Temperature SF PMCH
6
70⁰C 3.44x107 4.95x107
67⁰C 4.24x107 6.34x107
65⁰C 6.18x107 7.97x107
Table 3-11: HP-AL/S Column Characteristics
Column Name HP-AL/S
Column Type Capillary
Column Length 30 m
Inner Diameter 0.250 mm
Film Thickness 5 µm
The protocol established for the separation of PMCH in the selected column is detailed in
Table 3-12. PMCH contains a significant amount of impurities requiring the use of temperature
programming to remove the impurities from the column. The initial temperature of the column
is relatively low at 67⁰C and held for sufficient time to allow SF to elute. PMCH requires a
6
higher temperature for both elution, as well as separation from contaminants. The selected split
ratio of 50:1 allows for the development of Gaussian shaped peaks for both PMCH and SF
6
during analysis. In application, the split ratio may be adjusted and then corrected for quantitative
analysis to the split ratio determined in protocol. The resultant analysis of PMCH in tandem
with SF had desirable peak shapes and separation as seen in Figure 3-4. The developed protocol
6
is a balance of satisfactory separation and efficient analysis time.
35
|
Virginia Tech
|
is 350. PMCP and PMCB also have lower boiling points than that of PMCH: 47-53⁰C and 45⁰C,
respectively. PMCP, however, is incompatible with strong oxidizing agents and metals, and may
evolve into CO, CO2, or hydrogen fluoride if it encounters incompatible materials [73]. PMCB
must avoid products of combustion, heat, strong oxidizing agents, and must be kept cold when
stored [74]. PMCH is the most suitable PFT for underground mining applications based on the
conditions which must be avoided with the selection of either PMCP or PMCB.
3.5 Future Work
PMCH has been successfully established as a tracer element meeting the demands for use
in mine ventilation surveys. However, there is still much that is unknown about the
fluorocarbon. For instance, while studies have not suffered from sample loss, there is minimal
information available on PMCH vulnerabilities to condensation. Additionally, an efficient
method of release suitable for underground mining applications must be developed.
3.5.1 Sample Loss Considerations
Other tracer gas studies have not shown significant sample loss when using PMCH such
that further investigation into the matter is not warranted. This conclusion may be drawn based
on the results of other studies that sample loss will be negligible in underground mining
applications because PMCH has been used as a tracer in its vapor state in a variety of
environments.
The European Tracer Experiment implemented the use of PFTs in a study to model the
atmospheric transport across Europe. Tracer elements were released and traced for three days
during experimentation. While a great deal of information on the condensation habits of PMCH
is not readily available, the European Tracer Experiment used Equation 3.3 to model the
relationship between vapor pressure and temperature such that dew point may be determined.
37
|
Virginia Tech
|
P is the vapor pressure in units of mbar and T is the temperature in units of K. [72]. This
PMCH
relationship may serve as an indication of the possibility of sample loss to condensation should
the pressure of the surveyed region be available. However, the relationship does not serve to
indicate the amount of loss which may be endured should the dew point be reached. Based on
various studies, including the home air infiltration study [44], it may be concluded that the threat
of sample loss to condensation is minimal.
Equation 3.3
3.5.2 Methods of Release
One possible obstacle to bypass with the selection of PMCH as a novel tracer is the
method of release into a mining atmosphere. The European Tracer Experiment used a chimney
system to release vapor PMCH directly into the atmosphere continuously for a set amount of
time [72]. While a chimney method is not applicable in an underground mine, there are three
likely means of release to be considered. The first is with the use of a permeation tube which
will, when placed in an environment whose temperature is strictly controlled, release vapor
PMCH at a steady, known/calibrated rate [75]. The emission rate of PMCH from a permeation
tube would be limited by the temperature in which it is contained. A second means uses a
fluoroelastomer plug impregnated with a known mass of PMCH, and employed in the home
ventilation system analysis of air infiltration [44]. The plug naturally emitted PMCH vapor at a
slow rate over a period of more than five years. The third means of release is purchasing
cylinders containing vapor PMCH in nitrogen. Such containers are capable of releasing PFT
vapors up to 100 mg/min as controlled by a mass flow controller, but must be contained below
their dew point [70]. Each method of release has been successfully implemented in various
applications and could well be practical in a mine ventilation survey.
38
|
Virginia Tech
|
3.6 Conclusions
Improvement of the versatility of ventilation surveys is important in underground mining
applications. Improved tracer gas techniques can significantly improve the flexibility of
ventilation surveys. Radioactive gas and Freon gas tracers have been successfully used in the
past, but a novel tracer that may be used in active workings with a simple analysis method is
desirable. Two Freon gases, CF and C F were tested in tandem with SF on various columns
4 3 8 6
with various methods with little or no success. Peak separations between the Freon gases and
oxygen or SF6 were not readily achievable, nor were Gaussian shaped peaks captured.
Experimentation with PMCH and SF allowed for peak separation and the development of
6
Gaussian shaped peaks. PMCH is a favorable selection for a novel tracer to work in tandem with
SF due to its chemical stability, similar physical properties and detection limits to SF , and its
6 6
ability to be applied and integrated into an existing system. Additionally, PMCH has been
successfully utilized in other large-scale tracer gas studies. A method using a Shimadzu GC-
2014 with ECD and HP-AL/S column was developed to simultaneously detect SF and PMCH.
6
While there is yet much to be learned about the compound, future work will focus on the
development of an efficient method of release into an underground mine.
39
|
Virginia Tech
|
4 Gas Chromatograph – Electron Capture Detector Method for Solid Phase
Microextraction Sampling of Mine Gases
4.1 Abstract
Precise methods for the quantification of tracer gases in underground mine ventilation
studies are vital to the maintenance and design of mine ventilation systems, which in turn, play a
critical role in mine safety and productivity. This study was undertaken to determine if solid
phase microextraction (SPME) is a feasible sampling method for quantitative tracer gas analysis
as applied to underground mine ventilation. This paper details an optimized method for SPME
sampling techniques, as well as a simple GC-ECD method. Both methodologies are rapid and
reproducible. An equilibrium curve relating fiber exposure time under static conditions and a
methodology for creating a calibration curve using a SPME sampling method is also described.
4.2 Introduction
A tracer gas ventilation survey is performed by releasing a tracer which is easily detected
and analyzed, absent in normal mine air, chemically stable and inert, non-toxic and non-
explosive [11]. Tracer gas ventilation surveys are especially applicable in areas of a mine where
point measurement of velocity and quantity are difficult, such as inaccessible areas and areas
with highly irregular cross sections. The standard tracer gas employed by the mining industry is
sulfur hexafluoride (SF ) [2]. A tracer gas ventilation survey may be executed by a continuous
6
release of tracer or a pulse release of tracer [4]. When using the continuous release method,
sufficient time must be allowed for thorough mixing and air samples are taken downstream.
An integral step in the performance of a tracer gas survey is sampling. Sampling must be
completed in a standardized manner which allows for reproducible results, but is also convenient
with regard to application, storage and transportation, as well as cost effectiveness. The manner
40
|
Virginia Tech
|
of sampling must also be sensitive as the concentrations of tracer achieved in samples are often
quite small due to the relatively large quantities of air being sampled. Traditional sampling
methods applied in underground mines include glass syringe bottles with tight fitting rubber
stoppers [4], disposable plastic syringes and Tedlar bags [2], and evacuated containers [5]. A
comprehensive study of these common sampling methods was performed comparing the
precision, sensitivity, robustness, convenience, and cost of these sampling devices [6]. This
study concisely summarized the advantages and disadvantages of various sampling methods, but
a sampling method novel to the mining industry was also investigated: solid phase
microextraction (SPME). The results of the investigation showed that, while SPME fibers were
the least sensitive of the sampling methods, they allowed for the lowest amount of error.
SPME is a highly sensitive, simple and inexpensive sampling technique originally
designed for liquid and gaseous matrices. SPME works in two phases. Solute absorption from
the matrix into the fiber occurs during the first phase. A SPME fiber consists of a small amount
of solid extracting polymeric phase coated onto a small fused silica rod. SPME fibers can be
exposed directly to headspace of liquids and solids or solutions requiring minimal sample
preparation efforts [76]. Transfer of analytes from the SPME fiber into the GC system occurs
during the second phase. Analytes that have adsorbed onto the fiber are thermally desorbed into
the GC sample inlet [77].
The application of SPME as a sampling device in underground mines is encouraging
because it is a solvent free extraction method designed for rapid sampling at low concentrations.
The use of SPME is proven to be precise, but has not yet been proven to be sufficiently sensitive
for underground mine ventilation surveys [6]. However, SPME has been proven to be a rapid
sampling method for trace elements with a high molecular mass [78]. The sensitivity of SPME
41
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.