Science of the Total Environment 481 (2014) 317–334
Contents lists available at ScienceDirect
Science of the Total Environment
j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv
Prediction and innovative control strategies for oxygen and hazardousgases from diesel emission in underground mines
Jundika C. Kurnia a,b, Agus P. Sasmito c,⁎, Wai Yap Wong a, Arun S. Mujumdar a,c
a Minerals, Metals and Materials Technology Centre, National University of Singapore, Engineering Drive 1, Singapore 117576, Singaporeb Mechanical Engineering, Masdar Institute of Science and Technology, Masdar City, Abu Dhabi, P.O. Box 54224, United Arab Emiratesc Department of Mining and Materials Engineering, McGill University, 3450 University Street, Frank Dawson Adams Bldg Room 115, Montreal H3A2A7, QC, Canada
H I G H L I G H T S
• The dispersion of hazardous gases from diesel emission in underground mine is evaluated.• Oxygen concentration is analyzed for a miner to breathe.• Underground temperature is measured for thermal comfort.• Side tailpipe turned duct design performs the best.• Compromise between controlling hazardous gases and energy saving
Diesel engine iswidely used in undergroundminingmachines due to its efficiency, ease of maintenance, reliabil-ity and durability. However, it possesses significant danger to the miners and mining operations as it releaseshazardous gases (CO, NO, CO2) and fine particles which can be easily inhaled by theminers. Moreover, the dieselengine consumes significant amount of oxygen which can lead to insufficient oxygen supply for miners. It istherefore critical to maintain sufficient oxygen supply while keeping hazardous gas concentrations from dieselemission below the maximum allowable level. The objective of this study is to propose and to examine variousinnovative ventilation strategies to control oxygen and hazardous gas concentrations in undergroundmine to en-sure safety, productivity and cost related to energy consumption. Airflowdistribution, oxygen and hazardous gasdispersion aswell as ambient temperature within themining area are evaluated by utilizing the well-establishedcomputational fluid dynamics (CFD) approach. The results suggest that our newly proposed ventilation designperforms better as compared to the conventional design to handle hazardous gases from diesel emission.
For decades, internal combustion (IC) engine, especially the diesel en-gine, has been widely used in underground mines due to its high fuel ef-ficiency, ease ofmaintenance, reliability anddurability. It is amajor powerplant for various equipments utilized in underground mines, rangingfrom continuous miners to fans. Despite its important utilization in un-derground mine, diesel engines possess significant danger to the minersandmining operations. Diesel engines are themajor sources of hazardousgases (CO, CO2, NOX, SO2 and hydrocarbons (HC)) and submicron parti-cles (Park et al., 2012) commonly referred to as diesel particulatematters(DPMs) in underground mines (Burgaski et al., 2011). In addition, it
consumes significant amount of oxygen in the mining area which has al-ready been limited forminers to breathe. There is a growing evidence thatoverexposure to diesel emission over long periods of time adversely af-fects humanhealth in variousways. Prolonged exposure to diesel exhaustis believed to be carcinogenic (Volkwein et al., 2008). With advancementin undergroundmining operations, bigger andmore powerful engines aredeployed in underground mine to speed up and to increase productioncapacity which shall result in higher profit. These engines consumemore fuel and oxygen and release more hazardous gases. Hence, it is ofparamount importance to design an underground ventilation systemwhich can provide sufficient oxygen supply for the miners and engineswhile maintaining hazardous gases below the maximum allowable level.
To facilitate proper design of a ventilation system for undergroundmine, it is necessary to develop in-depth knowledge of the airflowpatterns and gases dispersion. Numerous studies on diesel emissionin underground mines have been reported. Zheng and Tien (2008) nu-merically investigated the airflow pattern and diesel particulate matter
318 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
(DPM) emission in undergroundmetal/nonmetal mine for single head-ing. They found that blowing face ventilation distributed DPM in signif-icantly smaller space than exhausting face ventilation. Mischler andColinet (2009) reported an evaluation of control technologies andmon-itoring instrumentations that are used to reduce DPM exposure of mineworkers. An integrated approach is proposed to control diesel partic-ulate emission with the highlight on the effective design of ventila-tion system. Haney and Saseen (2000) developed a model toestimate DPM exposures and to determine the impact of various die-sel particulate controls on occupational exposure to the DPM in un-derground mines. The model employed either in-mine measuredDPM concentrations or standardized emission data from enginemanufacturers through a series of calculations to estimate full shiftDPM concentrations.
Ray et al. (2004) employed computational fluid dynamics (CFD)modeling to investigate distribution of diesel emission gases in under-ground passenger rail road tunnel where train with locomotive dieselpass through. The mass flow rate and temperature of diesel exhaustwere modeled at locomotive exhaust port. Major concern of this studyis to evaluate concentration of NO2 gases when the train passes throughthe tunnel. They found that the application of mechanical ventilationsystem when NO2 level reaches 3 ppm prevented excessive NO2
concentration from building up. Another finding is that continuousoperation of mechanical ventilation system could eliminate dieselexhaust gas layers forming in the vicinity of diesel locomotive. Aninvestigation of various methods to reduce and to measure DPMaerosols in underground coal mines was conducted by the U.S. Bu-reau of Mines and was reported by Watts et al. (1989). They evalu-ated the effectiveness of the integrated dry system and dieselparticulate filters. They found that DPM emission can be reducedby up to 97% with little change in gaseous emission concentration.Stachulak et al. (2006) reported long-term field evaluation of dieselparticulate filter (DPF) system which is available in undergroundmining industry. They evaluated nine state-of-the-art DPF systembased on several criteria: filter media, means of DPF regeneration,efficiency and occurrence of unwanted emissions. The results wereused to evaluate the effects of DPF on the concentration of DPMand diesel exhaust gases. In addition, a review on diesel emissionand control which summarizes the latest developments in dieselemissions regarding regulations, engines, NOX (nitrogen oxides)control, diesel particulate matter (DPM) reductions, and hydrocar-bon (HC) and CO oxidation have been reported by Johnson (2009,2010). In risk prediction and hazardous gas mitigation field, severalstudies have also been conducted and reported. Amorim et al.(2013) evaluated the impact of urban trees over the dispersion ofhazardous gas of carbon monoxide (CO) emitted by traffic road uti-lizing computational fluid dynamics approach. The air quality canbe optimized based on the knowledge-base planning of greenspaces utilizing numerical simulation. Gallagher et al. (2013) inves-tigated the potential real world application of passive control sys-tems to reduce personal hazardous pollutant exposure of nitricoxide (NO) in an urban street canyon in Dublin, Ireland using CFDsimulation. Kountouriotis et al. (2014) simulated the dispersion ofvolatile organic compound (VOC) leak into the atmosphere due toevaporation of liquid fuel in petrol station; the wind speed and itsdirection as well as ambient temperature significantly affects thedispersion characteristic. Sasmito et al. (2013) evaluated someapproaches to improve ventilation system in underground minesand to dilute hazardous gas emission from coal mine. Majority ofthese studies, however, focuses only on the diesel particulate mat-ter (DPM) emission and dispersion of hazardous gases in indoorand outdoor environment. Hence, it is of major industrial interestto study the dispersion of oxygen and hazardous gases simulta-neously in underground mines together with mining machineryas well as evaluating thermal comfort for miners due to heat rejectionfrom mining machines.
This study addresses airflow distribution as well as the oxygenand hazardous gas dispersion within the mining area by utilizing com-putational fluid dynamics (CFD) approach. The objectives of this studyare as follows:
• To evaluate the effectiveness of a ventilation system installed inunderground mine tunnel where two diesel-powered equipments(continuous miner and shuttle) are in active operations, as presentedin Fig. 1. This would allow for risk assessment and mitigation for safe,healthy and productive underground mine environment.
• To investigate the effect of vehicle's tailpipe placement on the gasdistribution.
• To examine the effect of velocity of air supply from the blowing ductto reveal the minimum fresh air supply to maintain safe oxygen andhazardous gas level.
• To evaluate the effectiveness of suction duct in mitigating andcollecting hazardous gases from the mining area. Finally, the energyconsumption of ventilations scenarios is measured with regard tothe effectiveness and pumping power required.
2. Model formulations
A three dimensional model is developed in this work for an under-ground mine tunnel where continuous miner and shuttle car are in ac-tive operation (please refer to Figs. 2 and 3 for miner and shuttle cargeometrical detail, respectively). These two diesel-powered vehiclesare rated at 381 hp (≈284 kW) and 229 hp (≈170.76 kW), respective-ly. Details on underground mine and vehicle geometries are presentedin Tables 1 and 2.
2.1. Governing equations
In this study, the turbulent mass, momentum, energy and spe-cies transport equations subject to appropriate boundary condi-tions are solved. Oxygen and fuel are consumed by the dieselengine while exhaust gas is released from the tailpipe. Conserva-tion equations for mass, momentum, energy and species in vectorform are:
∇ � ρU ¼ 0; ð1Þ
∇ � ρUU ¼ −∇pþ∇ �μ þ μ tð Þ ∇Uþ ∇Uð ÞT
� �− 2
3½ μ þ μ tð Þ ∇ � Uð ÞI−ρkI�
� �þρg;
ð2Þ
∇ � ρcpUT� �
¼ ∇ � keff þcpμ t
Prt
� �∇T; ð3Þ
∇ � ρωiUð Þ ¼ ∇ � ρDi;eff þμ t
Sct
� �∇ωi ð4Þ
where ρ is the fluid density, U is the fluid velocity, p is the pressure,μ is the dynamic viscosity of the fluid, I is the identity or secondorder unit tensor, g is gravity acceleration, cp is the specific heat ofthe fluid, keff is thermal conductivity of the fluid, T is the tempera-ture, ωi is the mass fraction of species i (O2, H2O, CO2, CO, NO andN2), Di,eff is diffusivity of species i, μt is turbulent viscosity and Sctis the turbulent Schmidt number.
2.2. Constitutive relations
A species mixture comprising oxygen, water vapor, carbon dioxide,carbon monoxide, nitric oxide and nitrogen exists in the ventilation air
Fig. 1. Schematics of underground mining face with continuous miner and shuttle car in active operation: a) side tailpipe and b) top tailpipe configuration with forward blowing venti-lation; c) side tailpipe and d) top tailpipe configuration with turned blowing ventilation; and detail of e) side tailpipe and f) to tailpipe arrangement.
319J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
Blowing ductDiameter 0.6 mHeight from floor to the duct center 1.9 mSpace from the mining wall to the duct center 0.6 mSetback distance from the mining face 6 m
Height from floor to the duct center 2.5 mSpace from the mining wall to the duct center 3 mOpening location from mining face 8 and 20 m
321J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
in the tunnel. The interaction between the species is captured in themix-ture density which follows incompressible ideal gas law given by
ρ ¼ pMRT
; ð5Þ
where R is the universal gas constant andM refers to the mixture molarmass given by
M ¼ ωO2
MO2
þ ωH2O
MH2Oþ ωCO2
MCO2
þ ωCO
MCOþ ωNO
MNOþ ωN2
MN2
" #−1
: ð6Þ
Here, Mi is the molar mass of species i. Mass fraction of nitrogen iscalculated as
ωN2¼ 1− ωO2
þωH2O þωCO2þωCO þωNO
� �: ð7Þ
The fluid mixture viscosity is calculated using
μ ¼Xi
xiμ iXj
xiΦi; j
with i and j ¼ O2;H2O;CO2;CO;NO and N2 ð8Þ
where xi,j are the mole fraction of species i and j and
Φi; j ¼1ffiffiffi8
p 1þ Mi
Mj
!12
1þ μ i
μ j
!12 Mi
Mj
!14
24
352
: ð9Þ
The mole fractions are related to the mass fractions by
xi ¼ωiMMi
: ð10Þ
In this report, gas concentration is presented in mass fraction. Foreasy comparison, maximum allowable concentration mandated by theMSHA regulation will also be presented in terms of the mass fraction.
Table 2Geometric and operational properties of continuous miner and shuttle car.
To analyze energy consumption for each configuration, a pumpingpower required to drive the airflow will be calculated as
Ppump ¼ 1ηpump
!Q̇ΔP ð11Þ
where ηpump is the pump efficiency (assumed to be 70%),Q̇is the volumeflow rate, and ΔP is the pressure drop.
2.3. Turbulence model
The most commonly used turbulence model in engineering, stan-dard k–ε, is selected in this work. This model comprises of two-equationswhich solve for turbulent kinetic energy, k, and its rate of dis-sipation, ε, which is coupled to the turbulent viscosity.
∂∂t ρkð Þ þ∇ � ρUkð Þ ¼ ∇ � μ þ μ t
σk
� �∇k
� �þ Gk−ρε; ð11Þ
∂∂t ρεð Þ þ∇ � ρUεð Þ ¼ ∇ � μ þ μ t
σε
� �∇ε
� �þ C1ε
εkGk−C2ερ
ε2
k; ð12Þ
In above equations, Gk represents the generation of turbulence ki-netic energy due to the mean velocity gradients, C1ε and C2ε are con-stants, σk and σε are the turbulent Prandtl numbers for k and ε,respectively, and μt is turbulent viscosity given by
μ t ¼ ρCμk2
ε; ð13Þ
where C1ε, C2ε, Cμ, σk and σε are constant and the values are 1.44, 1.92,0.09, 1 and 1.3, respectively.
2.4. Diesel fuel
All diesel fuel consists of complexmixtures of aliphatic and aromatichydrocarbons. The aliphatic alkanes (paraffins) and cycloalkanes(naphthenes) are hydrogen saturated and compose approximately80–90% of the fuel oils. Aromatics (e.g., benzene) and olefins (e.g., sty-rene and indene) compose 10–20% and l%, respectively, of the fuel.Commercial diesel fuel is classified into three classes: 1D, 2D and 4Ddie-sel. The differences between these classes are on viscosity and pourpoint. 1D is a light distillate which consists primarily of hydrocarbonsin the C9–C16 range; 2D is heavier, usually blended and distillated withhydrocarbons in the C11–C20 range (Risher and Rhodes, 1995; Speight,2002).
In the United States, diesel fuel is controlled according to theAmerican Society for Testing and Materials Standard D975-97. By thisstandard, 1D and 2D diesel fuel used on roadsmust be of low sulfur con-tent. MSHA rules since 1997 have also required that all mining vehiclesand equipment must pass a low sulfur engine test before being used(Mine Safety and Health Administration). Majority engine dieselthat powered mining equipment is fueled with 2D fuel (Risher andRhodes, 1995; Speight, 2002).
2.5. Diesel engine combustions
In the combustion chamber the following reaction occur
CxH2xþ2 þ z 0:266O2ð Þ→aCO2 þ bH2O
where a = x, b = y/2, z = (a + b/2)/0.266.This equation presents stoichiometric combustion of a saturated hy-
drocarbon compound. While this reaction is only an approximation tothe real combustion of fuel (since only saturated hydrocarbons are
322 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
taken into account instead of complete aromatic compound and variousother chemical additives commonly added to the fuel), the reactionholds reasonably well as diesel fuel is predominantly C16 saturated hy-drocarbons (Pundir, 2007). In this study, 2D fuel combusted in dieselengines is approximated as C16H34 for simplification (Speight, 2002).
Fig. 4. Velocity profile in underground mine for va
2.6. Diesel engine emissions
Ideally, oxygen consumption and carbon dioxide emission should bemeasured directly from the engine to obtain precise figures. This directmeasurement, however, possesses a number of difficulties: (i) Emission
Fig. 5. Oxygen mass fraction at planes 0.1, 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 m from the mining face.
323J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
and consumption levels are dependent on a number of external factorssuch as fuel temperature prior to ignition and the injection systemused;and (ii) majority of mine operators and owners will not be willing toshare emission data of their existing fleet due to various reasons. Thus,for first approximation, a stoichiometric equation was selected to
estimate oxygen consumption and carbon dioxide emission in the solu-tion domain within the bounds of the current work. However, it is ac-knowledged that typical diesel engines burn at lean conditions, hencethe results need to be considered in terms of their relative differences,rather than as absolute concentration representative of a mine
Fig. 6. Carbon dioxide mass fraction at planes 0.1, 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 m from the mining face.
324 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
operating diesel engine. More realistic emissions data from diesel en-gine fleet in underground mining will be considered in future study toenhance accuracy of the prediction.
Fuel consumption of a diesel engine can be estimated as 0.110 gal/hp·hr (≈0.416 ltr/hp.hr) (Grisso et al., 2004). From this estimation
and given the power rate of the engine, the amount of oxygen consump-tion and carbon dioxide emission can be calculated by using the abovestoichiometric equation. Meanwhile, due to irregular carbon monoxideand nitric oxide emission from diesel engine, emission of these gases isestimated by using the maximum allowable level regulated by MSHA
Fig. 7. Carbon monoxide mass fraction at planes 0.1, 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 m from the mining face.
325J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
(i.e. 2.6 g/hp·h for CO and 6.9 g/hp·h for NO). The reason for this choiceis that diesel engine having NO and CO emission more than this valuewill not be allowed to operate.
During operation, diesel engine releases huge amount of heat to thesurroundings through exhaust, cooling system and its body. Given theengine efficiency and heating value of the fuel used, total heat loss
from the engine can be calculated. For this study, however, a simplerule-of-thumb to calculate the heat loss is implemented, i.e. only one-third of the energy input supplied to the engine (in the form of fuel) be-come useful energy (power produced by the engine), one third energyis wasted as heat loss through the exhaust and another one third is re-leases as heat loss through the cooling system (radiator) and engine
Fig. 8. Nitric dioxide mass fraction at planes 0.1, 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 m from the mining face.
326 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
body. Hence, given the power rating of the engine, heat loss from the ex-haust and heat loss from the cooling system and the body can easily becalculated.
2.7. Boundary conditions
The applicable boundary conditions are as follows:
(i) Atwalls: the standardwall function is used in all simulations; theunderground rock temperature is set at 31 °C which is in therange of type I mine temperature (Xiaojie et al., 2011).
(ii) At the inlet, air velocity of 2 m s−1 and temperature of 25 °C isprescribed.
(iii) At the outlet: stream-wise gradient of the temperature is set tozero and the pressure is set to standard atmospheric pressure(1 bar).
Fig. 9. Temperature distribution at planes 0.1, 2, 4, 6, 8, 10, 12, 14, 16, 18 and 20 m from the mining face.
327J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
(iv) At the blowing duct inlet: fresh air is supplied at velocity of 12 ms−1 (Parra et al., 2006) and at temperature of 25 °C. For blowingsuction configuration, blowing velocity of 6 m s−1 is prescribed.
(v) At the suction duct outlet: contaminated air is sucked at velocityof 12m s−1 and stream-wise gradient of the temperature is set tozero.
(vi) At the tailpipe outlet: mass flow inlet and heat flux correspondingto one-third of the energy input supplied to the engine (in the
formof fuel) is prescribed.Massflow is calculated from theoxygenconsumption and mass fraction of oxygen in air. Diesel engineemission is released at the rate calculated on Sections 2.5 and 2.6.
(vii) At the vehicle body: a heat flux corresponding to one-third of theenergy input supplied to the engine (in the form of fuel) is pre-scribed in the whole vehicle body.
(viii) At scrubber fan: to model the fan, its characteristic curve is intro-duced as an interfacial condition; the model is represented by a
328 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
polynomial function that is fitted to data from manufacturer forstatic pressure increases over the fan vis-à-vis the flow velocity,similar to our previous work (Sasmito et al., 2013).
Δpfan ¼ C1 ufanð Þ2 þ C2ufan þ C3 ð14Þ
where C1, C2 and C3 are constant whose values are –13.5 Pas2 m–2, 352.62 Pa s m−1, –692.95Pa, respectively.
3. Numerical methodology
The computational domains were created, meshed and labeledin Gambit 2.3.16. Three different amount of mesh of approximately1.2 × 106, 2.4 × 106 and 4.8 × 106 were implemented and comparedin terms of local pressure, velocities, and gas concentrations to ensurea mesh independent solution. It was found that the mesh amount ofaround 2.4 × 106 gives about 1% deviation compared to the mesh sizeof 4.8× 106;whereas, the results from themesh size of 1.2 × 106 deviateup to 12% as compared to those from thefinest one. Therefore, amesh ofaround 2.4 million elements was sufficient for the numerical investiga-tion purposes.
The governing equations together with the constitutive relations,turbulencemodel and boundary conditionswere solved using the finite
0.00
0.05
0.10
0.15
0.20
0.25
0 5 10 15 20
Ave
rage
oxy
gen
mas
s fr
actio
n
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (0.19 mass fraction)
00
0.19
0.20
0.20
0.21
0.21
0.22
0.22
0 5 10 15 20
Max
imum
oxy
gen
mas
s fr
actio
n
Distance from the mining face (m)
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (0.19 mass fraction)
b
0 5 10 15 20
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (0.19 mass fraction)
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
a
Fig. 10. Average (top) and maximum (bottom) oxygen mass fraction along the tunnel.
volume CFD code, Fluent 6.3.26. The equations were solved with theSemi-Implicit Pressure-Linked Equation (SIMPLE) algorithm, secondorder upwind discretization and Algebraic Multigrid Method (AGM).On average, each simulation required around 3000–5000 iterations tomeet convergence tolerance of 10−6 for all variables. Each run neededaround 8–12 h on a workstation with six core processor, requiring8–10 GB RAM.
4. Results and discussion
The mathematical model and flow behavior was validated in previ-ous work (Sasmito et al., 2013; Kurnia et al., 2014, in press) against ex-perimental data (Parra et al., 2006); for the sake of brevity, it is notrepeated here. Flow behavior, exhaust gas dispersion and temperaturedistribution in a mine tunnel where 2 mining equipment (continuousminer and shuttle car) is on active operation were investigated. In thefollowing, the effect of tailpipe placement and theminimum ventilationduct velocity to sufficiently maintain safe working environment in themining area is examined and discussed. Introduction of a suction ductand its effect on the exhaust gas dispersion is also presented anddiscussed.
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
0 5 10 15 20
Ave
rage
car
bond
ioxi
de m
ass
frac
tion
Distance from the mining face (m)
Side tailpipe normal duct Top tailpipe normal duct
Side tailpipe turned duct Top tailpipe turned duct
Suction Blowing suction
Threshold (0.015 mass fraction)
10-2
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
0 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
×
×
Max
imum
car
bond
ioxi
de m
ass
frac
tion
b
a
000 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct Top tailpipe normal duct
Side tailpipe turned duct Top tailpipe turned duct
Suction Blowing suction
Threshold (0.015 mass fraction)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
×
Fig. 11. Average (top) and maximum (bottom) carbon dioxide mass fraction along thetunnel.
329J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
4.1. Effect of ventilation design
4.1.1. Tailpipe placementFirst, the effect of tailpipe location on the exhaust gas dispersion and
temperature distribution in the mining area is explored and evaluated.Two common tailpipe placements are evaluated: side position(Fig. 1a) and top position (Fig. 1b). The computed velocity profiles (inthe form of velocity vector plot) are presented in Fig. 4a and b. It is ob-served that changing exhaust tailpipe only marginally affects the veloc-ity profile inside the tunnel; however, this is not the case when weclosely look at the oxygen concentration (Fig. 5a and b), exhaust gas dis-persion of CO2 (Fig. 6a and b), CO (Fig. 7a and b), NO (Fig. 8a and b) andtemperature distribution (Fig. 9a and b) as the effect becomes moreprominent. From Fig. 5, it is found that oxygen distribution for top tail-pipe configuration (Fig. 5b) is more uniform as compared to the sidetailpipe configuration (Fig. 5a). On the exhaust gas distribution, sidetailpipe configuration results in an accumulation spot with higherconcentration of CO2 (Fig. 6a), CO (Fig. 7a) and NO (Fig. 8a) gases ascompared to the top tailpipe (Figs. 6b, 7b, and 8b). In addition, side tail-pipe configuration (Fig. 9a) leads to a temperature hotspot on the rightfront of the shuttle car. Clearly, the top tailpipe configurations outper-form side tailpipe configurations in maintaining sufficient oxygen and
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
0 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (3.5 × 10-4 mass fraction)
10-5
00
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
0 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (3.5 ×
×
×
10-4 mass fraction)
10-5
Ave
rage
car
bon
mon
oxid
e m
ass
frac
tion
Max
imum
car
bon
mon
oxid
e m
ass
frac
tion
b
a
0 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (3.5 × 10-4 mass fraction)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (3.5 ×
×
10-4 mass fraction)
10-5
Fig. 12. Average (top) and maximum (bottom) carbon monoxide mass fraction along thetunnel. Threshold is not shown as it is significantly higher than the concentrationpresented.
keeping low hazardous exhaust gases in the mining area. From thisstudy, it can be seen that by simply moving the exhaust pipe from theside of the vehicle to the top, the concentration of hazardous gases inthe mine tunnel can be reduced up to 30%. It also reduces the tempera-ture inside the mine tunnel and increase oxygen concentration in themining area around 20%.
4.1.2. Flow directionFocusing on the blowing duct configurations, several questions
arise: (i)why is the fresh air from the ventilation duct blown to themin-ing facewhen no one is allowed toworking at this area? (ii)Whydo notwe direct the ventilation duct outlet and blow the fresh air to the areawhere the miners are working? Driven by this curiosity, we investigateexhaust gas dispersion and temperature distribution when the outlet ofthe ventilation duct is turned, as presented in Fig. 1c. Looking at the ve-locity profiles for side tailpipe turned blowing duct (Fig. 4c) and top tail-pipe turned blowing duct (Fig. 4d), as expected, changing the directionof the ventilation duct outlet significantly affects the airflow profiles ascompared to the normal duct designs (Fig. 4a and b). This is furtherreflected by the improvement of the available oxygen supply through-out the tunnel (Fig. 5c and d) and by the reduction of the dispersion ofhazardous gases of up to 40% (Figs. 6c and d; 7c and d; and 8c and d)
0.00
2.00
4.00
6.00
8.00
10.00
12.00
0 5 10 15 20
Ave
rage
nitr
ic o
xide
mas
s fr
actio
n
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (5 × 10-5 mass fraction)
10-5
000 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (5 × 10-5 mass fraction)
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
0 5 10 15 20
Max
imum
nitr
ic o
xide
mas
s fr
actio
n
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Threshold (5 ×
×
×
10-5 mass fraction)
10-5b
a
Fig. 13.Average (top) andmaximum(bottom) nitric oxidemass fraction along the tunnel.
330 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
as compared to the normal duct configurations (Figs. 6a and b; 7a and b;and 8a and b); and by the reduction of hot spot temperature (Fig. 9c andd). Closer inspection reveals that, for this particular case, among the twoturned duct designs, it is found that the top tailpipe configurations in-terestingly yield more uniform species distributions with lesser accu-mulation of hazardous gases and lesser hot spot temperature (around5 to 10% difference). Thus, it can be deduced that simply changing theventilation direction can enhance the removal of hazardous gases andimprove the safety for underground working environment — of coursefurther study and optimization is required for specific mine condition.
4.1.3. Exhaust ventilationAirflow pattern and gas dispersion in a tunnel when blowing venti-
lation system is utilized have been studied anddiscussed in the previoussection. Next, it is of interest to investigate the airflow pattern andgas dispersion when exhaust ventilation system is utilized instead ofblowing. In addition, the effectiveness of combination blowing and suc-tion ventilation systemwill be evaluated. Ventilation profiles when suc-tion and blowing-suction is applied are presented in Fig. 4e and f,respectively. Here, it is clear that the application of suction significantlyinfluences the airflow pattern in the mining area. As expected, thischange in ventilation system also significantly affects oxygen concen-tration (Fig. 5e and f) and exhaust gas distribution (Figs. 6e and f; 7e
310
315
320
325
330
335
340
345
350
355
0 5 10 15 20
Ave
rage
tem
pera
ture
(K
)
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
0 0
310
330
350
370
390
410
430
450
470
0 5 10 15 20
Max
imum
tem
pera
ture
(K
)
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
b
a
0 5 10 15 20
Distance from the mining face (m)
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Side tailpipe normal duct
Top tailpipe normal duct
Side tailpipe turned duct
Top tailpipe turned duct
Suction
Blowing suction
Fig. 14. Average (top) and maximum (bottom) temperature along the tunnel.
and f; and 8e and f) in the tunnel. While it could maintain relativelylower exhaust gas concentration in the tunnel as compared to side tail-pipe with normal blowing configurations, it could not keep sufficientoxygen level in the mining area. When combined with blowing ventila-tion (the blowing velocity is reduced to 6m s−1), the oxygen concentra-tion in the mining tunnel is improved. However, this combined systemperformance in keeping low exhaust gas is disappointing. It performsworse as compared to other configurations and it posses highest energyconsumption since two active ventilation system is used (fan for blow-ing and suction).
4.1.4. Overall performanceThus far, the distributions of oxygen and hazardous gases have been
qualitatively evaluated for various ventilation designs. In this section,the effectiveness of each design is compared and discussed in term ofthe average and maximum concentration of the oxygen (Fig. 10), CO2
(Fig. 11), CO (Fig. 12), NO (Fig. 13) and temperature (Fig. 14). Here sev-eral features are apparent; foremost among them is the side tailpipeturned duct performs best among others. It should be noted that, al-though the top tailpipe turned duct qualitatively looks betterwith lesseraccumulation of hazardous gases, the side tailpipe turned duct designquantitatively performs better due to lesser average hazardous gas con-centration. Looking closer to the average (Fig. 10 (top)) and maximum(Fig. 10 (bottom)) oxygen concentration, the side tailpipe turned ductgives rise to the highest oxygen concentration followed by top tailpipeturned duct with concentration peak of around 0.22 mass fraction atthe distance around 4 m from the mining face where the air is blown.This high oxygen concentration is beneficial for miners to breathe asthe concentration is above the threshold limit (0.19 mass fraction) aslisted in Table 3. While the lowest oxygen concentration is shown insuction ventilation design, followed by blowing-suction, side tailpipenormal duct and top tailpipe normal duct.
Fig. 11 presents the average and maximum concentration of carbondioxide mass fraction. The side tailpipe turned duct design yields thelowest concentration at the first 10 m from the mining face, followedby top tailpipe turned duct, top tailpipe normal duct, suction duct,blowing-suction and side tailpipe normal duct, respectively. Lookingat themaximumCO2 concentrationwith regard to themaximumallow-able limit of 0.5% v/v ~ 0.015mass fraction (Table 3), only three designs,i.e., side tailpipe turned duct, top tailpipe turned duct and top tailpipenormal duct, can meet the regulation within 6 m distance from miningface which is safe for miners to breathe. Similarly, for the dispersion ofcarbon monoxide (Fig. 12) and nitric oxide (Fig. 13), the side tailpipeturned duct design results in the lowest concentrationwith the concen-tration fall below the threshold limit (0.0035 mass fraction for CO and50 ppm mass fraction for NOX) within the maximum distance of 6 mfrom the mining face. Looking at the temperature distribution inFig. 14, side tailpipe turned duct design can provide thermal comfortfor miners with the average lowest temperature of around 325 K. Notethat additional cooling, for example by using water spray or reducingventilation air temperature further, can be considered to further reducethe ambient temperature for more comfortable environment. Overall, itcan be concluded that, in this particular case, the side tailpipe turnedduct design outperforms other designswith regard to the control of haz-ardous gases from diesel machine. Of course, further investigation isneeded to arrive at definite useful and practical design solution for spe-cific underground environment.
Table 3Threshold level for each gas (Hartman et al., 1997).
Gases Threshold
Oxygen ≥19.5% v/v (≈0.19 mass fraction)Carbon dioxide ≤0.5% v/v (≈0.015 mass fraction)Carbon monoxide ≤0.3% v/v (≈0.0035 mass fraction)Nitric oxide ≤30 mg m−3 (≈5 × 10−5 mass fraction)
Fig. 15. Average oxygen mass fraction along the tunnel for various configurations.
331J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
4.2. Effect of ventilation air velocity
For this investigation, the air velocity at the blowing duct outlet isvaried from 8 m s−1 to 16 m s−1 at 2 m s−1 interval. For combinedblowing and suction configuration, the suction velocity was kept
constant at 6 m s−1 while blowing velocity was varied. For suctiononly configuration, the suction velocity was varied. Velocities above14 m s−1 were dropped from testing as duct outlet air velocitiesabove this range does not give any significant improvement in main-taining oxygen levels and lowering exhaust gas concentrations (study
Fig. 16. Average carbon dioxide mass fraction along the tunnel for various configurations.
332 J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
on this was conducted but the results are not presented for the sake ofbrevity). This indicates that there is an effectiveness threshold in in-creasing duct outlet air velocities which is of important implicationsince this information can be made use of when trying to find an opti-mum supply velocity for design considerations.
As can be inferred from Figs. 15 and 16, in all configurations, theminimum safe air velocity is 12 m s−1. While decreasing the air veloci-ties, an interesting pattern is found between the side tailpipe and toptailpipe configurations. For the side tailpipe configuration, as velocityis decreased the effectiveness of the ventilation system in maintaining
333J.C. Kurnia et al. / Science of the Total Environment 481 (2014) 317–334
oxygen levels and lowering exhaust gas concentrations decreases in auniform manner. On the other hand, for the top tailpipe configuration,the effectiveness of the ventilation system decreases exponentiallywhen the air velocity is reduced below 12 m s−1. This is unfavorablein underground mine because in the event that duct air velocity dropsunexpectedly, a dangerous situation could build up in a very short peri-od. In all air velocities tested, the re-directed duct configuration offersbetter effectiveness compared to the front facing duct outlet configura-tion. For the combined blowing and suction configuration, increasingthe blowing air velocity does not show significant improvement in ef-fectiveness of lowering exhaust gas concentrations and the overall per-formance of this configuration fares poorlywhen compared to the othersystems. Moreover, the oxygen levels are very low when compared tothe other configurationmaking the overall performance of this configu-ration poor.
4.3. Energy consumption for various gas control strategies
Energy consumption is one of themost important aspects in design-ing ventilation system in underground mines. An optimummine venti-lation design should be able to control hazardous gases below themaximum allowable level, maintain sufficient oxygen for the minerwhile keeping energy consumption at low level. Table 4 summarizesenergy consumption for various mine ventilation design investigat-ed in this study. As expected, normal blowing duct offer the lowestenergy consumption due to its low pressure difference; whereas, ex-haust ventilation requires the highest energy consumption as com-pared to other ventilation designs. However, one needs to balancebetween energy consumption and the effectiveness for controllinghazardous gases; thus, the blowing turned duct design performs rea-sonable compromise between energy efficiency and ventilation per-formance: the pumping power is slightly higher than the blowingnormal duct; but it is still much lower than that of suction andblowing-suction designs.
5. Conclusions
A mathematical model for exhaust dispersion and temperature dis-tribution within the underground mining area has been presented andselected results were discussed. Effect of ventilation tailpipe placementon gas distributionwas examinedwith regard to theminers' safety, pro-ductivity and energy saving. Hence, the results are concluded as follows:
• Simply moving the tail pipe from bottom side to the top side can im-prove oxygen concentration in the mining area while reducing haz-ardous gas concentration.
• Turning the direction of ventilation airflow can further enhance thecontrol of hazardous gases.
• Installation of suction duct tomitigate hazardous gases frommining isunable to maintain adequate oxygen supply and is unable to keep ex-haust gas concentration at safe levels.
• Overall, the side tailpipe turned duct design performs the best amongother designs with reasonable energy consumption.
• The velocity of air supply from the blowing duct was examined to ob-tain minimum fresh air supply required to maintain safe working en-vironment for theminers. It was found that 12m s−1 is theminimumvelocity required to maintain sufficient oxygen supply and low haz-ardous gas concentration (14 m s−1 is the threshold velocity for opti-mum effectiveness).
This study provides some new ideas for designing a cost-efficientunderground mine ventilation system to control oxygen supply andhazardous gas emission. In the follow-up study, the application of anovel intermittent flow ventilation system to control exhaust gasconcentration in the tunnel could be examined. In addition, a morerepresentative internal combustion diesel engine will be consideredin future work to elucidate more detailed emission dispersion fromlight, medium and heavy duty diesel mining machineries in specificunderground mining types, viz. long wall, room and pillar and soforth.
Acknowledgments
This work was financially supported by the Singapore Economic De-velopment Board (EDB) through the Minerals, Metals and MaterialsTechnology Centre (M3TC) Research Grant R-261-501-013-414.
ReferencesAmorim JH, Rodrigues V, Tavares R, Valente J, Borrego C. CFDmodelling of the aerodynam-
ic effect of trees on urban air pollution dispersion. Sci Total Environ 2013;461–462:541–51.
Burgaski AD, Janisko SJ, Cauda EG, Noll JD, Mischler SE. Diesel Aerosols and Gases in Un-derground Mines: Guide to Exposure Assessment and Control. National Institute forOccupational Safety and Health, Office of Mine safety and Health Research:Pittsburgh, PA: National Institute for Occupational Safety and Health, Office of Minesafety and Health Research: ; 2011.
Gallagher J, Gill LW, McNabola A. The passive control of air pollution exposure in Dublin,Ireland: a combined measurement and modelling case study. Sci Total Environ2013;458–460:331–43.
Haney RA, Saseen GP. Estimation of diesel particulate concentration in undergroundmines. Min Eng Littleton 2000;52:60–4.
Hartman HL, Mutmansky JM, Ramani RV, Wang YJ. Mine Ventilation and Air Conditioning(3rd Ed). New York: Wiley; 1997.
Johnson TV. Diesel emission control in review. SAE Int J Fuel Lubr 2009;2:1–12.Johnson TV. Review of diesel emissions and control. SAE Int J Fuel Lubr 2010;3:
16–29.Kountouriotis A, Aleiferis PG, Charalambides AG. Numerical investigation of VOC levels in
the area of petrol stations. Sci Total Environ 2014;470–471:1205–24.Kurnia JC, Sasmito AP, Mujumdar AS. Dust dispersion and management in underground
mining faces. Int J Min Sci Technol 2014;24(1):39–44.Kurnia JC, Sasmito AP, Mujumdar AS. CFD simulation of methane dispersion and innova-
tive methane management in underground mining faces. Appl Math Model 2014.http://dx.doi.org/10.1016/j.apm.2013.11.067. [in press] [http://www.sciencedirect.com/science/article/pii/S0307904X13008111].
Mine Safety and Health Administration. Mining diesel engine regulations. Availableat http://www.dieselnet.com/standards/us/ohs.php. [Last accessed 28 January 2013].
Mischler SE, Colinet JF. Controlling and monitoring diesel emissions in undergroundmines in the United States. In: Panigrahi DC, editor. Proceedings of the 9th Interna-tional Mine Ventilation Congress. New Delhi, India: Oxford & IBH Publishing; 2009.p. 879–88.
Park D, Yoon Y, Kwon SB, Jeong W, Cho Y, Lee K. The effects of operating conditions onparticulate matter exhaust from diesel locomotive engines. Sci Total Environ2012;419:76–80.
Parra MT, Villafruela JM, Castro F, Mendez C. Numerical and experimental analy-sis of different ventilation systems in deep mines. Build Environ 2006;41:87–93.
Pundir BP. Engine emissions: pollutant formation and advances in control technology.Alpha Science InternationalPangbourne: Alpha Science International; 2007.
Ray RE, Kang K, Gilbey M. CFD analysis of diesel emission in a railroad tunnel. In: GanguliR, Bandopadhyay S, editors. Proceeding of the 10th North American Mine VentilationSymposium, Alaska USA. London: Balkema Publisher (Taylor & Francis Group); 2004.p. 117–28.
Risher JF, Rhodes SW. Toxicological profile for fuel oils. Agency for Toxic Substances andDisease Registry, U.S. Department Of Health And Human ServicesGeorgia: Agency forToxic Substances and Disease Registry, U.S. Department Of Health And Human Ser-vices; 1995.
Sasmito AP, Birgersson E, Ly HC, Mujumdar AS. Some approaches to improve ventilationsystem in underground coal mines environment — a computational fluid dynamicstudy. Tunn Undergr Space Technol 2013;34:82–95.
Stachulak JS, Conard BR, Bugarski AD, Schnakenberg Jr GH. DEEP project on evalua-tion of diesel particulate filters at Inco's Stobie Mine. In: Mutmansky JM,Ramani RV, editors. Proceeding of the 11th North American Mine VentilationSymposium, Pennsylvania, USA. London: Balkema Publisher (Taylor & FrancisGroup); 2006. p. 143–9.
Volkwein JC, Mischler SE, Davies B, Ellis C. Field measurement of diesel particulate matteremissions. Ann Occup Hyg 2008;52:99–105.
Watts Jr WF, Baumgard KJ, Cantrell BK, Rubow KL. Control of Diesel ParticulateMatter in Underground Coal Mines, Report of Investigations 9276. United States De-partment of The Interior, Bureau of Mines; 1989.
Xiaojie Y, Qiaoyun H, Jiewen P, Xiaowei S, Dinggui H, Chao L. Progress of heat-hazardtreatment in deep mines. Min Sci Technol (China) 2011;21:295–9.
Zheng Y, Tien JC. DPM dispersion study using CFD for underground metal/nonmetalmines. In: Wallace KG, editor. Proceeding of the 12th North American Mine Ven-tilation Symposium, Reno, Nevada, USA. Reno: University of Nevada; 2008.p. 281–6.
