Validation of the Ventgraph Program for use in Metal/Non-Metal Mines

1
Validation of the Ventgraph Program for use in
Metal/Non-Metal Mines
Christopher Pritchard MS P.E. Mining
Engineer NIOSH OMSHR Spokane, WA
2
Summary

• Introduction
• Description of 1972 Sunshine Mine and Fire Event
• Recreating the ventilation network
• Modeling network and mine fire
• Model verification
• What If Scenarios
• Summary and Conclusions

3
Introduction

• NIOSH Research Plan - 2008
• Reduce incidents and severity of fires in M/NM
  Mines
• Increase understanding of products of combustion
• Improve ventilation and escape planning
• Investigate and choose ventilation software
• Verify by using actual M/NM mine fire incident

4
Ventilation Software Requirements

• Mine Fire simulation capabilities
• Real time
• Non-compressible flow
• NVP
• Graphics
• Choice Ventgraph for Windows

5
Fire 1972 Sunshine Mine

• Historically significant
• Deep
• Complicated ventilation system
• Compressed air injection
• Multiple fans
• Drastic changes to ventilation system

6
Previous Sunshine Work

• 1977 USBM Report Hall/Ramani/Greuer
• Network model only
• Semi NVP
• No fire
• No contaminants
• No graphics

7
Sunshine Ventilation System
8
1972 Fire Event

• 1140 am Smoke detected 3700L _at_ No. 10 Shaft
• 1200 Mine evacuation begun by 3100L route,
  surface notified, 3700L fire door closed at
  Jewell Shaft
• 1245 last survivors hoisted to 3100L _at_10 SH,
  100 last hoist run.
• 115 No. 12 SH cover removed

9
Fire and Evacuation Sequence
10
Sunshine Pictures
Jewell Shaft entrance
Workings
11
Sunshine Pictures
Stope
10 Shaft Hoist at 3100 Level
12
Sunshine Pictures
No. 8 Shop area smoke detected
Blue Room 3700 Level
13
Sunshine Pictures
3700 Level 910R fire site
Fire Site in raise
14
Modeling Process

• Airway data from 1977 USBM report
• Scale airways from JW Andrews Sunshine maps
• Temperature data from USBM vent survey Fall,
  1971
• Barometric data from mine elevations

15
Modeling Development

• Initial run 1.6 accuracy with USBM
• Fire location 910 Raise between 3100 and 3700
  level
• Bulkhead failure sequence
• Problem with 3400 booster fan, redo with fan
  curve, 3.3 accuracy

16
Calibrate Model to Fire Events

• Key events
• 1135 am Smoke 3700 level
• 1200 smoke observed at surface
• 1210 smoke leaks into 3100 level

17
Fire Model Events

• Fire sequence model
• 1105 am wood fire 910R, 65m long, ramp up time 5
  minutes to Level 10
• 1110 fire to 100m, 910R Bulkhead partially
  failed, 08 Shop door open
• 1115 fire to 200m, 910R BH further failure
• 1120 fire constant, 910R BH wide open

18
(No Transcript)
19
Ventgraph Output
Smoke at approximate time of discovery 1140 am
20
Ventgraph Output

•  

Local Detail of Fire Area CO Levels at 1140 am
21
Verification

• Model - good correlation to observed smoke
  arrival times
• CO levels correlate well to escaping miner
  behavior
• Results of model give confidence to analyzing
  alternatives
• Run What-If Scenarios

22
What-If Scenarios

• Turn off 3400 Level Booster Fans
• Problem with implementation of task
• What-If Stop fan 1235 pm
• Clears most 3100 roadway in 15 min
• CO still high at 3100L hoist for 80 min
• Did not effect lower mine CO levels
• Would more 1245 cage miners survive?
• Allow rescuers to reach 3100L hoist more quickly?

23
What-If Scenarios

• Jewell Fire Door Closure Sequence
• Supervisors underground made best decision
  close 3700L door.
• Close 3100L door evacuating miners perish,
  lower levels may survive longer
• Best option - leave both doors open and shut down
  3400L booster fans

24
What-If Scenarios

• Opening No. 12 Shaft cover earlier
• Slight improvement in airflow to 4800L 0.9 2.1
  m3/s
• Essentially no difference to any other area of
  the mine

25
What-If Scenarios

• Placement of bulkheads
• Aid fire fighting
• Simulation short term accurate
• Fire can endanger mine infrastructure
• Planning may minimize damage
• Sunshine no combo of booster/doors works
• Stopping needed to be installed inby 910R (and
  fire) to be effective

26
Summary/Conclusions

• Ventgraph validated in Sunshine Scenario
• NVP
• Multiple fans
• Leakage
• Recirculation
• CO levels
• Issues/Problems
• Ventgraph fire Intensity Levels
• Booster fan flow deviation w/USBM model

27
Summary/Conclusions

• 3400 Level booster fan decision
• Contributed to 1977 MSHA ACT
• Valuable modeling/training tool
• Use before and during event

Survivors Tom Wilkinson and Ron Flory
28
NIOSH - OMSHRSpokane
Questions or Comments? CPritchard_at_cdc.gov 509-354-
8021