VERIFICATION AND CALIBRATION OF VENTILATION NETWORK MODELS

1

• VERIFICATION AND CALIBRATION OF VENTILATION
  NETWORK MODELS
• Frank von Glehn, Wynand Marx and Steven Bluhm
• BBE Consulting
• South Africa

2

• Two approaches to calibrate network
• Potential errors
• Some results using 2nd approach
• Real-time network analysis

3

• Network simulators have come a long way from
• Analogue engines
• Card input data on digital computers
• Early progams on personal or desktop computers
• 21st century systems, full graphics, real-time
  data, web based
• Powerful for what-if analyses

4
(No Transcript)
5

• Two main approaches to verify simulation
  networks
• Do full pressure survey and calculate an R or
  modified Atkinson k factor for each branch.
• Use standard k factors and only do spot
  measurements to check predictions. Modify chosen
  k factors until predictions match measurements.

6

• Detailed full-mine pressure surveys are important
  for
• Training in terms of basic ventilation practice,
  measurement techniques, instrument handling
• Understanding the mine on micro scale

7

• First approach
• Advantages
• First-hand knowledge of whole mine
• Individual k factors for each branch
• Disadvantages
• Requires big team to do measurements
• Requires many instruments
• Significant time spent on site

8

• Second approach
• Advantages
• Fewer people, instruments
• Short time
• Use modern PC power for efficient use of time
• Disadvantages
• Not every branch visited
• No direct check of prediction for each branch

9

• First approach measurement accuracy
• Consider 3 m x 3 m airway, 500 m long, average
  density 1.07 kg/m3, flow 37 m³/s, pressure drop
  100 Pa
• Calculated Atkinson k factor 0.01 Ns2/m4
• Accuracy of pressure reading is -20 Pa
• (2mm or 0.079 water gauge), 20
• Accuracy of flow reading is 10

10

• Atkinson equation dP k L C Q² ?
• A3 1.2

Atkinson equation dP k L C Q²
? A3 1.2
?k/k ?P/P 2 . ?Q/Q 20  2 . 10 40
11

• Second approach measurement accuracy
• Consider pressure measurement across door with dP
  3000 Pa (300 mm water gauge)
• Accuracy is still -20 Pa, but error lt 1
• Error in larger flow generally also less
• Define k at a higher level of precision,
• rather than attempting to measure it directly

12

• Examples of application

13

• Coal mine

14
(No Transcript)
15

• Deep gold mine

16
(No Transcript)
17

• Real-time network analysis
• Place instruments all over mine?

18

• Real-time network analysis
• Place instruments all over mine?
• Rather identify a few strategic locations and
  then predict what happens in rest of network
• Use second approach to calibrate network on
  on-going basis

19

• Application in Energy management projects
• Monitor what happens in network when reduce flow
  to mine
• Ensure that no critical conditions reached
• If necessary, flow can be re-set

20
(No Transcript)
21
(No Transcript)
22

• Conclusions
• Evaluate effect of k on the network rather than
  attempting to measure it directly
• Use measurements with low error potential to
  verify network
• Make efficient use of computing power to do
  what-if analyses