Darren Brady Manager OHECC Simtars Department of Mines and Energy

1. Problems With Determining Oxygen Deficiencies in Ratios Used for Assessing Spontaneous CombustionActivity • Darren Brady • Manager OHECC • Simtars • Department of Mines and Energy

3. Spontaneous Combustion Ratios Several ratios commonly used to indicate spontaneous combustion, compare products of oxidation with the amount of oxygen consumed.

4. Spontaneous Combustion Ratios These ratios are used to measure the intensity of any oxidation of the coal that may be occurring. As the coal gets hotter the oxidation reaction becomes more efficient and more of the oxygen is converted to products of oxidation, such as carbon monoxide and carbon dioxide.

5. Spontaneous Combustion Ratios Ratios such as Graham’s, Young’s and Jones-Trickett’s all divide products of combustion by the amount of oxygen consumed to give a quantifiable measure of how much oxygen was used to generate the amount of combustion products measured.

6. Oxygen Deficiency Oxygen deficiency is the term given to the amount of oxygen used (consumed/removed) from the inlet air stream by any activity as it undergoes reactions and interactions with the coal.

7. What Can Go Wrong? • More than one source of oxygen depletion

8. What Can Go Wrong? • Equation used for calculating the oxygen consumed by any oxidation/absorption-adsorption

9. What Can Go Wrong? • The measurement technique

10. What Can Go Wrong? • Instrument inaccuracies

11. What Can Go Wrong? • Unreliable for samples where oxygen deficiencies are less than 0.3%

12. More Than One Source of Oxygen Depletion • If there is more than one source of oxygen depletion then these ratios will be under estimated as it appears that more oxygen was used to produce the products than was really the case

13. Equations Graham’s ratio is often expressed as Where: = Graham’s ratio = final carbon monoxide concentration (%) = final nitrogen concentration (%) = final oxygen concentration (%) Equation 1

14. Equations • Enables calculation without actually knowing what the initial gas concentrations were. • The denominator in Equation 1 is the oxygen deficiency.

15. Initial Oxygen Calculation If initial gas entering an area has a fresh air ratio of 20.95% O2 to 79.02% N2 (20.95/79.02 = 0.265), Equation 2 can be used to calculate the initial O2 concentration by using the amount of N2determined to be present in the sample Equation 2 Where: = initial oxygen concentration (%) = final nitrogen concentration (%)

16. Initial Oxygen Calculation • Based on the assumption that nitrogen, being an inert gas, will not be consumed or created. • Only valid for samples where the initial gas has the same O2 to N2 ratio as fresh air and where N2 and Ar results are combined (79.02%). • Eliminates most problems with dilution because the measured N2 will also been diluted.

17. Oxygen Deficiency The measured oxygen concentration in the sample is then subtracted from the calculated initial oxygen to give the oxygen deficiency Equation 3 Where: = oxygen deficiency (%) = final nitrogen concentration (%) = final oxygen concentration (%)

18. Equations • Problems when the oxygen deficiency is large. • Analysis is done on a percentage volume basis, if O2 is being consumed/removed and nothing replaces it, the nitrogen concentration increases. • The elevated nitrogen concentration results in over estimation of initial oxygen concentration and therefore oxygen deficiency.

19. Equations *calculated assuming initial oxygen 20.95%.

20. Equations If initial gas results are available Graham’s ratio is often calculated using; Equation 4 Where: = Graham’s ratio = final carbon monoxide concentration (%) = initial carbon monoxide concentration (%) = initial oxygen concentration (%) = final oxygen concentration (%)

21. Equations • Used when a tube bundle sampling point located in an intake • Problems with calibration or drift of the oxygen analyser are negated as they are common to both measurements. • Any dilution with seam gas between locations is seen as oxygen deficiency and over estimates oxygen deficiency.

22. Equations Graham’s ratio calculations using Equation 4

23. Equations Where:

24. Equations Where: = Graham’s ratio = final carbon monoxide concentration (%) = initial carbon monoxide concentration (%) = final nitrogen concentration (%) = initial nitrogen concentration (%) = initial oxygen concentration (%) = final oxygen concentration (%)

25. Equations Graham’s ratio calculations using Equation 6

26. Equations • The use of the fresh air N2 concentration of 79.02% includes 0.9% Ar in the amount and is used for techniques that are unable to differentiate the two gases. • If the two are reported separately, the fresh air ratio is 20.95% oxygen to 78.1% nitrogen (20.95/78.1=0.268).

27. Equations • GC analysis determines Ar and N2 separately • Equations 1, 2 and 3 must be modified for GC results

28. Equations • Equation 1 becomes: Equation 7

29. Equations • Equation 2 becomes: Equation 8

30. Equations • Equation 3 becomes: Equation 9

31. Equations

32. Measurement Technique • Tube bundle and real time systems don’t measure N2 • It’s calculated by subtracting the sum of the measured gases from 100. • GC actually measures N2 • Influences which equation must be used

33. Real Time vs Tube Bundle Oxygen Measurements

34. Real Time vs Tube Bundle Oxygen Deficiencies

35. Real Time vs Tube Bundle Graham’s Ratio

36. Fresh Air Oxygen as Measured by Tube Bundle

37. Tube Bundle • Measurement of oxygen using paramagnetic analysers is flow rate dependent so flows from all tubes must be balanced.

38. Tube Bundle • Two locations with the same oxygen concentration could read differently because more resistance in one of the tubes results in a slower flow and subsequently a lower reading than a location with the same concentration but flowing through the instrument at a faster rate.

41. Instrument Inaccuracies • Slight inaccuracies in all measurements no matter how well the analysis is done and how good the instrument performing the analysis is. • These slight variations can cause problems in samples with no significant oxygen deficiency whenever we get a slightly higher O2 (or slightly lower N2 measurement by GC analysis), and apply the known fresh air O2 toN2 ratio to determine the oxygen deficiency.

42. Instrument Inaccuracies

43. Instrument Inaccuracies • It can appear that oxygen has actually been created (very unlikely underground). • Really indicates that the ratio has stayed the same. • Difference comes totally from the acceptable inaccuracies (tolerance) of the measurement technique.

44. Instrument Inaccuracies • Calibration gas suppliers certify each component as the likely concentration within limits eg the O2 concentration in a recently supplied certified calibration gas is 19.6±0.5%. • The true concentration may be as low as 19.1% or as high as 20.1%. • When used to calibrate an instrument to 19.6% any difference will result in all oxygen measurements being high or low, but analytically acceptable. • A change in calibration gas can lead to a step change in values measured by the sensor/instrument calibrated with that gas.

45. Instrument Inaccuracies Oxygen analysers, at best, are accurate to 1% of full scale. Thus a measured value of 20.7% for O2 would be +/- 0.2% A sample measured as: 10 ppm CO, 0.1% CO2, 20.7% O2 and 79.2% N2 (by difference) could thus vary between 20.5% and 20.9% O2 and conversely N2 would be between 79.4 and 79.0%

46. Instrument Inaccuracies Thus Graham’s Ratio would range between: GR =100 x 0.001 / (0.265 x 79.4 - 20.5) = 0.1 / (21.04 - 20.5) = 0.1 / 0.54 = 0.18 and GR = 0.001 x 100 / (0.265 x 79.0 -20.9) = 0.1 / (20.94 - 20.9) = 0.1 / 0.04 = 2.86

47. Oxygen Deficiencies Less Than 0.3% • When oxygen deficiencies are less than 0.3% the variation between readings can significantly affect the calculated ratios.

48. Oxygen Deficiencies Less Than 0.3%

49. Conclusions • Despite the problems, ratios incorporating oxygen deficiencies can still be useful but anyone doing interpretation must be aware of all of these implications.

50. Conclusions • Care must be taken when calculating oxygen deficiencies to ensure that the calculation is correct and representative for the sample and analysis technique.