|
QUESTION Where is the +/- 2 percent error written in terms of the accuracy of the volume and density measurements to determine the tons in a physical inventory? ANSWER The answer you are looking for isn’t in the difference between the found physical number vs. the books. The accuracy/error [i.e. the precision or repeatability, ed. note] is calculated based on the methodology and equipment. Very similar to the +/- in a scale. Both (volume and density) errors, once calculated are combined to come up with a total tonnage error for a physical inventory. These error calculations can be found in the ASTM Standards for density, volume, and tonnage. Both are based on standard deviations of “calculated” confidence intervals. National Map Accuracy Standards actually drive the error in volume to below 2%. The volume error is calculated based on the pile height and it’s mapped contour interval, therefore, the higher the pile with a 1 foot contour interval, the lower the error. The same thing can be said for density, the more tests taken should drive the standard deviation that is calculated lower, the lower the standard deviation and the higher number of tests gives you a lower calculated error. The errors are then combined in a square root of the sum of the squares as depicted in the ASTM standard. Giving you a total %error for tonnage. Here is a sample rough calculation: C = Ö(D2 +A2) Where: C = combined error (percent) D = percent density test error A = percent volumetric survey error Several factors are considered in the volumetric survey error calculation: 1. Aircraft flight height above the ground surface. 2. Focal length of the camera. 3. Stereo compilation instruments used to read and collect the digital data. 4. National Map accuracy standards. 5. Contour interval. (Compiled in 1 ft. intervals.) A series of variables based on the above factors results in a vertical accuracy of approximately 0.50 feet. The percent volumetric survey error is then calculated using the equation: A = (±0.50 feet/ average pile thickness) x 100% The average pile thickness for the Main Pile was calculated to be 51.28 feet. Using these numbers, the percent volumetric survey error is: AInactive = (±0.50 feet/ 51.28 feet) x 100% = ± 0.98% The percent density test error is calculated using the estimated error at the 98% confidence interval and the corresponding average wet density using the following equation: D = (estimated error at 98% confidence interval / average wet density) x 100% The average wet density (linear regression method) for the Main Pile was 67.6 pcf with an estimated error at the 98% confidence interval of 0.7 pcf. Using these values, the percent density test error is: DInactive = (±0.7 pcf / 67.6 pcf) x 100% = ±1.04% With D and A calculated, the combined error (percent) in tonnage for each method is: CInactive = Ö[(±1.04%)2 + (±0.98%)2] = ±1.43% With the combined error calculated for the tested pile, the following error statements can be made in regards to tonnage (wet weight basis) reported: Pile 386,470.0 tons ± 5,526.5 tons As you can see, these are based on statistics and this is used as an example only. The calculations are found in the ASTM standards. Therefore, the higher the piles and lower the density standard deviation (which can be accomplished with exceedingly increasing number of tests or consistent pile construction) the lower the overall PHYSICAL INVENTORY measurement error. |