Engineering and Mining Journal - Whether the market is copper, gold, nickel, iron ore, lead/zinc, PGM, diamonds or other commodities, E&MJ takes the lead in projecting trends, following development and reporting on the most efficient operating pr
Issue link: https://emj.epubxp.com/i/102812
BAUXITE sample values were used to determine the measurement uncertainty, which resulted as σmeasurement = 0.855%w. Hence, the analyzer accuracy can be seen in model 5b. Since this verified analyzer accuracy resulted close to the accuracy from the calibration step, and still within the ±0.5%w requirement from the plant operations, the analyzer was regarded as well calibrated. Operating Results To obtain useful results, it's vital that the facility staff properly maintain the equipment. There are many reasons why workers may be reluctant to take proper care of analyzer systems; classic excuses include: "It's difficult to stop the conveyor to take material samples," and "It's difficult to take too many material samples," or "We would waste too much time to recalibrate the analyzer" and "The analyzer is more complex than other instruments." Those issues must be addressed to keep an analyzer operating effectively. Although this application pertains to bauxite ore, most of the information presented here can be applied to different ores or bulk materials. Sidney A.A. Viana is a process control project specialist at Norsk HYDRO ASA's Paragominas, Brazil, bauxite operations. He can be contacted at sidney.viana@vale.com. The analyzer yields its moisture measurements to the plant process control system (PCS) through an analog 4–20 mA signal. Figure 6 shows a trend graph of the moisture content of the bauxite ore over a period of 20 minutes. Such variations are virtually impossible to detect from stop-belt sampling of ore, but are easily detected by an on-line analyzer. Trend graphs, like in Figure 6, are useful to understand short-term variations of the moisture content of the ore. To understand long-term variations, a histogram is best. Figure 7 shows a histogram of moisture for a period of 475 hours (≈ 20 days). The probability of the moisture being lower than 10.5%w was 1.11%; whereas the probability of being in the range 10.5–13.5%w was 95.81%; and the probability of being greater than 13.5%w was 3.08%. The histogram above clearly indicates that the moisture distribution is not Gaussian. Its asymmetrical form was expected because, from the plant process conditions, it is virtually impossible for the ore to have moisture values lower than 9%w, but there are many conditions that can increase the natural moisture content of the ore, such as rain or wash water on the conveyors. The statistics of this distribution are: median λ = 11.18%w, mean µ = 11.37%w, and standard deviation σ = 0.83%w. Before the implantation of the moisture analyzer, the nominal moisture value officially considered by the plant operations was 12%w. However, the median and mean values computed from the analyzer measurements should be more confident. Care is Crucial Microwave moisture analyzer systems can be used advantageously provided that they are designed and calibrated thoroughly. The key issues for a successful calibration of a microwave analyzer are the collection of representative sets of samples, and careful laboratory analysis, to provide consistent sample data for the calibration. www.e-mj.com JANUARY 2013 • E&MJ; 49