ANALYTICAL METHODS IN BENEFICIATION PROCESSES | |
ArticleName | Random sampling error experimental determination at processing plants |
DOI | 10.17580/or.2017.02.08 |
ArticleAuthor | Kozin V. Z., Komlev A. S. |
ArticleAuthorData | Ural State Mining University, Ekaterinburg, Russia: Kozin V. Z., Head of Chair, Doctor of Engineering Sciences, Professor, gmf.dek@ursmu.ru |
Abstract | The procedure for experimental determination of random sampling errors is described in the GOST standard 14180—80. In compliance with this procedure, from an ore shipment (in accordance with the standard, to be processed per shift) it is necessary to select a series of spot samples with an interval smaller than correlation time, thus making combined samples A and B non-informational with respect to determination of sampling errors. In the limiting case, when selecting a large number of spot samples, difference between samples A and B may be so small, as to exclude a possibility of sampling error calculation, because sampling error, in accordance with the standard comprising both sampling and samples preparation errors, will reflect only samples preparation error. The procedure application experience leads to the conclusion that samples analysis error is greater or equal to sampling error, which is incompliant with sampling conditions at processing plants. Correlation time for mass fraction of analyzed ore component at processing plants is 4–6 hours. In order to provide for sampling intervals exceeding correlation time, sampling is proposed to be done in 24 hours, forming samples A and B from two-hour samples, shifted in twelve-hour interval, and comprised of a number of spot samples, adopted at the plant in question. This permits to implement experimental determination of sampling errors using routine samples at the plant. A rational alternative for determination of sampling errors without forming samples A and B is proposed, completely complying with routine sampling at the plant. It consists in using routine replicate analyses for analysis error calculation, analytic formula of samples preparation scheme and routine sampling results’ variance calculation errors, completely avoiding the necessity for performance of special experiments. |
keywords | Random sampling errors, combined sampling plan, alternative experiments |
References | 1. Kozin V. Z. Minerals raw material sampling. Ekaterinburg: Publ. UGGU, 2011. 316 p. |
Language of full-text | russian |
Full content | Buy |