CERTIFICATE OF ANALYSIS SARM 77 FERROCHROME SLAG

Size: px
Start display at page:

Download "CERTIFICATE OF ANALYSIS SARM 77 FERROCHROME SLAG"

Transcription

1 SARM 77 Page 1 of 12 CERTIFICATE OF ANALYSIS SARM 77 FERROCHROME SLAG CERTIFIED REFERENCE MATERIAL Prepared by and Distributed by MINTEK P/Bag X3015 Randburg 2125 Republic of South Africa info@mintek.co.za Telephone: fax: STATUS OF CERTIFICATE This is the first issue of the certificate. 2. DATE OF ORIGINAL CERTIFICATION AVAILABILITY OF THE MATERIAL 100 g units of the powdered material are available. 4. SOURCE OF THE MATERIAL The material was supplied by AssmangChrome. 5. DESCRIPTION OF THE MATERIAL Ferro Chrome Slag. 6. INTENDED USE As a control sample in the analysis of samples of a similar type. Verification of analytical methods for analysis Ferrochrome Slag. As a reference material for the calibration of equipment used for analysing similar materials. 7. LEGAL NOTICE This certificate and reference material described in it has been prepared with due diligence. Mintek, or the Manager of the Analytical Science Division accept no liability for any decisions or actions taken following the use of the reference material. 8. STABILITY, TRANSPORTATION AND STORAGE INSTRUCTIONS Care must be taken to avoid undue vibration, since this could cause segregation within the container.

2 SARM 77 Page 2 of INSTRUCTIONS FOR THE CORRECT USE OF THE MATERIAL The material should be well mixed and dried at 110 C over night before subsamples are taken. The minimum sample mass for analysis is 0.2 grams. 10. METHOD OF PREPARATION OF THE REFERENCE MATERIAL Approximately 100 kg of material was reduced in particle size until 95% was less than 60 µm. This was achieved by using a ball mill. The sample was then split using a rotary splitter into 100 gram units. Graph 1 Graph 1: Particle size distribution cumulative distribution Q3(x) / % Sympatec WINDOX Designator Mean value Upper limit Lower limit Comparison curve Nr. 1 Nr particle size / µm 11. STATE OF HOMOGENEITY This reference material consist of 1000 of 100g units. 10 units were analysed FOUR times by XRF powder technique each for Al, Ca, Cr, Fe and Si. Analysis of variance at alpha 0.05 was used to test if there is a significant difference in means between the units. There was no difference except for Cr which passed at The overall relative standard deviation (RSD) expressed as a percent was Cr 1.6%, Fe 1.3%, Ca 1,3 %, Al 4% and Si 3.2% This Reference material has a sufficient homogeneity for its attended use. density distribution q3lg(x)

3 SARM 77 Page 3 of CERTIFIED AND TENTATIVE PROPERTY VALUES INCLUDING CONFIDENCE LIMITS AND INTER-LABORATORY STANDARD DEVIATION. Table 1 Certified and Tentative values Cr 2 O 3 % Ave Fe total Ave SiO2 Ave S Total Tentative 3 Value 95% Lab S N Median Al2O3 Average CaO Ave MgO Ave Notes 3 Relative standard deviation was to high 7.4 %. 4 Normality of the data was not assumed therefore the median was chosen as this is least affected by non normal distribution. The confidence interval was calculated using a Nonparametric two sided confidence intervals for the median of a continuous distribution based on order statistics. The median absolute deviation multiplied by the factor gives an estimate of standard deviation. NOTES: The Certified Value is an estimate of the "true" value based upon the best available data at the time of the certification. The 95% Confidence Interval for the Certified Value is the range of values having a 95% chance of containing the certified value, should the certification program be repeated an infinite number of times. The precision of the user's proposed analytical method must be taken into account when using this reference material. The Standard deviation shown is the inter-laboratory Standard deviation. Ideally a repeat analysis by a laboratory should fall in a range of 2.3 times the standard deviation from the accepted value.

4 SARM 77 Page 4 of VALUES OBTAINED BY INDIVIDUAL LABORATORIES Twelve laboratories submitted analytical results on this material. See 19 for all the results. 14. MEASUREMENT TECHNIQUES USED FOR THE CERTIFICATION Techniques used by the contributing laboratories were the following: Gravimetric analysis Optical emission spectroscopy with source Combustion Analysis Titration Infrared X-ray Fluorescence 15. TREATMENT OF THE NUMERICAL VALUES The data was assumed to belong to a normal distribution after examining the data statistically using tests for Normality, specifically kurtosis which examines the shape of the curve. The difference between the arithmetic mean and median were considered and in no cases was this greater than 0.7 % for certified values and 2.5 % for Tentative values. Large difference could suggest a skewed distribution. Outlier tests (Grubbs and Cochran's) were used to identify points which were far from the mean. These tests were never used twice on the same dataset and conservative approach was used before a result was rejected. Also results were plotted using the box and whisker method and this gives a visual indication if a value belongs to the population, see section 18. Relative standard deviation was considered before assigning a certified value. If the value was high tentative was the assigned. The sum of the values for the certificate is % Since this is a slag, metallic s could be present.

5 SARM 77 Page 5 of Participating Laboratories Mogale Alloys SGS Lakefield Research Africa (Pty) Ltd M&L laboratory services UIS Analytical Services (Pty) Ltd Mintek Set Point Laboratories Hernic Ferrochrome ASA Metals Zimasco (Pvt) Ltd Smancor Xstrata SA Wonderkop Works Tubatsa 17. NAME OF CERTIFYING OFFICERS Mr M. Mtakati, Manager Analytical Services Division, MINTEK. Mr S. A. Marsland, Head SARM PROGRAM, Analytical Service Division, Mintek

6 SARM 77 Page 6 of Data and Distribution KEY CHROME DISTRIBUTION OF RESULTS TIT XRF Titration X Ray Fluorescence Result Rejected 1 Failed Grubbs test Cr 2 O 3 % Ave lab14 k lab5.1 e lab2 b lab3 c TIT lab1 a XRF lab7 g lab4 d lab28 k lab21 j lab18 l lab17 i k e b c a g d k j l i

7 SARM 77 Page 7 of 12 IRON: DISTRIBUTION OF RESULTS KEY TIT XRF Titration X Ray Fluorescence Result Rejected 2 Failed Grubbs test Fe total Ave lab18 i lab7 f lab4 d lab14 g lab17 h lab21 j lab1 a XRF lab5.1 e lab28 l lab2 b lab26 k TIT lab3 c 2 TIT i f d g h j a e l b k 6.33 c

8 SARM 77 Page 8 of 12 SILICA: DISTRIBUTION OF RESULTS KEY XRF GRAV X Ray Fluorescence Gravimetric SiO 2 Ave lab18 j lab 5.1 e lab2 b lab1 a XRF lab3 c GRAV lab26 l GRAV lab21 m lab4 d lab7 g lab17 i lab14 h lab5.2 f GRAV lab28 k

9 SARM 77 Page 9 of 12 KEY SULPHER: DISTRIBUTION OF RESULTS COM IR Combustion Infrared S Total Tentative 3 Value 95% Lab S N Median lab5 e COM lab1 a COM lab3 c COM lab4 d lab14 g lab17 h COM lab26 k IR lab21 j COM lab2 b COM Relative standard deviation to high 7.4 %. 4 Normality of the data was no assumed therefore the median was chosen as this is least affected by non normal distribution. The confidence interval was calculated using a Nonparametric two sided confidence intervals for the median of a continuous distribution based on order statistics. The median absolute deviation multiplied by the factor gives an estimate of standard deviation e a c d g h k j b

10 SARM 77 Page 10 of 12 ALUMINIUM: DISTRIBUTION OF RESULTS XRF X Ray Fluorescence Result Rejected 5 Failed Grubbs test Al 2 O 3 Average lab26 j 5 XRF lab1 a XRF lab5.1 d lab14 f lab3 b lab17 g lab4 c lab21 i lab28 k lab18 h lab7 e i f d g h j a e l b k c

11 SARM 77 Page 11 of 12 CALCIUM: DISTRIBUTION OF RESULTS XRF X Ray Fluorescence CaO Ave lab4 d lab7 f lab18 i lab2 b lab14 g lab5.1 e lab28 j lab1 a XRF lab3 c lab26 k XRF lab17 h d f i b g e j a c k h

12 SARM 77 Page 12 of 12 MAGNESIUM: DISTRIBUTION OF RESULTS XRF X Ray Fluorescence MgO Certified Value 95% Lab S N Ave lab1 a XRF lab5.1 e lab7 f lab17 h lab26 k XRF lab21 l lab14 g lab28 j lab18 i a e f h k l g j i