Traditionally, calorific value detection has relied on laboratory methods, with a notable lack of rapid techniques for industrial applications. This paper proposes using EDXRF for the quick determination of calorific values, demonstrated through a case study on leaching slag and coke mixtures from zinc smelting. The investigation focuses on detecting calorific values in complex solid materials via X-ray fluorescence, highlighting key factors that influence method stability and accuracy. The results show that (1) Quantitative analysis of calorific values through X-ray fluorescence allows plotting numerical relationships between total spectral count rate, Compton scattering peak intensity (Mo-Ka-C), and Rayleigh scattering peak intensity (Mo-Ka). (2) Factors such as raw material particle size, water content, and instrumental parameters significantly affect detection stability; moisture is particularly critical. As water content increases from dry to 30 wt%, the correlation coefficient between total count and calorific value decreases from 0.8901 to 0.50342. (3) A predictive model correlating Mo–Ka–C intensity with the calorific value of leaching slag mixtures predicts a root mean squared error (RMSEP) of 0.59 MJ/kg and an average relative deviation (ARE) of 4.24 %, indicating improved prediction accuracy for estimating leaching slag's calorific value. The standard deviation (SD) of predicted values was 0.11 MJ/kg, surpassing national standards for repeatability.

中文翻译：

---

基于 EDXRF 的固体材料热值检测方法及影响机理研究


传统上，热值检测依赖于实验室方法，明显缺乏用于工业应用的快速技术。本文建议使用 EDXRF 快速测定热值，通过锌冶炼中浸出炉渣和焦炭混合物的案例研究进行了证明。该研究的重点是通过 X 射线荧光检测复杂固体材料中的热值，突出了影响方法稳定性和准确性的关键因素。结果表明，（1） 通过 X 射线荧光对热值进行定量分析，可以绘制总光谱计数率、康普顿散射峰强度 （Mo-Ka-C） 和瑞利散射峰强度 （Mo-Ka） 之间的数值关系。（2） 原料粒径、含水量、仪器参数等因素对检测稳定性有显著影响;水分尤为关键。随着含水量从干涸增加到 30 wt%，总计数与热值之间的相关系数从 0.8901 降低到 0.50342。（3） 将 Mo-Ka-C 强度与浸出渣混合物的热值相关联的预测模型预测均方根误差 （RMSEP） 为 0.59 MJ/kg，平均相对偏差 （ARE） 为 4.24 %，表明估计浸出渣热值的预测准确性提高。预测值的标准差 （SD） 为 0.11 MJ/kg，超过了重复性的国家标准。