Caking Processes in Granular NPK Fertilizer Gavin M. Walker,* T. Ronald A. Magee, Clive R. Holland, and Mohammed N. Ahmad Department of Chemical Engineering, The Queen’s University of Belfast, Belfast BT9 5AG, Northern Ireland, U.K. J. Neville Fox, Nicholas A. Moffatt, and Andrew G. Kells Irish Fertilizer Industries Ltd., Herdman Channel Road, Belfast BT3 9AP, Northern Ireland, U.K. This investigation into the caking of granular NPK fertilizer examines three specific areas, namely, accelerated caking tests, the role of unbound water, and the effect of ammonium chloride on caking. The caking propensity of granular fertilizer was characterized by two accelerated caking tests, with the caking propensity being correlated with the tensile properties of the fertilizer granules and the theory of capillary adhesion. The concentration of saturated fertilizer salts in free water found within the granules was determined and gave a good correlation with the caking propensity as determined by the accelerated caking tests. It also indicated that high levels of ammonium and chloride ions were found in the free water within fertilizer granules. X-ray diffraction analysis revealed that high concentrations of ammonium chloride were found on the surface of fertilizer having a high caking propensity. A further investigation using X-ray microanalysis indicated complete conversion of potassium chloride to ammonium chloride within fertilizer granules. Introduction International demand for fertilizers is continually increasing, forcing the fertilizer industry to respond by offering a wide range of high-quality grades. A modern fertilizer plant has stringent quality controls on its products, including NPK and particle size analyses; however, it is the caking propensity of granular fertilizer product which, in many cases, causes greatest concern. Most fertilizer production plants employ a number of features to prevent caking, such as drying, cooling, dust minimization, and product storage in moisture-proof bags. Although these measures significantly reduce fertilizer caking, they do not completely alleviate the problem. This work identifies physical and chemical techniques which can be used to compare caking and noncaking granular fertilizers and to identify factors which contribute to fertilizer caking. The physical analyses used include accelerated fertilizer caking tests; compressive testing of granules, and estimation of the free water content of the fertilizer. Further chemical analysis on the fertilizer utilized the following tech- niques: X-ray diffraction of fertilizer compounds, X-ray microanalysis, and chemical analysis of unbound water. The NPK analysis of the granular compound fertilizer investigated in this work is made up from ammonium nitrate, mono- and diammonium phosphate, and potas- sium chloride. Results and Discussion (i) Accelerated Caking Tests. The use of acceler- ated caking tests to characterize the caking propensity of granular product is widespread within the modern fertilizer industry due to the rapid determination of the caking propensity compared to the bag caking test (Thompson, 1972). This rapid determination means that accelerated caking tests can be used as a quality control analysis on fertilizer production lines and there- fore can be used as an aid in the study of factors that influence fertilizer caking. Fertilizer quality, in this study was characterized using two accelerated caking tests in which pressure is applied to a fertilizer sample for a predetermined time in a cylindrical press resulting in cake formation (Bloom and Sharpe, 1963). In the first accelerated caking test used in this work (ACT(I)), a known weight of fertilizer is subjected to a pressure of 30 psig in the cylindrical press for a predetermined period of time and uses the level of pressure required to shear the resultant cake as an indication of caking propensity. If the sample requires a pressure of >1 psig to shear the cake, the fertilizer is considered to be of poor quality and is liable to “caking” in storage. A fertilizer that is said to be “caking” correlates to product that after 12 months of storage is less than 99% (w.w.) free flowing product after a standard 50 kg bag has been dropped once from waist height. This accelerated caking test has been shown by experience to be an excellent indicator of the future caking propensity of granular compound fertilizer, with virtually no fertilizer having a ACT(I) value <1 psig, caking in storage. However, if the vast majority of fertilizer manufac- tured at a process plant is of significant quality as to be considered “noncaking” (i.e., ACT(I) < 1 psig), the ACT(I) test is of little use for quality control analysis. We have therefore developed a second accelerated caking test, ACT(II), for determining the effect of specific factors on caking. In this case a pressure of 60 psig is applied to the granules for a longer period of time than with the ACT(I) test, with the voidage of the resultant cake, the particle size distribution of the granular sample, and the surface tension of any coating oils used * Author to whom correspondence is addressed. Tel.: (01232) 274184. Fax: (01232) 381753. E-mail: g.walker@qub.ac.uk. 435 Ind. Eng. Chem. Res. 1998, 37, 435-438 S0888-5885(97)00387-4 CCC: $15.00 © 1998 American Chemical Society Published on Web 01/07/1998