Study On The Indicator System Of Activated Carbon For Drinking Water Treatment And The Principle Of Activated Carbon Blending

The technology of ozone-biological activated carbon（O₃-BAC）has become an effective method for advanced drinking water treatment,and it also puts forward higher requirements for the technical indicators and properties of activated carbon for drinking water treatment.In the paper,according to the role of activated carbon as adsorbent and biofilm carrier in drinking water treatment,the composition,structure and property characteristics of activated carbons were determined.Combined with the existing standards and indicators of activated carbon for drinking water treatment,the indicator system of activated carbon for advanced drinking water treatment was researched and constructed.Commercial granular activated carbon（GAC）samples for water treatment from Datong area,Taixi area and Xinjiang area,respectively,were chosen.The pore structure of activated carbon and their blends were characterized by N₂ adsorption desorption isotherm.The surface functional groups of activated carbon were analyzed by Boehm titration method.The adsorption properties of activated carbon and their blends,such as iodine number,methylene blue adsorption,tannic acid number and caramel adsorption,respectively,were analyzed by standard methods.The relationship between adsorption properties and surface chemistry and pore structure of activated carbon was investigated.Hydrodynamic performances of activated carbon and their blends were measured by cyclic experiments of downward adsorption-upward backwashing.The quantitative relationship of adsorption properties between the activated carbons and their blends was investigated by weighted average fitting method,linear fitting method and polynomial fitting method,respectively.The additivity of pore structure parameters of activated carbon blending was analyzed,and the prediction principle of adsorption properties of activated carbon blending was studied.The quantitative relationship of hydrodynamic performances between activated carbons and their blends was investigated by conventional methods,such as weighted average fitting method,linear fitting method and polynomial fitting method,respectively,and BP neural network model.The uniformity and stability of activated carbon blending during backwashing process were studied by iodine number.The main results of the paper are as follows:（1）The indicator system of activated carbon for O₃-BAC water treatment process should include physical and chemical properties such as heavy metal content（zinc,arsenic,chromium and lead）,water soluble matter,water content,p H value,particle size,floating rate,loading density,etc.,adsorption properties such as iodine number,caramel adsorption,adsorption ability for ECs,dynamic water purification capacity,etc.,and filter performances such as strength,hydrodynamic performances（bed expansion rate and bed pressure drop）and biofilm formation ability.（2）The pore structure of typical commercial coal-based activated carbon for drinking water treatment in China mainly develop in micropores with a certain degree of mesopore development.And the content of functional groups on the surface of activated carbon is less.The results of iodine number,methylene blue adsorption,tannic acid number and caramel adsorption of activated carbon show that there is no significant correlation between adsorption properties of activated carbon,with the exception of tannic acid number and caramel adsorption.The adsorption property indicators of activated carbon are mainly determined by pore structure parameters,and have little correlation with surface chemistry.The iodine number is mainly related to the volume of pore size 1.0～2.8 nm of activated carbon,and the linear correlation coefficient is 0.92.The methylene blue adsorption is mainly related to the volume of pore size 1.5～10 nm,and the linear correlation coefficient is 0.94.The tannin acid number is mainly related to the mesoporous volume,and the linear correlation coefficient is 0.91.The caramel adsorption is mainly related to the volume of pore size greater than 3.0 nm,and the linear correlation coefficient is 0.93.（3）The hydrodynamic perfoemance of activated carbon is influenced by the upward-flow superficial velocity.The bed expansion rate of activated carbon increases with the increase of superficial velocity.The slope of bed expansion rate curves increases gradually at a low bed expansion rate.When the bed expansion rate is higher than 10%,the bed expansion rate curves tend to be a straight line.The bed pressure drop of activated carbon increases initially and then decreases with the increase of superficial velocity,and the bed pressure drop there has peak value.Meanwhile,the particle size of activated carbon is also one of the important factors that influence its hydrodynamic performance.Under the same superficial velocity,12×40 mesh activated carbons have much higher bed expansion rate than that of 8×30 mesh activated carbon,and 12×40 mesh activated carbons firstly reach the maximum bed pressure drop.（4）The addition of activated carbon with high adsorption property can improve the adsorption property of activated carbon with low adsorption property.The weighted average fitting method,linear fitting method and polynomial fitting method,which are used to predict the iodine number,methylene blue adsorption,tannin acid number and caramel adsorption,yield mean relative errors of 0.29%～4.01%.The results show that the three fitting methods can effectively predict the adsorption properties indicators of activated carbon blending,and among them the weighted average fitting method is the most convenient fitting method.Meanwhile,the pore structure of activated carbon bleding is similar to that of original activated carbon.The relative errors between the measured values and the fitting values,which are calculated by the weighted average fitting method,are less than 7%.The pore structure characteristics of activated carbon are not changed in the process of blending,and those of activated carbon blending have good additivity.（5）There is a non-linear relationship of hydrodynamic performances between the activated carbon and their blends,which can be accurately predicted by BP neural network model.A three-layer BP neural network model for superficial velocity at 30%bed expansion rate yields mean relative errors of 2.17%,which is much lower than that of 5.53%,4.08%and 4.06%predicted by weighted average fitting method,linear fitting method and polynomial fitting method,respectively.The BP neural network model for maximum bed pressure drop yields mean relative errors of 1.37%,which is much lower than that of 4.31%,4.28%and 4.22%predicted by weighted average fitting method,linear fitting method and polynomial fitting method,respectively.（6）The results of iodine number of samples,which are collected from different heights of activated carbon bed,show that the stable state of activated carbon particles are affected by upward backwashing water,and rearrangement of activated carbon particles in different heights of filter happen during backwashing process.Parts of GACs,which possess better developed pore structure and greater adsorption capacity,are located at upper bed heights.After backwashing 10 min at 30%bed expansion rate,iodine number of activated carbon samples at different bed heights remain unchanged,and the particle rearrangement within the activated carbon bed is basically completed.The weighted average fitting method to predicted iodine number of activated carbon blending at different bed heights yield relative errors of less than 0.99%.Activated carbon blending at different bed heights after water backwashing are mixed by two original GACs through the initial mass ratio.And the water backwashing process almost have no effect on the composition proportion of activated carbon blending.