Nanofluids have attracted tremendous attention lately due to their promise as high thermal conductivity liquid and also due the inability of researchers all across the world in explaining the enhancement in the thermal conductivity. Various models and physics have been proposed and some of them have been quite successful in explaining the data, however none of the models in the literature take colloidal chemistry into account. Experimental data, however have shown dependence of thermal conductivity on pH and surface chemistry. In this paper we introduce a model which captures all the anomalies reported in the data 1) Effect of pH 2) effect of aging i.e. time 3) maxima in the thermal conductivity with respect to the diameter of the nanoparticles 4) increase and decrease in the ratio of the thermal conductivity of the nanofluids and the base fluids with increasing temperature. The model is based on the combination of aggregation kinetics with the physics of thermal transport.