Errors hydrocracking

The model was optimized with the lumped data from experiments (the unconverted Cie, the mono-branched and multi-branched Cie, and all the cracked products). Then the model was used to predict the carbon number distribution and iso-to-normal ratio for the various carbon numbers. Figures 5a and 5b show the parity plots for the Cie paraffin hydrocracking for both (a) the lumped observations and (b) the carbon number distribution at various conditions (T, P, LHSV and NH3). The agreement between the predicted and experimental data sets for all the high and low yield compounds is good, as all predictions are within the experimental error. The good prediction validates the fundamental nature of the model. 

Because of heat effects of the reactions, the calculated reactor temperature profiles from previous steps would show deviations from actual plant data. We tune the global activity factors again to ensure that the deviations of reactor temperature predictions are within tolerance. We repeat the calibration of reactor temperature profiles and mass yields ofhquid products several times until the errors of model predictions are within the acceptable tolerance. These back-and-forth procedures compose the first phase shown in Figure 6.13 which is a generalized guideline of initial calibration for the Aspen HYSYS Petroleum Refining HCR model. This follows because reactor temjjerature profiles and major liquid product yields are always crucial considerations for any hydrocracker. 

First, Equation 11.55 was solved in order to find the hydrocracking kinetic model parameters a, Oq, a, 8, and Then, by using these values. Equation 11.71 was resolved and the parameters p, and were found. All routines were written in MATLAB software. The criterion of minimization of the sum of square errors (SSE) obtained from the difference of calculated and experimental points was employed to determine all model parameters. More details of numerical solution of Equation 11.55 can be found elsewhere (Elizalde and Ancheyta, 2011). 

The comparison of simulated and experimental sulfur curves is shown in Figure 11.19. Similar to hydrocracking reaction, good agreement is also observed for HDS reaction with global absolute error less than 2%. Other statistical parameters also confirm the good correspondence between experimental and simulated results slope and intercept from parity plot of 1.003 and 1 x 10 ", respectively, and square correlation coefficient of 0.9996. 

---