OCR Text |
Show 10 Jordan Loftus SeD. Texaco Inc - August 24. 1989 where Nf equals one for firing from one side of the tube rows and equals two for firing on both sides. The reader is invited to compare the results using the above equations with Figure C-2 of the API- RP530 for black tubes. The local longitudinal to average flux, Fr and the convective flux, 9c are initially estimated as input by the user. Fortunately in the case of refinery heaters, the backward flux from the sink to the gas emitters are small in comparison with the forward flux. Thus the error in computing the flux profile is relatively insensitive to the sink temperature. The need to rerun is based on experience. Note however that these outside wall temperatures should not be used to evaluate the tube metallurgy and thickness. RP530 is used instead. D. Burners A simple burner model is used. parameters: Vendor data are used to set 1. the flame length, Fr. in feet Eq. 8. where F is a multiplier factor, Q is heat release in HH Btu/H and A and B are the best fit of vendors data (Figure 6). 2. the effective pressure drop coefficient, K, based on the minimum throat diameter: K - SP / 0.003 p V2 Eq. 9. where p and V are the air density and velocity in the throat and P, the pressure drop in inches of water. The burner shown in Figure 6 has K value of 2.0. Input data are K, F, A, B, air cone half angle, throat diameter or velocity. As oxygen to the burner approaches stoichiometric, the F factor increases hyperbolically. This is important when substituting LONOX burners for older burners. Heat release is taken as a linear function of flame length. Eq. 10. where Hr is the heat release up to the Z plane, Qro is the fuel burnt prior to entering the chamber, Qr is the total heat release by the burner,and Z is the flame path distance. |