Error Average

The average error is about 2% for tbe critical temperatures and pressures. The error increases with molecular weight and can reach 5%. 

This method is also applicable to other organic compounds for which one should refer to the original publication. The average error is around 2%. 

The average error with this method is about 20%. 

The accuracy of the conversion depends on the smoothness of the D 86 curve. Errors affect essentially the points in the low % distilled ranges. Average error is on the order of 5°C for conversion of a smooth curve. 

Table 4.5 shows the results for an example. These results differ significantly from those obtained by the method of Riazi for the initial and 10% distilled points. The reported average error for this method is about 3°C, except for the initial point where it reaches 12°C. 

The average error is about 4°C except for the initial and final points where it attains 12°C. I... 

This relation is used only for temperatures greater than 0°C. The average error is about 5 kJ/kg. Figure 4.5 gives the enthalpy for petroleum fractions whose is 11.8 as a function of temperature. For K, factors different from 11.8, a correction identical to that used for Cpi is used (to... 

The average error is about 7 kJ/kg. This model is valid only for ... 

The Mehrotra and ASTM methods apply with acceptable accuracy only for viscosities between 1 and 1000 mPa s. The average error is about 20%. The largest spreads are obtained at low and very high viscosities. 

The average error is about 30%. The relation can be used only if the reduced density is less than 2.5 and the reduced temperature of the mixture is greater than 0.80. 

The average error of this method is about 10%. The method is applicable... 

Usually, Tj is taken as the triple point and T2 as the normal boiling point. The average error is about 5%. 

The average error for these methods is about 10% at 20°C and about 15% at other temperatures. 

The method has an average error of 5% for all mixtures of hydrocarbons whose conductivities of its components are known. 

The effects of pressure are especially sensitive at high temperatures. The analytical expression [4.71] given by the API is limited to reduced temperatures less than 0.8. Its average error is about 5%. 

This method has an average error of 5 kJ/kg except in the critical region where the deviations can be up to 30 kJ/kg. 

This method is applicable for all gas mixtures. The average error being about 3%. 

This method applies only to liquids whose constituents have reduced temperatures less than 1. The average error is about 10% the most important differences are observed in mixtures of components belonging to different chemical families. 

This method is based on the expression proposed by Lee and Kesler in 1975. It applies mainly to light hydrocarbons. The average error is around 2% when the calculated vapor pressure is greater than 0.1 bar. 

This method is applicable to heavy fractions whose boiling point is greater than 200°C. The average error is around 10% for pressures between 0.001 and 10 bar. ... 

The accuracy depends on the fraction distilled it deviates particularly when determining the initial and final boiling points the average error can exceed 10°C. When calculating the ASTM D 86 curve for gasoline, it is better to use the Edmister (1948) relations. The Riazi and Edmister methods lead to very close results when they are applied to ASTM D 86 calculations for products such as gas oils and kerosene. 

The average error is around 30%. This formula applies to pure substances and mixtures. For pure hydrocarbons, it is preferable to refer to solubility charts published by the API if good accuracy is required. 

The model is predictive and uses a method of contributing groups to determine the parameters of interaction with water. It is generally used by simulation programs such as HYSIM or PR02. Nevertheless the accuracy of the model is limited and the average error is about 40%. Use the results with caution. 

Average error is about 5°C. The method should not be used for pour points less than 60°C. 

The average error of this simplified method is about 3°C and can reach 5°C. Table 4.22 shows an application of this method calculating the temperature of hydrate formation of a refinery gas at 14 bar. Table 4.23 gives an example applied to natural gas at 80 bar. Note that the presence of H2S increases the hydrate formation temperature. 

With regard to mass spectrometry, accuracy is not as high with an average error of 2.8 points, but on the other hand, the sample required is very small, being around 2 jl1. 

This expression, which has an average error of 2% or less, is vaUd up to the critical density. It also allows a corrective factor to be added to the critical pressure and temperature so that hydrogen and helium behavior can be accurately predicted (3,21). 

Foi 84 simple inoigaiiic substances and a variety of organic compounds, equation 69 produced an average error of 3.0% (87). The above correlations... 

This equation was empirically derived from 16 polar fluids and has an average error of 2.9%. A technique for estimating surface tension using nonretarded Hamaker constants (89) has also been presented. 

Heats of Vapon a/ion andFusion. A simple linear summation of most of the Lyderson groups (187) has been proposed for heat of vaporization at the normal boiling point and heat of fusion at atmospheric pressure for a wide variety of organic compounds (188). Average errors of 1.2 and 4.3% for group contribution-based estimations of heats of vaporization for selected n- and iso-alkanes, respectively, have been reported (215). 

---