Steam Cracking Coke Properties and Their Influence on Furnace Run Length Predictions: Experimental and Modeling Study

Abstract

Although coke formation is one of the most important problems that the steam cracking process still faces, the properties of the formed coke have hardly been studied over the past decade. For some of them, orders of magnitude differences have been reported. Therefore, state-of-the-art measurement techniques have been used on industrially relevant samples obtained at laboratory and industrial scale for ethane and naphtha cracking. The feedstock type does play an important role, reflected in, for example, factor 2 differences in BET surface areas for ethane and naphtha coke samples. Molar C/H ratios from 26 up to 59 are obtained by CHNS/O analysis, depending on the age of the coke and the proximity of the metal coil. The measured thermal conductivity values for ethane and naphtha coke, that is, 9.6 and 8.7 W m −1 K −1 at 1273 K, respectively, are approximately twice and three times the values reported by Dente and co-workers. A sensitivity analysis on the influence of these properties on the predicted run length of an industrial furnace gives huge differences, that is, from 2 weeks up to 212 days for naphtha furnace operation and 20 up to 227 days for ethane operation. This clearly indicates the importance and the need for more accurate assessment of coke properties to develop more predictive models for run length simulations.

Combined Catalytic and Pyrolytic Coking Model for Steam Cracking of Hydrocarbons

“…59,60 The heat flux from the furnace to the radiant tubes is required to perform the industrial furnace simulation, which acts as a set of boundary conditions for the simulation. The incident heat flux profile obtained by using the 1D furnace model 61,62 is implemented in this case. The framework of COILSIM1D consists of two parts: a microkinetic model and a reactor model.…”

“…These two severity indices unambiguously determine the product yields of the given feed. , The heat flux from the furnace to the radiant tubes is required to perform the industrial furnace simulation, which acts as a set of boundary conditions for the simulation. The incident heat flux profile obtained by using the 1D furnace model , is implemented in this case.…”

“…The density of coke deposition is taken as ρ c = 1.6 × 10 3 kg m –3 which is the same as the value used in the model development . The time step used in this simulation is 50 h. The thermal conductivity of coke deposition is taken as 8.6 W m –1 K –1 , while 24 ppm (wt %) DMDS is continuously added. The predicted run length with the current coking model is 2050 h or 85 days, which agrees with the observed run length.…”

Coke deposition and run length in industrial naphtha thermal cracking furnaces via a quasi‐steady state coupled CFD model

“…Although the measurement of coke properties is outside the scope of the current study and the values provided in the literature are used, an accurate assessment of coke properties is important due to its effect on run length. Geerts et al [ 6 ] analyzed the sensitivity of run length to the coke properties and stressed the importance of an accurate value of the coke thermal conductivity. They also showed that thermal conductivity of coke increased as the temperature increased, in which thermal conductivity of coke was measured as 8.1 and 8.6 W/(m · K) for two different times on‐stream of a naphtha cracker.…”

“…[ 1 ] Similarly, this was also reported in a recent review paper by Symoens et al [ 5 ] which discussed some aspects of coke formation and deposition on the coils. The thermal conductivity of coke was further illustrated in the work of Geerts et al [ 6 ] They report a huge differences in run length of naphtha cracking furnaces, for example, from 2 to 30 weeks, when coke thermal conductivity increases from 4.5 to 14 W/(m · K), respectively.…”