Delayed Coking Brochure

University of Tulsa Delayed Coking Project (TUDCP)
Phase VI accomplishments

Scope of Work

This page summarizes the accomplishments from Phase VI of the delayed coking project at the University listed by topic

Kinetics in the Coke Drum
Understanding the reaction and kinetics of a system is fundamental to optimizing the system. Coking kinetic studies were thus carried out in a 180-200 g/hr continuous mini-coker facility operated at conditions similar to the industrial delayed coker in order to better understand and characterize the reactions occurring at the industrial scale coker conditions. The products from the thermal reactions in the mini-coker were monitored from 30 minutes to 120 minutes of experiment time.

A four-lump kinetic model was developed from the mini-coker reactor data based on a pseudo-component reaction mechanism. First-order kinetic rate equations were developed for this mechanism based on yields and the material balances then the kinetic rate was expressed in terms of reaction temperature (T), pressure (P), and residence time (tv) for each feed stock. Rate constants and thus activation energies and pre-exponential factors were calculated at 0 & 40 psig. These correlate very well with the feed properties of the feedstock. These correlations have been built into a user-friendly excel model that helps predict the concentration of each pseudo-component as a function of time as well as the total quantity/composition of material in the reactor at each point in time.

This is a predictive model that can give valuable information about the reactivity of resid feedstock in the delayed coker at conditions closely mimicking the industrial delayed coker. It may also be a useful tool for making decisions in the case of an emergency shut in of the coker as well as choosing appropriate conditions to target specific volatile matter content of coke.

Yield model for Recycle
Delayed coker yield models are useful empirical correlations for predicting the output from the coke drum given feed properties and operating conditions. TUDCP’s yield models in the past did not include the effect of recycle on the delayed coking yields. A new delayed coker yield model was built to include the effect of recycle on the delayed coker yields. The pilot unit was retrofitted to allow the continuous recycling of HCGO produced during the run. The HCGO is currently generated at a cut point of 700°F+ at the system operating pressure for each feedstock.

Continuous HCGO recycle shows an increase in coke production and a decrease in liquid as the recycle through put ratio (TPR) increases. A non-linear product yield correlation was developed and validated for coke and liquid as a function of TPR, operating conditions and feedstock properties. The correlation proved to work well within ±5 wt% from actual experimental values. Extra enhancement may be required to account for the effect of higher asphaltene content in feedstock as well as include information for transitional coke making resids.

This model is a closer representation of the industrial delayed coker which has natural recycle albeit at different cut-points.

Fouling Studies
Fouling is another very important aspect of the delayed coking process. In the coker preheaters, fouling can cause increases in heating costs, high pressure drops and lost revenue due to having to shut down units to clean out furnaces. Two units are used to study fouling here at TUDCP- the fouling serperting coil which is a pilot scale furnace and the Hot Liquid Process Simulator (HLPS) a bench scale fouling screening unit. Preliminary fouling correlations were developed for both units as a function of feed properties and operating conditions where applicable.

Significant progress was made in quantitatively identifying the prevailing mechanism at the coking conditions (Asphaltene precipitation versus Coking). This will lay the foundation for a mechanistic fouling model. Studies were conducted using blends of vaccum resid with speciality feeds such as visbroken feed to evaluate the effect on fouling of unconventional coker feeds blended in with vacuum resid. This study opened up a potential new area of research related to the optimization of unconventional blending to minimize fouling. Also, in the pilot unit, gaseous hydrocarbons were found to effectively replace steam velocity without negatively impacting fouling. A unique tube-cutting device was built that allows for the splitting of the serpentine coil tubes so as to allow for in-situ viewing of foulant samples with minimal contamination from the cutting process. EDX and SEM technology was used to analyze foulant samples to identify and or support prevailing fouling mechanisms.

The mechanisms along with the new tools identified in this phase is expected to shed more light on the fundamental understanding of the fouling at coker furnace conditions.

Detailed Sub-Product Analysis
Understanding the nature of the coker liquid product is the key to maximizing the possible profits from the coker since certain liquids are more valuable than others. The sub-products from the coker fractionator are typically sent to downstream processing units; therefore, knowing the properties of these sub-fractions is valuable to ensure that they are fit to be sent to these units. The combined coker product liquid from the pilot unit is distilled in a BR distillation unit to collect heavy naptha, light coker gasoil (LCGO) and heavy coker gasoil (HCGO) fractions. These are then characterized and the yields correlated. PIONA is carried out on the heavy naptha fraction while API, viscosity and sulfur and analyzed on all fractions. The effect of recycle on the sub-product distribution is also studied.

High temperature & low pressure in the pilot coker decreased the naphtha fraction while the HCGO increased. On the other hand, low temperature & high pressure increased naphtha fraction while the HCGO decreased. Correlations were generated to predict the PIONA as a function of temperature, pressure and feedstock properties. The viscosity and API gravity were also correlated as functions of operating conditions and feed properties. The sulfur correlated mostly with feed sulfur but very weakly with run conditions. Recycle seemed to decreases the end-point and the sulfur content in the HCGO fraction. It also affects the MCR and API of the coker liquid product, increases the naphtha and LCGO, and decreases the HCGO.

The detailed liquid sub-product study is very valuable for modelling the delayed coker as a part of a refinery model. The accurate experimental information generated by this project will help further the correct modeling of the coker.

Additive Studies
The liquid product from the delayed coker is its most valuable product. While the effect of operating condition on coker yields is fairly well understood by our group, there is a limit to the operating range in the delayed coker as well as limiting factors such as fouling. Therefore, it would be valuable to identify an additive which could increase the liquid product and/or quality without being detrimental to the rest of the process or prohibitive economically. From this need, the additive study on “hydrogen donors” and “nano-particles” was born. Several pure and/or blends of light hydrocarbons were added into the pilot unit to evaluate their effectiveness as hydrogen donors. Nano-particles were also experimented with successfully in the mini-coker.

Significant amount of coke decreased and liquid increased with H-Donors and nano-particles (5 to 8 wt%). In some cases, the quality of the liquid improved compared to the case without additive. Lighter hydrocarbons effectively replaced steam velocity with equivalent fouling characteristics. A patent application was filed for H-donors to replace steam velocity and increase liquid yields.

At the levels of effects shown by the additives in this study, there could be significant economic benefit if the same effect is seen on the industrial scale.

Foaming Study:
Foaming and foam overs are problems frequently encountered in the delayed coker drum. Inadequate foam control can cause foam overs which contaminate valuable overhead coker products as well as potentially plug overhead lines. TUDCP studies foaming through two unique devices- a gamma densitometer installed around the steel coker drum that gives an indication of foaming by measuring the density throughout the length on the coker every 1.5 minutes and a continuous quartz reactor that gives a visual indication of foaming using an NIR camera. From these units, the effects of feed rate, temperature and pressure recorded in literature have all been observed.

In addition, it has been observed that paraffinic feedstocks foamed more eagerly and aggressively than aromatic feedstock. Also, the tendency of foaming to occur during steam stripping (re-foam) is reduced if denser coke beds are made. The tendency for re-foam was observed to decrease as pressure or temperature increased. A foam model was built and is continually enhanced to include new findings as well as account for AF decay and foam regrowth. The effect of recycle is also incorporated in the model.

Understanding the causes and possible prevention of out-of-control foaming, re-foams, AF effectiveness and decay will help optimize the coke drum to minimize problems caused by foaming.

Time Frame

Phase I - January 1999 to May 2002
Phase II - June 2002 to May 2005
Phase III - June 2005 to May 2008
Phase VI - June 2008 to May 2011
Phase V - June 2011 to May 2014
Phase VI - June 2014 to May 2017


Membership Fee
Industry Members $65,000/year

Industry Leveraging: 20 to 1