March 2018

Special Focus: Clean Fuels

Upgrading the bottom of the barrel

In the past few decades, new technologies, including both carbon rejection methods and catalytic conversion methods, have emerged.

Elshout, R., Energy Systems Engineering; Bailey, J., Brown, L., Nick, P., Consultant

In the past few decades, new technologies, including both carbon rejection methods and catalytic conversion methods, have emerged. Often, a combination of these two technologies has proven to be the best approach to upgrading operations. A combination of carbon rejection and the addition of hydrogen (H2) can be used to upgrade residuum fuels to make the low-sulfur gasoline and ultra-low-diesel (ULSD) fuels required by environmental regulations. Upgrading residual fuels uses either carbon rejection or H2 addition, or a combination of the two. Traditional carbon rejection technologies include delayed coking, visbreaking, fluid coking, solvent deasphalting and residual fluid catalytic cracking (RFCC). The key to optimizing refinery margins is often the technology that is used to upgrade the bottom of the barrel. With the advent of new catalysts, other upgrading technologies have remained the same. Relatively new heavy oil upgrading processes, along with improvements to existing technologies, are being used to help increase refining margins.

Some alternatives to carbon rejection are H2 addition technologies, which include ebullated bed hydrocracking andfixed-bed hydrotreating/hydrocracking, among others.

Residuum properties

The properties of the vacuum residuum in the feed are a key factor in determining which technologies will work best. These key properties include Conradson Carbon (Concarbon) residue, heavy metals such as nickel (Ni) and vanadium (V) that poison catalysts, the asphaltene content, sulfur and organic nitrogen. These key properties determine which technologies are feasible and which secondary upgrading technologies can be used to upgrade the feed into marketable fuels.

The composition of residuums varies as follows:

  • Carbon: 83%–87%
  • H2: 10%–14%
  • Nitrogen: 0.1%–2%
  • Oxygen: 0.1%– 1.5%
  • Sulfur: 0.5%–6%
  • Heavy metals (e.g., Ni and V) under 10 ppmw.

The market for both the products produced and the byproducts, such as coke, low-BTU fuel gas, low-sulfur fuel oils, paving or roofing asphalt, etc., is a strong determining factor in the evaluations. The market pays more for higher conversion process technologies. However, with the right set of economics, the incremental values of the products can provide sufficient economic incentives.

Several options of carbon rejection and H2 addition processes are evaluated here, as well as various combinations of these two basic process types. Process schematics, operating conditions, yields, catalyst and chemicals, and economics are presented.

Primary residuum upgrading technologies

The feed for these processes is vacuum residuum from the vacuum unit. It is assumed that a vacuum unit is employed to minimize the quantity of residuum to be upgraded. A light-vacuum gasoil (VGO), a heavy-VGO and a vacuum residuum are produced. The vacuum residuum is fed to one of the primary upgrading units.

Following the primary upgrader, a secondary upgrader that usually co-produces other low-value streams like VGOs will be required. Hydrocracking and FCC are the most commonly used secondary upgrading processes.

The semi-refined products are upgraded into environmentally friendly fuels using conventional technologies and well-known refinery processes that are used for processing the crude. These processes are shared between the conventional section of the refinery and the residual upgrading units to provide economy of scale.

Using block flow diagram simulators is useful for evaluating the cases. These programs contain databases of the feedstocks and models of the individual units, and simulate refined products blending into finished fuels, along with providing utility requirements.


Delayed coking

This technology is the most common residuum upgrading process used in a refinery. The delayed coker is the garbage can of the refinery. It is a brute force thermal process similar to visbreaking. Residuum is heated to approximately 482°C–510°C (900°F–950°F) and enters the main coker fractionator. The bottoms product from this fractionator then enters a coking drum to undergo the coking reactions. At least two coking drums continuously operate in a cyclic, batch-wise fashion. While the coking reactions are taking place in one drum, the other drum is having the solid coke product removed. Once the coke drum is emptied, the operating modes switch. As the flowrate of a delayed coker is increased, the number of coke drums is also increased to enhance the efficiency of the process.

FIG. 1. A schematic of the delayed coking process.
FIG. 1. A schematic of the delayed coking process.

Within the coke drum, a combination of temperature and time thermally decomposes the residuum feed into a broad mixture of materials from gases to naphtha, light and heavy gasoils. These drum products are then fed back to the main fractionator, which separates the products into gas, gasoline, light-coker gasoil and heavy-coker gasoil in a continuous fashion, and recycles the bottom resid product back to the coking drum(s). The coke left behind in the drum is removed by using high-pressure water. This process is referred to as “decoking” the drum. Coke exiting the bottom of the drum usually goes through a crusher car to pulverize it into ½-in chunks, which fall into a sluice and are carried to the dewatering area as water slurry. A schematic of the delayed coking process is shown in FIG. 1.

The solid coke product is sometimes calcined to remove volatile hydrocarbons before being transported to market. Environmental regulations prohibit the direct combustion of coke unless the combustion includes process facilities to recover hydrogen sulfur oxides and other pollutants. These facilities are included in the design of the continuous fluid coking process. One advantage of this process over the batch-wise delayed coking process is the recovery of energy from combusting some of the coke product.

Some delayed cokers process special composition feeds (i.e., highly aromatic), and can produce a more valuable, metallurgical grade of coke that is used to make anodes for the production of aluminum and other chemical products. The properties of the feedstock determine the applicability.

Typically, the gas produced is treated to remove hydrogen sulfide (H2S) and other impurities, then sent into the refinery fuel gas system. The coker naphtha is sent to the hydrotreater and catalytic reformer, either by itself or with other similar naphthas. The light gasoil (e.g., kerosine/diesel boiling range) can go to the hydrocracker for secondary upgrading. Typically, the heavy-coker gasoil goes to the FCCU for secondary upgrading.

The economics of a delayed coker depend on the sale price of the coke and the cost of transporting it to market. Coke is typically 20 wt% of the feed, and coke purchasers generally only pay the transportation cost to their facility.


This technology is very old, and has become useful only in areas where heavy fuel oils are used for power generation and fuel for ships. It is still used in some European countries to produce heavy fuel oil for boilers. Like the delayed coker, the visbreaker uses time and temperature to partially decompose the feed and reduce the viscosity. The primary difference between coking and visbreaking is that coking produces a solid bottoms product and recovers essentially all of the oil product. Visbreaking leaves a significant amount of oil in the bottoms product. For this reason, coking achieves profitable oil recoveries in the range of 55%–65%, whereas the less expensive visbreaking process achieves recoveries of approximately 25%–35%.

The coke deposits on the visbreaker’s heater tube walls and in the soaker drum must be periodically removed. This makes visbreaking a semi-continuous process, meaning that several parallel units are required with one in decoking mode.

The visbreaker produces small amounts of naphtha and gasoil products, as well. The offgas must be amine treated for H2S removal before entering the refinery fuel system. One potential application for visbreaking would be producing a pumpable oil from the tar sands (bitumen) upgrading process. Most tar sands deposits are located near the Arctic Circle. Transporting a syncrude composed of products of a crude, vacuum and visbreaker unit into a less viscous and transportable material has economic potential. Additional diluents may be required for this approach, as asphaltene compounds typically settle out from the bitumen in piping in cold climates.

Fluid coking and resid upgrading

Far fewer fluid cokers are in service than delayed cokers. A northern California refinery recently switched from fluid coking to delayed coking since fluid coke is less valuable, and a more difficult byproduct to handle, than delayed coke. The proprietary resid upgrading technologya utilizes a coke gasifier to turn the fluid coke that is not recycled back to the process into low-Btu gas (about 250 Btu), which cannot be economically mixed with the refinery fuel gas system. Typically, boilers or other dedicated heaters must be designed to use this very lean gas. Part of the resid upgrading operation includes the removal of H2S and sulfur oxides (SOx).

Another option for the gasification product is to further convert it into syngas [one part carbon monoxide (CO) and two parts H2]. Syngas is the starting feed for many petrochemical plant operations, such as an H2 plant.

The common part of the delayed coker or resid upgrading technology is the reactor-coking vessel. The fluid coke recirculates back to the reactor, making more coke by reacting with the feed. Some of the coke is combusted to supply heat for the coking reactions.

Solvent deasphalting (SDA)/demetallization

As previously mentioned, the residuum contains catalyst deactivators, which poisons catalysts in the downstream units. These contaminants include heavy metals such as Ni, V, asphaltenes and other catalyst poisons. Light hydrocarbons, such as propane and butane, are used as preferential solvents to separate the feed into the following semi-refined products using supercritical operating conditions: deasphalted oil (DAO), a resin and an asphaltene stream, which contains most of the contaminants. The DAO is very heavy but easy to crack, and can flow to the hydrocracker for upgrading. Typically, the resins are sent to an FCCU. The asphaltenes can be used as heavy fuel oil or made into roofing or paving asphalt.

RFCC process

This process is similar to conventional FCC, primarily with enhanced regeneration capacity. The feedstock is limited to an upper, maximum level of metals, which will poison the catalyst. The Concarbon number and asphaltene content determine the special requirements of the regenerator section. These carbon-rich compounds form coke on the catalyst, which must be removed during regeneration (burning with air), generating heat. Most of the heat generated is transferred by fluidization back to the fluidized reactor, where it is utilized. Steam generation coils can recover some of this heat, as well.

Due to the higher-than-normal heavy metals, the catalyst must be replaced to keep the equilibrium metals content within a tolerable level. The atmospheric residuum of light crudes, such as West Texas Intermediate (WTI), can be fed to an RFCCU. A simplified process flow diagram of an RFCCU is shown in FIG. 2.b In this scheme, the reactor and regenerator are configured side-by-side. The RFCCU is similar to a gasoil FCCU except that the regenerator must be designed to remove the higher heat loads resulting from the increased coke formation. Typically, steam coils or another type of steam generation system are used to remove heat.

FIG. 2. A schematic of the RFCC process.<sup>b</sup>
FIG. 2. A schematic of the RFCC process.b

The previously mentioned technologies are the most commonly used upgrading processes in the refining industry. In areas where it is environmentally permitted, vacuum residuum can be manufactured directly in the vacuum unit into No. 6 fuel oil for use in boilers and in marine fuel, as well as into paving asphalt. The asphalt can be further processed into roofing asphalt in an air blower, or cut back with solvent for custom road applications.

The semi-refined products from these upgrading technologies must be further processed to make marketable products. FCC and hydrocracking are used individually or in combination. The products produced from the secondary residual upgrading processes are then integrated into the refinery upgrading process units used for the lighter crude/vacuum unit semi-refined products.


Unlike carbon rejection technology, H2 addition uses catalyst, high pressure and an H2-rich atmosphere to break down large molecules into smaller ones, and hydrogenate them. Liquefied petroleum gas (LPG) is in the C3–C4 range, gasoline is in the C5–C9 range, kerosine is in the C9–C15 range, and diesel is in the C15–C20 range. All of these saleable products are blended to the required product specifications for new, low-sulfur gasoline,1 and new ULSD.2 The kerosine range material is used primarily for jet fuel and has less stringent environmental requirements. The key properties are flash point, freeze point and smoke point. Flash point is controlled by the front-end components; freeze point is controlled by the back-end components, and smoke point is controlled by the aromatics content and back-end components.

In H2 addition processes, H2 saturates the ends where carbon-to-carbon bonds are broken by hydrocracking reactions. H2 also converts sulfur to H2S, organic nitrogen to ammonia (NH3), and saturates most olefins and some naphthenes and aromatic compounds. Most of the processes use fixed-bed reactors loaded with different catalysts to perform the required functions. Typically, multiple beds are required to limit increases in exothermic temperature caused by quench H2 injection. Licensors provide their own proprietary catalysts or have catalyst manufacturers produce catalyst to their specifications. Some catalysts are generic and available from multiple catalyst vendors.

The design of the reactor must have sufficient wall thickness to withstand the high pressures. Typically, the base metal is a low-alloy cobalt/molybdenum (Co/Mo) or Ni/Mo carbon steel with sufficient amounts of Mo, Co or Ni to resist high-temperature corrosive attack. The base wall is then overlaid with austenitic stainless steel using a special welding procedure to ensure that the overlay does not disband. Since these heavy-wall alloy reactors require considerable time to fabricate, long lead times are required in the construction schedule.

Most hydroprocessing technologies use recycled H2 to provide sufficient partial pressure for the reactions, while suppressing coking reactions. Makeup H2 is added as required to replace the H2 consumed by the reactions.

As previously mentioned, both NH3 and H2S are produced from side reactions. To prevent the deposition (sublimation) of ammonium bisulfide (NH4HS) and chloride crystals in the downstream equipment, a water wash is included at temperatures where these reactions are predicted to occur. It is recommended that this injection system be designed to ensure that the wash water is adequately and uniformly distributed to keep these potential solid deposits dissolved in the wash water. Ultimately, this spent wash water is recovered in one of the separators and is sent to a sour water stripper. For each mole of NH3 in the reactor effluent, a mole of H2S combines with NH3 to produce a mole of NH4HS. The unreacted H2S is partially absorbed in the spent wash water.

Fixed-bed hydroprocessing unit

A process schematic of a typical fixed-bed hydroprocessing unit is shown in FIG. 3. These units use downflow, fixed-bed reactors to perform hydrotreating and minor hydrocracking. These units are typically a once-through process. The overall conversion can be up to 60%, with sulfur removal up to 90%. Depending on the concentration of heavy metals and Concarbon, the catalyst life can last up to 1 yr before a changeout is required. A common design of this process is the cat feed hydrotreating unit. These units can operate at up to 2,500 psig and at temperatures up to (371°C) 700°F. This process usually uses at least four reactors with different catalyst, such as desulfurization catalyst, organic nitrogen removal catalyst, saturation catalyst and crackability improvement catalyst.

FIG. 3. A schematic of a fixed-bed hydroprocessing unit.
FIG. 3. A schematic of a fixed-bed hydroprocessing unit.


The main fractionator removes the lighter products, including LPG, gasoline, jet fuel and diesel. The higher-boiling components flow to the FCCU. Due to new regulations for low-sulfur gasoline and ULSD, additional processing is required to provide a low-sulfur blendstock that can be blended with other fuel blendstocks.

Ebullated bed technology

In this technology, an ebullated bed consisting of recycled H2, feed and catalyst is circulated through a reactor. Pumps are used to increase reactor circulation through a center internal. An example of this technology is shown in FIG. 4.c Catalyst is continuously added through catalyst-addition ports that are connected to the reactor. The catalyst is continuously withdrawn to provide for a fresh makeup catalyst, as well as to maintain catalyst activity. Flash drums are used to separate vapors from cooled effluent and to feed the fractionation tower.

FIG. 4. A schematic of a proprietary ebullated-bed process technology.<sup>c</sup>
FIG. 4. A schematic of a proprietary ebullated-bed process technology.c


Comparison of technologies

A comparison of the various carbon rejection and H2 addition technologies is shown in TABLE 1; however, the table only discusses the relative merits of these technologies. These upgrading processes must then be integrated into the overall refinery. The three areas that are most affected and require debottlenecking include the sulfur recover units (SRUs), H2 production units and utilities (steam, fuel gas, electric power, cooling water circulation, etc.).


Ultimately, the final upgrading of the semi-refined products into fuels is required, which includes cat reforming, isomerization, fuel blending, hydrocracking and FCC. In the US, heavy fuel oils have nearly disappeared; however, some areas of Eastern Europe still produce fuel oils, such as bunker or No. 6 fuel oil, for power plants and fuel for ships.


A block flow diagram of a fully integrated refinery is shown in FIG. 5. The vacuum unit produces light vacuum gasoil and heavy-vacuum gasoil feedstocks for the hydrocracker and FCCU, respectively. The delayed coker upgrades the vacuum bottoms to produce light-coker gasoil and heavy-coker gasoil for the hydrocracker and FCCU. These units ultimately feed the final upgrading units to produce low-sulfur gasoline, jet fuel and ULSD.

FIG. 5. A block flow diagram of a fully integrated refinery.
FIG. 5. A block flow diagram of a fully integrated refinery.


Economic evaluation

Simple, before-tax payout is the means for screening various processing technologies early in the comparison stage of technology selection. Incremental revenues are divided into the incremental capital cost to determine the number of years to payout for each option. To determine the payout of a processing unit, subtract the unit’s feed costs from the net product revenues to determine the net product revenues. Then, the incremental cost of utilities, labor, catalyst and maintenance are subtracted to determine net revenues.

Conducting the evaluation in this manner avoids getting into the economics of the existing units. If the streams entering the new project perimeter are considered as feedstocks, and the semi-refined products leaving the “new project” area are considered products, then the economics can be determined on the upgrading of the products rather than the final products, which are a blend of existing and new operations. These revenues are based on in-house price values of the upgraded streams. When the streams leave the project perimeter, the economics can be greatly simplified. Using inter-refinery costs for feeds and products allows the evaluation to be made on a net upgrade basis and not on the whole refinery.

An additional upgrading technology

One such technology that is absent from the work is Eni's Slurry Technology (EST). In the ebullated-bed processes, the feed conversion can range from 60%–70%, depending on feed and process parameters, leaving 30%–40% of heavy residue for use as low-value fuel oil or bunker oil. The conversion rate is limited by the stability of the bottom.

A further step toward achieving the total conversion of the bottom of the barrel to middle distillates is the slurry hydrocracking technology (1). The slurry process operates in the presence of a catalyst with sub-micronic particles. The catalyst is more stable than the conventional hydrocracking catalyst, and it can be used in the presence of much heavier feedstocks. After the reaction and the separation sections, the catalyst remains in the residue of the vacuum distillation unit with the nickel (Ni) and vanadium (V) sulfides. Conversion higher than 95% is achieved by recycling the heavier unconverted fraction through the reactor, so that only this fraction remains in the reactor for a longer residence time.

The innovation of the new slurry technology consists of an oleo-soluble molybdenum octoate dissolved in the feedstock. The mixture is fed to the reactor, and H2 is supplied through a distributor located at the reactor bottom. Under these conditions, the catalyst precursor is converted to molybdenite, which is crystalline-layered MoS2, with an average particle size of a few nanometers.

The nano-sized hydrogenation catalysts and an original process scheme allow the complete feedstock conversion to valuable distillates, avoiding the production of residual byproducts, such as petcoke or heavy fuel oil. The heart of the process is a slurry reactor in which the heavy feed is hydrocracked to lighter products in the presence of nano-sized, molybdenum-based catalyst.

The feedstock conversion starts with thermal breakage of the C-C bonds and generation of free radicals that are suddenly quenched via H-uptake reactions, preventing the free radical recombination that could evolve to coke formation. The presence of an active Mo-based catalyst, such as molybdenum sulfide, promotes the H2 activation reaction.
The use of unsupported slurry catalysts is particularly useful in feedstock containing high concentrations of pollutants, particularly metals and asphaltenes. Contrary to the conventional supported catalysts that are utilized in fixed- and ebullated-bed reactors, the dispersed molybdenite is not susceptible to plugging problems due to the metals and coke deposits on the porous supports.

The upgraded oil withdrawn from the slurry reactor is sent to a separation section to recover gas, naphtha, and middle and vacuum distillates. The unconverted material, as well as the dispersed catalyst, are recycled back to the reactor and blended with the fresh feed. The recycle stream is reprocessed to achieve almost total conversion, while a small purge is necessary to eliminate metals buildup.

The process ensures complete metal removal (HDM), excellent Conradson Carbon Residue and sulfur reduction (HDCCR and HDS), and good denitrogenation (HDN). Another peculiar characteristic of the slurry technology process is the production of a high-quality vacuum gasoil (VGO) with low sulfur and aromatic content that can be further converted into transportation fuels or used as new marine bunker.

The first industrial unit started on October 14, 2013 at the Eni’s Refinery in Sannazzaro, demonstrating the possibility of making a correct scale up from the tests carried out on a pilot and demonstration plant. At the moment 3 licenses have been sold to three different oil companies (Total, Sinopec and Zhoushan); the Maoming plant is currently under construction and will be operational in 2020. As described in a previous article, EST is currently the best technology for converting the bottom of the barrel, much better than, for example, the delayed coking technology as it is able to achieve better economic results and lower environmental impacts.

This technology will be detailed in Giacomo Rispoli’s, Executive Vice President Portfolio Management for Eni, keynote address at IRPC Europe, which will be held in Milan, Italy in June.

(1) Refers to EST (Eni Slurry Technology)
(2) Sunset Boulevard of Delayed Coking


Bottom-of-the-barrel upgrading costs can run in the billions of dollars. Past evaluations have shown that the higher cost of the H2 addition technologies, when combined with carbon rejection technologies, can show robust payouts. This is primarily due to the increased volumes of more valuable products that are produced by the hydroprocessing processes. Carbon rejection technologies generally decrease the volume of valuable products, since carbon rejection reduces the net volume of the final products.


A detailed comparison of upgrading technologies is summarized in TABLE 2. A complete evaluation is needed to compare the various technologies. The necessary steps to select the best upgrading technology include:

  • Developing yields of the individual process technologies in-house, or by using a licensor.
  • Combining the upgrading technologies into the existing refinery using a block flow diagram simulator.
  • Determining the debottlenecking requirements of the existing units.
  • Determining the incremental modifications to offsite units, which can often be more than 50% of the onsite units.
  • Determining the operating costs, including utilities, catalyst and chemicals, maintenance, operations and labor.
  • Evaluating whether these modifications can be performed during various phases.
  • Carrying out evaluations that can usually be completed on several levels, including front-end loading (FEL)-0 (block flow diagrams and using cost vs. capacity curves for economics); FEL-1, which requires enough process engineering to obtain equipment pricing and conduct a factor estimate; and takeoff grade-type estimates using cost estimating software.
  • Conducting blending calculations, sulfur balances and H2 balances to discover the effect on the support units and utilities.
  • Ranking the various options and determining which technologies best meet the refinery’s capital cost money available. HP


  1. Elshout, R. V. and C. S. Bains, “USA motor fuels sulfur levels: Moving up a tier—Part 1:Gasoline sulfur,” Hydrocarbon Engineering, 2005.
  2. Elshout, R. V. and C. S. Bains, “Moving Up a Tier—Part 2:Ultra-low sulfur diesel,“ Hydrocarbon Engineering, 2005.


     a Flexicoking is a proprietary resid upgrading technology from ExxonMobil.
     b Refers to Honeywell UOP’s RFCC process.
     c Refers to CB&I’s LC-Fining technology.

The Authors

Related Articles

From the Archive