November 2017

Process Optimization

Heavy crude oil processing design and reliability

Revamping refineries to process heavier crude slates goes well beyond the requirements to meet equipment performance dictated by a shift in the quantity of lighter product yields to heavier products.

Martin, G., Liu, Z., Wang, S., GTC Technology US LLC

Revamping refineries to process heavier crude slates goes well beyond the requirements to meet equipment performance dictated by a shift in the quantity of lighter product yields to heavier products. This shift in product yield is easily seen by crude oil’s low-API gravity and the slope of its distillation curve. What is not as apparent is the influence on refinery processing units from high-viscosity metals, naphthenic acids, nitrogen (N2), sulfur, solids content and many other contaminants, as well as the difficulty in desalting heavy crudes.

Processing heavy crudes in refineries has caused performance and reliability issues. Understanding design requirements is of primary importance to properly select opportunity crudes, recognize their impact on unit performance and reliability, and realize the necessary unit modifications to overcome critical performance and reliability issues. The processing of heavier crudes increases equipment design and operating costs (e.g., higher-cost metallurgy and higher catalyst replacement costs from increased deactivation), and lowers refinery yields (e.g., higher production of coke from the coker unit). Ultimately, the benefit of processing these types of crudes is a financial decision. Ignoring the necessary equipment upgrades will lead to higher manpower costs in monitoring and troubleshooting operational headaches, higher equipment replacement costs, product yield losses and lost production due to increased downtime. Understanding and meeting the design and operating requirements for the characteristics of the crude slates to be processed will dictate a refiner’s return on investment (ROI).

Crude slates

Numerous production methods are now required in the production of crude oil. The production of heavy crudes mandates more complicated and energy-intensive production methods, and some of these methods can affect crude oil processing in the refinery. Some heavy crudes use cyclic steam injection (e.g., Kern River crude from California), while some unconventional crude production methods use steam to release the bitumen from the underground reservoirs, and require the use of diluents to meet pipeline viscosity and density specifications (e.g., Cold Lake crude from Canada). Long Lake heavy crude oil from Canada is a heavy sour synbit that is composed of steam-assisted gravity drainage (SAGD)-produced bitumen and synthetic crude upgraded from the same bitumen production. Borealis Heavy Blend from Canada is a diluted bitumen (dilbit) comprised of SAGD-produced bitumen and hydrotreated naphtha/conventional diluent. Unconventional methods are required to produce these types of heavier crudes, which typically have high viscosity, total acid number (TAN), sulfur, viscosity, asphaltenes, carbon residue, sulfur and metals values, and other contaminants. Exceptions to heavy crudes include Suncor Synthetic H, a Canadian heavy crude with a low resid, as it is an upgraded bottomless sour synthetic blend.

TABLE 1 provides a comparison of certain properties of heavy crudes to benchmark crudes. The properties tabulated are from past studies and do not necessarily reflect present crude properties, but they do provide an adequate representation of the variation of properties from benchmark crudes. West Texas Intermediate (WTI), Brent and Dubai (Fateh) are the commonly used crudes used to set crude oil prices. The properties listed in TABLE 1 provide an indication of the potentially corrosive/fouling nature, catalyst poisoning, increased coke production (lower refinery liquid yields) and other potentially negative effects on refinery operations that can accompany processing heavier crude slates.

Reliability and performance problems

Operational problems associated with processing heavy crudes have been extensive. Crude oils with higher concentrations of sulfur, N2, chlorides and other contaminates create challenges throughout refinery processing units. The crude unit desalter is commonly the initial point of many problems. Difficulties in adequate desalting have occurred from poor desalter efficiency, problems with stable emulsions, shorting of grids at lower temperatures, undercarry of oil and contaminants in the brine causing problems in the water (H2O) treatment plant, and difficulties in mud wash systems, among others. Improper desalting has led to operational problems throughout several downstream processes.

Atmospheric and vacuum crude units

Severe corrosion in some atmospheric crude column overheads—which is typically the result of a combination of chlorides and sulfur—has caused some units to become unreliable and require replacement after operating for only a few months. Severe fouling has led to the plugging of column internals in the top of the main fractionator and the associated piping circuits, causing reduced throughput, yield losses and unscheduled outages. The unique conditions found in this area of the unit—an acidic environment comprised of hydrochloric acid (HCl), hydrogen sulfide (H2S), H2O and salts—combined with phase changes, can have dramatic effects on the severity of corrosion.

Heavy crudes have a tendency to have higher filterable solids, sometimes higher than 300 lb/Mbbl. Filterable solids usually create stable emulsions, which increase problems with the desalter operations. Laydown of filterable solids in the atmospheric crude charge heater and other heavy oil heaters in downstream units has led to high tube wall temperatures, reducing run length. These solids, along with other corrosion products, can cause a reduction in heat transfer and a high-pressure drop in the crude preheat train exchangers. Due to poor desalting, the increase in sodium (Na) to the downstream units has led to increased coking in the vacuum unit charge heaters, as well as other heavy oil heaters. Heater design issues must be addressed to process a change in the feed slate.1,2 However, the prudent refiner will address these feedstock issues that impact the reliability of heaters processing heavy oil. Much of the fouling in atmospheric crude charge heaters does not come from coke formation, but is rather from a layer made up of filterable solids in the feed and salts. This kind of buildup is why pigging is often more effective than steam/air decoking.

The salts in crude are comprised mainly of Na, magnesium (Mg) and calcium chlorides (CaCl2), with a typical breakdown of 75%, 15% and 10%, respectively. The actual percentages vary by crude source. Sodium chloride (NaCl) is thermally and hydrolytically stable to approximately 800°F (427°C). Mg and CaCl2 are much less stable and will hydrolyze, forming HCl at much lower temperatures. These less-stable salts cause problems with corrosion in the crude unit preflash columns and their overhead systems, and in the top of the atmospheric crude column and its overhead condensing system. Na salts can hydrolyze in the atmospheric crude unit charge heater when oil film temperatures are high enough. However, since the bulk temperature is much less than the temperature at which hydrolysis occurs, only a small percentage of this salt is hydrolyzed.

The difficulties in desalting heavy crude slates has led to significant fouling and corrosion problems in the top of the atmospheric crude column, as well as its associated pumparound and product circuits, where proper designs were not implemented. Various factors also play a role in the potential for fouling and corrosion in this area of the crude unit. Units designed with heavy naphtha draws reduce the temperature in the top of the column. Amines in the crude unit feed from the upstream production site, or from reprocessing slops in the crude unit, will react with the HCl at these lower temperatures to form salts.

Additionally, salts can enter the column with the reflux from the overhead drum if oil/H2O separation is inadequate. Low reflux or top pumparound return temperatures that are below the H2O dewpoint—combined with HCl and salts—will form a very corrosive acidic solution that can lead to corrosion and fouling problems. The more stable Na salt begins to present a higher degree of problems at the vacuum unit and continues through the downstream units. Ammonia (NH3) is commonly injected into the vacuum column overheads to control pH and corrosion. The NH3 reacts with HCl to form ammonia chloride (NH4Cl) salts. It is not uncommon in units processing heavy crudes to have significant laydown of these salts in ejector system precondensers, and in other areas of the unit, if proper designs are not used.

The salt content of produced crudes varies by location. Both heavy and light crudes can have a high salt content. The main difference is the difficulty in desalting heavy crudes compared to light crudes. While producers may treat crudes to control the salt content, this does not mean that all crudes are desalted before shipping, or that the salt content will be constant. Some crude oils received by refineries have been as high as 300 lb/Mbbl. Even if a producer desalts the crude to less than 10 lb/Mbbl, this does not necessarily mean that the salt content will be the same when received at the refinery. An increase in salt content is likely during transportation by sea-going crude oil tankers.

Corrosion is also a problem in oilfield processing facilities, which often treat the crude to reduce corrosion problems or to meet transportation specifications, such as for pipelines. This pretreating of the crude can lead to unexpected contaminants (e.g., amines) in the feed to the refinery. Additionally, the well production process can introduce contaminants, including surfactant chemicals from tertiary recovery and organic chlorides from production chemicals, among others. Since they are unable to be desalted, organic chlorides in crude oil are a problem, requiring higher temperatures to produce HCl and, as a result, rarely impacting the atmospheric crude unit. However, organic chlorides will have an impact on the vacuum column and other downstream processes. Usually, the concentration of organic chlorides is low. Crudes or blended crudes that are high in asphaltenes can cause problems with precipitation in the desalter or crude preheat train and develop a stable desalter rag layer. Crudes high in asphaltenes that are blended with other crudes to provide a lower mixed crude asphaltene content can still result in asphaltene precipitation. High asphaltene content is common with heavy crude slates.

FIG. 1. Corrosion of a vessel tower attachment.
FIG. 1. Corrosion of a vessel tower attachment.

Crudes with a TAN greater than 0.5 mg KOH/g oil can cause problems if the metallurgy is not appropriate. Using the TAN value can prove insufficient, as it is a measure of all naphthenic acids in the crude, and it is known that not all of them are corrosive. This is why two different crudes with the same TAN value may not yield the same naphthenic acid corrosion behavior. Naphthenic acid corrosion occurs in areas where fluid velocity is high and organic acid vapors are concentrated. This type of corrosion primarily occurs above 390°F (199°C), and is not normally found downstream of the atmospheric and vacuum crude unit, since the reactors and heater temperatures in these downstream units are high enough to decompose the naphthenic acids. Although reports have indicated that some crudes have naphthenic acid corrosion in the kerosine fraction, it is not usually found until the diesel and heavier zones. The corrosive nature of crudes with TAN values between 0.5 mg KOH/g and 1.5 mg KOH/g have shown mixed results, but crudes with TAN values above 1.5 mg KOH/g should always be considered corrosive.

Other factors influence corrosion from naphthenic acids, such as the partial pressure of H2S. At high H2S partial pressures, a protective iron sulfide (FeS) film is laid down that impedes naphthenic acid corrosion. However, high velocities or areas of high turbulence will remove this protective layer and allow naphthenic acid corrosion. Under these conditions, with both types of corrosion occurring, the corrosion rate can be accelerated, leading to equipment failure. FIG. 1 shows naphthenic acid corrosion of a 38-in. thick stainless steel tower attachment in a vacuum tower, which appears as if the metal is dissolving. The naphthenic acids react with the metal surfaces and produce hydrocarbon-soluble metal naphthenates. The lack of a corrosion product in the area where corrosion is occurring is a feature of naphthenic acid corrosion.

FIG. 2. Naphthenic acid corrosion of structured packing.
FIG. 2. Naphthenic acid corrosion of structured packing.

FIG. 2 shows the top side of the wash zone packed bed in a vacuum column. A stainless steel with inadequate molybdenum content was used. Outside the areas where the spray from the spray header nozzles was wetting the packing, the top 8 in.–9 in. of structured packing was missing from naphthenic acid corrosion. The spray circles designated by the packing that had not been severely affected by corrosion show a lack of proper liquid distribution.

At low H2S partial pressures, naphthenic acid corrosion can be higher. Increased corrosion may occur if insufficient FeS film is developed for protection. While high H2S partial pressures can impede naphthenic acid corrosion, high velocity and turbulence that remove the protective film will accelerate corrosion. In this case, naphthenic acid and H2S corrosion are competing, and very high corrosion rates can occur. High-temperature sulfur corrosion is a general problem associated with processing units throughout a refinery. A general discussion on sulfur corrosion is provided at the end of this article.

Corrosion and fouling due to crude slate characteristics also influence the efficiency and reliability of distillation equipment. The fouling of distillation equipment has caused reduced fractionation, limitations in product draw rates, reduction in pumparound heat removal, premature flooding and reduced unit run length. An overall process design should be addressed to eliminate these problems, rather than design changes to overcome the symptoms of the problem. However, in certain situations, fouling resistance and efficiency requirements for distillation equipment design can be balanced and optimized for reliable unit performance.3

Fluid catalytic cracking units (FCCUs)

Poor crude desalting, when processing heavy crudes, has led to problems in FCCUs. Some FCCUs have experienced severe fouling in the top of the main fractionator. The salts formed deposits on the trays or packing, causing a high pressure drop, entrainment and flooding problems. The salts have fouled top pumparound circuits, causing reduced heat transfer, high pressure drop, circulation rate limitations, poor response to changes in reflux rate, the inability to control top temperature, premature flooding and increased corrosion. This predicament has led to undercutting gasoline production, loss of control of gasoline endpoint, reduced throughput and increased maintenance and operating costs.

Some refiners have experienced FCCU operating conditions so severe that they have installed electrical desalting for the atmospheric heavy gasoil stream that is fed to the FCCU. This setup is not necessarily the optimum method to solve the problem, but it does show the complications of processing some heavy crudes. Over the years, it has been more common for refiners to install permanent H2O wash systems in the top of the FCCU main fractionator to flush out salts that have precipitated in the top of the column. This online H2O wash of the main column is a means to overcome the salting problem and keep the unit running.

Heavy oil molecules with N2 in their molecular makeup are in all crudes, just at different proportions. At the high temperatures of the FCCU reactor, some of the N2 is converted into NH3 and hydrogen cyanide (HCN). Chlorides in the feed are converted to HCl. At lower temperatures, the NH3 and HCl will react to form salts that can create unexpected fouling problems. This problem occurs in other processing units, as well. Sulfur exists in heavy oil molecules and in all crudes, but at different proportions. Approximately 30%–50% of this sulfur is converted to H2S in the FCCU reactor. This combination of H2S, HCl, HCN, NH3, H2O and salts found in different concentrations, at different temperatures, creates areas where corrosion and fouling can be severe.

Phase changes and interactions between these compounds and the corrosion and salt products they form create a complex set of interactions, with a possibility of severe corrosion or fouling problems. For example, H2S can react with Fe to form FeS, which can produce a protective scale to limit additional corrosion. However, cyanide will react with FeS to form ferrocyanides, which are soluble in condensed water, allowing the protective scale to be removed for additional corrosion. High velocity and turbulence can also remove the scale, exposing the metal surface to further corrosion.

As H2O begins to condense in the main fractionator overhead system, HCl from the vapor stream goes into the solution, with the initial small quantity of H2O forming a very corrosive acidic solution. The HCl also reacts with NH3 in the vapor stream and forms ammonium chloride (NH4Cl), which is very hygroscopic, will react with H2O vapor (steam) and is very corrosive. NH4Cl salts can deposit on metal surfaces, causing severe localized corrosion. As seen on corrosion coupons, these salts will stick to metal surfaces. Removal of the salt from the coupon reveals pitting from corrosion. Laydown of these salts reduces heat transfer in heat exchangers, increases pressure drop and can lead to severe localized corrosion. H2S also reacts with NH3 to form ammonia bisulfide (NH4HS). This creates similar problems to NH4Cl. After adequate condensation by H2O or with a H2O wash system, the salts can be flushed out of the system. The pH of the H2O influences the rate of corrosion. The FeS scale on the metal surface provides a protective film, which hinders further corrosion. However, at higher chloride concentrations or low H2S partial pressures, the film is removed by the chloride ion faster than it can be regenerated.

Other crude contaminates in the FCCU feed can deactivate or poison catalysts, including metals such as vanadium (V) and nickel (Ni). V reacts with zeolite in the catalyst and destroys the crystal structure, which leads to a loss of catalyst activity. Ni, Fe and copper deposit on the catalyst and increase dehydrogenation reactions, thereby increasing the production of H2, the formation and condensation of aromatic molecules and the yield of coke at the expense of valuable liquid product yields. Fresh catalyst must be added to the unit to make up for losses and maintain the catalyst activity. Typical makeup rates are 0.15 lb/bbl of catalyst feed for gasoil units, and 1.5 lb/bbl for units processing resid feeds. Numerous factors influence catalyst makeup rates, and specific unit values vary significantly. Catalyst makeup rates influence refinery profitability through catalyst replacement costs and lower high-value product yields.

The effect of crude metals is complicated, due to different processing configurations such as gasoil FCCUs, gasoil FCCUs with the feed pretreated in gasoil hydrotreaters, atmospheric resid crackers, and FCCUs that are partial resid crackers. Regardless, high-metals crudes influence refinery catalyst costs and, potentially, product yields. Equipment process designs to control gasoil metals quality from atmospheric gasoil, vacuum gasoil and coker gasoil have more of an influence on gasoil FCCUs.4 Metal passivators can be added to the feed to reduce the negative effects of metals deposited on the catalyst, but this leads to higher operating costs. Nitrogen in the feed increases coke production, which lowers conversion and reduces liquid yield. However, this is only a temporary condition of catalyst poisoning, and is without continuous effect if the feed is switched to one with a low nitrogen content.

Hydrotreaters and hydroprocessing units

Processing crudes with increased contaminants, such as sulfur, and poor desalting experienced with heavy crudes has led to an increase in the deactivation of catalyst, as well as higher corrosion rates in downstream hydrotreaters and hydroprocessing units. Nitrogen and sulfur in the feed are converted into NH3 and H2S, respectively. Metallic salts in the feed can deactivate catalysts. Increased Na in the feed due to poor desalting in the crude unit will result in catalyst activity loss. It is now common for residual hydrotreaters to require Na content in the feed to be ≤ 1 ppmw to limit catalyst deactivation.

Nearly all of the sulfur and nitrogen in the hydrocracker’s feed is converted to H2S and NH3 in the reactor. The conversion in hydrotreaters is lower, but still significant. As seen in FCCUs, the combination of H2S, HCl, NH3, their salts and H2O found in different concentrations, at different temperatures, creates areas of the unit where corrosion and fouling can be severe. The sulfur, nitrogen and chloride in the feed, along with the reactor conversion, determine the net amount of H2S and NH3. Downstream from the reactor, at lower temperatures, these contaminates react further to produce ammonium salts that have created chronic problems with corrosion and fouling caused by NH3 salt deposition. Typically, diesel hydrotreaters, gasoil hydrotreaters, hydrocrackers and resid hydroprocessing units have large quantities of H2S and NH3 in the reactor effluent, which forms NH4HS. Downstream of the reactor, wash H2O systems are used to dissolve salts into an aqueous phase before the reactor effluent cools to the salt deposition temperature. This aqueous solution can be very corrosive.

Special areas of concern are sulfide-stress-corrosion cracking (SSC), hydrogen-induced-cracking (HIC), stress-oriented-hydrogen-induced-cracking (SOHIC) and blistering of metals. These concerns are due to high operating pressures, the presence of hydrogen, species such as cyanides (which promote the diffusion of atomic hydrogen into the steel), and other conditions in hydrotreaters and hydroprocessing units.


Poor desalting experienced with heavy crudes increases the Na in coker unit feed. Na increases coking in the coker unit heaters. To avoid coke slagging problems, refiners often limit the Na content in the fluid coker unit’s feed to < 15 ppmw. Heavy crudes tend to be high in asphaltenes and microcarbon residue. This makeup increases the propensity for coking problems in the charge heater and the bottom section of the main fractionator, and lowers liquid yields at the expense of increased coke production.5,6 As with the other units, high-sulfur crudes increase the corrosion rates in the unit.

FIG. 3. A typical two-stage desalter flow configuration.
FIG. 3. A typical two-stage desalter flow configuration.


The desalter is the best initial defense against corrosion, fouling and catalyst deactivation. Proper desalter design and operation will minimize corrosion, help control chemical costs, prolong equipment life, boost production, reduce maintenance and increase catalyst life. No single solution to eliminate these problems exists; however, their minimization begins with the desalter. A typical two-stage desalter flow configuration is shown in FIG. 3.

Desalter performance affects process unit corrosion and fouling problems. Corrosion caused by crude composition is primarily a function of chlorides, sulfur and naphthenic acid content. To understand the causes of downstream corrosion, it is necessary to understand the importance of desalter performance. If downstream corrosion is caused by sulfur or naphthenic acid, then modifications to performance will have a negligible influence on the corrosion problem. The desalter’s primary function is to remove chloride salts and suspended solids. The different available crudes have varying quantities of salts, mud, silt, clay and brine that can be removed in the desalter to more acceptable levels in the desalted crude.

Why do heavy crudes cause problems with desalting? The primary reasons are high viscosity, gravity and thermal conductivity, as well as high H2O, salt, asphaltenes and solids content. Heavy crudes require a well-designed, two- or three-stage desalting to achieve a good salt and basic sediment and H2O removal. An electrostatic desalter uses an electric field to excite droplets of brine so that they collide with other droplets and coalesce. Gravity then separates the H2O droplets from the bulk oil phase. Based on Stoke’s Law, the settling rate can be determined using Eq. 1.

Vs = 73.9 D2 (dw-do)/µ           (1)


Vs = Settling rate, fps
g = Gravity
D = Diameter of droplet, in.
dw = Density of water, lb/ft3
do = Density of oil, lb/ft3
µ = Viscosity of oil, centipoise (cP).

As determined by the equation, higher crude viscosities will make the separation of H2O from the oil more difficult. As viscosity is reduced, the flow resistance of H2O droplets falling through the oil layer is reduced. Higher oil densities will also make the separation of H2O from oil more difficult. Increasing the temperature to lower the oil density, and thereby increasing the difference between the densities of the oil and H2O, provides improved gravity separation. These two properties are a function of temperature, with viscosity dependence on temperature being exponential. The design of the grid system and its efficiency in promoting coalescence of the H2O droplets is important in achieving good separation. Higher oil viscosity reduces the ability for H2O droplets to migrate and coalesce. If H2O droplets do not coalesce effectively into larger droplets, then a lower bulk oil velocity and larger desalter are required, or efficient oil/H2O separation will not be achieved.

This is only the basic equation for settling H2O from the oil; it does not explain all difficulties in desalting heavy oils and properly sizing the desalter. This equation provides the directional effect of crude properties on oil/H2O separation, but it should not be used solely to size the desalter. Chronic problems with forming stable emulsions are seen with many heavy crudes. If the desalter size is based only on considering the effects of gravity and viscosity, then it will be too small to resolve the interface emulsion at a rate faster than it is being created.

Crudes with high asphaltenes can cause desalter problems. Asphaltenes can precipitate in the desalter, where they collect at the oil/H2O interface, stabilizing the emulsion. Rag layer removal headers are sometimes necessary to remove hard-to-break emulsions commonly experienced with heavy crudes. Stable emulsions formed by heavy crudes can limit the desalter’s mix valve pressure drop. The mix valve pressure drop creates shear forces to produce small droplet sizes to enable the proper contacting of oil with H2O. This process enables the salts to be dissolved into the H2O phase. Pressure drop adjustment is necessary to provide adequate salt removal, but an increase to improve mixing must be done without creating problems in breaking the emulsion at the desalter interface. Demulsifying chemicals are used to help break emulsions, but costs will always be a consideration. The grid system, along with the distribution of the oil, must be well-designed to provide good oil/H2O separation. Distribution has sometimes been overlooked in past designs—especially during revamps that reuse existing designs—and can have a big impact on overall desalter performance.

The maximum permissible desalter temperature is limited. The process stream must not contain any vaporization in the desalter, although this is less likely to be an issue with heavy crudes. The maximum temperature is commonly limited by the grid insulator bushings, where excessive temperatures will cause shorting and damage to the grid system.

While processing specific heavy crudes, some refineries have had problems with upsets and amperage runaways. As crude oil conductivity increases with temperature, so does the power requirement for the process. Reaching the necessary voltage gradient with higher-conductivity oil requires a higher electric current and correspondingly higher power consumption. Iranian Heavy crude has a higher thermal conductivity, which limits the temperature at which the desalter can operate. Many heavy crudes have an optimum desalter operating temperature in the range of 280°F–290°F (138°C–143°C), but some, like Iranian Heavy crude, are limited to even lower temperatures due to thermal conductivity limitations or excessive asphaltene precipitation. This can lead to bigger problems with extra-heavy crudes, where the permissible operating temperature is not high enough to provide a reasonable viscosity. Increasing the oil temperature will improve oil/H2O separation, improve H2O coalescing and assist in breaking emulsions, but the temperature limitations of the desalter grid system must be considered.

Good desalter performance requires good-quality H2O with low suspended solids, hardness and pH (preferably near 6). H2O sources used by refiners potentially have NH3 or amines that can react with the naphthenic acids in the crude, forming soaps that generate hard-to-break emulsions in the desalter. Maintaining good-quality H2O is important for all desalters, but when processing heavy crudes with high naphthenic acid content, it can become even more critical to ensure H2O quality.

Some heavy crudes have significant amounts of solids that require additional attention in the design of the mud wash system. The system must be designed to handle the larger-than-normal accumulation of undissolved solids in the bottom of the vessel.

Key parameters for desalting

Key parameters associated with desalting are:

  • Crude salt content
  • Crude viscosity (function of temperature and flexibility in the preheat train is needed to control temperature)
  • Crude filterable solids (poor filterable solids removal can lead to severe equipment fouling)
  • Crude asphaltenes (can precipitate out and lead to a more stable rag layer)
  • Crude naphthenic acid content (can cause soap formation affecting rag layer stability problems)
  • Desalter design/operation:
    • Desalter design (feed distribution, etc.)
    • Desalter size (cross-sectional area at centerline)
    • Grid design and transformer size
    • One-, two- or three-stage desalting
    • Amount and quality of H2O
    • H2O and oil mixing (mix valve pressure drop)
    • Mud washing to remove solids
    • Brine cooling heat exchanger design
    • Chemical treatment.

While the basic principle of desalting is the same for light and heavy crudes, issues with heavy crudes must be addressed. Inadequate desalting can lead to operating problems throughout the refinery’s processing units.

Caustic injection

For years, refiners have injected caustic sodium hydroxide downstream of the desalter to convert chlorides to the more stable inorganic salt NaCl. This process reduces the level of HCl in the crude column overhead system and essentially moves the containment problem to downstream processing units. It also increases the quantity of Na in the system that can affect downstream units. Na is a precursor to coking in the vacuum unit and coker heaters.

Na in downstream unit feeds due to poor crude unit desalting will result in catalyst activity loss. It is common for resid hydrotreaters to require the feed’s Na content to be ≤ 1 ppmw to limit catalyst deactivation. Crude caustic injection should be considered only when the desalter performance is causing severe crude unit corrosion and fouling problems. The impact to downstream units must be considered, as well.

High-temperature sulfur corrosion

High-sulfur crudes can be light- or heavy-API gravity crudes. The heavy crudes add to processing complexity. If no other corrosion processes are occurring, low-temperature sulfur corrosion is generally mild and normally only produces a protective FeS layer. The corrosion rate will be low in these cases, unless high velocity or turbulence results in the removal of the protective corrosion layer.

Most of the sulfur in crude oil is in the form of organic molecules. Not all sulfur compounds are corrosive. The most active are low-molecular weight compounds. H2S is one of the most problematic sulfur compounds in the refinery. Organic sulfur compounds are converted to H2S in processing unit heaters throughout the refinery. H2S corrosion is mainly electrochemical in nature, and can form many different types of FeS (pyrite, troilite, pyrrhotite, etc.). At elevated temperatures, sulfidic corrosion consumes a metal by the reaction between the metal surface and a liquid or gaseous hydrocarbon stream containing sulfur compounds. Sulfidic corrosion mainly occurs at temperatures above 450°F (232°C), and its rate accelerates as the temperature is increased. Above approximately 840°F (449°C), corrosion decreases rapidly, which is believed to be caused by the formation of a protective coke layer. Above 700°F (371°C), H2S can decompose into elemental sulfur.

Corrosion-resistant alloys or corrosion inhibitors are used to control sulfur corrosion. The type of alloy is selected based on sulfur content, temperature and the presence of hydrogen. Corrosion occurs at the metal surface, so metal temperatures should be used to predict corrosion rates. This is important in situations such as the selection of heater tube material, where the peak metal and oil film temperatures can be much higher than the bulk oil temperature.

Heavy crude processing

Processing heavy crudes is common, and so are the associated reliability problems. While some reliability issues can be resolved only by the proper selection of metallurgy, others can be resolved by design. Poor crude unit desalter performance is a prime example of causing reliability problems with potentially all downstream heavy oil processing units. For example, salt formation in the top section of the FCCU main fractionator has caused significant operating problems in some refineries. To enable operation without constant unscheduled outages, some refiners have added vessel height, modified the tower internals design and added a H2O draw tray so that they can periodically slump the tower and H2O wash the top section to remove the salts. This poor design choice does not address the source of the problem that is affecting not only the FCCU, but also other heavy oil processing units. Whether revamping to process heavy crudes or evaluating opportunity crudes to process, it is imperative to understand the impact on unit performance and reliability. HP


  1. Martin, G. R. and T. Barletta, “Vacuum unit fired heater coking—Avoid unscheduled shutdowns,” Petroleum Technology Quarterly, Spring 2001.
  2. Martin, G. R., “Heat-flux imbalances in fired heaters cause operating problems,” Hydrocarbon Processing, May 1998.
  3. Lee, S. H., “Balanced distillation equipment design,” Petroleum Technology Quarterly, Winter 2017.
  4. Golden, S. W. and G. R. Martin, “Revamping vacuum units for HVGO quality and cutpoint,” 1991 NPRA Annual Meeting, March 17–19, San Antonio, Texas, 1991.
  5. Golden, S. W. and G. R. Martin, “Analysis of a delayed coker packed column failure,” AIChE Spring National Meeting, March 19–23, Houston, Texas, 1995.
  6. Golden, S. W., N. Lieberman and G. R. Martin, “Correcting design errors can prevent coking in main fractionators,” Oil and Gas Journal, November 1994.

The Authors

Related Articles

From the Archive