AFPM Q&A—2: Refiners discuss gasoline processes

Oct. 7, 2019
During the 2018 American Fuel and Petrochemical Manufacturers Operations & Process Technology Summit, refiners addressed questions about gasoline processes selected and answered by industry experts from refining companies and other technology specialists.

During the 2018 American Fuel and Petrochemical Manufacturers Operations & Process Technology Summit (formerly Q&A and Technology Forum), Oct. 1-3, 2018, in Atlanta, Ga., US domestic and international refiners addressed questions about gasoline processes selected and answered by industry experts from refining companies and other technology specialists (see accompanying box).

This annual meeting addresses real problems and issues refiners face in their plants and provides an opportunity for members to sort through potential solutions in a discussion with panelists and other attendees.

This is the second of three installments based on edited responses in the 2018 official answer book. Part 1 in the series (OGJ, Sept. 2, 2019, p. 66) highlighted discussion surrounding hydroprocessing operations. The final installment (OGJ, Nov. 4, 2019) will focus on fluid catalytic cracking (FCC).

The only disclaimer for respondents was that they discuss their own experiences, their own views, and the views of their companies. What has worked for them in their plants or refineries might not be applicable to every situation, but it can provide sound guidance for what would work to address specific issues. 

The respondents

Lance Tallman, Citgo Petroleum Corp.
Kurt Detrick, Honeywell UOP LLC
Romain Lemoine, McDermott International Inc.
Greg Troutman, E.I. du Pont de Nemours and Co. (DuPont)
Shane Presley, DuPont
Gayl Mercado, Axens North America Inc.
Christian Arnoux, Valero Energy Corp.
Steve Philoon, Honeywell UOP
Richard Todd, Norton Engineering Inc.
Russ Wiltse, Honeywell UOP

What metallurgy works well and doesn’t work well in alkylation units? In what applications does the alternate metallurgy perform better-worse? 

Tallman: The standard metallurgy for use throughout hydrofluoric (HF) acid alkylation units is carbon steel (CS). Showing a high degree of corrosion resistance in both low- and very high-acid concentrations, CS is also useful for its relatively low cost compared to more exotic materials. In most applications throughout the unit, CS will display a general corrosion loss rather than the more-difficult-to-detect pitting or localized corrosion. When used in a stream that will contain HF acid, however, there are several restrictions placed on the steel composition (particularly chromium, nickel, and copper) to try and reduce the unique circumstances that can set up localized areas of accelerated HF corrosion. Despite the possibility of accelerated corrosion, CS has a long history of suitable use throughout US alkylation units and performs well in most streams present in these units. 

There are some streams that, through use and time, have required an alternate metallurgy to the standard approach of CS. These areas typically involve a higher temperature than are encountered in most of the unit. Most refiners use CS for HF-containing streams at operating temperatures below 150-200° F. At temperatures higher than this, or in areas in which previous experience with CS has provided unsuitable corrosion resistance, consideration is made for using a high-nickel alloy. The common choices for use vary between Monel 400 and various Hastelloy grades (C-276, B grades, etc.), with Monel used the most. These high-nickel materials provide more suitable corrosion resistance at the higher-temperature operation than is possible with CS. Monel 400 brings about its own set of challenges with the possibilities of cracking due to high residual stresses, particularly in services that contain oxygen or mercury contamination. A stress-relieving heat treatment is usually performed at initial fabrication to reduce the possibility of this catastrophic failure mechanism.

Most notable of the common materials used in refining that shouldn’t be used in HF alkylation units are stainless steels. While they can sufficiently resist corrosion in anhydrous environments containing HF acid, they perform worse than CS under aqueous conditions. All grades of stainless steel can undergo rapid corrosion losses (sometimes on the order of 100 mils/year, mpy) at certain conditions and should therefore be avoided. Other materials to avoid are those containing silicon, which will react with HF acid and form a highly corrosive fluorosilicic acid.

CS is used for most of the equipment and piping in sulfuric-acid alkylation units, including the reaction zone (contactors, acid settlers). CS doesn’t work well, however, in the presence of sulfuric acid with high temperatures or high velocities. The fresh- and spent-acid piping is typically constructed of 300 series SS, and-or Alloy 20; however, some units have successfully operated with CS piping in this service. Additionally, in the effluent-treating section and depropanizer-feed treatment system, the static mixers and upstream-downstream piping are typically upgraded to Alloy 20 and-or Hastelloy C. Although the contactors (shell and tube bundle) are constructed of CS, due to the highly turbulent service, internals such as the reactor impeller and distributors are upgraded to Hastelloy C, which greatly extends life of this equipment.

Monel is not recommended for material of construction in sulfuric-acid alkylation units, as it isn’t compatible with equipment and piping containing large amounts of sulfuric acid.

Detrick: API RP 751 has a good description of the experience with different materials in the HF alkylation unit in Section 3 and Annex D. 

CS is the most commonly used material in HF alkylation units in areas of relatively low-temperature and low-water content in the acid. The surface of fresh steel reacts with HF to form iron fluoride, but under the proper conditions (primarily low-water concentration in the acid), this layer of iron fluoride protects the underlying steel from further attack.

Monel is used in areas of the HF alkylation unit where temperatures are higher, or water content of the acid is higher (such as in the acid regenerator or rerun tower). Hastelloy (primarily C276) has also been used successfully in these areas of elevated temperature and-or water concentration.

316 Stainless steel vessels are used for transporting fresh anhydrous HF to the alkylation unit, but stainless steel should only be used for high-purity fresh acid. In commercial experience, stainless steels (both austenitic and martensitic) are aggressively attacked by the plant acid in an HF alkylation unit, and these stainless steels should never be used in the unit itself.

Lemoine: In less than 7 years, McDermott’s Chevron Lummus Global has successfully commercialized and licensed an advanced version of sulfuric-acid alkylation technology called CDAlky to clients worldwide. CDAlky technology focuses on effectively eliminating the root cause of drawbacks inherent to conventional alkylation units. A leading area of concern is excessive corrosion and the need for higher-grade metallurgy. 

Lummus’ CDAlky technology not only operates at lower temperature than conventional sulfuric-acid alkylation processes, but it also eliminates the need for a reactor effluent post-treatment section. The key to these technical breakthroughs resides in CDAlky reactor proprietary internals: AlkyPak and distributor plates. By eliminating caustic and water introduction in the alkylation section, and by maintaining a low-temperature operation, CDAlky technology can eliminate the need for higher-grade metallurgy. Recent turnaround inspections conducted on multiple CDAlky units have confirmed these technical benefits:

  • Low-temperature operation greatly reduces corrosion rates.
  • Neither reactor effluent wash nor post-treatment steps are required to remove sulfuric acid or sulfates from the alkylate.
  • No fouling material is observed in any fractionation reboilers.
  • Stainless steels and CS are suitable materials of construction.

Troutman: Although careful consideration is required when selecting appropriate metallurgy for any sulfuric-acid alkylation unit, a sound process design relegates the recommended use of exotic material to only a few areas. CS is by far the most common material found throughout the unit, as it’s both economical and capable of handling the range of temperatures and velocities observed in normal conditions as well as most alkylation-unit excursions. The typical hierarchy of metals found in a sulfuric-acid alkylation unit is as follows: CS > 316L Stainless Steel > Alloy 20 (in certain situations) > Alloy C276.

With respect to sulfuric acid, CS can be used if the velocity-turbulence through the piping or equipment isn’t high enough to disturb the protective passive layer of iron sulfate that forms on CS when contacted with sulfuric acid. Many refiners have utilized CS with a corrosion allowance of either 1/8-in. or 1/4-in. for normally flowing and normally not-flowing lines, respectively. In piping where high-velocity turbulence cannot be avoided (such as through valves), Alloy 20 or polytetrafluoroethylene (PTFE)-lined valving should be used to avoid excessive corrosion-erosion. When velocities high enough to avoid hydrogen grooving and low enough to avoid eroding the passive layer aren’t obtainable with sulfuric-acid piping, 316L stainless steel is a good substitute.

Areas where neutralization takes place will require at least Alloy 20, especially since most neutralization takes place in a static mixer where there’s high turbulence. While Alloy 20 is an excellent material in many applications, situations such as high temperature and certain contaminants or environments can cause it to activate and fail quickly.

For the reactor itself, CS is the preferred choice for the bulk components (shell, heads, tube bundle, etc.). CS is economical where process conditions (velocity, turbulence, and temperature) are typically mild. Since CS within the reactor forms a protective passive layer, a long life is typical. The reactor impeller, wear ring, feed nozzles, and other high velocity-turbulent areas can be constructed of Alloy C276 to dramatically improve the life span.

DuPont never recommends using Alloy 20 within a sulfuric-acid alkylation reactor due to process conditions that can cause the metal to become active and fail. In addition, areas where neutralization takes place with elevated temperatures (such as the mixing of sulfuric acid and hot alkaline water) are better suited for Alloy C276 rather than Alloy 20.

A few additional considerations:

  • For areas with high amounts of propane (such as the depropanizer overhead), depressurization could result in temperatures cold enough to require low-temperature CS.
  • CS is suitable for caustic environments (typically in the 10-12% range), but post-weld heat treating should be used in areas where temperatures are expected to be higher than 180° F.
  • Both stainless steel and Alloy 20 shouldn’t be used when large amounts of halide ions (such as Cl-) are present, as these may cause pitting corrosion.
  • Water isn’t corrosive by itself, but it can cause major corrosion issues if there is an area with a high probability of contacting trace sulfur dioxide (SO2) or sulfuric acid.

As for metallurgies that aren’t compatible in sulfuric-acid alkylation units, Monel isn’t compatible with equipment and piping containing large amounts of sulfuric acid. Monel has been shown, however, to be effective in hydrocarbon streams with weak acid. This is a benefit when considering HF-to-sulfuric acid alkylation conversion, as Monel is prevalent in HF alkylation units.

Nonmetallics such as PTFE, polypropylene, and glass have excellent corrosion resistance and can be considered for various portions of the plant. Fire resistance and physical strength also must be considered.

Corrosion is a complicated phenomenon and isn’t always predictable—or even explainable—by current metallurgical knowledge. DuPont continues to experiment and gain knowhow on the performance of specific materials in sulfuric-acid alkylation units. For the latest information, DuPont and-or a metallurgist should be involved in corrosion evaluations and upgrade decisions.

What type of release mitigation safety systems do you use for sulfuric acid and HF alkylation units? 

Tallman: For HF alylation units, Citgo employs the following: 

  • Rapid acid-deinventory systems to move HF acid from unit to remote acid storage drums (acid settler, reactor risers and coolers (includes unit shutdown), acid storage drum).
  • Water curtain surrounding high-acid area of unit (with autoactivation).
  • Water deluges on acid settler, remote acid storage drums, and acid-service pumps.
  • Elevated, remotely operable fire monitors with independent supplemental water supply.
  • Camera systems to improve response time.
  • Point source hydrocarbon detectors.
  • Point source HF detectors.
  • Bilevel perimeter laser HF detectors.
  • Remote capabilities to start, stop, isolate vent-to-flare and deluge acid-service pumps.
  • Remote control panel for acid truck unloading (isolation valves).
  • Closed-circuit monitoring of unit.

Sulfuric-acid alkylation units don’t require the same robust safety systems as HF alkylation units since sulfuric acid is a liquid at atmospheric conditions. The release-mitigation safety systems typically used are curbed and-or diked areas to contain the liquid sulfuric acid in event of a leak. This exists in the reaction section, blowdown section, and fresh-spent acid tanks. The same mitigation systems are in place in the effluent-treating section and the depropanizer feed-treating system where caustic is present. When operators respond to any release, they’re equipped with the appropriate personal protective equipment, which include slicker suit, face shield, and proper gloves. Hydrocarbon detectors also are present in and around the unit due to the presence of LPGs.

Presley: Sulfuric acid is a liquid at atmospheric conditions and will pool if released rather than form a vapor cloud. During the summer of 1991 an independent consultant performed a series of large-scale sulfuric acid leak tests. In all, thirty-six release tests were conducted, three tests utilizing alkylation equilibrium acid from an operating commercial alkylation unit and thirty-three tests using an acid-hydrocarbon emulsion. The tests involving equilibrium acid resulted in essentially 100% recovery of the acid released; the remaining tests resulted in an average 97.6% recovery of the acid released.1 

Because sulfuric acid will pool when released rather than forming a toxic vapor cloud, only basic mitigations are required. These mitigations typically include curbed and-or diked areas around sections of the unit containing sulfuric acid to contain the acid in the event of a leak. These areas typically include the reaction section, acid-blowdown section, and fresh-spent acid storage areas. In the event of small or minor leaks, sulfuric acid can be neutralized with soda ash before removal. Larger spills may require physical removal of the acid. Other release mitigation systems in place are not specific for sulfuric acid and are typical for other processing units containing LPG hydrocarbons.

How do you monitor and protect the heater tubes from overheating in high-temperature services such as catalytic reformer heaters? How is the tube-wall temperature monitored? 

Mercado: As a first step and low-investment solution, refiners will perform a visual check on the color of tubes to check for color differences or perform a thermal scan of the tubes. These solutions aren’t very accurate due to iron-scale formation and ceramic coating on the tubes. A simple visual check of tube color or thermal scan, however, can offer refiners a base line or trend for monitoring heater-tube temperatures.  

Tube-skin thermocouples can also be installed either during design or a turnaround. On reformer heaters, tube-skin thermocouples aren’t standard during design due to the large number of thermocouples that would be required for proper measurement. Additionally, these tube-skin temperatures may not be very reliable due to the thermocouples being burned out and-or the formation of coke on the probe.

Advanced tools with everyday capabilities are becoming more prevalent in the industry and will move the safety and reliability of the fired heaters to improved levels. Using data transfer and gathering improvements, software systems are beginning to replace the historical approaches that the industry has been using for some time. Axens’ Connect’In digital tool, which allows users to remotely monitor and analyze catalyst and unit performance, also allow engineers and operators to track key performance indicators on the heaters and detect possible issues concerning the equipment. The Connect’In software architecture allows users to automatically schedule data gathering, validate the data, and calculate the heater yield and efficiency. Within a single tool, users can analyze data trends, calculate remaining tube life, track instrumentation problems, and help diagnose issues.

Axens recommends the below maintenance approaches to minimize tubes from overheating:

  • Proper maintenance of heater burners to ensure normal flame patterns and help protect heater tubes from overheating. If overheating is suspected, some burners can be turned off or on depending on the tube-skin temperature patterns.
  • Reduction of the heat load on the limited heater, shifting the required heat load to other heaters that aren’t limited in heater-tube temperature. This, however, will result in a non-flat temperature profile.
  • Reducing excess oxygen to lower and protect the heater tubes from overheating.

Arnoux: Infrared (IR) scans are commonly used to monitor tube-skin temperature. Rust and scale on the tubes elevate the tube-skin temperature as registered on an IR scan. Ceramic-coating tubes during turnarounds may reduce the rust and scale on tubes. Be aware of concentrated corrosion-erosion at tube-weld attachments. The licensor can calculate the tube-skin temperature from process data. 

Philoon: UOP recommends using a calculation to estimate the temperature of the tube metal in the heaters of a naphtha reforming unit. The method described in API Standard 530 (“Calculation of Heater-tube Thickness in Petroleum Refineries”) or similar can be used.   

UOP’s experience is that knife-edge skin thermocouples can be useful as indicators of trends and step changes but are not reliably accurate indicators of the actual tube-metal temperature. Radiant heat from the flames can cause readings that are higher than the actual tube temperatures. Shielded skin thermocouples can provide a good indication of the tube-metal temperature, but the installation of the shield may create a hot spot on the tube that results in metal loss due to carburization.

Infrared pyrometers provide an indication that is higher than the actual temperature of the metal of the tube because they’re reading the temperature of the scale on the outside of the tube. When the tubes are clean, the indication will be accurate. This will be the case immediately after the start-up of a new unit or the restart after tube cleaning or tube replacement, or if the tubes are ceramic coated. These clean-tube readings provide field verification of the temperatures estimated by the calculation method. The increasing indication from the infrared readings, over time, in an operating unit when compared to the tube-metal temperature estimated using a calculation method provides direct indication of fouling on the tubes.

Todd: While tube-metal temperature thermocouples can provide indication of temperature trends over time, their long-term reliability is always problematic. It’s imperative, therefore, that owners-operators of reforming heaters obtain routine IR pyrometer readings of tube temperatures as a backup to the installed tube-metal temperature instruments. Even at the start of a run, IR-scan data is necessary to establish a base line for future comparisons. 

An IR scan of tubes in any heater doesn’t provide highly accurate readings for a myriad of reasons, some controllable and some not. It’s therefore important that the owner understands and controls those variables which can be controlled. Chief among controllable variables are the equipment used to conduct the IR scan, the operator of that equipment, and the target points for the scan. The ideal situation for collection of tube-metal temperature data by IR scan would be a single individual conducting all scans with the same equipment, shooting the same points on the heater tubes. While this is never totally practical, many operators have been successful in monitoring operating reformer heaters by controlling these parameters as closely as possible.

Several methods have been used to check the absolute accuracy of IR tube-metal temperature measurements, all with varying degrees of success. The first method uses a target tube (partial tube without cooling) fitted with an internal shielded thermocouple. This method provides a similar tube surface to an operating tube and a highly accurate tube-temperature measurement. The drawback is that this tube will operate at significantly higher temperatures than the operating tube, so the check temperature is typically 200 to 300° F. higher than the actual operating tube temperatures. Another method is to check the tube-surface temperature on the front face using a preformed hoop-contact thermocouple, which can be inserted into the firebox through a peep door and hooked around a tube to measure the front-surface temperature at or very near an IR target point.

As an alternative to routine IR scanning, several vendors can provide firebox cameras capable of continuous tube-metal temperature monitoring. These cameras are not in widespread use, but they can provide invaluable operating trends for units pushed to their limits.

What is the frequency of fixed-bed reforming and continuous catalytic reforming (CCR) recycle-gas compressor washing to remove salt deposits? What is the typical deposit composition? What is used to wash the compressor? 

Tallman: For fixed-bed reforming, the recycle compressor is water washed at every catalyst regeneration (typically once a year) after the chlorination step. In addition, if there’s any extended outage for mechanical work, the compressor will be water washed as a precaution anytime there’s an opportunity to do so. Steam condensate is used as the water source. The deposits are mostly ammonium chlorides and are readily removed by a hot condensate wash. On rare occasion, if the compressor develops a high vibration, there could be a need to perform a compressor wash. 

Arnoux: Nitrogen from the naphtha hydrotreater causes ammonium chloride salt build up in cold areas of reforming units. While intermittent water washing can be done to treat this, water is a reforming-catalyst poison and must be minimized. Compressor washing during turnarounds is a good practice. Rotating the recycle compressor during turnarounds also is a good practice. 

Wiltse: During normal operation of a fixed-bed reforming unit, the recycle-gas compressor internals may become coated with a fine powder-like deposit. This material is expected to mostly be ammonium chloride salts which are a by-product of having some organic nitrogen compounds in the reforming unit feed. These organic nitrogen compounds will break down across the catalyst bed and form ammonia. This ammonia can combine with the trace hydrochloric acid (HCl) in the recycle gas to form ammonium chloride salts which can deposit in low-temperature sections of the unit. These areas may include the combined-feed exchanger, product condenser, stabilizer column, and recycle-gas compressor. These ammonium chloride salts are water soluble and, as such, can be removed with a water wash.   

UOP recommends washing the recycle compressor during each regeneration before starting the carbon burn. The wash is recommended before the carbon burn to reduce the risk of compressor vibrations and corrosion. During carbon burn, the regeneration gas will be wet, and it’s possible to have some of these deposits spall off the rotor. This could cause balance problems that would result in high compressor vibrations. There’s also a concern for under-deposit corrosion if these salts adsorb moisture from wet regeneration gas.

UOP has observed refiners using a range of different solutions to water wash recycle-gas compressors. These mixtures are typically comprised of a basic solution with nonsudsing detergent. UOP typically recommends ~2% soda ash solution with 0.1-0.3% nonsudsing detergent such as Calgon, Cascade, tripotassium phosphate, etc. It’s important to discuss these chemicals and the wash procedure with the compressor vendor to ensure the wash won’t cause unexpected harm to the compressor. High concentrations of carbonate, for example, may damage aluminum components in the compressor. After the salts have been removed, the compressor should be flushed with clean water and treated with a corrosion inhibitor.


  1. Johnson, D.W., “Sulfuric Acid Release Report, National Petroleum Refiners Association annual meeting, San Antonio, Mar. 20-22, 1994.