January 2022

Bio-Based Processing

Impact of biofeed retrofits, coprocessing on refinery amine units, SWSs and SRUs—Part 1

Diesel with a portion of biologically sourced carbon is being produced at an increasing number of conventional crude oil refineries.

Diesel with a portion of biologically sourced carbon is being produced at an increasing number of conventional crude oil refineries. This is being done by coprocessing biofeedstocks in refinery hydrotreaters and fluidized catalytic crackers (FCCs) or through the installation of a dedicated biofeedstock hydrotreater to produce commercial biodiesel products. Generally, refineries are looking to technologies that allow them to easily incorporate biofeedstocks into their existing infrastructure. This article describes the impacts that this incorporation has on a refinery’s existing amine system, sour water stripper (SWS) and sulfur recovery unit (SRU), which are used to remove toxic contaminants from the hydrocarbon products, waste vapors and wastewater in the refinery. It also presents some of the operational and design options available to manage those impacts, as well as a specific case study. This article focuses on the operational changes in, and challenges for, amine, sulfur and sour water units in conventional oil refineries due to biodiesel production. This information will be useful for readers looking to predict the impact on these units or to resolve an operational issue resulting from biofeedstock coprocessing.

Refineries, meet biofeedstocks

Toxic hydrogen sulfide (H₂S) is present in hydrogen recycle streams, hydrocarbon products, effluent gas and water streams. Refineries must use a treatment scheme (FIG. 1) to remove it—a system that is mostly standardized across conventional refineries in service worldwide. The H₂S, carbon dioxide (CO₂) and ammonia (NH₃) are removed from hydrogen and hydrocarbon streams by water washing and amine solvent scrubbing. The amine and water are then stripped by using steam, and the resulting concentrated stream of H₂S, CO₂ and NH₃ is treated in a modified Claus plant commonly called an SRU. In the SRU, the H₂S is converted to elemental sulfur, and most of the NH₃ is broken down to N₂ and H₂O. CO₂ passes through the unit largely unreacted. The liquid sulfur product is further treated and then sold to the fertilizer and sulfuric acid industries. Many refineries will have an additional amine unit—a tail gas treating unit (TGTU)—to recover more H₂S and sulfur dioxide (SO₂) from the Claus tail gas to reduce SO₂ emissions to the atmosphere. While the general scheme is common to most refineries, there are many variations in configuration, technology, solvent type and equipment within the amine treating and SWS units, and also within the SRUs and TGTUs. H₂S degassing of the liquid sulfur product is not affected by biodiesel coprocessing and, as such, will not be covered in this article. Similarly, the impacts of biofeed coprocessing on the SRU incinerator are minor and can be disregarded.

With stricter limits on CO₂ emissions associated with conventional fuels, bio-based waste oils (such as vegetable oils and liquid intermediates of biomass conversion) have become an interesting subject as an alternative energy source for energy companies. One advantage of bio-based oils vs. crude oil is their low sulfur content; however, their high oxygen content (TABLE 1) presents several processing challenges.

Vegetable oils are the most common biofeedstock in oil refineries. They consist of triglycerides and fatty acids, which contain oxygen in their molecular structure. The presence of oxygen means that these components are already partially oxidized, so their energy density is lower. Although some bio-based oils can be used directly as a fuel, the process of upgrading for better-quality products is attractive. Upgrading bio-based oil requires the removal or reduction of oxygen atoms. Processing or coprocessing bio-based oils in refineries is a low-cost route because the processing infrastructure, supply chain and distribution network not only support but accelerate bio-based oil production.¹ In refineries, the most likely tie-in points for using bio-based products are the diesel hydrotreater (DHT) and the FCCU.

Alternatively, there are potentially cheaper future feedstocks, such as cellulosic biomass, which can be decomposed using pyrolysis to make bio-based pyrolysis oil. This can be used directly as a fuel for various heaters or further upgraded in the hydrotreater or the FCCU. Pyrolysis oil is dissimilar to its petroleum counterpart, containing approximately 300 different carbon molecules and up to 40% oxygen, which means that it requires significantly more (costly) hydrogen during hydrotreating than lipid-based biofeedstocks like vegetable oil.² Pyrolysis oil is acidic, less stable, contains high amounts of water and has low energy density (H/C ratio = 0.2). A project funded by the European Union (EU)—4Refinery—is investigating the processing of blended vegetable and pyrolysis oil feeds, as these are technically easier to handle in existing infrastructure than 100% pyrolysis oil. A summary of the different potential feedstock characteristics is provided in TABLE 1.

Coprocessing routes

Industrially, three routes to producing saleable diesel from a mix of biofeedstock and crude (FIG. 2) include:

1. Injecting a drop-in, lipid-based biofeed into the existing DHT
2. Using a purpose-built, standalone vegetable oil hydrotreater, the diesel from which is blended with conventional diesel to obtain the correct product properties (e.g., cloud point)
3. Incorporating a drop-in lipid (and potentially lignocellulosic) biofeed into the existing FCCU.

The first two routes rely on lipid-based biomass, such as vegetable oil, oil crops, algae or tallow—although pyrolysis and hydrothermal liquification oils are both under investigation for future use. Pinho et al. carried out test runs on a pilot-scale FCCU for coprocessing vacuum gasoil (VGO)—which is its normal feed—with pinewood chips.³ While lipid feedstocks can be converted relatively easily into biodiesel or green diesel, they are more expensive than lignocellulosic ones. Conversely, they generally contain much less oxygen, meaning less hydrogen demand during hydroprocessing. Furthermore, the upgrading technology for these compounds is already well-developed in industry and requires a relatively small investment when incorporated into existing refineries. The amount of upgrading required for lipid biomass is related to the desired degree of biodiesel blending.

Direct injection into the DHT

The first process route is the most common. In this route, the lipid biomass is pretreated (e.g., degummed) to remove contaminants before it is added to the DHT. The catalyzed, hydrogen-rich, high-temperature (300°C–400°C) and high-pressure (40 barg–100 barg) environment of the DHT simultaneously deoxygenates the biofeedstock molecules to produce H₂O and desulfurizes the petroleum diesel compounds to form H₂S. CO₂ is also formed through decarboxylation of the biofeed molecules. Methane, propane and carbon monoxide (CO) are also produced as byproducts. The hydrogenated chemical structure produced in this manner has long, unbranched hydrocarbon chains, which results in a biodiesel with a cloud point much greater than petroleum diesel. Therefore, this method of coprocessing will generally only permit a maximum of 10% biofeed before the blended product begins to exhibit cold flow issues.

The Preemraff Göteborg oil refinery in Sweden has successfully coprocessed up to 30% biofeedstock in its DHT as a summer-only fuel.⁴ However, this was done using a feedstock known as raw tall diesel, which is a processed bioproduct of pulp mills. This product is not equivalent to vegetable oil and is only available in sufficient quantities in some geographical locations.⁵ Note: These larger coprocessing proportions required some hydrotreater modifications and solutions to challenges related to hydrogen consumption, exotherm, corrosion and catalyst selection.

Injection into a standalone green diesel unit

The second process route is to produce hydrotreated vegetable oil (HVO) in a purpose-built standalone vegetable oil hydrotreater designed specifically for HVO production. The resulting HVO—known as hydrotreated esters and fatty acids (HEFA)—is then blended with the hydrotreated petroleum diesel. Certain processes that have successfully employed this approach are Honeywell UOP’s Green Diesel and Neste Oil’s NEXBTL processes. While not technically coprocessing, up to 30% HVO may be blended with conventional transport diesel to produce what is referred to as green diesel. This approach allows for a greater bio-content due to the extra processing in the standalone HVO hydrotreater.

As previously stated, one issue with directly injecting biofeed into the petroleum hydrotreater is the high cloud point of the resulting biodiesel. However, the HVO-specific hydrotreater is a two-stage process. In the first reaction stage, the fats are hydrotreated with hydrogen to saturate the double bonds and deoxygenate the molecules. The second hydrocracking/isomerization stage reduces the length and increases the branching of the alkane molecules. This second stage is also called dewaxing, as this process reduces the cloud point of the resulting HVO so that it is consistent with petroleum fuel specifications and has acceptable cold flow properties. The investment for an HVO standalone hydrotreater is greater than in the first processing approach and uses more hydrogen. However, the HVO product is closer to petroleum diesel and provides greater blending flexibility. The mass yield of HVO liquids from raw lipid material is approximately 80%, with the balance being composed of mostly propane, methane and oxygenated gases like CO₂ and CO. This approach produces significantly more propane than the first approach, as the hydrogenation reaction breaks the lipid esters from the propane backbone. A more detailed background is publicly available in the report by Kampman et al. for the EU commission.⁶

Injection into the FCCU

The third process route for coprocessing and blending is to co-feed a biofuel with the crude oil to the FCCUs, which operate at approximately 500°С and at low pressures (1 barg–2 barg). The purpose of the FCCU is to crack heavy hydrocarbon molecules into gasoline and some middle and heavy distillate fractions. This point in the refinery is a good candidate for coprocessing because the FCCU does not require expensive hydrogen for deoxygenating the feed. The catalyst is also more tolerant to contaminants than hydrotreating catalyst and can be regenerated onsite. Common problems faced in FCCU processing biofeeds are low yield, catalyst deactivation and reactor plugging.

The vegetable oil cracking mechanism in the FCCU is complex. Unlike in the hydrotreating process, where hydrogen is used to remove oxygenated hydrocarbons to produce water, the FCCU process rejects the oxygen atoms, which results in the production of H₂O, CO, CO₂ and other oxygenated hydrocarbons³ such as short-chain carboxylic acids, aldehydes, ketones and phenols.⁷ The concentration of these products depends on the bio-based oil type, the FCCU catalyst type, the operating parameters of the FCCU and other factors. TABLE 2 summarizes the impacts of the various processing methods on the quantity and quality of the important liquid and gas streams.

Impacts on amine treating

In conventional refining, hydrogen, fuel gas and LPG streams are contacted with an amine solvent to remove H₂S. The amine solvent will also remove CO₂ and, in some instances, other sulfur species. A simple process schematic of an amine system is shown in FIG. 3. The greatest impact of biodiesel production on amine treating is that the feed gas has more CO₂ and less H₂S.

Depending on the amine solvent used in the refinery, the extra CO₂ generated by processing the biofeed can reduce the quality of the acid gas feeding the SRU, especially at greater co-feed proportions. With primary and secondary amines, the extra CO₂ will normally be removed by the amine solvent to ppm levels in the treated gas or liquid, and there will be a comparable increase in CO₂ content in the amine acid gas to be processed in the SRU. Being a stronger acid than H₂S, CO₂ also requires more regeneration energy to strip from the solvent in the regenerator.

This extra duty load on the reboiler can potentially adversely affect the H₂S lean loading needed to meet product specification from the absorbers in the process loop. If the plant is already reboiler-limited, the quantity of coprocessed biofeed may need to be reduced.

If the plant uses a selective amine, such as methyl diethanolamine (MDEA), it has the potential to slip some of the extra CO₂ (since CO₂ can pass directly through the contactor without being removed), so that the effect of the extra CO₂ on the acid gas quality is not as significant as with non-selective amines. In some instances, it may be beneficial for an oil refinery to switch from ethanolamine (MEA), diethanolamine (DEA) or diglycolamine (DGA) to MDEA to offset the additional CO₂ and reduce energy consumption. However, it is more challenging for a weaker base like MDEA to meet treated gas specifications at low pressure than it is for primary and secondary amines. MDEA also has a higher hydrocarbon solubility than MEA, DEA or DGA and is more vulnerable to foaming as a result.

The presence of the extra CO₂ in the rich solvent can accelerate corrosion in the bottom of the regenerator and reboiler if the steam flow up the regenerator column is not optimized for the extra duty (FIG. 4). Higher CO₂ loadings can increase the risk of corrosion throughout the system. A detailed description of the mechanisms for this is provided in literature.¹² 

In addition to CO₂, CO will also increase in the feed gases to the amine system because of coprocessing. CO is known to react with MDEA and caustic solutions to form formate salts¹³, while CO reacts with primary amines (MEA and DGA) or secondary amines (DEA and DIPA) to form formamides and formate salts. Formates have been shown to be corrosive to carbon steel in liquid and vapor states. Formamides do not appear to be corrosive, but they do increase viscosity and reduce amine available for acid gas removal.¹³

Field test results from several facilities coprocessing biofeed in their DHT showed an increase in CO in the hydrogen gas being amine treated from zero to the order of 1,500 ppmv–4,500 ppmv, with one facility showing a CO content of more than 10,000 ppmv. Therefore, the partial pressure of CO in the amine unit will normally be less than 100 kPa with current coprocessing levels. Based on observations in synthesis gas-treating amine systems, significant formate formation from CO does not normally occur at partial pressures less than 400 kPa.

Current biodiesel feedstocks contain large quantities of long-chain fatty acids. In their unprocessed form, these will have a major negative impact on the surface tension and viscosity of an amine solvent. In turn, this may cause the amine solvent to foam (or emulsify in the case of LPG treating). Conventional silicon and polyglycol anti-foams have had some success in mitigation of this foam.

The FCCU biofeedstock coprocessing route appears especially vulnerable to foaming and emulsification problems. The authors are aware of five refineries coprocessing biofeedstock in their FCCUs, four of which experienced significant LPG emulsion issues as a result. Unlike the other two routes where hydrotreating is employed, there are many potential breakdown products that can be formed from catalytically cracking biofeedstocks. A summary of these is presented in TABLE 3. Tests on extracted phases of sour FCCU LPG have also indicated that high-oxygen content polymers may be present in the LPG.

Dealing with the emulsified FCCU LPG is a challenge, and conventional techniques (anti-foams and carbon beds) do not appear to be adequate. Tailored demulsifiers injected into the sour LPG feeding the column have shown some success at reducing the emulsification; however, this treats the symptoms rather than removing the cause. Some success has also been achieved using proprietary absorbents on the LPG and the amine sides. In practice, a combination of absorbents and demulsifiers on the FCCU LPG treater seems to be required to achieve FCCU biofeed coprocessing blends of more than a few percent.

Impacts on sour water stripping

Refinery biodiesel production often results in a significant increase in sour water volume as a result of the deoxygenation of biofeed with hydrogen and the water emulsified with the biofeedstock. The extra water may require debottlenecking of the sour water system or the installation of additional stripping capacity. The optimal strategy for debottlenecking is plant-specific and dependent on the pre-existing system. A holistic approach to the review of the complete refinery water system may reveal an opportunity to substantially reduce the use of fresh water.

A good process water strategy can have the following additional benefits:

• Reduced phenols in refinery effluent water
• Reduced fractionator overhead corrosion or less corrosion inhibitor use
• Reduced sodium in residue feeds, which may enable longer thermal cracker unit runs
• Reduced desalter emulsion issues
• Some spent caustic wastes could be neutralized in the sour water system.

With biodiesel production, the combined refinery sour water will contain more CO₂ and less H₂S and NH₃. There will also be a dilution effect because of the additional water produced. Consequently, more energy will be required to process wastewater generated from the biofeedstock, predominantly to increase the temperature of the extra water to stripping conditions. In most cases, the energy source is low-pressure steam, which is abundant in many facilities; however, the energy cost of stripping the wastewater should not be neglected when measuring the global benefits of a biodiesel project.

The co-authors’ company has found the following steps to be effective in optimizing refinery water systems:

1. Perform a detailed compositional analysis of the refinery process unit water streams feeding the existing SWS system and their actual or estimated flowrates.
2. Estimate increased water volume and its CO₂ content based on biofeedstock properties.
3. Route streams with less than 10 mg/l H₂S and 50 mg/l NH₃ directly to final effluent treating to free up stripping capacity. While, in theory, these streams can be reused, often the hydrocarbon content and water hardness render them unsuitable.
4. With multiple SWSs, segregate phenolic waters and recycle them to the desalter. In the desalter, most of the phenol will be extracted back into the crude oil where it can subsequently be broken down in downstream hydrotreating.
5. Supplement desalter water with unstripped water from the crude overhead if capacity allows.
6. Use water streams with few strong cations and anions (e.g., some tail gas unit quench towers) as amine system makeup water.
7. Utilize a semi-stripped stream drawn from near the bottom of the SWS column (with approximately 200 ppmw NH₃) to potentially replace neutralizing chemicals injected into fractionator overheads for the desired pH of about 6.5, which saves on operating expenditures (OPEX).
8. Maximize the reuse of water from the low-pressure to high-pressure side on an FCCU, while not exceeding 25 ppmw hydrogen cyanide (HCN) and 2 wt% ammonium hydrosulfide (NH₄HS) in the high-pressure wash water.
9. Note: In-situ sulfidic spent caustic neutralization in an SWS is possible, but good pH control is essential and recycling sodium-containing stripped water upstream can have a negative effect on desalter performance (increased emulsification), thermal cracker time between decoking and FCCU catalyst life. Careful study is needed before doing this.

In many instances, the additional water from biodiesel production will require an additional SWS unit. The optimal process location for this is not necessarily for this new water but will be site-specific. For instance, if there is not a pre-existing segregated phenolic water system, this may be the opportunity to install one on the catalytic and thermal cracking units, which are the source of phenols in the refinery water. This can save OPEX on expensive phenolic treatment of the water.

A new stripper also does not have to follow a conventional design. A stripper requiring a low capital expenditure (CAPEX), proposed by Lieberman, sends the hot stripped water directly to the plant desalter and can incur less than half of the final installed cost of a conventional unit (FIG. 5).¹⁴ This low-cost stripper does have some limitations, the most significant being its sensitivity to reboiler duty. In one study, a 2% increase in reboiler duty increased SWS acid gas volume by about 30% (from water additional vapor), which can be a challenge for the SRU. This configuration works best when the low-CAPEX style stripper is contributing a relatively small proportion to the combined amine and sour water acid gases, and when the SRU is operating at a healthy load. This may not be suitable for facilities with a small ratio of amine acid gas (AAG) to sour water acid gas (SWAG); therefore, it is critical to confirm the SRU minimum turndown and to review burner controls.

There is often a range of trace-level strong ions in the various feed water sources to the sour water system. In conventional refining, these are normally sodium, chloride, sulfate and short-chain organic acids. Phosphate is present in most biofeedstocks, and, while pretreatment of biofeedstock will remove the bulk of the phosphate, trace amounts will emerge in the wastewater. While the concentration of these trace ions is usually small (< 500 mg/l), they can pose a significant operational challenge, as they will form an ionic bond with H₂S or NH₃ that renders them unstrippable at conventional SWS conditions. Counter-dosing with an acid or base to neutralize them is common practice. Judicious addition of base in the optimal location can also reduce the energy requirements for stripping the water. Detailed discussions of this and other operational challenges in sour water stripping have been published in literature.¹⁵

Wastewater can also contain significant quantities of hydrocarbons and particulates, and may vary significantly in appearance, depending on the degree of contamination in the source waters (FIG. 6). Hydrocarbons that are not stripped from the water will often—in conjunction with particulates—form a fouling material that adversely affects SWS performance.¹⁵ 

In addition, lighter hydrocarbons that enter the SWS will be removed and sent with the overhead gas to the SRU. This may create operational problems for the SRU and be detrimental to its performance. Even a small increase in heavier hydrocarbons in the SWAG can result in a large increase in the amount of air required for combustion in the SRU reaction furnace, a phenomenon that is aggravated when the hydrocarbon content of the water is not stable. The following may occur in the SRU because of hydrocarbons in the sour water:

• Decreased hydraulic capacity
• Unstable plant operation due to variable air demand
• Damage to the SRU reaction furnace caused by rapid temperature variations
• Reduced efficiency because of poor H₂S-to-SO₂ ratio control
• Decreased recovery efficiency due to increased carbon disulfide (CS₂) formation
• Increased operating pressure from catalyst bed plugging by soot deposition from incomplete hydrocarbon combustion
• Decreased catalyst activity because of catalyst blockage from aromatic cracking
• Reduced sulfur quality due to soot in the sulfur product, resulting in “black sulfur”
• Soot fouling of the sulfur plant waste heat boiler and condenser systems.

These effects are extensively documented in literature.¹⁶

Hydrocarbons in water streams can be present in three forms: free, dissolved and emulsified.¹⁷ In most cases, biofeed coprocessing is done without resizing water separation vessels on the FCCU or HDT, which means less residence time for water and liquid hydrocarbon separation because of the increased water volume associated with processing biofeeds. Additionally, less time for separation can leave more liquid hydrocarbon in the wastewater from the FCCU or HDT.

Minimizing the hydrocarbon content of the sour water feed may be achieved by the following actions:

• Removing free hydrocarbons in the inlet three-phase separator; in general, 25 min of residence time with a liquid level at 50%–60% of drum height is optimal
• Installing a sufficiently large feed stabilization tank with hydrocarbon skimming facilities (sometimes, these tanks have several days of residence time, during which some emulsified hydrocarbons will come out of emulsion and may be removed)
• Installing a coalescer to remove emulsified hydrocarbons—typically, this requires pre-filters (FIG. 7)
  
  

  FIG. 7. Optimal wastewater filtration and coalescing scheme.¹⁶
• Using a hydro-cyclone separator to reject a hydrocarbon-rich light phase back to the feed stabilization tank.

Part 2

Part 2 will be published in the February issue. HP

NOTES

This work was first presented at the 28th European Biomass Conference & Exhibition (EUBCE 2020).

ACKNOWLEDGMENTS

The authors would like to thank MPR Services for sharing information on FCCU breakdown products, Jeroen Engels for his assistance with graphics and formatting and Peter Seville for the proofing of this paper.

LITERATURE CITED

1. Van Dyk, S., J. Su, J. D. McMillan and J. N. Saddler, “‘Drop-In’ Biofuels: The Key Role that Co-Processing will Play in Its Production,” IEA Bioenergy, 2019.
2. Karatzos, S., J. D. McMillan and J. N. Saddler, “The Potential and Challenges of Drop-In Biofuels,” IEA Bioenergy, 2014.
3. Pinho, A. R., M. B. B. Almeida, F. L. Mendes, L. C. Casavechia, M. S. Talmadge, C. M. Kinchin and H. L. Chum, “Fast Pyrolysis Oil from Pinewood Chips Co-Processing with Vacuum Gas Oil in an FCC Unit for Second Generation Fuel Production,” Fuel, 2017.
4. Egeberg, R., K. Knudsen, S. Nyström, E. L. Grennfelt and K. Efraimsson, “Industrial-Scale Production of Renewable Diesel,” Petroleum Technology Quarterly, 2011.
5. Van Dyk, S., J. Su, J. D. McMillan, J. N. Saddler and J. Su, “Potential Synergies of Drop-In Biofuel Production with Further Co-Processing at Oil Refineries,” Biofuels, 2019.
6. Kampman, B., R. Verbeek, A. van Grinsven, P. van Mensch, H. Croezen and A. Patuleia, “Options to Increase EU Biofuels Volumes Beyond the Current Blending Limits,” European Commission Study, 2013.
7. Bielansky, P., A. Weinert, C. Schönberger and A. Reichhold, “Catalytic Conversion of Vegetable Oils in a Continuous FCC Pilot Plant,” Fuel Processing Technology, 2011.
8. Ng, S. H., N. E. Heshka, Y. Zheng, Q. Wei and F. Ding, “FCC Co-Processing Oil Sands Heavy Gas Oil and Canola Oil. 3. Some Cracking Characteristics,” Green Energy & Environment, 2019.
9. Bryden, K., G. Weatherbee and E. T. Habib, “Flexible Pilot Plant Technology for Evaluation of Unconventional Feedstocks and Processes,” GRACE Catalyst Technologies, January 2013.
10. Al-Sabawi, M., J. Chen and S. Ng, “Fluid Catalytic Cracking of Biomass-Derived Oils and Their Blends,” Energy Fuels, July 2012.
11. Watkins, B., C. Olsen, K. Sutovich, J. Deady, N. Petti and S. Wellach, “New Opportunities,” Hydrocarbon Engineering, January 2009.
12. Sheilan, M. and R. F. Smith, “Hydraulic-Flow Effect on Amine Plant Corrosion,” Oil & Gas Journal, 1985.
13. Kim, C. J., A. M. Palmer and G. E. Milliman, “Absorption of Carbon Monoxide into Aqueous Solutions of Potassium Carbonate, Methyldiethanolamine, and Diethylethanolamine,” Industrial & Engineering Chemistry Research, 1988.
14. Lieberman, N., “Sour water strippers: Design and operation,” Petroleum Technology Quarterly, 2Q 2013.
15. Le Grange, P., “Operational Challenges in Sour Water Stripping,” Petroleum Technology Quarterly, 2Q 2019.
16. Klint, B., “Hydrocarbon Destruction in the Claus SRU Reaction Furnace,” Sulphur Recovery Textbook, Sulphur Experts, 2005–2012.
17. Sheilan, M., B. Spooner, E. van Hoorn, D. Engel, P. le Grange and F. Derakhshan, “Reducing Hydrocarbons in Sour Water Stripper Acid Gas,” Laurance Reid Gas Conditioning Conference, 2014.

The Authors

Related Articles

From the Archive

Comments