B. Watkins, ART Hydroprocessing, Chicago, Illinois; and C. Baillie, ART Hydroprocessing, Worms, Germany
Hydroprocessing investments in sustainable fuels are gaining momentum, with a pipeline of up to $50 B expected by 2025 involving more than 250 projects, the majority of which focus on hydroprocessed esters and fatty acids (HEFA).1 HEFA is a source of biodiesel in the form of renewable diesel (RD) and is also a source of sustainable aviation fuel (SAF). In fact, HEFA is now the only commercial route used at scale to produce SAF, and is also expected to be a major contributor to the SAF pool in the long term.
Against this backdrop, the authors’ company has launched a new catalyst seriesa. These are tailored catalyst systems for the hydroprocessing of renewable feeds that enable customers to process a wide range of biofeeds with the maximum yields of RD and SAF. This article discusses how the challenges associated with producing renewable fuels through the hydroprocessing of biofeeds can be addressed through such specialized catalyst technologies.
Managing contaminants. The hydroprocessing of biofeeds presents considerable challenges compared to petroleum-based feeds, and consequently requires specialized catalysts. Contaminant removal is a particular key consideration, and the authors’ company has drawn upon its experience and expertise in resid catalyst technologies to develop a new series of catalysts that allows a wider and different set of poisons to be captured. Biofeedstocks can be high in phosphorus, alkali/alkaline earth metals such as sodium, calcium, potassium and magnesium, as well as iron.
It is important to consider the various contaminants to avoid unfortunate surprises like pressure drop build-up or unexpected catalyst deactivation, both of which can result in shortened cycle length and unexpected turnarounds. In some cases, guard catalysts can be employed to help mitigate some of the problems caused by catalyst poisons, and guidelines are suggested to help minimize the potential impact and allow the hydroprocessing system to perform at peak values for an extended time.
Of the major contaminants, phosphorous is the highest concern. Phosphorus is found in the phospholipids present in biologically produced oils and is also used in the process to make several grades of edible oils—since there is no regulated limit on this, their levels can vary significantly from batch to batch in the processing. Phosphorus is believed to be attracted to the alumina surface and has the ability to combine with the molybdenum on the catalyst surface. The binding with the metal sulfide sites dramatically reduces catalyst activity and alters the acidic chemistry of the sites, changing the oxygen selectivity of the catalyst. The activity loss from phosphorus on the main hydrodeoxygenation (HDO) catalysts is roughly Δ10°F (Δ5.5°C) per 1 wt% on the catalyst. It is recommended to keep levels low to extend run length, as higher temperatures tend to decrease product yield due to excessive carbon monoxide (CO)/carbon dioxide (CO2) production.
Sodium is commonly present in animal oils, and is a severe catalyst poison that can cause significant activity loss even at low levels by promoting the sintering of catalytic metals and neutralizing acid sites. Depending on the source of sodium, the signs of poisoning include rapid activity loss and an increase in pressure drop. Sodium has a larger impact on catalyst activity losses, where 1 wt% on the catalyst translates to roughly Δ30°F (Δ17°C) loss in activity. Often, high sodium content found in the raw oils can be separated out via a water wash that will dissolve the sodium often found as salt from the cooking process. Calcium is another contaminant present in animal oils. It is a similar poison to sodium with some evidence suggesting that it has an even bigger impact on deactivation, with roughly 1 wt% calcium on the catalyst resulting in Δ50°F (Δ28°C) activity loss.
Olefin saturation and deoxygenation selectivity. In general, biofeeds consist of triglycerides, which are molecules containing a three-carbon glycerol-based backbone connected to unsaturated fatty acid chains (FIG. 1).
In addition to containing olefins, they also have a high oxygen content, which can typically be between 10 wt% and 12 wt%. Olefin saturation and deoxygenation tend to occur nearly simultaneously under easy hydroprocessing conditions, which poses various challenges.
Olefins release a high amount of heat upon undergoing hydrogenation (hydrogen addition) to form the saturated paraffinic chains. Since the molecule chains are fairly long, an uncontrolled reaction with excessive heat can cause the fatty acid chains to polymerize into high carbon-content molecules or aromatics, forming large molecules that can deposit on the catalyst leading to pressure drop buildup.
Removing oxygen also generates significant heat—doing so in an uncontrolled manner leads to a non-kinetically controlled system conducive to carbon-carbon bond cleavage and results in unwanted side products and lower product yields. Therefore, it is essential to use catalysts that can control the rate of these reactions and avoid the negative consequences of rapid and excessive heat formation.
The authors’ company has performed an extensive amount of R&D to better understand the pathways for oxygen removal, as well as the impact of operating conditions and catalysts on deoxygenation selectivity. Oxygen removal can proceed via three different reaction pathways, as shown in FIG. 2.1 The desired reaction is deoxygenation, in which oxygen is removed in the form of water only. When deoxygenation occurs, all renewable carbons within the fatty acid chain are preserved in the paraffinic product, resulting in maximum renewable carbon retention and liquid yield.
The other two routes, decarbonylation and decarboxylation, are competing reactions that should be minimized. In these two routes, oxygen is also removed, albeit in the form of carbon monoxide (CO) and carbon dioxide (CO2). This results in a yield loss as the paraffinic products contain one less carbon for use in energy production. The lower renewable carbon retention through carbon-carbon bond breaking creates CO and CO2 byproducts, which can inhibit downstream catalysts, as well. The extra side reactions that can occur also consume nearly the same amount of hydrogen as compared to selective deoxygenation without the liquid gain.
Fortuitously, nature assists us in efforts to conveniently measure the selectivity of oxygen removal. The fatty acids in most triglycerides typically have an even number of carbon atoms, with fatty acids consisting predominantly of 16–18 carbons. The desired route for the deoxygenation reaction preserves all of these carbons in the product, resulting in paraffins with an even number of carbons, while the undesired decarbonylation and decarboxylation reactions lead to a loss of a single carbon, resulting in paraffins with an odd number of carbons. Therefore, through carbon tracing using ASTM D2887,2 the ratios of even to odd numbered carbon chains can quickly help distinguish the selectivity towards deoxygenation, as shown in FIG. 3.
Key findings of extensive pilot plant research into understanding the operational factors that impact oxygen and nitrogen removal when processing biofeeds are shown in FIG. 4. To help maximize the selective routes for oxygen removal, lower liquid hour space velocity (LHSV) and a higher ratio of hydrogen-to-oil is highly favorable for oxygen conversion. Complete oxygen conversion can typically be achieved at much lower temperatures than traditional fossil fuel operations, although a hydrogen-to-oil ratio of 5–6 times the consumption is recommended on a fresh feed basis. Complete nitrogen conversion requires between a Δ30°F–Δ50°F (Δ17°C–Δ28°C) higher reactor temperature coupled with a more nitrogen-focused choice of catalyst to produce the low nitrogen levels needed for downstream processing. A catalyst with superior nitrogen removal enables a lower start-of-run weighted average bed temperature (SOR WABT) by allowing the deoxygenation reaction to proceed at lower temperatures and operate in the kinetic region for the catalyst used.3
American Petroleum Institute (API) density measurement is also an effective way to help gain an indication of selective oxygen removal. TABLE 1 highlights the API for various triglycerides (TGA), free fatty acids (FFA) and paraffins. FFAs tend to have slightly lower densities (higher API) than their TGA counterparts, with a big decrease in density observed for the paraffin products. In addition, C17- paraffins have a lower density than C18- paraffins. This has assisted the authors’ company’s catalyst development studies, as it can quickly pivot to alternative formulations when it observes that product density is too low.
Placing the right catalyst in the right place is critical for optimum yields. FIG. 5 examines the products generated through a catalyst system and can shed light on the levels of oxygen selectivity. The chart on the left represents the top 20% of the catalyst system. As the operating temperatures of that region are increased, an increase in propane yield is observed. This is an indication that it is possible to kinetically control the cleavage of the carbon-oxygen bonds. With too high a temperature—even in a small layer of catalyst—the presence of CO and CO2 can be observed, indicating that this layer has moved away from a kinetically controlled system for HDO to a more thermodynamic system that can break carbon-carbon bonds.4,5
The chart on the right of FIG. 5 represents the top 40% of the catalyst system. Under these conditions, the overall system is operating at a slightly higher temperature, and higher conversions can be achieved while still maintaining very high levels of selectivity. Through this work, it becomes clear that catalyst staging and proper operating temperature windows can dramatically alter the system’s ability to maintain maximum yields. Poor selectivities are observed when operating too hot, even in a very well-designed system.
Using the correct catalyst system is critical to achieve optimum yields for a given process, as the choice of catalyst must consider unit limitations as well as feedstock variation. In the chart on the left of FIG. 6, two different catalyst systemsa are shown, both of which ultimately provide the target product density and HDO to achieve the required selectivity.
The two catalysts systems are different in their activity levels and the speed at which they can achieve full oxygen removal. This knowledge of the combination of speed and efficiency for a tailored system can be applied for different units depending on specific constraints. For example, a unit constrained by LHSV would be trying to achieve a slightly faster reaction than what would be achievable at a larger unit, and would benefit more from System B than System A.
The chart on the right in FIG. 6 shows the impact that different biofeeds can have on the operation. Each biofeed requires slightly different conditions to achieve the target oxygen and nitrogen levels, which can be addressed through the properly designed catalyst system.
Achieving cloud point specifications. As highlighted in FIG. 2, the hydroprocessing of triglycerides generates straight chain paraffins, and these molecules have poor cold flow properties. Therefore, cloud point reduction catalysts are required so the resulting RD and SAF comply with the finished diesel and jet fuel cloud point specifications.
When selecting these catalysts, the inability to prevent cracking to smaller molecules can lead to significant liquid yield loss. Traditional dewaxing catalysts crack normal paraffins—while this does provide the required reduction in cloud point, it also significantly lowers diesel and jet yields. The authors’ company’s catalyst technologyb results in higher product yields by suppressing cracking reactions, and achieves the required cloud point through the isomerization of normal paraffins to iso-paraffins. The company’s latest generation of this catalyst technologyb has been employed in more than 20 commercial units for the dewaxing of fossil-derived molecules. Building upon this industry-leading performance, the company has now commercialized new dewaxing catalysts specifically for producing renewable fuels.
FIG. 7 shows the evolution of the company’s dewaxing catalysts for renewable fuel production. The chart on the left of FIG. 7 highlights how the latest isomerization catalystsc provide increased yields at a given cloud point compared to conventional catalysts that were developed for petroleum-based fuels. The production of SAF requires high dewaxing severity to achieve the required cloud point properties, which can result in lower SAF yields. The chart on the right shows that the isomerization catalystsc minimize naphtha formation formed by undesired cracking, with the desired isomerization reactions resulting in > 90 vol% SAF yield. Production of RD requires lower dewaxing severity and yields of > 97 vol% are possible.
Commercial experience summary. The company’s catalystsa have been successfully applied for the hydroprocessing of a wide range of biofeeds, including refined and unrefined vegetable oils, greases and animal fats, animal fats and used/waste cooking oils. They have been applied in ultra-low sulfur diesel (ULSD) units co-processing biofeeds, single-stage units (sour service) processing 100% biofeeds, as well as the more common two-stage units (sweet service) for processing 100% biofeeds. The company’s first commercial reference for 100% biofeed applications dates back to 2012. Since then, its catalysts have been used in more than 26 hydroprocessing cycles, with more than 15 repeat customer fills, in some of the highest profile operating units in North America, Europe and Asia-Pacific. This has allowed the company to build up extensive knowledge around processing renewable feeds and their impact on unit operation and yields and provides it with the basis to continue to develop new catalystsd for contaminant capture and selective deoxygenation, as well as isomerization catalystsc to maximize the yields of both RD and SAF. HP
NOTES
a ART’s ENDEAVOR™
b Chevron’s ISODEWAXING® catalyst
c Chevron’s EnHance™
d ART’s EnRich®
LITERATURE CITED
BRIAN WATKINS has more than 30 yr of experience in various areas of hydroprocessing, and has held a number of technical research, management and technical engineering support positions at Grace and Advanced Refining Technologies, as well as managing the pilot plant facilities in Chicago. He now serves as Global Technology Manager, Distillate and Renewable SME (subject matter expert) for ART Hydroprocessing in Chicago, Illinois, where he provides customer support, technical advice and performance monitoring to refiners globally. Watkins holds a BS degree in chemistry from Western Illinois University in Macomb, Illinois, and has written and presented numerous technical papers at AFPM events, CLG symposiums and various other locations globally.
COLIN Baillie is the Strategic Manager for Distillate Hydrotreating and Renewables at ART, a joint venture between Grace and Chevron. He joined Grace in 2006 and has 18 yr of experience working in the field of FCC and hydroprocessing catalysts. Previously, Dr. Baillie obtained a PhD in chemistry from the University of Liverpool, UK, and an MBA from the Open University, UK.