March 2022

Process Optimization

How hydroprocessing feed filtration system design impacts process reliability and efficiency

Oil refining is an exceptionally complex industry, with much of the inherent complexity stemming from wide regional variances in crude quality and composition, plant capabilities and the types of products produced.

Oil refining is an exceptionally complex industry, with much of the inherent complexity stemming from wide regional variances in crude quality and composition, plant capabilities and the types of products produced. This creates the need for a broad range of equipment, processes and technologies, which can vary in quality, performance and design.

When considering process filtration, some refiners explore innovative solutions for solids contaminant control, while others simply elect to maintain the status quo. Low-efficiency or undersized filtration systems can lead to frequent and prolonged process upsets, as well as downtime due to equipment fouling, repeated filter changeouts and/or higher process-related operating costs. Additionally, frequent filter changeouts result in higher direct consumable costs, along with indirect costs related to safety, labor, inventory and disposal.

The most significant cost in substandard filtration lies in potential damage to the catalyst beds of hydrotreaters and hydrocrackers. Many operators do not fully realize the impact of inefficient or insufficient filtration in these units, and they simply view filtration as an insignificant piece of equipment on a piping and instrumentation diagram (P&ID). However, insufficient filtration, specifically the lack of particle removal efficiency, has a substantial impact on operational reliability and costs.

Hydrotreaters are units that use hydrogen to remove impurities (such as sulfur) from petroleum cuts (FIG. 1). Hydrocrackers are units that use hydrogen for the conversion of heavy cuts into lighter fractions such as naphtha, kerosene and gasoils (FIG. 2). Both types of reactors use a fixed bed of catalysts (along with pressure, temperature and hydrogen) to cause the desired chemical reaction.

FIG. 1. Hydrotreaters remove impurities, such as sulfur or nitrogen, and are designed for specific operations. Hydrotreater catalysts are composed of a porous alumina support with a coating of metallic sulfides and are approximately 0.0625 in. in diameter. Common applications include gasoline, naphtha, kerosene and gasoils, as well as biofuels, which present a broader range of applications and challenges.

FIG. 2. Hydrocrackers are used for converting heavy feeds (such as vacuum gasoil) into more valuable middle distillates. Depending on the crude slate and processing, there can be an estimated 20% yield in light and heavy gasoline (sent to the reformer), along with the same percentage in kerosene.

These units are the heart of the refinery and are critical to downstream processes and product specifications. Fixed-bed reactors are often set up in series and are used for low-to-medium metal content in feeds. These reactors have graded beds and progress from hydrotreatment to hydrocracking.

Hydroprocessing filtration

The purpose of feed filter filtration for both hydrocrackers and hydrotreaters is to prevent fouling of the catalyst bed. Historically, filter designs vary because the quality and characteristics of crude oil differ greatly by region and classification. For example, the sulfur content in sweet light crude is generally 0%–0.5% of weight, while in lower-quality heavy sour crude, it can be upwards of 5%.

This is significant because in the next 20 yr–30 yr, the market expects to witness a growing movement toward lower-quality crudes, as well as toward bio-sourced feedstocks from plant-based and animal-based oils. These changes will result in completely different sets of filtration challenges, and these trends may also be accompanied by stricter environmental specifications—all of which place an increased demand on catalyst bed protection.

Nevertheless, protecting the catalyst bed by providing reliable and predictable effluent fluid quality is the primary goal of any properly designed filtration system. For hydroprocessing units, the feed filter provides a sacrificial system that removes undesirable contaminants that can cause reactor fouling. A well-designed feed filter system will prevent problems in both hydrotreater and hydrocracker applications by allowing the reactors to reach full catalyst life and to enter scheduled turnarounds as planned. Although they are replaceable, filters should be given serious consideration, within the context of ensuring trouble-free refining operations, to optimize refining margins and minimize downtime. Unfortunately, this careful consideration does not always occur, as feed filters are often viewed as a “commodity.” As a result, operators tend to default to antiquated and/or undersized designs, or they simply choose the lowest-cost option. Because of the importance of catalyst bed protection, a more comprehensive view should be taken when determining the selection of a filter.

Determining a properly sized filtration system reflects a struggle between capital and operational expenditures (CAPEX/OPEX). CAPEX budgets may call for smaller vessels, which can result in poor performance and higher changeout frequency. OPEX is better served by a filter system sized on criteria that includes inlet particle size distribution, suspended solids concentration, higher filter surface area, higher filter particle retention efficiencies and total cost of ownership.

The economic relationship between the filter and reactor

To truly evaluate a superior value in a filtration system, it should be viewed as a component of the larger refining process. Ideally, a reactor goes into the scheduled turnaround for catalyst reload every 3 yr to 7 yr, depending on the unit. The reactor should reach terminal differential pressure and low-level activity simultaneously. However, many refiners struggle with this and, due to excessive differential pressure, are forced to shut down operations prematurely. The cause of this is poor feed filter performance, which leads to reactor plugging and inefficient reactions because of contaminate loading from the process feedstock. Signs of this include:

• Reaching catalyst bed differential pressure prematurely
• Requiring scraping of the guard bed
• Restricted throughput
• Leaving the bypass line open to help throughput
• Experiencing reactor hot spots due to channeling.

In addition to poor conversion and lower productivity, bed loading can cause the formation of coke, which will ultimately lead to an early catalyst reload. To prevent this costly measure, it is critical to have proper feed filters and to seek out innovative solutions that enhance the performance and longevity of the reactor.

In evaluating the economics for a feed filter system, the first thing to consider is whether the catalyst bed is being protected. While the catalyst is technically a variable cost, it is easy to obtain the historical expenditures at any given refinery. For example, one major refiner’s spend on catalysts and chemicals is displayed in TABLE 1.

Outside of the one-time cost of the filter vessel, there are many things to consider when evaluating the total cost of filtration. These may vary greatly from one application to the next, but the following can typically be found to some degree in any system:

• Costs of plant downtime, including loss of production and equipment replacement, if the right filtration technology is not deployed
• Labor costs
• Total annual direct filter cartridge cost = total changeouts × cost per changeout
• Filter disposal costs
• Operator hazard-related costs
• Cost of shipping and storing filters
• Other consumable costs per changeout (e.g., fluid losses, vessel seals)
• Rental equipment for
  catalyst change.

Viewed as an integral part of a larger process design, the selection of a feed filter system at the design phase becomes a more significant consideration. Feed filter performance is crucial for the refiner. A lack of performance can result in millions of dollars in direct material and labor costs, as well as unplanned production losses.

Designing filters for value and improved catalyst performance

A correctly sized, efficient feed filter system provides the most effective way to condition the feedstock prior to reaching the reactor. The feed filter should be located as close to the reactor as economically feasible to ensure that a minimal amount of contaminants is selected between the two. The location for temperature considerations is important, as well. For heavier feedstocks, an increase in the temperature will lower viscosity and cause asphaltenes to go into the solution, thus extending filter life. However, this trade-off could result in a higher-cost design for the filters. While the filters may have a bypass line, this should only be opened in rare or abnormal conditions.

Additional important design and installation criteria include:

• Filter retention efficiencies consistent with the recommendations of the catalyst manufacturer or process licensor—usually 10 µm–20 µm at 99%–99.98% efficiency, depending on catalyst diameter and reactor packing density
• Ergonomically designed vessels that are duplexed (2 × 100%) for ease of changeout
• Vessel sizing to accommodate upset conditions to avoid bypass or rate reductions
• System sized on both flow and potential contaminant loading
  • No greater than 0.5 gpm/ft² flux rate to maximize cartridge dirt holding capacity (DHC)
  • Consideration of total suspended solids and particle size distribution
  • Consideration of ancillary contaminants, such as asphaltenes, as these can be captured by the filter
• Media tested according to industry standard ASTM F795-88 (1993), including single-pass, initial efficiencies to ensure consistency and performance integrity
• Pressure drop based on actual operating conditions
• Considerations for future throughput potential
• Media, support layers, component, and elastomer chemical and thermal compatibility
• Vessel design that has an optimized number and geometry of elements to minimize wasted space
• Filter element design with an optimum effective surface area.

Recent innovations such as finer fiber diameters, higher media porosity and greater gradient density have led to important benefits, including lower pressure loss, higher DHC, reduced particle capture size and improved efficiencies. However, the performance of traditional cylindrical filters can only be pushed so far. Much of this is due to the inherent limitations of their physical shape. Traditional cylindrical filters leave considerable dead space in a filter vessel, particularly when multiple filters are placed in a housing. A trapezoidal design minimizes the dead space and maximizes the effective filter media surface area in a pressure vessel (FIG. 3).

FIG. 3. A proprietary trapezoidal-shaped cartridged filtera (right) increases the usable surface area up to 169% vs. traditional cylindrical element designs (left).

By minimizing dead space, trapezoidal filters can provide up to 176% more effective surface area than cylindrical filters, depending on the configuration of the filter (TABLE 2). This results in lower pressure drops and longer online life. The density of the trapezoidal configuration also results in fewer filters to stock, change out and dispose. Additionally, the use of trapezoidal filters in place of cylindrical elements has the potential to reduce costs in areas such as shipping, storage and changeout, thereby presenting operating units with a significant overall cost reduction.

Greater surface area equals longer filter cartridge life. The increase in surface area has a direct impact on filter cartridge life. This is because a filter cartridge’s DHC is directly related to the amount of usable surface area in the cartridge itself. When the surface area is increased, the corresponding decrease in fluid flow per unit area drives up the DHC. By doubling the surface area, a filter system can increase filter life by up to four times, as illustrated in Eq. 1:

Le = (Ae / Ao)ⁿ              (1)

where:

Le = Extended filter life
Ae = Extended surface area
Ao = Original surface area
n = The ability of the solid to form a porous cake.

When various factors—cartridge life, reliable catalyst bed performance and uptime—are taken into consideration, a clearer view of the importance of filter selection begins to take shape. After several months of successful operations, one facility using the proprietary trapezoidal filters^a recorded measurable benefits in several key areas, including an 84% increase in effective surface area; a 400%–500% increase in filter life; lower direct consumable costs; fewer process upsets; and reduced labor, shipping and disposal costs.

Takeaway

For refiners that place a priority on minimizing downtime, optimizing performance, protecting the longevity of catalyst reactors and minimizing the overall lifecycle costs of their filtration systems, upgrading the quality and construction of the filter system represents a giant step toward achieving these goals. Proper feed filter filtration can prevent millions of dollars in lost production, along with direct losses related to material and labor costs.

An important step toward optimizing performance for the industry is to include ASTM F795-88 (1993) testing standards to determine cartridge filter efficiency. Furthermore, by requiring filter manufacturers to adhere to a recommended sizing specification of 0.5 gpm/ft² flow per unit area of media or less, operators can put all manufacturers on a level playing field and evaluate products accordingly.

As a result, a more accurate measure of true lifecycle cost can be evaluated, which incorporates the filter life (as determined by DHC) and ancillary costs, such as shipping, storage and changeout times. Furthermore, migrating away from existing cylindrical filters toward more innovative solutions (such as trapezoidal filters) can help operators significantly improve both value and efficiency. This will move them toward a new standard in overall filtration system performance. HP

NOTE

^a Filtration Technology Corp.’s Invicta® filter

The Authors

Related Articles

From the Archive

Comments

Comments

{{ error }}

Add Comment

{{ comment.comment.Name }} Editor • {{ comment.timeAgo }}

{{ comment.comment.Text }}