Hydrogen needs will increase two to five times as the world turns its attention to cleaning up engine exhaust, says Anthony J. Nagy of Muse, Stancil & Co., Dallas.
Nagy addressed the subject of fuel trends and hydrogen needs at Foster Wheeler USA Corp.'s Hydrogen Plant Conference, June 2-4, in Orlando.
The conference was attended by more than 100 people from 12 different countries. Drawing on knowledge from over 1 billion scfd of total installed hydrogen plant capacity, Foster Wheeler experts presented papers in the fields of steam reforming, partial oxidation (with all feedstocks, from natural gas to resids and coal), and steam reformer design.
Other industry specialists gave papers on refinery balances, markets, coal feedstocks, utility systems, and components for hydrogen plants.
Major topics of the conference included future refinery hydrogen needs and hydrogen management.
HYDROGEN NEEDS
Environmental regulations are dictating trends in fuel composition. Muse, Stancil used these trends and linear-program (LP) models to predict current and future hydrogen needs and balances for petroleum refineries in the U.S. and Western Europe.
The mounting environmental concerns of the 1980s culminated in 1990 with the passage of the Clean Air Act Amendments (CAAA) in the U.S. Western Europe entered the decade by introducing a reformulated fuel in Finland in 1991.
What conclusions can be drawn, based on the past 20 years' experience?
According to Nagy, public opinion will hasten the demand for cleaner fuels. California will continue to lead the reformulated fuel revolution and the remainder of the U.S. will follow. Western Europe, which has historically lingered behind the U.S., is reducing that time lag. Throughout the world, air pollution is increasingly being recognized as a serious problem and countries are examining steps to cleaner fuels. These steps will likely be patterned after changes already tested elsewhere.
REFINERY CHANGES
To determine the effect of fuel trends on hydrogen needs, refineries typical of Western Europe and the U.S. were modeled in a computerized LP.
These "average" refineries utilize a feedstock of 100,000 b/d of Arabian Light crude oil.
The process configurations shown in Figs. 1 and 2 were derived from reported average processing capabilities. 1 The configurations represent the respective averages for Western Europe and the U.S.
The ratio of premium-to-regular gasoline sales was limited to 1991 industry averages for the "current" refinery models.
The "future" LP models were constrained by an estimate of projected gasoline sales ratios.
For future refinery scenarios, processes were added and configured to meet fuel specifications expected for the year 2000.
WESTERN EUROPE
Table 1 shows typical current and expected future processing scenarios for a Western European refinery.
As the need for heavy fuel oil declines, vacuum distillation will increase to provide more feed for fluid catalytic cracking (FCC).
The role of the naphtha catalytic reforming unit will be expanded through increased capacity and severity at reduced operating pressure to provide high-octane blendstock to replace tetraethyl lead (TEL). An upper limit on reformer severity will be established by the availability and economics of methyl tertiary butyl ether (MTBE) for octane blending.
To meet reduced gas oil sulfur specifications, hydrotreating capacity will almost double, said Nagy. Cracked distillate that is included in the gas oil pool will require hydrotreating for stabilization, as well as sulfur and nitrogen removal.
The expected trend toward premium fuels will require investment in hydrocracking. Sulfur plant capacity will require significant expansion.
The current refinery's surplus of hydrogen will be used, and a supplemental hydrogen supply will be required for the future.
UNITED STATES
Some of the expected trends in Western European refining are already incorporated into the average U.S. refinery. Essentially all gasoline in the U.S. is unleaded, Rvp reductions are in place, and fuel-oil production is much lower than in Western Europe.
Table 2 summarizes the typical current and expected future U.S. refinery processes.
The need for vacuum distillation will decrease as atmospheric resid is desulfurized and cracked to lighter products. Isomerization and naphtha catalytic reforming capacities remain essentially constant as reformer pressure is reduced.
MTBE and tertiary amyl methyl ether (TAME) production is increased to utilize surplus light olefins as Rvp decreases and oxygenate requirements increase. Distillate hydrotreating increases by 88% and hydrocracking by 50% to provide higher-quality fuels.
Vacuum gas oil catalytic cracking declines and resid catalytic cracking increases. Both vacuum gas oil and resid feeds to the crackers require desulfurization. Sulfur-plant capacity doubles to allow almost complete removal of sulfur from all liquid products.
Supplemental hydrogen needs double as the current supplemental hydrogen sources are expanded.
PRODUCT MIX
The changes expected in Western European and U.S. refining configurations will cause a shift in the production ratios of various fuels. Table 3 shows the differential yields of products between the modeled current and future refineries.
Significant changes will occur in refining and marketing as Western Europe phases out TEL and the U.S. begins using reformulated fuels.
HYDROGEN BALANCE
Nagy also told conference delegates that the coming environmental revolution, combined with hydrogen-deficient crude oils and reduced demand for heavy fuel oil, will accelerate the demand for hydrogen to the turn of the century and beyond.
ln Western Europe, hydrogen demand is expected to increase at a compound growth rate of approximately 23%/year between 1992 and 2000.
The U.S. likewise will experience a substantial increase in hydrogen demand, which will more than double by the turn of the century. The compound growth rate is expected to be about 10%/year. Figs. 3 and 4 illustrate hydrogen supply and demand for the modeled Western European and U.S. refineries.
WESTERN EUROPE
The average refinery in Western Europe depends primarily on naphtha catalytic reforming for hydrogen production. The reforming capacity currently provides a surplus of hydrogen amounting to about 50% of reformer production. This surplus hydrogen is sent to the refinery fuel system.
The current demand for hydrogen is in naphtha, kerosine, and gas oil desulfurization. As Western Europe accelerates on its path of environmental concern, Nagy expects hydrogen demand to increase by a factor of five.
This demand will utilize all surplus hydrogen and require the addition of supplemental hydrogen sources. Hydrogen will be utilized further to desulfurize naphtha, kerosine, and gas oil with the gas-oil fraction receiving the most emphasis.
Hydrocracking--currently a minor process in Western Europe--will increase in importance and expand to produce higher-quality hydrogen rich fuels for an increasing motor-fuel demand. With the European Community (EC) and the rest of the world phasing out TEL, the need for octane via naphtha catalytic reforming will increase.
Nagy foresees that Western Europe will significantly expand naphtha reforming capacity to almost double its present size. This expansion will be in the form of low-pressure processes that produce higher ratios of hydrogen per barrel of feed than do currently installed processes.
Although additional low-pressure reforming will produce substantial amounts of hydrogen, it will be insufficient to satisfy the faster-rising demand.
UNITED STATES
The U.S. has essentially phased out the use of TEL for octane enhancement and, for the past 15-20 years, has faced a continuing reduction in the hydrogen content of crude oil.
Because gasoline is the motor fuel of choice, U.S. refineries have substantial naphtha reforming capacity, which meets about two thirds of current hydrogen demand, The other third is supplied by supplemental hydrogen production.
Pending environmental regulations will cause a reduction in reformer severity and feedstock availability as benzene and aromatics are controlled. Surprisingly, these changes will have very little effect on hydrogen supply from the naphtha catalytic reformer.
This fact is attributed to a reduction in reforming process pressure as older high-pressure units are phased out in preference of low-pressure processes that produce more hydrogen.
Future hydrogen needs are expected to double from the current level. This increase in demand will be divided almost equally between hydrocracking and hydrodesulfurization. To produce adequate supply, current supplemental production will have to be increased by a factor of about five.
The average U.S. refinery has some hydrocracking capacity. The hydrocracker is the largest consumer of hydrogen in the U.S. refinery, with more than half of produced hydrogen being consumed by the process. Hydrogen demand for hydrocracking is expected to grow by 50%. As the U.S. approaches the turn of the century, Nagy expects that a significant investment will have been required to expand desulfurization processes for all products. The LP model projects that future desulfurization will require five times more hydrogen than the current level.
Of course, many different factors come into play in assessing an individual refiner's needs, Among these is crude slate, which can contain crude types varying from sweet, paraffinic, Pennsylvania grade to the heaviest and most sour of the world's known crude oils.
In addition, each refiner has a unique process configuration. The processes may be the same, but the quantities, mechanical aspects, and interaction between process units are different.
Seasonal variation and geographical location are important factors that dictate a refiner's product mix. Each refinery must address these variations to determine the competitive solutions that will solve the problems of the environmental revolution.
Notwithstanding, hydrogen will play an important and ever-increasing role in every refiner's capability and flexibility.
HYDROGEN MANAGEMENT
John G. Pomfret, manager, chemical processing for Foster Wheeler, defined "hydrogen management" as the balance of hydrogen within the refinery, taking into consideration the sum of the hydrogen-producing units and the distribution of hydrogen to the hydrogen-consuming units.
In his presentation at the conference, Pomfret addressed four broad areas within this subject:
- The impact on hydrogen-producing units of likely changes to fuel specifications caused by current U.S. legislation
- The changes in refinery hydrogen consumption as a result of these changes in fuel specifications
- The effects of hydrogen purity on processing steps on processing steps
- The integration of refinery hydrogen use and the role of various methods of purifying hydrogen-rich streams.
Pomfret sees hydrogen management as an important criterion in future refinery planning. He contends that, as a result of the CAAA, hydrogen consumption will increase and hydrogen by-product production could decrease.
In a broad sense, the manufacture of environmentally acceptable products will also affect worldwide hydrogen production from petroleum refineries.
Changes involved in the increase in hydrogen demand will include benzene, olefins, aromatics, and sulfur reduction in the gasoline pool; aromatics and sulfur reduction in the diesel pool; and sulfur reduction in fuel oil.
One of the first steps in addressing these changes is to develop a strategy to manage hydrogen within the refinery.
HYDROGEN COPRODUCTION
The primary source of hydrogen within the refinery is the catalytic reformer. This unit could diminish in importance as a source of hydrogen because:
- Gasoline reformulation somewhat de-emphasizes the octane-production role of the reforming unit because of the use of MTBE to meet the requirement of 2% oxygen in the gasoline pool.
- The benzene balance in the refinery will need to be managed to meet 1 LV % maximum benzene in the gasoline pool, possibly leading to lower capacities on reformer units (caused by removal of benzene precursors). Alternately, lower-severity reaction conditions could produce less aromatics and increase reformate yield while reducing hydrogen production.
- Any reduction in the 90% distillation point (T90) would require lowering the end point of the feed to the reformer, with consequent reductions in the quantity of reformer feed, and thus hydrogen coproduction.
HYDROGEN USE
The potential effects of changing refinery operations to meet the CAAA could increase hydrogen demand substantially through:
- An increase in hydrotreating requirements to meet tighter sulfur specifications. Individual refiners may decide to limit sulfur in gasoline to meet the requirements for reductions of volatile organic compounds (VOCs).
- The addition of benzene and benzene precursors--removed from reformer feed by prefractionation--to the feed to an isomerization unit.
- The reduction of T90 to reduce exhaust VOC levels. Available options to reduce VOCs, such as hydrocracking the heavy end of FCC gasoline, would further increase refineries' hydrogen needs.
- The production of cyclohexane via saturation of excess benzene in the gasoline pool.
These scenarios also add to the potential increase in hydrogen demand for processes such as hydrocracking, which lead to lighter products.
Pomfret says each refiner should analyze the impact of these issues on the hydrogen balance within his refinery and develop a management plan that best meets his individual needs. Any changes to a specific refinery scheme will be constrained by cost, refining hardware, market factors, and feedstock options.
HYDROGEN PURITY
The purity of makeup hydrogen significantly affects the design and cost of hydrogen processing units, even when the purity difference may appear small, e.g., 97 vol % vs. 99.9 + vol %. Consequently, it is necessary to consider hydrogen makeup an integral part of the hydrogen management system.
Influences of hydrogen purity on the overall system are:
- Lower-purity hydrogen makeup introduces more inerts into the hydroprocessor. These inerts must be rejected from the system either by purging from the hydrogen recycle loop or by removal as dissolved gases.
- Losses of hydrogen increase with purges, as purge gas contains 1-10 moles H2/mole inert. Consequently, there is a compounding effect, with lower purities increasing hydrogen makeup.
Foster Wheeler completed a study of the effect of hydrogen purity on hydrocracking characteristics and design for two purities: 2
Case 1-97%: This was a typical purity of hydrogen product before the advent of pressure-swing absorption (PSA) units and is in the range of modern membrane-based purification systems.
Case 2-99.9%: This purity can typically be expected from PSA hydrogen-production facilities.
For the study, a mild hydrocracker was used, with feed characteristics of 22,000 b/sd throughput capacity, 19.0 API, 3.0 wt % sulfur, 2,500 ppm nitrogen, and 35 vol % conversion.
Major process parameters, such as yield structure, reactor and flash gas temperatures, chemical hydrogen consumption, and hydrogen partial pressure, were assumed to be constant.
The purge rates and reactor-outlet pressure were changed as required to maintain hydrogen partial pressure.
Table 4 shows a comparison of Cases 1 and 2, with a sensitivity of Case 1 (Case 1A). In Case 1A, hydrogen partial pressure and hydrogen makeup requirements are maintained the same as in Case 2 by an increase in the operating pressure of 13%.
Incremental investment costs (Table 5) are caused by:
- Case 1--Increase in makeup compression equipment size
- Case 1A--Increase in design pressure of equipment and higher makeup compression-discharge pressure.
Incremental operating costs are caused by:
- Case 1--Increase in makeup hydrogen flow
- Case 1A--Increase in operating pressure for feedstocks.
Table 5 demonstrates the importance of considering hydrogen purity as an integral part of hydroprocessor design. The direct use of catalytic reformer offgas as a makeup (typically 70-90 mole % hydrogen) would only increase the effects shown in Table 5.
MAKEUP HYDROGEN
Recent studies have assessed the various methods for hydrogen recovery from refinery gas (primarily for reformer offgas and purge streams from hydroprocessing units). These studies have involved a comparison of technologies for hydrogen recovery from refinery offgases within an integrated refinery system.
Various purge gas streams originated from process units such as gas oil hydrotreaters, diesel hydrotreaters, and isomerization units. These streams were considered for recovery of hydrogen by either PSA or membrane-type systems.
From these analyses, certain general statements can be made:
- The effects of hydrogen makeup purity should be assessed on individual consuming units as described because membrane units typically produce a less-pure hydrogen product than do PSA units.
- Mechanisms for recycling hydrogen to users should be assessed, as membrane plants produce lower-pressure hydrogen product than do PSA plants, introducing a potential penalty for capital and operating (compression energy) costs.
- The fuel-gas system that could receive the tail gases from the units should be assessed, particularly from a minimum acceptable operating pressure point of view. This assessment allows the maximum available driving force to be considered for PSA unit design and any penalties for tail-gas purge recompression and/or loss of PSA unit hydrogen recovery to be evaluated.
- Any requirement of potential pretreatment of feed offgases should be assessed because of possible sensitivity of membranes to the presence of saturated impurities (such as water, H2S, aromatics, and paraffins) close to their dew points.
Balancing the disadvantage described previously for membrane-type systems is:
- Greater potential hydrogen recovery (for example, 95% vs. 88%)
- Lower capital cost of the unit at smaller (
- Greater tail-gas pressure leaving the unit.
Each refinery scheme has unique aspects that could affect the decisions as to the most appropriate purification schemes. These aspects include:
- Impurities in the offgases that can affect feed pretreatment requirements
- Existing compressor availability for conversion to duty as hydrogen/purge gas booster compressors
- Pressure balance to match requirements of end users
- Lower-pressure fuel gas systems (for example, 10 psig) employed
- Hydrogen streams available from multiple sources that, from a layout and characteristics point of view, could be considered alone, together, or grouped according to components and pressure.
Depending on these factors, both membranes and PSA units have been chosen for similar typical duties as the most applicable for a particular situation of offgas purification.
Once the physical options for the hydrogen management system have been conceptually established, qualitative analyses should be completed, including:
- Identification of any likely changes in the quality of hydrogen-rich streams, caused by process variances
- Turndown requirements
- Sensitivity of downstream processes to hydrogen purity
- Likely future expansion needs.
Pomfret summarized the advantages and disadvantages of membrane vs. PSA units for the following important qualitative considerations:
Feedstock conditions. The PSA unit is the most flexible process in terms of its ability to maintain purity and recovery under changing conditions. The membrane unit can maintain product purity for small feed composition changes, but maintaining product purity can significantly affect hydrogen recovery.
Turndown. Although both PSA and membrane systems have good turndown characteristics, it is critical that turndown requirements be analyzed. For example, in times of protracted turndown, one tactic for membrane systems is to isolate one of the modules. Consequently, one consideration of design is the number of modules in the system.
Expansion. Because of their modular nature, membrane systems are ideally suited for expansion. It is possible to expand PSA units, but considerations for expansions can lead to added cost in the initial phase of the project.
Product purity. Product purity has an impact on consuming process units. The 99 + mole % product from PSA units has advantages, particularly for hydroprocessor installations.
CRYOGENIC PROCESSES
Cryogenic processes also have been considered during studies, particularly where:
- By-product recovery, such as olefins recovery for integrated petrochemical operations, is to be considered within the hydrogen recovery scheme
- Larger (50 MMscfd) hydrogen-recovery rates are available
- Lower hydrogen content (30-75 vol %), high-pressure, large volume streams are available
- Greater hydrogen recovery (92-97%) is required.
During the analysis of cryogenic systems, the feed compositions of components that potentially freeze (such as CO2, heavy hydrocarbons, and benzene) should be assessed in detail to ascertain whether pretreatment steps are required.
REFERENCES
- Thrash, Lou Ann, Worldwide Refining Survey, OGJ, Dec. 23, 1991, p. 39.
- Miller, Marvin H., Lacatena, Jerome J., and Miller, Geoffrey Q., "Hydrogen for Hydroprocessing Operations," National Petroleum Refiners Association annual meeting, 1987.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
