The Hydrotreating Process Part 2

Process Variables

During hydrotreating operation at different scales (laboratory, microreactor, bench scale, pilot scale, and commercially), there are four main process variables that have a great impact on reaction conversion and selectivity as well as on the activity and stability of the catalyst: (1) total pressure and hydrogen partial pressure, (2) reaction temperature, (3) H2/oil ratio and recycle gas rate, and (4) space velocity and fresh feed rate. Typical operating conditions for the hydrotreating of various petroleum fractions are given in Table 3.1.

TABLE 3.1. Typical Operating Conditions and Hydrogen Consumption During the HDT of Various Feeds

Type of Feed	Temperature (°C)	Pressure (psig)	LHSV
Naphtha	280-425	200-800	1.5-5.0
Gas oil	340-425	800-1600	0.5-1.5
Resid	340-450	2000-3000	0.2-1.0
			H2 Consumption
	Source	API Gravity	(scf/bbl)
Naphtha	—	—	100-700
Gas oil	—	—	300-800
Resid	—	—	500-2000
AR	Venezuela	15.3-17.2	425-730
VR	Venezuela	4.5-7.5	825-950
AR	West Texas	17.7-17.9	520-670
VR	West Texas	10.0-13.8	675-1200
AR	Khafji	15.1-15.7	725-800
VR	Khafji	5.0	1000-1100
AR	Kuwait	15.7-17.2	470-815
VR	Kuwait	5.5-8.0	290-1200

Total Pressure and Hydrogen Partial Pressure The total pressure of a hydrotreating unit is determined by the reactor design and is controlled by the pressure that is maintained at the high- pressure separator (HPS). Inlet or outlet hydrogen partial pressure is calculated by multiplying the total pressure by the hydrogen purity (H2 mole fraction) of the recycle gas. Definition of the value of total reactor pressure is decided depending primarily on the nature of the feed and the amount of impurities to be removed (i.e., the quality of the feed and the quality of the product desired). In general, when a hydrotreater is operated at high hydrogen partial pressures, the following main effects are obtained (Mehra and Al-Abdulal, 2005; Gruia, 2006):

• Longer catalyst cycle life

• Capability for processing heavier feeds

• Higher throughput capability

• Higher conversion capability

• Better distillate quality

• Purge gas elimination

Since the catalyst deactivation rate will be increased substantially and catalyst cycle life reduced at low reactor pressure due to coke formation, a hydro-processing reactor must be operated at H2 partial pressure very close to the design value. Although it is highly desirable to operate a reactor at the highest allowable pressure, equipment limitations restrict the operation at a pressure close to or a little bit higher than the design value. Given this situation, the only way to increase hydrogen partial pressure is by increasing the purity of the recycle gas, which can be achieved either by increasing the H2 purity of the makeup hydrogen, or by venting gas off the HPS or reducing the temperature of the HPS (Gruia, 2006).

At higher H2 partial pressures, the removal of impurities is easier; however, reactors become more expensive and hydrogen consumption increases, which can become a significant cost factor for the refinery. New units are being designed to operate under higher H2 partial pressure atmosphere by working at higher total pressure.

The performance of any hydrotreating reactor and process is limited by the hydrogen partial pressure at the inlet to the reactor. The higher the hydrogen partial pressure, the better the hydrotreating reactor performance. The overall effect of increasing the partial pressure of the hydrogen is to increase the extent of the conversion (Speight, 2000). This has been confirmed extensively by studies conducted with model compounds for HDS, HDN, HDA, and so on, reactions as well as with real feeds (light distillates, middle distillates, heavy oils, etc.) at microscale, benchscale, and pilot plants. Figure 3.5 exemplifies the effect of hydrogen partial pressure and reaction temperature on sulfur removal and saturation of polyaromatic hydrocarbons (PAHs) (Binghan and Christensen, 2000- Chen et al., 2003). It is clearly seen that PAHs react quite readily, but their conversion is thermodynamically limited, and neither increasing the temperature nor increasing the H2 partial pressure reduces the PAH content in the product to values lower than 2wt%.

The presence of heteroatom compounds with different reactivities in a hydrotreating feed makes, for example, HDS of refractory multiring sulfur compounds very difficult, with a high hydrogen demand, a pathway that first goes through prehydrogenation of one of the aromatic rings. If H2 partial pressure is not at the value required, the following problems will be faced during hydrotreating (Ho, 2003 ):

• The slow HDN rate of nitrogen compounds blocks off virtually all active sites that are available for HDS.

• The HDS rate of refractory sulfur compounds may be limited by a ther-modynamically mandated low hydrogenation rate.

• The catalyst surface may be starved or adsorbed hydrogen.

In commercial operation, hydrogen partial pressure is obtained primarily by feeding the proper amount of makeup gas. The increase in catalyst activity for achieving higher impurities removal and conversion rates would require significant modifications in hydrotreating reactor operation, primarily through the use of higher pressure and also by increasing the hydrogen rate and purity, reducing the space velocity, and proper selection of catalyst. Pressure requirements would depend, of course, on the feedstock quality and on the product quality target of each refinery. Also, if all aromatics need to be hydrogenated, a higher pressure is needed in the reactor compared with that of a conventional operating mode. The level of pressure required for such product specifications will be limited by the cost and availability of the technology.

Figure 3.5. Effect of H2 partial pressure on sulfur removal and aromatic saturation.

Reaction Temperature Reactor temperature generally determines the types of compounds that can be removed from the petroleum feed and also establishes the working life of the catalyst. Increasing the temperature increases reaction rates and thus the removal of impurities. However, similar to reactor pressure, there are limits to the maximum allowable temperature, since depending on the feed above a certain thermal cracking value of the hydrocarbon constituents becomes more prominent, which can lead to the formation of considerable amounts of low- molecular- weight hydrocarbon liquids and gases, and also to catalyst deactivation much more quickly than at lower temperatures. Thermal cracking also produces olefins, which when hydroge-nated, release heat, increasing the temperatures further as well as the thermal cracking rates (hot spots). Finally, this condition inside the reactor provokes temperatures higher than the safe upper limits for the reactor walls (Speight, 2000). The effect of reaction temperature on impurities removal is illustrated in Figure 3.5 together with that of pressure.

Most hydrotreating reactions are exothermic in nature, which makes the commercial reactor temperature increase as feed passes through the catalyst bed. For experimental reactors (e.g., micro- and bench-scale reactors), an isothermal condition is commonly achieved, but for adiabatic (commercial) reactors, the outlet reactor temperature will be higher than the inlet reactor temperature. To determine the average temperature of adiabatic reactors, the weight-average bed temperature (WABT) is typically used. WABT can easily be calculated if the reactor is provided with various temperature indicators (TIs) located in different zones of the catalytic bed, by the following equations (Stefanidis et al., 2005):

where WABT; is the average temperature of each catalytic bed between two TIs and T/n and Tout are the inlet and outlet temperatures in each catalytic bed, respectively. The global WABT is calculated with

where N is the number of catalyst beds and Wct is the weight fraction of catalyst in each bed with respect to the total.

Equation (3.1) is used instead of a common arithmetic average since it takes into account the common nonlinear gradient of temperature observed in hydrotreating reactions. In Eq. (3.1) it is assumed that in the last two-thirds of the reactor length the temperature is closer to Tout, while in the first one – third of the reactor length the prevailing temperature value is closer to Tin.

WABT is frequently used during operation for process control purposes; in such cases Eqs. (3.1) and (3.2) are preferably expressed as only one linear equation as a function of TI values as follows:

where at are constant values determined by solving Eqs. (3.1) and (3.2). Note that the sum of all at values must be equal to unity. Equation (3.3) can easily be programmed into the process control system and the WABT can be reported in real time.

Figure 3.6. Life of catalyst and required increase in WABT for hydrotreatment of various feeds.

A common practice during commercial operation of hydrotreating units is to increase the reactor temperature steadily to compensate for catalyst deac-tivation and to produce a constant- quality product. This production policy requires the unit to be operated at a different WABT value during time on stream, which are known as start-of-run (SOR) temperature (WABTSOR) and end – of – run (EOR) temperature (WABTEOR). In fact, when designing a hydrotreating plant, simulations have to be carried out for at least two cases: at SOR and EOR conditions. Properties of the feed, the desired quality of the product, and the reactor design are the main parameters that define the values of WABTsor and WABTeor as well as the temperature increase during opera-tion.Typically,WABTEOR – WABTSOR = 30°C. High-metal-content feeds require the temperature to be increased more frequently, of course. When WABT reaches a value close to the maximum designed, the catalyst has to be replaced. Figure 3.6 summarizes the typical increases in WABT and catalyst life, depending on the type of feed. For the HDS of naphtha, a long catalyst life is observed and the increase in WABT over time is unimportant. However, during the HDT of heavy oils, WABT has to be increased constantly so that catalyst deac-tivation is compensated and the product is produced at constant quality.

H2/Oil Ratio and Recycle Gas Rate The H2/oil ratio in standard cubic feet (scf) per barrel (bbl) is determined by Another unit frequently used to report the H2/oil ratio is m3/bbl, obtained by multiplying the H–oil ratio (in scf/bbl) by a conversion factor (0.028317). A

where MWoil and MWH2 are the molecular weights of the oil to be hydrotreated and of hydrogen, respectively, and poil and pH2 are the densities of the oil and hydrogen, respectively (pH2 at 15°C and 1 atm is 0.0898kg/cm2).

Apart from economic considerations, gas recycle is used to compensate for hydrogen consumption and hence to maintain the hydrogen partial pressure within the reactor. Use of a high excess of hydrogen (i.e., an elevated H2/oil ratio) ensures adequate conversion and impurities removal due to efficient physical contact of the hydrogen with the catalyst and hydrocarbon; also, carbon deposition is minimized, which reduces the rate of catalyst deactiva-tion. The latter is actually the main reason to work in a high-hydrogen-concentration atmosphere; otherwise, the catalyst can deactivate faster, due to coking. Another important benefit of operating at high H2 partial pressure is the reduction in the SOR temperature of the reactor, which increases the cycle life of the catalyst. However, there is a limit in the value of the H;/oil ratio, since above a certain gas rate, the change in hydrogen partial pressure will be relatively small and no further benefits will be obtained. In fact, higher gas rates than necessary incur extra heating and cooling rates, which may become more important than other advantages.

Increasing the recycle-gas rate increases the H2/oil ratio and the hydrogen partial pressure in the reactor. Apart from this, the objective of the gas rate is to strip volatile products from the reactor liquids, thus affecting the concentration of various components in the reactive liquid phase. The H2 partial pressure and H2/oil ratio must be maintained very close to the design value; otherwise, the catalyst life will be affected adversely.

The hydrogen loop in a hydrotreating unit involves several streams, as can be seen in Figure 3.7. The reactor effluent stream is separated in a high pressure separator (HPS) into liquid hydrotreated products and noncondens-able H2-rich gases (typically, 78 to 83 mol% H2 plus H2S and other light gases, CH4, C2H6, C3H8, butanes, traces of pentanes). Hydrogen sulfide, either formed via HDS or present in the reactor feed, is commonly separated with an amine contactor to increase the purity of the H2. Following use of an amine contactor, lighter hydrocarbon gases are still present in the recycle gas stream; part of this stream (10 to 15%) is purged to the fuel gas system or sent to a hydrogen purification process (e.g., PSA: pressure swing adsorption) if the unit is provided with such a plant, from which about 20% is lost to the fuel gas system. The other portion of gases leaving the amine contactor is compressed and recycled to the top of the reactor and/or used for quenching (80 to 85mol% H2). The separated H;-rich gases from the hydrogen purification process are mixed with the makeup hydrogen and recycled to the top of the reactor.

Molar H2/oil ratio can also be calculated from the volumetric H;/oil ratio by means of the following equation:

Figure 3.7. Hydrogen loop in a hydrotreating unit.

Depending on the source of makeup hydrogen, it is typically available at 96 to 99.9 mol% H2 purity. The separated hydrotreated liquids from the HPS are sent to other sections for further processing (Mehra and Al-Abdulal, 2005).

When working under ultralow- sulfur fuel production conditions or with high-sulfur feedstocks, the concentration of H2S in the recycle gas stream can achieve high values, which consequently reduces the hydrogen purity of the recycle gas stream and hydrogen partial pressure. This highly concentrated H2S atmosphere inhibits the HDS reaction, as shown in Figure 3.8. According to various authors, HDS activity is reduced around 3 to 5% for each 1 vol% of H2S in the recycle gas stream, which means roughly that an increase of 3 to 5% in the catalyst amount is needed to balance this situation. Even a low H2S content results in an increase (e.g., 0.3mol% can reduce the reaction rate about 5%). The operation of HDS units with 9% H2S content in the recycle gas stream would require about 15 to 20% more catalyst to achieve the same results than when the H2S concentration is 0%. In addition, when H2 purity is increased, the SOR temperature can be lowered about 9°C, and the run length can be extended about 30%.

The recycle gas from the HPS is generally water-washed to remove ammonia, preventing the formation of ammonium sulfide, which might form blockages in the reactor effluent cooler, and is then sent to the sour water plant to remove H2S. If a scrubbed recycle gas is not available, the reactor temperature must be increased to offset the H2S inhibition, the effect of which is greater at higher total reactor pressure.

Hydrogen consumption is another very important process parameter, since it determines the amount of hydrogen makeup. Hydrogen consumption during hydrotreating depends on feedstock properties and impurity removals. As a feed becomes much heavier it will require substantially more addition of hydrogen to reach the product quality desired. Table 3.1 shows typical hydrogen consumption values for hydrotreating of different hydrocarbon feeds.

Total hydrogen consumption is a summation of chemical hydrogen consumption and dissolved hydrogen (calculated from vapor-liquid equilibrium), assuming that any mechanical hydrogen loss is negligible. The most common approach to calculating H2 consumption not only at the commercial level but also at different experimental scales is by means of a hydrogen balance in gas streams. That is the amount of hydrogen entering the reactor minus the amount of hydrogen exiting the reactor. Another way to do so is by determining the hydrogen content in the liquid feed and products. Liquid products, which have been hydrogenated, must have a higher hydrogen content than that of liquid feed. The difference is the hydrogen added to the feed (i.e., hydrogen consumption). There are also rules of thumb for quick calculations, which employ typical hydrogen consumption values reported in the literature (Edgar, 1993; Speight, 1999), but the values obtained must be taken with care due to their empirical nature, yielding only approximations to the real amount of hydrogen consumed.

Space Velocity and Fresh Feed Rate Space velocity is a process variable normally used to relate the amount of catalyst loaded within a reactor to the amount of feed. Space velocity in normally expressed on a volume basis poil and pcat are the densities of the hydrocarbon feed and the catalyst, respectively. When using LHSV as a process parameter, pcat is not important. However, in the case of WHSV, pcat becomes relevant since it can vary depending on how the catalyst was loaded to the reactor. For example, for dense loading, more catalyst is loaded in the same reactor volume and the WHSV value will be different than with nondense loading, although the LHSV value will be the same in both cases.

Figure 3.8. Effect of H2S on product sulfur content during HDT of middle distillates over

In certain cases, space velocity is also used as the GHSV, which is calculated as follows:

In a hydrotreating process, the space velocity is used as the LHSV. The space velocity is inversely proportional to the residence time. Therefore, an increase in space velocity indicates a decrease in residence time and thus in reaction severity. Figure 3.9 shows the influence of LHSV on the sulfur content of products obtained during the hydrotreating of different middle distillates. It is clearly observed from this figure that a decrease in LHSV results in diminished sulfur content in the product. Operating at a higher space-velocity value (a higher feed rate for a given amount of catalyst) requires a higher reactor temperature to maintain the same impurity removal (i.e., product quality), resulting in an increased deactivation rate, thus reducing the catalyst life.