Refiners gave heavy attention to hydroprocessing operations at the most recent National Petroleum Refiners Association annual question and answer session oil refining and petrochemical technology.
Among the topics covered were diesel color, blending to meet diesel sulfur specs, and ammonia injection in hydrocracking units. The panelists (see box) also related their experiences with increasing vacuum gas oil conversion in hydrocracking operations.
The 1992 NPRA Q&A Session was held Oct. 14-16, in Anaheim, Calif. For details on this meeting's format, see OGJ, Feb. 22, p. 45.
DIESEL COLOR
In view of more stringent specifications expected for diesel fuel, future refining would require severe hydrotreating where some color degradation may occur. What are the proven technologies and catalysts already in service for producing diesel fuel meeting future specifications?
Malik: At our Corpus Christi refinery we operate a mixed-distillate hydrotreater at a space velocity of 2.2/hr and a total pressure of 900 lb at the separator. We are currently producing a diesel boiling range product at 800-1,000 ppm of sulfur and a kerosine product of less than 100 ppm of sulfur. We do not believe that severe hydrotreating would be necessary for us to meet the future 500 ppm of sulfur specification.
The key to avoiding color degradation is to keep the space velocity low by design, increase hydrogen partial pressure, keep recycle high, and use a high hydro-denitrogenation (HDN) activity catalyst. By operating at lower temperatures at low space velocity, the onset of color degradation is delayed. We currently have an all nickel-molybdenum catalyst in our reactor which aids in color-body removal. Additionally, by operating at lower temperatures, we prevent color degradation by avoiding the recombination reactions due to cracking at the higher temperatures, and the removal of any chemical or corrosion inhibitors in the feed.
The hydrotreating technology for producing low-sulfur diesel is available from several process licensors and engineering and construction type companies. Likewise, a high-activity hydrotreating catalyst is available from all vendors in the hydrotreating catalyst business.
Quinn: Amoco feels the following can be used to meet future diesel fuel specifications:
- Limit the amount of high boiling components in the feed. Studies have shown that these are especially difficult to desulfurize when they boil above 600 F.
- Operate the hydrotreater at low temperature to prevent color degradation.
- Reduce space velocity while maintaining high hydrogen partial pressure throughout the reactor.
- Use either nickel-molybdenum or nickel-tungsten catalysts to meet possible aromatics limits.
Warwick: Many refiners will probably have difficulty meeting the more stringent diesel specifications without some kind of increased severity capability built into their units. The reactor temperature difference between 0.3 and 0.05 wt % sulfur is about 60 F. on many feedstocks. There are better catalysts available today, but not to compensate for 60 F.
We assume many refiners will run with shorter cycles or add capital to meet the future requirements. If they have not already done so, it is getting late to meet the 1993 deadline. NPRA paper AM-91-39 contains a very good discussion of diesel fuel color.
Lavergne: I want to emphasize that whatever you do, you need to keep the temperature requirements at the start-of-run the same or less, otherwise you are going to have a shorter cycle.
George: We have conducted extensive pilot-plant testing using a wide range of feedstocks and operating conditions including:
Feedstocks:
Straight-run diesel
100% light cycle oil
(LCO)
Blends of straight-run
diesel/LCO/CLGO
Kerosine
Feedstocks studied had sulfur varying between 0.4 to 1.5 wt % and end point between 540 F. and 725 F.
Operating conditions:
Liquid hourly space velocity (LHSV) = 0.5-2.5 hr-1
Hydrogen partial pressure = 400-900 psig
Weighted average bed temperature (WABT) = 530-720 F. with the lower temperatures for kerosine hydrodesulfurization (HDS).
Catalyst systems include high activity CoMo and high activity NiMo plus stacked beds.
I should know better than to generalize, but the following general remarks emerge:
- Reducing the product sulfur from 0.14 to 0.05 wt % requires an additional 40 F. operating temperature for a feed sulfur of 1.3 wt %.
- For a blend of straight-run diesel and LCO with a sulfur of 1.3 wt % and end point of 680 F. a WABT of 645 F. is required to achieve 0.05 wt % sulfur in the product, other unit conditions being LHSV = 1.0 hr-1 and unit pressure of 450 psig.
- Lowering the feed end point will significantly reduce the operating severity requirement. For example, lowering the feed end point to 620 F. from 680 F. would lower the WABT by 30 F.
- Increasing hydrogen partial pressure beyond a minimum level has a marginal impact on catalyst HDS activity but significant impact on catalyst stability and activity for HDN and aromatic saturation.
- Reducing the LHSV by adding catalyst volume has a significant impact on catalyst activity and stability. For example, lowering the LHSV from 2 hr-1 to 1 hr-1 will reduce the operating temperature by about 30 F.
- Kerosine with an end point of 540 F. can be desulfurized to 0.05 wt % at 530 F. and LHSV = 1.5 hr-1.
For certain feedstocks and operating conditions, the application of NiMo or stacked bed NiMo/CoMo catalysts can meet 0.05 wt % sulfur product requirement. The products from the isothermal pilot plants were tested for color. Product color was largely
Due to a large delta temperature obtained commercially when processing cracked stocks, isothermal pilot plant testing may not give a representative picture of commercial units. Commercially, color deterioration may be observed at lower WABT than 720 F. since the bottom part of the catalyst bed could be operating at temperatures in the range of 720-740 F., especially when processing LCO.
Greg Dearwater (Profimatics Inc.): Color degradation of diesel fuels results when hydrotreater catalyst temperature is raised beyond the threshold temperature. Until that point is reached, catalyst temperature increases have little or no effect on the product color. Obviously refiners, in trying to meet more stringent sulfur targets, may approach or reach the temperature.
Possible solutions are running as high a hydrogen partial pressure as possible, cutting feed rate to lower temperature, using a catalyst with a higher hydrogenation activity, and with a multibed unit, minimization of the peak temperature in any bed. In attaining low aromatic contents in diesel streams, the conditions necessary for high aromatic conversion-for example higher pressures, higher saturation activity catalyst-should directionally be those that minimize color degradation.
DIESEL BLENDING
In the future, diesel blending will become more critical, as well as downtime of hydrotreating units. At 0.05 wt % sulfur in the blend, what will be the target sulfur level in the hydrotreater and how can refiners cope with blending sulfur at 0.05 wt %?
Cabrera: Clearly, as all of you know, the problem is not only the sulfur, but you are asked to make low-sulfur diesel at 0.05 wt %, make a 45 minimum cetane index and also reduce the aromatics.
At least at this time the current regulations only apply to high-speed transportation, automotive diesel. Perhaps not all the diesel pools or diesel products will be affected. This is the case today but it may change.
We have heard a lot about the variables which are involved. We have concluded, however, that there is no defined correlation for everybody. I have a few comments on what UOP thinks will determine the severity of desulfurization.
Clearly the biggest contributors-both from a type of sulfur and the distribution of the sulfur in the boiling range-are the fluid catalytic cracking (FCC) light cycle oil, which has been addressed as far as reducing end point is concerned, and stocks derived from thermal processing operations such as coking and visbreaking.
Unless you have a very unusual crude situation where the native sulfur in the diesel is exorbitantly high and also carries an unusual boiling-range distribution, the severity of the hydrotreater is going to be determined by the amount of light cycle oil and/or thermal stocks you are going to blend into the diesel pool.
Deeper desulfurization certainly can be achieved in the existing hydrotreaters as several of the panelists have eluded to, and clearly the key variable is the space velocity, which either means taking a capacity penalty or adding equipment and/or catalyst. Fortunately, as better catalysts continue to appear, their use can provide deeper levels of desulfurization without significant degradation of the diesel properties.
The cetane index and aromatics reduction requirements are the toughest issues. Our feeling is that there will be very little cetane improvement that can be achieved at the very severe conditions required to achieve deep levels of desulfurization.
George: Most refiners are targeting hydrotreated sulfur at 0.05 wt %. However, a number of refiners are designing units to achieve 0.04 wt % sulfur or even lower to cope with the anticipated blending requirements associated with future needs to process higher-sulfur crudes. In this way, they are designing some future blending margins into the refinery.
Lavergne: I will address the target-sulfur part of the question. The precision of our existing X-ray sulfur test is approximately 0.01 wt % sulfur. We feel this can be reduced to probably 50 ppm by counting the ionic responses in the test for a longer period of time.
Combined with the use of statistical process control or equivalent techniques on the hydrotreater process, we think this will lead refiners to target for 0.04 to 0.045 wt % sulfur on hydrotreated products. Higher product sulfur targets, of course, will lead to longer run lengths on the unit due to the lower start of cycle temperatures.
Malik: For our 900 psi mixed-distillate hydrotreater operating at 2.2 liquid hourly space velocity, we expect the fractionator bottom, which is a No. 2 oil product, to run down at about 600 ppm total sulfur. The kerosine product will contain less than 100 ppm of sulfur.
The two, when run in a common line to storage tank, will have less than 450 ppm of sulfur. This will present very minimal blending opportunities and will provide a cushion on the sulfur specification to work out the new operating philosophy for regulatory compliance. We may revisit our operating philosophy when the October 1993 specifications for on-road diesels become effective.
Warwick: We will probably target for 0.04 wt % sulfur. The hydrotreaters will have to be monitored in much greater detail than before. Operators and engineers will need to be well-schooled to make unit changes quickly for units with large feed changes.
On-line sulfur analyzers will be a must. Their accuracy at the low levels are not very good; however, they can detect trends which operations can respond to. As far as blending, refiners will just have to cope. Tankage will have to be made available for slops. More efficient operation will be a must.
Robert Shepard (Setpoint Inc.): This is a general question to the panel or to the floor. Is sulfur recombination associated with processing olefinic feeds in these diesels to achieve this low-sulfur level an issue? Do you need to have a primary reactor to saturate the olefins followed by a higher level of hydrotreating in order to get the sulfur out?
George: I believe if you were to use, for instance, a nickel-molybdenum or cobalt-molybdenum catalyst, that over the conditions we are discussing, most olefins would have disappeared. So I would not expect significant recombination of the olefins.
Billy Jarratt (Diamond Shamrock Inc.): Is there any comment or clarification on what the regulations are going to be? Are we going to have a quarterly sulfur bank to deal with or not?
Higgins: I can comment on the off-highway diesel. The U.S. Environmental Protection Agency recently initiated activity to look at regulating off-highway diesel. At this point they determined they would let it go for a few years. They want to see what the experience is on the on-highway program. I think off-highway regulations are a good 5 to 6 years away.
HYDROCRACKING
What are the measures taken to achieve total conversion of vacuum gas oils without causing any problem due to buildup of polynuclear aromatics (PNAs)? What analytical methods are available to the refinery to minimize fractionator bottoms purge rate? Do adsorbents work? If so, is the adsorbent regenerable?
Cabrera: Polynuclear aromatics are an inherent product of the hydrocracking reactions. Large polynuclear aromatic molecules which UOP calls HPNAs, precipitate and create serious problems in the cold section of the unit.
The analytical method used to track and quantify these molecules is a fluorescent spectroscopy method. UOP uses a fluorescent spectrometer with a proprietary data analysis procedure. The color of the recycled liquid can also be used to give a rough indication of whether the HPNAs are present and is a method used for qualitative-type comparisons only.
Various methods are practiced commercially to control HPNA buildup. However, total conversion of a heavy vacuum gas oil feedstock, while maintaining absolute control of HPNAs, is only possible with UOP's proprietary HPNA adsorption technology. Basically, the system consists of an adsorbent that selectively removes the HPNA from the streams in question and therefore prevents the plugging of the cold section of the unit and allows the unit to operate at full conversion. The adsorbent we use is not regenerable.
George: To increase conversion of vacuum gas oils and hydrocrackers, the following methods should be considered to control and minimize buildup of PNAs:
- A bleed stream should be sent to such units as FCC which will enhance FCC conversion to gasoline.
- Routing of the bleed stream back to the vacuum flasher to remove the very heavy molecular weight PNAs.
- Avoid operating the pretreat reactor outlet at high temperatures and at the wrong side of aromatic equilibrium curve. Take all possible measures to improve hydrogen partial pressure, e.g., high treat-gas rate and purity and high operating pressure.
- Application of a deep hydrogenation catalyst in the high pressure hydrocracker reactor loop to fully saturate polynuclear aromatics. Zeolyst Enterprises offers Z704A as a deep aromatic saturation catalyst.
Tools available to refiners for measurements of PNAs include ultraviolet (UV) absorption. More sophisticated laboratory techniques, such as high-pressure liquid chromatography, are used by the laboratory people to provide more detailed characterization of the type of multiring aromatic structures such as coronenes and ovalenes.
I agree that by far the quickest and cheapest method is to use one's 20/20 vision and check for color on the recycle liquid. Usually color changes from yellow to orange to red indicate progressively higher PNA.
Warwick: We process distillate (i.e., FCC LCO, virgin coker) to extinction, but do not try to process high end point material because of polynuclear aromatic buildup. When we do test for polynuclear aromatics, we use an in-house UV method.
It is simple and available in many refineries. Many patents have been issued on processing the recycle stream; however, I know of no one actually practicing this.
Greg Dearwater (Profimatics Inc.): We have seen two approaches to the PNA problem. First, one refiner has taken a slipstream of hydrocracker fractionator bottoms back to the vacuum unit to purge the polynuclear aromatics. The second solution we have seen is the first solution taken to the limit. This refiner takes the entire hydrocracker fractionator bottom stream back to the vacuum unit. Obviously, this can be justified economically only in a location where fuel gas values are low.
AMMONIA INJECTION
What experience do you have with NH3 injection during operation, with the aim of improving middle-distillate selectivity? Are there any drawbacks besides a higher required operating temperature?
George: First we need to distinguish between two modes of hydrocracker design. A series-flow unit has the pretreat reactor and the cracking reactor in series with all the effluents-the liquid plus the gas, including ammonia-passing directly from the pretreat reactor to the cracking reactor. A two-stage mode has interstage separation/fractionation between the pretreat and the cracking reactors.
In the two-stage hydrocracker, a small level of ammonia between 10 and 50 ppm is found in the recycle gas, depending on the washwater rate and the phase-separator temperature. It is in the two-stage mode of operation that we have found, in pilot-plant tests as well as commercial experience, that ammonia injection up to a 100 ppm level has a positive effect on increasing the jet and diesel yield selectivity.
While pilot-plant tests clearly show the negative impact of ammonia injection on operating severity, our commercial experiences show a more practical aspect of the impact of a loss of an ammonia injection pump. Basically, ammonia desorbs rapidly from the catalyst, especially when operating at higher temperatures (above 650 F.) and could give rise to a temperature exotherm.
I would Like to refer to Mr. Walliser's comments in the 1984 NPRA Q&A Session where temperature excursion was observed due to problems with the water-injection pump that could result in variation in ammonia concentrations in the recycle gas.
Cabrera: The chemical constituents contained in the environment within the hydrocracking reactor all affect the types of cracking reaction that can or will occur, as well as the cracking activity of that particular catalysts. Ammonia is no exception and it certainly is a constituent that, when present, can affect the performance of the hydrocracking catalyst.
The effect of ammonia on a given catalyst will depend on several factors and Mr. George has covered some of those. To re-emphasize, certainly the type of hydrocracking catalyst used, the relative concentration level in the reactor, and the process configuration all influence reactor performance. We consider these particular effects to be proprietary and part of the knowledge of the integrated catalyst unit system.
The controlled addition of ammonia at the correct level can affect the cracking selectivity and it is necessary to maintain good and reliable control since you are affecting not only the stability of the cracking reactor, as has been mentioned, but it also affects the product yield.
The question of whether or not a particular catalyst system will benefit from the controlled addition or removal of ammonia from the reactor environment is a case-specific issue and has to be addressed depending upon the catalyst type and unit configuration. We should emphasize that there can be negative effects on hydrocracking operations either by adding or removing ammonia from the system.
Lavergne: Our experience in heavy gas oil operation has been that the addition of ammonia has resulted in a 10% relative increase in jet, fuel yield. We have found ammonia to demonstrate a much greater influence on yield distribution with heavy feeds vs. light feeds. Higher required operating temperature with ammonia addition results in slightly higher energy costs.
Purchase, storage, and pumping of ammonia may also be necessary. The effect of ammonia on yield selectivity will also result in a lower volume gain on feed. Depending on individual refinery pricing, the net value of using ammonia may be positive or negative.
Warwick: Ammonia injection can be used to improve distillate selectivity on zeolite catalysts. The most significant drawback is higher operating temperature.
Safety is always an issue with ammonia injection. Special pumps and piping are usually required. You also must have easy access to ammonia.
Alternatively, nitrogen compounds that quickly decompose to form ammonia (i.e., alkyl amines) can be used, but the cost per volume of ammonia injected is usually higher.
Copyright 1993 Oil & Gas Journal. All Rights Reserved.
