PATRICK GRIPKA (Criterion Catalysts & Technologies)
We have many customer examples of FCCPT units and mild hydrocracking (MHC) units directly producing on-spec ULSD with Criterion catalyst systems. Many refiners who are making 30 to 50 ppm sulfur diesel from hydrotreating and/or mild hydrocracking unitsare able to blend this into the diesel pool and still have the whole pool meet the less than 15 ppm diesel sulfur specification. In these cases, there is typically overtreating in another hydroprocessing unit.
Assuming unit pressure and recycle gas, rates are maximized, in order to reduce the sulfur to the ULSD level, the immediate options a refinery has are feed rate, feed composition, and feed cutpoint. All are essentially managing the overall feed difficulty and processing requirement in the hydrotreater. Depending on refinery flexibility, it may be feasible to move a portion of the most difficult feed stream to another hydrotreating unit so the subsequent easier feed will produce a lower sulfur diesel. Another solution is to lower the feed endpoint and send the hardest to treat molecules to another hydroprocessing unit. However, these options are less than ideal as there are often economic penalties due to non-optimal operation across these multiple units. The Best Practice is to improve the catalyst technology in the heavy gas oil hydrotreater or mild hydrocracker and produce ULSD directly.
I have two examples that demonstrate how catalyst technology has been used to increase overall ULSD make within an FCCPT unit. In both examples, the refiner chose the option of reducing feed cutpoint in previous cycles to meet the desired diesel properties so on-spec ULSD was being produced. However, use of improved catalyst technology has allowed the cutpoint to be increased in subsequent cycles.
Example 1: Potential to Significantly Increase the Endpoint and Increase the Amount of ULSD Make
The graph below shows product sulfur level and the endpoint of the diesel product from an FCCPT unit. Two cycles from an FCCPT unit are represented, and the catalyst was upgraded to a CENTERA™ catalyst system in Cycle 2. Both cycles were consistently producing diesel in the 1-to 30 ppm range; however, the big difference is that the diesel T90 in Cycle 1 was limited to the 525 to 575°F range, which is a very light diesel (almost a kerosene), but the diesel T90 in Cycle 2 is in the 625 to 670°F range, which represents a 100°F increase in diesel T90. Feed qualities and operating conditions are similar between the two cycles. The improved catalyst performance has resulted in an overall increase in recoverable diesel. As noted above, one of the biggest levers to maximize cutpoint is proper catalyst selection.
Example 2: Potential to Continue Producing Higher Amounts of ULSD as the Cycle Progresses
The graph below shows percent diesel yield versus days onstream; the red dots represent the previous cycle; and the blue dots represent the subsequent cycle with an improved catalyst system. Here the FCCPT unit processes all the coker gas oil –heavy and light –in the refinery. The yield of ULSD removed in the fractionator is very important to the economics of the refinery. The improved CENTERA™ catalyst system resulted in a more stable ULSD yield profile as the previous cycles exhibited a drop-offin ULSD production at the end of the cycle due to the unit’s inability to maintain product sulfur at the higher endpoint. Over the cycle with the improved CENTERA™ catalyst system, the average yield of diesel was 5% higher, with an average 10°F higher diesel fraction T-95compared to the previous cycle.
In terms of diesel quality, the diesel produced from the FCCPT or MHC unit routinely meets U.S. specs. As shown below, the diesel cetane for this unit meets/exceeds U.S. specifications and was similar for both cycles and exhibited stable quality over the cycle.
DAVID VANNAUKER (Haldor Topsoe, Inc.)
Diesel sulfur has a profile with the sulfur concentration increasing as the boiling point increases. A simple option is to produce a wider-cut jet stream that can be used as a blend stock with ULSD. Another option is to change the next catalyst load to increase the HDS functionality, which has been proven to enable the full wide-cut ULSD. When units increase the EP, there is a potential for higher pour/cloud point. Employing dewaxing or isomerization catalysts can resolve this issue.
SAM LORDO (Nalco Champion)
To achieve a good online cleaning (with the exchanger bypassed by not pulled), it is imperative to have nozzles on the inlet and outlets that are large enough to facilitate the circulation of a heat cleaning solution or steam with a cleaner. Recovery of heat duty is depended on the velocity of the cleaner, the foulant material, and whether the cleaning solution actually came in contact with the foulant (100% fouled-off exchangers are hard to clean online). Typically, online cleaning can be used successful and recover 60 to 80% of the flow or duty. Offline cleaning is more thorough, if done properly and recovery is routinely higher.
RON PARISE (Nalco Champion)
We have seen good success through the continuous application of one of our cleaning chemicals at low dosages with some online exchangers, particularly B/E (bottom entry) exchangers in desalters. For offline cleaning, it is critical to have sufficiently sized injection nozzles/circulation nozzles (at least 2” diameter) to afford enough turbulence with the cleaning solution to provided effective cleaning of the exchanger.
CHRIS CLAESEN (Nalco Champion)
The answer is partially the same as the one given to Question 28. First, the root cause needs to be determined. If the dP is caused by corrosion products due to corrosion in the upstream refinery units, the corrosion in these units can be reduced by applying the proper corrosion control program. If the dP is caused by gum formation, the gum formation needs to be controlled by applying a program in the feed and storage system. While we recommend that the focus be on prevention measures, we have successfully reduced hydrotreater dP on many units by applying an online cleaning program.
DENNIS HAYNES (Nalco Champion)
A hydrotreater processing import naphtha was experiencing reactor bed pressure drop well into the run. It was suspected that particulate iron sulfide was a main component of the material restricting the upper portion of the reactor. It was not possible to stop the pressure drop increase in this case, but application of a dispersant resulted in a reduction in the rate of pressure drop increase to the extent that the unit could make it to turnaround. For chemical additives, it is important that due diligence is done in the evaluation of the fouling mechanisms so that the appropriate inhibitors can be fit for purpose.
WALTER MILITELLO, PhD (Nalco Champion)
Several units have shown some degree of fouling, usually impacting the heat transfer performance of heat exchangers and producing hydraulic restrictions at HT reactors. Fouling nature can be organic or inorganic, depending on feed contaminants and operating conditions. The most successful approach to minimize pressure drop buildup is the application of top reactor dispersant agents. The function of those agents is to disperse the foulant material, thereby reducing its adhesivity and compacting. Often those agents are able to break fouling crust, if injected properly and at the correct dosage and duration. If it is not possible to implement the dosing system during the plant operations, an alternative solution is the injection of dispersants from the preheat train. The immediate benefit is to recover heat transfer across feed/effluent exchangers; because usually, the fouling problem also impacts the cross-exchangers. But careful dosing strategy and accurate monitoring must be in place to avoid significant fouling material transportation from heat train to reactor, making the hydraulic situation worse. Recent experience of dispersant around a Saudi plant, showed a FF (fouling factor) reduction of cross-exchangers from 0.007 ft2hr°F/BTU growth every month, to a growth rate of less than 0.001 ft2hr°F/BTU per month. The reactor norm-dP(flow corrected) growth during the chemical program was only 0.12 psig per day, compared to previous cycle where norm-dP grew at a rate of 0.22 psig per day, resulting in a premature cat skimming after 11 months from fresh cat startup.
RALPH WAGNER (Dorf Ketal Chemicals LLC)
The fouling precursor can be classified into two groups: organic and inorganic. Organic precursors are due to the presence of unsaturated in cracked feedstock which forms polymers and coke, depending on the operating conditions. Inorganic precursors are due to the presence of a corrosion product coming from the unit upstream. Both precursors agglomerate and deposit on the preheat exchanger surface; and at a break-even point, they get carried to the reactor, further aggravating the fouling scenario. Dorf Ketal offers a range of antifoulant chemistries specific to the precursor type. Typically, an antioxidant is used to terminate the polymer formation, while dispersant chemistry uses steric barrier to limit the particle size and deposition. Dorf Ketal has successfully treated naphtha hydrotreaters worldwide for organic and inorganic fouling. Dorf Ketal’s FeS agglomerate has been proven successful for inorganic fouling and has resulted in run-lengths being increased by several months with sustained pressure drop.