Designing vacuum units
When designing vacuum units for processing heavy Canadian crudes, reliability costs can be high if the feedstock’s thermal instability is not fully appreciated
Scott W Golden and Tony Barletta, Process Consulting Services
Viewed : 4394
New vacuum unit capacity is currently being added at several North American refineries to process higher amounts of heavy Canadian crude. Many other refiners are evaluating revamps or new units. Ultimately, coker unit capacity will determine the maximum amount of heavy Canadian crude refiners can process. Nonetheless, maintaining a high vacuum gas oil (VGO) yield throughout the run will reduce the coker charge rate, decrease the amount of coke produced per barrel of crude, and permit maximum processing of heavy crudes.
Heavy Canadian crude oil properties influence vacuum unit reliability, VGO yield, gas oil quality and run length. Few refiners currently process high percentages of these crudes in their crude blend. Furthermore, oil sands- based crudes are less thermally stable than conventional crudes, with asphaltene precipitation and vacuum heater tube coking just two of the many challenges. Operating at a low gas oil cutpoint is one solution, but this approach reduces the VGO yield, increases the coker charge rate and raises coke production per barrel of crude.
Heavy Canadian crude oils
The implications of crude source on vacuum unit design cannot be overstated. Heavy Canadian oils consist of conventional heavy and oil sands-based crude oils. Cold Lake and Western Canadian Select are examples of conventional heavy crude. Oil sands-based crudes include Albian Heavy, Christina Lake, McKay River and others that have not yet come into production. Total conventional heavy Canadian crude production today is approximately 500 Mbpd, which is expected to drop slowly until 2010 and then sharply decline. Conversely, oil sands-based production will increase from about 1000 Mbpd today to about 2500 Mbpd or more by 2015, depending on how many of the planned upgrader projects are actually built, when they reach full production or whether there is sufficient blendstock for the bitumen.
Applying conventional design practices to new units that will process heavy Canadian crudes will result in many unwanted surprises. Oil sands-based crude oils are generally less thermally stable than conventional crudes, some are very unstable, most have high volatile vanadium and nickel content in the VGO boiling range, many have high solids content and several have very high total acid number (TAN) in the VGO boiling range. Rapid equipment coking and asphaltene precipitation are major problems facing refiners. In some instances, heater outlet temperatures have been reduced below 700°F to obtain reasonable heater run lengths at the expense of a low gas oil cutpoint.
Crude processing flexibility will largely depend on equipment metallurgy. Some oil sands-based crude oils produce heavy vacuum gas oil (HVGO) product with TAN values of 6–8. Processing high percentages requires 317L SS in many of the circuits. Since the crude supply situation is still developing, metallurgical selection may be the difference between processing low-cost crudes and having to compete with other refiners for the low-TAN feedstocks. If most new units are built to process only low-TAN crudes to minimise investment, crude flexibility will be limited or high corrosion rates in some circuits will be the consequence of processing high-TAN crudes. Price differentials between low- and high-TAN heavy sour Canadian crude will dictate whether metallurgy upgrades will pay off.
Crude composition variability
Conventional and oil sands heavy Canadian crudes are blends of condensate or upgrader products and very heavy crude or bitumen. Many are currently produced in limited volumes. Thus, it is prudent to assume crude composition will likely be different than the current assays when commercial production begins. The same has been true of extra-heavy Venezuelan blends; there have been significant variations in compositions. Lower API gravity crude forces a higher heater outlet temperature to achieve the same VGO product cutpoint. Even though the crude source is the same, gravity variability requires operating flexibility to achieve the same distillate cutpoint.
In addition, laboratory data show that a 10°F change in temperature in the laboratory distillation device dramatically alters the rate of thermal cracking on some of these crudes. Even though laboratory devices do not reflect true cracking tendencies in the vacuum heater, they do show directionally that there is an important stability difference between conventional heavy crude oils like Cold Lake and some of the oil sands-based crudes.
Typical vacuum unit design assumes crude quality is fixed based on a given assay. An alternate approach assumes large quality variability during the design phase and determines the operating flexibility and capital cost to maintain product yield, quality and run length. Evaluating critical operating variables and estimating the VGO yield over a range of feedstock qualities is prudent. Refiners processing heavy and extra-heavy Venezuelan and Maya crude oils have experienced significant performance differences between the design and actual operation. Many of these early heavy oil projects were designed at a benchmark 1050°F TPB cutpoint, yet several achieved less than 1000°F when commissioned. These units produced 3 vol% less VGO and higher coker feed rates than design. In some cases, the actual crude charge rate was reduced by 8–10% to stay in balance with coker capacities.
Reducing crude oil cost is the major incentive driving these new crude and vacuum unit projects. Yet the reality is that higher capital cost designs are needed to process these crudes reliably over a four-to-five-year run length. Designing a low-cost unit will reduce project revenues, because product yields are lower and run lengths are shorter between turnarounds.
Vacuum unit reliability
Few refiners have experience of processing high volumes of either conventional or oil sands-based heavy Canadian crude oils. And these few operate at TBP cutpoints below 950°F. In some instances, major equipment design changes were needed to reliably produce high volumes of acceptable-quality VGO for four-to-five-year run lengths. In several cases, vacuum heater and column wash sections coked in less than one year. One refiner had less than one year’s run on the gas oil hydrotreater due to HVGO product vanadium content of 10 ppmw or more. Additional reactors were added to remove the high metals prior to the desulphurisation reactors so that reasonable run lengths were achieved in the hydrotreater.
Add your rating:
Current Rating: 2