Myth of high cutpoint in dry vacuum units
Accurate measures of true boiling point cutpoints are essential to correctly evaluate vacuum unit performance
SCOTT GOLDEN, TONY BARLETTA and STEVE WHITE
Process Consulting Services
Viewed : 8677
Refinery vacuum units maximise higher value vacuum gas oil (VGO) and minimise vacuum residue (VR). Dry unit designs (see Figure 1) are being used to process a variety of crude blends from very light crudes to heavy Canadian blends. Designers claim these dry units can achieve VGO true boiling point (TBP) cutpoints of 1050-1105°F (565-595°C). Dry units have lower operating costs and marginally lower capital investment versus those using steam in the heater coils and/or steam stripping the VR; therefore, they are often mistaken for the most profitable design. Dry unit operation may be justified when processing light crude blends or when the crude contains less VGO boiling range material making lower TBP cutpoints acceptable. When processing medium or heavier crudes or thermally unstable crudes, dry vacuum units produce much lower VGO yield and higher VR production. VR is often coker feed where 10-20% of the VGO boiling range material is turned into coke; this reduces refinery liquid volume yield.
Vacuum unit performance
Over the last 18 years, rigorous field testing of more than 100 CDU/VDUs processing crudes ranging from West Texas Intermediate (WTI), Urals, Arab Light, Canadian bitumen, Venezuelan extra-heavy to everything in between shows high TBP cutpoint claims are rarely achieved. Nearly every unit tested had an actual VGO TBP cutpoint 40-100°F lower than design or had run lengths as short as six to 24 months. High VGO cutpoint claims continue, but actual performance has shown that rarely are these claims valid across a four to six year run. It is possible to run a high heater outlet temperature and operate with little or no overflash for a short period of time. However this combination usually results in a shortened vacuum heater run length or coked wash bed. There are many documented cases where actual VGO yields were 3-6 vol% lower on whole crude than claimed, and heater and column run lengths less than two years are common. If crude is worth $100/barrel, destroying VGO in a coker and taking unscheduled shutdowns to pig heaters or replace the wash section reduces refinery revenue.
Low cost is a simple measure, but it is often not the most profitable. Dry vacuum units should be used when the operating cost savings are worth more than lost revenue from reduced VGO and refinery liquid volume yields. Simple metrics that use ejector steam consumption as a measure of performance are very misleading. Refiners are in the business of selling liquid product, not producing coke or fuel oil. Dry vacuum units always produce less VGO than units designed with steam (see Figure 2); no competent designer can argue that dry units cut deeper. Processing moderate to heavy crude blends and thermally unstable crudes requires steam to achieve moderate to high TBP cutpoint throughout a four to six year run. Dry vacuum units have been used with heavy crudes and several have TBP cutpoints less than 930°F (498°C), even though the stated designs claims higher than 1000°F (537°C).
Definition of VGO TBP cutpoint
VGO TBP cutpoint is a refining industry metric to compare vacuum unit performance. Higher cutpoint means more VGO and less VR production. Unfortunately, reported TBP cutpoints are often not correct. TBP cutpoint is based on the whole crude TBP distillation curve; the temperature on the Y-axis of the curve corresponds to yield of VR as a volume percent of crude (see Figure 3). The whole crude TBP curve must be accurate and the unit material balance good to determine actual cutpoint, otherwise the calculated value is not correct. While TBP cutpoint is a simple concept, the data needed to calculate it correctly is not readily available from typical refinery product analyses. Consequently, shortcut methods and other practices have been developed to estimate TBP cutpoint. Unfortunately, these methods are very misleading.
Generating an accurate whole crude TBP curve requires good temperature and volume data in the VGO boiling range. For an operating unit, VGO and VR distillations must be accurate but are generally not available or are often inaccurate due to inherent difficulties with VR distillation tests. Most refineries process many crudes, therefore crude blend percentages are often estimates, and crude assays are not useful in measuring operating performance.
Crudes contain significantly different amounts of VGO boiling range material; the slopes of the distillation curves in the 975-1100°F (524-593°C) range are very different. For example, Middle East Arab Light contains much less VGO than many of the heavy crudes, therefore incentives to cut deep vary by crude type. Designing for high TBP cutpoint on Arab Light yields less incremental VGO than other crudes. Table 1 shows the VGO volumes for Arab Light, Arab Heavy, Merey, Maya, and Canadian bitumen blends between 975-1100°F (524-593°C).
Table 1 shows that VGO TBP cutpoint is a poor measure of incremental VGO recovery because VGO content is a function of crude type. The steeper the distillation curve, the less the incremental VGO content for a given cutpoint change. Higher slope TBP curves require incrementally higher heater outlet temperature, lower column operating pressure or more steam to recover incremental VGO product. Crudes with less VGO have less incentive to cut deeper than crudes containing more recoverable VGO and each incremental 1 vol% yield increases costs.
Many refiners and some technology providers confuse TBP cutpoint with other VGO yield metrics. VGO whole crude TBP cutpoint is not equal to the HVGO product 95 vol% temperature or an average of HVGO and VR product 95%/5% temperatures. It also is not the atmospheric equivalent temperature (AET) calculated by the Maxwell-Bonnell method using the measured flash zone temperature and pressure. For a given feed, flash zone pressure and temperature are a relative measure of VGO product yield. They do not correspond to whole crude VGO TBP cutpoint. The Maxwell-Bonnell conversion of ASTM D5236 Potstill measurements bears little relationship to VGO cutpoint on an operating unit. Crude type and upstream atmospheric column distillate yield determine vacuum unit feed composition. Heavier feeds result in lower cutpoint for the same flash zone temperature and pressure, thus using flash zone pressure and temperature to calculate cutpoint is incorrect. Vacuum units stripping VR can yield 3-10 vol% additional VGO on whole crude even when operating at higher flash zone pressures than a dry vacuum unit. The Maxwell-Bonnell AET method is not the same as whole crude TBP cutpoint.
Dry vacuum unit: fundamental principles
Dry vacuum units rely solely on operating pressure and temperature to vaporise the feed. They do not use coil or stripping steam to influence oil partial pressure. Heater outlet pressure is set by flash zone pressure and transfer line pressure drop. Maximum outlet temperature is set by oil’s thermal stability and heater design. Low stability crudes such as northern Athabasca mined bitumens crack at 680-690°F (360-365°C) outlet temperature, even with a well-designed double-fired heater that minimises oil film temperature rise, whereas Arab Light can be operated at much higher temperatures with a well- designed conventional heater. Poor heater design lowers maximum heater outlet temperature, independently of crude type, because it is results in higher peak temperatures. Dry vacuum designs do not use coil steam to control radiant section oil residence time, therefore heater outlet temperature must be lower than in designs using coil steam to control oil residence time. Heater oil mass flux should always be optimised to control oil film temperature rise to minimise cracking.
Add your rating:
Current Rating: 4