operate high enough to avoid ammonium bisulphide and other amine sulphur compound laydown to avoid exces- sive corrosion rates and internals plugging. Temperature cannot simply be minimised to reduce condensable oil. Some crude columns have almost no stripping section efficiency due to poor internals design, so assuming an arbitrary efficiency can result in an undersized vacuum system due to a high condensable oil rate. Well-designed crude columns have more stripping trays and proper strip- ping tray design with stripping steam rates of 8-10 lbs steam/barrel of atmospheric column bottoms. A well- designed stripping section will result in lower condensable oil in the vacuum overhead stream. Where appropriate, atmospheric crude column stripping section modifications should be considered as part of vacuum system modifica - tions. In this example in Table 1 , vacuum unit feed was supplied from three different atmospheric crude columns. One has a well-designed stripping section and two others have poor stripping efficiency; hence the condensable oil rate is relatively high. Ultimately, the process loads provided to the ejector manufacturer will be used to size the ejectors and condens - ers. Care must be given to providing accurate information with adequate design margin applied to the cracked gas flow rate. Ejector operation and maximum discharge pressure (MDP) Actual ejector curves and design basis process loads (see Table 1) are used to illustrate the importance of the first-stage inter-condenser and how its failure to perform as intended can result in performance break. A ‘thought experiment’ stress test on the first stage of this system is used to highlight how failure of the long air baffle can lead to a higher flow rate to the second-stage ejector and, ulti - mately, performance break. The ejector system designed for the process loads shown in Table 1 results in a first-stage ejector suction pressure of 19 mmHg absolute and MDP of 98 mmHg absolute. As long as the operating discharge pressure is at or below 98 mmHg absolute, the suction pressure is solely determined by process load. Assuming the design motive steam rate and conditions, and operating pressure below its MDP, the first-stage suction pressure is not affected by operating discharge pressure. When operating discharge pressure exceeds MDP, the suction pressure increases due to per - formance break (see Figure 6 ). The ejector performance curve sets the relationship between process load, suction pressure, and MDP. Figures 7 and 8 are actual first- and second-stage ejector perfor - mance curves that will be used to quantify how long air baffle bypass affects first-stage ejector discharge pressure and performance break. The suction and MPD performance curves show the process load on the y-axis and the resul- tant suction and MDP on the X-axis. Long air baffle bypass raises the second-stage process load and first-stage inter- condenser gas outlet temperature while increasing second- stage suction pressure and first-stage ejector discharge pressure.
Break
30 19
Design (no break)
Vacuum column overhead
1st stage ejector
98
MDP
100
97
89
1st stage inter-condenser
91
105
100
Pressure, mmHgA Temperature, ˚F
2nd stage ejector
Condensate
Figure 6 System performance break
35000
30000
25000
20000
15000
10000
5000
100 0 Suction pressure Ps, discharge pressure Pd (mmHg) 0 60 80 40 20
120
Ps: suction curve Pd: discharge
Curve suction Discharge
Capacity
Figure 7 First-stage ejector performance curve
14000
12000
10000
8000
6000
4000
2000
250 0 Suction pressure Ps, discharge pressure Pd (mmHg) 0 150 200 100 50
300
Ps: suction curve Pd: discharge
Curve suction Discharge
Capacity
Figure 8 Second-stage ejector performance curve
32
PTQ Q1 2023
www.digitalrefining.com
Powered by FlippingBook