2004
2007
2010
2013
2019 Looking for NH removal/low reactivity compromise
2019
2023
Looking for even lower initial reactivity & easier start-up
Need for NH removal
Looking back to no initial reactivity
D-680
D-680
Hybrid 1
Hybrid 1
Hybrid 2
3A-MS
3A-MS
Dynocel 680
Dynocel 680
3A-MS
3A-MS
3A-MS
3A-MS
3A-MS
Figure 2 Evolution of C₃ dryer loading with time
peak rapidly disappeared during following the regeneration cycles, and the 3A-type molecular sieves/Hybrid 1 combina- tion exhibited excellent efficiency in both water and ammo - nia removal throughout the three-year campaign until the next turnaround. Therefore, in 2010, the same combination was loaded. During the first regeneration after start-up, a temperature peak was again detected in the treater. This time, a severe temperature excursion led to the outlet temperature exceed- ing the inlet temperature (see Figure 3b ). This indicated that a polymerisation reaction had taken place in the adsorber, caused by the undesired reactivity of the freshly loaded Hybrid I adsorbent. 2013-2016 : To avoid any reaction taking place in the dry- ers, they were loaded as in the past with 3A-type molecular sieves only. However, as in 2007, the need for NH 3 removal forced PCK to include a hybrid adsorbent layer for the next change-out. In 2016, it was therefore decided to repeat a 3A-type molecular sieves/hybrid combination, but this time with a higher safety level. 2016-2019: In accordance with the Hybrid 1 supplier, a set of measures was established to impede the polymerisation reaction during fresh adsorbent regeneration. These meas- ures included: • Installation of temperature measurements in the hybrid adsorbent bed and automatic pressure relief to flare when temperature excursion is recognised • Establishment of a special procedure for the first regen - eration: use of propane and reduction of regeneration
temperature with olefinic C₃ cut down to 180°C instead of > 220°C • Use of alternative hybrid adsorbent from the same supplier but with reduced initial catalytic activity (Hybrid 2). Together, the measures allowed for safer operation, but the procedure of using propane for the first regeneration required a considerable logistic effort, including several extra hours during the start-up phase (see Figure 3c) . In addition, during the three-year service life, Hybrid 2 product showed undesired reactivity in the case of accidental overheating of the adsorbent. Figure 4, left shows a severe exotherm that took place in the Hybrid 2 bed after being heated to 275°C during regeneration. Thanks to presence of thermocouples and release valves, no safety issues took place, but this led PCK to look for an even safer option. This option was found within Dynocel 680 adsorbent due to its previous track record from other projects at the refinery, including the adsorbent’s intrinsic characteristics and ease of use (no pre-load step needed), longer guaranteed lifetime, and competitive price set. 2019-2023: Consequently, in 2019, dryers were loaded with Dynocel 680 only. The first regeneration was run at a reduced temperature of 190°C, which resulted in the efficient removal of contaminants with no signs of an exothermal reaction (see Figure 3d ). The same was true when run at the higher temperature of 220°C. Even in the case of overheating of the adsorbent, no exotherm was detected in the adsorbent bed (see Figure 4, right ), confirming its low level of reactivity. 2023 to date: Due to the safety, reliability, and overall
d. 2019 - Dynocel 680
a. 2004 - 3A - MS
b. 2010 - Hybrid 1 & 3A - MS
c. 2016 - Hybrid 2 & 3A - MS
300
300
300
300
250
250
250
250
200
200
200
200
150
150
150
150
100
100
100
100
50
50
50
50
0
0
0
0
0
Time (h) 72 12 24 36 48 60
0
Time (h) 72 12 24 36 48 60
0
Time (h) 72 12 24 36 48 60
0
Time (h) 72 12 24 36 48 60
Figure 3 Temperature profile during first and following regenerations. Light grey : inlet regeneration gas temperature, Coloured : outlet regeneration gas temperature
54
PTQ Q4 2023
www.digitalrefining.com
Powered by FlippingBook