Microsoft Word - 46gerke.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 69, 2018 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors: Elisabetta Brunazzi, Eva Sorensen Copyright © 2018, AIDIC Servizi S.r.l. ISBN 978-88-95608-66-2; ISSN 2283-9216 C2 Splitter Revamp Surpasses Capacity and Efficiency of 4- Pass Trays Daniel R. Summersa, Waldo de Villiersb, William L. Greenc aSulzer Chemtech USA, Inc., Tulsa, Oklahoma, USA bShell Global Solutions (US) Inc., Houston, TX, USA cShell Chemical Company, Norco, LA, USA A C2 Splitter in Norco, Louisiana was successfully revamped with Shell HiFi™ Plus trays in 2015 for increased tray capacity. The HiFi Plus tray is a joint development between Sulzer and Shell and is relatively new to the marketplace. These trays are intended for large diameter and high pressure/high liquid loaded applications such as those found in Ethylene Production. Benchmarking the capacity and performance of such trays will be invaluable to the distillation industry. Operating data were taken both before and after the revamp and will be presented in the paper along with tray hydraulic information. This paper will address two main issues, capacity and performance of Shell HiFi™ Plus trays and the very important topic of comparing the new trays against the old 4-Pass tray design. Discussions will include the observed excellent tray efficiency, operational hydraulics and pressure drop of the new trays. In addition, unique features of the HiFi Plus tray will be discussed that enabled this tray efficiency to be achieved. 1. Introduction Sulzer Chemtech and Shell have been providing tower internals for over 60 years. However, since March of 2000, Sulzer has been aligned with Shell Global Solutions to provide Shell HiFi™ Plus trays to the Chemical and Refining Industries. One of the opportunities that was enabled through this alliance was to revamp the C2 Splitter (Item PV-1767) in Norco, LA in 2015. Prior to this revamp, the tower had operated for more than 35 years with conventional 4-Pass trays. The tower consisted of 154 stainless steel (410s) round valve trays in a 4877mm ID vessel situated at 508mm tray spacings. The objective of the revamp was to fully utilize available hot side capacity and to provide additional tower capacity for future debottlenecking. In addition polymer grade ethylene purity had to be maintained with minimal ethylene losses in the bottoms. The tower is highly heat integrated into the whole Olefins Unit. The feed to the C2 Splitter is a vapor stream from the overhead of the upstream Deethanizer. The feed point was located 37 trays from the bottom of the tower prior to the revamp. The ethylene product is drawn from a chimney tray located 10 trays from the top of the tower. There is also a side reboiler and the liquid draw and vapor return was located 32 trays from the bottom of the tower. 2. Operating Data Prior to the Revamp Prior to the revamp, operating data were taken to examine the capacity and performance of the existing 4- Pass trays. The data were taken during several days of steady state operation in April 2015. Two particular days were “pushed” to high reflux rates and had steady operation for two days straight on the 26th and 27th of April. Since there was no feed analysis, the feed composition is based on the summation of the three product streams; Vent, Ethylene Products and Ethane Recycle. The overall material balance over those two days was excellent and within 3%, see column 1 in the below Raw Operating Data Table 1. Please NOTE - All flow rates in all the Tables have been “normalized” to match 1000.0 Kg/hr as the feed flow to the plant before revamp, as shown in Table 2 as the Feed. This “normalization” was enacted at the request of the operating plant. The tower was simulated using a proprietary Pro-II model using a Soave-Redlich-Kwong Equation of State VLE model. The data used to determine the proprietary interaction parameters was from Barclay, Flebbe DOI: 10.3303/CET1869118 Please cite this article as: Summers D., De Villiers W., Green W., 2018, C2 splitter revamp surpasses capacity and efficiency of 4-pass trays, Chemical Engineering Transactions, 69, 703-708 DOI: 10.3303/CET1869118 703 and Manley (1982). This model was established within Sulzer's organization in the late 1990's and calibrated to actual operating data from several C2 Splitters. The resulting Heat and Material (H&M) Balance is shown in Table 2. The difference in feed rate shown in Table 1 (compared to Table 2) is an indication of the unit’s material balance, which is excellent and less than 5.6%. From this H&M balance, the internal loads and physical properties for all the 4-pass trays in the tower were generated and the key loads in each tower section examined with Sulzer’s Hydraulics software program. Table 3 shows the results from an analysis for all four of the key tower sections. As can be seen in Table 3, the Jet flood is very high in the Section above the feed (trays 38-144) and the downcomer velocity is very high in the Pasteurization Section. Because of the high loads in these two sections, the tower was highly loaded in 2015 and the plant was correct in their assessment that this tower was a bottleneck. Also of note, the observed pressure drop of the existing trays was also very high, see Table 1, at 1.1KPa per tray. This very high pressure drop (including the vapor head) could be considered an indication of flooding across some of the trays. Table 1: Raw Operating Data C2 Splitter PV-1767 Item Description Tag No. Units April 2015 Before Revamp Feb 2017 After Revamp Feed Rate* (normalized) FF-5758A kg/hr 1055.83 1142.1 Feed Temp TI-6147 °C -9.5 -2.1 Top Pressure PT-5416 KPag 2194.5 2325.6 Delta-P PDI-5414 Kpa 170.3 99.3 Bottom Flow Rate FT-5164 kg/hr 405.55 360.75 Reflux Drum Pressure PI-5415 KPag 2193.9 2324.9 Reflux Drum Temperature TI-6257 °C -27.1 -20.9 Reflux Flow FI-5420 kg/hr 3503.1 4003.17 Vent Flow FI-5425 kg/hr 42.09 46.34 Side Reb. Duty Calculated MMKJ/Hr 22.1 25.3 Ethylene Draw FC-5154 kg/hr 576.4 722.2 Ethylene Product - Ethane AI-5417A ppm 486.7 504.4 Ethylene Product - Methane AI-5417B ppm 37.34 1.3 Splitter Vent - Hydrogen AI-5424A % 0 0.93 Splitter Vent - Methane AI-5424B % 1.9943 0.075 Bottom Product - Ethylene AI-5427 % 1.8808 4.52 Table 2: Heat and Material Balance – Pre-Revamp Simulation Results (*normalized flows) Composition, wgt% Feed Vent Ethylene Product Ethane Bottoms Hydrogen 0.0003% 0.0075% 0 0 N2 0.0453% 1.2658% 0.0017% 0 Methane 0.0498% 1.1160% 0.0049% 0 Ethylene 62.40% 97.606% 99.942% 1.751% Ethane 37.15% 0.0046% 0.0522% 97.318% Propylene 0.2131% 0 0 0.559% Propane 0.0000% 0 0 0 IsoButane and Heavier 0.1421% 0 0 0.372% Total (kg/hr)* 1000.0* 42.09 576.47 381.44 Phase Vapor Vapor Liquid Liquid Temperature, °C -9.86 -25.7 -23.9 0.5 Pressure, KPag 2550 2193.1 2203.5 2364 Reflux Drum Pressure 2193.1 KPag Tower Top Pressure 2193.8 KPag Reflux Rate 3503.25 kg/hr Reflux (Bubble Point) Temp. -25.7 °C Top Temperature -24.5 °C 704 Table 3: Tray Hydraulics of the Pre-revamp 4-Pass Trays Trays Tray Count Section Name Jet Flood % Weir Loading m3/m-hr DC Velocity m/sec Pressure Drop KPa/Tray 145-154 10 Pasteurization 82 73.0 0.106 0.543 38-144 107 Above Feed 87 63.7 0.092 0.520 33-37 5 Above Side Reboiler 65 63.7 0.092 0.375 1-32 32 Bottom 54 55.2 0.079 0.355 3. Revamp For the Ethylene plant to realize the premised improvement in ethylene production, several features to the plant were upgraded in 2015. One of them was to debottleneck the C2 Splitter, PV-1767. Shell did a full study of various feed slates to the tower as well as optimizing the feed point and side reboiler location. This study included the full revamp of all trays on a tray-for-tray basis from the very old 4-Pass trays to the modern Shell HiFi Plus trays. The reboiler was also upgraded. The old trays from 1978 had straight side downcomers and box-type straight center and off-center downcomers that wasted a significant amount of tower cross-sectional area. In addition, these trays were designed in an era when 4-pass tray balancing was not even considered important. Four pass tray balancing became very important in the new millennium as shown by the referenced article by Summers in 2009. There was a significant opportunity for a vast capacity improvement of this tower. With adjustments to upstream cracking, the feed ethylene purity would become higher in ethylene content as well as higher in overall throughput. The feed tray was moved higher in the tower to above tray 45 (counted from the bottom) and the side reboiler draw and return location moved up to tray 37 as well. Note that the old liquid withdraw nozzles for the side reboiler were used after the revamp and that internal piping was used to get the liquid from tray 37 down to tray 32 for withdraw. HiFi Plus trays were chosen because of positive past experience in C2 Splitter service as depicted in the referenced publication by Summers, et al from 2007. These trays are comprised of multiple sloped and truncated downcomers with active tray deck area between. The prominent feature of the HiFi tray is its long weir length which enables high liquid loaded systems to operate at reduced weir loading and achieve high capacity, see Figure 1. The other prominent feature is a defined flow path length which enables this tray to achieve very good tray efficiencies. Figure 1: Two Half-Tray Trial Layouts of the HiFi Plus Tray for PV-1767 The revamp of the tower took a just little more than 30 days and was not encumbered by a single safety incident. To expedite the revamp of all 154 trays, elevators (instead of cranes) were employed to keep tray parts moving to and from the tower even in inclement weather, see Figure 3. The elevators also enabled, 705 during the peak of activity, to have internals installed concurrently at five working zones in the column. In addition to a HiFi revamp, Chimney trays and all feed devices were changed out as well. To ensure adequate spacings for capacity two active trays were converted to chimney trays resulting in 152 HiFi Plus trays in the tower. 4. Operating Data after the Revamp After the HiFi tray revamp, operating data were again taken to examine the capacity and performance of the new trays. The data were taken over several days of steady operation in early 2017, see Figure 2. One particular day, February 7, 2017 was very steady for not just the reflux, but the other flow rates, and was chosen for this study. Again the feed composition is based on the summation of the three product streams; Vent, Ethylene Products and Ethane Recycle. The overall material balance over this day of operation was excellent and within 1%, see column 2 in the Raw Operating Data Table 1. To match the reflux rate on this date, Murphree Tray efficiency values where employed in Pro-II to the trays in the tower and the overall efficiency adjusted. The trays showed a higher efficiency (82.6%) than the old 4-Pass trays (78.3%). The resulting Heat and Material (H&M) Balance is shown in Table 4. The difference in feed rate shown in Table 1 (compared to Table 4) is an indication of the unit’s material balance, which is excellent and less than 1.2%. From this H&M balance the internal loads and physical properties for all the HiFi trays in the tower were generated and the key loads in each tower section examined with Shell’s Hydraulics software program. Table 5 shows the results from that analysis for all four of the key tower sections. As can be seen in Table 5, the Jet flood and the downcomer velocities in the two upper sections are no longer a problem. In fact, the tower is quite under loaded and has plenty of extra capacity. There was one thing to note, the pressure drop across the trays reduced significantly from 170.2 KPa before the revamp to 99.2 KPa after. This is also a clear indication there is plenty of extra capacity in this tower. The calculated pressure drop in Table 5 is 57.8 KPa. The vapor head across the 91.5 meter tall tower is approximately 42 KPa. This gives us a 99.8 KPa predicted pressure drop that matches the observed value quite accurately. Condenser and refrigeration capacity limitations were intentionally not addressed in this project, and have, as expected, limited the utilization of the full hydraulic HiFi tray capacity to date. Figure 2: Steady Sate Determination – Operating Reflux data after the Revamp Reflux Flow Rate 706 Figure 3: C2 Splitter being Revamped – Note the construction Elevator attached to the Vessel 707 Table 4: Heat and Material Balance – After Revamp Operation Simulation Results (*normalized flows) Composition, wgt% Feed Vent Ethylene Product Ethane Bottoms Hydrogen 0.0003% 0.0074% 0 0 N2 0.0525% 1.2504% 0.0018% 0 Methane 0.0009% 0.0198% 0.0001% 0 Ethylene 69.30% 98.717% 99.944% 4.2135% Ethane 30.34% 0.0057% 0.0540% 94.840% Propylene 0.1833% 0 0 0.5736% Propane 0.0000% 0 0 0 IsoButane and Heavier 0.1192% 0 0 0.3731% Total (kg/hr)* 1129.3* 46.34 722.2 360.8 Phase Vapor Vapor Liquid Liquid Temperature, °C 3.4 -22.8 -21.9 0.83 Pressure, KPag 2757 2324 2329.6 2428.1 Reflux Drum Pressure 2324.0 KPag Tower Top Pressure 2324.8 KPag Reflux Rate 4003.66 kg/hr Reflux (Bubble Point) Temp. -22.8 °C Top Temperature -22.1 °C Table 5: Tray Hydraulics of the HiFi Plus Trays Trays Tray Count Section Name Jet Flood % Weir Loading m3/m-hr DC Velocity m/sec Pressure Drop KPa/Tray 145-154 10 Pasteurization 66.2 43.8 0.0735 0.415 47*-144 98 Above Feed 64.7 36.6 0.0606 0.391 38-45 8 Above Side Reboiler 53.0 36.6 0.0600 0.364 1-37 37 Bottom 44.8 32.2 0.083 0.336 *Tray 46 became a chimney tray 5. Conclusions The revamp of the Norco, Louisiana C2 Splitter (PV-1767) with 152 Shell HiFi Plus trays resulted in a significant improvement of tray efficiency and capacity. The plant, independently, reports that the energy usage in both the side reboiler and bottom reboilers decreased by 6% on an energy per mass of product basis. However, a comparison of the total reboiler duty between Tables 2 and 4 indicate that the decrease is closer to 15% based on the data presented here. The tower also has about 20 to 25% extra capacity available for the future based on employing a maximum jet flood of 85% as the column’s limitation. Tray efficiency is very consistent with previous HiFi Plus applications in high-pressure service. Predicted pressure drop for the HiFi Plus trays match operational values extremely well. References Barclay D.A., Flebbe J.L., Manley D.B., 1982, Relative Volatilities of the Ethane-Ethylene System from Total Pressure Measurements, J. Chem. Eng. Data, 27, 135–142. Summers D.R., 2009, Four Pass Tray Design Techniques, Distillation 2009 Topical Proceedings, AIChE 2009 Spring National Meeting. Summers D.R., Bernard A., de Villiers W., 2007, High Capacity Tray Revamp of a C2 Splitter, Distillation 2007 Topical Conference Proceedings, AIChE 2007 Spring National Meeting. 708