Trays, process changes improve frac plant operation

Pervaiz Nasir, Robert Andring
Shell Midstream Enterprises
Houston
Jose Bravo
Shell Oil Products Co.
Houston
Ross Mowrey
ABB Randall
Houston
Tom Gaskin
AET Technologies Inc.
Houston

Modifications at Shell Midstream Enterprises' Norco, La., NGL fractionation plant are ongoing but have already resulted in improved performance and efficiencies. (Photograph courtesy of Shell Midstream Enterprises, Houston)

The Shell Norco, La., NGL fractionation plant is being revamped from a nominal feed capacity of 43,000 b/d to 70,000 b/d.

This was achieved with existing columns and process heaters but also with addition of several heat-integration measures and replacement of the column internals with Shell's high capacity HiFi and CS trays.

Operating conditions in the deisobutanizer column were modified to optimize its capacity, taking advantage of better separation performance. The combination of these measures has resulted in a 17% reduction in the amount of energy per unit of feed required to run the plant.

Modifications

Before the revamp, the plant had demonstrated the capacity to process up to 43,000 b/d of the original design mixture ( Table 1 [24,801 bytes]). The process flow sheet describing the original configuration ( Fig. 1 [79,074 bytes]) indicates a conventional series arrangement of distillation columns with very modest heat integration.

The capacity of the plant was dictated by the capacity of the debutanizer and the deisobutanizer columns. In general, such peripheral equipment as pumps, condensers, and reboilers were well matched to the individual towers.

Table 2 [34,131 bytes] summarizes some of the more important operating parameters of the various towers.

The plant was to be revamped for more capacity in two stages. The first stage in 1997 would take the plant to a feed capacity of 60,000 b/d with a heavier feed coming from new deepwater developments in the Gulf of Mexico.

The second stage, to be completed by December of this year, will take the plant to its ultimate feed capacity of 70,000 b/d.

Energy efficiency improvements were also included in the revamp to reduce the energy consumption per unit of feed by about 17%. Product specifications were to remain the same, and the feed design basis is described in Table 1.

The proposed changes to the plant (Fig. 2 [85,288 bytes]) included heat integration of the debutanizer to the deisobutanizer with the overhead condensing duty of the debutanizer providing a significant fraction of the heat required for the deisobutanizer.

Other heat-recovery changes included preheating of the raw make using depropanizer condensing duty and subcooling of refrigeration propane with de-ethanizer overhead.

Additional cooling-water capacity was required as well as water exchangers to replace air coolers in some services that required tight temperature approaches.

New low-NOx burners were retrofitted to the hot-oil heaters (furnaces) that supply the entire heat load of the plant. A larger refrigeration chiller was added in the first stage and additional refrigeration compression capacity was to be added in the second stage of the project, to be completed late this year.

All columns in the plant were to be retrayed with high efficiency and capacity Shell HiFi and CS trays to replace the conventional multipass valve trays in the plant towers.

Retrofit

Retrofitting the debutanizer and deisobutanizer required the most creative process engineering approach. The following describes the most important issues around the integration and debottlenecking of these towers:

Increase debutanizer pressure: Provides hotter overhead that can be used to reboil deisobutanizer.
Decrease deisobutanizer pressure: Provides cooler bottoms that can be reboiled by debutanizer overhead; increases relative volatility of isobutane requiring lower reflux; decreases local hydraulic capacity available in tower shell.
Increase deisobutanizer tray count: Decreases tray spacing and local hydraulic capacity; increases theoretical stage count; and reduces reflux required.

The combination of the latter two items results in large energy savings and increased overall feed capacity as a result of reduced internal column traffic. Additional hydraulic capacity in the deisobutanizer at reduced pressure and tray spacing can only be achieved with use of Shell CS sieve trays.

Interesting in this analysis is that pressure and tray-spacing reduction are normally considered to reduce column capacity. In the case of the deisobutanizer, the gains from improved separation performance with more trays at lower pressure overcome the loss of hydraulic capacity by requiring lower reflux.

When this is combined with the superior flooding capacity of the Shell CS sieve trays, one can debottleneck the deisobutanizer by moving in the seemingly counter-intuitive direction of lower tray spacing (more trays) and lower pressure (better volatility).

The first three towers in the sequence were retrayed with Shell HiFi trays (Fig. 3 [99,509 bytes]) and the deisobutanizer used Shell CS trays (Fig. 4 [68,323 bytes]). This combination maximized the overall plant capacity.

The maximum capacity of the debutanizer and deisobutanizer with the new internals was set at 65,000 b/d of feed, whereas the ultimate capacity of the de-ethanizer and the depropanizer was now deemed to be 75,000 b/d of feed.

The overall plant ultimate capacity with additional refrigeration was set at 70,000 b/d. This meant that at maximum rates, a 5,000-b/d portion of the depropanizer bottoms would have to be sent to the adjacent refinery for processing.

Revamp results

The project was engineered, procured, and installed in 12 months.

Project execution and installation were performed on schedule. In addition, more than 240,000 man-hr were spent without a single OSHA-recordable incident (OSHA = U.S. Occupational Safety & Health Administration).

Total plant downtime for the retraying and piping tie-ins was about 30 days. During this time, 20,000 b/d of raw make were shipped to jug storage for later processing.

The fractionation plant has performed very well since the start-up. Implementation of Phase 2 is under way with completion set for December 1998, and the capacity expected out of Phase 1 has been exceeded (Table 3 [58,466 bytes]).

Some of the improvements to date include:

Maximum demonstrated capacity for all towers of 63,000 b/d of plant feed. This is 5% greater than revamp Phase 1 expectation and represents a feed-capacity increase of 46.5%.
Reduction in energy utilization from 2,470 BTU/gal to 2,060 BTU/gal for the same composition feed; or an improvement in energy efficiency of 17%.
All product specifications have been maintained.

The Authors

Pervaiz Nasir is a staff chemical engineer in the operations support group of Shell Midstream Enterprises. He holds a BS from Middle East Technical University, Ankara, an MS from University of Alberta, Edmonton, and a PhD from Rice University, Houston, all in chemical engineering. Jose L. Bravo is senior staff engineer with Shell Oil Products Co., Houston, responsible for separations process technology. Previously, he worked for Jaeger Products, Houston; the Separations Research Program, the University of Texas, Austin; and S&B Engineers, Houston. Bravo holds an MSChE from the University of Texas at Austin.

Robert Andring is a project manager in the engineering and construction group of Shell Midstream Enterprises. He holds a BS in electrical engineering from Washington State University, Pullman. He is a registered professional engineer in Texas.

Thomas K. Gaskin is a principal process engineer with Advanced Extraction Technologies Inc., Houston, and was lead process engineer for ABB Randall (1992-98) and senior process engineer for Shell Oil Co. (1977-92). He holds a BS (1977) in chemical engineering from Clarkson University (formerly Clarkson College of Technology), Potsdam, N.Y., and is a registered professional engineer in Texas.

Ross E. Mowrey is principal process engineer for ABB Randall Corp., Houston. He was previously employed by Black & Veatch Pritchard Inc. Mowrey holds a BS in chemical engineering from the University of Texas at Austin and is a member of he AIChE.