logo


Oct-2010

Debottlenecking a refinery fuel 
gas absorber

A refinery fuel gas absorber was revamped to increase capacity while maintaining H2S in the product well below specification

Darius Remesat Koch-Glitsch Canada
Michael Beshara Irving Oil Refining

Viewed : 5772


Article Summary

The purpose of a fuel gas absorber is to selectively remove components, primarily H2S and to a lesser extent CO2, using a solvent (amines) that absorbs these specific components. The product fuel gas can then be burned with reduced environmental impact.

One of the fuel gas absorbers at the Irving Oil Refinery in Saint John, New Brunswick, Canada, had a maximum sustainable rate of approximately 980 mscfh. Increasing the gas flow rate beyond this point had resulted in increased column differential pressure (an indicator of the onset of column flooding) and amine carryover (increasing operating cost and operational challenges). This column was limiting the ability to increase overall plant capacity, since Irving Oil Refining has strict operating requirements for environmental stewardship.

Design objectives and path
Irving Oil Refining wanted to process as much material through the column while maintaining product quality (H2S in fuel gas not to exceed 50 ppm[v]) with minimum modifications to the plant during a planned shutdown in the autumn of 2009. It considered multiple options to debottleneck the column and settled on studying changes to column internals for increased throughput while maintaining or improving product quality. The newly designed high-performance trays would need to address the following criteria:
• The new design will take into consideration the foaming tendency of amine
• The expected rich amine loading shall not exceed API guidelines for carbon steel in specific amine service at the anticipated temperatures. 

Based on past successes with high-capacity trays at the site and from other references,1,2,3 Irving 
Oil Refining commissioned Koch-Glitsch to:
• Model the operation of the fuel gas absorber (C14001) and validate current operation versus design, based on a comprehensive unit test run conducted in January 2009 and on existing internal drawings
• Recommend and model internal changes to increase column capacity while retaining 60% turndown capability (maximum throughput with given constraints is desired) 
• Limit the extent of modifications to reusing the existing tray ring supports, including downcomers. Tray number and spacing to be retained, with 25 trays in total at 2ft spacing
• Retain current absorbent (amine at 25–30 wt%) and limit the flow and temperature that can be provided with existing equipment, such as recirculation pumps and exchangers
• Revamp work to fit within the set turnaround schedule.

Methods and tools
The first and most important step in any revamp study is to generate an accurate characterisation of the process.3 The test run performed in January 2009 gathered data using calibrated instrumentation, creating a closed mass and energy balance. The next step is to take the data from the test run and to create a representative model of the plant that can be used to predict the future performance with the new tower internals.

Choice of modelling program
Numerous programs are available to assist in representing a column that uses amines to remove H2S and CO2 from fuel gas streams. From the authors’ experience, rate-based models provide the best overall representation for new columns in this service, especially for packed columns. As an example, the rigorous, mass transfer rate approach used for all column calculations eliminates the need for empirical adjustments to simulate new applications correctly.

However, for column revamps, especially with trays, the use of an equilibrium-based model that has the necessary, proven adjustable parameters from operating experience is a suitable alternative to rate-based models, provided the necessary specific equipment characteristics of the high-capacity tray can be appropriately represented in the simulation model.

The simulation user needs to be sensitive to the fact that even the most sophisticated equilibrium-stage model uses only two of five elements employed in the rate-based model; namely, mass and energy balances around an entire ideal stage, plus thermodynamic-phase equilibrium. Programs that include reaction kinetics by empirical modelling via an adjustable parameter (H2S and CO2 tray efficiencies and/or liquid residence times) that forces the simulation 
to reproduce a conventionally 
operated column’s treated gas composition can only be effective 
if comprehensive operating 
experience has been gained and validated. In addition, the equilibrium-based program should have a reliable feature to include tray efficiencies to convert ideal stages into actual trays so that the tray characteristics can be represented pre- and post-revamp.

VMGSim4 uses specific mass transfer multipliers that can be tuned to match plant data and provides the ability to use tray and component efficiencies in the model. As a result, VMGSim has been used successfully to model existing plants and to accurately predict tray revamps in this service.

Of note, the solvents used in amine absorbers are rarely pure solutions of water and amine. Contaminants entering with the feed gas or makeup water can change the chemistry of the solvent significantly. This can both worsen and, in some cases, enhance the absorption efficiency. To improve the accuracy of the simulation, the impact of heat-stable salts and other contaminants on the performance of the amine should be factored into the evaluation.

Process evaluation
A simulation using VMGSim 
(equilibrium-based model) with an appropriate amine thermodynamic package (validated with both Protreat and Ratefrac rate-based models) was developed based on plant data provided from January 2009. The fuel gas absorber was running at ~921 mscfh charge to the unit. Simulation cases were run at:
• 921 mscfh to match plant data
• 980 mscfh demonstrated sustainable limit of absorber column performance
• 1175 mscfh based on expected acid gas loading limit of 0.6 (moles acid gas/moles amine).


Add your rating:

Current Rating: 3


Your rate: