Improving performance through low-cost modification of tower internals

Low-cost revamps of tower internals improve production from existing assets with a payback period of less than a year

Darius Remesat
Koch-Glitsch

Viewed : 4694

Article Summary

During tough economic times, operators are challenged with increasing margins and decreasing capital expenditures. However, operators can pursue low-cost/high-return opportunities that are within their existing plant capital budgets and will improve product recovery, increase capacity and/or improve reliability (for example, by increasing run length). These projects do not require large-scale engineering support, but can be accomplished with the assistance of knowledgeable professionals who have the skills and experience appropriate to the specific project. Many of these opportunities are uncovered during normal maintenance planning, where justification for the project is based on incremental cost to upgrade tower internals set against the cost to simply replace the internals in kind. As a result, most tower internal revamps can be justified as standalone projects (with less than six months’ payback time) at the plant level.

General approach
To capture these low-cost/high-return projects, the following tools, tests, data and analyses should be included in the project scope:
• Accurate feed characterisation
• Detailed data from plant operations and test runs
• Inside-out” design approach1
• Up-to-date design guidelines
• Appropriate simulation thermodynamics and topology
• Simulate actual tower internals characteristics
• Simulate actual trays (with tray efficiency) not just theoretical stages
• Computational fluid dynamic (CFD) analysis and/or tower gamma scans where beneficial.

Proper feed characterisation and detailed test run data are essential to obtain an adequate representation in simulation form. They provide the basis for making recommendations on revamps. Data should be compiled, reconciled and regressed to minimise errors between plant data and the subsequent simulation.

Once a satisfactory data set is developed, the simulation strategy needs to be set. The inside-out design approach,1 together with appropriate thermodynamics and simulation topology, has been used successfully in applying tower internal characteristics into the simulation. This methodology helps to develop a representative model that effectively predicts future tower performance post-revamp and provides detailed operating conditions to design each tower internal. For certain critical separation applications (for instance, vacuum columns, coker and FCC main fractionators), using CFD analysis to model the behaviour of certain tower internals under revamp conditions is crucial in setting an optimal design.

Using an iterative process to modify the equipment’s characteristics (tower diameter, heat exchanger heat duty, control valve opening, pump and compressor capacity) within the simulation and in the tower internals design procedure can help to squeeze as much capacity or increase recovery within the associated equipment’s operating envelope. The intent in this exercise is to capture as much improvement in performance from a tower internal modification without requiring expenditure to debottleneck other equipment. In addition, to maintain an overall low project cost for the revamp, the use of mechanical considerations such as minimal (or no) welding, hinged-joint active area panels that reduce bolting requirements, and modular construction can reduce installation time.

Case studies
A simple and cost-effective way to improve recovery (separation) within a distillation column is to increase contact between liquid and vapour by improving the vapour/liquid contact for a given device and/or increasing the number of devices in the separation column.

Figure 1 illustrates the concept of increased stage count to improve separation or to reduce the duty requirements on a column for a given separation. The graph shows that as the number of stages increases, the heat duty for the column decreases. In general, in a revamp for a set heat duty, the amount of separation increases when the number of trays increases (as long as flood point is not reached).

Case study 1: improve hydrocracker fractionator recovery
The operator was replacing tower internals to improve diesel recovery and prepare for future unit flow increases of 15%. The changes to the column were made during the normal maintenance schedule.

The hydrocracker fractionator separates the hydrocracker reactor outlet into light gas (C4 and lighter), heavy and/or light naphtha, diesel, kerosene and bottoms. During this process, the hydrocarbon is reacted over catalyst in the presence of hydrogen.
To increase diesel recovery and to handle 15% higher flow, the stripping and wash section were revamped with Superfrac trays. Numerous refiners have successfully used this specific revamp design.^2,3

In the stripping section located below the feed to the column, two trays were added to the existing six trays. The Superfrac trays increased capacity and improved vapour/liquid interaction. Figure 2 illustrates the revamped stripping section. Since the liquid and vapour rates were much lower than the rate needed, the design used a can with an internal diameter 40% of the full tower diameter. The can arrangement provides much better liquid and vapour contact, as Table 1 indicates. Prior to the revamp, the tray efficiency used to match the stripping section performance was 14% compared to a more typical value of approximately 25%. The difference in efficiency can be attributed to the overly large existing tower diameter. The can arrangement matches the cross-sectional area to the post-revamp flow rates. Also, for this particular case, the can arrangement shortened installation time, because of the modular construction and reduced part handling and bolting needed within the column. Parallel baffles can alternatively be used when the diameter difference between existing and suggested revamp is small.

Due to elevated flow rates in the wash zone, Superfrac trays replaced sieve trays on a one-for-one basis. The improvement in tray efficiency as a result of using the can arrangement and Superfrac trays was 12.5 percentage points per tray — nearly double the existing tray efficiency.