Can We Count on Good Turndown in Two-Pass Moving-Valve Trays?

November
,
2018

At turndown, the efficiency of two-pass moving-valve trays may be lower than expected due to maldistribution of the vapor within the trays. This article explores this maldistribution and describes ways to alleviate it.

The moving-valve tray is one of the most common high-capacity tray types in distillation and absorption columns in the oil refining, petrochemical, and chemical industries. It provides excellent turndown — typically down to 20–25% of the maximum operating loads (1) — with good efficiency, because at low vapor loads the valves close, preventing liquid weep that is detrimental to tray efficiency. Excellent turndown is a common reason for designers and operators to favor moving-valve trays.

Recently, a few cases were reported where turndown of two-pass moving-valve trays fell short of expectations (2). This is of great concern, as most trays larger than 8 ft in diameter are two-pass or multipass. Although a few designers believe that maldistribution is the main culprit, the nature and mechanism of this maldistribution have remained a mystery.

To explore this maldistribution mechanism, this article applies the Fluor Multipass Maldistribution Model (MMM) (3) to discover multiple steady-state vapor/liquid distributions in two-pass moving-valve trays at turndown. Due to the symmetry of two-pass trays, a perfect split of both vapor and liquid between the passes is always a possible and well-known steady-state distribution. The other, previously unknown, steady states have uneven vapor distribution, in some cases highly so, with wide variations in liquid/vapor (L/V) ratios from pass to pass. This severe maldistribution is detrimental to tray efficiency.

The maldistribution described here is due to neither poor initial distribution of vapor or liquid from an inlet, nor differences in tray geometry between passes. The cause of the maldistribution is that at turndown, the valves close, and while closing, the dry (friction) pressure drop becomes independent of the vapor flowrate. This makes it easy for the vapor to swing from one pass to another, forming several alternative “preferred-path” steady states.

Using our model, we found that varying weir loads and uniform weeping have only a slight effect on the steady-state multiplicity. The multiplicity appears to be a vapor-driven phenomenon with little effect on or influence by the liquid. The magnitude of this turndown maldistribution can be largely reduced by using valves of different weights. This technique, practiced in many facilities, should be considered when potential for this type of maldistribution exists.

This article describes our investigation and findings. Additional details are available in Ref. 4.

Using the multipass maldistribution model

The tower modeled is 8 ft in diameter and contains six two-pass trays that are 24 in. apart (Figure 1). The tray decks are 14 gauge, with round moving valves that have a thickness of 16 gauge and a leg length of 7/16 in. The open slot area is 11% of the bubbling area.

image

Figure 1. The analysis described in this article is based on a tower with a diameter of 8 ft that contains six two-pass moving valve trays at 24-in. tray spacing. (Note: DC is short for downcomer.)

The model used in this article is an extension of our model to calculate maldistribution on four-pass trays (3). It is a two-pass maldistribution model that has six trays, hydraulically linked top to bottom, with a liquid feed (false downcomer) to the top tray and a vapor feed to the bottom tray. Total vapor and liquid loadings are constant throughout the tray section.

The calculations were performed for the separation of cyclohexane from normal heptane at 24 psia and total reflux. Abundant data for this system in single-pass sieve and valve trays have been published (5–7).

At pressures less than 100 psia, maximum tray capacity is usually limited by jet flooding. Jet flooding occurs when the intertray vapor velocity is high enough to uplift the bulk of the liquid drops populating the intertray spacing, causing liquid accumulation on the trays above.

Turndown is expressed relative to the jet-flood vapor velocity. Trays are usually designed to operate at 80–90% of the jet-flood vapor velocity. Turndown, expressed as a percentage of the actual vapor velocity to the jet-flood velocity, typically occurs at 20–50% of jet flood (i.e., 20–50% of the jet-flood vapor velocities).

Two equalities are required for a hydraulically balanced distribution in two-pass trays (Figure 2):

  • pressure drop. On trays with liquid flow from center to side, the center downcomer divides the vapor space, and the pressure on either side does not need to be equal. On trays with liquid flow from side to center, the vapor space is continuous, and the pressure equalizes in the shared space, i.e., every two trays. For a balanced solution, the total pressure drop (through both trays) on either side of the shared downcomer must be equal.
  • downcomer backup. Because the center downcomer is shared, the calculated downcomer backup for both sides must be equal.
image

Figure 2. Several key terms are used in the multipass maldistribution model, including hdc (height of liquid in the downcomer), L (liquid flowrate), V (vapor flowrate), and ΔP (pressure drop).

Our hydraulics equations are in the published literature and are the same as presented in our earlier article, “Preventing Maldistribution in Multi-Pass Trays” (3). For pressure drop calculations, we use the Klein method (8) and the Glitsch method (9), as we have had a lot of success with both. For downcomer backup, we use the classic downcomer backup equation in Perry’s Chemical Engineers’ Handbook (1) with the Klein method, and the Glitsch downcomer backup equation with the Glitsch method.

We closely analyzed the vapor and liquid distributions at 18%, 22%, 29%, 36%, and 40% of the estimated jet-flood point.

Steady-state multiplicity

Two-pass trays are symmetrical, so a perfectly balanced solution with a 50%/50% split of vapor and the same for liquid is always a hydraulically balanced solution. Our MMM analysis found two additional hydraulically balanced vapor and liquid distributions for two-pass moving-valve trays at some low loadings, a phenomenon we refer to as steady-state multiplicity. Such multiplicity is evident when using the Klein pressure drop method (8), but not when using the Glitsch pressure drop method (9), as described later.

For a moving-valve tray, the dry pressure drop increases with gas flowrate until the valves begin to open — the closed balance point (Figure 3). Then, as the gas flowrate continues to increase, the dry pressure drop stays constant (for the Klein method) from the point when the first valve opens to the point when all valves are open — the open balance point. Dry pressure drop again increases with the gas flowrate once all valves are open.

image

Figure 3. In a moving-valve tray, the dry pressure drop increases with gas flowrate to the point that the valves begin to open — this is called the closed balance point. Pressure drop remains constant from there until the point at which all valves are open — this is called the open balance point.

Multiple solutions with the Klein pressure drop method

Figure 4 shows the dry pressure drop, pressure drop through the liquid, and total pressure drop through Trays 4 and 5 on either side (Pass A and Pass B) of the center downcomer from Tray 5. The loading is 29% of the estimated jet-flood point, and the liquid is evenly distributed between Passes A and B for both trays. The x-axis is the fraction of vapor to Pass A, with the remainder going to Pass B.

image

Figure 4. The dry pressure drop, pressure drop through the liquid, and total pressure drop through Trays 4 and 5 were calculated using the Klein pressure drop method, with Pass A shown in red and Pass B shown in blue. The loading is 29% of the estimated jet-flood point.

The dashed lines represent the calculated dry pressure drop through both trays for either pass. The Pass A dry pressure drop resembles the dry pressure drop characteristic in Figure 3, and Pass B is its mirror image. The dotted lines represent the pressure drop through the tray...

Author Bios: 

Matthew R. Olsson

Matthew R. Olsson was part of the Distillation Expertise Team at Fluor Corp., in both Sugar Land, TX, and Aliso Viejo, CA. He has now moved to Eastman Chemical Co. in Longview, TX, where he is a process design engineer. He has ten years of experience in design, troubleshooting, and revamping fractionation processes and equipment. Olsson obtained his BS in chemical engineering from Texas A&M Univ. (College Station)....Read more

Henry Z. Kister

Henry Z. Kister is a Fluor Corp. senior fellow and director of fractionation technology (phone 1-949-349-4679, email henry.kister@fluor.com). He has over 30 years experience in design, troubleshooting, revamping, field consulting, control and startup of fractionation processes and equipment. He is the author of three books, the distillation equipment chapter in Perry’s Chemical Engineers' Handbook, and over 100 articles, and has taught the IChemE-sponsored “Practical Distillation Technology” course more than 350 times. A recipient of several...Read more

Would you like to access the complete CEP Article?

No problem. You just have to complete the following steps.

You have completed 0 of 2 steps.

  1. Log in

    You must be logged in to view this content. Log in now.

  2. AIChE Membership

    You must be an AIChE member to view this article. Join now.

Copyright Permissions: 

Would you like to reuse content from CEP Magazine? It’s easy to request permission to reuse content. Simply click here to connect instantly to licensing services, where you can choose from a list of options regarding how you would like to reuse the desired content and complete the transaction.