Perspectives: Plant InSites

Distillation Columns: Get Some Inside Information

A gravity-flow loop can enable measuring internal liquid flow in a column

By Andrew Sloley

Throughput or fractionation performance in specific systems may vary dramatically depending upon how precisely the internal liquid rate can be controlled close to its limit. Four basic methods can determine the liquid rate within a column:

• calculating heat and material balances using heat input, heat output and external stream rates and properties;
• taking a total draw and then returning liquid through a metered pumped-flow loop;
• taking a total draw and then returning liquid through a metered gravity-flow loop; and
• measuring a column liquid level to infer flow.

Each of these has pros and cons; several or all four might work for a given system. Physical layout and data precision impose particular requirements for each option. Specific systems tend to favor one option or another. Getting the details right counts.

Here, let’s specifically look at taking a total draw and then returning the liquid through a metered gravity-flow loop.

Figure 1 is a partial elevation sketch of a 42-in.-diameter distillation tower. Feed enters the tower between Bed 2 and Bed 3. Beneath Bed 2, a collector tray takes liquid out as a total draw. A gravity-fed loop is used to measure the flow rate of liquid from that collector tray.

I first applied this configuration nearly 25 years ago in a specialty lubricants distillation unit as part of a larger revamp. Discussing that job after the fact, another senior engineer emphatically stated that this type of flow measurement system never could work. He had seen many attempts but none had succeeded. This surprised me because the client was very pleased with its performance. So, we sat down together to review what was built.

Figure 1 highlights seven key features:

1. Collector tray. This serves to gather the liquid. For a packed tower, it’s a default choice. Most applications of this configuration will have a low liquid rate. (A future column will cover the “why.”) Therefore, the tray must be fully welded. A collector tray is a good idea even in a tower with trays. However, with the correct modifications, a special tray configuration may suffice.

2. Nozzle sized for self-vented flow. Vapor bubbles entrained in the liquid reduce system capacity and give incorrect flow-meter readings. Self-vented nozzles have a low enough liquid velocity that vapor can escape and return to the tower. (See: “Assess the Gravity of the Situation.")

3. Sufficient static head. H1 is the minimum distance required to generate enough static head for pressure drop to drive flow through the system. This sets the minimum distance between the draw nozzle and return nozzle elevations. Pipe diameter often is reduced downstream of the self-venting nozzle to cut costs. In that case, H1 must be based on the elevation of the pipe diameter reduction, as shown in Figure 1. In an effort to minimize this height, high-β-value orifice plates often are used to decrease pressure drop.

4. Adequate upstream pipe run. H2 is the minimum upstream pipe run to get well-conditioned flow needed for accurate metering. With orifice flow meters, the higher the β value, the longer the pipe run required. (See: “Think Straight about Orifice Plates.”)

5. Flow instrument in liquid-filled location. The meter can be installed in either a vertical (as shown) or horizontal position. In all cases, the piping configuration must ensure the line is filled with liquid.

6. Long-enough downstream pipe run. H3 is the minimum downstream pipe run necessary for well-conditioned flow for meter accuracy.

7. Suitable distance between the return and lowest-pressure points. H4 is the minimum interval from the return elevation to the lowest-pressure point (vena contracta) to prevent vaporization downstream of the orifice plate. The vena contracta occurs some distance after the orifice plate. Most times, it’s close enough that the orifice plate location can be used to set elevations. Nevertheless, engineers with a good grasp of fundamentals never forget the difference. Avoiding vaporization is important to getting accurate flow measurements. Either H3 or H4 may set the minimum distance from the flow meter to the return nozzle.

One extra point not shown — insulation — is important. In hot systems, a little heat loss makes H4 less likely to set the system. In cold systems, heat gain easily may make H4 dominant. More insulation helps. Vaporization in the system also dramatically may boost pressure drop, increasing H1 as well.

These points apply equally to both packed and trayed towers.

Going back to the discussion with my colleague: After reviewing the system, he asked: “Just how tall was this loop?” The final system with its vertically mounted meter was nearly 30-ft high. Gravity-feed flow-measurement systems require ample height. A horizontal mounting may make H2 and H3 unimportant but H1 and H4 never go away.


ANDREW SLOLEY is a Chemical Processing contributing editor. You can email him at ASloley@putman.net

More from this perspective...

Title

Don’t let parallel pumps cross you up

Small differences with parallel pumps can lead to big difficulties, according to Andrew Sloley in this month's Plant InSites column.

03/28/2007

Keep others from making “classic” mistakes

A heat exchanger problem provides a textbook example of design limitations, according to Contributing Editor Andrew Sloley.

04/24/2007

Teaming two solutions can be counterproductive

Sometimes things just don’t add up as expected. Andrew Sloley discusses how sometimes two solutions aren't always better than one solution.

06/06/2007

Properly protect centrifugal pumps

Consider various factors when selecting how to guard against low flow, Andrew Sloley says in this month's Plant Insites column.

07/11/2007

Don’t look at precondensers in a vacuum

Consider the impact of the downstream vacuum ejector on precondenser performance, advises Andrew Sloley, in this month's Plant InSites column.

08/20/2007

Correct those disturbing lapses

Keeping liquid in a reflux drum still can prevent problems, says Andrew Sloley, in this month's Plant InSites column.

08/31/2007

Do some heavy thinking about light loads

Cutting steam demand can get you into hot water, advises Contributing Editor Andrew W. Sloley, contributing editor, in this month's Plant InSites column.

10/01/2007

Carefully check vendor references

Determine the differences between a cited installation and yours, Andrew Sloley, contributing editor, says in this month's Plant InSites column.

10/30/2007

Keep parallel pumps in line

Mind the curves when using different size pumps.

02/13/2008

Don’t stack the deck against yourself

Always consider future expansions before finalizing layouts

03/14/2008

Plant InSites: Cast a cold eye on columns

Subcooling can complicate and compromise tower performance

05/23/2008

Understand the cold facts about subcooling

Systems sometimes can surmount major layout errors

06/04/2008

Use water to check vessel integrity

Hydrostatic testing offers advantages but requires care

06/25/2008

Right Vacuum Control Choice Takes the Pressure Off

Choosing the proper recycle stream for regulation is crucial

08/07/2008

Plant InSites: Look Beyond the Lore

Don’t rely on recollections about how a unit had performed

10/10/2008

Full Vacuum Rating Isn't an Empty Benefit

It often can provide important but underappreciated advantages for vessels

10/16/2008

Stripper Bares All

Analysis reveals how operating practices compounded design mistakes

11/21/2008

Steam Systems: Simple Solutions Can Prompt Complex Problems

Steam systems are especially susceptible to developing difficulties

12/23/2008

Troubleshooting: Take to the Field

Troubleshooting demands getting into the plant to inspect units.

01/30/2009

Take Control Down a Level

Sometimes you can design out the need for instrumentation.

02/23/2009