The
Distillation Group, Inc. Home Services Newsletter Contact Distillation Technical Software Training Questions Links Search Navigation

Control of Parallel Overhead Condensers

Methanol Distillation

To: asloley@distillationgroup.com
Subject: Parallel condensors in Methanol Distillation column
Date: Tue, 27 Aug 2002 14:04:30 +0600
From: D. B. S. Asian Chemical Company

Dear Mr. Sloley:

We are carrying out revamp of our Methanol distillation column with new trays. The existing condenser is limiting. In the column we intend to install a new condenser in parallel with the existing condenser. The new condenser is for and additional 25% capacity. We would like to know the precautions to be taken for
designing/installing a new condenser to take care of hydraulics and pressure control of the column at various loads.

D. B. S.

To: DBS
Subject: Parallel condensors in Methanol Distillation column
Date: Wed, 28 Aug 2002
From: asloley@distillationgroup.com


Do you mean that you are:

A. Installing a second stab-in condenser in parallel with an existing stab-in bundle?
B. Installing a second bundle in parallel in a conventional overhead system?

A. Sloley

To: Andrew Sloley <asloley@distillationgroup.com>
Subject: Re: Parallel condensors in Methanol Distillation column
Date: Thu, 29 Aug 2002 11:33:38 +0500
From: DBS

Thanks for the prompt reply.

We are installing a second condenser (condenser already available from another plant) in parallel with the existing condenser. Hence, vapour from the column will be distributed into the two condensers. We are worried about following:

HIGH LOAD OPERATION
Pressure drop calculated across 1st condenser is 0.15 bar, condensing capacity 15000 kg/hr vapour. Pressure drop across parallel condenser is 0.1 bar, condensing capacity 5000 kg/hr.

Now when the plant operates at 20000 kg/hr of vapour from the tower, ideally the vapour to the existing condenser should be 15000 kg/hr & new condenser should be 5000 kg/hr. But as pressure drop across the new condenser is low, more vapour will be pushed through this condenser and due to limitation of condensing capacity of only 5000 kg/hr, we envisage uncondensed vapour from new exchanger.

To control the distribution, we are thinking to install a low pressure drop control valve on the vapour line to the new condenser.

What is your experience with parallel condensors ?

If we use this valve & during operation this valve is throttled very much, is there any chance of vacuum creation in the condenser.

I am attaching the schematic of the system for better understanding.

To: DBS
Subject: Parallel condensors in Methanol Distillation column
Date: Wed, 28 Aug 2002
From: asloley@distillationgroup.com

Overall, my experience with parallel condensers has been very satisfactory. In any unit revamp you have to balance between adding condensers in parallel and adding them in series. Both choices have different circumstances that favor them.

The best situation for parallel exchangers is to have identical exchangers with a symmetrical piping network. This solves most major flow distribution problems. However, non-identical exchangers in parallel can also work. It is common practice to put a low-pressure drop (often butterfly) valve to prevent excessive flow through an exchanger that has a lower pressure drop at desired operating conditions. I have done this many times myself. The real question in this configuration is what you control the butterfly valve with. We will come back to this point later.

Your other question was could the system in the new exchanger run in vacuum if the flow to the new exchanger went to a very low rate. Based on your simplified piping diagram (Figure 1), the answer is no. If you have low flow to the second exchanger and it acts as a total condenser and starts to create a pressure in the shell lower than the existing exchanger, you will get vapor flow from the drum to the new exchanger. If your existing pressure control system is adequate to prevent vacuum in the drum it will continue to operate adequately.

Figure 1
Proposed overhead system

Control of your two exchanger system is the real question. Figure 2 shows adding a butterfly valve to the feed to the new exchanger. Figure 3 shows the alternate location for this valve, in the outlet line to the drum. While the line is larger, I recommend the location in the line to the drum (Figure 2). In your configuration, putting the control valve in the outlet line (Figure 3) has the potential to create two problems. First, you could hold a liquid leg on the valve that partially floods the new condenser and reduces its capacity below your desired level. This could be avoided by careful place of the exchanger and valve relative to the reflux drum. Second, you use a nitrogen injection upstream of the condensers to prevent drum pressure from going too low. The nitrogen may serve two purposes: A-it partially blankets the heat-transfer surface of the exchangers dropping heat transfer and B-adding vapor to the drum to increase its pressure. The nitrogen guarantees that the flow in the outlet lines from the condenser to the reflux drum is two-phase. Control valves in two-phase flow have erratic performance and many maintenance problems. This configuration should be avoided. Of course, you could always move the nitrogen make-up to the drum directly.

Figure 2
Recommended location for flow control valve

Figure 3
Potential problems with flow control in outlet line

Staying with the good practice of having the control valve in the feed line to the new exchanger, the simplest system would be to use a flow orifice in the lines to the condensers to measure flow and use a flow ratio control on the valve to maintain a constant split between the exchangers (Figure 4). You need to be careful in this case to prevent condensation that forms upstream of the orifice plates from accumulating. Liquid pools upstream of the orifice make the flow readings inaccurate. Additionally, in some piping layouts a liquid leg upstream of an orifice can accumulate and significantly raised the tower pressure. On one memorable troubleshooting occasion, pressure field pressure readings revealed a 7 meter tall liquid stack created by combining a poor piping layout with a readily condensable overhead and an orifice plate.

Figure 4
Flow control to parallel exchangers

The tempting alternative of using a DP cell across the exchangers with a differential pressure ratio control (Figure 5) should be resisted in your case. You have only a 0.15 bar (2.2 psi) pressure drop across one exchanger and a 0.10 bar (1.5 psi) pressure drop across the other. You would be trying to control on a small number derived from the difference between two larger numbers. This is a classic error that is always warned against but often made. The differential pressures would also decrease at low flow rates, making the situation worse.

Figure 5
Differential pressure control to parallel exchangers
(Only recommended if differential pressures are high)

If using direct flow control is unattractive, you can control to have a differential temperature of zero between two thermocouples placed downstream of the condensers (Figure 6). I have used this system with great success in several installations.

Figure 6
Outlet temperature control

Best Regards.
Andrew Sloley
DGI

Images have been sized for full screen display on an 800x600 monitor.

This page updated 30 August 2002.
© 2002 Andrew W. Sloley. All rights reserved.