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

Fluid Heating Systems

Heat input choices for a vaporizing system

Anhydrous ammonia feed to a selective catalytic reduction system

Subject: Thermosyphon reboiler considerations
Date: Tue, 27 Aug 2002 19:39:15 -0400


We have a proposed application for a thermosyphon reboiler (TB) that will be associated with an anhydrous ammonia system. The configuration consists of the following:

  1. 35,000 gallon anhydrous ammonia storage tank
  2. TB
  3. Selective catalytic reduction system (SCR).

The SCR is installed in the flue gas stream (temp ~ 550 F) at a coal fired power plant. The catalyst reacts ammonia (NH3) with nitrogen oxides (mostly NO) to produce N2 and H2O (nitrogen oxides contribute to ozone formation).

Some SCR installations feed the anhydrous ammonia (which actually contains 0.2 to 0.5 weight % water) from the storage tanks to a vaporizer/dilution air system (this dilutes the ammonia vapor to 5% ammonia by volume), and then into the flue gas stream upstream of the catalyst bed. In these situations, there is no holdup in the vaporizers, so what goes in goes out.

The TB installation proposes to simply circulate ammonia vapor (two-phase flow?) back into the storage tank. The vapor from the ammonia storage tank is then fed to the SCR system.

A typical operating condition would be 75 F temperature in the storage tank, and 1000 lb feed of vapor from the tank to the SCR system. The purpose of the TB would be to maintain the pressure in the tank.

I am concerned about the gradual buildup of water in the storage tank, and the TB. At 75 F and "anhydrous ammonia" with 0.5 wt % water, the saturation pressure is about 140 psia, and the concentration of water in the vapor phase is very low (0.000006 wt %), as opposed to the 0.5% is the liquid mixture.

Thus, the vapor that is fed out of the tank is almost pure ammonia. In this case, the water must accumulate in the storage tank. Thus the physical properties of the fluid that is feeding the TB will change over time, the net effect being a higher boiling point. In addition, the density of the feed from the tank changes with increasing water concentration. The vapor density should also change, primarily due to the density of the liquid phase (I am assuming here [maybe a bad one] that there is two phase flow at the discharge of the TB).

It seems to me that in this situation, unless the tank is purged from time to time, then there is the potential for the flow out of the TB will decrease over time and eventually "stall" unless a careful design is made.

There is also the problem with the periodic replenishment of the ammonia supply in the storage tank. This is done from time to time, and there is the issue that when you mix essentially anhydrous ammonia (i.e. 0.5% water) from the supply truck there is an exothermic reaction, which will cause the bulk contents in the tank to heat up, the internal pressure to increase, and for venting to occur (these things are designed with safety relief valves in the vapor space).

Can you comment on the applicability of a TB for this application? My real concern is the potential for the TB to stall unless there is significant over design to account for the change in the feed properties over time. Am I off base here? The accumulation of water in the tank is unavoidable, simple vapor liquid equilibrium stuff. I don't know anything about TB.

Thanks in advance,

W. P.
North American Power Plant
(edited for clarity)

To: W. P.
Subject: Thermosyphon reboiler considerations
Date: Wed, 28 Aug 2002


You have outlined a number of technical issues here. These include:

  1. Suitability of a thermosyphon design versus other design choices.
  2. Trace component buildup in the anhydrous ammonia tank.
  3. Exothermic heat of mixing driven vaporization when adding anhydrous ammonia to the tank.

Most of the comments will address the first issue but some comments will be made on points two and three, however, a full discussion of these points is out of our scope here. As a general note, you could certainly make a thermosyphon work if you wanted to. Nevertheless, a thermosyphon seems unnecessarily complicated for the service unless other design factors not noted in your question come into play. Everything else being equal, I'd probably favor a kettle type or a stab-in type for this system.

As a basis we will assume a new installation where the major selecting factors will be economic and technical rather than having to accommodate the equipment into an existing storage and vaporization system. Essentially, we can look at five major types of systems:

  1. In-line vaporizer (similar to the 'conventional' design you discussed).
  2. Stab-in bundle
  3. Thermosyphon
  4. Forced circulation
  5. Kettle

In-Line Vaporizer

An in-line vaporizer takes a flow controlled stream from the vaporizer and vaporizes it directly.

Disadvantages: ??

Advantages: no buildup of water in the system: tank can be at grade.

Figure 1
In-line vaporizer

Stab-In Bundle

A stab-in bundle is an exchanger bundle with its tubesheet attached to a flange on the ammonia vessel (or tank). These are commonly used in tank heating applications and in some distillation towers.

Disadvantages: requires space in the vessel to disengage liquid and vapor: in many vessels will set a minimum liquid inventory (the bundle must be covered) that may not be wanted: does nothing to help with water accumulation issue.

Advantages: simple: minimal to no piping required: no moving parts: tank can be at grade.

Figure 2
Stab-in bundle


A thermosyphon is a natural circulation loop that drives circulation around the loop by the head difference between the inlet side of the loop and the outlet side of the loop. In principle it is possible to have a liquid circulation loop that works by density differences between a liquid a two temperatures. However, practical thermosyphon loops in the refining and chemical industry always have a two-phase (liquid-vapor mixture) to reduce density dramatically on the outlet side of the loop to generate the required fluid driving force. Thermosyphon exchangers have to be located below the feed liquid pool.

Disadvantages: exchanger must be located below tank, this may force tank to be elevated: complex loop hydraulics, as a rule, the more distance between the minimum liquid level in the tank and the thermosyphon return to the tank, the more distance required between the tank and the thermosyphon: no help for the water build-up. As you note, composition differences that change the boiling point of the liquid can stall this system if the heating medium is not hot enough to maintain vaporization in the loop. Density differences on the inlet side should have a minor affect unless you eventually build up extremely high water contents. In general, the higher density liquid on the supply side of the exchanger will help the loop circulate rather than stalling the system.

Advantages: no moving parts.

Figure 3
Thermosyphon system:
The larger the difference in draw and return points (H1) the larger the clear liquid driving force is required (H2) to assure that the system flows at all conditions.

Forced Circulation

A forced circulation system uses a pump to force fluid flow at all conditions.

Disadvantages: more equipment: higher operating cost: no help with water build up.

Advantages: tank can be mounted at grade or with a small elevation (for NPSH): no complex fluid loop.

Figure 4
Forced circulation system


A kettle system uses a kettle exchanger to vaporize liquid in it. Normally, kettle exchangers include an internal overflow weir that creates a still liquid area inside the exchanger. Level control on the still liquid sump is a common method of product control from these exchangers. In your case, the kettle would not have the internal weir and the liquid level on the exchanger would be used to control flow from the tank. As long as sufficient pressure differential was available, no pump would be needed between the tank and the kettle exchanger. If pressure differentials were unfavorable, a pump would be needed.

Disadvantages: larger exchanger vessel needed

Advantages: no complex flow hydraulics: moves water concentration point out of the main storage tank and into the exchanger, this may make purging water from the system easier.

Figure 13
Kettle system:
System shown is on pressure balance for flow from tank.
With a pump (or with sufficient elevation difference) the system could have the flow to SCR on pressure control on the tank.

Andrew Sloley

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

This page updated 28 August 2002.
© 2002 Andrew W. Sloley. All rights reserved.