Evaporation of the tube-side coefficient in a heat exchanger

Hoplite

I've got a problem. I'm trying to design a reactor with internal cooling provided by water flowing through tubes directly within the reactor. Basically, it's like a shell and tube heat exchanger with an internal heat source.

The problem is that the cooling water is to be evaporated within the reactor, and I don't know how to account for the resulting increase in volume. If I pump the coolant in at a standard liquid speed of 2m/s, then the exist vapour will be at several kilometers per second. This is no doubt well above the speed of sound in steam and would obviously damage the pipes.

So, is it safe to just include a single expansion joint in the tubes within the reactor so that the inlet and outlet speeds are within safe range, or would that just cause other problems?

I can't find anything in the literature about evaporation of coolant streams, but I'm sure it's done in industry.

Astronuc

How did you figure that? I think the flow resistance would prevent that. One cannot do a simple mass balance and change in specific volume.

What are the parameters of the cooling system.

One may need to pressurize it, in which case, it is normal to have an accumulator on such a system to account for changes in volume.

Depending on the temperature, cooling with liquid metal (NaK) might be better, but that has it's own complications.

Hoplite

I just realised I used the word "coefficient" in the title instead of coolant for some reason.

Well, if the mass of water entering the tubes is equal to the mass exiting, then the flow speed should be inversely proportional to the density. Of course, it not so possible to pump steam at that speed, but that’s my point.
260C isothermal reactor temperature with 110MW of cooling water required.
Do you mean to pressurise it in order to prevent evaporation, to reduce the exit steam’s volume? I guess that might be the only option. Thanks.
The decision to use water has already been made though. The steam has already been earmarked for use elsewhere within the process.

Q_Goest

Hi Hoplite,
Yes, that's exactly the reason higher pressure is desirable as Astronuc suggests. If your reactor is at 260 C, consider a temperature just below that which can give you a decent dT between the reactor and water. Let's use 20 C just for starters. The final temperature depends considerably on heat transfer from the tubes to reactor gas. You want the reactor gas to tube heat transfer coefficient to be high enough to perform the cooling required.

Given a water saturation temperature of 240 C (260 C - 20 C) the saturation pressure for water is about 485 psia. Thus the saturated vapor density is about 1.045 lbm/ft3. You can control to this temperature by putting a thermocouple on the discharge and varying flow either using a control valve or VFD on your pump.

We do something similar for cryogenic systems in which a vaporizer is used to warm a stream of cryogenic liquid to ambient temperature. In this case, the discharge piping of the vaporizer is sized to handle the pumped flow assuming warm gas on the outlet of the vaporizer. I think you'd want to do the same thing, size the steam pipe to handle the vapor required. I'd think your company, assuming you work in industry, already has these sorts of tools available to do these calculations efficiently. Do you work in industry, or is this just a hypothetical question for a school project?

Astronuc

I was wondering about the points made by Q_Goest.

This might be of use -

http://www.thermexcel.com/english/tables/eaubou1.htm
http://www.thermexcel.com/english/tables/vapeau1.htm


If this is a school or hypothetical project, that is one thing, but if this is an actual industrial design project, then must observe the applicable standards for design of Boiler and Pressure Vessels (e.g. ASME BPV Code).

Are the requirements for the steam supply 110 MW of steam at 260°C? Then one must also know the superheat or is it at saturation.

Hoplite

Thanks for your help, guys.

It's actually the final design project for my degree, so it has to be designed in detail, but without the information we'd have available in industry.