Bulk Temperature of Diluted Caustic at Mixing Point

[gkennedy34]

I've modeled heat of solution effects in Aspen Plus for diluting 50% caustic to 10% in an attempt to determine resultant temperatures at varying concentrations. This temperature is being compared against a piping specification to ensure the piping is adequate for this service. The resultant temperature curve for diluting the 50% caustic ventures into the area of the piping specification "service graph" where it requires stress relieving welds and bends. The maximum temperature is around 30% dilution (by weight), but the final temperature at 10% is well below the stress relief requirement zone.

In my mind, as long as the full amount of dilution water is delivered at the mixing point, the bulk temperature of the liquid should remain below the "stress relief" zone. In other words, the enthalpy of the combined streams is instantaneous and the heat of solution effects will raise the temperature to its final resultant temperature.

Is my thinking flawed?

Thanks.

gk

Side note: I compared Aspen Plus' results with a empirical reference curve and found very good agreement with dilution from 30% to 10%. Aspen Plus' resultant temperatures from 50% to 30% were significantly higher. Since Aspen's numbers were on the conservative side, I think I am okay (as long as my thinking above is not flawed). Once I got "okay" agreement with the reference curve I changed my dilution water and 50% caustic stream temperatures to model the system I was interested in.

 

[kenvlach]

Caustic solution (50 wt%) doesn't mix that well with water; it goes to the bottom & requires stirring. So, you may get localized overheating at the mixing interface.

People mixing caustic soda pellets with water in plastic tanks have burned holes through the bottoms. Solution is far better, though, but it should enter through a diffuser or other mixer.

OxyChem's Caustic Soda Handbook may be helpful:

http://www.oxy.com/oxychem/Products/caustic_soda/literature/caustic[1].pdf

 

[gkennedy34]

Thank you for the link, kenvlach. I saw a graph in there I had not seen before which will allow me to double-check some of my Aspen numbers.

Yes, we would likely use a diffuser or mixer of some sort.

Thanks again.

 

[stevethermo]

As you may remember from a lab class, if doing the mixing in a batch process, such as in a tank, add the water first, then add the caustic, while mixing vigorously, and you shouldn't have a problem. Cold water is better.

An Enthalpy chart for aqueous solutions, such as the one in the Oxy link, is very useful. (A talented engineer could develop one from data generated in Aspen Plus.) The temperature follows a line connected from the pure water side of the diagram to the original caustic concentration. The temperature path upon mixing depends on which fluid is in the tank first, and ends at the final concentration. Using this type of chart will let you know whether or not to start with the solution first, or the water first. Some charts also have pressure curves, so that you will know if you are getting close to the boiling point of the solution.

If mixing with the caustic first, some part of the mixture will be at the peak temperature on the chart. There is no way around it.

If mixing in a pipe, injecting the caustic into the middle of the water stream would be better than the alternative, but settling may occur. I would consider a static mixer to minimize this. Some parts of the system may still need to be heat treated.

 

[cub3bead]

Look at the Resistoflex web page. They make plastic lined dilution tees for both acids and caustic.

 

[lastone]

At my facility we offload 45% KOH from a railcar. The 45% KOH is pumped from the railcar into a mixing tee where it is diluted to 20% wt with water. A setpoint is entered for the KOH flow, and the water flow is ratio controlled to supply the correct dilution ratio. Just downstream of the mixing tee the temperature is read. If a high temperature alarm is reached, a signal is sent to automatically shut down the KOH offloading pump and the water flow. The system works well. I don't know if the mixing tee does a perfect job in the pipe or not, but since it goes to a tank that runs on constant circulation, I don't have an issue with imperfect mixing. We also take a sample about 10 minutes after the offloading begins to ensure we are making the proper ratio.