4 replies to this topic

[pawan]

Hi all, I am really stuck at this.

There is a CSTR for EO addition reaction. A ring type sparger for liquid EO (Ehtylene oxide) is given at the bottom. It is thru 1" pipe with 3 mm 57 no holes.

My objective is to find out the maximum possible flow of EO thru this sparger.

When I calculated the same considering Liquid EO & using Orifice equations I found very high flow but Actually we are restricted to only ~1500 Kg/hr EO rate.

So I checked heat transfer eqn for flow thru dip pipe & found that it will generate ~5% vaporization in the sparger feed pipe due to high rxr temp of 160°C. ( EO NBP is 10.5°C ).

Now the problem is that I could not find any correlation for two phase flow estimate in orifices.

Interesting thing is that when I consider 100% vapor EO as feed then it is coming ~1584 Kg/hr. But by heat transfer equation it should not be more than 5%.


Any suggestion......Any kind.........is welcome for solving this.

[mishra.anand72@gmail.com]

If holes are not plugged, max. flow possible is 1500 because of two phase flow phenomenon happening.

[joerd]

If the 5% means 5 mole-% or 5 weight-%, I would encourage you to calculate the volume-% vapor. This will show you that on a volume basis you have mostly vapor (about 75%). Then you can calculate the flow through the orifice using the orifice equations for vapor, using the mixture properties (calculated as the weighted average of vapor and liquid). Let us know what you find.

[Art Montemayor]

Pawan:

I have gone through this type of problem before, in batch kettle reactors with 2,000 and 4,000 gallons capacity. We also sparged Ethylene Oxide into our kettle reactor – but had no troubles.

I don’t know what kind of orifice equation you are using and under what assumptions, but if you consider the normal situation in trying to introduce a compressed, liquefied gas into a lower pressure vessel you will find that several things are happening. I also don’t know what kind of heat transfer calculation leads you to believe that you have only 5% vaporization – and is that % weight or % volume?

There are no credible 2-phase flow predictions through orifice plates, to my knowledge. And I doubt if you will find any that make sense in real-life applications.

If you want to “find out the maximum possible flow of EO thru this sparger”, then you are going to have to assume either 100% EO liquid or 100% EO vapor going through it. The easiest thing to do is to vaporize your EO prior to introducing it into the sparger – i.e., sparging only vapor EO. This allows you to fix the maximum amount of EO, using choke flow as a criterion. To assume that you are going to have 100% liquid flow through the orifice, you have to make sure that you are throttling back enough to maintain the EO in the line in the liquid state – and you will find that this is practically impossible unless you replace the orifice plate with a back pressure throttle valve. That’s why it is impractical to try to sparge in liquid EO. No matter what type of operation you have, you must be basically doing the following in order to sparge in EO to the lower-pressure reactor:

1. You are taking liquid EO from the main EO storage tank (where it is stored as a saturated liquid);
2. You are pumping the amount of liquid EO required for sparging by using a conventional pump;
3. You are metering the EO that is going to the reactor for sparging;
4. You are throttling the metered EO downstream in order to maintain the upstream EO in a constant, liquid state;
5. The fact that you have to throttle the EO means you have flashed the liquid in an adiabatic, isenthalpic manner;
6. The result of the flashing is a 2-phase mixture that either is directly sparged into the reactor or is vaporized before being sparged. In the former case, the cooled mixture acts as a heat sink for an exothermic reaction and introduces temperature control problems within the reactor during the sparging.

The best and most efficient manner of controlling the sparging rate and the reactor temperature is to feed vaporized EO at a set, controlled temperature – and that usually is in the vaporized state, just as Joerd infers in his post.

I hope this experience helps you.

[pawan]

Dear Art
Thanks for your time & advice.

I am attaching the rough schematic for better clarity of the issue.

I have used conventional restriction orifice equations for liquid. For heat transfer I simply used nusselt equation developed based on simulation models by some other producer for the same type of product & reactor i.e. Nu = 0.036 * Re^0.8 * Pr ^1/3 * (L/d)^0.055

Where L is the submerged length of dip pipe (responsible for heat transfer from hot reactor content to feed liquid EO) d is dip pipe ID. This gives me 5% Wt% EO vaporization which reduces the mix density from 860 Kg/m3 to ~266 Kg/M3 based on reciprocal law of mix density.

So regarding your point

There are no credible 2-phase flow predictions through orifice plates, to my knowledge. And I doubt if you will find any that make sense in real-life applications

in this case, the liquid EO is not completely vaporized when inlet P = 13 Ata & T = 10-13°C which are measured values from the system. SO it MUST be two phase flow. This is necessary as my system pressure including liquid head is ~3.1 Atm in the reactor which is ~41°C of Saturation temp of EO. So either it should get completely vaporized in the pipe by heat transfer equal to 1500 Kg/Hr OR its mixed flow.

Based on above Heat txr eqn I calculated a rate of 20 kW while for 1500 Kg/hr it comes out 821 kW @ 547 KJ/Kg latent heat of EO.

By all this I am trying to discuss the real life problem.

Because of high pressure system design probably reducing the pressure for vaporization & increasing the flow is not possible in the practical situation as EO cant be sparged alone.

Further I am also confused (Even after so much exp) that the above heat txr equation is giving a Temp of EO as 45°C while Saturation temp at system pressure is 41°C as mentioned above. But heat txr rate is not sufficient for vaporization of 1500 Kg/hr. This is the confusion as I am unable to visualize the problem part here.

Attached Files

•  Doc3.doc   19KB   77 downloads