Piping for Pilot Plant

[Alphite]

I'm trying to size piping to several jacketed tanks (25L, 40L, 60L). Everything will be heated with low pressure steam, and cooled with chilled water.

I calculated the approximate cooling time, and my heat transfer value U appears to be correct based on actual results from a similar tank. I then used the standard m-Cp-deltaT to approximate the cooling load (dividing by cooling time to get kW). To do this, I assumed the heat loss for the water inside the tank was the same as the heat increase in the fluid inside the jacket. I read somewhere that a 5C increase was typical for cooling water, and back-calculated the cooling water flowrate based on this.

If this is all making sense, I am now a little uncertain as to what the pipe size should be. The chilled water flowrate is higher than I expected (6-8 gpm). For a 40L tank, I was assuming it'd be around 3/8 or 1/2 inch NPT. After much searching, even most 3/4 inch lines only average 4-5 gpm, but a larger line seems ridiculous. I think there's an error in my calculations, but I'm not sure what I'm missing.

At the same time, a single control valve will be used for either steam or chilled water. I've seen this work elsewhere, but it also seems a little strange, since steam is so much more efficient at transferring heat?

Is there a resource or guide that explains this? The run from the header is fairly short, so I'm not worried about friction losses. Any help or advice would be appreciated. I can share excel calculations if it helps.

 

[BronYrAur]

The single valve is concerning. Are you dumping the steam and chilled water down the drain once it passes through the jacket? Not good for either system. When you say chilled water, do you just mean domestic cold water?

 

[mk3223]

IMO, the required flow of steam or water depend on the efficiency of the heat transfer of the jacketed tank. And, the flow of the water may be increased for 3/4" pipe (say, to 8 gpm) if the supply pressure can be higher.
The other option is, instead of the jacketed tank, to consider the process with an external heat exchanger to improve the heat transfer efficiency.

 

[Alphite]

Thanks for the replies!

Bronyraur:

I have to confirm, it's possible the previous systems had two separate control valves, one for steam in the top, and another for chilled water in the bottom. Everything is behind a wall and there are no drawings...

There is a separate return line for chilled water and condensate, they are recovered and the tank jacket is drained before switching between the two. It is actually chilled water, it passes through a heat exchanger and comes out around 35-40F, supply pressure is 30 psig.

mk3223

I used the heat transfer coefficient for similar tanks (determined experimentally), and estimated that 6-8 gpm are required for this 40L tank. I guess this is reasonable?

My main concern is how to confirm that a given pipe size will supply the correct flowrate. The entire plant system is not well defined, so it'd be hard to do a complete pressure balance. I'm hoping that there are at least general rules of thumb that I can use, until more data is gathered (1/2 = 2 gpm, 3/4 = 5 gpm, etc). What also concerns me is that a 30L tank supplier uses a 3/4" inlet, and a different 85L supplier uses 3/8" inlet. There seems to be no consistency...

One last question...The control valve for an existing tank is a v-notch, and it was about 50% open for most of the cooling. Does this mean that only 50% of the flow is supplied, or, if there's enough pressure, the same flowrate is delivered, it just moves faster through the valve? I will research on my own, but hopefully it's an easy answer...

 

[mk3223]

Without knowing the detail, I guessed one of the reason for the small pipe in the large tank may be because the smaller thermal load was required for the process. you should identify and compare the operation parameters for both processes. The higher thermal load required, the higher flow needed as well as the larger pipe at one fixed pressure drop.
The type of the control valve could be designed either linear, equal percentage, quick opening, etc. The 50% opening may not equivalent to the 50% of the flow. You should verify the valve type first and find out the flow character in your system.

 

[MJCronin]

Talk with the prospective plant operator.

While your modest pipe sizes seem reasonable for maintaining temperatures in a cooking process....

However, if the operator assumes that you will be required to heat a tank up from ambient to some elevated temperature in a reasonable time (less than a week).... then you may need more steam.

Same thing for the cooling circuit .... It has been my experience that plant staff will complain bitterly if the steam system cannot heat things up in a very short period of time .... they will bitch that the system is undersized and "engineered wrong"

Address their expectations up front...

MJCronin
Sr. Process Engineer

 

[lilliput1]

Be careful in your calculation. Note that jacketed vessel heat transfer can not be counter flow so you will not be able to attain as high a log mean temperature difference as typical counter flow heat exchangers or coils.