Variable Dowtherm System Return Pressure

[gte682n]

Hello, I have a dowtherm system (40MMBTU/hr) that supplies 6 heat exchangers at 60 psig/660 degF. The Dowtherm control valves are all upstream of each exchanger, so the heating fluid comes in superheated to each exchanger. The system's return pressure is dictated by the largest Dowtherm user. When this user requires a lot of heat, it's control valve is more open, and the return pressure is higher. When the heat requirement is low, the control valve is pinched back and the return pressure is lower. A consequence of the lower return pressure is that the dowtherm has more superheat since the supply pressure/temp is the same, but more pressure drop is taken at each control valve. This has a noticeable effect on reduced heat transfer.

Is there a recommended way to stabilize the system return pressure throughout various system heat loads? The condensate return of each exchanger is gravity drained to a receiver, then pumped back to the heater. There are no liquid seal legs on any of the users. I have read elsewhere that this could help stabilize the return pressure, but do not understand how.

Thanks.

 

[Latexman]

Are you using dowtherm as a common noun or a proper noun? I.e., is Dow your supplier? The reason I ask, I asked our supplier to review our P&IDs, procedures, visit the plants, and make improvement recommendations. And, WOW! They really did a good job. They helped us solve some issues we were having. Maybe Dow provides this service too. You have nothing to lose and lots to gain by asking.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[don1980]

For good heat transfer control, the exchanger needs have its own separate condensate collection pot on the condensing (Dowtherm) side. When you're transferring heat into a stream by condensing a vapor, the condensing side pressure needs to be able to fluctuate, as demanded by the system. A condensate pot allows this to happen. This is how a condensing exchanger is controlled - by delta-T changes on the hot (condensing) side which occur due to pressure changes. This will be obvious to you when you look at the governing equation (Q=UAdT). The U and A are fixed values, so in order to vary the Q you need to be able to vary the dT. That's done by allowing the pressure to fluctuate. When you need more Q, the inlet valve opens further, which in turn causes the pressure to increase on the condensing side. This higher condensing pressure results in a higher dT for the exchanger. As your system is now, this doesn't (and can't) occur.

 

[StoneCold]

So if the Control valves are on the inlet side of the condensers, and the condensate runs into collection vessels that pump back to the heater, then the exchangers must be running at slightly higher than atmospheric pressure. So to raise the pressure up in the exchangers you need either the control valves on the outlet of the exchanger or to put pressure on the condensate receiver and add a liquid leg to the exchangers. Then they will be running at a higher pressure but not blinded by nitrogen, or whatever gas you use as a blanket for the condensate receivers. How much above the normal boiling point is 660F?

Regards
StoneCold

 

[gte682n]

The condensate receiver operates at 45 psig and 580 deg F, so it is subcooled liquid. There is no inert blanket on the vessel. The system pressure is dictated by the temperature requirement of the Dowtherm exiting the heater, i.e. saturated vapor at 660 F = 60 psig. Return pressure is set by the user with the highest outlet pressure. Normal BP is 495 deg F.

So, a liquid leg would create additional back pressure on each exchanger, but wouldn't it still vary with swings in the largest user?

Thanks

 

[Compositepro]

So now you have clarified that your heating system supplies saturated hot vapor and returns sub-cooled liquid. This is basically a steam system except for using Dowtherm instead of water. So where are your "steam" traps?

 

[gte682n]

No traps anywhere on the system. Condensate gravity drains to the receiver.

 

[Compositepro]

I think that is your problem. Would you design a steam heating system that way? If you did, how do you think that would work? You would have the problems you have described.