Dear all,
Please find attached a flowsheet file, simulating evaporating surfaces of a boiler as a part of a larger flowsheet.
When using a fixed vapor fraction for the drum, the flowsheet converges without any problem. When however the heat duty is fixed, convergence fails.
could you please have a look, and give me a clue on the cause?
thanks, and enjoy your weekend!
regards, Henk
Flash convergence issue
Moderator: jasper
Flash convergence issue
- Attachments
-
- EvapSurf.fsd
- (138.36 KiB) Downloaded 803 times
Re: Flash convergence issue
Nice icons!
Looks to me like the flash is the only unit operation in the loop that puts a constraint on temperature (implicitly, the outlet streams must be at the temperature at which the product splits are at specified fraction of the feed). Putting up a heat duty specification at the flash does not constrain temperature here; if the feed gets warmer, so do the products.
With nothing in the loop to constrain temperature temperature will just keep moving in one direction.
Putting a temperature constraint on the flash also does not make sense, as this is a pure compound at its boiling point. At this temperature everything may go over the top or over the bottom, both satisfy the imposed constraint.
Note that putting a heat duty on the flash, and a temperature spec anywhere else in all the loops should make that it converges fine. This is however not straight forward, as there are three loops into FlashMix, two of which are easy to put a temperature condition on, but if we do not do so on the loop that ends in stream 91, we get a similar situation; increasing T of stream 91 increases T of stream 25, and as nothing into the loop back to stream 91 fixes the temperature, this will not converge.
Looks to me like the flash is the only unit operation in the loop that puts a constraint on temperature (implicitly, the outlet streams must be at the temperature at which the product splits are at specified fraction of the feed). Putting up a heat duty specification at the flash does not constrain temperature here; if the feed gets warmer, so do the products.
With nothing in the loop to constrain temperature temperature will just keep moving in one direction.
Putting a temperature constraint on the flash also does not make sense, as this is a pure compound at its boiling point. At this temperature everything may go over the top or over the bottom, both satisfy the imposed constraint.
Note that putting a heat duty on the flash, and a temperature spec anywhere else in all the loops should make that it converges fine. This is however not straight forward, as there are three loops into FlashMix, two of which are easy to put a temperature condition on, but if we do not do so on the loop that ends in stream 91, we get a similar situation; increasing T of stream 91 increases T of stream 25, and as nothing into the loop back to stream 91 fixes the temperature, this will not converge.