Dear Daniel,
I am trying to model an oil and gas central processing facility in DWSIM and I keep encountering problems with compressors. I have attached a sample simulation file that demonstrates the issue. I have tried this calculation on multiple computers with different solver settings. If GPU acceleration is disabled, then there is one type of error - failed flash with NaN for all component mole fractions. If GPU acceleration is enabled (in static mode only CUDA works, OpenGL immediately errors out; both work well in dynamics based on running sample files), there is a different type of error - "Sequence contains no elements". Screenshots of the errors are attached. "не число" is just NaN in Russian. I am using the latest 64 bit version of DWSIM (6.3 U5).
Also, I want to add a bit of a backstory on how the error occurs. The attached file is just part of the flowsheet that I am creating. Compressors works just fine initially, but at some point in time, they begin to error out. I have experienced this behavior in several different simulations, but they are using similar component sets. In the full flowsheet, this problem arises for 2 out of the 3 compressors used. Additionally, the Phase Envelope utility will error out (for example on the INLET stream in the attached file) with NaN mole fractions for all components. It's possible that these errors are related. I am using the PR EoS for the calculation and sometimes switching a compressor to SRK will revive it for a little while before the same problem begins to occur.
Will be grateful for any help!
Best regards,
Mikhail Evtushenko
Can you remove Helium from the simulation?
Removing Helium from the compound list does fix the compressor problem. So this is a viable workaround for now, but is this something that can be solved in the long term? Also, could this potentially be solved by creating a custom Helium compound or will it result in the same error?
I ran into another problem with the same simulation. I have a gas stream going into a low temperature separator at -40C. The separated liquid comes out at -40, but the vapor for some reason comes out at a different temperature. It is +77C in the attached file, but it changes every single time the flowsheet is recalculated. I have observed values from -15 to +78 C by recalculating this example. I have noticed this on at least one other separator in the simulation, but it is much more difficult to check on some of the other equipment in the simulation.
And another question, is there any way to expose variables from utilities/plugins to the flowsheet (spreadsheet/adjust)? First of all, this would be Reid Vapor Pressure from the Petroleum Properties Utility and HC/Water dew points from the Natural Gas Properties plugin.
Once again, thanks for any help.
Best regards,
Mikhail Evtushenko
I have been playing around with "Compressor Bug.dwxmz" above. I have a solution. My notes are in 'Miscellaneous under Settings.
Dear Frank/Daniel,
I have a bit of a follow up on the issues that I brought up. I spent quite a bit of time trying to figure out what was wrong and the biggest factor is that at some point in time I switched to the VLLE flash calculation type instead of the default (SVLLE). All of the separation issues that I was having are related to the VLLE flash mode. I recommend that it is either removed from DWSIM or at least be clearly marked as deprecated as the VLLE flash calculation is prone to errors, is unstable is calculation, and takes longer to calculate than SVLLE. I wrote a whole wall of text with examples before figuring it out, so I can share if need be.
As far as compressors with Helium, they only work in a certain temperature range for those particular working conditions. This is not a deal breaker but would be nice to get an update. I think that I will move the question about parameters from plugins/utilities to the general forum along with some others that I have accumulated.
Best regards,
Mikhail Evtushenko
Let's use the GitHub tracker for a better experience: https://github.com/DanWBR/dwsim6/issues