Brittleness/high dependency of `high_rate_wastewater_treatment_system` on feed

[mretier]

Hi BIOSTEAM team,
I'd like to use the high_rate_wastewater_treatment_system for wastewater treatment.
However, the system fails to converge in, for me, unintuitive ways.

Below is a contained example that fails to converge:

import biosteam as bst
import thermosteam as tmo

chemicals = bst.Chemicals(['CO2', 'AceticAcid', 'FormicAcid', 'H2O', 'O2', 'Methanol', 'Methane', 'N2'])

bst.settings.set_thermo(chemicals)

feed = bst.Stream(
	CO2=0.0468,
	AceticAcid=0.624,
	FormicAcid=0.00169,
	H2O=1.03e5,
	O2=0.0323,
	Methanol=4.39,
	T=273.15+37,
)

wwt_sys = bst.create_high_rate_wastewater_treatment_system(ins=feed)
wwt_sys.simulate()

Troubleshooting to figure out the feed components that caused failure to converge resulted in the following:

1. No convergence when removing O2, CO2 or both O2 and CO2
2. Convergence when removing Methanol from feed, but the solution has negative O2 flows in the outs of R603 and S601
3. Convergence when increasing O2 by 3 to 3.0323. In this case no negative O2 flows are observed
4. No convergence when changing water inflow by orders of magnitude

Am I misunderstanding something fundamental here?
Thanks for any input!

Best,
M

[yalinli2]

Hi @mretier , thanks for reporting the issue and did the detective work.

The core issue here is that the gases (CO2, CH4, O2, N2) are not locked at gas phase, so they incorrectly changed the moisture content calculation of the streams, and because of the belt thickener has moisture content thresholds (if moisture content too low, this unit will be skipped), so it was skipped and broke the recycle loop.

Locking the chemicals at gas phase should fix the error:

CO2 = bst.Chemical('CO2', phase='g')
O2 = bst.Chemical('O2', phase='g')
Methane = bst.Chemical('Methane', phase='g')
N2 = bst.Chemical('N2', phase='g')

chemicals = bst.Chemicals([CO2, 'AceticAcid', 'FormicAcid', O2, 'H2O', 'Methanol', Methane, N2])

bst.settings.set_thermo(chemicals)

feed = bst.Stream(
     AceticAcid=62.4,
     FormicAcid=16.9,
     H2O=1.03e5,
     Methanol=4.39,
     T=273.15+37,
)

wwt_sys = bst.create_high_rate_wastewater_treatment_system(ins=feed)
wwt_sys.simulate()

This being said, there were some internal bugs in the polishing filter that caused the negative O2 flow, I had some codes to back-calculate the O2/air needs, but apparently it wasn't robust enough.

Other thoughts: for the feed, it has no solids and little organics, the COD of this stream is only ~134 mg/L, this is even lower than municipal wastewater, and you don't need the high-rate WWT process, it's designed for COD on the level of 10x-100x of that.

I'll send in a PR for the bug fix, and then close this issue when it's done. Thanks!