Author Topic: Different default parametrization in ASU Tank vs Multiprobe sensor blocks  (Read 309 times)

arspr

  • Posts: 16
    • View Profile
This is a possible child issue I've detected while working on this one: https://forum.mikepoweredbydhi.com/index.php/topic,2470.msg28846.html#msg28846

I don't know if there are clear reasons but WEST default values are not fully coherent between ASU and Multiprobe sensor blocks (at least in ASM1Temp and ASM2dModTemp).

Different values I've found:

ASM1Temp
  • f_P is 0.08 in ASU tanks but 0.2 in Multiprobe sensors. As far as I know this is somehow not relevant: f_P in Multiprobe sensors is just used for BOD estimation, while f_P is quite important for ASU tank reactions, (decay of biomass).
  • By the way, ASU tanks don't have an i_N_S_I value. Therefore their TKN and TN values are slightly underestimated.

ASM2dModTemp
  • i_N_X_I is 0.02 in ASU tanks but 0.3 in Multiprobe sensors. I think this affects meassured Nitrogen continuity: multiprobe sensors are going to say there's quite more Nitrogen than the actually evaluated in the biological reactions in ASU tanks.
  • BTW, f_P, well f_X_I, is fully coherent here with a 0.1 value in both blocks.

Am I right?

Enrico Remigi

  • Posts: 320
    • View Profile
ASM1Temp
... ASU tanks don't have an i_N_S_I value. Therefore their TKN and TN values are slightly underestimated
Correct. Will fix it with the next SP.

ASM2dModTemp
i_N_X_I is 0.02 in ASU tanks but 0.3 in Multiprobe sensors.
True, but the actual value of the parameter may differ depending on the location of the sensor. And it can be changed at the level of the user interface.
The best way to ensure that the same parameter has the same value (if appropriate) across the layout is to create a top-level parameter.
Enrico U. Remigi
Wastewater Process Modeller
DK Urban
Kortrijk (Belgium)

Enrico Remigi

  • Posts: 320
    • View Profile
As a follow up to this issue: we have revised the implementation in ASM1 and confirm there is an inconsistency in the way T(K)N is calculated in the influent fractionation model, in the tank, in the sensor and in the effluent (defractionation) model.
Case 1
For the influent and effluent, the default model does NOT account for the nitrogen fraction of inert soluble COD. Same for the tank model. The sensor however does account for the term, with a factor (i_N_S_I) that is set to 0.01.
One easy way to solve this is to set i_N_S_I = 0 in the sensor. Then the 4 models will be fully consistent.
Case 2
However, if the inert organic nitrogen was not negligible, one should select the alternative fractionation and defractionation models (WEST.ASM1-2.Input.Layout.xml and WEST.ASM1-2.Output.Layout.xml) which incorporate the term; and use the i_N_S_I term in the sensor. The only real issue though is that the tank model (Gujer Matrix) would not allow for the same term to be taken into account.

In the upcoming Update 1, rel.2020/b], this will be sorted out:
  • the Gujer matrix of ASM1 will incorporate the S_I term for T(K)N
  • the default value of i_N_S_I will be set to 0.0 both in the matrix and in the sensor model

If inert organic nitrogen is negligible -> use the default model
If inert organic nitrogen is NOT negligible -> 1) use the alternative fractionation and defractionation ("-2") and 2) set i_N_S_I to the desired value (e.g. 0.01) both in the matrix and in the sensor
Enrico U. Remigi
Wastewater Process Modeller
DK Urban
Kortrijk (Belgium)