Skip to main content

I am interested to know which product is being produced in chemical reactor in series… I know the current product at the exit of production, but not which product is being run in each reactor. I know the residence time in the different reactors. Is there any possibility to create a formula indicating which product is being produced in each reactor by applying a variable time shift which depends on the residence time?

 

Hi Alberto,

If I understand correctly, the residence time is fixed and known. In that case I think the most straightforward solution would be to take the product tag (which gives the product at the end of Reactor N) and shift it backwards in the active tag menu to match another reactor:

  • For Reactor N-1, put a timeshift of the negative (=shift backwards) of the residence time of Reactor N
  • For Reactor N-2, a shift of -(residence time N + residence time N-1)
  • … 

If you save a TrendHub view with the shifted tag, it will be remembered as shifted for that view. You could thus create a view for each reactor, with the shifted product tag specifying the product at the end of that reactor.

Let us know should this solution not fit your use case!

Wouter


hi Daniel,

unfortunately no, residence time can be follows by production rate (kg/h), density (kg/m3) and reactor volume (m3)… but production rate is changing depending on market needs, raw material availability and even which kind of product (each product is reacted longer or shorten in each reactor to achieve a differentiated performance).

Thus, residence time can be followed and calculated for each reactor, but it is not fixed.


I see! While I believe it would be possible to derive the product in the earlier reactors from the volumetric flow data, that would require a highly specific calculation that in my view would only be possible using the advanced custom calculations feature (https://userguide.trendminer.com/en/tag-builder--custom-calculations.html).

In my experience, if it is in any way possible, it is better to have such tags for the other reactors created on a lower level in the data (DCS, Historian), than to calculate them later in TrendMiner.

Of course if the product type is really only determined at the last reactor, and there is no way of knowing the final product type in the upstream reactors, a calculation based on the volumetric flow (some sort of chain of backwards totalizing integrations of the volumetric flows to see where it matches the reactor volumes) becomes the only possibility.


Reply