Hey everyone,
I am trying to fix our diluent trim control. Here is a quick description of the process and control:
Shipping oil from our booster pumps is mixed with butane (a set ratio entered by operations that changes seasonally), and diluent (light viscosity oil to bring our heavy oil blend viscosity down).
Here are the current process values obtained from insight:
Ultimate gain: 3.45
Ultimate period: 308.00 sec
Process dead time: 43.2 sec
Process gain: 2.2
Process time constant: 376.7 sec
Please keep in mind that the viscosity analyzer, AIC-3603 gets its signal from our custody transfer company that ships our oil. The meter is very far downstream of where we add the diluent and butane (trim site), and that is why we have a dead time. We have no viscosity measurement upstream of the shipping booster pumps, so we don't know the viscosity beforehand, and can only trim afterwards.
Also of note is the fact that we ship from our sales tank, so it usually has time to settle as we have other tanks fro the live process to feed into.
My main question is what kind of control philosophy should we employ for this process? We are currently using the AIC-3630 with a CAS set point to the diluent flow meter, 5-FIC-103. That set point is then multiplied with a characterize block, which I am not sure where they got those numbers from.
Any help with this process would be much appreciated, as I am not sure where to start.
Here is a trend with a modified PV scale so you can see the PV step changes better.
We try to maintain a set point of 380 cP, but our current control is lacking.
Jayme