Methods of monitoring and controlling the melt index of a polyolefin product during production
Methods for producing polyolefin polymers may use a predictive melt index regression to estimate the melt index of the polyolefin during production based on the composition of the gas phase and, optionally, the concentration of catalyst in the reactor or reactor operating conditions. Such predictive melt index regression may include multiple terms to account for concentration of ICA in the reactor, optionally concentration of hydrogen in the reactor, optionally concentration of comonomer in the reactor, optionally the catalyst composition, and optionally reactor operating conditions. One or more terms may independently be represented by a smoothing function that incorporates a time constant.
1. A method comprising:
providing a predictive melt index regression derived at least in part on data from a previous production run of a first polyolefin having a first melt index formed by reacting an olefin monomer with a catalyst system in the presence of an induced condensing agent (ICA) and optionally hydrogen, wherein the predictive melt index regression is based on an effect of a concentration of the ICA in the reactor and optionally an effect of a concentration of hydrogen in the reactor on the melt index of the first polyolefin, and wherein a smoothing function using a time constant represents the concentration of the ICA in the predictive melt index regression;
contacting in a fluidized bed gas phase reactor the olefin monomer with the catalyst system in the presence of an ICA and optionally hydrogen to produce a second polyolefin having a second melt index;
monitoring a reactor ICA concentration and optionally a reactor hydrogen concentration;
calculating a predicted melt index for the second polyolefin using the predictive melt index regression based on a change to the reactor ICA concentration; and
adjusting the reactor ICA concentration based on the predicted melt index to maintain the second melt index within or move the second melt index to within a melt index threshold range.
2. The method of claim 1 , wherein terms of the predictive melt index regression associated with the concentration of the ICA and the concentration of the hydrogen independently include a smoothing function that incorporates a time constant.
3. The method of claim 1 , wherein the catalyst system includes a high molecular weight (HMW) catalyst and a low molecular weight (LMW) catalyst, wherein the method further includes:
monitoring a mole ratio of the BMW catalyst and the LMW catalyst (HMW mol :LMW mol ) in the reactor, and
calculating the predictive melt index regression further based on an effect of the HMW mol :LMW mol on the melt index of the polyolefin.
4. The method of claim 3 , further comprising adjusting the mole ratio of the HMW catalyst and the LMW catalyst (HMW mol :LMW mol ) in the reactor based on the predicted melt index.
5. The method of claim 1 , wherein the time constant associated with the concentration of the ICA is about 100 minutes to about 300 minutes.
6. The method of claim 1 , wherein the smoothing function is an exponential smoothing function.
7. The method of claim 1 , wherein the predictive melt index regression includes an exponential function.
8. The method of claim 1 , wherein the concentration of the ICA is represented by a partial pressure of the ICA (ICA PP ).
9. The method of claim 1 , wherein the concentration of the hydrogen is represented by a mole percent of total reactor gas ratio of the hydrogen to the olefin monomer (H2 mol :olefin mol ).
10. The method of claim 1 , wherein the predictive melt index regression includes at least one additional term associated with reactor temperature or polymer residence time.
11. A method comprising:
providing a predictive melt index regression derived at least in part on data from a previous production run of a first polyolefin having a first melt index formed by reacting an olefin monomer and at least one comonomer with a catalyst system in the presence of an induced condensing agent (ICA) and optionally hydrogen, wherein the predictive melt index regression is based on an effect of a concentration of the ICA in the reactor, an effect of a concentration of the comonomer in the reactor, and optionally an effect of a concentration of hydrogen in the reactor on the melt index of the first polyolefin, and wherein a smoothing function using a time constant represents the concentration of the ICA in the predictive melt index regression;
contacting in a fluidized bed gas phase reactor the olefin monomer and the at least one comonomer with the catalyst system in the presence of the ICA and optionally the hydrogen to produce a second polyolefin having a second melt index;
monitoring a reactor ICA concentration and optionally a reactor hydrogen concentration;
calculating a predicted melt index for the second polyolefin using the predictive melt index regression based on a change to the reactor ICA concentration; and
adjusting the reactor ICA concentration based on the predicted melt index to maintain the second melt index within or move the second melt index to within a melt index threshold range.
12. The method of claim 11 , wherein terms of the predictive melt index regression associated with the concentration of the ICA, the concentration of the comonomer, and the concentration of the hydrogen, if used, each independently include a smoothing function that incorporates a time constant.
13. The method of claim 11 , wherein the catalyst system includes a high molecular weight (HMW) catalyst and a low molecular weight (LMW) catalyst, wherein the method further includes:
monitoring a mole ratio of the HMW catalyst and the LMW catalyst (HMW mol :LMW mol ) in the reactor, and
calculating the predictive melt index regression further based on an effect of the HMW mol :LMW mol on the melt index of the polyolefin.
14. The method of claim 11 , further comprising adjusting the mole ratio of the HMW catalyst and the LMW catalyst (HMW mol :LMW mol ) in the reactor based on the predicted melt index.
15. The method of claim 11 , wherein the time constant associated with the concentration of the ICA is about 100 minutes to about 300 minutes.
16. The method of claim 11 , wherein the smoothing function is an exponential smoothing function.
17. The method of claim 11 , wherein the predictive melt index regression is an exponential function.
18. The method of claim 11 , wherein the concentration of the ICA is represented by a partial pressure of the ICA (ICA PP ).
19. The method of claim 11 , wherein the concentration of the hydrogen is represented by a mole percent of total reactor gas ratio of the hydrogen to the olefin monomer (H2 mol :olefin mol ).