Controlling local fluid age in a stirred reactor

Patent metadata

Field	Value
Publication number	US-10543473-B2
Application number	US-201515319414-A
Country	US
Kind code	B2
Filing date	Jun 4, 2015
Priority date	Jun 24, 2014
Publication date	Jan 28, 2020
Grant date	Jan 28, 2020

Abstract

In a stirred tank chemical reactor the mean age of the reactor contents affects a number of properties of the product, including for example the homogeneity of the product. The mean average age of the reactor contents can be determined by constructing a transparent model of the reactor and filling it with a fluid containing a fluorescent dye and having flow properties comparable to those of the reactor in use. A light is shone on the fluid as it is stirred under reaction conditions and a clear fluid flow into the model. Pictures are taken of the reactor contents and the mean fluid age of the contents of the model are determined relative to the exit age of the contents. This approach can be applied to determine for example which reactor ports to use, what agitator to use, what flow rates to use to improve reactor function.

First claim

The invention claimed is:

1. A method to determine a preferred inlet location for a continuous stirred tank reactor comprising:
1) preparing a transparent model reactor comprising, an agitator, a number of alternate inlet ports, and at least one planar exterior surface, which model reactor will be used to generate data to be stored from subsequent steps of the method;
2) filling the model reactor with a test fluid having a viscosity chosen to produce the desired Reynolds number, which fluid contains a uniform concentration of a tracer dye which fluoresces under a known wavelength of incident light;
3) flowing six times the model reactor volume of test fluid without dye into and through the model reactor using a selected inlet port with the agitator running at a constant speed to provide a constant Reynolds number to introduce a step change in concentration of the tracer dye in the model reactor;
4) while the dye is being washed out of the model reactor and during mixing, subjecting the model reactor contents to a plane of laser light through the planar exterior surface of the model reactor, whose wavelength causes the dye to fluoresce at a different wavelength from that of the laser;
5) taking digital images of the fluorescent light from the model reactor through a filter to filter out the incident laser light at a rate synchronized with the rate of rotation of the agitator;
6) calculating a normalized dye concentration in each pixel of each image;
7) integrating the normalized concentration in each pixel over all images in time to calculate an average local fluid age of each pixel;
8) calculating a standard deviation and a mean of the local ages and a coefficient of variance;
9) iteratively repeating steps 1 through 8 for different inlets to the model reactor, wherein the calculated normalized dye concentration, calculated average local fluid age, calculated standard deviation and mean of the local ages, and calculated coefficient of variance become stored data;
10) comparing the stored data to determine the inlet which provides at least one of:
i) a normalized dye concentration from 0.5 to 1.3;
ii) a coefficient of variance from 0.01 to 0.1; and
iii) a minimum standard deviation in the average local age of the model reactor contents; and
iv) a minimum standard deviation in the average age of the model reactor contents, and

11) selecting the input port, which provides the preferred values from step 8.

2. A method to determine a preferred agitator for a continuous stirred tank reactor (CSTR) reactor comprising:
1) preparing a transparent model reactor comprising, an agitator, a number of alternate inlet ports, and at least one planar exterior surface, which model reactor will be used to generate data to be stored from subsequent steps of the method;
2) filling the model reactor with a test fluid having a viscosity chosen to produce the desired Reynolds number, which fluid contains a uniform concentration of a tracer dye which fluoresces under a known wavelength of incident light;
3) flowing six times the model reactor volume of test fluid without dye into and through the model reactor using a selected inlet port and agitator running at a constant speed to provide a constant Reynolds number to introduce a step change in concentration of the tracer dye in the model reactor;
4) while the dye is being washed out of the model reactor and during mixing, subjecting the model reactor contents to a plane of laser light through the planar exterior surface of the model reactor, whose wavelength causes the dye to fluoresce at a different wavelength from that of the laser;
5) taking digital images of the fluorescent light from the model reactor through a filter to filter out the incident laser light at a rate synchronized with the rate of rotation of the agitator;
6) calculating a normalized dye concentration in each pixel of each image;
7) integrating the normalized concentration in each pixel over all images in time to calculate an average local fluid age of each pixel;
8) calculating a standard deviation and a mean of the local ages and a coefficient of variance;
9) iteratively repeating steps 1 through 8 for different agitators, wherein the calculated normalized dye concentration, calculated average local fluid age, calculated standard deviation and mean of the local ages, and calculated coefficient of variance become stored data;
10) comparing the stored data to determine the agitator which provides at least one of:
i) a normalized dye concentration from 0.5 to 1.3;
ii) a coefficient of variance from 0.01 to 0.1; and
iii) a minimum standard deviation in the average age of the model reactor contents, and

11) selecting the agitator which provides the preferred values from step 8.

3. A method to determine a preferred feed flow rate to a continuous stirred tank reactor (CSTR) reactor comprising:
1) preparing a transparent model reactor comprising, an agitator, a number of alternate inlet ports, and at least one planar exterior surface, which model reactor will be used to generate data to be stored from subsequent steps of the method;
2) filling the model reactor with a test fluid having a viscosity chosen to produce the desired Reynolds number, which fluid contains a uniform concentration of a tracer dye which fluoresces under a known wavelength of incident light;
3) flowing six times the model reactor volume of test fluid without dye into and through the model reactor at a constant flow rate using a selected feed port and agitator running at a constant speed to provide a constant Reynolds number to introduce a step change in concentration of the tracer dye in the model reactor;
4) while the dye is being washed out of the model reactor and during mixing, subjecting the model reactor contents to a plane of laser light through the planar exterior surface of the model reactor, whose wavelength causes the dye to fluoresce at a different wavelength from that of the laser;
5) taking digital images of the fluorescent light from the model reactor through a filter to filter out the incident laser light at a rate synchronized with the rate of rotation of the agitator;
6) calculating a normalized dye concentration in each pixel of each image;
7) integrating the normalized concentration in each pixel over all images in time to calculate an average local fluid age of each pixel;
8) calculating a standard deviation and a mean of the local ages and a coefficient of variance;
9) iteratively repeating steps 1 through 8 for different feed flow rates to the model reactor, wherein the calculated normalized dye concentration, calculated average local fluid age, calculated standard deviation and mean of the local ages, and calculated coefficient of variance become stored data;
10) comparing the stored data to determine the model reactor feed rate which provides at least one of:
i) a normalized dye concentration from 0.5 to 1.3;
ii) a coefficient of variance from 0.01 to 0.1; and
iii) a minimum standard deviation in the average age of the model reactor contents, and

11) selecting the model reactor feed rate which provides the preferred values from step 8.

4. A method to determine a preferred agitator speed for a continuous stirred tank reactor (CSTR) reactor comprising:
1) preparing a transparent model reactor comprising, an agitator, a number of alternate inlet ports, and at least one planar exterior surface, which model reactor will be used to generate data to be stored from subsequent steps of the method;
2) filling the model reactor with a test fluid having a viscosity chosen to produce the desired Reynolds number, which fluid contains a uniform concentration of a tracer dye which fluoresces under a known wavelength of incident light;
3) flowing six times the model reactor volume of test fluid without dye into

Assignees

• Nova Chem Int Sa

Inventors

• Hawryluk Andrew

Classifications

• B01J2219/00015

  Scale-up · CPC title

• B01J2219/00011

  Laboratory-scale plants · CPC title

• G01N21/6408

  with measurement of decay time, time resolved fluorescence · CPC title

• B01F2215/0409

  Relationships between different variables defining features or parameters of the apparatus or process · CPC title

• B01J2219/00166

  controlling the residence time inside the reactor vessel · CPC title