Sensor network design and inverse modeling for reactor condition monitoring

Patent metadata

Field	Value
Publication number	US-10401164-B2
Application number	US-201314037953-A
Country	US
Kind code	B2
Filing date	Sep 26, 2013
Priority date	Oct 16, 2012
Publication date	Sep 3, 2019
Grant date	Sep 3, 2019

Abstract

Methods and systems for detecting a condition within a component of a process plant, including obtaining a first and second set of parameter measurements from a first and second plurality of sensor locations along a first dimension of an outside surface of a component, processing the first and second set of parameter measurements to develop a continuous surface condition profile of the component using a predetermined model. The predetermined model includes a forward solution to an equation describing the condition and is linearly separable in at least two dimensions corresponding to the first and second dimension. The model includes an inverse solution to the equation to provide a set of coefficients corresponding to a set of basis functions of the forward solution based on the first and second set of parameter measurements.

First claim

The invention claimed is:

1. A method for detecting a condition within a component of a process plant, comprising:
providing an array of a plurality of sensors on an outside surface of the component, wherein providing the array of the plurality of sensors comprises:
locating a first plurality of sensors of the array of the plurality of sensors at a first plurality of sensor locations along a first dimension of the outside surface of the component;
locating a second plurality of sensors of the array of the plurality of sensors at a second plurality of sensor locations along a second dimension of the outside surface of the component; and
locating at least one additional sensor of the array of the plurality of sensors on at least one additional sensor location along the outside surface of the component;

obtaining a first set of parameter measurements from the first plurality of sensors at the first plurality of sensor locations along the first dimension of the outside surface of the component;
obtaining at least a second set of parameter measurements from the second plurality of sensors at the second plurality of sensor locations along the second dimension of the outside surface of the component;
processing the first set of parameter measurements and the second set of parameter measurements, using a computer, to develop a continuous surface condition profile of the component using a predetermined model, wherein the array of the plurality of sensors is operatively connected to the computer, the model comprising:
a forward solution to an equation describing a condition within the component, the forward solution being linearly separable in at least two dimensions, the at least two dimensions corresponding to the first and second dimensions of the outer surface of the component, and
an inverse solution to the equation to provide a set of coefficients, corresponding to a set of basis functions of the forward solution, based on the first and second set of parameter measurements for solving the forward solution;

obtaining a third set of parameter measurements from the at least one additional sensor location along the outside surface of the component;
comparing the continuous surface condition profile with the third set of parameter measurements at locations of the continuous surface condition profile corresponding to the at least one additional sensor location, to thereby determine an accuracy of the continuous surface condition profile;
identifying from the continuous surface condition profile the condition within the component; and
controlling the operation of the component in response to the condition identified from the continuous surface condition profile.

2. The method of claim 1 , further comprising displaying, on a display device, a representation of the continuous surface condition profile.

3. The method of claim 1 , further comprising:
providing an alert when at least a portion of the continuous surface condition profile exceeds a predetermined threshold of the condition.

4. A method for detecting a condition within a component of a process plant, comprising:
providing an array of a plurality of sensors on an outside surface of the component, wherein providing the array of the plurality of sensors comprises:
locating a first plurality of sensors of the array of the plurality of sensors at a first plurality of sensor locations along a first dimension of the outside surface of the component; and
locating a second plurality of sensors of the array of the plurality of sensors at a second plurality of sensor locations along a second dimension of the outside surface of the component;

obtaining a first set of parameter measurements from the first plurality of sensors at the first plurality of sensor locations along the first dimension of the outside surface of the component;
obtaining at least a second set of parameter measurements from the second plurality of sensors at the second plurality of sensor locations along the second dimension of the outside surface of the component;
processing the first set of parameter measurements and the second set of parameter measurements, using a computer, to develop a continuous surface condition profile of the component using a predetermined model, wherein the array of the plurality of sensors is operatively connected to the computer, wherein processing the first set of parameter measurements and the second set of parameter measurements includes predetermining how many coefficients to include in the set of coefficients corresponding to the set of basis functions, wherein predetermining how many coefficients includes:
obtaining a first calibration set of parameter measurements from at least one of the first plurality of sensor locations or the second plurality of sensor locations; and
applying a transform to the first calibration set to identify a set of coefficients corresponding to spatial frequencies along the first dimension with energy above a predetermined threshold, the model comprising:
a forward solution to an equation describing a condition within the component, the forward solution being linearly separable in at least two dimensions, the at least two dimensions corresponding to the first and second dimensions of the outer surface of the component, and
an inverse solution to the equation to provide a set of coefficients, corresponding to a set of basis functions of the forward solution, based on the first and second set of parameter measurements for solving the forward solution; and

identifying from the continuous surface condition profile the condition within the component.

5. The method of claim 4 , wherein the component includes a cylindrical vessel having a central axis, an inner wall diameter and an outer surface diameter; and wherein the first dimension is parallel to the central axis and disposed at the outer surface diameter, and the second dimension is disposed along a circumference at the outer surface diameter of the cylindrical vessel.

6. The method of claim 4 , wherein obtaining a first set of parameter measurements consists of obtaining the first set of parameter measurements from the first plurality of sensor locations along the first dimension only; and wherein obtaining at least a second set of parameter measurements consists of obtaining the second set of parameter measurements from the second plurality of sensor locations along the second dimension only.

7. The method of claim 4 , wherein the first set of parameter measurements and the second set of parameter measurements comprise temperature measurements.

8. The method of claim 4 , wherein the condition with the component of a process plant includes a temperature of a wall of the component.

9. The method of claim 4 , wherein the predetermined model is based a heat diffusion equation.

10. The method of claim 4 , wherein the forward solution is a linear combination of basis functions multiplied by the set of coefficients.

11. The method of claim 4 , wherein the set of coefficients is generated based on a predetermined relationship between the first and second set of parameter measurements.

12. The method of claim 4 , wherein applying a transform to the first calibration set to identify a set of coefficients corresponding to spatial frequencies along the first dimension with energy above a predetermined threshold includes applying a discrete cosine transform to the first calibration set to identify a set of coefficients corresponding to spatial frequencies along the first dimension with energy above a predetermined threshold, wherein predetermining how many coefficients further includes:
obtaining at least a second calibration set of parameter mea

Assignees

• Song Limin
• Kumaran Krishnan
• Sengupta Bhaskar
• Wang Jingbo
• Exxonmobil Res & Eng Co

Inventors

• Song Limin
• Kumaran Krishnan
• Sengupta Bhaskar
• Wang Jingbo

Classifications

• G05B23/0254 Primary

  based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks · CPC title

• G01B21/20 Primary

  for measuring contours or curvatures, e.g. determining profile · CPC title