Soft Sensor for Crude Stabilization H2S and RVP Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The petroleum industry faces challenges in achieving effective quality control and energy optimization in crude stabilization due to the difficulty in obtaining real-time measurements of critical parameters like H2S and RVP, primarily because of the limitations and inaccuracies associated with manual sampling and laboratory analysis.
Innovation Solution
A computer-implemented method is used to generate and update a soft sensor for crude stabilization, employing a hybrid approach combining model-driven and data-driven techniques to develop a model-free multivariable nonlinear soft sensor, which computes real-time H2S and RVP values from easily measurable process variables, thereby enabling automatic control of the stabilization process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual sampling and laboratory analysis are used to measure critical parameters, then measurement equipment and infrastructure requirements are reduced, but measurement precision and real-time monitoring capability deteriorate
Solution Approach 1:
The patent replaces manual sampling and laboratory analysis (mechanical/physical measurement systems) with a soft sensor system that uses mathematical models and computational algorithms to estimate critical parameters. This substitution enables real-time monitoring with high measurement precision while reducing the need for complex physical measurement infrastructure.
Solution Approach 2:
The patent introduces soft sensors as an intermediary system that bridges the gap between easily measurable process variables and difficult-to-measure critical parameters. The soft sensor uses mathematical models to compute H2S concentration and RVP values from readily available process data, providing accurate measurements without requiring direct physical sensors for these parameters.
2Reliability
If real-time measurements of critical parameters are obtained, then quality control and energy optimization are improved, but measurement and detection difficulty increases
Solution Approach 1:
The patent replaces difficult physical measurement systems with computational models. The soft sensor system uses mathematical relationships to infer critical parameters (H2S, RVP) from easily measurable variables, achieving reliable quality control without the technical challenges of direct real-time measurement.
Solution Approach 2:
The patent creates virtual copies of critical parameter measurements through soft sensors. Instead of directly measuring H2S concentration and RVP in real-time, the system computes surrogate values based on mathematical models and process data, providing reliable quality control information without the measurement difficulties associated with direct sensing.
3Productivity
If soft sensor is used for automatic control, then productivity and automation level are improved, but device complexity increases
Solution Approach 1:
The patent implements a self-service control system where the soft sensor automatically computes critical parameters and feeds them to the control system without requiring manual intervention. The system uses readily available process data to generate real-time estimates of H2S and RVP, enabling automatic quality control and optimization while improving productivity.
Solution Approach 2:
The soft sensor acts as an intermediary computational layer between process measurements and control decisions. It automatically transforms easily measurable process variables into reliable estimates of critical parameters, enabling automated control systems to make informed decisions without requiring complex direct measurement infrastructure.
Data Source
AI summary
A global theoretical graphical representation of a soft sensor is generated based on a cursory model, where the soft sensor is used to control crude stabilization. A plurality of local real-life graphical representations are generated for the soft sensor, each of the plurality of local real-life graphical representations corresponding to a respective local regime. A global real-life graphical representation is generated for the soft sensor by combining the plurality of local real-life graphical representations. A set of numerical values for the soft sensor are generated based on the global real-life graphical representation. The soft sensor is updated based on lab results and a crude stabilization operation is controlled using the soft sensor.