Crude Distillation Dynamic Optimization During Feedstock Transitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current crude distillation unit (CDU) optimization systems struggle with dynamic optimization due to computational demands, reliance on steady-state assumptions, and inability to handle rapid changes in crude composition, leading to sub-optimal operations and production of off-spec products.
Innovation Solution
A system and method utilizing real-time plant data and non-linear time-dependent functions, combined with autoregressive exogenous models and neural networks, to dynamically optimize CDU operations, predict yields, and adjust parameters for optimal product quality and profit, without requiring crude assay or laboratory analyzers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If steady-state real-time optimization (RTO) tools are used to maximize profit, then optimization capability is provided, but the system cannot handle dynamic changes in crude composition and requires steady-state assumptions
Solution Approach 1:
The patent transitions from steady-state RTO to dynamic optimization that explicitly models time-varying crude composition changes. The system uses dynamic process models and real-time crude characterization to adapt optimization decisions during crude transitions, eliminating the need for steady-state assumptions and enabling continuous optimization during dynamic operations.
Solution Approach 2:
The system implements real-time crude characterization by measuring and tracking composition parameters such as API gravity, sulfur content, and nitrogen content. These parameter changes are fed into the optimization model to dynamically adjust operating conditions, allowing the system to adapt to varying crude qualities without requiring steady-state conditions.
2Productivity
If advanced process control (APC) is used to optimize operations, then control capability is provided, but performance significantly reduces during crude switch operations
Solution Approach 1:
The system performs preliminary crude characterization and prediction before crude switch operations begin. By measuring composition parameters in real-time and predicting upcoming changes, the optimization system can proactively adjust operating conditions in anticipation of crude transitions, maintaining high performance throughout the transition process rather than reacting after performance degrades.
3Measurement precision
If first principles-based optimization models are used, then accurate optimization is provided, but computational time increases and real-time optimization becomes difficult
Solution Approach 1:
The patent extracts and measures only the critical composition parameters (API gravity, sulfur, nitrogen) that most significantly impact optimization decisions. By focusing on these key parameters rather than full compositional analysis, the system maintains high optimization accuracy while dramatically reducing computational burden and enabling real-time implementation.
Solution Approach 2:
The system applies different levels of measurement and modeling detail to different parts of the optimization problem. Real-time composition measurements are applied where most critical (feed composition), while other parameters use standardized models or less frequent measurements, optimizing the balance between accuracy and computational efficiency.
4Measurement precision
If online measurement for crude True Boiling Point (TBP) is implemented, then accurate feedstock characterization is achieved, but system complexity and cost increase
Solution Approach 1:
The system uses a multi-functional measurement approach where a single online analyzer measures multiple critical parameters (API gravity, sulfur content, nitrogen content) simultaneously. This universal measurement device provides comprehensive feedstock characterization without the complexity of multiple specialized instruments or full TBP distillation curves.
Data Source
AI summary
The present invention relates to a system and method for real-time dynamic optimization of crude distillation unit (CDU). The present invention makes use of a combination of CDU and plant data such as crude density, flow, temperature profile and pressure profile of column, other operating conditions, and column heat balance to calculate product yields and product properties. The present invention makes use of real-time CDU or plant measurements, and its non-linear time-dependent functions to represent change in feedstock (product) characteristics, and predict the optimal value of CDU input data to optimize the CDU, and product yields and product properties. This provides a robust optimized solution that significantly reduces the computational time, accounts for dynamic change in crude composition and the unsteady nature of the operation during crude transitions.