Abstract
Abstract The naphtha reforming process converts low-octane gasoline blending components to high-octane components for use in high-performance gasoline fuels. The reformer also has an important function as the producer of hydrogen to the refinery hydrotreaters. A process model based on a unit model structure, is used for estimation of the process condition using data reconciliation. Measurements are classified as redundant or non-redundant and the model variables are classified as observable, barely observable or unobservable. The computed uncertainty of the measured and unmeasured variables shows that even if a variable is observable it may have a very large uncertainty and may thereby be practically unobservable. The process condition at 21 data points, sampled from two years of operation, was reconciled and used to optimize the process operation. There are large seasonal variations in the reformer product price and two operational cases are studied. In case 1, the product price is high and throughput is maximized with respect to process and product quality constraints. In case 2, the product price is low and the throughput is minimized with respect to a low constraint on the hydrogen production. Based on the characteristics of the optimal operation, a “self optimizing” control structure is suggested for each of the two operational cases.
Highlights
In all three cases above, good agreement with plant data was reported. These models are used for simulation and design purposes except in Taskar and Riggs (1997) where optimal operation during a catalyst cycle, is considered
The Combined Gaussian distribution is described by the following objective function where J(ym, z) is the objective function for data reconciliation, f (z) = 0 represents the process model, Arz = br is used to specify known values and zr min ≤ z ≤ zr max physical constraints
Mary, data reconciliation is based on the Combined Gaussian objective (4), whereas the Gaussian objective (2) is used for analysis of the uncertainty in the estimate
Summary
It is referred to the references and the thesis of Lid (2007). The Combined Gaussian distribution is described by the following objective function where J(ym, z) is the objective function for data reconciliation, f (z) = 0 represents the process model, Arz = br is used to specify known values and zr min ≤ z ≤ zr max physical constraints. The ny measured values are collected in the measurement vector ym. If the measurement error is normally distributed N(μ, σ) which has two adjustable parameters, p and b. Lid and Skogestad, “Data reconciliation and optimal operation of a catalytic naphtha reformer”. Mary, data reconciliation is based on the Combined Gaussian objective (4), whereas the Gaussian objective (2) is used for analysis of the uncertainty in the estimate
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Modeling, Identification and Control: A Norwegian Research Bulletin
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
