This Thesis deals with the problem of the quality control and monitoring and the related energy consuming aspects in industrial multicomponent distillation columns. These have important relevance especially in case of units subjected to important changes in operating conditions which manifests in fluctuation of the product concentration. Due to the low reliability of the seldom available and delayed analytic composition measurements, they cannot be employed for an efficient online monitoring of the separation. Instead, this Thesis proposes the use of composition observers with passive innovation mechanism and driven by temperature measurements. The passive structure allows to limit the number of ODEs to be online integrated, which is a central issue to obtain a product suitable for the industrial implementation. The design of this passive structure involves the choice of number and location of the temperature sensors, the choice of the innovated component and the set of modeled components in the (possibly simplified) estimation model. This issue is addressed with a simple and systematic methodology that employs steady state information about the per-component temperature gradient of the column and detectability measures and conditions. Differently from the majority of the works in literature, the proposed methodology allows also to assess the feasibility of obtaining good estimation performance using the available temperature sensors in real columns. Even the analytic composition measurements cannot be employed for quality control in feedback control loops mainly due to delay problems. Instead, temperature feedback controllers are usually used. But even if the temperature sensor is well located (and in this Thesis a criterion for sensor location is suggested based on an extension of the slope criterion to the multicomponent case), when the operating conditions undergo changes, the indirect regulation of the product quality cannot be achieved, with alternating periods of under and over-purification that also adversely affect the energy requirement for the column functioning. Thus, in this Thesis an application oriented solution is proposed consisting in adding a feedforward temperature setpoint compensation that guarantee a close quality regulation, including a more efficient management of the energy, under the changing operating modes of the column. The methodologies and techniques proposed and described in this Thesis are tested with an industrial multicomponent IC4-NC4 splitter located at the Sarlux Refinery (Sarroch, Italy), which, motivated by the good results, has already implemented the composition estimator and is going to implement the feedforward setpoint compensator.
Quality regulation and energy saving through control and monitoring techniques for industrial multicomponent distillation columns
PORRU, MARCELLA
2015-03-26
Abstract
This Thesis deals with the problem of the quality control and monitoring and the related energy consuming aspects in industrial multicomponent distillation columns. These have important relevance especially in case of units subjected to important changes in operating conditions which manifests in fluctuation of the product concentration. Due to the low reliability of the seldom available and delayed analytic composition measurements, they cannot be employed for an efficient online monitoring of the separation. Instead, this Thesis proposes the use of composition observers with passive innovation mechanism and driven by temperature measurements. The passive structure allows to limit the number of ODEs to be online integrated, which is a central issue to obtain a product suitable for the industrial implementation. The design of this passive structure involves the choice of number and location of the temperature sensors, the choice of the innovated component and the set of modeled components in the (possibly simplified) estimation model. This issue is addressed with a simple and systematic methodology that employs steady state information about the per-component temperature gradient of the column and detectability measures and conditions. Differently from the majority of the works in literature, the proposed methodology allows also to assess the feasibility of obtaining good estimation performance using the available temperature sensors in real columns. Even the analytic composition measurements cannot be employed for quality control in feedback control loops mainly due to delay problems. Instead, temperature feedback controllers are usually used. But even if the temperature sensor is well located (and in this Thesis a criterion for sensor location is suggested based on an extension of the slope criterion to the multicomponent case), when the operating conditions undergo changes, the indirect regulation of the product quality cannot be achieved, with alternating periods of under and over-purification that also adversely affect the energy requirement for the column functioning. Thus, in this Thesis an application oriented solution is proposed consisting in adding a feedforward temperature setpoint compensation that guarantee a close quality regulation, including a more efficient management of the energy, under the changing operating modes of the column. The methodologies and techniques proposed and described in this Thesis are tested with an industrial multicomponent IC4-NC4 splitter located at the Sarlux Refinery (Sarroch, Italy), which, motivated by the good results, has already implemented the composition estimator and is going to implement the feedforward setpoint compensator.File | Dimensione | Formato | |
---|---|---|---|
PhD_Thesis_Porru.pdf
accesso aperto
Tipologia:
Tesi di dottorato
Dimensione
11.92 MB
Formato
Adobe PDF
|
11.92 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.