Vasiliki Kazantzi

Optimal synthesis and scheduling of hybrid dynamic/steady-state property integration networks

Abstract This paper introduces a new problem in the emerging area of property integration. The problem involves synthesizing a network of property-modifying units and scheduling operating schemes. Data and constraints for the process sources, sinks, and interceptors are given in terms of properties. Sources are process streams to be allocated. Sinks are process units that can accept the sources. Interceptors are new units to be added to the process to intercept the sources and adjust their properties to meet the sink requirements. The interceptors may be steady-state or dynamic units whose performance is expressed in terms of input–output relations for designated properties. In addition to flowrate bounds, the constraints on acceptable feeds to the sinks are described as lower and upper bounds on designated properties. A conceptual framework is first developed to serve as the basis for a mathematical program coordinating the synthesis of the interception and allocation networks as well as the operational scheduling of the system. A source–interception–sink representation is developed to embed structural configurations of interest. The time domain is decomposed into repeating episodes. Within each episode, the interceptors are allowed to modify properties of sources, remain idle, or undergo regeneration. The regeneration process restores the interceptor to its initial condition. A mixed-integer nonlinear programming formulation is developed to minimize the total annualized cost of the system, synthesize the network, and determine its optimal operating schedule. A case study is solved to demonstrate the new problem and the devised solution algorithm.

Managing uncertainties in a safety-constrained process system for solvent selection and usage: an optimization approach with technical, economic, and risk factors

This paper addresses the problem of managing uncertainties in a safety-constrained process system for economic performance enhancement. Within such a context, a typical solvent selection problem involves a number of different solvents with nominal property values that are utilized in various process units and requires the minimization of the total operating cost while satisfying certain technical performance criteria and process safety constraints. Practically, in any process system, property values of streams are not exact; they are usually functions of operating variables and market conditions that change over time inevitably introducing irreducible uncertainties in system performance. A key aim of the present study is to systematically explore the effect of volatility in solvent prices on the economic performance of the process. Appropriate sensitivity analysis and Monte Carlo simulation work have been carried out to assist the decision maker in taking into account the continuously changing market conditions, while identifying operationally safe feasibility regions for solvents with different risk characteristics in the underlying optimization problem. The aforementioned uncertain inputs are shown to cause shifts of the associated Pareto front of optimal solutions toward feasibility regions that can be characterized in a more realistic manner. Finally, an illustrative case study that uses the permissible exposure limit as a risk factor is considered to evaluate the proposed method.

Property clustering and group contribution for process and molecular design

Abstract In this work, property clustering techniques and group contribution methods are combined to enable simultaneous consideration of process performance requirements and molecular property constraints. Using this methodology, the process design problem is solved to identify the property targets corresponding to the desired process performance. A significant advantage of the developed methodology is that for problems that can be satisfactorily described by three properties, the process and molecular design problems can be simultaneously solved visually, irrespective of how many molecular fragments are included in the search space. On the ternary cluster diagram, the target properties are represented as individual points if given as discrete values or as a region if given as intervals. The structure and identity of candidate components is then identified by combining or “mixing” molecular fragments until the resulting properties match the targets.

Property-based integration for sustainable development

Abstract This paper provides a systematic approach for optimal resource allocation, unit manipulation, and waste reduction using integrated component-less design. In particular, we identify a new procedure for determining optimal modifications in the design and operating variables of the process so as to optimize the allocation of process resources and minimize waste discharge. Interval arithmetic tools are used to derive rigorous bounds on the process performance, when all allowable changes in design and operating variables are considered. These bounds are mapped into a “trust region of clusters”, which represents the feasible search domain. In addition, material substitution strategies are considered for optimizing both the process and the fresh properties. A case study is also presented to illustrate the applicability of the proposed approach.

Optimal Scheduling of Biodiesel Plants through Property-based Integration with Oil Refineries

Abstract This paper addresses the design and scheduling problem of biodiesel plants in conjunction with typical oil refineries via blending of biodiesel and petro-diesel. The feedstocks are often seasonal and their availability and cost usually vary with time. A multi-period scheduling framework is formulated as an optimization problem to determine the optimal feedstock utilization and blending of biodiesel with petro-diesel using a property-integration framework. A case study is solved to illustrate the applicability of the devised approach.

A process design approach to manage the uncertainty of industrial flaring during abnormal operations

Abstract Flare management challenges are related to the flaring uncertainty during abnormal situations. In this work, a multi-objective optimization framework is upgraded with multi-period optimization and Monte Carlo simulation to incorporate the risk associated with uncertain flare events. An ethylene plant is used to present the developed framework. Using the ethylene process historical flaring data, Monte Carlo simulation generates probabilistic values for flaring events and event duration. Here, cogeneration unit (COGEN) is considered as the flare reduction alternative. The results of the formulations are presented as a set of Pareto fronts providing insights into the competing techno-economic and environmental objectives. Sensitivity analysis on the factors for the case suggests that some factors such as CO2 tax savings are severely affected by minor variations in flaring profiles, whereas others such as the fixed and operating costs are less sensitive. Hence, using this approach, the decision maker gains techno-economic-environmental insights regarding the flare reduction alternative (COGEN).

A pinch analysis approach to project selection problem

Project selection is a very important decision that every firm has to take; in fact, this decision plays a major role in the prosperity of the firm. Pinch analysis, which was initially developed to conserve energy and improve energy efficiency in industrial process, is now being extended to non-conventional areas. In this paper, pinch analysis is applied to select multiple independent projects from a large pool of viable projects, subject to budget constraints. The underlying mathematical optimization problem is discussed and a graphical approach to obtain optimal insightful solutions is presented. Applicability of the proposed methodology is demonstrated through an illustrative example of energy conservation in the Indian cement industry.

Managing Uncertain Industrial Flares during Abnormal Process Operations using an Integrated Optimization and Monte Carlo Simulation Approach

Abstract In this work, an integrated optimization framework with Monte Carlo (MC) simulation techniques is suggested for the systematic synthesis of energy alternative tools, such as cogeneration (COGEN) systems, which can effectively manage industrial flares with uncertain occurrence patterns. The optimization model that was previously developed is now extended to incorporate the risk associated with the uncertain nature of the flaring events that are probabilistically characterized based on empirically meaningful historical samples. The model aims at minimizing the total annualized cost including fixed and operating costs of the system, the value of by- and co-products (i.e., power, excess heat), and regulatory taxes/credits associated with Green House Gases (GHGs). A base case ethylene production plant is presented to illustrate the applicability of the proposed approach and highlight trade-offs between different performance objectives (economic, environmental and energy-related). The results show that some of the examined factors (i.e., CO2 tax savings) can be severely affected by small variations in flaring profiles, whereas others are only slightly affected by such variability (i.e., power vs. heat generation curves, fixed and operating costs). Therefore, the uncertain nature of flaring events may be of high importance in process performance and should be inevitably considered during abnormal situation management.