A new Lagrangian decomposition approach applied to the integration of refinery planning and crude-oil scheduling

The aim of this paper is to introduce a methodology to solve a large-scale mixed-integer nonlinear program (MINLP) integrating the two main optimization problems appearing in the oil refining industry: refinery planning and crude-oil operations scheduling. The proposed approach consists of using Lagrangian decomposition to efficiently integrate both problems. The main advantage of this technique is to solve each problem separately. A new hybrid dual problem is introduced to update the Lagrange multipliers. It uses the classical concepts of cutting planes, subgradient, and boxstep. The proposed approach is compared to a basic sequential approach and to standard MINLP solvers. The results obtained on a case study and a larger refinery problem show that the new Lagrangian decomposition algorithm is more robust than the other approaches and produces better solutions in reasonable times.     

A novel two-stage Lagrangian decomposition approach for refinery production scheduling with operational transitions in mode switching☆

To address large scale industrial processes, a novel Lagrangian scheme is proposed to decompose a refinery scheduling problem with operational transitions in mode switching into a production subproblem and a blending and delivery subproblem. To accelerate the convergence of Lagrange multipliers, some auxiliary constraints are added in the blending and delivery subproblem. A speed-up scheme is presented to increase the efficiency for solving the production subproblem. An initialization scheme of Lagrange multipliers and a heuristic algorithm to find feasible solutions are designed. Computational results on three cases with different lengths of time hori-zons and different numbers of orders show that the proposed Lagrangian scheme is effective and efficient.     

A new continuous-time formulation for scheduling crude oil operations

In today's competitive business climate characterized by uncertain oil markets, responding effectively and speedily to market forces, while maintaining reliable operations, is crucial to a refinery's bottom line. Optimal crude oil scheduling enables cost reduction by using cheaper crudes intelligently, minimizing crude changeovers, and avoiding ship demurrage. So far, only discrete-time formulations have stood up to the challenge of this important, nonlinear problem. A continuous-time formulation would portend numerous advantages, however, existing work in this area has just begun to scratch the surface. In this paper, we present the first complete continuous-time mixed integer linear programming (MILP) formulation for the short-term scheduling of operations in a refinery that receives crude from very large crude carriers via a high-volume single buoy mooring pipeline. This novel formulation accounts for real-world operational practices. We use an iterative algorithm to eliminate the crude composition discrepancy that has proven to be the Achilles heel for existing formulations. While it does not guarantee global optimality, the algorithm needs only MILP solutions and obtains excellent maximum-profit schedules for industrial problems with up to 7 days of scheduling horizon. We also report the first comparison of discrete- vs. continuous-time formulations for this complex problem.     

A decision tree based decomposition method for oil refinery scheduling

Refinery scheduling attracts increasing concerns in both academic and industrial communities in recent years. However, due to the complexity of refinery processes, little has been reported for success use in real world refineries. In academic studies, refinery scheduling is usually treated as an integrated, large-scale optimization problem, though such complex optimization problems are extremely difficult to solve. In this paper, we proposed a way to exploit the prior knowledge existing in refineries, and developed a decision making system to guide the scheduling process. For a real world fuel oil oriented refinery, ten adjusting process scales are predetermined. A C4.5 decision tree works based on the finished oil demand plan to classify the corresponding category (i.e. adjusting scale). Then, a specific sub-scheduling problem with respect to the determined adjusting scale is solved. The proposed strategy is demonstrated with a scheduling case originated from a real world refinery.     

New approach for scheduling crude oil operations

Scheduling of crude oil operations is crucial to petroleum refining, which includes determining the times and sequences of crude oil unloading, blending, and CDU feeding. In the last decades, many approaches have been proposed for solving this problem, but they either suffered from composition discrepancy [Lee et al. 1996. Mixed-integer linear programming model for refinery short-term scheduling of crude oil unloading with inventory management. Industrial and Engineering Chemistry Research 35, 1630-1641; Jia et al., 2003. Refinery short-term scheduling using continuous time formulation: crude-oil operations. Industrial and Engineering Chemistry Research 42, 3085-3097; Jia and Ierapetritou, 2004. Efficient short-term scheduling of refinery operations based on a continuous time formulation. Computer and Chemical Engineering 28, 1001-1019] or led to infeasible solutions for some cases [Reddy et al., 2004a. Novel solution approach for optimizing crude oil operations. A.I.Ch.E. Journal 50(6), 1177-1197; 2004b. A new continuous-time formulation for scheduling crude oil operations. Chemical Engineering Science 59, 1325-1341]. In this paper, coastal and marine-access refineries with simplified workflow are considered. Unlike existing approaches, the new approach can avoid composition discrepancy without using iterative algorithm and find better solution effectively. In this approach, a new mixed integer non-linear programming (MINLP) formulation is set up for crude oil scheduling firstly, and then some heuristic rules collected from expert experience are proposed to linearize bilinear terms and prefix some binary variables in the MINLP model. Thus, crude oil scheduling can be expressed as a complete mixed integer linear programming (MILP) model with fewer binary variables. To illustrate the advantage of the new approach, four typical examples are solved with three models. The new model is compared with the most effective models (RKS(a) and RKS(b) models) presented by Reddy et al. [2004a. Novel solution approach for optimizing crude oil operations. A.I.Ch.E. Journal 50(6), 1177-1197; 2004b. A new continuous-time formulation for scheduling crude oil operations. Chemical Engineering Science 59, 1325-1341], which proves that the new approach is valid and feasible in most small-size and medium-size problems.     

Supply chain planning and scheduling integration using Lagrangian decomposition in a knowledge management environment

The integration of planning and scheduling decisions in rigorous mathematical models usually results in large scale problems. In order to tackle the problem complexity, decomposition techniques based on duality and information flows between a master and a set of subproblems are widely applied. In this sense, ontologies improve information sharing and communication in enterprises and can even represent holistic mathematical models facilitating the use of analytic tools and providing higher flexibility for model building. In this work, we exploit this ontologies’ capability to address the optimal integration of planning and scheduling using a Lagrangian decomposition approach. Scheduling/planning sub-problems are created for each facility/supply chain entity and their dual solution information is shared by means of the ontological framework. Two case studies based on a STN representation of supply chain planning and scheduling models are presented to emphasize the advantages and limitations of the proposed approach.     

An adjustable robust optimization approach to scheduling of continuous industrial processes providing interruptible load

To ensure the stability of the power grid, backup capacities are called upon when electricity supply does not meet demand due to unexpected changes in the grid. As part of the demand response efforts in recent years, large electricity consumers are encouraged by financial incentives to provide such operating reserve in the form of load reduction capacities (interruptible load). However, a major challenge lies in the uncertainty that one does not know in advance when load reduction will be requested. In this work, we develop a scheduling model for continuous industrial processes providing interruptible load. An adjustable robust optimization approach, which incorporates recourse decisions using linear decision rules, is applied to model the uncertainty. The proposed model is applied to an illustrative example as well as a real-world air separation case. The results show the benefits from selling interruptible load and the value of considering recourse in the decision-making.     

Time representations and mathematical models for process scheduling problems

During the last 15 years, many mathematical models have been developed in order to solve process operation scheduling problems, using discrete or continuous-time representations. In this paper, we present a unified representation and modeling approach for process scheduling problems. Four different time representations are presented with corresponding strengthened formulations that rely on exploiting the non-overlapping graph structure of these problems through maximum cliques and bicliques. These formulations are compared, and applied to single-stage and multi-stage batch scheduling problems, as well as crude-oil operations scheduling problems. We introduce three solution methods that can be used to achieve global optimality or obtain near-optimal solutions depending on the stopping criterion used. Computational results show that the multi-operation sequencing time representation is superior to the others as it allows efficient symmetry-breaking and requires fewer priority-slots, thus leading to smaller model sizes.     

A novel network-based continuous-time representation for process scheduling: Part II. General framework

In the first part of this series of papers we presented a new network-based continuous-time representation for the short-term scheduling of batch processes, which overcomes numerous shortcomings of existing approaches. In this second part, we discuss how this representation can be extended to address aspects such as: (i) preventive maintenance activities on unary resources (e.g., processing and storage units) that were planned ahead of time; (ii) resource-constrained changeover activities on processing and shared storage units; (iii) non-instantaneous resource-constrained material transfer activities; (iv) intermediate deliveries of raw materials and shipments of finished products at predefined times; and (v) scenarios where part of the schedule is fixed because it has been programmed in the previous scheduling horizon. The proposed integrated framework can be used to address a wide variety of process scheduling problems, many of which are intractable with existing tools.     

An augmented Lagrangian interior-point approach for large-scale NLP problems on graphics processing units

The demand for fast solution of nonlinear optimization problems, coupled with the emergence of new concurrent computing architectures, drives the need for parallel algorithms to solve challenging nonlinear programming (NLP) problems. In this paper, we propose an augmented Lagrangian interior-point approach for general NLP problems that solves in parallel on a Graphics processing unit (GPU). The algorithm is iterative at three levels. The first level replaces the original problem by a sequence of bound-constrained optimization problems using an augmented Lagrangian method. Each of these bound-constrained problems is solved using a nonlinear interior-point method. Inside the interior-point method, the barrier sub-problems are solved using a variation of Newton's method, where the linear system is solved using a preconditioned conjugate gradient (PCG) method, which is implemented efficiently on a GPU in parallel. This algorithm shows an order of magnitude speedup on several test problems from the COPS test set.     

Rigorous approach to scheduling of sterile drug product manufacturing

Optimizing the scheduling of liquid drug product manufacturing is paramount for pharmaceutical companies in their increasingly competitive environment and requires the modelling of industry-specific constraints. Such constraints include: (i) changing sequence-dependent setup times; (ii) maintaining a sterile production environment (e.g., through sterile holding times); (iii) periods with limited or no plant activity (e.g., no workforce during weekends); and (iv) demand timing (i.e., delivery deadline and release date constraints). In this work, an immediate precedence model is formulated to optimize the scheduling of liquid drug product manufacturing, considering the industry-specific constraints. The primary objective is to minimize the production makespan.Four case studies comprising up to 38 batches from a real multi-product facility illustrate the performance of the rigorous optimization approach. The makespan could be reduced by up to 7.9% compared to expert schedules.     

A continuous-time formulation for refinery production scheduling problems involving operational transitions in mode switching

In the refinery scheduling, operational transitions in mode switching are of great significance to formulate dynamic nature of production and obtain efficient schedules. The discrete-time formulation meets two main challenges in modeling:discrete approximation of time and large size of mixed-integer linear problem (MILP). In this article, a continuous-time refinery scheduling model, which involves transitions of mode switching, is presented due to these challenges. To reduce the difficulty in solving large scale MILPs resulting from the sequencing constraints, the global event-based formulation is chosen. Both transition constraints and production transitions are introduced and the numbers of key variables and constraints in both of the discrete-time and continuous-time formulations are analyzed and compared. Three cases with different lengths of time horizons and different numbers of orders are studied to show the efficiency of the proposed model.     

Genetic algorithm for large-size multi-stage batch plant scheduling

This paper presents a heuristic approach based on genetic algorithm (GA) for solving large-size multi-stage multi-product scheduling problem (MMSP) in batch plant. The proposed approach is suitable for different scheduling objectives, such as total process time, total flow time, etc. In the algorithm, solutions to the problem are represented by chromosomes that will be evolved by GA. A chromosome consists of order sequences corresponding to the processing stages. These order sequences are then assigned to processing units according to assignment strategies such as forward or backward assignment, active scheduling technique or similar technique, and some heuristic rules. All these measures greatly reduce unnecessary search space and increase the search speed. In addition, a penalty method for handling the constraints in the problem, e.g., the forbidden changeovers, is adopted, which avoids the infeasibility during the GA search and further greatly increases the search speed.     

A discrete-time scheduling model for continuous power-intensive process networks with various power contracts

Increased volatility in electricity prices and new emerging demand side management opportunities call for efficient tools for the optimal operation of power-intensive processes. In this work, a general discrete-time model is proposed for the scheduling of power-intensive process networks with various power contracts. The proposed model consists of a network of processes represented by Convex Region Surrogate models that are incorporated in a mode-based scheduling formulation, for which a block contract model is considered that allows the modeling of a large variety of commonly used power contracts. The resulting mixed-integer linear programming model is applied to an illustrative example as well as to a real-world industrial test case. The results demonstrate the model's capability in representing the operational flexibility in a process network and different electricity pricing structures. Moreover, because of its computational efficiency, the model holds much promise for its use in a real industrial setting.     

A rule-based genetic algorithm for the scheduling of single-stage multi-product batch plants with parallel units

This paper presents a heuristic rule-based genetic algorithm (GA) for large-size single-stage multi-product scheduling problems (SMSP) in batch plants with parallel units. SMSP have been widely studied by the researchers. Most of them used mixed-integer linear programming (MILP) formulation to solve the problems. With the problem size increasing, the computational effort of MILP increases greatly. Therefore, it is very difficult for MILP to obtain acceptable solutions to large-size problems within reasonable time. To solve large-size problems, the preferred method in industry is the use of scheduling rules. However, due to the constraints in SMSP, the simple rule-based method may not guarantee the feasibility and quality of the solution. In this study, a random search based on heuristic rules was proposed first. Through exploring a set of random solutions, better feasible solutions can be achieved. To improve the quality of the random solutions, a genetic algorithm-based on heuristic rules has been proposed. The heuristic rules play a very important role in cutting down the solution space and reducing the search time. Through comparative study, the proposed method demonstrates promising performance in solving large-size SMSP.     

Capacity expansion planning through augmented Lagrangian optimization and scenario decomposition

Stochastic programming is a typical method for addressing the uncertainties in capacity expansion planning problem. However, the corresponding deterministic equivalent model is often intractable with considerable number of uncertainty scenarios especially for stochastic integer programming (SIP) based formulations. In this article, a hybrid solution framework consisting of augmented Lagrangian optimization and scenario decomposition algorithm is proposed to solve the SIP problem. The method divides the solution procedure into two phases, where traditional linearization based decomposition strategy and global optimization technique are applied to solve the relaxation problem successively. Using the proposed solution framework, a feasible solution of the original problem can be obtained after the first solution phase whereas the optimal solution is obtained after the second solution phase. The effectiveness of the proposed strategy is verified through a numerical example of two stage stochastic integer program and the capacity expansion planning examples. © 2011 American Institute of Chemical Engineers AIChE J, 2012     

Three-scale integrated optimization model of furnace simulation,cyclic scheduling,and supply chain of ethylene plants

In order to explore the potential of profit margin improvement, a novel three-scale integrated optimization model of furnace simulation, cyclic scheduling, and supply chain of ethylene plants is proposed and evaluated. A decoupling strategy is proposed for the solution of the three-scale model, which uses our previously proposed reactor scale model for operation optimization and then transfers the obtained results as a parameter table in the joint MILP optimization of plant-supply chain scale for cyclic scheduling. This optimization framework simplifies the fundamental mixed-integer nonlinear programming (MINLP) into several sub-models, and improves the interpretability and extendibility. In the evaluation of an industrial case, a profit increase at a percentage of 3.25% is attained in optimization compared to the practical operations. Further sensitivity analysis is carried out for strategy evolving study when price policy, supply chain, and production requirement parameters are varied. These results could provide useful suggestions for petrochemical enterprises on thermal cracking production.     

On deterministic online scheduling: Major considerations,paradoxes and remedies

Despite research in the area, the relationship between the (open-loop) optimization problem and the quality of the (closed-loop) implemented schedule is poorly understood. Accordingly, we first show that open-loop and closed-loop scheduling are two different problems, even in the deterministic case. Thereafter, we investigate attributes of the open-loop problem and the rescheduling algorithm that affect closed-loop schedule quality. We find that it is important to reschedule periodically even when there are no “trigger” events. We show that solving the open-loop problem suboptimally does not lead to poor closed-loop solutions; instead, suboptimal solutions are corrected through feedback. We also observe that there exist thresholds for rescheduling frequency and moving horizon length, operating outside of which leads to substantial performance deterioration. Fourth, we show that the design attributes work in conjunction, hence, studying them simultaneously is important. Finally, we explore objective function modifications and constraint addition as methods to improve performance.     

Continuous‐time modeling and global optimization approach for scheduling of crude oil operations

Scheduling of crude oil operations is a critical and complicated component of overall refinery operations, because crude oil costs account for about 80% of the refinery turnover. Moreover, blending with less expensive crudes can significantly increase profit margins. The mathematical modeling of blending different crudes in storage tanks results in many bilinear terms, which transforms the problem into a challenging, nonconvex, and mixed‐integer nonlinear programming (MINLP) optimization model. Two primary contributions have been made. First, the authors developed a novel unit‐specific event‐based continuous‐time MINLP formulation for this problem. Then they incorporated realistic operational features such as single buoy mooring (SBM), multiple jetties, multiparcel vessels, single‐parcel vessels, crude blending, brine settling, crude segregation, and multiple tanks feeding one crude distillation unit at one time and vice versa. In addition, 15 important volume‐based or weight‐based crude property indices are also considered. Second, they exploited recent advances in piecewise‐linear underestimation of bilinear terms within a branch‐and‐bound algorithm to globally optimize the MINLP problem. It is shown that the continuous‐time model results in substantially fewer bilinear terms. Several examples taken from the work of Li et al. are used to illustrate that (1) better solutions are obtained and (2) ε‐global optimality can be attained using the proposed branch‐and‐bound global optimization algorithm with piecewise‐linear underestimations of the bilinear terms. © 2011 American Institute of Chemical Engineers AIChE J, 2012