Nouri J. Samsatli

Object-oriented dynamic supply-chain modelling incorporated with production scheduling

Abstract In this paper we present a new modelling approach for realistic simulation of supply-chains. It is based on an object-oriented architecture, which enables flexible specification of the supply-chain configuration along with its operational decisions and policies. A model of a generic supply-chain node is developed to capture the features present in all supply-chain entities. The generic node models in detail activities such as inventory control, manufacturing processes and order handling. The supply-chain model is constructed by linking generic nodes and specifying the physical and business attributes of each supply-chain member. The generic-node model may also be linked to external software for greater accuracy, e.g., detailed production scheduling or optimisation. The model provides a fully dynamic simulation of the supply-chain and the effect of various uncertainties can be evaluated through Monte Carlo simulation and other, more efficient, sampling techniques (not described here). A case study is presented to illustrate the applicability of the model. The case study demonstrates how the effect of policy changes on the supply-chain performance under uncertainty can be evaluated before implementation.

A general spatio-temporal model of energy systems with a detailed account of transport and storage

Abstract This paper presents a general spatio-temporal model of energy systems comprising technologies for generation/conversion, transport and storage and infrastructures for transport. The model determines the optimal network structure (e.g. location and size of technologies and their interconnections through transport infrastructures) and its operation (e.g. rate of utilisation of technologies and transport flows) considering simultaneously the short-term dynamics and a long-term planning horizon. Here, we address one of the main challenges of solving a large scale MILP model: tractability. This issue is mainly caused by the need to include a wide range of time scales in the model: yearly (or decadal) intervals to include investment decisions; seasonal intervals to account for e.g. seasonal variations in demand and availability of resources; and hourly (or shorter) intervals to model the dynamics of storage technologies and to account for intermittency of renewable resources and demand. To exacerbate the problem, the spatial aspects also need to be fine enough to locate and size the technologies properly and to model the transport of resources, which depend on the location of demand and availability of resources. The model uses an efficient representation of time that exploits periodicity in system properties via a non-uniform hierarchical time discretisation. A decomposition method is also proposed wherein the large problem is broken down into 3 sub-problems that are then solved iteratively until the objective function is no longer improved. These methods significantly improve the computational efficiency without sacrificing temporal and spatial detail. The applicability of the model is illustrated using a case study in which the least-cost design and operation of a hydrogen network is determined such that the hourly transport demand of the different regions of an island is met by the intermittent and remotely located wind energy.

A flexible and generic approach to dynamic modelling of supply chains

In this paper, we present a new modelling approach for realistic supply chain simulation. The model provides an experimental environment for informed comparison between different supply chain policies. A basic simulation model for a generic node, from which a supply chain network can be built, has been developed using an object-oriented approach. This generic model allows the incorporation of the information and physical systems and decision-making policies used by each node. The object-oriented approach gives the flexibility in specifying the supply chain configuration and operation decisions, and policies. Stochastic simulations are achieved by applying Latin Supercube Sampling to the uncertain variables in descending order of importance, which reduces the number of simulations required. We also present a case study to show that the model is applicable to a real-life situation for dynamic stochastic studies.

An Optimization Based Design Procedure for Biochemical Processes: Part II: Detailed Scheduling

The first part of this paper presented an overall approach to optimization based design of biochemical processes. The procedure is divided into two stages. In the first stage (described in part I of this paper 1 ), the processing rates and conditions of the unit operations and equipment capacities are determined through dynamic optimization, while simple scheduling considerations are accounted for. The second stage, described here, deals with detailed scheduling and design adjustments, required for an accurate determination of the sequence and timing of the unit operations, using the processing rates from the first stage. A scheduling model is developed to take account of batch integrity. The design procedure is demonstrated using an example of a process common in the biochemical industries.

An Optimization Based Design Procedure for Biochemical Processes: Part I: Preliminary Design and Operation

A two stage optimization based design procedure for biochemical processes is presented. In the first stage, the processing rates and conditions of the unit operations and equipment capacities are determined through dynamic optimization, while simple scheduling considerations are accounted for. The second stage deals with detailed scheduling and design adjustments, required for an accurate determination of the sequence and timing of the unit operations, using the processing rates from the first stage. A scheduling model is developed to take account of batch integrity and is described in the second part of this paper. The design procedure is demonstrated using an example of a process common in the biochemical industries. The production of an intra-cellular enzyme requires a fermentation stage (or stages) followed by a number of primary separation stages, with high resolution steps for final purification. In this example the process is considered up to and including the primary separation stages only. The process consists of nine stages including four typical unit operations.

Robust optimal fermentation operating policies

Abstract The bioprocess industry is starting to face new commercial pressures leading to a greater emphasis on improving manufacturing, and delivering relatively low-volume, high value-added products at costs acceptable to health-care providers. A major challenge in this industry is the rapid development of new integrated processes for optimal large scale production. Accurate predictive models of biochemical unit and process operations are required if optimal design is to be achieved without extensive pilot plant trials (Richardson and Peacock, 1994). In the literature there exist many models of unit operations used in bioprocesses. In principle, these should be particularly useful in determining the optimal operation of the process, but a wide range of uncertainty associated with each model presents a number of problems. The main obstacle is that using an operating policy which has been optimised for the nominal model parameter values in an uncertain system can lead to dramatic changes in the performance of the process. It is preferable to use a policy in which these changes are kept to a minimum, while maintaining a good performance in the nominal case. Therefore a comprehensive design of the process must account explicitly for these uncertainties. This paper concentrates on the optimisation of fermenter operating policies. Here, we consider an appropriate definition of “robustness” for biochemical processes, and go on to describe one means of ensuring such robustness during the optimisation of the dynamic operation of a formenter. We use an illustrative example to contrast, using stochastic simulation, the proposed robust approach with a deterministic design based on nominal parameter values.

Optimal integrated design of biochemical processes

Abstract Integrated optimal design problems for biochemical processes which consider simultaneously all aspects necessary for a rigourous design typically result in mathematical formulations which would almost certainly be intractable. We overcome this problem by decomposing the design problem into two stages. The first stage is a dynamic optimisation where equipment sizes, operating conditions, processing rates and times, etc. are determined assuming a finite intermediate storage policy and a fixed series of unit operations. The majority of the design parameters are then fixed for the second stage, which determines the optimum operating schedule and intermediate storage capacity for the plant. Further refinements to the equipment design may also be made. To enforce batch integrity and take advantage of a number of simplifications not present in most scheduling packages, a new scheduling formation is developed for the second stage. This paper briefly describes the two stage design procedure, outlining the key features of both subproblems. An example problem is shown with a summary of the results.

Optimisation of policy parameters in supply chain applications

In this paper, three approaches for finding optimal parameters for supply chain systems are described and evaluated. The first approach is based on rigorous mixed integer linear programming concepts. Although using this approach guarantees finding the optimal solution for the policy parameters, the drawback of applying such a model to supply chain networks is the excessive computational requirements. The second approach is to combine a detailed stochastic simulation model with a simulated annealing algorithm. While optimality is no longer guaranteed with this approach, far larger problems should be tractable. Finally, a gradient-based approach is suggested, in which a higher-level optimiser utilises the results from the same detailed stochastic simulation model to obtain the optimal parameters. Results obtained when introducing small deviations to the policy parameters are used to estimate the gradients, which are used by the higher-level model. Two case studies are considered: a simple reorder-point warehouse example and an industrial-scale multi-site production distribution problem. The first example is used to compare the effectiveness of the simulated annealing and gradient-based approaches with the rigorous mixed integer linear programming (MILP) formulation. In the industrial example, only simulated annealing and simple gradient-based methods are applied, as the MILP model becomes intractable. Here, dynamic safety stock levels and order quantities are primarily investigated.

Optimal design of continuous sterilisation networks

Abstract The design of continuous sterilisation networks involves the selection of design variables, such as the sterilisation temperature, heat exchanger sizes, etc., to satisfy the prime objective of destroying sufficient micro-organisms in the fermentation medium. Secondary objectives include minimising nutrient degradation and minimising annual costs. Common design practices for continuous sterilisation networks rely on simple heuristics which typically do not account for any suspended solids and do not guarantee cost optimality. In this paper, an optimal design procedure is formulated which guarantees sterility both in the liquid and solid phases, whilst avoiding excessive degradation of any labile nutrients present, and takes account of capital and operating costs. The results of four optimal design problems are presented and compared with a typical heuristic design.

Fower-to-hydrogen and hydrogen-to-X pathways: Opportunities for next generation energy systems

Energy systems are evolving rapidly around the world, driven mainly by CO2-e reduction targets. This has led to opportunities for integrated low carbon electricity-and-fuel systems founded on large scale “Power-to-Hydrogen, Hydrogen-to-X” (PtH-HtX). Power-to-Hydrogen (PtH) refers to large scale electrolysis. Hydrogen-to-X (HtX) refers to a range of high value products and services. If these pathways start with low-carbon electricity, then the fuel consumed at the downstream end also low-carbon. Use of intermittently low valued power lowers all production costs. This paper specifically identifies the main pathways and interconnections in a way that overcomes the ambiguities inherent in the term “Power-to-Gas”. In turn, this provides solid and easier to understand foundations for building legal and regulatory frameworks for new business opportunities along the lengths of the numerous pathways from supply to consumption.