Alexandros Koulouris

Throughput analysis and debottlenecking of integrated batch chemical processes

Abstract In batch plants, as we attempt to increase annual throughput, we run into bottlenecks that are either equipment or resource (e.g. utilities, labor) related. Throughput bottlenecks are equipment or resources that limit the amount of product produced per batch. Scheduling bottlenecks are units or resources that limit the number of batches that can be processed per year or campaign. In this paper, we present a systematic simulation-based methodology for identifying bottlenecks in batch processing and a strategy for eliminating them. The decision-making process is based on a set of operation-related variables that are used to indicate the presence of bottlenecks. The implementation of the methodology is demonstrated with the use of a synthetic pharmaceutical example.

Simulation-based reactive scheduling in tomato processing plant with raw material uncertainty

Abstract Production scheduling in a real plant that processes tomatoes into various types of paste under raw material supply uncertainty is studied in this paper. Using simulation-based, finite-capacity scheduling software, a model of the production process is developed and used to generate feasible production plans that satisfy the tomato inventory constraint under an assumed supply profile. During actual production, real data on tomato supply are daily fed into the schedule and the feasibility of the plan is reassessed and, when necessary, updated to satisfy constraints imposed by the inventory and the limited shelf-life of the raw material. In this way, the production planner has a continuously updated view of the overall production plan and can make well-informed and timely decisions in order to meet the production objectives.

Learning to solve mass balance problems through a web-based simulation environment

Abstract Being able to solve mass and energy balance problems on any given process is a key skill required for students in chemical, or, in general, process engineering. Yet, many students have serious problems solving such problems because the necessary equations do not come in a ready-to-use form; instead, the general and abstract balance equations must be tailored by the student to the problem and the data at hand before they can be used. To solve mass balance problems, students must first recognize the input/output structure of the system(s) around which balance equations can be written, identify the components present and, from all possible balance equations, choose those that, based on the available data, will lead to results in the fastest and more efficient way. In a typical textbook problem, there is no predefined structure in the input data (as it exists, for example, in a process simulator) and the solution process is completely data-driven. We have developed a web-based environment that attempts to imitate the process of solving mass balances by hand. The webpage embeds a solution algorithm which identifies in a stepwise fashion the smallest subset of equations which, based on the available data, can yield results (even partial). A student will not do degrees of freedom analysis on the entire data set nor can he/she solve large sets of equations with many unknowns; the solution algorithm respects those restrictions. Two web pages embodying this problem-solving environment have been developed; one around a single continuous process with a collection of options on its input/output structure, and one around a batch process. Visual feedback is provided for the user to identify variables whose values can be calculated with the available data and the equations that can be used in the calculation. Through these web pages, users can solve their own problems or practice on predefined problems displayed at the bottom of the page in the form of multi-choice quizzes. The expectation is that such an environment of free, student-centred experimentation can provide structure and guidance on the process of formulating and solving mass balance problems.

Scheduling challenges in biopharmaceutical manufacturing

Abstract For a scheduling tool to be acceptable and effective in biopharmaceutical manufacturing it must embrace the richness of constraints that exist in the biomanufacturing floor. It must also provide a way to quickly develop and modify feasible schedules. This paper discusses the unique challenges that characterize scheduling in biopharmaceutical manufacturing and sketches the features of a tool that can effectively meet these challenges.

Optimization of biopharmaceutical manufacturing with scheduling tools - Experiences from the real world

Abstract This paper presents industrial experience with a resource-constrained batch process scheduling program. The batch process representation is loosely based on the ISA S88 batch process standard. This representation allows the import of batch process information from other software, e.g. batch process simulators. The scheduling algorithm is a non-optimization approach that proceeds in two steps. First a bottleneck analysis is done to determine a lower bound on the process cycle time, and all the batches are scheduled accordingly. Second, if conflicts remain, they are resolved by applying progressively aggressive modifications to the schedule. This approach to scheduling was tested on several biotech processes. These processes consist of a sequence of batch steps performed with dedicated equipment. The scheduling challenges in biotech processes lie in the ancillary operations: media and buffer preparation, vessel and line cleaning, and chromatography column preparation. Such operations may use shared resources and may serve to couple process suites with otherwise dedicated equipment. These considerations are further complicated by variability in process durations. Three case studies, which are based on a process for the manufacture of monoclonal antibodies (MABs), illustrate the value of a constrained-resource scheduling tool for biotech processes. In the first case study, the scheduling tool shows that auxiliary cleaning equipment can limit batch production. A second case study shows how scheduling tools can calculate the size of a purified water system. A third case study illustrates how to use scheduling tools to mitigate the effects of process variability.

Cycle Time and Throughput Analysis in Batch Biochemical and Pharmaceutical Processing

Abstract Throughput analysis of batch processes aims at identifying opportunities for increasing annual production by identifying and eliminating size and scheduling bottlenecks. Batch size and cycle time are identified as the key parameters in throughput analysis and the relationship between them is analytically investigated in the context of a single-product batch plant. A framework for performing throughput analysis is developed and demonstrated with the use of a synthetic pharmaceutical production example.

Modeling and Evaluation of Biological Wastewater Treatment Facilities - What Role can Process Simulators Play?

Abstract The role of process simulation in designing, evaluating, and optimizing wastewater treatment facilities is discussed. Alternatives for controlling VOC emissions from treatment plants and removing dissolved solids from clarified effluent streams are evaluated