Anna Lindholm

Plant-wide utility disturbance management in the process industry

Utilities, such as steam and cooling water, are often shared between production areas at large-scale sites. A disturbance in the supply of a utility is therefore likely to affect a large part of a site, and cause great loss of revenue. This study focuses on identifying disturbances in utilities and estimating the economical effects of such disturbances. A general method for reducing the loss of revenue due to utility disturbances, the utility disturbance management (UDM) method, is presented. Modeling of the effects of utility disturbances on production is needed to complete all steps of the method. In this paper, a simple on/off modeling approach is suggested to quickly obtain key performance indicators that may be used for decision support for proactive disturbance management. A matrix representation of a site and its utilities is introduced to simplify the computations. The UDM method is applied to an industrial case at Perstorp, Sweden.

A General Method for Handling Disturbances on Utilities in the Process Industry

Utilities, such as steam or cooling water, have shown to play an important role within the process industry, since a malfunctioning utility is a plant-wide disturbance that can lead to large revenue losses due to reduced production quantities. This work focuses on identifying disturbances on utilities that give economical consequences. Measures of utility availability and area availability are introduced and used for estimating the ratio of disturbances on utilities. A generic method for handling disturbances on utilities is presented, which could be applied using site models of different level of detail. Some modeling approaches for modeling a site are described and the framework of the general method is demonstrated with a case study example at Perstorp AB, Sweden. (Less)

Hierarchical scheduling and utility disturbance management in the process industry

The integration of scheduling and control in the process industry is a topic that has been frequently discussed during the recent years, but many challenges remain in order to achieve integrated solutions that can be implemented for large-scale industrial sites. In this paper we consider production control under disturbances in the supply of utilities at integrated sites together with the integration towards production scheduling. Utilities, such as steam and cooling water, are often shared between the production areas of a site, which enables formulation of an optimization problem for determining the optimal supply of utilities to each area at the occurrence of a disturbance. Optimization in two timescales is suggested to handle the scheduling and disturbance management problems in a hierarchical fashion. The suggested structure has been discussed with companies within the chemical process industry. A simple example is provided to show how the structure may be used.

Hierarchical scheduling and disturbance management in the process industry

The integration of scheduling and control in the process industry is a topic that has been frequently discussed during the recent years, but many challenges remain in order to obtain integrated solutions that may be implemented at large-scale industrial sites. This paper introduces a general framework for production scheduling (PS) and detailed production scheduling (DPS) using a two-level hierarchical approach. The PS activity generates a monthly production schedule based on information on orders and forecasts, and the DPS activity handles disturbances in production on an hourly basis. The focus is on disturbances in the supply of utilities, which often cause great losses at process industrial sites. The research has been conducted in close collaboration with Perstorp, a world-leading company within several sectors of the specialty chemicals market. A specification list provided by Perstorp has been used as a starting point for formulating the PS and DPS activities as optimization problems. An example that is inspired by a real industrial site is presented to show how the PS and DPS may operate and how the integration of these two functions behaves

Formulating an Optimization Problem for Minimization of Losses due to Utilities

Utilities, such as steam and cooling water, are often shared between several production areas at industrial sites, and the effects of disturbances in utilities could thus be hard to predict. In addition, production areas could be connected because of the product flow at the site. This paper introduces a simple modeling approach for modeling the relation between utility operation and production. Using this modeling approach, an optimization problem can be formulated with the objective to minimize the economical losses due to disturbances in utilities by controlling the production of all areas at a site. The formulation of the problem is general, and thus the optimization can be performed for any site with similar structure. The results are useful for investigating the impact of plant-wide disturbances in utilities, and can provide decision support for how to control the production at utility disturbances. To enable online advise to operators on how to control the production, the posed optimization problem is solved in receding horizon fashion.

A METHOD FOR IMPROVING PLANT AVAILABILITY WITH RESPECT TO UTILITIES USING BUFFER TANKS

Disturbances on utilities, such as steam and cooling water, often cause large revenue losses at industrial sites. These disturbances are often hard to handle, since they commonly are plant-wide disturbances that affect more than one production area at the site. Also, the production areas of the site are often dependent on other production areas because of the flow of product through the site, which makes the effect of disturbances on utilities hard to predict. In this paper, a simple method for decreasing the revenue loss due to disturbances on utilities is presented, where production areas are modeled as either operating at full speed or not operating (on/off), and buffer tanks between areas at the site are utilized. Both choice of levels in the buffer tank and control of the product flow at the site during a disturbance are discussed. (Less)

A general method for defining and structuring buffer management problems

In an industrial plant, availability is an important factor since increased availability often gives an increase of final production, which in many cases means an increased profit for the company. The purpose of using buffer tanks is to increase the availability either by separating production units from each other or by minimizing flow variations. However, the methods for achieving this goal is not trivial, and depend on the specific characteristics of the problem. This paper contributes to structuring the general buffer management problem for continuous chemical plants and suggests methods for solving some specific problems, presented as a case study at Perstorp AB, Sweden.

Hierarchical Production Scheduling - A Case Study at Perstorp

Planning and scheduling are functions that have large economic impact in the chemical process industry. For integrated sites with many interconnected production areas, obtaining production schedules that respect all production-related constraints is a complex task. One important issue is the constraints due to disturbances in utilities, such as steam and cooling water. These are often site-wide disturbances that may make it impossible to maintain desired production rates in several production areas at a site. In this study, scheduling at two levels of the functional hierarchy at a site of a world lead chemical industry, Perstorp, is handled. The activities are denoted production scheduling (PS) and detailed production scheduling (DPS). Real data of incoming orders and utility disturbances are used to produce a production schedule and detailed production schedule for one month. The PS and DPS problems are formulated as optimization problems, where production-related constraints such as production rate constraints, inventory limitations, and start-up costs are included. The objective functions of the PS and DPS problems are formulated to reflect the importance of different issues at the site. The procedure aims to show how the hierarchical optimization framework may be used to provide decision support for how to operate the production at a site in order to maximize profit while minimizing the effects of site-wide disturbances.

A Tool for Utility Disturbance Management

Disturbances in the supply of utilities may lead to large economical losses at industrial sites. In order to take appropriate decisions on for which utilities improvement efforts should be made, a tool for evaluating the economical effects of utility disturbances is needed. This paper presents a tool for quickly estimating the revenue loss caused by each utility, using simple matrix operations. A matrix representation of a site is introduced to be able to quickly get key performance indicators for any site. The revenue loss is estimated using an on/off production modeling approach.

Empirical Models for Utility Disturbances in the Process Industry

Abstract In this paper, a general method for identifying common disturbances in the supply of utilities to a process industrial site is presented, with focus on the chemical process industry. The method aims at finding typical utility disturbance trajectories and define them by simple measures, such as duration and time between failures. Statistical models are suggested for each of the measures. The estimated disturbance trajectories may be used as inputs to optimization problems for determining the optimal supply of utilities to each area of a site during a disturbance. Use of such optimization procedures can improve both proactive and reactive disturbance management for utility disturbances. Industrial data for some utilities, among others steam and cooling water, have been used to retrieve disturbance models for utilities at a specific site. These models may also give some information on the general characteristics of utility disturbances.