ESCAPE, Industrial Forum, May 28th 2002

CAPE challenges in pulp and paper industry

Risto Ritala

Long term challenges of the pulp and paper industry

Pulp and paper industry - as most mature process industries - faces the following four challenges of being simultaneously

• flexible towards customers,
• efficient in use of the capital intensive assets,
• ecologically sustainable both when using resources and utilities and with respect to waste handling,
• and attractive working places for future workforce.

These challenges can be met with the help of decision support tools in all business processes: design, production, management of the supply chain, and in customer service. The decision support systems complement with advanced mathematical methods the users’ detailed knowledge about the pulp and paper processes.

Characteristics of (Finnish) pulp and paper industry and its business environment

The main characteristics affecting top-level decision making in pulp and paper industry are:

• shareholders expect a low-risk 12-18% return on capital employed (ROCE)
• industry is very capital intensive (a green field production site is an investment of 400..500 Million Euro)
• raw material costs and logistics cost are both well above 25% of total costs
• consolidation has been rapid in 1990s and is expected to continue: as a result production lines can be focused to a narrower range of products
• the industry has been strongly cyclic and even after the recent consolidation the cycles are still quite strong
• even though products are considered bulk commodities, the product portfolio has diversified and is expected to diversify further requiring increased flexibility of the operations
• process design has been outsourced to engineering consultancy companies for at least the last 20 years
• development of new or improved unit process is considered as the task of the process equipment vendors
• core competencies of the pulp and paper industry are: efficient operation of production processes, and customer relationship management.

The industry invested heavily on new production sites in 1980s and early 1990s. Since then the emphasis has been on practices and systems supporting the development of production efficiency. Finnish based industry in particular has been successful in developing production efficiency into the main competitive factor, as is demonstrated by its leading role in the consolidation of the industry.

CAPE challenge 1: Optimal design of the dynamics

Modern production lines produce paper web that is 10 m wide, 50 um thick, runs at the speed of 25 m/s and the web may carry a coating 5-10 um thick. A single production line may produce up to 400.000 tons per year. The production process is quite simple in unit operations - mostly dilution/mixing/thickening of flows - but complex in the way the unit processes are connected.

All production lines produce many grades: due to consolidation the product spectrum is narrower but due to market requirement for more diversified products the spectrum is denser. The main quality control loops are MIMO model based controllers supported by hundreds of PI controllers.

Present process designs are based on principles of 1970s when stability was of higher importance than flexibility and when no advanced control algorithms, such MPC, were available. The present practice is to design first the process and only afterwards the control system. The challenge is to achieve considerably better combination of stability and flexibility through systematic optimization of process dynamics. The approach should be applicable both to design of new production lines and to retrofits.

The application of present MIDO methods may be applicable to this problem, but it has not been proven to work in practice.

The critical requirements for the (conceptual) design optimization are:

• system consists of up to 10 tanks and 20-30 controllers including MPC controllers
• each flow carries 4-10 (mutually non-reacting) substances
• dynamic scenarios are based on market demand variations of the products
• the economic performance is the objective for optimization
process structure and parameters, and control architecture, algorithms and parameters are to be optimized.

CAPE challenge 2: Management of the supply chain as a dynamic system

The supply chain logistics costs constitute over 25% of total costs of paper industry. Contrary to the large effort put into developing production efficiency, the optimization of supply chain has received less attention and is thus expected to have a larger marginal of improvement.

When managing the supply chain - both incoming raw materials and outgoing products - the decision maker must make a tradeoff between minimizing the capital tied to warehoused products and maximizing the reliability of deliveries. At present, main paper producing companies control a large part of the supply chain and thus have full access to all information in the chain, such as warehouse levels. Furthermore, the relationship to the largest customers is tight partnership so that the paper producer may actually run the warehouse of the customer. The information is available through enterprise resource planning systems (ERPs) which, however, have little to offer in terms of tools for tactical decision making.

In particular the outgoing supply chain of products is subject to market fluctuations, cyclic and random. Therefore the supply chain is a dynamic stochastic system. The methods of operational optimization of production processes can be applied to supply chain. The problem to solve is to maximize economic performance subject to that the probability of customer running out of product is given (for each customer) and assuming that the optimization has access to all information in the supply chain up to the warehouse making customer deliveries. Furthermore, as entirely central decision making often leads to low employee motivation, the implementation of the decision support must be decentralized and based, for example of dynamic coordination benefits/costs at local level.

To our knowledge such optimizing decision support tools do not exist.

The critical requirements are:

• covers 1-4 production sites, 3-10 warehouses and 5-20 customers
• producers the order basis to production sites
• deals with transport delays, limited transport and warehouse capacity etc
• optimization carried out daily with 30-60 days optimization horizon.

this file in MS-Word format