Value chain management for commodities: a case study from the chemical industry

M. Kannegiesser, H.O. Gunther, P. van Beek, Martin Grunow, C. Habla

    Research output: Contribution to journalJournal articleResearchpeer-review

    Abstract

    We present a planning model for chemical commodities related to an industry case. Commodities are standard chemicals characterized by sales and supply volatility in volume and value. Increasing and volatile prices of crude oil-dependent raw materials require coordination of sales and supply decisions by volume and value throughout the value chain to ensure profitability. Contract and spot demand differentiation with volatile and uncertain spot prices, spot sales quantity flexibility, spot sales price-quantity functions and variable raw material consumption rates in production are problem specifics to be considered. Existing chemical industry planning models are limited to production and distribution decisions to minimize costs or makespan. Demand-oriented models focus on uncertainty in demand quantities not in prices. We develop an integrated model to optimize profit by coordinating sales quantity, price and supply decisions throughout the value chain. A two-phase optimization approach supports robust planning ensuring minimum profitability even in case of worst-case spot sales price scenarios. Model evaluations with industry case data demonstrate the impact of elasticities, variable raw material consumption rates and price uncertainties on planned profit and volumes.
    Original languageEnglish
    JournalOR Spectrum - Quantitative Approaches in Management
    Volume31
    Issue number1
    Pages (from-to)63-93
    ISSN0171-6468
    DOIs
    Publication statusPublished - 2009

    Keywords

    • Demand uncertainty
    • Sales and supply network planning
    • Value chain management
    • Commodities
    • Chemical industry

    Fingerprint Dive into the research topics of 'Value chain management for commodities: a case study from the chemical industry'. Together they form a unique fingerprint.

    Cite this