A STOCHASTIC PROGRAMMING APPROACH FOR ANNUAL DELIVERY PROGRAM CREATION IN LNG SUPPLY CHAINS

ARMANDO GUARNASCHELLI; HECTOR ENRIQUE SALOMONE

Roma

Conferencia; EWGSO Conference ECSO-2017 ?Stochastic Optimization in Service Science?; 2017

EURO Working Group on Stochastic Optimization (EWGSO)

In this work, a new modeling approach to create annual delivery programs (ADP) for LiquefiedNatural Gas (LNG) supply chains is introduced. An ADP is a fulfilling, inventoryand delivery plan for a 12-month operation horizon of a LNG producer. It intends to satisfycustomers demand under long and short term contracts and possibly spot sales. This work,quantifies the impact of uncertainty both in LNG demand and sale price [1].LNG sale is generally regulated by contracts [2]. In this work, a variety of contract types isconsidered through parametrization, including, overstock, understock and shipment cancellationclauses for customers [3]. These are economic clauses that impact on the annualbenefit obtained by the LNG producer. The approach also includes the concept of spotmarket trade which takes place regularly in the LNG business, allowing to alleviate LNGstock fluctuations at production facilities.To consider uncertainty, the deterministic equivalent of a two-stage stochastic program isdescribed, the scenario generation and solution approach is discussed. The model is validatedusing a real-world example, results and the value of incorporating uncertainty in priceand demand (Value of the stochastic solution) are given and discussed. Conclusions andfuture research directions are given.