PLANNING AND SCHEDULING FOR PETROLEUM …

1 ISSN 0104-6632 Printed in Brazil Vol. 19, No. 02, pp. 207 - 228, April - June 2002*To whom correspondence should be addressedBrazilian Journalof ChemicalEngineeringPLANNING AND SCHEDULINGFOR PETROLEUM refineries USINGMATHEMATICAL , ,2 and *1 Department of Chemical Engineering, University of S o Paulo05508-900, S o Paulo - SP, : Petr leo Brasileiro S/A(Received: December 20, 2001 ; Accepted: April 17, 2002)Abstract - The objective of this paper is the development and solution of nonlinear and mixed-integer(MIP) optimization models for real-world PLANNING and SCHEDULING problems in PETROLEUM , we present a nonlinear PLANNING model that represents a general refinery topology and allowsimplementation of nonlinear process models as well as blending relations. The optimization model isable to define new operating points, thus increasing the production of the more valuable products andsimultaneously satisfying all specification constraints.

2 The second part addresses SCHEDULING problemsin oil refineries , which are formulated as MIP optimization models and rely on both continuous anddiscrete time representations. Three practical applications closely related to the current refinery scenarioare presented. The first one addresses the problem of crude oil inventory management of a refinery thatreceives several types of crude oil delivered exclusively by a single oil pipeline. Subsequently, twooptimization models intended to define the optimal production policy, inventory control and distributionare proposed and solved for the fuel oil and asphalt plant. Finally, the PLANNING model of Moro et al.(1998) is extended in order to sequence decisions at the SCHEDULING level in the liquefied PETROLEUM gas(LPG) area for maximization of the production of petrochemical-grade propane and product : optimization, PLANNING , SCHEDULING , operations research, mixed-integer eighties were characterized by the emergenceof international markets and the development ofglobal competition.

3 The chemical processingindustry had to restructure in order to competesuccessfully in this new scenario and bettereconomic performance with more efficient plantoperation has been achieved (Moro et al., 1998).Implementation of advanced control systems inoil refineries generated significant gains inproductivity of the plant units. These resultsincreased the demand for more complex automationsystems that take into account production a result, unit optimizers were , the optimization of production unitsdoes not assure the global economic optimization ofthe plant. The objectives of individual units areusually conflicting and thus contribute to suboptimaland many times infeasible operation. The lack ofcomputational technology for production schedulingis the main obstacle to the integration of productionobjectives into process operations (Barton et al.)

4 ,1998). A more efficient approach would incorporatecurrent and future constraints in the synthesis ofproduction schedules. The short-term , and Journal of Chemical Engineeringobjectives must be translated into operatingconditions for the processing units. Such anapproach supplies an analytical tool for the effect ofeconomic disturbances in the performance of theproduction system and provides mechanisms toaccount for commercial and paper describes the approach taken in thedevelopment of optimization models for productionplanning and SCHEDULING of oil refineries . The plantis divided into subsystems, which although coupled,allow development of the representation of the mainscheduling activities within relevant time final objective is to develop strategies forincorporating these models in an automated planningand SCHEDULING system that generates paper is organized as follows: first, anoverview of PLANNING and SCHEDULING activities in oilrefineries is introduced.

5 Developments in mixed-integer representations for nonlinear planningmodels are presented, followed by a discussion ofoptimization work in refinery SCHEDULING withapplications in crude oil management, productionand distribution of oil products, such as fueloil, asphalt and LPG. Finally, conclusions are drawnand current as well as future developments OF PLANNING ANDSCHEDULING IN OIL REFINERIESThe potential benefits of optimization for processoperations in oil refineries with applications of linearprogramming in crude blending and product poolinghave long been observed (Symonds, 1955). Oilrefinery management is increasingly concerned withimproving the PLANNING of their operations. Themajor factor, among others, is the dynamic nature ofthe economic environment. Companies must assessthe potential impact of variations in demands forfinal product specifications, prices and crude oilcompositions or even be able to explore immediatemarket opportunities (Magalh es et al.)

6 , 1998).Coxhead (1994) identifies several applications ofplanning models in the refinery and oil industry,such as crude selection, crude allocation for multiplerefineries, partnership models for raw materialsupply and operations availability of LP-based commercial softwarefor refinery production PLANNING , such as PIMS(Process Industry Modeling System - Bechtel, 1993),has allowed the development of general productionplans for the whole refinery, which can beinterpreted as general trends. As pointed out byPelham and Pharris (1996), PLANNING technology canbe considered well developed and majorbreakthroughs should not be expected. The majoradvances in this area will be based on modelrefinement, notably through the use of nonlinearprogramming, as in Picaseno-Gamiz (1989) and,more recently, Moro et al. (1998) and Pinto andMoro (2000).Bodington (1992) also mentions the lack ofsystematic methodologies for handling nonlinearblending relations.

7 Ramage (1998) refers tononlinear programming (NLP, MINLP) as anecessary tool for the refineries of the 21st century,as a result of the significant progress made in thenineties (Viswanathan and Grossmann, 1990; P rn etal., 1999).On the other hand, there are few commercial toolsfor production SCHEDULING and these do not allow arigorous representation of plant particularities (Rigbyet al., 1995; Moro et al., 1998). For that reason, refineries are developing in-house tools stronglybased on simulation (Steinschorn and Hofferl, 1997;Magalh es et al., 1998) in order to obtain essentialinformation for a given system (Moro and Pinto,1998). In the open literature there are specificapplications based on mathematical programming,such as crude oil unloading and gasoline blending(Bodington, 1992; Rigby et al., 1995; Shah, 1996;Lee et al., 1996). Ballintjin (1993), who comparescontinuous and mixed-integer linear formulationsand points out the low applicability of models basedonly on continuous variables, also discusses the lackof rigorous models for refinery has also been recognized that the integration ofnew technologies into process operations is anessential profitability factor and that this can only beachieved through appropriate PLANNING (Cutler andAyala, 1993; Macchietto, 1993).

8 According to asurvey of hydrocarbon processing companies,management pointed to sales and PLANNING , planningand operations management and PLANNING anddistribution (Bodington, 1995) as major areas forprocess integration. Mansfield et al. (1993) discussthe issue of integration of the process control,optimization and PLANNING activities into gasolineblending. Bodington and Shobrys (1996) andSteinschorn and Hofferl (1997) point out theimportance of on-line integration of PLANNING , SCHEDULING and and SCHEDULING for PETROLEUM refineries 209 Brazilian Journal of Chemical Engineering, Vol. 19, No. 02, pp. 207 - 228, April - June 2002 PLANNING MODEL This work focuses on the development ofnonlinear PLANNING models for refinery activities involve optimization of rawmaterial supply, processing and subsequentcommercialization of final products over one orseveral time periods.

9 Moro et al. (1998) developed a nonlinearplanning model for refinery production that canrepresent a general topology. The model relies on ageneral representation of refinery processing units inwhich nonlinear equations are considered. The unitmodels are composed of blending relations andprocess equations. Also, the unit variables mustsatisfy bound constraints, which consist of productspecifications, maximum and minimum unit feedflow rates and limits on operating typical oil refinery generates several streamsthat are blended in order to specify a commercialproduct. Furthermore, there are products of differentgrades that must satisfy market demands. The modelassumes the existence of several processing units,which produce a variety of intermediate streams withdifferent properties that can be blended to constitutethe desired products. The topology of the refinery isdefined by sets that specify connections betweenstreams and model of a typical unit is represented by thefollowing variables:i) Feed flow rate: this is the combination of the ratesof every incoming ) Feed properties: these are derived from the mixingof individual streams calculated through blendingalgorithms that are generally ) Unit operating variables: variables such as heateroutlet temperature and reaction temperature are usedto control unit performance.

10 These variables usuallyinfluence product flow rates and properties in a verynonlinear ) Product flow rates: each product stream flow rateis a function of the feed flow rate, the feed propertiesand the operating variables. It is important to notethat since each product stream can be sent to variousdestinations, it may be further split into ) Product properties: these are functions of the feedproperties and unit operating real-world application was developed forproduction PLANNING at the REVAP refinery in dos Campos (SP, Brazil), as illustrated in Figure1 and described in detail in Pinto and Moro (2000).This refinery has one crude distillation unit (CD1),one vacuum distillation unit (VD1), one FCC unit(FCC), one propane-deasphalting unit (PDA), threehydrotreating units (two for kerosene and one fordiesel, referred to as HT1, HT2 and HT3), one C3/C4separation unit (DEP) and one MTBE productionunit (UMTBE).