MACRo 2015- 5th International Conference on Recent Achievements in Mechatronics, Automation, Computer Science and Robotics
Simulator for Distribution Scheduling in Downstream István HECKL1*, Péter István BORBÁS2, Béla SZOMBATHELYI1, Márton FRITS1 1
Department of Computer Science, University of Pannonia Egyetem u. 10, Veszprém 8200, Hungary 2 Supply Chain Management, MOL Plc. Olajmunkás u. 2, Százhalombatta 2443, Hungary *Corresponding author:
[email protected]
Manuscript received January 12, 2015, revised February 9, 2015.
Abstract: A novel simulator for distribution scheduling in downstream of the oil industry is introduced in this work to identify inconsistencies in the proposed short term schedule. The simulation system is actively used in the MOL Plc which is a major regional oil company in Central Europe. It takes into account production and storage sites, tanks (mobile and immobile part), and of course the pipelines. The pipeline network features branching points, the pipe diameter may change between two sites, there are parallel pipelines between two sites, and the direction of a pipe can be reversed. Different products, next to each other, are transported through a single pipe. The movement of the product in the pipeline is driven by the recently pumped product. The exact locations of the products within the pipes are calculated by this custom simulator. The simulation has three phases: providing the initial data and the input, such as the actual contents of the pipes, tanks, and the planned product transfers; the simulation itself; and the presentation of the results. The initial data come from human operators and other IT systems. The simulator highlights the potential error in the schedule. The changing content of each pipe and the levels of the tanks can be displayed. If inconsistency occurs, the schedule has to be modified and the simulation has to be restarted. The simulator is extended to be able to handle barge and train transportation, and spots errors regarding foreign inventory. Keywords: Modelling, simulation, automation, oil industry, scheduling.
1.
Introduction
Process simulation has great significance in the industry. A lot of systems are too large to completely see through its inner operations. Thus, we cannot predict accurately what will happen for certain inputs. This has usually two causes. First, the system in question can have a great number of components each with plenty of properties. Thus, it is challenging or impossible to keep in mind all the data. Second, the complexity of the system is high. A single input may affect a number of components; these components affect other components, etc. Our initial input can cause unwanted side-effects. That is 73
10.1515/macro-2015-0007
Unauthenticated Download Date | 12/13/17 8:24 PM
74
I. Heckl, P. I. Borbás, B. Szombathelyi, M. Frits
where a simulator can be a great help. A simulator calculates the behavior of a system, i.e., determines the future states of the system if the inputs and the current state are given. Moreover, a simulator helps us to identify bottlenecks in the system by running different scenarios. Process simulation is especially important for the oil industry. In the downstream division, a lot of products are to be produced. The job of the Supply Chain Management (SCM) departments in the oil companies is to coordinate the transport and the storage of both the raw materials and the products. MOL Plc is the major Hungarian oil company. A major point of the MOL Group strategy is to extend and enhance the adaptation of the Supply Chain Management philosophy continuously. The SCM department of MOL had not such a simulator program which would totally satisfy all of its needs. Our main goal is to develop and extend a novel process simulation system which enhances the SCM operation of MOL. The introduction is followed by the problem definition and detailed literature review in the next section. The third section briefly describes the capabilities and the operation of the proposed simulator. The last section summarizes the results.
2.
Problem definition and literature review
A simulator is introduced in this work which is capable to: (i) keep track of the products and raw materials regarding the SCM department, (ii) help the design of the pipeline schedule, (iii) and simulate and visualize the changing contents of the product pipelines. The proposed system can be regarded as a decision support tool as well. Making decision in the oil industry is a highly complex process. Decisions are made at different hierarchy levels and different areas within the SCM itself. Still, a significant number of chemical and oil companies are held back by obsolete decision making processes; see (Lasschuit and Thijssen, 2004). Modeling, simulation, scheduling, and production planning are important research topics in the oil industry. An overview has been presented in (Crama, 2001) about different production planning approaches in the process industry. Herrán-González et al. (2009) modeled and simulated a gas distribution pipe network. Their main focus was on the gas ducts. A decision support tool has been presented in (Lasschuit and Thijssen, 2004) for supply chain scheduling and planning both in the chemical and the oil industry. The proposed system gives a coherent framework taking into account realtime market and operational data which allows consistent economic and operational steering. A general framework for modeling petroleum supply chains has been introduced in (Neiro and Pinto, 2004). Specific models have been proposed for pipes, tanks, and refineries. The topology of the concomitant superstructure is highly complex. Based on its topology a MINLP problem is formulated. Rejowski and Pinto (2008) proposed to generalize the previous framework from discrete to continuous timeframe. Genge et al. (2013) use Emulab and Simulink software tools to explore the dependencies between critical infrastructures.
Unauthenticated Download Date | 12/13/17 8:24 PM
Simulator for distribution scheduling in downstream
Demand
Validated product pipeline schedule
Planned product pipeline schedule
Update states
Operator
Planned blending order schedule
75
ProdSim Pipes
Tank
Layout
Validated blending order schedule
Reports
Figure 1: The operation of the ProdSim simulator.
3.
The ProdSim simulator
The proposed system is termed as ProdSim. It refers to product pipeline and tank simulation. Figure 1 illustrates the operation of ProdSim. This figure is not complete because it does not show the numerous connecting IT systems. ProdSim alone would not be useful, it has to exchange data with Orion (handles production), Logir (keeps track the pipe contents), Fir (manages other important data of the refinery), SAP, DDM (discrete demand management which is responsible for sales) among others. ProdSim has to know the layout of the pipeline network and the properties of the depots. In the modeling step, information is to be given about the products, pipes, tanks, refineries, and the sites. By doing this, we map a real life system into a conceptual model. The pipeline network features branching points, the pipe diameter may change between two sites, there are parallel pipelines between two sites, and the direction of a pipe can be reversed. ProdSim also knows the current state of the system, i.e., the content of tanks and the product pipelines. This information is monitored regularly. If everything goes according to the plans then the projected and the actual data should match. The demands are collected by salesman. These demands are known in advance. A demand description includes the product type, the delivery time, the delivery method, the amount. There are two types of sites: depots and refineries. Only refineries can create products, depots only store and distribute the final products. The products can be transported by various methods including railway tank car, barges, product pipelines, and trucks. The demands are flexible, meaning amount and delivery time may change over time. This uncertainty has to be handled. The operator of the ProdSim determines the future blending order schedule and the product pipeline schedule. The blending order schedule is the production plan for the different products. The product pipeline schedule defines when, where, and in which order are products transported.
Unauthenticated Download Date | 12/13/17 8:24 PM
76
I. Heckl, P. I. Borbás, B. Szombathelyi, M. Frits
The output of ProdSim is the validated product pipeline and blending order schedule. The operator plans the schedules and justifies them with ProdSim that has been designed exactly to this task. The validated product pipeline schedule and its result can be given as reports or the changing content of the tanks and the product pipelines can be displayed in a visualization module. ProdSim has the following functional parts: the administration module, the simulation engine, the visualization module, the interfaces, and the connecting database. Performing a simulation has three main steps: (i) specifying the input to ProdSim by the operator (these are the planned product pipeline schedule, blending order schedule, and the updates of the current system state), (ii) executing the simulation, (iii) overviewing and analyzing the results. If the simulator indicates an error then the schedule has to be modified and the simulation has to be rerun. Originally the simulator focused only to the transportation in pipelines but it is extended with transportation in barge and train. The other major improvement is the handling of foreign inventory. Tanks do not contain only MOL products. For example, government owned products (strategic reserve) are stored as well. Now the simulator can warn us, if such products are scheduled to sell.
Figure 2: Screenshot of scheduler of module of ProdSim.
4.
Summary
The ProdSim system is capable to validate the planned product pipeline schedule and the blending order schedule. The result of the simulation can be presented in reports both in the program itself and in excel table, or in the visualization module. Figure 2 shows a screenshot of the scheduler module. The upper part shows product transports
Unauthenticated Download Date | 12/13/17 8:24 PM
Simulator for distribution scheduling in downstream
77
(the arrows), the middle displays the content of the selected pipelines, and the lower part shows the levels of certain tanks. The usefulness of the ProdSim is demonstrated by the fact that it is actively used by a major regional oil company (MOL). It made possible to decrease the average level of the inventory, moreover the planning horizon for the short term scheduling increased from 4-7 days to a month. ProdSim will be further extended in the future. Specifically, the internal logistics of the sites will be taken into account. Moreover, a heuristics module is planned, which can assist greatly the work of the operator.
Acknowledgements We acknowledge the financial support of the Hungarian State and the European Union under the TAMOP-4.2.2.A-11/1/ KONV-2012-0072.
References [1] Crama, Y., Pochet, Y., Wera, Y., 2001, A Discussion of production planning [2]
[3]
[4]
[5]
[6]
[7]
approaches in the process industry, Core Discussion Papers 2001/42 Genge, B., C. Siaterlis, M. Hohenadel, 2013, AMICI: An Assessment Platform for Multi-Domain Security Experimentation on Critical Infrastructures, Lecture Notes in Computer Science 7722, 228-239. Herrán-González, A., De La Cruz, J. M., De Andrés-Toro, B., Risco-Martín, J. L., 2009, Modeling and simulation of a gas distribution pipeline network. Applied Mathematical Modelling 33, 1584–1600. Lasschuit, W., Thijssen, N., 2004, Supporting supply chain planning and scheduling decisions in the oil and chemical industry, Computers & Chemical Engineering 28, Issues 6-7, 863-870. Herrán-González, A., De La Cruz, J. M., De Andrés-Toro, B., Risco-Martín, J. L., 2009, Modeling and simulation of a gas distribution pipeline network, Applied Mathematical Modelling 33, 1584–1600. Neiro, S. M. S., Pinto, J. M., 2004, A general modeling framework for the operational planning of petroleum supply chains, Computers & Chemical Engineering 28, 871-896. Rejowski Jr, R., Pinto, J. M., 2008, A novel continuous time representation for the scheduling of pipeline systems with pumping yield rate constraints, Computers & Chemical Engineering 32, Issues 4-5, 5 1042-1066.
Unauthenticated Download Date | 12/13/17 8:24 PM
