Endüstriyel enerji yönetiminde değer akış haritalarının kullanımı

Abstract

Hızla gelişen dünyada enerji kaynaklarının kullanımı dikkatle ele alınması gereken bir konudur. Özellikle fosil yakıtların kullanımı, hem günümüz hem de gelecek açısından önem arz etmektedir. Başta kömür, doğalgaz ve elektrik olmak üzere enerji kaynaklarının en yoğun olarak kullanıldığı alanlardan birisi imalat sektörüdür. Bu yüzden, imalat işletmelerinde enerji yönetimi, hem işletmeler hemde ülkeler açısından kritik öneme sahiptir. Günümüzde yasal düzenlemeler ve teşvik programları ile artırılmaya çalışılan verimlilik oranları, henüz istenilen oranlara yaklaşamamıştır. Dolayısıyla endüstriyel enerji verimliliği ve tasarrufunu artıracak yeni politikalar ve araçlar geliştirilmeye devam edecektir. Enerji verimliliği çalışmaları, enerji üretimi ve tüketiminin çevresel etkileri de göz önünde bulundurularak ele alındığında daha da kritik hale gelmektedir.İmalat sektöründe her türlü israfın önlenmesi esasına dayanan yalın üretim felsefesi, imalat yönetiminde kullanılacak bazı yeni yaklaşım ve araçları beraberinde getirmiştir. Yalın üretimin dayanağı olan beş temel ilke değer, değer akışı, akış, çekme ve mükemmelliktir. Bu ilkelerin hayata geçirilmesi sürecinde kullanılan araçların en önemlilerinden birisi değer akış haritaları pratik ve etkili bir kağıt-kalem tekniğidir. Değer akış haritaları, bir ürünün imalatı sürecinde değer yaratılan ve yaratılmayan işlemlerin tespit edilmesini sağlayarak iyileştirme yapılabilecek noktaların belirlenmesini mümkün kılar. Değer akışı haritalama sürecinde tesiste yapılan etütlerle oluşturulan mevcut durum haritası hazırlandıktan sonra, belirlenen noktalarda yapılacak iyileştirmelerle işletmenin ulaşacağı durumu gösteren gelecek durum değer akış haritası hazırlanır. Bu aşamada tespit edilen birden fazla iyileştirmenin hepsinin aynı anda gerçekleştirilmesi kaynak eksikliği sebebiyle genellikle pek mümkün olmaz. Dolayısıyla, belirlenen çeşitli olasılıkların sınanması ve bunların performanslarının birbirleri ile karşılaştırılması gereklidir. Senaryo hazılamak üzere kullanılabilecek araçlardan birisi, bir tür bilişsel haritalama tekniği olan Bayes (İnanç) Ağları'dır. Mevcut durum haritaları üzerinde tespit edilen iyileştirme noktalarına ait israfın önlenmesine yönelik veri ve bilgileri kullanarak gelecek durum değer akış haritasının hazırlanması aşamasında Bayes Ağları destek olabilir. Bu tez çalışması kapsamında, değer akışı haritalamada gelecek durum haritasına geçiş aşamasında Bayes inanç ağlarının, enerji israflarının önlenerek verimliliğin artırılmasında kullanıldığı bir yöntem geliştirilmiş ve yöntemin kullanılabilirliğini sınayan basit bir uygulama çalışması yapılmıştır.

In a rapidly developing world, usage of energy resources requires a special attention. Especially, usage of fossil fuels is critical for both today and future. Manufacturing industry is one of the most intense users of energy sources like coal, natural gas and electricity. Because of this fact, energy management in industrial plants is crucial for both companies and countries. Today, industrial energy efficiency is tried to improve by official regulations and government promotions but still could not reach an acceptable level. So, it is urgent to develop new policies, technologies and tools to improve industrial energy efficiency and savings. Efficient use of energy is extremely critical in energy consumption, and the saving potential in different industries allows research on efficiency improvement projects. Today, main focus of the industrial energy efficiency studies are on the equipments that use intense energy. We can see samples of studies focused on variable speed drivers, waste heat recovery, high efficient motor utilization, preventing the leaks in air compressors and preventing the pressure drop. A typical energy efficiency study in industry field focuses on determining energy intensive proseses, analysis of enegy bills, determining theoretical minimum energy usage, determing key performans indicators, searching new proseses and technologies. Importance of energy efficiency became clearer if it is handled with the environmental impacts of energy production and consumption. International regulations and customer preferences make companies to be more sensitive about energy usage.Lean production is an engineering approach which is based on avoiding all kind of non-value adding activities and supplies new approaches and tools. Lean production is the way of putting entire value stream for specific products relentlessly in the foreground and rethinking every aspect of jobs, careers, functions, and firms in order to correctly specify value and make it flow continuously along the whole length of the stream as pulled by the customer in pursuit of perfection. Five main principles of lean production are value, value stream, flow, pull system and perfection. Some special tools are developed while realizing these principles. Some of the these important tools are value and value stream mapping, determining and preventing all kind of wastes (material, time, human sources etc.) takt time, kanban and visualization, 5S, standard work, pull system, creating one peace flow, continuous improvement and perfection, process kaizens, reducing setup and setting times, reducing maintenance times and cellular manufacturing.Value Stream mapping is one of the most important tools of lean manufacturing and beneficial in order to identify value adding and non-value adding tasks and processes. Value stream mapping allows observing the flow of material and information as a product or service makes its way through the value chain. While mapping values stream, firstly, current state maps are prepared by hands on experiences in the plant and then future state map, which shows target improvements, is prepared. It is useful for determining the bottlenecks but achieving the targets is a challenge.While preparing future state maps and deciding on improvements, it is generally impossible to focus on all of the waste sources; so it is needed to try some scenarios and compare the overall performances. Bayesian (belief) Networks are one of these tools, which can be used for observing the outcomes of different scenarios. Bayesian Networks can exploit the data and knowledge about sources of waste and support the preparation of future state maps.Combining industrial energy efficiency and the phlosophy of lean thinking can be made for many reasons in different ways. In order to determine and lean energy wastes, lean tools would be beneficial. Furthermore, some lean tools can be transformed to energy efficiency tools. Non-value adding energy usages, reducing energy due to work-in-processs, reducing over production and faulty production, organizing kaizens focused on energy efficiency and cellular energy usage would be typical examples for these efforts.This thesis study suggests a framework to use value stream maps to detect non-value adding energy consumptions and using belief networks to establish future state energy-value stream maps. Study contributes to the energy efficiency research field by suggesting a solution for energy efficiency bottleneck handling.The framework is constructed in three steps. In the first step, lean energy consumption analysis in the production company so that focus is given only on the energy consumption of value adding activities. In the second step, energy efficiency analysis is applied to design the current state energy-value-stream-map (E-VSM) focusing only on the energy related activities. In the last and third step, by the help of E-VSM, analysis by determining the factors or parameters for the Bayesian Networks are defined. Survey with experts would allow finding conditional probabilities for the bottlenecks. Bayesian Network will lead for developing different scenarios for constructing the future state E-VSMs.A sample application is performed to demonstrate the three steps of the suggested framework. The case is a small-medium size enterprice (SME) producing shaped and covered chipboard. The enterprice produces table tops and stool parts. The production in the facility is performed by semi-automatic machines and the process is observed to be lean. Characteristics of the manufacturing workshop allow lean applications naturally. Production is run in a small area, the communication between work stations can be done by human voice and a simple pull system exists. There is an important waste in a critical point. The boards are thicker than required even though the size is standardized by the Turkish Standards Institute according to fragility. It makes products heavier and requires more drying and cooking times where energy is used intensively. Also the chopping process is performed in three steps although there exist a technology that does it in one step. In the enterprice, the serial production has just started and the machine setting and employee experience are to be improved.In the E-VSM, it is observed that during the processes of chip drying and hot press non-value-adding energy usage (NVA-EU) exists. This parameter is chosen as the decision node of Bayesian Map. Thus, with the change of NVA-EU, the change of energy efficiency and the other parameters will be observed. Eight parameters are considered in expert interviews and a conditional probabilistic table is prepared. Bayesian Network is transferred to NETICA software to try scenarios. Developed Bayesian Network which is a simple one with a single parameter, non-value adding energy usage: NVA-EU. Changes in the current state E-VSM create the scenarios. With the change of NVA-EU, a serious amount of change in energy efficiency (EE) is observed in the software analysis. So, it seems reasonable for the enterprice to focus on non value adding energy usage.To sum up, in this thesis study, a framework for using VSMs for energy efficiency improvements is proposed. In order to create future state E-VSMs, Bayesian networks are used. It is observed with a small sample application that the proposed framework is applicable and useful. Especially for the situations where there is not enough quantitative data, and using expert beliefs is inevitable, combining E-VSMs and Bayesian Networks is proven to be beneficial in order to determine improvement areas. SME?s are commonly faced with this kind of situations and proposed framework is not difficult to be used. Yet, in order to construct detailed scenarios, more parameters are to be used in Bayesian Networks. These can be driven both from production and other energy related concepts like energy consumption and energy intensity.In order to develop the proposed frame work, a numerical evaluation is obligatory. Further work will be done to measure the performance. It is only possible to minimize the energy usage after observing the performances. Inferences and benefits of Bayesian Nets will be more beneficial with the scenarios created in line with the performances.