Bu çalışmada, artan elektrik talebinin mevcut ve yeni kurulacak santraller ile karşılanabilmesi için bir üretim genişleme planlaması problemi ele alınmıştır. Bu problem için tek amaçlı lineer olmayan bir model önerilmiş, modelin uygun ortamda çözümü ile enerji yatırımına ilişkin değerlendirmelerde bulunulmuştur. Elektrik üretimine ilişkin yatırım problemlerini ele aldıkları konular ve kullandıkları yöntemler ile çeşitlendirmek mümkündür. Güncel hususların da modellere eklenmesi bu problemleri daha gerçekçi hale getirmektedir. Bu çalışmanın amacı Türkiye'nin artan elektrik talebini karşılamak amacıyla ulaşması gereken enerji karmasını incelemektir. Bu amaçla, çeşitli kaynakların elektrik üretimindeki paylarının 2022-2035 yılları arasında nasıl olacağı, hangi kaynaklardan ve hangi güçte yeni tesislerin kurulacağı ve bunların emisyon salınımına etkisi araştırılmıştır. Planlamada karşılanamayan talep olmayacak şekilde tercih edilen elektrik talebi, tesislerin inşa süreleri, kullanılabilirlik faktörleri de göz önüne alınarak modele dahil edilmiştir. Yeni kurulacak tesislerin kapasitesi ve bu tesisin kurulma kararı olan ikili değişkenin, yatırım maliyeti hesaplanırken birlikte bulunmaları amaç fonksiyonunu lineer olmayan hale getirmektedir. Tek amaçlı, lineer olmayan ve karışık tam sayılı olarak ele alınan problemin çözümü için Python programlama dili ve geniş kütüphanesinden faydalanılmıştır. Çözücü olarak kısıt tam sayı programlama çözücüsü olan SCIP tercih edilmiştir. Toplam 22 ayrı durumun değerlendirildiği başlıca iki senaryo vardır. Senaryo 1, 2021 Aralık verileri ile mevcut kurulu güç üzerinden, Senaryo 2 ise 2023 yılında Akkuyu Nükleer Santralinin devrede olacağı şartı üzerinden çeşitli durumları inceler. Bu durumlardan başlıcaları ithal edilen enerji oranının sınırlandırılması, yenilenebilir kaynakların üretimdeki payının arttırılması, net sıfır emisyon hedefi doğrultusunda emisyon salınımının azaltılmasıdır. Uygun optimallik aralığındaki senaryo çıktıları not edilmiş ve enerji yatırımına dair yorumlarda bulunulmuştur. Nükleer santralin varlığının plan maliyetini ve plan süresince salınan emisyonu düşürdüğü istatistiksel test ile gösterilmiştir. Çalışmada rüzgâr santrallerinin artan elektrik talebini karşılamada öncelikle tercih edildikleri, doğal gaz santrallerinin ithal enerji olduğu göz ardı edildiğinde elektrik üretim maliyetini düşürdükleri, güneş santrallerinin kullanılabilirlik faktörlerinin küçük olması nedeniyle elektrik üretiminde payının düşük olduğu ifade edilmiştir. Model çıktıları geçtiğimiz yıllarda duyurulan doğal gaz keşifleri, yenilenebilir enerji kaynaklarının desteklendiği finansal mekanizma ve küresel anlamda destekçisi olduğumuz net sıfır taahhüdü ile birlikte ele alınarak yorumlarda bulunulmuştur.
Undoubtedly, energy is the most controversial issue on the world's agenda. Technological developments, politics, crises, and even wars frame discussions about energy subjects. There is a significant evolution in politics and economy; Covid-19, The Russian-Ukraine War, and digital transformation impact both global balance and daily routines. The pandemic has hit the electricity supply-demand balance and shaken energy markets. This situation has shown that global warming enforcement should be more realistic and stricter restrictions for a more sustainable society. The decision-makers should distribute energy mix to generate electricity to meet energy demand and react quickly to price fluctuations. The use of renewable sources and decarbonization policies are a must for sustainability targets. However, natural gas and nuclear power are also necessary for a predictable and manageable market. While the rate of renewable energy use is increasing, the electricity market structure has evolved with the prosumers and storage facilities. Electric cars involved in energy networks and institutions' digital transformation have brought new dimensions. Therefore, macro and micro scales' energy planning and policy design are crucial. With population and income per capita growth, Turkey is named one of the biggest twentieth economies in the world. Although the share of renewable sources in electricity production, the carbon dioxide emission per capita rate has also increased over the years. As one of the developing countries, Turkey should plan its energy policy meticulously. Energy investments should be comprised of energy efficiency and energy density. There are two different approaches to examining energy investments. The first one is energy models that aim mainly to bring economic conclusions. The other is comprised of more technical algorithms and suggests mathematical solutions. The mix of examining the economic sides of the energy system and the technical properties of the power system are ideal. Generation expansion planning (GEP) problems deal with energy policies and investments, aiming to meet energy demand by adding new plants. GEP problems are differentiated by topics and methods, and integrating recent developments such as stochasticity into the models makes GEPs more complex but more comprehensive. This study focuses on a GEP problem to meet increasing electricity demand by existing generating units and newly built facilities. A single nonlinear model is proposed, then solved in an appropriate environment to make conclusions on energy investments. Predicting energy demand accurately is the first step to designing an investment plan for expansion planning. When the literature is examined, there are many studies focusing on demand forecasting. This thesis tries several options for predicting electricity demand between 2009 and 2016, and ridge regression is found to be far better than other techniques. After that, the plan period of the study (2022-2035) is predicted with ridge regression. Although the predicted demand for 2022 is closer to the actual electricity consumption than the official projections, the higher demand values calculated by Turkish Electricity Transmission Corporation (TEIAS) are preferred to avoid unmet energy demand. The target is here to predict Turkey's energy mix to meet the electricity growth and analyze the impact of this energy mix. The ratio of different energy sources in electricity production between 2022 and 2035 will be found. Then which units at what capacity will be built and the emission ratio will be investigated. While there is no unmet demand in the proposed model, construction time and availability factors of generating units are involved. The capacity of newly built units and the binary variable of these units -considered to be constructed or not- make the objective function nonlinear. A single objective mixed integer nonlinear problem is coded in Python 3.9.12 using the Pyomo software and solved via SCIP 7.0 solver. The optimality gap value is accepted between 1-10%. When it is required, the model is moved to Google Collaboratory, where accelerated compute environments are offered. There are two main scenarios, and these evaluate 22 cases in total. Scenario 1 is based on Turkey's installed capacity of December 2021. Scenario 2 assumes that Turkey's first nuclear power plant, Akkuyu NPP, is ready to generate electricity in 2023. Both two scenarios examine several cases, including reducing the energy import of Turkey, increasing the renewable share in electricity production, and achieving net zero emission targets. In addition, domestic coal production, a renewable production support mechanism called YEKDEM, and also CO2 emissions from electricity production are discussed in the study. The output of the study with the acceptable optimality gap is noted, and then conclusions about energy investments are made. The existence of nuclear power in the energy mix is found to reduce the cost of energy investment plans and also the emission rates. Wind turbines are preferred to meet increasing electricity demand with their low prices. Apart from the fact that Turkey has very limited gas reserves and it is a large natural gas importer country, the efficiency rate of natural gas while burning fuel into electricity puts it into the energy mix in most of the scenarios. However, solar energy plants with their low availability factors are included relatively low in energy investment plans. Distributed energy sources, which are mainly unlicensed projects, are also evaluated in the thesis. However, the policy encouraging the use of distributed sources is not found to reduce the plan cost or total emission significantly. Moreover, ongoing projects of power plants are also evaluated in scenarios. Their completion rates are also taken into account. Although those projects increase the plan cost, when they are embedded into the energy mix, they will lower emission rates. Since electric car use is increasing, battery costs are declining, electricity storage facilities are being improved, and carbon capture technologies are being supported, the proposed model can be enhanced and rerun. Especially, Turkey has taken some action on energy efficiency. The effect of such studies on the cost and the emissions that will be conducted on energy efficiency fields would be diversified the study. This study neglects the transmission line investments. However, the integration of transmission expansion planning can enhance the model. Moreover, adding stochasticity will bring more realistic conclusions but make a more challenging solution.
Except where otherwise noted, this item's license is described as http://creativecommons.org/licenses/by/4.0/
