聯合國環境大會於2022年決議制定全球塑膠公約，這是全球因應塑膠污染問題的多邊協議，透過政府間談判委員會擬定具法律效力的全球協議，最終化為每個國家的國家政策，希望藉由塑膠的生產管理、限制特定化學品與一次性產品使用、增加塑膠回收率、強化塑膠循環經濟等手段，最終達到消除塑膠污染的目標。目前最新進度為2025年8月瑞士日內瓦會議，因各方無法達成共識而流會。
　　塑膠產業是煉化產業的延伸與下游應用，煉化提供基礎原料，塑膠則轉化為生活必需品，然而，這種緊密關聯也讓塑膠產業受到原油市場波動、國際政策與環境議題的強烈影響。未來塑膠與煉化產業的關鍵在於：如何在維持供應鏈穩定的同時，推動減塑、循環利用與綠色轉型。由於臺灣是以出口為導向的經濟體，塑膠管理的國家計畫對煉化產業之發展與產業競爭力影響重大，爰此，本文將解析聯合國塑膠公約意涵、統整主要國家因應塑膠公約的政策、分析塑膠公約對臺灣煉化產業的影響，並彙整塑膠源頭替代材料與塑膠回收技術的現況與推動障礙，以及在限塑政策推動下臺灣煉化業的發展方向，期能找出塑膠污染問題的解決方法與未來管理方向，以供政府能在兼顧煉化產業的健全發展之下，制定因應全球塑膠公約的妥善國家政策。
　　The United Nations Environment Assembly resolved in 2022 to establish the Global Plastics Treaty, a multilateral agreement to address plastic pollution. Through an intergovernmental negotiating committee, a legally binding global agreement is drafted, ultimately becoming national policies for each country. The aim is to eliminate plastic pollution through measures such as managing plastic production, restricting the use of specific chemicals and single-use products, increasing plastic recycling rates, and strengthening the plastic circular economy. The latest progress was the Geneva, Switzerland meeting in August 2025, which was adjourned due to a lack of consensus.
　　The plastics industry is an extension and downstream application of the refining and chemical industry. Refining provides basic raw materials, while plastics are transformed into daily necessities. However, this close connection also makes the plastics industry highly susceptible to fluctuations in the crude oil market, international policies, and environmental issues. The key to the future of the plastics and refining industries lies in how to maintain supply chain stability while promoting plastic reduction, recycling, and green transformation. As Taiwan is an export-oriented economy, national plans for plastics management have a significant impact on the development and competitiveness of the refining and chemical industry. Therefore, this article will analyze the implications of the United Nations Convention on Plastics, summarize the policies of major countries in response to the Convention, analyze the impact of the Convention on Taiwan's refining and chemical industry, and compile the current status and obstacles to the promotion of alternative materials for plastic sources and plastic recycling technologies. It will also discuss the development direction of Taiwan's refining and chemical industry under the promotion of plastic restriction policies, in order to find solutions to the plastic pollution problem and future management directions, so that the government can formulate appropriate national policies in response to the global Plastics Convention while ensuring the healthy development of the refining and chemical industry.

　　台灣位處環太平洋火環帶上，國內火山包含大屯火山群、龜山島及綠島。火山地區地溫梯度高，地底下若有水源就可能發展成有潛力的地熱儲集層。早年大屯山區探勘發現地下儲集層溫度最高293℃極具發展潛能。然而地下流體大多含高鹽度且呈酸性，這些流體對井下套管及井口裝置造成嚴重腐蝕，要將其開採發電並不容易。
　　台灣中油公司肩負能源轉型之責，積極開發國內地熱能源，本研究針對大屯山酸性地熱井開發策略進行綜整分析，首先彙整大屯山過往腐蝕問題，再介紹國際成功開發酸性地熱井的方法，接著進行耐蝕合金篩選實務探討，最後提出酸性地熱井工安問題及開發建議。研究成果期望可作為國內酸性地熱井開採之參考。
　　Taiwan is located along the Pacific Ring of Fire. The country's volcanoes include the Tatun Volcano Group, Guishan Island, and Green Island. Volcanic areas have high geothermal gradients, and if there is an underground water source, it may develop into a potential geothermal reservoir. Early exploration in the Tatun mountain area discovered underground reservoirs with temperatures reaching up to 293°C, showing great development potential. However, the underground fluids mostly have high salinity and are acidic, causing severe corrosion to downhole casings and wellhead equipment, making power generation extraction challenging.
　　CPC Corporation Taiwan, bearing the responsibility of energy transition, actively develops domestic geothermal energy. This study conducts a comprehensive analysis of development strategies for the acidic geothermal wells in the Tatun Mountains. It first summarizes past corrosion problems in the Tatun area, then introduces international successful methods for developing acidic geothermal wells, followed by a practical discussion on the selection of corrosion-resistant alloys. Finally, it proposes occupational safety issues and development recommendations for acidic geothermal wells. The research results are expected to serve as a reference for the domestic extraction of acidic geothermal wells.

　　台塑石化公司煉油廠共有兩套甲基第三丁基醚單元(Methyl Tertiary Butyl Ether, MTBE)，係將入料四碳混合物中之異丁烯與甲醇反應生成高價產品MTBE，以供作汽油摻配及化學品原料使用。本案以第二套MTBE單元(MTBE#2)為改善標的，利用歷史數據資料進行大數據分析，對過去已知操作經驗，了解設備操作限制，進一步針對製程優化操作與控制，透過製程操作達到確保品質條件下節能的目的。
　　Formosa Petrochemical Corporation has two Methyl Tertiary Butyl Ether (MTBE) units. These units produce the high-value product MTBE，used for gasoline blending and as a chemical feedstock by reacting isobutylene present in the C₄ feedstock with methanol. This project focuses on the second MTBE unit (MTBE #2) and leverages historical process data for big-data analysis. By examining past operational experience and identifying equipment constraints, the goal is to optimize process operations and control. Through these process adjustments, we aim to conserve energy while maintaining product quality.

　　隨著鋰離子電池持續朝高能量密度發展，安全性議題逐漸浮現。澳洲電動車事故調查機構EV FireSafe指出近15年來有570起電動車火災案例，分析主因多為熱失控現象。熱失控一般是從負極側的固態電解質界面膜(Solid Electrolyte Interface, SEI)分解產生放熱反應觸發，經後續一系列連鎖反應，使溫度急升，最終起火爆炸。
　　本研究採用源自蝦蟹殼的幾丁聚醣包覆於軟碳或人工石墨負極，並利用戊二醛交聯提升膜的穩定性，這層預先包覆的人工固態電解質膜(Artificial Solid Electrolyte Interface, A-SEI)，可大幅降低84%以上SEI分解的放熱量，進而提升安全性。此外，半電池測試5C快充能力也可提升2.5%以上，5C充放循環壽命提升15%以上。為了提升電性，進一步將磺酸根接枝上幾丁聚醣，經磺酸化後的幾丁聚醣有助於鋰離子傳遞與均勻沉積，有效地延長循環壽命47%。接著，以正極NCM811搭配硫化-幾丁聚醣鍍膜的複合負極(軟碳10%+人工石墨90%)組成1Ah全電池進行SAE J2464穿刺實驗，鍍膜電池未起火爆炸燃燒，成功通過安全評估。藉由生質高分子修飾負極表面，不僅提升電池安全性，更同步提升倍率性能與循環壽命。此技術不僅符合綠色環保及循環再利用趨勢，亦為中油核心技術負極材料提供價值加乘，並降低中油未來充換電服務的運營風險，展現出與一般市售電池市場不同的競爭優勢。
　　With the continuous development of higher energy density in lithium-ion batteries, safety issues have been gradually emerging. An Australian electric vehicle accident investigation agency, EV FireSafe, reports 570 electric vehicle fires incidents in the past 15 years, with the main cause attributed to thermal runaway. Thermal runaway is typically triggered by the decomposition of the solid electrolyte interface (SEI) on the anode side, generating an exothermic reaction. This subsequent chain reaction causes a rapid temperature rise, ultimately leading to fire and explosion.
　　To mitigate this phenomenon, this study utilizes chitosan derived from shrimp and crab shells to coat a soft carbon or artificial graphite anode, and utilizes glutaraldehyde cross-linking to enhance the membrane's stability. This pre-coated artificial solid electrolyte interface (A-SEI) significantly reduces the heat released by SEI decomposition by over 84%, thereby improving safety. Furthermore, half-cell testing shows an improvement of over 2.5% in 5C fast charging capability and over 15% in 5C cycle life. To improve electrical properties, sulfonate groups are further grafted onto chitosan. The sulfonated chitosan facilitates lithium ion transfer and uniform deposition, effectively extending cycle life by 47%. A 1Ah full-cell composed of an NCM811 positive electrode and a sulfonated-chitosan-coated composite negative electrode (10% soft carbon + 90% artificial graphite) is then subjected to the SAE J2464 puncture test. The coated full-cell does not ignite or explode, successfully passing the safety assessment. Surface modification of negative electrode with biopolymer not only enhances battery safety but also improves rate capability and cycle life. This technology not only corresponds with the trends of green environmental protection and circular reuse, but also adds value to CPC's core technology of negative electrode materials, reduces the operational risks of CPC's future charging and swapping services, and demonstrates a competitive advantage different from the general and commercial battery market.

　　本研究旨在利用Chitinibacter tainanensis發酵黑水虻（Hermetia illucens）衍生成分，評估將蟲體中的幾丁質轉化為N-乙醯葡萄糖胺（N-acetylglucosamine, NAG）的可行性。實驗選用三種材料，包括脫脂蟲粉（A組）、全脂蟲粉（B組）及蟲皮（C組），並設計六組發酵配方（A4、A8、B4、B8、C4和C8），每組均添加100 mL BH培養基，於控制條件下發酵72小時，期間持續監測pH值與Brix值的變化，最後透過高效液相層析儀（HPLC）分析NAG生成量。結果顯示，除全脂蟲粉（B組）外，其餘組別在72小時後均有顯著生成NAG（A4: 4.33 g/L； C4: 3.97 g/L； C8: 2.93 g/L），其中A4與C4組的NAG轉化率皆超過20%。推測全脂蟲粉因油脂含量高而干擾菌株對幾丁質的分解效率。
　　本研究結果表明，Chitinibacter tainanensis可有效將黑水虻脫脂蟲粉與蟲皮的幾丁質轉化為NAG，且在液體比例較高的培養條件下，具有較佳的幾丁質水解與利用效率，為未來NAG的生產及新型態黑水虻飼料提供了新選擇與工業化應用參考。
　　This study aimed to evaluate the feasibility of using Chitinibacter tainanensis to ferment black soldier fly (Hermetia illucens) derived substrates for converting chitin in the insect body into N-acetylglucosamine (NAG). Three types of materials were selected for the experiment, including defatted larval powder (Group A), full-fat larval powder (Group B), and exoskeleton (Group C). Six fermentation formulations were designed (A4, A8, B4, B8, C4 and C8) using these types of materials. Each group was supplemented with 100 mL BH medium and fermented under controlled conditions for 72 hours, during which pH and Brix values were continuously monitored. NAG production was finally analyzed using high-performance liquid chromatography (HPLC).
　　The results showed that, except for the full-fat larval powder (Group B), all other groups exhibited significant NAG production after 72 hours (A4: 4.33 g/L； C4: 3.97 g/L； C8: 2.93 g/L). Among them, the conversion rates of chitin to NAG in groups A4 and C4 both above 20%, suggesting that the high lipid content in the full-fat larval powder interfered with the strain’s efficiency in chitin degradation. This study demonstrates that Chitinibacter tainanensis can effectively convert chitin from defatted larval powder and exoskeleton of black soldier flies into NAG, and higher liquid-to-solid ratios in the culture medium improve chitin hydrolysis and utilization efficiency. These findings provide a new substrate option and industrial reference for future NAG production.

　　為協助台灣中油公司油品行銷事業部業務單位快速進行台灣海運燃油訂價，本研究以機器學習貝氏(Bayesian)模型為核心建立船用燃油(MF-180)價格(以下簡稱CPC價格)模型，並與XGBoost(eXtreme Gradient Boosting)模型進行比較，以評估不同演算法之表現。
　　利用鄰近六港口(日本、南韓、香港、新加坡、上海及舟山)海運燃油價格歷史報價資料，並納入CPC價格前三日的滯後特徵(Lag Features)以增強時間序列訊息。透過均方根誤差(Root Mean Square Error, RMSE)、平均絕對誤差(Mean Absolute Error, MAE)與反正切絕對百分比誤差(Mean Arctangent Absolute Percentage Error, MAAPE)三項指標進行評估，結果顯示貝氏模型在準確性與穩定性上表現最佳，MAAPE僅0.88%，而XGBoost模型MAAPE為1.1%，且出現過度擬合現象，反映其泛化能力不足之情形。特徵重要性分析進一步揭示前一日CPC價格為最關鍵特徵，其次為舟山港與新加坡港報價，且在兩種模型中排序一致，顯示CPC價格高度自相關，並受特定港口價格影響。
　　此外，以全自動化流程實現資料定時擷取與更新，並透過Power BI儀表板呈現視覺化結果，供使用者能迅速掌握每日預測結果。以全年250個工作天計算，可節省約125小時人力，提升業務單位對海運燃油價格訂定的效率，降低人工彙整資料的負擔。
　　To assist the Marketing Business Division of CPC Corp., Taiwan in promptly pricing marine fuel, this study develops a Marine Residual Fuels 180 price forecasting model centered on a Bayesian model. The model’s performance is compared with that of the XGBoost (eXtreme Gradient Boosting) model to evaluate the effectiveness ossf different predictive approaches.
　　Historical price data from neighboring ports—Japan, South Korea, Hong Kong, Singapore, Shanghai, and Zhoushan—were utilized, alongside lag features incorporating CPC prices from the previous three days to enhance time-series representation. The model evaluation was conducted using three metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Arctangent Absolute Percentage Error (MAAPE). The Bayesian model outperformed in both accuracy and stability, achieving a MAAPE of only 0.88%, while the XGBoost model yielded a MAAPE of 1.1% and exhibited signs of overfitting, indicating weaker generalization capability.
　　Feature importance analysis further revealed that the previous day’s CPC price was the most influential factor, followed by prices at Zhoushan and Singapore ports. This ranking was consistent across both models, suggesting a strong autocorrelation in CPC prices and their susceptibility to specific port prices.
　　In addition, a fully automated process was developed to support scheduled acquisition and updating of neighboring port prices. Forecast results are visualized through the Power BI dashboard, enabling users to quickly grasp daily pricing insights. Assuming 250 working days per year, the solution yields an annual manpower saving of approximately 125 hours, thereby enhancing pricing efficiency, reducing the workload associated with manual data processing, and improves overall decision-making quality.