一、本學會配合115年專題論壇之舉行，將辦理優秀論文徵選，研究主題為「全球油氣供應鏈重整的新格局與挑戰」，分能源政策與規劃、探勘開發、煉製與石化、天然氣產業、新及再生能源、營運管理等與主題相關之理論、實證與應用研究，中英文學術論文與應用報告皆可。
二、本年優秀論文徵選不接受已經刊載過之文稿，惟於研討會發表過之論文，經適當增補其內容者為例外。
三、本年優秀論文徵選投稿截止期限為115年8月31日，預計於10月下旬公布優秀論文得獎名單，得獎論文作者將於11月舉辦之專題論壇接受頒獎，論文擇優者刊登於石油季刊。論文徵選審查將邀請台灣產、官、學、研界專家學者組成評審委員會進行匿名審查工作，決定獲獎名次。
四、論文評選依能源政策與規劃、探勘開發、煉製與石化、天然氣產業、新及再生能源、營運管理等類進行評比，優勝者頒發獎座、獎牌及獎金：
(一) 獎項說明如下：
1.台灣中油特優論文獎：預計由各類論文中選出 1 篇為原則，頒發台灣中油特優論文獎獎座乙座及獎金新臺幣叁萬元。
2.金獎：預計各類選出 1篇為原則，頒發金獎獎牌乙面及獎金新臺幣貳萬元。
3.銀獎：預計各類選出 2篇為原則，頒發銀獎獎牌乙面及獎金新臺幣壹萬貳千元。
4.佳作：預計各類選出 2篇為原則，頒發佳作獎牌乙面及獎金新臺幣捌千元。
(二) 各獎項若任一類未達得獎標準，名額則可從缺或不足額錄取，其缺額得由評審會議建議，經理監事會議議決後可彈性調整至其他類使用。
五、投稿電子郵件信箱地址為cpi.org@msa.hinet.net，請以電子郵件附件方式寄送稿件之WORD檔及PDF檔各一份。為便利編審作業，請於稿件PDF檔中移除作者之相關資料 (如姓名、服務機關、電話及電子郵件信箱等)。
六、投稿時請附上「生成式人工智慧（AI）使用聲明書」(格式如後附)。
七、徵稿格式說明請參閱石油季刊最後一頁稿約說明，若有任何疑問，請電洽 02-28201255 石油季刊編輯部李小姐。

中國石油學會115年優秀論文徵選稿件
生成式人工智慧（AI）使用聲明書
本人（或本研究團隊）謹此聲明，於本次投稿名稱《_______________________________》撰寫與研究過程中，對於生成式人工智慧（如 ChatGPT、Gemini、Claude、Copilot 等工具）之使用情形，請勾選下述適用項目：
☐ 本文未使用任何生成式人工智慧工具於資料蒐集、資料分析、內容撰寫、語言潤飾或其他任何研究與寫作過程中。
☐ 本文部分使用生成式人工智慧工具。
☐ 本文廣泛使用生成式人工智慧工具，所有生成內容皆經作者審核，並確保無侵犯智慧財產權或違反學術倫理之情事。

作者姓名（簽名）：__________________________
聯絡信箱：_________________________________
日 期：_________年_________月_________日

　　隨著美國和以色列在2026年2月28日無預警對伊朗展開攻擊，德黑蘭當局幾乎完全封鎖荷姆茲海峽的通行，該海峽是全球約5分之1的石油和天然氣定期運輸的通道，海峽的封鎖已導致全球能源價格飆升。中東局勢持續動盪，對全球能源及石化供應鏈成嚴重衝擊，石化原料、成品及相關衍生物缺貨及漲價頻傳，不論食、衣、住、行及科技產業均受到影響。
　　石油被譽為現代工業的血液，對於缺乏自有天然資源的臺灣而言，能源的穩定供應更是國家安全與經濟發展的命脈。臺灣高達90%以上的原油與石油產品依賴進口，這些能源的進口不僅是一個單純的商業採購問題，更牽涉到複雜的地緣政治、國際航運風險以及全球市場的價格波動。臺灣原油與天然氣的進口來源經過多年的戰略調整，目前已形成相對穩定的多元化供應版圖，主要由國營的台灣中油與民營的台塑石供負責採購與煉製。面對全球能源市場的重整與低碳轉型浪潮，本報告將深入剖析臺灣原油與關鍵石油產品其油氣供應鏈的結構與策略、天然氣的進口來源與安全存量、國內兩大石油巨頭的採購策略，以及近期備受國際關注的能源供應鏈隱憂，同時聚焦於政策調整、產業韌性與永續發展，並探討如何在全球油氣供應鏈重塑中強化臺灣低碳能源轉型的穩定性與競爭力。
　　Following the unexpected attack by the United States and Israel on Iran on February 28, 2026, Tehran virtually blocked the Strait of Hormuz, a vital passage for approximately one-fifth of the world's regular oil and gas transport. This move caused global energy prices to soar. The ongoing instability in the Middle East has severely impacted the global energy and petrochemical supply chains, leading to frequent shortages and price increases of petrochemical raw materials, finished products, and related derivatives, affecting industries across food, clothing, housing, transportation, and technology.
　　Oil is considered the lifeblood of modern industry. For Taiwan, which lacks its own natural resources, a stable energy supply is crucial for national security and economic development. Taiwan relies on imports for over 90% of its crude oil and petroleum products. This is not merely a simple commercial procurement issue but also involves complex geopolitical factors, international shipping risks, and global market price fluctuations. This report will delve into the challenges Taiwan faces in the global energy market restructuring and low-carbon transition for crude oil and key petroleum products. Taiwan must re-examine the structure and strategies of its oil and gas supply chain, its natural gas import sources, the procurement strategies of its two major oil giants, and the recent international concerns regarding energy supply chain vulnerabilities. After years of strategic adjustments, Taiwan's crude oil and natural gas import sources have formed a relatively stable and diversified supply structure, primarily procured and refined by the state-owned CPC Corporation and the private Formosa Petrochemical Corporation. The report also focuses on policy adjustments, industry resilience, and sustainable development, exploring how to strengthen the stability and competitiveness of Taiwan's low-carbon energy transition in the reshaping of the global oil and gas supply chain.

　　位於非洲之O油田自民國109年2月投產，截至114年12月已累計生產原油16.8百萬桶。生產數據表明，O油田目前已進入產能衰退期，面對原油產量遞減，透過擴大油源以維持油田產能已成當務之急。本研究之契機在於114年新鑽鑿之M-16井於原被忽略的Ma地層M1砂層發現輕質原油，為油田增產帶來新契機。
　　本計畫整合114年鑽鑿之M-14、15及16三口新井資料，完成M構造構造地質模式更新，其中新發現之Ma地層M1砂層貢獻約21.4百萬桶。
　　Ma地層M1砂層砂體呈透鏡狀分布，側向連續性較差。鑑於其連通性存在不確定性，本研究分別設定M1砂層全區連通、西側獨立及僅西側高區含油等三種假設情境，進行水平井生產與換層生產之數值模擬。結果顯示，在西側獨立的情境下，水平井之累積產量預測為82.5萬桶，顯著優於換層生產的37.2萬桶。綜合評估，水平井為維持產能峰值與加速回收之最佳方案；惟考量地質風險，建議在正式鑽探水平井前，透過試油氣作業釐清M1砂層之實際油氣範圍，以降低開發風險。
　　The O Oil Field, located in Africa., It has reached a cumulative crude oil production of 16.8 million barrels in December 2025. The catalyst for this study was the discovery of light crude oil in the previously overlooked M1 sand of the Ma formation by the M-16 well drilled in 2025, which presents a new opportunity for production enhancement.
　　This project integrates data from three new wells drilled in 2025 （M-14, M-15, and M-16） to update the structural geological model of the M structure. The updated STOIIP for the newly discovered M1 sand of the Ma formation contributing approximately 21.4 million barrels.
　　The results indicate that under the "independent western sector" scenario, the predicted cumulative production for a horizontal well is 825,000 barrels, significantly outperforming the 372,000 barrels yielded by recompletion.
　　A comprehensive evaluation suggests that horizontal drilling is the optimal strategy for maintaining peak productivity and accelerating recovery. However, in light of geological risks, it is recommended that well testing operations be conducted prior to the formal drilling of horizontal wells to clarify the actual hydrocarbon extent of the M1 sand and mitigate development risks.

　　當前全球正面臨由溫室效應所引發的暖化危機，而二氧化碳作為主要溫室氣體，如何降低其排放與實現資源化利用已成為永續發展的重要課題。隨著碳中和與循環經濟議題的推動，台灣中油積極投入碳捕集、利用與封存技術（Carbon capture, utilization and storage, CCUS）的研究，以因應能源轉型與環境需求。傳統聚氨酯材料雖應用廣泛，但其合成高度依賴有毒異氰酸酯，對環境與人類健康構成風險，且已受到REACH與RoHS等國際法規的嚴格限制。在此背景下，本研究開發一種環保固碳複合材料，利用CO₂與環氧化物反應生成環碳酸酯，再與二胺聚合製得非異氰酸酯聚氨酯（Non-Isocyanate Polyurethane, NIPU）。該材料具有三大綠色化學特色：第一個是避免使用異氰酸酯，降低毒性風險並提升製程安全；其次為直接利用CO₂作為原料，符合碳捕集與利用的永續理念；最後，反應效率高且無副產物，符合原子經濟性原則。此外，NIPU的β-羥基氨基甲酸酯結構可與環氧樹脂形成新型複合網絡，不僅增強對金屬與陶瓷等極性基材的附著力與機械強度，拓展其在塗料與黏著劑的應用潛力，亦賦予材料自我修復等增益特性，有助於延長使用壽命並減緩環境衝擊。
　　The world is currently facing a severe warming crisis driven by the greenhouse effect, with carbon dioxide as the primary greenhouse gas. Reducing its emissions and achieving resource utilization have become critical challenges for sustainable development. In response to the global push for carbon neutrality and the circular economy, CPC Corporation, Taiwan, has actively invested in research on carbon capture, utilization, and storage (CCUS) to address both energy transition and environmental demands. Although conventional polyurethane materials are widely used, their synthesis relies heavily on toxic isocyanates, posing risks to human health and the environment, and they are increasingly restricted by international regulations such as REACH and RoHS. Against this backdrop, this study develops an environmentally friendly carbon-fixing composite material by converting CO₂ and epoxides into cyclic carbonates, which are subsequently polymerized with diamines to produce non-isocyanate polyurethanes (NIPUs) and further combined with epoxy resins. This material embodies three major principles of green chemistry: first, it eliminates the use of isocyanates, thereby reducing toxicity and enhancing process safety； second, it directly utilizes CO₂ as a raw material, aligning with the concept of CCU for sustainability； and third, it achieves high reaction efficiency with no byproducts, consistent with the principle of atom economy. Moreover, the β-hydroxyurethane structure of NIPU enables the formation of novel hybrid networks with epoxy resins, which not only enhance adhesion and mechanical strength on polar substrates such as metals and ceramics but also expand potential applications in coatings and adhesives. In addition, this structure imparts self-healing and other functional properties, enabling recyclability and reusability, thereby extending products’lifetime and mitigating environmental impact.

　　2050年淨零碳排已成為全球目標，各國積極發展再生能源、儲能與綠氫技術。本研究整合太陽能發電、釩氧化還原液流電池（VRFB）儲能及陰離子交換膜（AEM）水電解產氫，作為實現淨零排放的可行策略。由於液流電池與電解系統在膜材設計上具相似性，具高度整合潛力。
　　VRFB具有良好電化學可逆性、長循環壽命及高效率。本研究針對聚苯咪唑（PBI）與磺化聚醯亞胺（SPI）膜進行改質。結果顯示，摻雜共價有機框架（COF）的PBI複合膜可提升質子導電度並降低釩離子滲透性，在40–100 mA/cm²下展現優異電壓、庫倫及能量效率，表現優於Nafion。
　　SPI膜透過引入脂環二胺（TCDDA）進行共聚，提升微相分離與自由體積，使釩離子通透率降至9.79×10⁻⁸ cm²/min，庫倫效率超過90%。AEM方面，利用降冰烯衍生物製備不同交聯膜並探討其電解性能，具低溫操作、無需貴金屬與高壽命優勢。
　　本研究成功整合儲能與電解系統，將間歇性太陽能轉為可長期儲存之綠氫。未來將透過化學交聯抑制膜材膨潤，在維持導電度下提升穩定性，以發展高效能PV/VRFB/AEM整合系統。
　　Achieving net-zero carbon emissions by 2050 has become a global goal, driving the devel-opment of renewable energy, energy storage, and green hydrogen technologies. This study inte-grates photovoltaic (PV) power generation, Vanadium Redox Flow Battery (VRFB) energy stor-age, and Anion Exchange Membrane (AEM) water electrolysis for hydrogen production as a fea-sible strategy toward net-zero emissions. The similar membrane requirements of VRFB and AEM systems provide significant potential for integration. For VRFB applications, polybenzimidazole (PBI) and sulfonated polyimide (SPI) membranes were modified and evaluated. COF-incorpo-rated PBI composite membranes exhibited enhanced proton conductivity and reduced vanadium ion permeability, delivering superior voltage, coulombic, and energy efficiencies at 40–100 mA cm⁻²compared with Nafion membranes. In addition, TCDDA-modified SPI membranes improved microphase separation and free volume, reducing vanadium permeability to 9.79 × 10⁻⁸ cm²min⁻¹ while maintaining coulombic efficiencies above 90%. For AEM water electrolysis, crosslinked membranes based on norbornene derivatives were prepared and investigated for hydrogen pro-duction performance. The system offers advantages including low-temperature operation, the elimination of precious metal catalysts, and long service life. This work successfully integrates energy storage and electrolysis systems to convert intermittent solar energy into storable green hydrogen. Future efforts will focus on suppressing membrane swelling through chemical cross-linking to enhance stability while maintaining conductivity, thereby advancing the development of high-performance PV/VRFB/AEM integrated energy systems.

　　本研究以生質物料為碳源製備硬碳，經高溫碳化與酸洗處理後，獲得適合用於鋰/鈉離子電池負極的材料。將生質硬碳導入鋰離子電池之測試結果顯示，首次充放電具穩定的庫倫效率；在 1.0–3.0C 的倍率條件下，容量保持率超過95%，恢復至0.5C時容量回復率達99.7%。長循環測試顯示，經1,000次充放電後容量保持率為92%，庫倫效率穩定於99.6%以上。在鋰離子電池製作方面，本研究製備生質硬碳–磷酸鋰鐵電池，容量約19Ah，電池內阻低於2.0mΩ，顯示批次間一致性良好。進一步組裝的4S14P之電池模組搭配電池管理系統，於0.2C充放電條件下，能量轉換效率超過98%。而由模組、控制系統與軟體整合成的儲能櫃系統進行電力調控測試，系統能於離峰用電時段充電及儲電，並於尖峰用電時段支援放電，具有移峰填谷的效果；並測試不斷電功能，儲能櫃系統能在電網電力中斷後10毫秒內切換至電池模組供電，以確保負載持續運作，維持供電不中斷。綜合測試結果，生質硬碳與其低碳製程，能滿足鋰離子電池負極在容量保持率、倍率性能及長循環壽命上的需求，並經電池模組與儲能櫃驗證，生質硬碳具備應用於電動載具及電網級儲能系統的可行性。
　　Biomass-derived hard carbon was synthesized from agricultural biomass through high-temperature carbonization and acid washing, and evaluated as an anode material for lithium- and sodium-ion batteries. In lithium-ion cells, the material exhibited stable coulombic efficiency during initial cycling. At charge–discharge rates of 1.0–3.0C, capacity retention exceeded 95%, and when returned to 0.5C the capacity recovered to 99.7%. In durability testing, the capacity retention was 92% with coulombic efficiency consistently above 99.6% after 1,000 cycles at 1C.
　　Full-cell tests were performed using biomass-derived hard carbon–LiFePO₄ cells, with an average capacity of ~19Ah per cell and internal resistance below 2.0 mΩ, indicating good batch-to-batch consistency. A 4S14P module was then assembled and integrated with a dual-balancing battery management system. The module provided stable output under high-current operation and achieved an energy conversion efficiency above 98% at 0.2C. For system-level validation, the module was incorporated into an energy storage cabinet with control hardware and software. The cabinet managed peak-load discharges and switched to battery backup within 10 ms during grid interruptions, ensuring continuous operation.
　　These results demonstrate that biomass-derived hard carbon, with its renewable origin and low-carbon processing, satisfies the requirements for capacity retention, rate capability, and long-term cycling stability. Validation at both module and system levels confirms its suitability for electric vehicles and grid-scale energy storage systems.

　　本研究主要是透過人工智慧(artificial intelligence，簡稱AI)模型，以物件偵測(object detection)技術針對台灣中油股份有限公司自有的太陽光電系統進行影像辨識及分析，其中利用無人機蒐集AI模型在訓練階段僅需要的紅外線(IR)熱影像資料。我們藉由批次樣本數(batch size)為3筆資料且經30個訓練週期(epoch)訓練所產生的AI模型，成功地就資料集裡的IR熱影像照片，分辨出太陽能板陣列和熱斑缺陷的類別與位置，尤其該AI模型對於後者重點物件的辨識能力指標分數(F1 score)高達0.95。本研究進一步設計出一個簡易型的使用者介面(user interface)系統，其中整合具有物件偵測功能之AI模型，據以輔助此AI模型在影像辨識工作上之執行及分析。
　　In this study, the object detection towards solar photovoltaic systems of CPC Corporation, Taiwan is accomplished using artificial intelligence (AI) models, merely trained by employing relevant infrared (IR) images. Especially, unmanned aerial vehicles are used to collect such images. Batch size and epochs of 3 and 30, respectively, are prepared for the AI model training. Thus, this AI model can successfully recognize solar panels and hot spot defects from the datasets of IR images. A high F1 score of 0.95 is given by the AI model based on the detection for the defects. Moreover, a user interface system is further established to assist AI models with the image recognition and analysis in this study.