108年石油事業傑出貢獻獎
得獎人：台塑石化公司總經理曹明

108年石油技術卓越獎
　　本學會為選拔及表揚在石油技術上有特殊成就者，特設置石油技術卓越獎。本獎項分為探採類、煉製類、石化類、天然氣類、營運類及其他類，每類各選一位，如當屆無適當候選者則該類從缺。
　　108年度各單位推薦人選均相當優秀，經本學會獎章委員會評審結果，通過8位人選特別在煉製類與石化類各增列一位，復經理監事聯席會議審議結果，無異議通過。

　　重油加氫脫硫工場(RDS)產出之低硫燃料的高流動點問題，不利於運送，導致客戶投訴。一般在工場操作上都知道觸媒活性會隨使用而降低，因而必須適時增加媒床溫度犧牲觸媒壽命來換取可接受的活性。然而，流動點異常的原因來源廣泛，在一些狀況下，仍然可以藉由不增加成本的條件下，而能解決問題。本研究透過貝氏網絡分析高流動點異常，並依此改善操作過程，以利於產出高品質的低硫燃料油。首先，本文描述RDS工場的生產過程和貝氏網絡分析法。其次，收集大林煉油廠第三RDS工場的現場操作數據，結合專家知識構建流動點貝氏網絡分析模型，並藉由Netica由實際數據中實現其貝氏網路的條件機率數值，以推論RDS流動點異常的原因。最後，透過貝氏網絡的敏感性分析發現，在使用C觸媒情況下，低硫燃油流動點異常時，最有可能的原因可能是媒床溫度設定不足。此外，影響分析發現，在極差操作環境下，使用觸媒C不合格流動點比率高達87.5%。在最佳操作條件下，選用觸媒B不合格流動點比率則降為0.37%。本研究發現，在不影響觸媒壽命的情況下，稍微提升一點媒床溫度，配上本研究所建議的操作條件，適當降低進料流動點，減少輕質成品的取出比例，可以在不增加成本的條件下，穩定低硫燃油的流動點。本研究的成果可以提供油品製程異常監控，並幫助進一步的深入分析與應用。

　　The transportation problems caused by the high pour point of low-sulfur fuel produced in RDS leads to considerable customers’ complaints. However, the reason for the pour point anomaly is still inconclusive. The paper analyzes the anomalies of the pour point by Bayesian Network and optimizes the strategy of operation in order to produce high quality low-sulfur fuels oil. Firstly, the features of RDS process are described. Secondly, we collect recent 4 cycles of operational data from KOR#3 RDS Talin Refinery and build a Bayesian Network Model based on expert knowledge. Next, this paper illustrates the Bayesian network-based methodology and realizes the complete Bayesian Network Analysis by Netica. Then, the preliminary problem of the pour point anomaly in the RDS unit is discussed. Finally, this study tackles sensitivity analysis for pour point anomaly. It is found that the temperature of catalyst bed affects the pour point of low-sulfur fuel most, followed by the catalyst characteristics. In addition, the high pour point generated by the catalyst C (87.5%) is much higher than that done by the catalyst B (0.37%). Research findings help to optimize process which could ensure oil quality and future research.

　　近年來，大數據在各領域火熱發展，尤其如電子商務數據支撐了快速大量交易，電商購物節屢屢刷新紀錄，數據的大量膨脹，早已經超越「人工」可以處理的境界。大量繁務的工作將借重可自行學習、理解和遇到新情況時決定如何反應的人工智慧(AI)來幫忙，盡可能與人類一樣的思考邏輯與行為模式。
　　然而，AI可以解答所有問題嗎?石化產業適合應用AI嗎?石化產業要從哪一個項目優先著手做起?石化產業一直以來為民生產業不可或缺的上游產業，在科技尚未找尋到可替換石化材料的現在，石化產業對於上、中、下游帶動的效果不可小覷，並於相關產業鏈扮演重要角色。但在全球節能減碳、綠色商品、環保意識高漲的趨勢下，全球石化產業近年來面臨市場需求與消費型態改變、數位資訊驅動新型態的商業模式等挑戰，數位轉型、大數據與AI在石化產業應用可否為產業帶來新契機，沉睡已久的巨人在AI的浪潮下甦醒。

　　For the past few years, big data analytics flourish in various fields, especially in e-commerce, since it could sustain massive and rapid transactions. The volume of data has exceeded the ability that human could handle； hence, numerous and tedious works could rely on Artificial Intelligence (AI) since they contain familiar logic patterns as humans behave.
　　However, AI might not be the panacea for all industry. The modern technology has not yet discovered the alternative material of petrochemical； therefore, the petrochemical industry plays an essential role for all livelihood firm. Under the rising awareness of environmental, energy efficiency and carbon reduction； the petrochemical business are facing two challenges, changes in consumption pattern and new business models launched by digital information. New opportunity for the petrochemical industry could be created via the digital transformation, big data and AI, which causes the sleeping giant to awake.

　　第四次工業革命之後，各種工業生產技術的發展相當迅速。在物聯網（Internet of Things, IoT）的技術基礎下，工業界發展出了許多的新興應用，也誕生了許多新穎的技術名詞，例如：大數據（Big Data）、雲計算（Cloud Computing）、邊緣運算（Edge Computing）、工業智能（Industrial Artificial Intelligence, Industrial AI）等。這些技術大部分應用在機台產業，化工產業要如何找到適合的技術導入？本文以人工智慧的發展史為出發點，簡單說明化工系統與一般機台產業的差異，並以相關文獻及案例介紹在化工產業可能的相關運用，例如：專家系統運用在高爐冶煉將鐵礦石還原成生鐵、運用數據分析方法判斷氣體洩漏的特徵等。相信可以為化工產業帶來新的應用方向。

　　In the fourth industrial revolution, production technology has developed quite rapidly. Based on the technology of the internet of things (IoT), the new applications in the industries are increasingly developing. Recently, lots of new technical terms such as big data, cloud computing, edge computing, or industrial AI (Artificial Intelligence) have been created and most of them are used in machine industry. Another issue is how to find the suitable technologies to be used in chemical industry.
　　Firstly, this article introduces the development history of artificial intelligence. Then we explain the difference between the chemical system and the general machine industry. Last, we will introduce some applications in the chemical industry according to relevant publications and cases. For example, expert systems have been applied in the blast furnace of the steel industry, data analysis methods have been used to predict the characteristics of gas leaks. This article try to propose some development directions for the chemical industry.

　　近年來，隨著硬體設備和演算法的大幅發展，人工智慧(AI)在人類日常生活已逐漸具體實現。近期科學家期望的透過人工智慧演算法的導入，使一些工作可利用機器去完成，提昇效率及降低成本，在石油產業中約可增加79%的附加價值。
　　現階段國際石油公司主要的兩種人工智慧應用模式：分別是「智能機器人(intelligent robot)」和「虛擬助手(virtual assistant)」。埃克森美孚(ExxonMobil)從2016到2021這五年內在此領域預計投入2,500萬美元，並與麻省理工學院 (Massachusetts Institute of Technology, MIT)合作，設計運用於海洋探勘的智能機器人；道達爾石油公司（Total）宣布與谷歌雲(Google Cloud)達成協議，共同開發人工智慧井下資料分析系統，以改善探勘流程。
　　台灣中油公司希望透過人工智慧的導入，增進產能與強化工安，因此，中油探採研究所積極邀請資料科學家蒞所指導與演講，同仁對於人工智慧的應用已有基礎的認識，探研所已研擬數個人工智慧相關專案，配合進階的教育訓練課程，同仁們更加了解人工智慧的應用與發展。並藉由資料科學家對人工智慧的軟硬體需求評估及規劃，預期在未來相關的計畫，能夠有顯著的助益。

　　With the rapid advance of hardware and algorithms, recently artificial intelligence (AI) has been gradually applied to our daily life. There are several stages in AI development history. At this stage, scientists hope to use high-level AI algorithms to enable computers to perform some tasks, previously done by man. Such technologies have been proven to significantly increase work efficiency and reduce costs, with an average adding value of approximately 79% in the petroleum industry.
　　Today, the most popular AI applications in the international oil companies appear to be "Intelligent Robots" and "Virtual Assistants", such as ExxonMobil, investing 25 million dollars in this field, has announced that it will work with Massachusetts Institute of Technology (MIT) to design AI robots for ocean exploration；Total Oil announced an agreement with Google Cloud to jointly develop an artificial intelligence system for well log data analysis to improve the exploration process.
　　CPC Corporation, Taiwan hopes to increase production capacity and industrial safety through the introduction of AI and Big Data Technology. As a result, the Exploration and Development Research Institute (EDRI), CPC Corporation, Taiwan hopes that colleagues will have a basic understanding of the application of AI through speeches by the invited data scientists. Now, EDRI has proposed several projects which are related to AI. We hope that through more comprehensive AI education training courses, EDRI’s personnel will have a better understanding of the application and development of AI. Besides, we also hope that through the software and hardware requirement evaluation of the application of AI, the future EDRI’s AI-related projects can be carried out more smoothly.

　　工業4.0(Industry 4.0)帶動智慧製造(Smart Manufacturing)的浪潮，以工業物聯網(IoT)及人工智慧(AI)實現(AIoT)之智慧工廠，提供感應器數據收集連動，優化生產效能，或在異常發生前提早因應，增加生產效益。文中主要是利用台塑FORMOSA分散式製程控制系統(Distribution Control System)，搭配RtWX工業物聯網(Industrial Internet of Thing)，結合人工智慧(Artificial Intelligence)，以智慧物聯網(AI+IIoT=AIoT)，達成人機協作的數位分身(Digital Twins)，實現虛實整合(Cyber Physical Systems)的智慧製造。

　　In Industry 4.0, smart manufacturing has been called to leverage industrial internet of thing (IIoT) and artificial intelligence (AI) to fulfill smart factory where collecting sensor data to optimize production or prevention failure. This paper describes the FORMOSA Distribution Control System (DCS) cooperates with RtWX Industrial Internet of Thing (IIoT) which combines Artificial Intelligence (AI) using the AIoT (e.g. AI+IoT) technology to realize the digital twins of the cyber physical systems in smart manufacturing.

　　人工智慧是以機器模擬人類智慧的思考過程，尤其是利用電腦系統。它運機器學習去解決複雜的人類問題，並可提升人類生活品質，例如人工智慧的運用在一些領域如醫療、交通、飲食、娛樂及天氣預測等。
　　台灣中油公司行銷事業部共有22座供油服務中心及650餘座加油站，並有油罐車運輸油品，可提供快速、準確、親切的加油服務。隨著AI時代的來臨，將可強化加油服務，確保市場占有率、降低成本、追求成長，增加利潤及提高績效。
　　AI技術在「油品行銷及輸儲」整合上的運用，如「設備資訊監控系統」使用在供油中心油槽監控、管線監控及加油站油槽監控與派車，「門禁管理(車牌辨識)系統」使用在員工車輛、油罐車管理、承攬商管理及加油站管理，「油罐車運途管理系統」使用在人臉偵測、防疲勞駕駛、車道偏離預警暨前方碰撞預警系統及GPS定位，「車隊卡系統」針對車隊消費模式進行分析等，都是運用AI技術強化油品行銷及輸儲。

　　AI (Artificial Intelligence) is the simulation of human intelligence processes by machines, especially computer systems. It uses machine learning to solve complex human problems. It could better the quality of human life in various fields such as its application in medical care, transportation, food, entertainment or weather forecast.
　　There are 22 oil supply centers, 650 plus gas stations and many tank trucks in the Marketing Division, CPC Corporation, Taiwan. For past several decades, CPC has been providing customers with excellent service of fuel refilling in Taiwan. With the fast development of AI, the gas stations’ service improves day after day. The AI technology can help CPC to raise the share of marketing, lower cost, strengthen growth, increase profit and better the performance.
　　The applications of AI technology in oil products marketing, transportation & storage are very common. The “Device Information Monitoring System" in the oil supply center is used to monitor oil tank of the center, transportation pipeline, oil tank of gas station and arrangement of tank trucks. The "Access Control (License Plate Recognition) System” is used to identify staff’s vehicles, arrangement of oil tank truck, management of contractor and operation of gas station. The “Tank Truck Transit Management System” is used in face identification, fatigue-driving prevention, lane departure warning and forward collision warning and GPS positioning. The "Fleet Card System" is used to analyze the behavior of the fleet consumption. The AI technology could improve the marketing, transportation and storage.

　　人工智慧（AI）已成為全球科技發展趨勢最熱門關鍵技術之一。由於這一波AI的核心是機器學習，驅動機器學習的核心是充足的資料，有充足的資料才能學習。有效的資料量越多、分析更精準，AI學習成效才會更好。從這個觀點來看，台灣最有機會發展AI應用的是製造業、醫療及金融等三大領域。分析台灣產業結構，擁有資料最充足的首推製造業，台灣製造業的群體廣泛、產業鏈完整，相對其他領域，更具備充足的資料量來源，可滿足AI機器學習及應用開發。
　　本年專題論壇首先闡述當前石油產業面臨的關鍵問題，包括油氣產業未來走向、國際原油價格展望、IMO 2020之衝擊分析、全球電動車未來發展、煉油技術發生重大轉變、天然氣供需趨勢與挑戰、5G 的技術挑戰與機會等要項。其次，探討大數據與人工智慧技術在石油產業的應用，包括國際油公司在上、中、下游的應用範疇，並以其實例作為國內的借鏡，期能促進石油產業開發技術升級換代，最後研析台灣石油公司在引進人工智慧與大數據應用於探採、煉製、石化、輸儲與行銷面向的成效。