面對歐盟CBAM與國內碳費等碳定價壓力，現有研究針對製造業建構涵蓋組織盤查、產品碳足跡與製程優化之整合框架者仍有限。本研究以台灣某金屬零件製造商（A公司）為實證對象，依 ISO 14064-1 執行全廠溫室氣體盤查；依 ISO 14067「Gate-to-Gate」邊界量化20項核心銷歐產品之廠內製程碳排熱點；並透過物聯網（IoT）電力監測，比較傳統離散製程與多軸CNC加工之節能差異，結合作業基礎成本制（ABC）轉換節電效益為碳成本降低指標。 量化結果顯示，A公司2024年類別一與類別二排放量約10，079公噸CO₂e，其中92.03%來自外購電力。CNC加工作業平均貢獻65.29%之廠內製程碳排放，為核心熱點。實證顯示，多站離散作業整合至先進高階設備可削減非切削待機能耗：產品A改用8軸複合機台後單位碳排降46.96%，產品C導入雙主軸車床後降19.58%。然而，產品B升級8軸自動車床後，因基礎運轉功率遽增，單位碳排反升167.6%，揭示「能效陷阱」—高規格設備之非切削待機功耗，在特定條件下，可能抵銷乃至反轉效率優勢。 多情境財務模型估算，產品A與C之製程優化每年可節電約41.9萬度，並在納入電費、國內碳費與CBAM碳邊境成本之基準至高風險情境下，每年創造新臺幣233萬至273萬元之碳成本降低效益。本研究顯示，製造業節能關鍵在於消除非切削待機能耗，而非僅追求設備效率提升。企業應以碳排強度為核心建立機台指派策略，將法規遵循轉化為可量化之氣候財務韌性與成本減損能力。;In the face of intensifying carbon pricing pressures， such as the European Union′s Carbon Border Adjustment Mechanism (CBAM) and domestic carbon fees， existing research offers limited integrated frameworks for the manufacturing sector that simultaneously encompass organizational greenhouse gas (GHG) inventories， product carbon footprints (PCFs)， and manufacturing process optimization. This study empirically investigates a Taiwanese metal component manufacturer (Company A). A plant-wide GHG inventory was conducted in accordance with ISO 14064-1， and the in-plant process carbon emission hotspots of 20 core products exported to Europe were quantified using a "gate-to-gate" boundary under ISO 14067. Furthermore， Internet of Things (IoT) power monitoring was deployed to compare the energy-saving differentials between traditional discrete processes and multi-axis CNC machining. Finally， Activity-Based Costing (ABC) was integrated to convert electricity-saving benefits into carbon cost reduction indicators. The quantitative results indicate that Company A′s Category 1 and Category 2 emissions totaled approximately 10，079 metric tons of CO₂e in 2024， with purchased electricity accounting for 92.03% of the total. CNC machining operations contributed an average of 65.29% to the in-plant process carbon emissions， identifying it as the primary hotspot. Empirical evidence demonstrates that consolidating multi-station discrete operations into advanced high-end machinery can significantly mitigate non-cutting standby energy consumption; specifically， the carbon intensity of Product A decreased by 46.96% after transitioning to an 8-axis multi-tasking machine， while that of Product C fell by 19.58% upon implementing a twin-spindle lathe. Conversely， after upgrading the production of Product B to an 8-axis automatic lathe， its unit carbon emissions surged by 167.6% due to a drastic increase in base operating power. This phenomenon reveals an "energy efficiency trap，" wherein the high non-cutting standby power consumption of over-specified equipment can， under specific operational conditions， offset or even reverse the intended efficiency gains. Multi-scenario financial modeling estimates that the process optimizations for Products A and C can achieve annual electricity savings of approximately 419，000 kWh. Under scenarios ranging from baseline to high-risk—incorporating electricity tariffs， domestic carbon fees， and CBAM border costs—this reduction translates into an annual carbon cost reduction benefit of NT$2.33 million to NT$2.73 million. This study highlights that the key to energy conservation in manufacturing lies in eliminating non-cutting standby energy consumption rather than solely pursuing nominal equipment efficiency improvements. Enterprises should establish machine allocation strategies centered on carbon intensity， thereby transforming regulatory compliance into quantifiable climate financial resilience and cost-mitigation capabilities.
