環保專案成果報告查詢系統

專案計畫編號
經費年度	112
計畫經費	11,100 千元
專案開始日期	2023/03/10
專案結束日期	2024/12/05
專案主持人	黃致霖
主辦單位	嘉義縣環境保護局
承辦人	侯雅馨
執行單位	思維環境科技有限公司
專案分類	大氣空氣
中文關鍵字	空氣品質感測器、多元智慧監控
英文關鍵字	air quality sensors
協同主持人	江嘉凌；鐘文舜
共同主持人
計畫聯絡人	劉栩彤
計畫聯絡電話	04-23141323
計畫聯絡信箱	m131393@greenideas.com.tw

中文摘要	本計畫推動工作包含於嘉義縣轄內300台空氣品質感測器數據品質管理及維運工作，透過群集、衰減、動態校正作業瞭解感測器設備狀況、亦運用空品感測數據與空氣品質物聯網資訊進行結合，分析本縣空污熱區，釐清可疑污染對象，統計期中報告執行期間自112年3月10日至113年12月31日之工作執行成果說明如下：本計畫針對感測器點位進行最適化規劃，並完成12個點位的調整，完成18個行政區感測器覆蓋率達100%。其次，建置「空品數據中心」系統，提供數據租用服務，並辦理相關教育訓練。感測器巡查作業每半年執行一次，目前已完成4次巡檢，以及每季執行一次10%感測器巡檢比對作業，已完成8件次皆全數符合環境部現場比對標準。而每月進行群集分析、衰減分析以及動態校正，皆完成22次分析作業，針對有異常或離群設備即刻派員現場確認是否有污染源或是更換故障設備；分析結果顯示於空品良季（4-9月）因環境濃度較低，故測站內設備與環境濃度趨勢變化差異較大，在空品不良季改善許多。此外，針對感測器故障進行檢修，因感測元件使用壽命介於3-5年，本縣於110年全面更換國產感測器至今已有3年，故112年感測主機板異常及PM2.5元件異常佔比較高為27.7%，本工作團隊即時派員更換元件後，相比113年故障率已明顯下降。本計畫透過感測器數據及污染告警系統結合，於LINE進行「污染發生空品感測器編號、污染發生時間、初步判斷污染源、產生污染事件警告、污染持續時間、事件平均值、事件最高值」，分析該次告警可疑污染源，並通報相關計畫前往稽查。在112年至113年運用告警機制，分析現場巡查32件次，其中29件次現場有明顯污染來源，告警準確度為90.1%；現場巡查之可疑污染來源分別為露燃44.8%、工地揚塵為24.1%、工廠逸散為13.8%、移動源及其它分別為13.8%、3.4%。感測器數據分析報告已完成8件次，提供之可疑污染源名單，或現場巡查發現污染排放情事之彙整及通報後，經各相關管制計畫前往稽查，統計至113年12月31日，合計處分案件共11件次，其中露天燃燒共4件、工廠違規6件，工地揚塵1件，處分金額合計新臺幣362萬8,750元。本計畫針對敏感族群如老人及孩童為對象，已完成辦理5場空氣品質及污染認知宣導會，參加人數超過1,600人。在數據上傳品質部分，感測器資料上傳完整率達97.1%，顯示數據傳輸穩定度佳。為提升工作人員專業知能，以及針對各項工作內容依據實際執行現況提出後續精進作為，使本計畫順利推動如期達成環保局交辦之相關業務，邀請專家學者以感測器於輔助環境治理為方向，已完成辦理4場次之教育訓練。於多元智慧監控部分，本工作團隊針對常受鄰近居民陳情之重點工業區周圍架設高值啟動自動影像系統，於112年10月8日監測上O資源管道爆炸觸發啟動1件次，管制計畫進廠稽查發現該廠因進行鍋爐氣電共生程序作業致流體化床鍋爐（E501）發生爆炸，產生大量黑色濃煙散布於大氣中，故罰鍰新臺幣100,000元整。針對常受鄰近居民陳情之重點工業區或陳情熱點，如民眾異味陳情、畜牧場異味陳情及縣市周界異味陳情周圍架設空氣異味微型感測器，統計113年12月31日前已執行10點次異味監測，同時監測輔導改善前後空品狀況。其中永O畜牧場，管制計畫於分析異味熱點時段進場確實有發現業者因每日僅於上午清理1次排泄物，導致夜間累積異味較重已請業者改善，後續分析異味監測數據已明顯下降。
中文關鍵字	空氣品質感測器、多元智慧監控

中文摘要

本計畫推動工作包含於嘉義縣轄內300台空氣品質感測器數據品質管理及維運工作，透過群集、衰減、動態校正作業瞭解感測器設備狀況、亦運用空品感測數據與空氣品質物聯網資訊進行結合，分析本縣空污熱區，釐清可疑污染對象，統計期中報告執行期間自112年3月10日至113年12月31日之工作執行成果說明如下：本計畫針對感測器點位進行最適化規劃，並完成12個點位的調整，完成18個行政區感測器覆蓋率達100%。其次，建置「空品數據中心」系統，提供數據租用服務，並辦理相關教育訓練。感測器巡查作業每半年執行一次，目前已完成4次巡檢，以及每季執行一次10%感測器巡檢比對作業，已完成8件次皆全數符合環境部現場比對標準。而每月進行群集分析、衰減分析以及動態校正，皆完成22次分析作業，針對有異常或離群設備即刻派員現場確認是否有污染源或是更換故障設備；分析結果顯示於空品良季（4-9月）因環境濃度較低，故測站內設備與環境濃度趨勢變化差異較大，在空品不良季改善許多。此外，針對感測器故障進行檢修，因感測元件使用壽命介於3-5年，本縣於110年全面更換國產感測器至今已有3年，故112年感測主機板異常及PM2.5元件異常佔比較高為27.7%，本工作團隊即時派員更換元件後，相比113年故障率已明顯下降。本計畫透過感測器數據及污染告警系統結合，於LINE進行「污染發生空品感測器編號、污染發生時間、初步判斷污染源、產生污染事件警告、污染持續時間、事件平均值、事件最高值」，分析該次告警可疑污染源，並通報相關計畫前往稽查。在112年至113年運用告警機制，分析現場巡查32件次，其中29件次現場有明顯污染來源，告警準確度為90.1%；現場巡查之可疑污染來源分別為露燃44.8%、工地揚塵為24.1%、工廠逸散為13.8%、移動源及其它分別為13.8%、3.4%。感測器數據分析報告已完成8件次，提供之可疑污染源名單，或現場巡查發現污染排放情事之彙整及通報後，經各相關管制計畫前往稽查，統計至113年12月31日，合計處分案件共11件次，其中露天燃燒共4件、工廠違規6件，工地揚塵1件，處分金額合計新臺幣362萬8,750元。本計畫針對敏感族群如老人及孩童為對象，已完成辦理5場空氣品質及污染認知宣導會，參加人數超過1,600人。在數據上傳品質部分，感測器資料上傳完整率達97.1%，顯示數據傳輸穩定度佳。為提升工作人員專業知能，以及針對各項工作內容依據實際執行現況提出後續精進作為，使本計畫順利推動如期達成環保局交辦之相關業務，邀請專家學者以感測器於輔助環境治理為方向，已完成辦理4場次之教育訓練。於多元智慧監控部分，本工作團隊針對常受鄰近居民陳情之重點工業區周圍架設高值啟動自動影像系統，於112年10月8日監測上O資源管道爆炸觸發啟動1件次，管制計畫進廠稽查發現該廠因進行鍋爐氣電共生程序作業致流體化床鍋爐（E501）發生爆炸，產生大量黑色濃煙散布於大氣中，故罰鍰新臺幣100,000元整。針對常受鄰近居民陳情之重點工業區或陳情熱點，如民眾異味陳情、畜牧場異味陳情及縣市周界異味陳情周圍架設空氣異味微型感測器，統計113年12月31日前已執行10點次異味監測，同時監測輔導改善前後空品狀況。其中永O畜牧場，管制計畫於分析異味熱點時段進場確實有發現業者因每日僅於上午清理1次排泄物，導致夜間累積異味較重已請業者改善，後續分析異味監測數據已明顯下降。

中文關鍵字

空氣品質感測器、多元智慧監控

類型	檔名	檔案大小	說明	最新版公開日期
期末報告	112-113年度嘉義縣精進空品感測器物聯網發展計畫(定稿)V3.pdf	52MB		2025/2/5 上午 12:00:00

摘要	This project already planned the optimal sites and adjusted 12 sites. The coverage rate of sensors in 18 districts reached 100%. In addition, the “Air Quality Data Center” system was established, the date rental service was provided, and the education training course was held. The sensors were inspected 4 times every six months. Executed 10% of sensors sampling inspection 8 times in each season and all results match the standard of the Ministry of Environment. Executed clustering, decay, dynamic calibration, and operations in each month and finished 22 times analysis. For those unusual or outlier sensors, staff immediately check for pollution on-site or replace faulty equipment. The analysis results show that during the good air quality season (April to September), due to the lower ambient concentration, the equipment in the measuring station and the environmental concentration trend change are significantly different, and there is a lot of improvement during the bad air quality season. In addition, the reason of failure sensor repairmen is the product lifespan of the sensor components is only 3 to 5 years and the county replaced all domestic sensors since 2021. Therefore, the abnormality rates of the main board and PM2.5 components detected in 2023 were 27.7%. After the project team arrange worker to replace components, the abnormality rate in 2024 decreased significantly compared to 2023. This project combines sensing data and the pollution alert system to analyze air pollution events and send the message through LINE with information on the “number of sensors, time, preliminary determination of pollution source, duration of contamination, average value of event and the highest value of event” and inform related project to suspected. With the alert system to analyze 32 on-site inspections from 2023 to 2024. There are 29 on-site cases with obvious pollution sources, the correct alarm rate is 90.1%. The suspect sources of on-site inspection include open burning are 44.8%, factory violations are 24.1%, factory emissions are 13.8%, and mobile sources and other pollution at 13.8% and 3.4%. The sensor data analysis report was completed 8 times and provided the suspicious pollution list or found the compilation and notification of pollution emissions by on-site inspection. After the relevant control programs inspected the on-site situation, a total of 11 cases of sanctions were calculated, including 4 cases of open burning, 6 cases of factory violations, and 1 case of construction site dust. The total amount of penalty was NT$ 3,628,750 until December 31, 2024. This project held the promotion of air quality and air pollution 5 times for sensitive groups such as the elderly and children and over 1,600 people participated. In terms of the quality of data upload, the sensor data upload completeness rate reached 97.1% which means data transmission is stable. The project team held the education course 4 times with experts or scholars teaching the use of sensors to assist in environmental management as the topic to enhance the professional knowledge of staff and propose follow-up improvement measures based on the actual implementation status of each work content. So that this plan can be smoothly promoted and the relevant tasks assigned by the Environmental Protection Bureau can be completed as scheduled. The project team set up the high-value trigger automatic imaging system around industrial areas which were petitioned by citizens. The sensor monitored the Shang O resource pipeline explosion 1 time on October 8, 2023. It’s the fluidized bed boiler exploded due to the operation of the boiler gas-electricity cogeneration process and spread a large amount of black smoke into the atmosphere. The factory was fine NTD 100,000 after inspection. Micro air odor sensors are installed around key industrial areas or hotspots that are often complained about by nearby residents, such as complaints from the public, complaints from livestock farms, and complaints from county and city boundaries. Statistics show that as of December 31, 2024, 10 odor monitoring sites have been implemented, and the air quality conditions before and after monitoring and improvement have been provided. Among them, Yong O Livestock Farm analyzed the odor hotspots and found that the operator only cleaned up the excrement once a day in the morning, which led to the accumulation of heavy odor at night. The operator was asked to improve it. Subsequent analysis of the odor monitoring data showed that decline.
英文關鍵字	air quality sensors

摘要

This project already planned the optimal sites and adjusted 12 sites. The coverage rate of sensors in 18 districts reached 100%. In addition, the “Air Quality Data Center” system was established, the date rental service was provided, and the education training course was held. The sensors were inspected 4 times every six months. Executed 10% of sensors sampling inspection 8 times in each season and all results match the standard of the Ministry of Environment. Executed clustering, decay, dynamic calibration, and operations in each month and finished 22 times analysis. For those unusual or outlier sensors, staff immediately check for pollution on-site or replace faulty equipment. The analysis results show that during the good air quality season (April to September), due to the lower ambient concentration, the equipment in the measuring station and the environmental concentration trend change are significantly different, and there is a lot of improvement during the bad air quality season. In addition, the reason of failure sensor repairmen is the product lifespan of the sensor components is only 3 to 5 years and the county replaced all domestic sensors since 2021. Therefore, the abnormality rates of the main board and PM2.5 components detected in 2023 were 27.7%. After the project team arrange worker to replace components, the abnormality rate in 2024 decreased significantly compared to 2023. This project combines sensing data and the pollution alert system to analyze air pollution events and send the message through LINE with information on the “number of sensors, time, preliminary determination of pollution source, duration of contamination, average value of event and the highest value of event” and inform related project to suspected. With the alert system to analyze 32 on-site inspections from 2023 to 2024. There are 29 on-site cases with obvious pollution sources, the correct alarm rate is 90.1%. The suspect sources of on-site inspection include open burning are 44.8%, factory violations are 24.1%, factory emissions are 13.8%, and mobile sources and other pollution at 13.8% and 3.4%. The sensor data analysis report was completed 8 times and provided the suspicious pollution list or found the compilation and notification of pollution emissions by on-site inspection. After the relevant control programs inspected the on-site situation, a total of 11 cases of sanctions were calculated, including 4 cases of open burning, 6 cases of factory violations, and 1 case of construction site dust. The total amount of penalty was NT$ 3,628,750 until December 31, 2024. This project held the promotion of air quality and air pollution 5 times for sensitive groups such as the elderly and children and over 1,600 people participated. In terms of the quality of data upload, the sensor data upload completeness rate reached 97.1% which means data transmission is stable. The project team held the education course 4 times with experts or scholars teaching the use of sensors to assist in environmental management as the topic to enhance the professional knowledge of staff and propose follow-up improvement measures based on the actual implementation status of each work content. So that this plan can be smoothly promoted and the relevant tasks assigned by the Environmental Protection Bureau can be completed as scheduled. The project team set up the high-value trigger automatic imaging system around industrial areas which were petitioned by citizens. The sensor monitored the Shang O resource pipeline explosion 1 time on October 8, 2023. It’s the fluidized bed boiler exploded due to the operation of the boiler gas-electricity cogeneration process and spread a large amount of black smoke into the atmosphere. The factory was fine NTD 100,000 after inspection. Micro air odor sensors are installed around key industrial areas or hotspots that are often complained about by nearby residents, such as complaints from the public, complaints from livestock farms, and complaints from county and city boundaries. Statistics show that as of December 31, 2024, 10 odor monitoring sites have been implemented, and the air quality conditions before and after monitoring and improvement have been provided. Among them, Yong O Livestock Farm analyzed the odor hotspots and found that the operator only cleaned up the excrement once a day in the morning, which led to the accumulation of heavy odor at night. The operator was asked to improve it. Subsequent analysis of the odor monitoring data showed that decline.

英文關鍵字

air quality sensors