環境資源報告成果查詢系統

108年度花蓮縣營建工程暨噪音稽查管制計畫

中文摘要 依據合約規範工作項目及內容,配合貴局之要求,於108年度內完成規定之工作項目並提出工作成果報告,每月初提送上月執行成果報表,而各項作業至108年12月31日止,營建工地完成簽立工地自主管理同意書共計55處(年度新增37處)。在道路認養方面,目前累計61處(年度新增42處),洗掃長度達47,479公里,營建工程巡查共完成4,447處次,空污費共徵收2,734件,稽催件數達477件,其餘各項工作1至12月執行成果依序說明如下: (一)巡查管制作業 統計108年度工地污染管制巡查共完成4,447次,其中0缺失共計3,929次佔88%,10點以下缺失共計424次佔10%,10點以上嚴重缺失共計94次佔2%。各月份巡查處次與本計畫執行初期所定之巡查目標相較,累計達成率已達100%。其中各鄉鎮市巡查件數以污染負荷最大地區花蓮市1,245件及吉安鄉793件最多,其次依序為玉里鎮(371件)及新城鄉(346件)。平均每月巡查件數為371件。以工程類別來看,則以其他工程2,578件與建築(RC)工程770件,為主要巡查重點。 就稽查件數統計,目前共稽查266處次,以建築房屋RC工程佔最多,其次為其他工程。就各鄉鎮之稽查件數統計,以花蓮市及吉安鄉為主要稽查重點區。 為執行營建工地噪音管制,目前本計畫人員至12月底為止建築工程進行113次營建工地噪音主動稽查量測,依據現場主要施工機具進行噪音量測,分別有怪手、吊車、破碎機、預拌車等易發生噪音源之機具,依據法規量測Leq與Lmax,執行迄今共查獲9處工地於施工期間曾有音量超標情形,其噪音來源為佈樁作業及結構體灌漿時所產生,後續經與業者溝通改善並於改善期限內再次複查,已無超標情形。統計該113處營建工地噪音量測結果,Leq量測範圍在55.0~83.7dB(A)之間,Lmax量測範圍在66.1~96.2dB(A)之間。Lmax值較高之工程大多集中在開挖及佈樁作業。 (二)一、二級工程管制率及法規符合度 統計108年度之工地共計列管4,084處,在4,084處列管第一至三級營建工程中屬於一級工程有1,038處,佔總數之25%;二級工程有1,7,84處佔總數44%;三級工程1,262處佔總數之31%,而至108年12月31日止,針對工程巡查狀況共計巡查2,969處,其中第一級工程共現勘巡查846處,納管率為82%,第二級工程巡查1,391處,納管率為78%,整體花蓮縣營建工程納管率為73%。 而在法規符合度上,一級工程法規符合度達到93%,二級工程法規符合度達到89%,總法規符合度達到91%,後續仍將持續加強管制相關工作。 (三)營建工程空氣污染防制費徵收業務 統計108年度本縣營建工程空氣污染防制費共徵收5,639萬6,716元,其中以疏濬工程共計徵收2,047萬5,405元最多,佔總徵收金額36%,次之係其他營建工程營建工程共計徵收1,679萬1,236元,佔總金額30%。107年營建工程空氣污染防制費共徵收4,645萬6,875元,108年徵收金額較去年同期增加993萬9,841元。 統計105年至108年間以營建工程徵收類別分析。105年度空污費徵收以道路工程為主,而106年度新增疏濬工程11件,總計徵收金額為新台幣為2,021萬5,066元,107年度以其他營建工程及疏濬工程為主要徵收對象,共徵收2,890萬8,961元,108年統計12月空污費徵收以其他營建工程及疏濬工程為主要徵收對象,共計徵收3,726萬6,641元。綜合以上,本縣營建工程空污費徵收工程類別仍以其他營建工程及疏濬工程為主要徵收對象。由營建工程徵收鄉鎮別分佈來看,花蓮縣主要營建工程分布於花蓮市,其它鄉鎮則分布於吉安鄉、新城鄉。 108年1月至12月期間,縣境內因合約經費減少異動申請退費的件數最多,合計共有301件,共退費63萬9,906元;其次為因工期縮短申請退費有214件,共退費155萬3,711元。以退費金額來看,以「工期縮短」申請退費的金額為最高,共計申請退費有214件,共退費為155萬3,711元。因合約經費減少異動申請退費仍屬多數,但其退費金額卻不若因「土石外運減少」申請退費案件致空污費退費來的高,係因空污費計徵方式有別之故。 (四)營建工程空污費稽催作業 統計108年度由營建工程資料庫系統統計營建空污費申報未依期限繳納者共計有118處工地,分別為其他類工程95件、空污費繳費金額為新台幣64萬1,185元;管線開挖工程13件、空污費繳費金額為新台幣15萬1,590元;道路工程7件、空污費繳費金額為新台幣23萬2,826元;建築工程(SRC)2件、空污費繳費金額為新台幣1萬1,232元;建築工程(拆除)1件、空污費繳費金額為新台幣1,311元,逾期繳納空污費金額總計為新台幣103萬8,144元,經人員進行催繳作業後已完成100%到繳。平均各月份營建空污費主動到繳率為98%及次月進行催繳作業之達成率則為100%,僅有少數案件因遺忘而遲延繳納情形。 (五)機動車輛噪音管制暨非游離輻射源調查 計畫期間已執行65場次之機動車輛攔檢、巡(查)作業,由警方依據車輛噪音感受程度予以攔查檢測,共計攔檢、巡(查) 645輛次,檢測作業相關記錄皆彙整後提送至環保局執行後續程序,統一將不合格車輛列為通知到檢對象。108年度共計發文到檢117輛次,其中51輛次到檢完成,3輛次移至外縣市,報廢2輛次,過戶則有1輛次,改善完成率50%。 除依工作項目執行攔檢(查)作業外,並定期依據噪音車檢舉網站及警方寧靜專案資料進行彙整,篩選經陳情之可疑噪音車,協助辦理通知到檢之行政程序及現場檢測作業,計畫期間已執行通知到檢作業共計51輛汽機車,其中33輛為路邊攔查檢及校園稽巡查部分,噪音車陳情網站部分為16輛,外縣市移入則有2輛。 本年度(108年)應執行本縣非游離輻射量測作業達50點次以上,計畫期間迄今已完成50點次。 (六)事業廢棄物流向管制 統計108年度針對縣轄內營建廢棄物清除及處理機構共查核238件次皆符合規定,並完成4次公民營廢棄物處理機構深度查核,除第一季現勘場區設施與許可不符,依法裁處新臺幣6,000元罰鍰,其餘3次查核皆符合規定,本縣轄內GPS列管車輛稽(巡)查每季30車次,Q1:37車次、Q2:47車次、Q3:49車次、Q4:38車次皆符合操作規定,行政院環境保護署事業廢棄物清運機具即時監控系統勾稽每週一次,共計51次1,223筆。廢棄物列管工廠(場)事業及營建工地共查核1,300件,於108年8月30日完成大型事業深度查核1次,現場符合規定。廢棄物清理計畫書營建工程部分請空污費櫃檯每週提送當週新申報營建工地名單,共提送53次2,706件,新增列管102件,廢棄物清理計畫書審查情形:已核備94件、審查中2件、展延1件、未提送2件、未提送(期程未達)2件、不列管1件,解列部分已提送解列判定表94件已核備92件、審查中2件。 (七)創新作法-出入口道路髒污 AI 輔助辨識 工地出入口道路髒污最常遭民眾陳情之缺失樣態,為加強營建工程污染管制,雖然可要求工地人員立即洗掃改善,減輕污染衝擊,由於環保稽查人員無法長時間的盯梢,部分營建工地抱著僥倖的心態,虛偽應付環保人員的稽查。因此,選定一處公共工程即時監視系統做為創新示範工地,撰寫AI辨識系統,以作為未來管制利器,將針對出入口CCTV影像,抓取固定區域中顏色相似範圍之圖片檔,建立基礎RGB色階比對數值,包括出入口道路髒污優、可、劣等級,降雨情境、車輛進出情境、人力灑水、車輛洗掃情境等,個特定出入口即可建立一RGB色階辨識準則,後續CCTV擷取影像經比對此辨識準則,可由AI輔助便識出入口道路髒污程度,透過簡訊等方式提醒管理者查看該時段道路髒污情形,落實預防及污染改善以促進稽巡查管制效率,紓解人力管制負荷。 1.測試對象 疏濬工程及環評工程等重大營建工地,於審查階段均有要求於工地出入口裝設CCTV,為加值CCTV的運用,選定一處營建工程做為示範場所,作為推動科技化環境管制的相關政策研擬、業務推動、居民溝通之重要參考依據,本(108)年度初步篩選蘇花改工程及疏濬工程,調查後發現疏濬工程CCTV皆為封閉式,僅蘇花改工程為開放式平台可供連網,因此選定目前即有可連網架設平台之CCTV蘇花改工程,作為影像擷取測試的對象。 2.污染減量推估 AI輔助辨識出入口道路髒污程度變化,工地如能落實道路洗掃髒污管控工作,掃街污染削減量效益推估推估每日4次洗掃長度合計至少2公里/日,依據環保署減量公式(削減量=洗掃街長度×減量係數)推估,TSP可削減0.028公噸,PM10 為0.005公噸,PM2.5 為0.001公噸,一年至少將可減量TSP 10.074公噸,PM10 1.898公噸,PM2.5 0.443公噸。 3.執行成果 加值運用大型營建工程自主架設之CCTV,程式自動擷取影像圖片,導入AI輔助辨識出入口道路髒污程度,強化營建工地污染即時管制,預定提升辨識工地之管理辦法法規符合度及污染削減率,有效改善本縣道路髒污情形,進而減少粒狀污染物排放,改善空氣品質及地球永續發展之目標,目前執行至11月底蘇花改道路髒污辨識影像,選定來源使用台9線蘇花公路中仁隧道接續工程八支攝影機以及台9線蘇花公路仁水隧道新建工程的11支攝影機。其中有數隻為隧道內之攝影機,因光線原因不予採用。 經過機器學習之後,建立道路污染辨識API,利用此模型,可將監視影像截圖後傳送至本辨識模組,辨識模組回傳1、4、10等分數。10分者立即通報業者或洗掃街車進行洗掃,8小時內累積12分以上(3個4分)也進行類似通報。 以工地落實工地出入口洗掃作業頻率(長度500公尺、一天4次)進行估算,預估一天洗掃貢獻量為PM10削減0.0052公噸。 設定分數之對應作為,假如分數為10分,立即以Line機器人或行動電話簡訊通知業主進行洗掃,4小時內累積5張畫面4分時,以Line機器人或行動電話簡訊通知業主進行洗掃4小時內累積5張畫面10分時,要求工地提出改善作為,並拍照回傳。此API提供精確度及喚回度同為77.78%,已可用於明年CCTV錄影監控網路設備,針對重大污染與屢遭陳情之對象進行影像分析。 (八)其他 人力及物力支援:皆依局內指示協助辦理。 臨時交辦事項:皆依局內指示協助辦理。 九、統計至108年12月31日止,合約達成率依工作權重加權計算後為100%,整體進度管控達成率100%。
中文關鍵字 營建工程、空氣污染、污染防制、噪音管制、非游離輻射、營建廢棄物流向管制

基本資訊

專案計畫編號 108-02-05 經費年度 108 計畫經費 14250 千元
專案開始日期 2019/01/01 專案結束日期 2019/12/31 專案主持人 趙嘉雍
主辦單位 花蓮縣環境保護局 承辦人 黎虹宛 執行單位 康廷工程顧問企業有限公司

成果下載

類型 檔名 檔案大小 說明
期末報告 108年花蓮營建期末報告-定稿.pdf 80MB

108 Hualien County Construction Project and Noise Audit Control Plan

英文摘要 According to the standard work items and contents of the contract, and in accordance with the requirements of your bureau, the required work items will be completed within 108 years and a work results report will be submitted. The results report of the previous month's execution will be submitted at the beginning of each month, and each operation will be completed until December 31, 108 As of now, the construction site has completed a total of 55 written consents for the independent management of the construction site (37 new sites in the year). In terms of road adoption, there are currently 61 sites (an increase of 42 in the year), with a sweep length of 47,479 kilometers. A total of 4,447 inspections of construction projects have been completed. A total of 2,734 air pollution charges have been collected. The implementation results of the various tasks from January to December are described as follows: (I) Inspection and control operations According to statistics, there were 4,447 inspections of construction site pollution control in the year of fiscal year 2010, of which 3,929 were missing 0, accounting for 88%, 424 were missing below 10, accounting for 10%, and 94 were serious missing, accounting for 2%. Compared with the inspection targets set at the beginning of the implementation of this plan, the cumulative inspection rate in each month has reached 100%. Among them, the number of inspections in each township was 1,245 in Hualien City and Ji'an Township in the area with the largest pollution load, followed by Yuli Town (371 cases) and Xincheng Township (346 cases). The average number of inspections per month is 371. In terms of project categories, 2,578 other projects and 770 construction (RC) projects were the main inspection focus. Statistics on the number of audits have currently totaled 266 inspections, with construction and housing RC projects accounting for the most, followed by other projects. According to the statistics of the number of inspections in each township, Hualien City and Ji'an Township are the main inspection areas. In order to implement construction site noise control, the planners have conducted 113 active construction site noise inspections and measurements until the end of December, and conducted noise measurements based on the main construction equipment at the site, including strange hands, cranes, crushers, and Noise-prone machines such as truck mixers have measured Leq and Lmax in accordance with regulations. A total of 9 construction sites have been found to have exceeded the volume during construction so far. The source of noise is generated during pile placement and structure grouting. After communication and improvement with the industry and re-examination within the improvement period, there is no excess. The noise measurement results of the 113 construction sites are counted. The measurement range of Leq is between 55.0 and 83.7 dB (A), and the measurement range of Lmax is between 66.1 and 96.2 dB (A). Most projects with higher Lmax values ​​are concentrated in excavation and pile laying operations. (II) Control rate of first and second-level projects and compliance with regulations According to the statistics, there were 4,084 construction sites in 108 years, among which 1,038 were first-level and third-level construction projects, which were 1,038, accounting for 25% of the total; 1,7,84 second-level projects accounted for the total 44. 1,262 third-level projects accounted for 31% of the total. As of December 31, 108, a total of 2,969 sites were inspected for the inspection status of the project. Among them, 846 first-stage projects were inspected and inspected, and the management rate was 82%. The second-level project inspected 1,391 sites, and the management rate was 78%. The overall management rate of Hualien County construction projects was 73%. In terms of compliance with laws and regulations, the compliance of primary engineering regulations reached 93%, the compliance of secondary engineering regulations reached 89%, and the compliance of general regulations reached 91%. Control-related work will continue to be strengthened in the future. (3) Collection of Air Pollution Control Fees for Construction Projects According to statistics, in fiscal year 108, air pollution control fees for construction projects in the county totaled RMB 56,396,716, of which dredging projects totaled RMB 20.47575,405, which accounted for the largest collection of 36%, followed by other construction projects, which totaled 16.79 million. 1,236 yuan, accounting for 30% of the total amount. A total of 46,456,875 yuan was levied on air pollution control fees for construction projects in 107 years, and the amount collected in 108 increased by 9,399,841 yuan over the same period last year. Statistics analysis of construction project collection category from 105 to 108 years. The collection of air pollution charges in 105 was mainly based on road works, while 11 new dredging works were added in 106, and the total amount collected was NT $ 20,025,000. 28.9 million yuan 8,961 yuan. In December of 108 statistics, the collection of air pollution charges was mainly for other construction projects and dredging projects, and a total of 37.266 million yuan was levied. To sum up, the types of air pollution charges for construction projects in this county are still subject to other construction projects and dredging projects. Judging from the distribution of townships and villages collected by construction projects, the main construction projects in Hualien County are located in Hualien City, and other townships are located in Ji'an Township and Xincheng Township. From January to December 108, the county had the largest number of applications for refunds due to reductions in contract funds, with a total of 301, a total of 639,906 refunds; followed by 214 applications for refunds due to shortened construction periods The refund is RMB 1.553,711. In terms of the refund amount, the refund amount for the "shortened construction period" is the highest. A total of 214 refunds were applied for a total of RMB 1.553711. Due to the reduction of contract funds, there are still many applications for refunds, but the amount of refunds is not as high as the air pollution charges due to "reduction in earth and rock transportation". For no other reason. (IV) Inspection of air pollution charges for construction projects According to the statistics of the construction project database system in fiscal year 108, there were 118 construction sites that had failed to pay the deadlines, which were 95 other types of projects, and the amount of air pollution fees paid was NT $ 64,185. 13 works, air pollution fee payment amount is NT $ 1,590; 7 road works, air pollution charge payment amount is NT $ 2,32,826; 2 construction works (SRC), air pollution charge payment amount is new NT $ 11,232; 1 construction project (demolition), air pollution fee payment amount is NT $ 1,311, overdue payment of air pollution fee amounted to NT $ 1,038,144, 100 people have completed 100 % Paid. On average, the active payment rate of construction air pollution fees in each month was 98% and the fulfillment rate of the next month's call-up operations was 100%. Only a few cases were delayed due to forgetting. (V) Motor Vehicle Noise Control and Investigation of Non-Ionizing Radiation Sources During the planning period, 65 inspections and inspections of motor vehicles have been carried out. The police will inspect and inspect according to the noise level of the vehicles. A total of 645 inspections and inspections will be conducted. The relevant records of inspections are compiled. Submit it to the Environmental Protection Agency for follow-up procedures, and unqualified vehicles will be listed as the subject of notification. In 108, a total of 117 documents were issued for inspection, of which 51 were completed, 3 were moved to other counties and cities, 2 were scrapped, and 1 was transferred, improving the completion rate by 50%. In addition to carrying out inspection (check) operations according to work items, and regularly compiling according to the noise car report website and police quiet project data, screening of suspicious noise cars that have been reported, assisting in the administrative procedures for notification of inspection and on-site testing operations During the painting period, a total of 51 steam locomotives have been notified and inspected, of which 33 are roadside inspections and campus inspections, 16 are noise car sites, and 2 are moved in from other counties. This year (108), the county should perform non-ionizing radiation measurement operations more than 50 times, and 50 points have been completed so far during the planning period.   (VI) Regulation of the flow of business waste According to statistics, in 1988, a total of 238 inspections of waste disposal and disposal agencies under the jurisdiction of the county were found to be in compliance with the requirements, and 4 in-depth inspections of waste disposal agencies of civilian camps were completed. A fine of NT $ 6,000 was fined, and the remaining three checks were in compliance with the regulations. The GPS audited vehicle inspection (patrol) within the county jurisdiction inspected 30 trips per season, Q1: 37 trips, Q2: 47 trips, Q3: The 49 trips and Q4: 38 trips are in compliance with the operating requirements. The real-time monitoring system for the waste removal and removal equipment of the Environmental Protection Agency of the Executive Yuan is checked once a week, with a total of 51 1,223 records. A total of 1,300 waste pipe factory (field) businesses and construction sites were inspected. On August 30, 108, a large-scale business in-depth inspection was completed and the site met the requirements. For the construction project of the waste cleaning plan, please ask the air pollution charge counter to submit a new construction site list for the week, a total of 53 2,706 submissions, 102 new pipelines, and review of the waste cleaning plan: 94 cases have been prepared, 2 are under review, 1 has been extended, 2 cases have not been submitted, 2 cases have not been submitted (the deadline has not been reached), 1 case has not been listed, and 94 cases have been submitted to the delisting judgment form. 92 cases have been prepared and 2 are under review. (VII) Innovative Practices-AI-assisted identification of dirty roads at entrances and exits Dirty roads at the entrances and exits of construction sites are most often lacking in public sentiment. In order to strengthen the pollution control of construction projects, although construction site personnel can be required to immediately clean up and improve the impact of pollution, environmental protection auditors can not be followed for a long time. With a fluke mentality, it is hypocritical to cope with the inspection by environmental protection personnel. Therefore, a public engineering real-time monitoring system was selected as an innovative demonstration site, and an AI identification system was written as a future control tool. It will capture the image files of similar ranges in fixed areas for CCTV images at the entrance and exit to establish the basic RGB color gradation. The comparison value includes the dirty, excellent, acceptable, and inferior grades of roads at the entrance and exit, rainfall scenarios, vehicle entering and exiting scenarios, human watering, and vehicle washing and cleaning scenarios. A specific entrance and exit can establish an RGB color gradation identification criterion for subsequent CCTV capture. The image is compared with this identification criterion, and the degree of dirt on the road at the entrance can be recognized by the assistance of AI. The manager is reminded to check the pollution of the road during the period by text messages, etc., and implement prevention and pollution improvement to promote the efficiency of inspection and control, and relieve human control. load. 2. Estimation of pollution reduction The AI ​​assists in identifying changes in the degree of pollution at the entrance road. If the site can implement the road cleaning and pollution control work, the estimated benefit of street pollution reduction is estimated to be at least 2 km / day for a total of 4 cleaning times per day. The formula (reduction amount = length of scrubbing street x reduction factor) estimates that TSP can reduce 0.028 metric tons, PM10 is 0.005 metric tons, PM2.5 is 0.001 metric tons, and at least one year can reduce TSP 10.074 metric tons, PM10 1.898 metric tons, PM2. 5 0.443 metric tons. 3. Implementation results The value-added uses CCTV independently erected by large-scale construction projects, the program automatically captures image pictures, introduces AI to help identify the degree of pollution at the entrance road, strengthens the real-time control of pollution at construction sites, and plans to improve the compliance with the management methods and regulations and the pollution reduction rate for identifying construction sites. Effectively improve the county's road pollution, thereby reducing particulate pollutant emissions, improving air quality and sustainable development of the earth. Currently, the implementation of Suhua ’s road pollution identification image has been implemented until the end of November, and the selected source is Taiwan 9 line Suhua Eight cameras for the continuous project of the highway Zhongren Tunnel and 11 cameras for the new construction of the Renshui Tunnel of the Suhua Highway on the 9th line. Several of them are cameras in the tunnel, which are not used due to light. After machine learning, a road pollution identification API is established. Using this model, screenshots of surveillance images can be sent to this identification module, and the identification module returns scores of 1, 4, and 10. A 10-point person immediately notified the operator or a street sweeper to perform cleaning, and a similar report was also made for 12 or more points (3 4 points) within 8 hours. Based on the frequency of site cleaning operations (500 meters in length, 4 times a day), it is estimated that the daily cleaning contribution will be reduced by 0.0052 metric tons for PM10. Correspond to the set score. If the score is 10, immediately notify the owner to perform cleaning by using a line robot or mobile phone newsletter. When 4 screens are accumulated within 4 hours, 4 points will be notified by the line robot or mobile phone newsletter. 4 When 5 frames were accumulated for 10 minutes within 10 hours, the construction site was asked to propose improvement actions and take photos to return. This API provides accuracy and recall of 77.78%. It can be used in CCTV video surveillance network equipment next year for image analysis of major pollution and frequently-seen objects. (Eight) other Human and material support: All are assisted in accordance with the instructions of the Bureau. Temporary assignments: follow the instructions of the bureau to assist in handling. Nine, statistics until December 31, 108, the contract completion rate is 100% after weighted calculation based on the work weight, and the overall progress control completion rate is 100%.
英文關鍵字 Construction project, Air pollution, Pollution prevention, Noise control, Non-ionizing radiation, Construction waste flow control