應用風險地圖探討污染土地再生管理策略

The development of land-use regeneration strategies on contaminated site by using risk maps

陳怡君(I-Chun Chen)

指導教授 ： 馬鴻文

國立臺灣大學/工學院/環境工程學研究所/博士(2013年)

https://doi.org/10.6342/NTU.2013.00792

摘要

污染土地再利用(以下簡稱褐地再生) 為複雜而難解習題，執行關鍵涵蓋經濟面、社會面、環保面、法律面等需求，因此污染責任釐清、政府法規彈性誘因、污染場址清除範圍、以及開發資本支持…等議題，為近幾年褐地再生研究焦點。然而，諸多研究僅說明關鍵問題，卻尚無研究方法探討如何降低關鍵因子不確定性，此部分則為本研究主要探討範疇。另一方面，國內目前面臨污染場址管理不僅單一場址整治問題，同時亦面臨大範圍的土壤及地下水污染場址整治或土地再利用等問題，其涉及土地利用類型與受體暴露途徑複雜，管理者並非能以保守點估計風險值或機率風險值提出完整場址管理決策，若採以最保守性可接受致癌風險(10-6)以及非致癌商數(1)作為整治目標考量，導致整治決策者無法更彈性考量其他因素，例如污染土地用途變更成都市化讓污染土地可以永續發展。若污染場址整治決策僅考量環保議題，則大範圍污染土地因整治經費籌措困難而無法活絡。
有鑑於此，應有詳細估算污染土地造成健康風險及土地價值減損變動趨勢研究方法，以探討污染土地造成風險空間分布特性。依據國外文獻指出，褐地再生關鍵成功因素為「環境風險」及「經濟誘因」兩項因子最受利害相關者重視，同時污染土地因鄰近居民健康風險危害所衍生污名化之風險成本難以估價，亦導致污染土地價值鑑價考量時主要不確定性影響，因此本研究透過地理資訊系統結合健康風險評估，將健康風險評估納入褐地再生管理決策中，建置污染場址風險地圖模式(ArcGIS for Risk assessment management, 簡稱Arc-RAM)，建立彈性風險值與土地價值間關係，以量化風險估算污染土地再利用之不確定性影響。
本研究第一部分透過蒐集風險問卷參數建立專一性污染場址風險地圖，探討族群暴露習慣所呈現污染場址之風險地圖分布差異，本研究透過機率值分析風險地圖空間分布，研究結果與傳統蒙地卡羅第三層次風險評估比對，發現風險地圖空間分布可以突顯污染範圍內暴露生活習慣差異，以解決蒙地卡羅無法分析風險空間分布問題。第二部分依據風險地圖繪製不同機率個別風險值(Individual risk, IR)，顯示大範圍污染場址涵蓋各村里風險等級分布與影響暴露人口數，依照不同機率值呈現個別風險值(IR)下所影響各村里族群風險(Population risk, PR)變動情形，探討族群風險值(PR)與個別風險值(IR)相關性，依據累積機率10%; 50%; 90%; 100%設定，以族群風險變動趨勢圖作為決策可接受風險基準參考，進一步探討族群風險與土地價值減損關係。第三部份依據不同機率族群風險值(PR)或個別風險值(IR)設定彈性風險，估算彈性風險範圍所造成各村里土地價值影響(Land values at different risk levels, LVrv)，依照風險變動情況下估算土地價值減損(Damaged land values of risk level, DLVrv)，DLVrv變動斜率越大則代表該村里(或場址)具有污染土地再利用潛在誘因。第四部份則是引用Bell模式設定不同彈性風險整治目標情境，以估算污染土地鑑價之改善污染計畫誘因性及風險性成本，並且估算風險性成本所佔污染土地減損價值比例，透過案例研究發現污染場址在無考量整治過程逕行污染土地再利用，估算國內風險成本比率(42.79%)比起國外來得高，尤其個案場址之住宅區因健康風險影響導致劃定風險範圍程度大，此外土地價值減損程度介於假設情境一(R＞1×10-6)與情境二(R＞1×10-5)之間，依據本研究個案成果顯示在不損及鄰近居民健康風險下，將彈性風險設定為1×10-5造成土地價值減損幅度較小，因此透過本研究分析流程應須共同考量土地利用永續與居民健康保護，則可提出較經濟與高效益之褐地再生管理政策。
本研究結果證明建立污染場址專一性風險地圖評析方法，可以解決褐地再生涉及「土地價值」與「環境風險」等複雜不確定性影響，突破以往風險評估單點估計與蒙地卡羅不確定性分析方法，以風險地圖分布狀況探討污染場址周圍族群風險變化趨勢，並從空間角度共同考量環境風險與土地價值兩者變動趨勢，並進一步估算不同風險等級所影響土地價值減損，透過本研究可以清楚評析國內推動褐地再生管理決策機制，應以彈性風險設定條件與土地管理共同擬定較經濟與有效彈性管理策略，並以系統化機制降低褐地再生之風險成本不確定性影響，以避免褐地再生議題於環保政策與土地再利用脫鉤情形。

關鍵字

褐地再生 ； 風險地圖 ； 族群風險 ； 污染土地減損 ； 不確定性 ； 彈性風險管理

並列摘要

Land reuse of contaminated land (referred to as “brownfield regeneration”) is a complex and obscure issue. The crucial implementation of brownfield regeneration involves different aspects, including pollution liability, flexible regulations, contamination conditions, financial support, etc. Although many studies provide descriptions of critical problems, none of the research methods explore the issue of how to reduce the uncertainties of key factors. In addition, the problem that contaminated site management in Taiwan faces currently is not just single site remediation but large scale problems of soil and groundwater pollution remediation and land reuse. These problems involve different types of land categories and complicate exposure pathway of receptor. Complete site management decision cannot be made only based on conservative estimation of value at individual risk or probability risk. If the most conservative level of acceptable cancer risk (10-6) and non-carcinogenic quotient (1) are adopted to achieve the goal of remediation, decision makers of remediation will be unable to consider more flexibly in adopting other factors. For example, changing contaminated land use into urbanization to improve the sustainability development for contaminated sites. Then, a wide range of contaminated land will not be successfully resolved due to financing difficulties.
Therefore, a research method estimate the contaminated land which leads to health risk and damage to land value should be available. It can explore the characteristics of risk space distribution in contaminated land. Environmental risk and economic incentives are the two critical success factors of brownfield regeneration in the past studies. In addition, the cost of risk derived from stigmatization due to health hazard of nearby residents caused by contaminated land is incalculable. It also contributes to main uncertainties in assessing damage values of contaminated land. Therefore, we constructed a contaminated site risk map model (ArcGIS for risk assessment management, abbreviated as Arc-RAM) by combining the four major steps of risk assessment with the application of a geographical information (GIS) system. Risk maps estimate uncertainties of contaminated land reuse in order to establish relationship between flexible value at risk and land values.
The first part of this research establishes a specific risk map of contaminated sites through collected risk questionnaire parameters. It can explore differences of risk map distribution of contaminated sites. The space distribution of risk map is analyzed using probability values. The results of this research are compared to the traditional Monte Carlo third-level risk assessment. We discover that risk map space distribution can highlight the differences of exposure habits within the scope of contamination and can resolve the analysis problem of risk space distribution which Monte Carlo risk assessment is unable to achieve. The second part of this research draws different population risk (PR) of probability based on risk maps to show villages’ affected population exposure. In accordance with different probability values under individual risk (IR), it is presented the correlation between population risk (PR) and individual risk (IR). According to the settings of cumulative probabilities of 10%, 50%, 90% and 100%, a risk variation drawing of population risk (PR) is regarded as a reference of acceptable risk benchmark for decision making. It further explores the relationship between population risk (PR) and damage to land value. The third part is to set flexible risk based on different probability risk of PR or IR in order to estimate the impact of land value in each village (LVrv) by flexible risk scope. The damaged land values of risk level (DLVrv) are estimated according to risk change. The greater the change slope DLVrv the more potential inducement of contaminated land reuse of a village (or site). The fourth part is to set different situations of flexible risk remediation goal by referencing Bell model in order to estimate improvement of contaminated plan incentives and cost of risk of contaminated land assessment. It also estimates the proportion of damage value of contaminated land accounted for by cost of risk in this research. If contaminated site reuse is directly conducted without remediation, it is estimated that domestic cost of risk ratio (42.79 %) is much higher the ratio of foreign countries. In particular, case of residential areas has greater degree of delimited scope of risk. In addition, flexibility risk is set to 1×10-5 is lower on damage to land values than other risk condition and directly redevelopment. It proves that if Taiwan wants to strengthen land management policy of brownfield regeneration, sustainability of land utilization and protection of the health of residents must be considered at the same time.
The results of this research prove that establishment of specific risk map analysis of contaminated site is able to resolve complex uncertainties with respect to “land value” and “environmental risk” derived from brownfield regeneration. Given consideration to flexible risk setting condition and land value, this research formulates the economic and effective strategy of flexible management. Moreover, this research takes a systematic mechanism to reduce uncertainty of risk cost of brownfield regeneration.

並列關鍵字

Brownfield regeneration ； Risk maps ； Population risk ； Land-value damage ； Uncertainty ； Flexible risk management

參考文獻

林子欽、陳振惟，2009 ，受污染工業土地之風險、污名與價值：美國經驗的啟示”，住宅學報，第十八卷，第二期，第23-44頁

環保署，2006，土壤及地下水污染場址健康風險評估評析方法及撰寫指引

徐世榮、蕭新煌，2006 ，污染場址再利用政策之研究：美國與台灣之比較”，都市與計劃，第三十三卷，第二期，第143-167頁

Adams, D., Disberry, A.N, Hutchison, N., Munjoma, T., 2000. Taxes, subsidies and the behaviour of brownfield owners. Land Use Policy, 17, 135-145.

Alastair A., Jim B., Stanley M., Bill D., Suzanne H., 2000. The financing of urban regeneration. Land Use Policy, 17, 147-156.