EUROPEAN COMMISSION DIRECTORATE-GENERAL...

1488-REG/R/03/B

EUROPEAN COMMISSION

DIRECTORATE-GENERAL ENVIRONMENT

STUDY ON THE VALUATION AND RESTORATION OFDAMAGE TO NATURAL RESOURCES FOR THE PURPOSE

OF ENVIRONMENTAL LIABILITY

B4-3040/2000/265781/MAR/B3

FINAL REPORT

BY

MACALISTER ELLIOTT AND PARTNERS LTDAND THE

ECONOMICS FOR THE ENVIRONMENT CONSULTANCY LTD

MAY 2001

MacAlister Elliott and Partners Ltd56 High Street

LymingtonHampshire SO41 9AH

United KingdomTel: +44 1590 679016Fax: +44 1590 671573

E-mail: [email protected]: http://www.macalister-elliott.com

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B

Final Report MEP and EFTEC

TABLE OF CONTENTS

EEXXEECCUUTTIIVVEE SSUUMMMMAARRYY

PPAARRTT AA:: EEVVAALLUUAATTIIOONN MMEETTHHOODDOOLLOOGGIIEESS

1 STUDY BACKGROUND AND OBJECTIVES....................................................... 11.1 BACKGROUND................................................................................................................ 11.2 AIMS AND OBJECTIVES................................................................................................... 21.3 SCOPE............................................................................................................................. 2

2 OVERVIEW AND APPROACH TO THE ISSUES................................................ 32.1 MONETARY COMPENSATION BASED ON THE VALUE OF DAMAGE.................................. 32.2 RESOURCE COMPENSATION............................................................................................ 42.3 STRUCTURE OF THE REPORT........................................................................................... 4

3 NATURAL RESOURCE DAMAGE ASSESSMENT AND RESTORATION..... 73.1 PRE-INCIDENT RESOURCE STATUS................................................................................. 73.2 DETERMINING THE SCALE AND ‘SIGNIFICANCE’ OF DAMAGE ...................................... 10

3.2.1 Scale of Damage.................................................................................................. 103.2.2 Significance of Damage ...................................................................................... 113.2.3 Establishing Conservation Objectives ................................................................ 13

3.3 APPLYING NATURAL RESOURCE DAMAGE ASSESSMENT IN PRACTICE......................... 20

4 PRIMARY RESTORATION OPTIONS................................................................ 234.1 SETTING PRIMARY RESTORATION TARGETS................................................................. 244.2 IDENTIFICATION AND CATEGORISATION OF TECHNICAL OPTIONS FOR PRIMARY

RESTORATION............................................................................................................... 284.3 CATEGORISATION AND SELECTION OF RESTORATION OPTIONS.................................... 29

4.3.1 Categorisation of Restoration Options ............................................................... 304.3.2 Selection of Restoration Options......................................................................... 33

4.4 ESTIMATION OF INTERIM LOSSES ................................................................................. 364.5 ROLE OF COST-BENEFIT ANALYSIS .............................................................................. 36

5 COMPENSATORY RESTORATION OPTIONS................................................. 385.1 IDENTIFY THE OBJECTIVES OF COMPENSATORY MEASURES......................................... 385.2 RESOURCE COMPENSATION.......................................................................................... 41

5.2.1 Identifying Compensatory Restoration Projects ................................................. 415.2.2 Classifying and Selecting Compensatory Restoration Projects.......................... 425.2.3 Scaling Restoration Options ............................................................................... 435.2.4 Implications for Monetary Value of Liability...................................................... 47

5.3 MONETARY COMPENSATION ........................................................................................ 47

6 CONCLUSIONS AND RECOMMENDATIONS.................................................. 48

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


PPAARRTT BB:: TTHHEE CCAASSEE SSTTUUDDIIEESS

7 AZNALCÓLLAR MINE TOXIC SPILLAGE....................................................... 537.1 SITE DESCRIPTION........................................................................................................ 537.2 SITE SERVICES.............................................................................................................. 547.3 INCIDENT DESCRIPTION................................................................................................ 557.4 SCALE OF DAMAGE ...................................................................................................... 557.5 CASE STUDY ASSESSMENT OF DAMAGE TO THE DOÑANA NATIONAL PARK AND

CORRIDOR ECOLOGICA DE RIO GUADIAMAR NATURA 2000 SITES............................... 567.5.1 Scope of Damage Assessment ............................................................................. 567.5.2 Establishing Conservation Objectives ................................................................ 567.5.3 Conclusion of Significance Assessment............................................................... 62

7.6 ACTUAL RESTORATION OF DOÑANA NATIONAL PARK AND ESTABLISHMENT OF THEGREEN CORRIDOR ........................................................................................................ 62

7.7 COMPARISON OF ACTUAL AND POSSIBLE RESTORATION ACTIVITIES........................... 637.8 OUTCOME OF SETTLEMENT .......................................................................................... 65

8 SEA EMPRESS OIL SPILL .................................................................................... 668.1 SITE DESCRIPTION........................................................................................................ 668.2 SITE SERVICES.............................................................................................................. 678.3 INCIDENT DESCRIPTION................................................................................................ 688.4 SCALE OF DAMAGE ...................................................................................................... 688.5 IMPACT ASSESSMENT................................................................................................... 698.6 CONCLUSION OF DAMAGE ASSESSMENT ...................................................................... 718.7 ACTUAL RESTORATION OF PEMBROKESHIRE MARINE AND CARMARTHEN BAY AND

ESTUARIES NATURA 200 SITES .................................................................................... 728.8 COMPARISON OF ACTUAL AND POSSIBLE RESTORATION ACTIVITIES........................... 728.9 OUTCOME OF SETTLEMENT .......................................................................................... 74

9 EXXON VALDEZ OIL SPILL................................................................................ 769.1 SITE DESCRIPTION........................................................................................................ 769.2 SITE SERVICES.............................................................................................................. 769.3 INCIDENT DESCRIPTION................................................................................................ 769.4 SCALE OF DAMAGE ...................................................................................................... 779.5 CLEAN-UP ACTIVITIES.................................................................................................. 779.6 IMPACT ASSESSMENT FOR NATURAL RESOURCES........................................................ 779.7 RECOVERY OBJECTIVES FOR HUMAN SERVICES........................................................... 789.8 OUTCOME OF SETTLEMENT .......................................................................................... 789.9 RISK AND UNCERTAINTY – IMPLICATIONS FOR LIABILITY ........................................... 799.10 COMPARISON OF EXXON VALDEZ AND SEA EMPRESS OIL SPILLS................................ 79

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


FIGURES, TABLES AND BOXESFiguresFIGURE 2-1: DAMAGE ASSESSMENT AND CHOOSING THE RESTORATION OPTIONS..................................................5FIGURE 3-1: WETLAND ECOSYSTEM SERVICES AND SOURCES OF VALUE .............................................................10FIGURE 4-1: PRIMARY RESTORATION PROCESS PATHWAYS ..................................................................................23FIGURE 5-1: CHOOSING COMPENSATORY RESTORATION OPTIONS........................................................................39FIGURE 5-2: DAMAGE, INTERIM LOSSES AND RESTORATION OPTIONS..................................................................40FIGURE 5-3: PRIMARY AND COMPENSATORY RESTORATION SCALING COMPONENTS OF PANTHER CREEK...........46FIGURE 7-1: LOCATION OF DOÑANA NATIONAL PARK ..........................................................................................53FIGURE 8-1: PROTECTED AREAS OF SOUTH-WEST WALES......................................................................................66

TablesTABLE 3-1: IMPACTS ON LAGOONS OF MINERAL EXTRACTION AGAINST CSAC CONSERVATION OBJECTIVES .....16TABLE 3-2: MATRIX FOR ASSESSING SIGNIFICANCE OF IMPACTS ON NATURA 2000 SITES AGAINST

CONSERVATION OBJECTIVES .............................................................................................................17TABLE 4-1: CONDITION TABLE FOR TILIO ACERION RAVINE WOODLAND IN THE UK...........................................25TABLE 4-2: OPTIONS FOR PRIMARY RESTORATION OF NATURAL RESOURCE DAMAGE .........................................29TABLE 7-1: ANNEX I HABITATS PRESENT WITHIN THE DOÑANA NATIONAL PARK PSCI.......................................57TABLE 7-2: ANNEX I BIRD SPECIES PRESENT IN SIGNIFICANT NUMBERS (>P) WITHIN THE DOÑANA NATIONAL

PARK PSCI .........................................................................................................................................58TABLE 7-3: CASE STUDY DAMAGE ASSESSMENT OF AZNALCOLLAR MINE SPILL ON ANNEX I HABITATS IN THE

DOÑANA NATIONAL PARK AND CORREDOR ECOLOGICO DE RIO GUADIAMAR NATURA 2000 SITES.59TABLE 7-4: CASE STUDY DAMAGE ASSESSMENT OF AZNALCOLLAR MINE SPILL ON ANNEX II SPECIES IN THE

DOÑANA NATIONAL PARK AND CORREDOR ECOLOGICO DE RIO GUADIAMAR NATURA 2000 SITES.60TABLE 7-5: CASE STUDY DAMAGE ASSESSMENT OF AZNALCOLLAR MINE SPILL ON ANNEX I BIRDS IN THE

DOÑANA NATIONAL PARK AND CORREDOR ECOLOGICO DE RIO GUADIAMAR NATURA 2000 SITES.61TABLE 8-1: CASE STUDY DAMAGE ASSESSMENT OF SEA EMPRESS OIL SPILL ON ANNEX I HABITATS IN THE

PEMBROKESHIRE MARINE AND CARMARTHEN BAY AND ESTUARIES NATURA 2000 SITES ...............69TABLE 8-2: CASE STUDY DAMAGE ASSESSMENT OF SEA EMPRESS OIL SPILL ON ANNEX II SPECIES, SPA AND

BIRD POPULATIONS IN THE PEMBROKESHIRE MARINE AND CARMARTHEN BAY AND ESTUARIESNATURA 2000 SITES ..........................................................................................................................70

TABLE 8-3: SUMMARY OF TOTAL COSTS RESULTING FROM SEA EMPRESS OIL SPILL (£ MILLION)..........................74TABLE 9-1: USE OF PAYMENTS MADE BY EXXON AS PART OF CIVIL AND CRIMINAL SETTLEMENTS .....................79

BoxesBOX 3-1: BLACKBIRD MINE CASE STUDY - SITE DESCRIPTION...............................................................................9BOX 3-2: BLACKBIRD MINE CASE STUDY - SCALE OF DAMAGE ...........................................................................11BOX 3-3: BLACKBIRD MINE CASE STUDY - IMPACT ASSESSMENT ........................................................................15BOX 4-1: THEORETICAL EXAMPLE OF SETTING PRIMARY RESTORATION OBJECTIVES FOLLOWING DAMAGE TO TILIO

ACERION RAVINE WOODLAND IN THE UK .............................................................................................27BOX 4-2: BLACKBIRD MINE CASE STUDY - PRIMARY RESTORATION OBJECTIVES................................................27BOX 4-3: RATCLIFFE CRITERIA FOR NATURE CONSERVATION EVALUATION ........................................................30BOX 4-4: BLACKBIRD MINE CASE STUDY - SELECTION OF PRIMARY RESTORATION PROJECTS ............................35BOX 5-1: BLACKBIRD MINE CASE STUDY - IDENTIFYING COMPENSATORY RESTORATION PROJECTS ..................45BOX 5-2: BLACKBIRD MINE CASE STUDY - OUTCOME OF THE SETTLEMENT ........................................................47

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Acronyms and Terms UsedBT Benefits TransferCBA Cost Benefit AnalysisCEA Cost Effectiveness AnalysisCERCLA Comprehensive Environmental Response, Compensation and Liability ActCM Choice ModellingCR Contingent RankingCV Contingent ValuationDARP Damage Assessment and Restoration ProgrammeDETR Department of the Environment, Transport and the Regions (of the United

Kingdom)EA Environment Agency (of England and Wales)EC European CommissionEEPSEA Economy and Environment Program for South East AsiaEIA Environmental Impact AssessmentEU European UnionEVRI Environmental Valuation Reference InventoryHP Hedonic PricingIOSC International Oil Spill Conference (Seattle, March 8-11, 1999)IPPC Integrated pollution prevention and controlNRDA Natural Resource Damage AssessmentNOAA National Oceanic and Atmospheric AdministrationOPA Oil Pollution ActpSCI Proposed Site of Community ImportanceRC Restoration CostRP Revealed Preference TechniquesSAC Special Area of ConservationSEO Sociedad Española de OrnitologíaSP Stated Preference TechniquesSPA Special Protection AreaSSSI Site of Special Scientific InterestTC Travel Cost MethodTEV Total Economic ValueUSEPA United States Environmental Protection AgencyWTA Willingness to Accept CompensationWTP Willingness to Pay

Report contributorsEce Ozdemiroglu Economics for the Environment Consultancy Ltd, United KingdomTannis Hett Economics for the Environment Consultancy Ltd, United KingdomJonathan Cox MacAlister Elliott and Partners Ltd, United KingdomDiana Tingley MacAlister Elliott and Partners Ltd, United KingdomDonald Reid MacAlister Elliott and Partners Ltd, United KingdomTim Huntington MacAlister Elliott and Partners Ltd, United KingdomNick Hanley The University of Glasgow, United KingdomAlan Randall The Ohio State University, USAEdward Brans Vrije University Amsterdam, The Netherlands

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


EXECUTIVE SUMMARYI. Study Background and ObjectivesVarious options of developing a liability regime for damage to biodiversity or naturalresources were explored in a White Paper (EC, 2000). This study is one of those that arecommissioned to provide further information on different aspects of such a liability schemeand aims to answer three inter-related questions:

• How to define ‘significant’ damage to natural resources and decide on the ‘minimumlevel of restoration’;

• How, and to what extent, monetary valuation techniques can be used to estimate theeconomic value of damage to natural resources; and

• How, and to what extent, cost-benefit analysis can be used to choose between restorationoptions.

Although initially the scope of the study was limited to the damage to Natura 2000 sites, themethodology presented in the study is in principle applicable to other sites of natureconservation value. It is not clear whether the term ‘biodiversity’ or ‘natural resources’ willbe used as the scope of a potential Directive on environmental liability. The term‘biodiversity’ refers to “the variability among living organisms from all sources including,inter alia, terrestrial, marine and other aquatic ecosystems and the ecological complexes ofwhich they are part” (Convention of Biological Diversity, Article 2). On the other hand, theterm ‘natural resources’ refer to “land, fish, wildlife, biota, air, water, groundwater, anddrinking water supplies” (Oil Pollution Act of the USA). The wider-concept of naturalresources is used in this study provided that it does not refer to the privately owned resources,which are covered in the current liability regime.

In trying to answer these questions, the study reviews the relevant literature in Europe and theUSA and the relevant aspects of the litigation experience in the USA, namely the NaturalResource Damage Assessment (NRDA) guidance. It also provides three case studies (basedon the Sea Empress oil spill off the coast of Wales in 1996, the Aznalcóllar mine toxicspillage affecting the Doñana National Park in 1998 and the Exxon Valdez oil spill in Alaskain 1989) to illustrate the framework recommended in the study and the issues raised. A fourthcase study is used throughout Chapters 3 to 5 focusing on the Blackbird Mine pollutionincident in Idaho, USA.

The framework used in this study for assessing the damage to natural resources and choosingbetween different restoration options would be familiar to those using tools for environmentalimpact assessment and/or cost benefit analysis. It is also based on the guidance for NRDA inthe USA (NOAA, 1997). This framework consists of three main steps:

• Damage assessment and significance;

• Primary restoration options, and

• Compensatory restoration options.The rest of this executive summary outlines the main points of importance in each step, andthe role of economic valuation and cost-benefit analysis within the overall framework.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


II. Damage Assessment and SignificanceThis stage of the framework is concerned with the following:

• Definition of the status of the resource prior to the incident causing damage;

• Assessment of the scale of damage;

• Impact assessment; and

• Determining whether damage is ‘significant’.Chapter 3 of the main report provides further details on this stage of the framework.

The status of the resource prior to the incident of concern is also known as the ‘baseline’condition of the resource. The baseline is defined not only in terms of the type and quantityof the resource, but also the services the resource provides. These services could be related tothe ecological functions or the various uses made of the resource. The focus of thisframework is on the uses that are not under private ownership or commercial, since these arealready covered by the existing liability regime.

The inclusion of the concept of ‘services’ in the definition of baseline prepares the ground foreconomic analysis later on, should this be desired, since economic values, or in other words,people’s preferences1 for natural resources are affected by the services provided by theseresources. In this context, economic value can also arise from people’s preferences for theconservation of natural resources that are not related to the use they make of the services butfrom the knowledge that such services exist. This type of value is referred to as passive ornon-use value.

Natural resource services cannot be clearly identified without reference to the populationbenefiting from these services and, hence, impacted by the damage to the natural resources.This population could be the residents in and around the damaged area, visitors to the area, ora group which holds non-use values and which is not necessarily restricted by geographicalboundaries.

Assessment of the scale of damage is another aspect of this step of the framework. Thisassessment identifies and quantifies the damage in terms of: its geographical scale; whetheror not it leads to the damage/loss of habitats and/or species; and whether it is acute (such asan oil spill) or chronic (such as a long-term leakage from a hazardous waste facility).Impact assessment is concerned with the identification and quantification of the impacts ofthe damage on the affected habitat and species population in terms of its geographical scale,and whether or not the impact is temporary (reversible) or permanent (irreversible). Thisstage is crucial both for identifying the restoration options and estimating the economic valueof damage if this is required.

The final step of this stage is determining whether the damage exceeds a significancethreshold, the guidance for which can be found in the EU Habitats Directive. Given the site-specific characteristics of natural resources and damage, it is not possible to be prescriptive.

1 Preferences are expressed as willingness to pay (WTP) to secure an improvement or to avoid a degradation aswell as a willingness to accept compensation (WTA) to forgo an improvement or to suffer a degradation.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


For Natura 2000 sites, significance can be judged by assessing whether the damage has hadan adverse effect upon the integrity of the site, where integrity can be defined as:

‘The coherence of the site’s ecological structure and function, across its whole area, or thehabitats, complexes of habitats and/or populations of species for which the site is or will beclassified.’A similar approach could be adopted for other sites of national or even regional importance,provided that conservation objectives and attributes that define these have been developed foreach of these sites.

III. Primary Restoration OptionsThis stage of the framework addresses primary restoration, that is, actions that aim to restorethe damaged resource and, if possible, return the resource to baseline (pre-incident)conditions. It is concerned with the following:• Setting restoration targets;• Identifying primary restoration options;• Selecting primary restoration options; and• Estimating interim losses.Chapter 4 of the main report provides further details on this stage of the framework.

The guidance on NRDA in the USA sets the target for the primary restoration options asrestoring the resource to its pre-damage status, in other words, the baseline. Guidance for thiscan also be found for Natura 2000 sites. For example, in the UK ‘favourable (conservation)status’ tables are used to determine the characteristics of the Natura 2000 sites.

Although technical options for primary restoration are numerous, it is possible to group themin four categories:

• No intervention: Where possible, restoration should seek to assist natural processes torestore the damage inflicted on habitats or species populations. A non-interventionapproach might be appropriate where sites are particularly sensitive to machinery andfurther physical disturbance or are otherwise inaccessible. This can happen, for example,with oil pollution damage to salt marshes, where the soft nature of the substrate andsensitivity of the vegetation to oil-dispersing chemicals makes it difficult or impossible toremove oil from such habitats. The best option in such circumstances is, therefore, toleave the oil to degrade naturally.

• Limited level of intervention: In other instances, limited intervention will be appropriateto restore ecological relationships. This might, for example, involve planting grasses,trees or shrubs that provide an improved structure to allow for the natural re-colonisationof other elements of the habitat that have been lost.

• Full-scale reconstruction: This might include intensive removal of contaminants,replacement of soils, replanting of habitats and re-introduction of species populations.Such full-scale restoration projects are uncommon and there is always a danger of tryingto re-create a facsimile of the lost habitat that will be false.

• Monitoring and surveillance: coupled with all restoration strategies is a need to monitorhabitats and species populations to ensure that restoration targets are met. Techniques formonitoring will vary from site to site and may include remote sensing from satelliteimagery or air photography through to detailed vegetation and species surveys usingstandard ecological monitoring techniques.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Selection of primary restoration options should be the result of an evaluation process basedon, but not limited to, the following criteria:

• The cost to carry out the option;

• The length of time it will it take for the restoration to be effective;

• The extent to which each option is expected to return the damaged resource to itsbaseline;

• The likelihood of success of each option;

• The extent to which each option will prevent future damage (flowing from the initialincident), and avoid collateral damage as a result of implementing the option;

• The extent to which each option generates benefits for the damaged and/or other naturalresources beyond returning the damaged resource to its baseline; and

• The effect of each alternative on public health and safety.If there is more than one primary restoration option that can achieve the restoration target,and they are equally preferable according to ecological criteria, the option with the least costshould be chosen. The process of choosing the least cost option is known as cost-effectiveness analysis (CEA) and involves the comparison of the present value (discounted)of the costs of the options, where costs include those for undertaking damage assessment andimplementing restoration (such as cleaning, and species and habitat restoration).

In some cases, the cost of the chosen primary restoration option may be deemed to be“excessive”. Without a benchmark against which the costs can be compared, it is not possibleto decide whether the costs are indeed excessive or not. This benchmark is the benefits ofprimary restoration. The benefits of restoration can be defined as the restored ability of thedamaged resource to provide the services mentioned above (see Section V below for anoverview). The process of measuring and discounting the costs and benefits of an option andcomparing costs and benefits of an action is referred to as the cost-benefit analysis (CBA). Ifthe option passes the cost-benefit test (i.e. the benefits of the option exceed the costs), then itshould be implemented.

Finally for this step, interim losses refer to the reduction in natural resources and the servicesthey provide, relative to baseline, which occur from the onset of an incident until completerecovery of the injured resources. Even where full recovery is possible, interim losses ofnatural resources and the services they provide will occur, for the simple reason that recoverycannot happen instantaneously. Note that interim losses occur over an infinite time period ifprimary restoration is not possible, or the damage is irreversible. The objective ofcompensatory restoration (Section IV below) is to compensate the public for these interimlosses. Therefore, the identification and quantification of these interim losses, which areinextricably linked to the primary restoration option chosen, are fundamental to the selectionof compensatory measures.

IV. Compensatory Restoration OptionsThis stage of the framework is concerned with the following:• Setting the objectives for compensatory restoration options;• Monetary compensation and/or resource compensation;• Identifying the compensatory options; and• Selecting the compensatory options.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Chapter 5 of the main report provides further details on this stage of the framework.

The objective of compensatory restoration is to compensate the public for the loss ofresources and services (interim losses) which are incurred during the recovery period asdefined above. Compensatory restoration may take place at the damaged site or off-site.Together, primary and compensatory measures aim to fully compensate the public for thedamage incurred. Providing compensation, whether it be in monetary or resource terms,involves some degree of substitution between resources. Compensatory restoration projectswill, at the very least, involve trade-offs over time, i.e. resources or services are provided inthe future to compensate for resources or services which are lost today. However, other typesof substitutions may well also be involved: in space (projects may take place in a differentgeographical location); in the types of services provided (e.g. ecological functions orrecreational opportunities); in the populations who gain services from restoration projectscompared to those who lose from the damage.

Considering that technical options for resource compensation can be numerous, the NRDAguidance in the USA (NOAA, 1997) develops four classes of technical options forconsideration:

Class I: Same type, same quality and comparable value;

Class II: Same type, same or different quality and not of comparable value;

Class III: Comparable type and quality; and

Class IV: Not of comparable type and quality.

Type, in this context refers to the type of the damaged resources or services. Considerationshould also be given to the capacity of the resource to generate the compensatory servicesand the opportunity to provide these services. For example, compensation for the damagedservices of a beach should be by enhancing or creating another beach rather than creating,say, an inland lake. This is where identifying the population affected by the initial damage isalso crucial.

The most desirable category of compensatory restoration options is class I, while class IV isnot recommended for implementation. A number of different approaches can be used todesign, select and determine the scale compensatory restoration options, which arediscussed below:

• Service-to-service approach: This approach is only suitable for Class I options, as itassumes that the public would be willing to accept a one-to-one trade-off between theservices that are lost due to damage and the services that are created throughcompensatory restoration. It only seems reasonable to make this assumption if thereplacement resources are of the same type, quality and of comparable value. Theapproach requires the identification of the services lost in the interim and then designs thecompensatory restoration option accordingly. Parameters to take into account include thetime when the restoration project begins, the time until the project provides full services,the productivity of the project through time, the relative productivity of the created orenhanced resources and services compared to the damaged resources and services, andthe population affected by resource losses compared to the population that gains fromcompensatory restoration. The size of the appropriate compensatory option is determinedby equating the present discounted value of services gained from restoration to thepresent discounted value of interim losses. If there are more than one option that canprovide the service-to-service match, the one with the least cost should be chosen (CEA).

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


• Value-to-value approach: This approach can be used for scaling of Class II and IIoptions, i.e. when the assumption of a one-to-one match between lost services andcompensatory services is not necessarily valid. The approach estimates the economicvalue of interim losses and the economic value of the services generated by thecompensatory restoration option. The scale of the restoration option is then determinedsuch that the compensatory services provided are of equal economic value to the interimlosses. This approach requires the use of economic valuation techniques to express thelosses and gains in either resource or monetary terms. If there is more than one option thatcan provide the value-to-value match, the one with the least cost should be chosen (CEA).

• Value-to-cost approach: Within this approach, restoration is scaled by equating the costof the restoration plan to the value (in monetary terms) of losses due to the injury. Thisapproach is only appropriate when the valuation of lost services is practicable, but thevaluation of the replacement natural resources and services cannot be performed within areasonable time frame or at a reasonable cost. In general, this approach is only suitablewhen damage is relatively minor.

The experience in the USA shows that typically service-to-service approach is implementedwherever possible, i.e. whenever Class I restoration options are available. When this is notpossible, the damaged and compensatory resources and services are measured and comparedin monetary or resource terms, but it may be possible to implement the value-to-valueapproach with a single survey instrument that elicits economic values for both interim lossesand compensatory services at the same time. When this is not possible due to time andresource restrictions, value-to-cost approach is used.

V. The Role of Economic Valuation and Cost-Benefit AnalysisThe main focus of this study is to discuss the potential role of economic valuation methodsand the potential role of CBA within a liability regime. It is important to note that the twohave related but separate roles. The principles of economic valuation are used to define thebaseline (the resource-service-value link mentioned above) and to estimate the value of theinterim losses, even if a CBA framework is not used for choosing between (primary orcompensatory) restoration options. On the other hand, if CBA is used, economic valuationmethods need to be implemented to estimate the benefits of restoration unless there is a goodjustification for using non-monetary expressions of benefits.

The economic valuation techniques include: (i) stated preference techniques which rely oncarefully structured surveys to elicit people’s preferences about natural resources; and (ii)revealed preference techniques which use data from selected actual markets (in this contextespecially recreational behaviour) to extrapolate people’s preferences for natural resourceswhich are assumed to be reflected in these actual markets. When it is not possible toimplement an original valuation study, estimates from the relevant literature can be borrowedto use in the context of the damage assessment in hand. This process is referred to as benefitstransfer and is another way to derive monetary expressions of damage to natural resources inthe current context. Finally, if it is not possible to estimate monetary expressions of naturaldamage, scoring and weighting techniques can be used.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The following criteria can be taken into account when deciding whether or not, and if so how,to use different types of valuation techniques (see Chapter 6 and Annex A for details):• Likely magnitude of the damage: The more severe the magnitude of the damage to a

natural resource, the more important it is that the valuation of damage is carried outthoroughly to ensure full compensation. Moreover, in the case of severe damage therequired assessment is likely to be more complex, and it may well be the case that non-use values are affected. Original studies, in particular stated preference techniques, aretherefore likely to be the most appropriate techniques for use in this context.

• Critical importance of the environmental resource impacted, the significance of theimpact and the type of value to be measured: the more important the resource and themore significant the impact, the greater the need for as comprehensive an analysis aspossible. For example, if non-use values need to be estimated, the only techniques ofrelevance are the stated preference techniques.

• Feasibility of compensatory restoration with resources of the same type, same quality andof comparable value: the less similar and the more distant the resources identified forcompensatory restoration, the harder it will be to be reasonably sure that restoration reallydoes provide an appropriate level of compensation without conducting valuation. Wheredamage is relatively severe and the resource concerned is unique or of critical importance,there may be a strong case for a thorough investigation of preferences to provide someassurance that the scale of restoration is appropriate to provide full compensation. Statedpreference techniques such as contingent valuation or choice modelling are likely toprovide the most accurate information for this purpose.

• Applicability: the purposes for which the valuation techniques are implemented determinewhich technique should be chosen. For example, if the purpose is to have ballpark figuresabout use values, benefits transfer could be sufficient.

• Time and data available for analysis: availability of data about the physical measure ofenvironmental impacts is a concern for all valuation techniques. The availability ofeconomic valuation data is typically not a concern for stated preference techniques whichcollect their own data.

• The cost of the valuation exercise depends on the complexity of the damage andrestoration options which affect the data requirements, the complexity of thequestionnaire design, the size of the sample and the complexity of the data analysis.However, the crucial issue here is not the absolute cost of a valuation exercise but itsincremental cost in terms of additional information it provides and the increased accuracyand reliability of the results produced at the end of the assessment process.

• Whether the results of a valuation exercise are legally defensible depends on howstrongly a valuation technique is grounded in theory and how well it is implemented inthe particular study of concern. In general, the fewer assumptions required for theexercise, the more likely the results are to stand up to challenge.

• The fact that the valuation exercise is likely to take place after the incident causingdamage, complex designs would be necessary to account for possible strategic and protestbehaviour of the affected population.

• Differences in the estimates of people’s preferences (WTP and/or WTA) estimatesobtained by different studies have been cause for concern for some. However, in mostcases, such differences are to be expected as they result from different aspects ofeconomic value being estimated or different populations (such as users versus non-users)being covered by the studies. Although some of these differences could be symptomaticof inconsistencies with a study, there are guidelines to ensure that such inconsistencies areminimised (see, for example, NOAA, 1993 and EFTEC, 2001).

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


A similar list of considerations can also be presented for the choice between different levelsof analysis used to choose primary and/or compensatory restoration options (see Annex B fordetails):• The choice between CEA and CBA is largely affected by whether the cost of the

restoration option identified by CEA is deemed to be ‘excessive’. If the cost is notdeemed excessive, then CEA is sufficient. Otherwise, CBA needs to be implemented.

• CEA does not require the measurement of the benefits of restoration so long as therestoration target is identified and agreed. On the other hand, CBA requires the benefits ofrestoration to be expressed in monetary units for direct comparison with the costs ofrestoration.

• Acknowledgement and incorporation of risk and uncertainty attached to differentrestoration options are necessary regardless of whether CEA or CBA is implemented.Some ways in which risk and uncertainty can be dealt with are complex and possibly notfeasible given their information requirements. However, others such as sensitivityanalysis have relatively less information requirements but can add significantly toexplaining the uncertainties and hence improve the quality of the resulting decision.

• The discount rate used for CEA and CBA has been the subject of ongoing debate.Currently, the European Member States use a range of discount rates, ranging from 3% to8%, while the European Commission employs a rate of 4%. There is some evidence thatthe ‘social rate of discount’ is towards the lower end of this range, while the opportunitycost of capital is somewhat higher. While choice of the discount rate to be used inanalysis is ultimately a political decision, for consistency in the implementation of thelegislation across the EU it may be desirable to ensure the rate chosen is consistent acrossMember States. The effect of the chosen rate on the final results may be tested throughsensitivity analysis.

In short, the choice about different levels of analysis is site and event specific and depends onfactors such as the scale of the damage, importance of the damaged resource, the scale of theaffected population and so on. Such factors affect the desired level of accuracy androbustness, and information, time and resource requirements. It is not possible at this stage tomake recommendations that would apply to every possible case in the future.

Finally, the level of difficulty with any analysis depends on the analysts. As with any otherinterdisciplinary work, assessment of damage, choice of restoration options and assessmentof costs and benefits require experts from different disciplines to be involved in the process.A minimum requirement would be ecologists, economists and legal professionals.

VI. Report StructureThis report consists of nine chapters and seven annexes:Chapter 1 outlines the study background and objectives;Chapter 2 presents an overview and approach to the relevant issues;Chapter 3 discusses the steps involved in damage assessment and determination ofsignificance of damage;Chapter 4 discusses the design and selection of primary restoration options;Chapter 5 discusses the design and selection of compensatory restoration options;Chapter 6 summarises the conclusions and recommendations of the study; andChapters 7, 8 and 9 present the case studies of the 1996 Sea Empress oil spill off the coastof Wales, the 1988 Aznalcóllar mine toxic spillage affecting the Doñana National Park andthe 1989 Exxon Valdez oil spill in Alaska, respectively.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Annex A outlines the concept of economic value of damage and presents the variouseconomic valuation techniques;Annex B provides details about cost-benefit analysis, cost effectiveness, and multi-criteriaanalysis, and how they may be used in practice to decide between compensatory options;Annex C discusses the differences between economic valuation techniques and expectationsof the differences in monetary estimates obtained using different techniques;Annex D outlines the guidelines for and experience with NRDA in the USA;Annex E provides copies of the Standard Data Form used for Natura 2000 sites; andAnnex F presents the references used in the report and annexes.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B

Final Report MEP and EFTEC Page 1

1 STUDY BACKGROUND AND OBJECTIVES

1.1 BACKGROUNDThe European Commission is seeking to establish an environmental liability regime in orderto implement the key environmental principles set out in the EC Treaty and boost theimplementation of EC environmental law. The approach taken, that of ensuring the adequaterestoration of damaged environments, enshrines the basic principle that the polluter shouldpay for remedying the damage that he may have caused, providing that (i) the polluter can beidentified, (ii) the damage quantified and (iii) a causal link established. Various options fordeveloping such a regime were explored in a White Paper2 adopted by the Commission inFebruary 2000. A number of the key elements of the proposed regime are as follows:• No retroactivity.• Damage covered – The regime is intended to apply to “environmental damage” (i.e.

damage to natural resources and damage in the form of contaminated sites) and to“traditional damage” (i.e. injury to health and damage to property).

• Activities covered – It is proposed to link liability to existing EC environmentallegislation, or EC legislation having an environmental objective among others (such asthe IPPC and the revised Seveso II Directives), thus ensuring better implementation of theexisting legislation and acting as a disincentive for poor environmental management.

• Types of liability – Strict liability is intended to apply to damage caused by inherentlydangerous activities and fault-based liability to damage to natural resources caused by anon-dangerous activity. In the case of natural resource damage caused by non-dangerousactivities, if the fault of the causer cannot be established, the State may be responsible forthe restoration costs.

• Liable parties – This is intended to be those legally recognised persons controlling theactivities by which the damage is caused.

• Criteria for natural resource damage – This is the key subject of this study. It is proposedto limit liability for natural resource damage to the Natura 2000 network (i.e. those sitesprotected under the Habitats and the Wild Birds Directives3) where there is significantdamage (to be defined in the study) and where restoration can be undertaken at areasonable cost (also to be defined in the study). The White Paper allows for alternativemeasures to be taken if restoration is not possible. Flexible cost-effective mechanisms forthe valuation of natural resources are required, with ‘benefits transfer’ techniques apossible candidate for investigation.

• Other relevant aspects – (i) where compensation is paid for environmental damage, thereis to be an obligation to spend the money on restoration; (ii) in cases of environmentaldamage, the State should have primary responsibility for acting; (iii) public interestgroups should be entitled to act on a subsidiary basis if the State fails to act, or fails to actproperly; (iv) in urgent cases, public interest groups should be entitled to seek aninjunction against the polluter and carry out preventive measures; and (v) the Commissionintends to clarify whether the proposed regime could apply to areas which are alreadycovered by international law (such as in the case of liability and compensation for damagecaused by oil spills at sea).

2 White Paper on environmental liability COM (2000) 66, dated 9 February 2000. Directorate General for theEnvironment, Legal Affairs Unit (DG Environment B3)3 EU Directive on the Conservation of Wild Birds (Council Directive 79/409/EEC) and the EU Directive on theConservation of Natural Habitats and of Wild Fauna and Flora (Council Directive 92/43/EEC)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


1.2 AIMS AND OBJECTIVESThe objectives of the study are to give guidance to the European Commission on:

(i) how to define ‘significant damage’ to natural resources and a ‘minimum level ofrestoration’ in such a directive;

(ii) how, or to what extent, monetary valuation techniques can be used to estimate theeconomic value of biodiversity damage; and

(iii) how, or to what extent, the valuation of damages to natural resources should beincluded in a future directive on liability.

The potential role of economic valuation in this context is twofold. First, it may be used todetermine the total value of environmental damage incurred in a given incident, for thepurposes of determining the maximum amount of financial liability of the responsible party.Second, economic valuation may allow for comparison and evaluation of restoration options.This allows the environmental benefits of different restoration options to be evaluated in thesame terms as the financial costs, thereby facilitating a decision about whether any restorationis desirable and, if so, which restoration option is most desirable.

This study aims to assess the applicability and adequacy of economic valuation techniquesfor use in these contexts. It also aims to assess the process of ‘benefits transfer’, i.e. using theresults of an existing study to estimate the economic value in a different context.

Valuation of damage for the purposes of application of the liability regime becomes relevantabove a minimum threshold, as only ‘significant’ damage is intended to be covered by thelegislation. This study therefore aims to develop criteria to define significant damage.Definition of a minimum level of restoration is also a priority. While the White Paper statesthat ‘restoration should aim at the return of the natural resource to its state before the damageoccurred’, this may often be impossible, or possible only at extreme cost. Considerations suchas ecosystems functions and presumed future use of the resource will therefore play a role inthis process.

The analysis of the aspects detailed above forms the primary goals of this study, and form thesubject of Chapters 2 to 6. The concepts and issues are illustrated through three case studies,in Chapters 7 to 9. Throughout, the report aims to highlight and discuss relevant experiencefrom the USA, in particular with regards to their legislation on liability for and restoration ofdamages to natural resources, and the costs associated with the implementation of thisparticular aspect of their liability regime.

1.3 SCOPEThe purpose of this study is to provide the European Commission with guidance on how, andto what extent, valuation methods capable of estimating damage to natural resources could beincluded in a future directive on liability. It is proposed in the White Paper that the regime belinked to existing and future relevant legislation on the protection of the environment. Themethodology used has therefore been designed to reflect the likely scope of such a regime,such as the confinement to areas already covered under Community legislation to conservebiodiversity, namely the Wild Birds and Habitats Directives.

Whilst it is recognised that some activities, for example natural resource damage resultingfrom the use of genetically modified organisms (GMOs) may need to be covered by theproposed liability regime, these are not covered in this report. For further information on thelikely scope of the regime, the reader is directed to Chapter 4 of the White Paper.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


2 OVERVIEW AND APPROACH TO THE ISSUES

At present, liability regimes for damage to nature or natural resources do not exist in theEuropean Community to any meaningful extent. The aim of this report is to advise onpossible structures of a liability regime that ensures appropriate compensation for anydamage caused, but which avoids disproportionate costs of natural resource restoration. Inparticular, this report focuses on the potential role for cost-benefit analysis and economicvaluation in this process.

The ideal outcome of a liability regime would be a solution that provides full compensation tothe public for damages to natural resources, at the least cost to the liable party. In principle,there are three possible options to provide compensation for damages to natural resources:

(i) monetary compensation: compensation according to the ‘value of damage’ caused tonatural resources;

(ii) resource (‘in-kind’) compensation: investment in resource restoration projects, whichcould be on-site and/or off-site; and

(iii) mixture of in-kind compensation (partial restoration) and a monetary payment.

The liability regime covering damage to natural resources could be framed either as anobligation to pay for the damage incurred in order to compensate injured parties (monetarycompensation), or as an obligation to compensate these parties ‘in kind’ through restorationand investment in natural resources, both on and off-site (resource compensation). Thisdistinction is important for the remainder of this report, and some implications and furtherconcepts are therefore worth explaining in more detail here.

2.1 MONETARY COMPENSATION BASED ON THE VALUE OF DAMAGEEstimation of the ‘value of damage’ to natural resources for liability purposes can in principlebe made in monetary terms, using economic valuation techniques. This involves identifyingthe injured parties who suffer from the damage, and by how much. Injured parties in thiscontext might include people who ‘use’ the resource directly, for example, for recreationalpurposes. They could also include those who use the resource indirectly: for example, aforest might be used as a source of watershed protection. Finally, and very importantly inthis context, injured parties also include those people who care about the continued existenceof the resource, without actually using it: for example, people who want to preserve aresource for future generations. Economic techniques may be used to estimate the ‘value’ ofthese damages in money terms, as explained in other sections of this report.

The important point to note is that the value of damage, as defined in this sense, isindependent of the costs of cleaning up or restoration after an incident. While the value ofdamage is based on public preferences for an environmental state, costs of clean-up andrestoration are based on the technical options available. It is therefore possible that the valueof damage may be greater than or less than the costs of restoration. One objective of theproposed legislation is to avoid spending on restoration that is disproportionate to the value ofdamage: estimating the economic value of damage may therefore be useful in this context.

If liability is defined to be the value of damage caused and monetary compensation is sought,economic techniques will be needed to estimate this value in monetary terms. Definingliability as an obligation to provide resource compensation may or may not require the use ofeconomic valuation techniques as discussed in the next section.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


2.2 RESOURCE COMPENSATIONCompensation for damages to natural resources can also, in principle, be provided ‘in-kind’through restoration of the damaged resource (primary restoration) and other compensatoryrestoration projects on- or off-site (compensatory restoration). An important aspect of theproposed legislation is the intention to use recovered funds for the purpose of restorationand/or investment in other sites for the conservation of natural resources. If this is how fundsare to be spent – not as direct money compensation – then the problem of ‘compensating thepublic’ becomes one of finding the appropriate level, or scale, of restoration action. Liabilityin this case would be the cost of the appropriate compensatory restoration action, or actions,after the scale of action has been determined.

In determining the appropriate scale of restoration actions, there are several issues toconsider. Liability cannot be framed simply in terms of an obligation to pay for restoration ofthe damaged resource. This is true for several reasons. First, restoration to originalconditions may or may not be possible, even allowing for time and natural recovery. If it isnot possible, the costs of resource restoration to original conditions would be infinite.Therefore, some scope for trade-offs between resources must be taken into account with anyin-kind compensation regime. In other words, both on-site and off-site restoration projectsmay need to be considered.

Second, even where full restoration on-site is possible, this will not fully compensate thepublic for losses due to damage for the simple reason that restoration cannot happeninstantaneously. There will be losses during the recovery period, termed interim losses,which also need to be considered in the required amount of compensation. The concept ofinterim losses also applies where damage is irreparable, the only difference being that, in thiscase, the time period over which losses are incurred is infinite. Where compensation is in-kind, measures over and above restoration to original conditions on-site must be implementedto offset these interim losses. Such restoration measures are termed compensatoryrestoration. Much of the challenge in designing appropriate in-kind compensation packagesis in the estimation of these interim losses and the identification of relevant compensatoryprojects at the appropriate scale. Welfare economic techniques, including monetary valuationtechniques, remain relevant in this process as public preferences for ‘trade-offs’ betweeninjured and restored resources need to be assessed.

2.3 STRUCTURE OF THE REPORTThe structure of this report follows, as far as possible, the ‘time line’ of actions and decisionswhich must be taken after an incident occurs. Possible legal structures and implications androles for cost-benefit analysis are discussed throughout. Note that this report does not dealwith remedial actions such as immediate clean-up of contaminants and other emergencymeasures take prior to restoration options. Figure 2.1 overleaf outlines the framework forassessing damage and choosing restoration options. It also refers to the chapters and annexesof the report that are relevant for parts of the framework. In addition to these chapters andannexes, the report presents three case studies to illustrate the concepts being presented. Thecase studies presented in Chapters 7 to 9, focus on: the 1996 Sea Empress oil spill off thecoast of Wales; the 1988 Aznalcóllar mine toxic spillage affecting the Doñana National Park;and the 1989 Exxon Valdez oil spill in Alaska. A fourth case study (Blackbird Minepollution) is interspersed throughout Chapters 3 to 6. An Annex is also included summarisingthe USA experience of the liability system (Annex D.3).

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Figure 2-1: Damage Assessment and Choosing the Restoration Options

Damage Assessmentand Significance

● Pre-incident resource status● Scale of damage● Impact assessment● 'Significance' of damage See Chapter 3

Primary RestorationOptions

(to restore the initialdamage)

● Setting primary restoration targets● Identification and categorisation of technical options● Selection of restoration options● Estimation of interim losses● Cost-effectiveness and cost-benefit analyses See Chapter 4, Annex A, Annex B and Annex C

CompensatoryRestoration Options(to supplement primaryrestoration options when

these are not sufficient andto compensate for interim

losses)

● Objectives of compensatory measures● Monetary compensation and the value of damage● Resource compensation measures● Classification and selection of compensatory measures● Assessing the scale of compensatory measures● Cost-effectiveness and cost-benefit analyses See Chapter 5, Annex A, Annex B and Annex C

Chapter 3 is concerned with assessment of damage due to an incident, and criteria toestablish whether damage is ‘significant’. The status of the resource prior to the incident ofconcern is defined in terms of its ecological importance, condition, status, and usage. This isimportant in establishing the ‘baseline’ for the analysis as well as identifying what kind ofuse and/or non-use values may be attached to the resource.

Assessment of the scale of damage identifies and assesses the damage in terms of itsgeographical scale, whether or not it leads to the damage/loss of habitats and/or species andwhether or not it is acute (such as an oil spill accidents) or chronic (such as a long-termleakage from a hazardous waste facility). Impact assessment is concerned with theidentification and assessment of the ‘impacts’ of damage on the affected habitat and thespecies population in terms of its geographical scale, and whether or not the impact istemporary (reversible) or permanent (irreversible). This is crucial for the identification of therestoration options as well as measuring the economic value of the impacts. This is followedby a discussion of relevant criteria for a defining a significance threshold.Chapter 4 is concerned with primary restoration options, that is, actions which aim torestore the damaged resource and, if possible, return the resource to baseline (pre-incident)conditions. This chapter outlines how to construct an inventory of possible restorationoptions, and analysis in terms of expected environmental changes and the times at which theyare expected to occur.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Once an inventory of possible options is complete, selecting a course of action may be doneusing various different scientific and/or economic criteria. Possible ways of selectingprimary restoration options, and their advantages and disadvantages are discussed here. Thissection will also focus on the categorisation and estimation of restoration costs, including:the cost of assessing the significance of the damage; the cost of intervention; and the costs offuture monitoring and surveillance against restoration targets.

Different restoration options are likely to have different impacts over time, which will haveimplications for the desirability of each option and also for the size of interim lossesassociated with any given option. Interim losses are likely to be specific to options for otherreasons (e.g. different chemicals used in the clean-up of an oil spill may lead toenvironmental damages themselves, thus affecting interim values). Quantification of interimlosses is discussed here, as this is important for selection of compensatory options (in Chapter5).

Finally, Chapter 4 addresses the potential role of economic criteria in the selection ofprimary restoration options. This includes cost-benefit analysis, cost-effectiveness analysisand scoring techniques.

Compensatory restoration options are the subject of Chapter 5. This chapter looks in detailat possible ways of estimating and rectifying damage, through monetary or resourcecompensation, and the advantages and disadvantages of the different methods available. Theobjectives of compensation are discussed. The chapter is divided into two broad sections.The first section examines resource compensation options: identification of suitable options,methods of comparing the magnitude of the benefits of options with the magnitude of interimlosses, estimation of compensatory restoration costs, selection and scaling of compensatoryrestoration options. The second section deals with estimating the monetary value of damage,the methods available, and the uses of these estimates in terms of option selection ordetermination of compensation.Chapter 6, the last of Part A of the Report provides our recommendations and conclusions.Part B of the report presents three case studies: the Aznalcollar Mine Toxic Waste Spill inSpain, the MV Sea Empress Oil Spill off the coast of Wales and the MV Exxon Valdez OilSpill in Alaska. The case studies follow the structure of the damage assessment andrestoration presented in Figure 2.1.

In addition to the main body of the report, several annexes are dedicated to the discussion ofsome technical issues and supplementary material:• Annex A outlines the concept of economic value of damage in more detail, and presents

the various economic valuation techniques available to estimate this value.• Annex B provides details about various economic decision-making tools, and how they

may be used in practice to decide between compensatory options. These include cost-benefit analysis, cost effectiveness, and multi-criteria analysis.

• Annex C is a technical annex dedicated to a discussion of the differences betweeneconomic valuation techniques and expectations of the differences in monetary valuesestimated using different techniques.

• Annex D outlines the experience with Natural Resource Damage Assessment for thepurposes of liability, the techniques used and insights from the experience in the USAover the past ten years.

• Annex E provides copies of the Standard Data Form for the two European case studies• Annex F presents the references used in the report and annexes.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


3 NATURAL RESOURCE DAMAGE ASSESSMENT ANDRESTORATION

This chapter presents the first step of a damage assessment, namely, the assessment of pre-incident resource status (Section 3.1), and determination of the scale and significance ofdamage (Section 3.2). Section 3.3 discusses some issues regarding the application of naturalresource damage assessments in practice.

The case study of the Blackbird Mine Hazardous Waste Site in the USA is used through thenext three chapters to highlight the points raised and guidance provided. Summaries of thecase study are presented in boxes under the relevant section headings but for a full review ofthe case study, please see Chapman et al, 1998. This study is available from the NOAAwebsite (http://www.darp.noaa.gov/).

3.1 PRE-INCIDENT RESOURCE STATUSThe need to restore natural resource damage defined by this project is limited to damageincurred to sites of European importance to nature conservation. These have been defined byreference to sites identified by Member States for inclusion in the European network of sitestermed Natura 2000. Selection of Natura 2000 sites is undertaken by Member States inconsultation with the Commission. They contain examples of habitats or populations ofspecies that meet criteria identified in that appropriate legislation. Although Natura 2000sites are the ‘jewels’ of nature conservation, the conservation of these most important sitescannot hope to achieve biodiversity conservation as this will rely upon wider measures andthe protection of other sites at a national level. The application of a liability regime fornatural resource damage to Natura 2000 sites is, therefore, a first step in ensuring theconservation and protection of the most important centres for natural resources, but is not apanacea. As a next step Member States may wish to consider how examples of habitats andspecies populations that are of European significance but not included in the Natura 2000 sitenetwork should be considered by a liability regime, or if it should be extended to nationallyimportant sites.

The need to extend the scope of the liability regime can be justified, as there are manyinstances where habitats and species populations are listed as being of European Importancein the Habitats Directive and Birds Directive, but are not included in the Natura 2000network. This can be illustrated by reference to the oak woodlands in the west of the UK.Considerable areas of oak woodland occur on the Atlantic western coastline of the UK thatconform to the Annex I habitat ‘old oak woodland with Ilex and Belchnum of the BritishIsles’. Only the best examples of this woodland type have been selected for inclusion withinthe Natura 2000 network. As a consequence, there are large areas of this internationallyimportant habitat that are outside of the designated area and, hence, would not be protectedby the liability regime as currently envisaged.

EU Directive on the Conservation of Wild Birds (Council Directive 79/409/EEC)The Birds Directive (Article 4 onwards) requires Member States to take special conservationmeasures to conserve the habitat of two specific groups of birds;1. Species listed in Annex 1 of the Directive; and2. Populations of regularly occurring migratory birds.

Member States are required to:

‘classify in particular the most suitable territories in number and size as special protectionareas’ (SPA).

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


EU Directive on the Conservation of Natural Habitats and of Wild Fauna and Flora(Council Directive 92/43/EEC)This Directive, commonly referred to as the Habitats Directive, is closely related to the BirdsDirective, and aims to harmonise the conservation measures for birds with all other species ofwild fauna and flora and natural habitats.Article 2 of the Directive states:

‘the aim of this Directive shall be to contribute towards ensuring bio-diversity through theconservation of natural habitats and of wild fauna and flora within the European territory ofMember States’.Article 3 of the Directive introduces the concept of a network of special areas ofconservation to be termed Natura 2000. This network is to be composed of sites hosting thenatural habitat types listed in Annex I of the Directive and habitats of species listed in AnnexII that are considered of European Importance.

Article 3 also states that the Natura 2000 network will include special protection areas (SPA)classified by Member States under the Birds Directive, so harmonising the protection to birdsand other species with that of natural habitats with the intention to:-

‘enable natural habitats and species habitats concerned to be maintained or, whereappropriate, restored to a favourable conservation status’.This latter clause is important as it places an obligation on Member States not only tomaintain Natura 2000 sites but also, where appropriate, to restore them to favourable status.Article 4 introduces the concept of Priority Habitats and Priority Species. These are habitatsor species listed in Annex I or II of the Directive that are especially endangered andconsequently demand special conservation measures, additional to those adopted for otherhabitats and species listed on Annex I or II.

Information Provided by Member States Relating to Natura 2000 SitesArticle 4 of the Habitats Directive requires Member States to submit a list of proposed Natura2000 sites to the European Commission. For each site, Member States must indicate whichnatural habitat types in Annex I and which species in Annex II the site hosts. Thisinformation is provided on a Standard Data Form (97/266/EC). In addition to the basic listsof habitats and species, the Standard Data Form also provides information on:• the area of habitat within the site and an estimate of the proportion of the national

resource of that habitat the site contains;• the size of the species population within the site and the proportion of the national

population the site contains; and• an assessment of the quality of the habitat and species populations concerned in terms of

their conservation of ecological structure and function and restoration possibilities.

The information contained in the Standard Data Form, therefore, provides a minimum levelof ecological data for all Natura 2000 sites within the Community. Examples are provided inAnnex D of this report. In most instances, this will be supported by much more detailedinformation at a national level, contained in management plans and other monitoring reports.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The consistency of information provided in the Standard Data Form is, however, of vitalimportance in assessing the significance of an impact on a Natura 2000 site, as:1 it provides the basis for the establishment of ‘conservation objectives’ for each Natura

2000 site; and2 it provides a minimum level of information on the relative importance of the habitats

and species populations of European importance that are present, both within the siteand in the national context.

Box 3.1 below describes the Blackbird Mine site and the incident that took place there. Thecase study is particularly interesting in the context of this stage of the assessment since therewas little historical data about the damaged resources prior to the incident.

Box 3-1: Blackbird Mine Case Study - Site Description

Blackbird Mine is an inactive mine site located in east-central Idaho. It is situated in thePanther Creek drainage, a major tributary of the Salmon River, which is a principal sub-basin ofthe Snake River, which flows into the Columbia River. The mine is situated between twodrainage basins, Big Deer Creek and Blackbird Creek, both of which flow into Panther Creek.

Blackbird Mine

State of Idaho

Big Deer Creek

BlackbirdCreek

PantherCreek

SalmonRiver

SnakeRiver

ColombiaRiver

The mine site consists of approximately 360 hectares (ha.) of private land, and about 4 000 ha.of unpatented mine claims held by private corporations on National Forest System lands.Ninety-nine percent of the Panther Creek basin is National Forest, and less than 1% is privatelyowned.

Blackbird Mine Case Study: Incident DescriptionMining of cobalt and copper began at the site in the 1890s and continued until the 1960s.Several studies over the past 25 years document the release of hazardous substances includingcobalt, copper, nickel and zinc, from the mine site, and identify actual or potential sources ofthose releases into Panther Creek and its tributaries.

Sources of hazardous substances: include waste rock and tailings piles, the open pit, road fillcontaining waste rock, dredge spoils, and the underground mine workings.

Releases of substances: occur through erosion and leaching from waste rock and tailings piles;discharges from mine openings, and; discharges of contaminated ground water from seeps andsprings.

In 1992, the State of Idaho initiated a Natural Resource Damage Assessment for the BlackbirdMine, and filed a natural resource damage claim pursuant to CERCLA. Subsequently, theUnited States on behalf of NOAA and the USFS joined suit.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


3.2 DETERMINING THE SCALE AND ‘SIGNIFICANCE’ OF DAMAGE

3.2.1 Scale of DamageAssessing the ‘scale’ of damage in a natural resource damage assessment is primarilyconcerned with identifying the impacts to resources, and to the services they provide. Thisassessment needs to take into account the effect on both ecological services provided by theresource, and the human-related services. Examples of the former include, for example, geo-hydrological functions, production/habitat, end ecosystem integrity; examples of the lattercould include recreation, commercial activities, and health-related services. Figure 3.1provides a summary example of the types of ecological and human services provided by awetland.

Figure 3-1: Wetland Ecosystem Services and Sources of ValueEcological Services Human ServicesGeo-hydrological:• Floodwater storage and conveyance• groundwater recharge and discharge• pollution assimilation• sediment trapping and control• nutrient cycling• shoreline stabilisation

Recreational:• beach use / swimming• fishing, boating• wildlife viewing• hunting

Production/Habitat:• fish and shellfish habitats• habitat for fur-bearers, waterfowl & other wildlife• food production• oxygen production• organic material• timber production• pollination• maintenance of gene pools• maintenance of plant populations

Commercial / public or private:• drinking water• waterway navigation• hydropower generation• irrigation / commercial process water• property protection• agriculture, timber• fishing, trapping, fur-bearers

Ecosystem Integrity:• natural open space• climate regulation• biodiversity storehouse• carbon cycling• resistance and resilience

Cultural / historical:• religious / spiritual uses• cultural uses• historical

Scientific:• pharmaceutical (health)• increase productivity

Health:• morbidity / mortality reductions due to provision

of clean air, water and food

Non-use value:• Species, habitats, ecosystems• Genetic, species diversity and resilience• Life support: carbon/nutrient cycles.

Preliminary assessment of the scale of damages is a necessary step towards determiningwhether damage may be regarded as ‘significant’, and therefore whether a full resourcedamage assessment is required.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Box 3.2 presents the scale of damage in the case of the Blackbird Mine incident.

Box 3-2: Blackbird Mine Case Study - Scale of DamageThe Panther Creek Drainage contains approximately 400 miles of perennial streams andincludes nearly 100 miles of streams suitable for anadramous fish. Highly contaminateddischarge from the mine affects habitat in the lower 25 miles of Panther Creek, and presents apassage barrier that blocks access to remaining upstream habitat.Surface water resources downstream of the mine were found to suffer injury from copper andcobalt releases. The injured resources included surface water, streambed fauna, resident andanadramous fish, ecosystem services and human services. Damage occurred over a numberof years, while mining activities were taking place and after they ceased. For the purposes ofliability, only those losses occurring after 1980, the year that CERCLA was enacted, wereconsidered. Damage to the ecosystem was deemed to be reversible, but only through activeintervention: natural recovery would not be sufficient to return ecosystem services to baselinelevel.

3.2.2 Significance of DamageFundamental to the determination of liability for damage to natural resources is the need toascertain the significance of the damage. This is reflected in the White Paper, which states:

‘There should be a minimum threshold for triggering the regime: only significant damageshould be covered. Criteria for this should be derived, in the first place from theinterpretation of this notion in the context of the Habitats Directive.’

Tests of SignificanceThe significance of impacts is a concept widely used in environmental assessment. Forexample, EU Directive 85/337/EEC on environmental assessment states in Article 1 that ‘thisDirective shall apply to the assessment of the environmental effects of those public andprivate projects which are likely to have significant effects on the environment’. There hasbeen much debate as to how to determine when an predicted environmental impact isconsidered likely to be ‘significant’. In some instances, existing environmental qualitystandards may provide a framework for evaluation of acceptable and unacceptable limits ofpolluting substances in air or water. The application of existing standards for determiningsignificance is, however, not commonly applicable to ecosystem, habitat or speciespopulation impacts, especially as many such standards are determined on the basis of publichealth rather than ecological requirements. Other methods of testing significance may bemore subjective and include considerations of:• the extent and magnitude of the impact;• the duration of the impact, i.e. whether it is short term or long term;• whether impacts are reversible or irreversible;• the sensitivity and rarity of the resources impacted; and• compatibility with environmental policies.In some instances it can be helpful to attach scores to these criteria to indicate levels ofsignificance. Scaling and weighting can be useful in distinguishing between the relativeimpacts of alternative development proposals, for example, the effect three different pipelineroutes would have on ecology. Some examples of scoring techniques, including the Hessianand Andalusian approaches, are presented in Annex B.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


A test of significance can also used to determine the value of a site or region for a particularpopulation of a species or for its representation of a certain habitat. For example, theassessment of bird populations is undertaken using the 1% criterion. This assumes that a sitesupporting more than 1% of an international population of a bird species is of internationalimportance for that species. While there is no fundamental biological reason to take 1% of apopulation as the threshold level, long experience and evaluation have found this percentageto be useful in giving an appropriate degree of protection to populations, and in definition ofecologically sensible sites. The 1% criterion has therefore gained wide acceptance throughoutthe world, as well as with the Contracting Parties of the Ramsar Convention on theConservation of Wetlands of International Importance.

Tests of significance also run through into the identification of sites as candidates forinclusion in the EU Natura 2000 network. For example, populations of species listed inAnnex II of the EU Habitats Directive occurring within a candidate Natura 2000 site areevaluated according to the ratio of the population in the site to that within the nationalterritory. Populations are graded A – D according to the progressive model as defined in theStandard Data Form (97/266/EC):A: 100% ≥ p > 15%B: 15% ≥ p > 2%C: 2% ≥ p > 0%D: non-significant population

where p = the population and the percentage intervals equate to ratio of the population in thesite to that within the national territory.

The concept of ‘significance’ is, therefore, widely used both in assessing environmentalimpact and the valuation of a species populations and representation of habitats within a site.

Significance Tests in the Habitats DirectiveArticle 6 of the Habitats Directive gives guidance on the approach Member States should taketo development proposals affecting a Natura 2000 site. This requires ‘competent nationalauthorities’ to agree to a plan or project that is likely to have a significant effect upon aNatura 2000 site only after having ascertained that it will not adversely affect the integrity ofthe site concerned. Only in exceptional circumstances can such a plan or project beimplemented following a negative assessment of the effects of a proposal. Assessmentsundertaken in this way are termed ‘Appropriate Assessments’ and are becoming anincreasingly important procedure in the development of plans or projects affecting Natura2000 sites. As with environmental impact assessments, appropriate assessments seek topredict the impacts of a proposed development (plan or project) on a known natural resource.This may be a habitat or habitat complex (ecosystem), a population of a certain species or anassemblage of species. In circumstances where insufficient information is available toconduct such an assessment, it is a requirement of the proponent of the plan or project toprovide sufficient information to the competent authority to undertake the appropriateassessment.

Article 6 of the Habitats Directive provides a framework for predicting impacts resultingfrom a proposed plan or project. This same framework could be used to assess thesignificance of environmental damage, for example following a catastrophic pollutionincident, upon a Natura 2000 site. In this situation there it may not, however, be possible togather information on the pre-incident state of the environment. If such an assessment is tobe undertaken it will therefore be necessary to ensure there is a standard level of informationon the size and distribution of species populations and habitats within the Nature 2000 site

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


concerned. This can often be obtained from Management Plans developed for theconservation of individual Natura 2000 sites. However, in the absence of such plans it shouldbe possible to obtain sufficient information on any Natura 2000 sites within the EuropeanUnion by reference to the information each Member State is required to forward to theEuropean Commission on the Standard Data Form (97/266/EC). A copy of such a form forthe Carmarthen Bay and Estuaries candidate SAC is included in Annex E.

The proposed test to determine if environmental damage is considered significant is,therefore, based upon an assessment to determine if the damage has had an adverse effectupon the integrity of the Natura 2000 site concerned. As with assessments made underArticle 6 of the Habitats Directive, it is proposed that the natural resource damage assessmentwould be undertaken in light of the site’s nature conservation objectives.

3.2.3 Establishing Conservation ObjectivesRecently published guidance from the European Commission (EC, 2000) suggests thatconservation objectives for a Natura 2000 site can be determined by reference to theinformation provided on the Standard Data Form (see Annex E). In some instances, MemberStates have published such conservation objectives for Natura 2000 sites. In other instances,it may be necessary to agree these with the national nature conservation agencies prior toundertaking the assessment.

Information in the Standard Data Form will, however, only identify the habitats and speciespopulations which need to be assessed. The test of significant damage also needs to relate tothe effects the damage has on the integrity of the site.

Definitions of Natura 2000 Site IntegrityThe integrity of a Natura 2000 site has been defined in the UK (UK DoE, 1994) as:

‘The coherence of the site’s ecological structure and function, across its whole area, or thehabitats, complexes of habitats and/or populations of species for which the site is or will beclassified.’This definition is useful in that it refers to the site’s ecological structure and function and theneed to conserve these across the whole area of the site. This echoes the information on thestructure and function, and extent of habitats and species populations, provided in theStandard Data Form. It does not, however, provide us with sufficient detail to determine ifthis integrity has been adversely affected.

More detailed information on the assessment of site integrity can be provided by reference tothe definitions of favourable conservation status given in Article 1 of the Habitats Directive.The definitions given in the Directive relate to the status of habitats and species populationsat a national and Community level but can equally be applied to a site. In summary, thisarticle states that the conservation status of a species population can be considered favourableif:

• population dynamics data on the species concerned indicate that it is maintaining itself ona long-term basis as a viable component of its natural habitats;

• the natural range of the species is neither being reduced nor is likely to be reduced for theforeseeable future; and

• there is, and will probably continue to be, a sufficiently large habitat for the species tomaintain its populations on a long term basis.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Conservation status of a habitat can equally be considered favourable if:• its natural range, and the areas it covers within that range, are stable or increasing;• the species structure and functions which are necessary for its long term maintenance exist

and are likely to continue to exist; and• the conservation status of its typical species is favourable, as defined above.These definitions can be used at a site level to enable an assessment to be made of thesignificance of the impact from a damaging incident.

Assessing the Significance of Natural Resource DamageThe assessment of significance can be made by following a set of simple steps.1 Identify habitats and species populations that contribute to a site’s nature conservation

objectives by reference to the Standard Data Form for Natura 2000 sites. Habitats andspecies populations that occur in non-significant levels (graded D) will not form part ofthe conservation objectives. The standard data form used for the collection ofinformation under the Habitats directive may be found in Annex E.

2 Identify the impacts to habitats and species populations resulting from the damageincident. These might include direct or indirect impacts as well as short and long termimpacts. An example of this is the context of the Blackbird Mine case study is given inBox 3.3.

3 Assess the impacts against the conservation objectives of the site to determine if thesehave an adverse affect on the site’s integrity, using the definitions of site integrity givenabove. Ecological features of constituent habitats and species populations that contributeto site integrity can be defined by reference to the definitions of favourable conservationstatus given in Article 1 of the Habitats Directive. The most convenient method formaking such an assessment is through the use of a matrix in which impacts are assessedagainst the features that define favourable conservation status. A real example of such amatrix used in an Appropriate Assessment is reproduced in Table 3-1. This exampleconcerns a mineral extraction and associated waste disposal operation adjacent to a seriesof saline lagoons within a candidate SAC (pSCI) on the south coast of England. The‘conservation objectives’ for the lagoon habitat have been defined by a series ofattributes. These equate to the features that define ‘favourable conservation status’described above and relate to one of three categories:-

a) The size or extent of the habitat;b) The ecological structure of the habitat (in terms of species populations and

dynamics); orc) The ecological function of the habitat.

A theoretical example using a more objective use of impact assessment is produced inTable 3-2 (follows Table 3-1). In this example, impacts can be assessed as being Major orMinor, Short term or Long term, Reversible or Irreversible. They may also be direct orindirect, and cumulative or non-cumulative. The final significance of the impact isdetermined from a consideration of all these elements.

It would be appropriate also to consider if any of the species or habitats affected arePriority habitats or species as defined by the Habitats Directive.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Box 3-3: Blackbird Mine Case Study - Impact Assessment

Injury assessment was quantified by comparing the quality and quantity of resourcescompared to baseline. In this case, some historical data did exist, but was lacking for manyvariables of interest. Baseline conditions were established using both existing historical dataand ‘control areas’, which were unaffected by the discharge but otherwise similar to theassessment areas or resources. Major impacts to resources included injuries to:

Water quality: Exceedence of federal water quality criteria for both copper and cobalt.

Streambed fauna: Streambed fauna downstream of the Blackbird Mine releases containedconcentrations of arsenic, cobalt and copper up to twice the levels of upstream referencesareas. Dramatic reductions in biomass and species composition occurred, indicative ofeffects of impacts due to metals exposure. Injury to streambed components resulted in theloss of food for resident trout, anadramous species and other aquatic animals.

Fish: Fish injury occurred for both resident and anadramous species. Densities of troutupstream from Blackbird Creek (and mining influences) were found to be 35-50 times higherthan densities immediately below, and seven times higher than densities 20 miles below.Laboratory studies determined that copper and cobalt are lethal to trout at concentrations lessthan those measured in the Panther Creek drainage. Furthermore, most Salmon River basinstreams support populations of the anadramous steelhead, and it is likely that Panther Creekalso had a substantial run that was eliminated due to the release of hazardous substances. TheSnake River basin steelhead was listed as a threatened species under the ESA in 1997.Releases from the mine also contributed to the decline, and ultimately the elimination, ofchinook salmon runs in Panther Creek. Historically, Panther Creek supported approximately2,000 spawners annually. Downstream impacts, such as habitat loss, contributed to some ofthe decline. However, it was estimated that, in the absence of impaired water quality, PantherCreek and its tributaries could support runs of 200 adult salmon annually.

Ecosystem services: The loss of chinook salmon in Palmer Creek also represents a loss inecosystem services. Most Salmon River basin streams are relatively nutrient-poor, dependenton the return of salmon and their death after spawning to return nutrients from the ocean intothe headwaters of streams. Salmon carcasses support populations of a variety of mammalsand birds. In addition, as the carcasses decompose, they provide a nutrient base for thestream, stimulating primary production, promoting growth of aquatic plants,macroinvertebrates, small fish, etc., and providing energy for the stream to support anothergeneration of salmon.

Human services: Panther Creek is one of the more accessible streams in the county, and onceprovided recreational opportunities including fishing, birding and picnicking. Salmon fishingwas closed in 1957 to preserve the remaining run. Furthermore, Panther Creek is a streamthat has provided subsistence fishing for Native American Tribes. The Native Americans ofthe region depended on the fish resource as a food, barter and religious base for their culture.

4 A conclusion as to the impact of the damage on the integrity of the Natura 2000 site can bemade by considering the overall outcome of the matrix. As with assessments made underArticle 6 to predict likely impacts of a proposed plan or project, an assessment of damagesignificance using this method will be reliant upon the expertise and impartiality of theauthority undertaking the assessment. Such an assessment will have significant legalimplications and may be open to challenge. It must therefore be robust to externalexamination. To provide the necessary level of transparency, clear guidance as to thestructure and content of such an assessment will be required, and each stage in theassessment process must be documented to provide a clear ‘audit trail’.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 3-1: Impacts on Lagoons of Mineral Extraction Against cSAC Conservation ObjectivesSite attributes which contributeto favourable condition of SAC Impact

Extent of habitat Extent of habitat is currently unchanged although predicteddraw-down of ground water may reduce extent ofPennington and Oxey Lagoons during summer.

Inflow and mix of water from freshand saltwater sources

Current and predicted draw-down of saline ground waterlevels are likely to reduce saline inputs, resulting inincreased surface water drainage into all three lagoons(both from dewatering and natural surface water drainage)with greatest impact upon those nearest to the mineralworkings.

Stability and variation in the extentof lagoonal wetland (water levelvariation)

Water levels are likely to be reduced during the summermonths as reduced ground water levels and minimalfreshwater inputs lead to drying, particularly of Penningtonand Oxey Lagoons.

Salinity variation across the SACand within each lagoon

In winter, reduced ground water levels will lead to reducedsalinity in all three lagoons, with lagoons nearest to themineral workings being most affected. Reduced lagoonlevels in summer may lead to hypersalinity, particularly inPennington and Oxey Lagoon.

Water quality Unlikely to be impact although as mineral extractionapproaches ‘dilute and disperse’ waste disposal sites,possibility of contamination of dewatering discharge withpollutants.

Presence and population levels ofhighly specialised or scarceinvertebrate species

Populations of specialised saline lagoon species havealready been lost from Pennington-Keyhaven Lagoon atleast partially as a consequence of mineral working. Otherpopulations within Pennington and Oxey Lagoons are alsothreatened. Populations of aquatic insect species ofbrackish water may increase but will not be of equivalentvalue as lost species.

Presence and population levels ofhighly specialised or scarce plantspecies

Populations of scarce plant species within the cSAC, suchas Ruppia cirrhosa, would be threatened by markedlyreduced salinity in Pennington-Keyhaven Lagoon.

Conclusion of Assessment: Given the significant contribution these lagoons make to theoverall nature conservation value of the Solent and Isle of Wight Lagoons cSAC, it isconsidered that these impacts constitute an adverse impact to the integrity of the site.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 3-2: Matrix for Assessing Significance of Impacts on Natura 2000 Sites against Conservation Objectives

ConservationObjective

Its natural range and areas itcovers is stable or increasing

Its species structure and functionsexist and are likely to continue to

exist

The conservation status of itstypical species is favourable

Habitats(Annex I)

Impact Nature Significance Impact Nature Significance Impact Nature Significance

Sand dunes Adverse St, R Minor Adverse St, R Minor Adverse St, R Minor

Marine reefs Adverse St, R Minor Adverse St, R Minor Adverse St, R Minor

Intertidal sandflats Adverse St, R Minor Adverse St, R Minor Adverse St, R Minor

Mudflats Adverse Lt, R Major Adverse Lt, R Major Adverse Lt, R Major

Saline lagoons* Adverse Lt, IR Major Adverse Lt, Ir. Major Adverse Lt, Ir. Major

Species(Annex II & SPA)

Population is maintaining itself ona long-term viable basis

Natural range of the species isneither being reduced nor is likely

to be reduced

Sufficiently large habitat remainsto maintain populations on a long

term basis

Otter Adverse St, R Minor None - - Adverse St, R Minor

Atlantic salmon Beneficial St, R None None - - None - -

Populations ofregularly occurringmigratory birds

Adverse St, R Minor None - - Adverse St, R Minor

Key: St = Short term, Lt = Long term, R = Reversible, IR = Irreversible, * = Priority habitats

Definitions of significance usedMinor significance: the impact would have a significant adverse effect on the ecology of the feature, but the level of the effect is such that the resource would be capable ofabsorbing this impact.

Major significance: the impact would have a significant adverse effect on the ecology of the feature. Such an impact would present a measurable long term and permanentthreat to the viability of the resource within the Natura 2000 site.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Weaknesses of the ApproachThe method of assessing significance of damage has deliberately avoided heavy reliance onscoring or ranking of impacts to provide a quasi-objective assessment. Such assessmentsmay appear to be more scientific and hence defensible but, in practice, ranking or scoringsystems are rarely applicable across a range of different sites or circumstances. Although theproposed assessment methodology is heavily reliant upon a more subjective approach, this isclearly founded within a rigorous framework that relates to the ecological integrity of the siteconcerned. To this extent the proposed assessment of damage significance the ‘damageassessment’ is no different from an assessment of a proposed plan or project undertaken aspart of an Environmental Impact Assessment under Directive 85/337/EEC, or an AppropriateAssessment undertaken in accordance with Article 6 of the Habitats Directive. In all theseinstances, the findings of the assessment must be based upon a clear structure and theirconclusions must be sufficiently robust to withstand a legal challenge.

Information Required to Make an Assessment of Damage to Natural ResourcesIn assessing the impact of a proposed development, it is possible to undertake surveys ofhabitats and species populations to determine exactly what areas of habitat are likely to beaffected or what populations of a species using a particular part of a site are likely to beimpacted by a proposed development. In such assessments, the likely impacts on thesefeatures have to be predicted. By contrast, when assessing the impacts of damage to a site,the impacts are mostly evident, although some long-term or cumulative impacts might not beimmediately obvious. Equally, there is much less scope for surveys of habitat and speciesthat have been damaged. To assess the damage, there is therefore a need to have goodbaseline information on the extent and distribution or population size of habitats and speciesfor each Natura 2000 site. Such information is invariably incomplete and will vary betweensites, habitats and species concerned. Reference to management plans and other monitoringand surveillance information may assist in reconstructing the ‘before’ situation. Theinformation and supporting maps that accompany the Standard Data Form provide at least aminimum level of consistent information that can be used to form the basis of such anassessment. There is also likely to be a need to undertake additional field surveys to identifythe former extent of habitat and, where possible, to reconstruct the distribution of habitats andspecies populations that have been impacted.

Setting Restoration TargetsThe assessment of natural resource damage may conclude that the ecological integrity of theNatura 2000 site has been adversely affected. Under an environmental liability regime,targets will need to be set for the restoration of damaged habitats and species populations.Given that Member States are required to maintain or, where appropriate, restore Natura 2000sites to favourable conservation status, this must be a fundamental objective for restoration ofthe damaged ecosystem. Specific restoration targets, and the methods by which these can beachieved, will have to be developed for each individual damage incident or event. It ispossible, however, to determine a process by which such targets can be defined.

Article 3 of the Habitats Directive sets an objective for the Natura 2000 network ofmaintaining and, where appropriate, restoring habitats and species populations of Europeanimportance to achieve favourable conservation status in their natural range. Decisions onwhen to restore habitats and species populations must be based upon an assessment of theircurrent condition against predetermined targets. In the UK this has been done through thedevelopment of ‘favourable condition’ tables for each habitat and species population within

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


each Natura 2000 site. The ‘favourable condition’ used is equivalent to the favourableconservation status referred to in the Habitats Directive and, hence, there is a direct linkbetween the concept of favourable condition and site integrity as defined above. The tablesestablish attributes of the habitat or species population concerned. These correspond to thefeatures by which favourable conservation status and site integrity are defined (extent,ecological structure and function and status of typical species). For each attribute, measuresby which these can be assessed are provided, and specific targets defined. An example ofsuch a condition table is reproduced in Table 3-2.

It must be emphasised that such condition tables have been developed initially to assess thecondition of Natura 2000 sites under current management to form the basis of monitoringprogrammes. In many instances, the result of such monitoring will be a requirement tochange or alter the management of the site to restore favourable conservation status. Thespecific targets provided in the ‘favourable condition’ tables are however equally useful indefining restoration targets for Natura 2000 sites following a damage incident.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


3.3 APPLYING NATURAL RESOURCE DAMAGE ASSESSMENT IN PRACTICEThe Habitats Directive and, more specifically, recent practice in its implementation, provide auseful basis for developing a methodology for assessing the significance of natural resourcedamage where it affects a Natura 2000 site. This is an essential pre-requisite towardsdetermining whether the environmental liability regime will be triggered, making theperpetrator of the natural resource damage liable to pay for its restoration. A similar processcould be used at the national level to assess damage to other sites, habitats or speciespopulations of nature conservation value, providing sufficient baseline ecological data isavailable to undertake a natural resource damage assessment.

Preliminary Assessment of the Likelihood of Significant DamageThis study focuses on the assessment of significant natural resource damage to Natura 2000sites. These are sites which have been submitted by Member States to the EuropeanCommission as proposed Sites of Community Importance (pSCI) or sites that have been fullydesignated as Special Areas of Conservation (SAC) in accordance with Habitats Directive.Sites classified by Member States as Special Protection Areas (SPA) under the Wild BirdsDirective are also included within this definition of Natura 2000 sites.

For all sites within the Natura 2000 network, a baseline of ecological information is availablein the Standard Data Form.

Damage to such sites needs to be ‘significant’ before a liability regime is triggered. Todetermine if damage is significant an assessment needs to be made, but this will only beappropriate in certain instances. It would be impractical to undertake such an assessment forminor damage incidents that affect small parts of a site, or whose impact is likely to be ofshort duration. There are, however, instances where small scale damage might be part of acumulative process, in which case a threshold needs to be determined when a series of smallscale impacts constitute sufficient damage to require a full damage assessment.

The first stage in considering when to trigger the natural resource damage assessment couldtherefore be some consideration of likely significant effect. This will be based on best expertjudgement related to a number of key variables. These might include the following:• The magnitude of the damage inflicted on the site in terms of the area of habitat or

proportion of a species population impacted;• The relative importance of the habitat or species population at a site, national and

international levels including reference to priority habitats and species;• The likely duration of the damage (short or long-term);• The likely response of the habitat or species population to the damage (is it likely to

recover quickly without significant intervention?); and• Whether the damage is part of an ongoing process of cumulative damage to the site or a

one-off event.

The consideration of these aspects of likely significance need to be recorded by the Stateauthorities to demonstrate why a decision to undertake a full natural resource damageassessment was taken or not.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Who Would be Most Appropriate to Undertake the Natural Resource Damage Assessment?Having determined that a full assessment of the damage to natural resources is needed, it maybe reasonable for the appropriate ‘competent authority’ to undertake the assessment. Thiswould provide a degree of impartiality in the assessment and would mirror arrangements formaking an appropriate assessment of proposed plans and projects required under Article 6 ofthe Habitats Directive. The appropriate competent authority may be the state environmentagency or nature conservation department. Funding arrangements will need to be decided, bywhich the cost of obtaining the necessary information to make the assessment can bereclaimed by the competent authority. For instance, this might include the cost of survey torecord the impacts of a damage incident on habitats and species populations of Europeanimportance and the staff time and resources needed to undertake the assessment.

Developing Conservation Objectives for an Assessment of Natural Resource DamageThe assessment of natural resource damage proposed in this regime could be made againstthe site’s nature conservation objectives. These can be developed by:1 reference to the habitats and species populations evaluated as having more than ‘non-

significant’ presence within the Natura 2000 site;2 the application of the definitions of favourable conservation status given in Article 1

of the Habitats Directive; and3 reference to any pre-existing conservation objectives and targets published by national

authorities such as the ‘favourable condition’ tables in the UK.

Information Needed to Undertake a Natural Resource Damage AssessmentWhen assessing the impact of proposed plans or projects, it is possible to undertake detailedecological surveys of a threatened site to determine its value for different habitats and speciespopulations. An assessment is then made of the likely or potential impact of thedevelopment. In assessing a damage incident, the magnitude of the damage may be moreobviously apparent, for example in terms of length of coastline impacted by and oil spill ornumbers of fish or birds killed. Other impacts may be less easily recorded, for example, sub-lethal effects of pollutants on species populations. Assessing the importance of such impactsis, however, dependent upon the availability of pre-existing ecological survey data. Examplesinclude the spatial distribution of Annex 1 habitats within the damaged area, or thepopulations of Annex 1 species using this area for all or part of their life-cycle. If thisinformation has not been gathered in a systematic way then it will be difficult or impossibleto undertake a sound assessment of the significance of the damage. For an environmentalliability regime to be applicable, it will therefore be imperative that Member States undertakeregular surveillance of the conservation status of the habitats and species populations ofEuropean importance within the Natura 2000 network. Undertaking this level of surveillanceis already a requirement of Article 11 of the Habitats Directive. The creation of a liabilityregime further increases the importance of fully implementing this article.

The Process of Making an Assessment of Natural Resource DamagePrior to undertaking a full assessment of natural resource damage, a competent nationalauthority should take a view as to the likely significance of the impact. If it is concluded thatthe impact is likely to be significant, then a full assessment could be undertaken. This test oflikely significant effect should therefore be seen as a ‘coarse filter’ by which obviouslyminimal and insignificant levels of damage are ruled out of the liability regime, unless anaccumulation of small scale damage incidents would satisfy the test of significance.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


If it is concluded that the damage is likely to be significant, a full assessment is triggered. Ifthe conclusion is that there has been significant damage that requires restoration, this couldhave important legal and financial consequences for those responsible for causing thedamage. It will therefore be imperative that such an assessment is seen to be both impartialand scientifically rigorous in its approach. Consideration will need to be given tomechanisms by which the impartiality, transparency and scientific integrity of the assessmentcan be assured.

The mechanism by which the assessment is made could follow the established structure of anEnvironmental Impact Assessment (EIA) as outlined in Directive 85/337/EEC (as amended).For its application to assessing natural resource damage the assessment could thereforecontain the following elements:

1 A description of the damage incident in terms of the type and extent of damage, thelocation of the damage, the area of land or water affected, and the duration of thedamage. As any subsequent liability will fall on the operator of the activity causing thedamage, dates for the commencement and termination of the damage should be given asprecisely as possible. This will assist in determining which of a possible series ofoperators will attract liability;

2 A description of the features of the Natura 2000 site or sites impacted upon, with specialemphasis on those features of European importance as defined by Article 2 of theHabitats Directive; and

3 An assessment of the impact of the damage on the features of European importancewithin the affected site, with special reference to the conservation objectives for thesehabitats and species populations. In certain circumstances, this assessment might alsoinclude an assessment of cumulative damage to a site where small incremental amountsof damage of similar nature have been inflicted to a site over a period of time.

The final stage of an EIA is generally to consider measures which might be implemented toreduce or off-set the impacts of the proposed development. Such mitigation proposals areclearly not applicable in assessing the impact of natural resource damage. However, thiselement of the assessment might be adapted to consider restoration options and methodsneeded to restore the Natura 2000 network to favourable conservation status. This isconsidered in more detail in Chapter 4.

Conclusions of the Natural Resource Damage AssessmentThe conclusion of such an assessment must be to determine if significant damage has beencaused to natural resources in terms of habitats and species populations of Europeanimportance.

Although many impacts may be reversible in the long-term, for example following an oil spillincident, the short term impacts may be severe. Given the wide public concern for the needto ensure proper and relatively rapid action to restore damage to ecosystems, it is likely thateven though damage may be reversible in the long-term, the overall impact is likely to beassessed as being significant.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


4 PRIMARY RESTORATION OPTIONS

This chapter presents the second step of a damage assessment, namely, the identification ofprimary restoration options and how to choose between them. Figure 4.1 below shows aflow-chart outlining the actions need to be taken within this step of the assessment, and refersto the sections of this chapter that provide more discussion about each action.

Figure 4-1: Primary Restoration Process Pathways

Only necessary if restoration targets are not mandatoryand / or there is an opportunity for cost discussions

Identify Primary Restoration TargetsSection 4.1

Identify and Categorise PrimaryRestoration Options

Section 4.2

PrimaryRestoration IS

Possible

PrimaryRestoration ISNOT Possible

Select Primary Restoration OptionUsing Cost-Effectiveness Analysis

Section 4.3

Cost of thechosen

restorationoption is not"excessive"

��

Cost of thechosen primary

restorationoption is

"excessive"

Identify Compensatory RestorationOptions

Chapter 5

Estimate Interim LossesSection 4.4

Assessment ofPrimary RestorationOptions Stops Here

��

Cost-BenefitAnalysis

Section 4.5

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


4.1 SETTING PRIMARY RESTORATION TARGETSThe assessment of natural resource damage may conclude that the ecological integrity of theNatura 2000 site has been adversely affected. Under an environmental liability regime,targets will need to be set for the restoration of damaged habitats and species populations.Given that Member States are required to maintain or, where appropriate, restore Natura 2000sites to favourable conservation status, restoration of the damaged habitat or speciespopulation must be the fundamental objective for restoration of the damaged ecosystem. Thisis termed primary restoration. However, in some instances it will not be possible to restorethe damaged habitat. For example, some habitats such as ancient woodland have evolvedover many thousands of years and have developed a complex relationship between soil types,hydrology, micro-climate and other environmental variables, and the distribution of plant andanimals species. It is technically impossible to recreate such a complex habitat once it hasbeen destroyed. Specific restoration targets and the methods by which these can be achievedwill have to be developed for each individual damage incident or event. It is possible,however, to determine a process by which such targets can be defined.

Article 3 of the Habitats Directive sets an objective for the Natura 2000 network ofmaintaining and where appropriate restoring habitats and species populations of Europeanimportance to achieve favourable conservation status in their natural range. Decisions onwhen to restore habitats and species populations must be based upon an assessment of theircurrent condition made against predetermined targets. In the UK this has been done throughthe development of ‘favourable condition’ tables for each habitat and species populationwithin each Natura 2000 site. The term ‘favourable condition’ used in this example isequivalent to ‘favourable conservation status’ referred to in the Habitats Directive and hencethere is a direct link between the concept of favourable condition and site integrity as definedabove.

The favourable condition tables establish attributes of the habitat or species populationconcerned. These correspond to the features by which favourable conservation status and siteintegrity are defined (extent, ecological structure and function and status of typical species).For each attribute, measures by which these can be assessed are provided and specific targetsdefined. An example of such a condition table is reproduced in Table 4-1 overleaf.

It must be emphasised that such condition tables should be developed initially to assess thecondition of Natura 2000 sites under current management to form the basis of monitoringprogrammes. These should be prepared for all Natura 2000 sites by Member States as part oftheir monitoring and surveillance requirements under the Habitats Directive. In manyinstances, the result of such monitoring will be a requirement to change or alter themanagement of the site to restore favourable conservation status. The specific targetsprovided in the ‘favourable condition’ tables are however equally useful in definingrestoration targets for Natura 2000 sites following a damage incident. See Box 4-1 for anexample using targets taken from the condition table in Table 3-2. Box 4.2 presents asummary of the primary restoration objectives for the Blackbird Mine case.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 4-1: Condition Table for Tilio Acerion Ravine Woodland in the UK(NVC W8, W9 part, Annex I habitat)

Attributes Measures Targets Comments

1. Area Extent/location ofstands

• No loss of ancient semi-natural stands• At least current area of recent semi-

natural stands maintained, although theirlocation may alter.

• At least the area of ancient woodlandretained.

• Stand loss due to natural processes e.g. in minimum intervention stands may beacceptable.

• Stand destruction may occur if the understorey and ground flora are irretrievablydamaged even if the canopy remains intact.

• Loss = 0.5 ha or 0.5% of the stand area, whichever is the smaller.• 20% canopy cover is conventionally taken as the lower limit for an area to be

considered as woodland.• Area and location of stands may be assessed remotely or by site visit.

2. Naturalprocesses andstructuraldevelopment

Age/size classvariation withinand betweenstands; presenceof open space andold trees; deadwood lying on theground; standingdead trees

• At least the current level of structuraldiversity maintained.

• Understorey (2-5m) present over at least20% of total stand area (except inparkland).

• Ground flora present over at least 50% ofarea

• Canopy cover present over 30-90 % ofstand area (except in parkland stands).

• Age class structure appropriate to the site,its history and management.

• A minimum of 3 fallen lying trees >20cm diameter per ha and 4 trees per haallowed to die standing.

• Any changes leading to exceedance of these limits due to natural processes are likelyto be acceptable.

• There is generally a good structural variety in these stands although veteran trees maybe under-represented because of past treatment and the unstable nature of some sites.

• The ground flora may appear sparse in places late in the season where colonies ofAllium have died back. Its composition may be variable ( see attribute 5).

• In coppiced stands a lower canopy cover (of standards) can be accepted, as will alsobe the case in parkland.

• See JNCC guidance note for the sorts of age structure likely to be appropriate fordifferent types of management regime.

• Assess this attribute by field survey.

3.Regenerationpotential

Successfulestablishment ofyoung stems ingaps or on theedge of a stand

• Signs of seedlings growing through tosaplings to young trees at sufficientdensity to maintain canopy density over a10 yr period (or equivalent regrowth fromcoppice stumps).

• No more than 20% of areas regeneratedby planting.

• All planting material of locally nativestock

• No planting in sites where it has notoccurred in the last 15 years.

• A proportion of gaps at any one time may develop into permanent open space; equallysome current permanent open space/glades may in time regenerate to closed canopy.

• Regeneration may often occur on the edges of woods rather then in gaps within it.• The density of regeneration considered sufficient is clearly less in parkland sites than

in high forest; in coppice most of the regeneration will be as stump regrowth. SeeJNCC Guidance Note on likely desirable levels of regeneration.

• The minimum level of regeneration to be acceptable from a nature conservationviewpoint is likely to be much less than that needed where wood production is also anobjective.

• Assess this attribute by walking through the wood in spring/summer.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Attributes Measures Targets Comments

4.Composition

Cover of nativeversus non-nativespecies (alllayers)Death,destruction orreplacement ofnative woodlandspecies througheffects of non-native fauna orexternal unnaturalfactors

• At least the current level of site-nativespecies maintained.

• At least 90% of cover in any one layer ofsite-native or acceptable naturalisedspecies.

• Death, destruction or replacement ofnative woodland species through effectsof introduced fauna or other externalunnatural factors not more than 10% bynumber or area in a five year period.

• In sites where there might be uncertainty as to what counts as site-native or as anacceptable naturalised species this must be made clear (e.g. the position of sycamore).

• Where cover in any one layer is less than 100% then the 90% target applies to thearea actually covered by that layer.

• Factors leading to the death or replacement of woodland species could includepollution, including eutrophication from adjacent farmland; new diseases (Dutch elmdisease where it has not already struck).

• Damage to species by non-native species that does not lead to their death orreplacement by non woodland species (e.g. damage from squirrels to trees that non-the -less survive) is not necessarily unacceptable in nature conservation terms.

• Excessive browsing/grazing by even native ungulates may be considered an unnaturalexternal factor where it leads to undesirable shifts in the composition/structure of thestand, although this may be picked up by attributes 2 or 5 anyway.

• Assess this attribute by a walk through the site.

5. Species,habitats,structurescharacteristicof the site.

Ground flora typeDistinctive anddesirableelements for agiven site e.g..lime, locallyuncommonspecies such asConvallariamajalis; veterantrees or richlichen,invertebrateassemblages.Patches ofassociatedhabitats andtransitions e.g. toalder wood, yewgroves, species-rich grassland

• 80% of ground flora cover referable torelevant NVC community (usually W8,W9)

• Distinctive elements maintained atcurrent levels and in current locations(where appropriate).

• Patches and transitions maintained inextent and where appropriate location.

• Changes leading to these targets not being met may be acceptable where this is due tonatural processes.

• Distinctive elements and patches should be marked on maps for ease of checking inthe field wherever possible.

• If there are species groups/assemblages that cannot be assessed directly on a generalsite visit then surrogate features should be given where possible, e.g. dead woodconcentrations for associated invertebrates.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Box 4-1: Theoretical example of setting primary restoration objectives following damage toTilio Acerion Ravine woodland in the UKIncident: Lorry carrying toxic chemicals crashes on mountain road, causing damage to 6ha of TilioAcerion Ravine woodland on steep slopes below the road.

Damage: Many trees and shrubs physically damaged by impact with lorry. Large area of woodlandsoil saturated in toxic chemicals, destroying woodland ground flora and associated fauna. Majordisruption to woodland structure through loss of canopy. Potentially long term impact to woodlandsoils due to contamination with toxic chemicals.

Primary Restoration Targets:

• Restore structural diversity to woodland. Targets to include:i. Maintenance of at least the current level of structural diversity;ii. Canopy cover presence over 30-90 % of stand area, understorey (2-5m) presence over at

least 20% of total stand area; andiii. Signs of seedlings growing through to saplings to young trees at sufficient density to

maintain canopy density over a 10 yr period (or equivalent regrowth from coppice stumps).

• Restore chemical and physical properties of woodland soils. Targets to include:

i. 80% of ground flora cover referable to relevant NVC community (usually W8, W9);ii. Maintenance of distinctive elements at current levels and in current locations; andiii. Maintenance of patches and transitions in extent and, where appropriate, location.

• Prevent invasion with in-appropriate plant species and soil erosion. Targets to include:

i. Maintenance of at least the current level of site-native species; andii. At least 90% of cover in any one layer of site-native or acceptable naturalised species.

Box 4-2: Blackbird Mine Case Study - Primary Restoration Objectives

The objective of primary restoration is to restore injured resources to their baseline levels. In thiscase it was determined that primary restoration of Panther Creek would be possible, in other words,with the implementation of appropriate restoration activities the damage would be reversible.However, restoration would only be possible over a lengthy time scale.

The trustees selected naturally spawning chinook salmon as the metric for measuring restorationsuccess. This was on the assumption of a high degree of correlation between salmon vitality andoverall ecosystem health, so that as the salmon population is restored, other resources would berestored as well. In fact, other resources would recover on their own shortly after water qualityrestoration, while salmon would not.

Primary restoration options were identified in order to achieve two objectives:

• restoration of water quality: This was the first requirement for restoration, and was necessarybefore any biological restoration could take place; and

• restoration of chinook salmon populations: Following restoration of water quality, optionsaimed at restoring chinook salmon populations to baseline levels could be implemented.

Restoration of water quality is classified as a clean-up activity, and therefore not included in theprimary restoration options, but as a pre-requisite to primary restoration. The EPA is overseeingclean-up at the site, and remedial action is expected to restore water quality by the year 2005.Selection of primary restoration options therefore focused on measures to restore salmonpopulations, to be implemented after this time.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


4.2 IDENTIFICATION AND CATEGORISATION OF TECHNICAL OPTIONSFOR PRIMARY RESTORATION

Should the assessment of damage to natural resources conclude that the impact has beensignificant, the first priority is to seek to achieve primary restoration i.e. restoration of thedamaged habitat. Decisions must be made regarding the method used to restore theecological integrity of the Natura 2000 site. Specific targets and methods employed will varydepending on the nature of the habitats and species populations affected, the nature of thedamage inflicted upon them and the technical options available for restoration. It willtherefore be necessary to design a restoration programme for each significant damageincident. Although the restoration programme must be tailor-made, it is possible to define arange of options that might be applicable.

Where possible, restoration should seek to assist natural processes to restore the damageinflicted to habitats or species populations. At one extreme, this might lead to a decision toleave the damaged habitats completely alone to allow natural processes to restore the damage.This non-intervention approach might be appropriate where sites are particularly sensitive tomachinery and further physical disturbance or are otherwise inaccessible. Attempts to clean-up and restore such habitats might therefore result in greater damage than if they had beenleft alone. This can happen, for example, with oil pollution damage to saltmarshes where thesoft nature of the substrate and sensitivity of the vegetation to oil-dispersing chemicals makesit difficult or impossible to remove oil from such habitats. The best option in suchcircumstances is, therefore, to leave the oil to degrade naturally. However, due to the oftenanoxic conditions found within a short distance from the surface of such soft sediment coastalhabitats, oil can remain trapped within the substrate for a very long time and can becomeenvironmentally ‘active’ again following erosion or other disturbance. In other instances, forexample damage to woodland or forest areas, physical damage resulting in the loss of treeand shrub cover might also be best left to recover naturally. Woodlands are often adapted tosuch periodic catastrophic events and will restore themselves through natural regeneration ifleft alone. There are many examples of damaged woodlands being further damaged by mis-guided restoration projects that have resulted in soil compaction and introduction ofinappropriate species and genetic types.

In other instances, a limited level of intervention will be appropriate to restore ecologicalrelationships. This might involve, for example, the planting of grasses, trees or shrubs thatprovide an improved structure to allow for the natural re-colonisation of other elements of thehabitat that have been lost. The restoration intervention in these circumstances will be theminimum needed to allow natural processes to restore the ecosystem. The planting istherefore intended to provide an ecological framework within which natural process canoperate. This might be appropriate, for example, in the restoration of damage to coastal sanddunes, where some level of ecological stability is required to prevent wind-erosion to thedunes, which might best be provided by planting of marram grass Ammophila arenaria.

In some cases, a much more interventionist strategy will be required where ecologicalfunction has been so disrupted that full-scale reconstruction is needed. This might includeintensive removal of contaminants, replacement of soils, replanting of habitats and re-introduction of species populations. Such full-scale restoration projects are uncommon andthere is always a danger of trying to re-create a facsimile of the lost habitat that will alwaysbe false. Despite this, there are certain habitats that lend themselves to this approach morethan others. For example, damaged river sections can often be fully restored byreconstructing the correct mix of profile, water quality and flow regime. Other relativelysimple habitats can also be fully restored, for example reed bed habitats formed by mono-specific stands of species such as common reed Phragmites communis.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


In all the above scenarios it will be vital to develop and implement a properly planned andresourced monitoring and surveillance strategy to ensure restoration targets are met.

The various options for restoration are summarised in Table 4-2 below:

Table 4-2: Options for Primary Restoration of Natural resource DamageNon-intervention Limited-intervention Full-intervention

• Do nothing and leavenatural process to restorehabitat and speciespopulations

• Habitat management toencourage natural processto restore habitat (e.g.limited planting, controlinvasive species, removecontaminants, nutrientstripping from soil orwater)

• Full clean up• Habitat recreation (e.g.

full planting, soilrestoration/ replacement,reintroduction of species)

Monitoring and surveillance

4.3 CATEGORISATION AND SELECTION OF RESTORATION OPTIONSThe likely effect on restoration strategies of the attribution of liability to the person orpersons responsible for significant natural resource damage must be borne in mind indetermining the appropriate strategy. For example, the legal systems of some Member Statesinclude the principle that loss should be minimised. Accordingly competent authorities mayhave to bear in mind that they may have to justify the strategy pursued on economic as wellas ecological grounds. There will also be a justifiable wish on the part of the party or partiescausing the damage to have their liability quantified definitively at as early a date as possible.Periods of post-restoration monitoring, giving rise to additional ‘long tail’ costs, may meetresistance.

Measures of nature conservation valueIn deciding upon restoration options, it is worth considering the fundamental aspects of a sitethat are considered important to nature and natural resource conservation. One method ofevaluating sites, including habitats and species populations, was developed in the UK byDerek Ratcliffe (1977) in a Nature Conservation Review. This provided an initial assessmentand identification of all the most important wildlife sites in the UK. The criteria used for theselection of these was based upon a set of 10 criteria listed in Box 4.3. These ‘Ratcliffe’criteria have subsequently been adopted by a wide range of organisations and used in anumber of applications, including management planning and site evaluation. Whenconsidering restoration options it is important to take account of these fundamental aspects ofnature conservation evaluation. These are touched upon in greater detail in the followingreview of restoration options.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Box 4-3: Ratcliffe Criteria for Nature Conservation Evaluation

1. Size: In general, larger sites are more highly valued than smaller ones, all else beingequal. Amongst aspects of size to be considered are the relative size of the site comparedwith sites of similar type, the extent of individual components of the site, and whether thesite is of sufficient size that small changes within will not lead to the loss of the site'svalue.

2. Diversity: One of the most important site attributes is variety in numbers of bothcommunities and species, which are usually closely related, and in turn depend largely ona diversity of habitat.

3. Naturalness: Ecosystems least modified by man tend to be rated more highly. However,the vast majority of sites of conservation value have been influenced by man's activitiesto some extent. The degree and nature of this influence should be noted.

4. Rarity: Rarity is concerned with communities and habitats as well as individual species.The presence of one or more rare components on a site, gives it higher value than anothercomparable site with no rarities.

5. Fragility: This reflects the degree of sensitivity of habitats, communities and species toenvironmental change. Fragile sites often represent ecosystems which are highlyfragmented, dwindling or difficult to re-create.

6. Typicalness: The typical and commonplace within a field of ecological variation are alsoof value.

7. Recorded history: The existence of a scientific record of long-standing addsconsiderably to the value of a site. Note should also be made of recorded land-usepractices.

8. Position in an ecological unit: In the event of two sites representing a certain formationbeing of equivalent intrinsic value, the close proximity of one site to a highly ratedexample of another type increases the value of that site.

9. Potential value: Certain sites could, through appropriate management or even naturalchange, eventually develop a nature conservation interest substantially greater than thatexisting at present. Note may also be made of those factors which would limit suchpotential being achieved.

10. Intrinsic appeal: While science may view all creatures as equal, pragmatism dictates thatin nature conservation it is realistic to give more weight to the more popular appeal ofsome species or groups than others.

4.3.1 Categorisation of Restoration OptionsFull InterventionFull restoration intervention of a badly damaged site is likely to be highly expensive. Theresultant habitat may never fully recover although much of its ecological function can berestored. Conservation of genetic diversity is considered fundamental to natural resourceconservation. As many Natura 2000 sites support habitats and ecosystems that have evolvedover many thousands of years, the genetic integrity of these sites is of vital importance.Where full intervention is needed, great care will be needed therefore to conserve the geneticintegrity of the damaged site. This might require the development of specific plant andanimal propagation projects in preparation for re-introduction.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Damage to a habitat can also create temporary new habitats that might be exploited byspecies that are of high nature conservation value in their own right. For instance, whereforest cover is damage as a result of ‘acid rain’ the resultant forest clearings may be colonisedby birds associated with such clearings such as the black grouse Tetrao tetrix, a species listedon Annex I of the EU Birds Directive. Restoration of former forest cover might restore thehabitat to a pre-damage state but would result in the loss of the black grouse habitat. In othercases, the cessation of hunting or fishing activity post-damage may also allow speciespopulations to recover. In such circumstances, it will be important to restore the habitat topre-damage condition even if this means some loss of populations that have taken advantageof the damage or management intervention following the damage.

The cost of full intervention options will need to include:• cost of undertaking the damage assessment and preparing restoration and monitoring

strategy;• cleaning: the cost of cleaning in full intervention options can be considerable, for example

removal of oil from contaminated beaches. The ecological advantage of cleaning willvary immensely between sites, but a full cleaning option might be appropriate on coastsediment shores composed of sand or shingle, where habitats are adapted to a high degreeof habitat mobility and are liable to re-establish rapidly after cleaning. On soft sedimentcoasts, including mudflats and saltmarshes, cleaning is both more difficult and can lead toreduced levels of natural restoration;

• costs of species population restoration: this might include captive breeding and re-introduction programmes or simply appropriate habitat restoration, such as creation ofwoodland glade micro-habitats for woodland butterflies or re-introduction of appropriatestock grazing to create specific grassland sward height and structure;

• habitat restoration: this may require propagation of appropriate trees, shrubs, grasses andother plants from locally derived genetic stock, growth of these plants in a suitablenursery environment and replanting. This may need to be combined with appropriate soilrestoration, restoration of hydrology and water quality and, in certain circumstances, mayrequire the re-formation of geomorphological features, such as riverine structures(meanders, pools); and

• cost of implementing the monitoring and surveillance strategy.

Limited InterventionLimited intervention is probably the preferable option following most damage incidents. Itrelies on providing sufficient ecological amelioration to enable natural processes to restorethe damaged habitat or species population. As with full intervention, it will be important toensure the genetic integrity of the site is maintained when resorting to planting or re-introduction of species and might therefore require the development of specific propagationor breeding programmes. Also in common with the full intervention option, there may besome species that benefit from the damaged ecosystem that will decline in population size ordistribution following restoration. Decisions on restoration techniques should however focuson restoring the overall ecosystem balance to pre-damage levels even if this disadvantages toopportunist species. Costs involved will need to be calculated along similar lines to those forfull intervention but are likely to be smaller in magnitude.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


This option may be cheaper, but this is not the reason why such a limited intervention ispreferable to a full intervention option. The most ecologically sustainable and valuablehabitats and species populations are often those that have a high degree of naturalness (seeBox 4-3). By allowing natural processes to restore a habitat around the basic frameworkprovided by limited intervention, the resultant habitat or ecosystem (complex of habitats) willbe of greater nature conservation value than one that has simply a facsimile of the original.Taking the limited intervention option will probably take a longer time to achieve restorationand hence there will be greater interim losses. This would increase the necessary scale ofcompensatory restoration options.

Cost items for limited intervention are likely to be similar to those for full intervention.However, the magnitude of the costs involved is likely to be much reduced. For example,there is unlikely to be a need to undertake captive breeding programmes for speciesrestoration and the amount of plant propagation needed will also be much reduced or evenabsent. Cleaning costs may, however, be as significant or even more significant than in totalrestoration, as in this option the objective is to restore the physical structure of the damagedhabitat to allow natural processes to restore the biological diversity. Clean-up and somephysical habitat manipulation may, therefore, be all that is required, but this will need to bedone to a very high standard. The time needed for natural processes to restore a habitatfollowing limited intervention will, however, be much greater and hence interim losses willincrease. Despite this, limited intervention options have ecological advantages over fullintervention in that the final habitat will be far more natural and hence of higher natureconservation value.

Non InterventionNon-intervention is unlikely to be politically acceptable in some instances, as there will be aneed to show that some action is being undertaken to restore damage to a site. Despite this itmay be that non-intervention is the best long-term restoration option, especially where sitesare inaccessible or fragile. Non-intervention costs are likely to be the least expensive but willinclude:

• cost of undertaking damage assessment and preparing a restoration / monitoring strategy;and

• cost of implementing the monitoring and surveillance strategy.

Although the cost items may be limited, the length of post damage monitoring andsurveillance may be longer than in other options and hence more expensive. The rate atwhich such habitats recover is also likely to be slow and, hence, interim losses and the scaleof compensatory habitat restoration may be greater.

The nature conservation value of habitats that have recovered through natural processes islikely to be higher since planting or importation genetic material, soil or other material can beavoided.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Monitoring and SurveillanceCoupled with all restoration strategies is a need to ensure that habitats and speciespopulations are properly monitored to ensure restoration targets are being met. This willform a fundamental feature of any restoration plan or strategy following a damage incident.Techniques for monitoring will vary from site to site, and may include remote sensing fromsatellite imagery or air photography through to detailed vegetation and species surveys usingstandard ecological monitoring techniques. Such monitoring strategies need to be ofsufficient duration to ensure restoration targets are achieved. There may, however, bejustifiable resistance (from the damaging party) to monitoring that is not well targeted at thedmamged habitats and species populations of European importance. Monitoring techniquesmust also be cost effective. A balance will need to be reached between the value of theinformation obtained and the cost of obtaining it.

To achieve the necessary level of monitoring it is likely that some agreed ecologicalmonitoring protocol will need to be developed. This could provide a structured frameworkwithin which monitoring objectives can be defined, methods of monitoring agreed andsystems for storage and retrieval can be developed – including the need for database designand data ownership.

4.3.2 Selection of Restoration OptionsThe cost of achieving primary restoration will obviously vary immensely depending upon themagnitude of the damage, the complexity of restoration and the restoration option adopted.There are however some fundamental costs that will be common to each of the restorationoptions.

To summarise, the selection of primary restoration options should be the result of anevaluation process based on, but not limited to, the following criteria:• The cost to carry out the option as detailed above for different types of restoration

options;• The extent to which each option is expected to return the damaged resource to its

baseline;• The likelihood of success of each option;• The extent to which each option will prevent future damage (flowing from the initial

incident), and avoid collateral damage as a result of implementing the option;• The extent to which each option benefits more than one natural resource and/or service;

and• The effect of each alternative on public health and safety.These criteria are based on the current experience with damage assessment in Europe and therelevant discussions in the USA Oil Pollution Act of 1990 (Part 990 - Natural ResourceDamage Assessments) but are also typical of standard project appraisal guidance. It is notpossible to have a general rule that says one criterion is more or less important than another,since this is likely to depend on the type and scale of damage and the resource and therelative performance of restoration options against the restoration target.Some combination of the above criteria can be sufficient to select the preferred primaryrestoration option. If the selection procedure concludes that there is more than one option thatmeets the restoration target, and that the options are similar in terms of other selectioncriteria, the final decision could be based on the cost of the option, i.e. selecting the mostcost-effective alternative.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The relevant economic appraisal technique for choosing the best primary restoration option atthis stage is cost-effectiveness analysis (CEA). CEA can be used for two purposes: (1)minimising the cost of primary restoration and (2) maximising the benefit of primaryrestoration. The first purpose implies that the target for primary restoration is predeterminedand that there is more than one option available to achieve this target. In this case, the optionthat achieves the restoration target at the least cost is the best option according to CEA. Thesecond purpose implies that the budget for primary restoration is predetermined and thatdifferent options achieve different levels of restoration. In this case, the option that achievesthe greatest scale of restoration for the given budget is the best option according to CEA.Formally, the two purposes serve the same economic principle: whether the restoration targetis the constraint or whether, instead, the constraint is the restoration budget, we are tocombine the different options so as to get the most for our money, or spending the least toachieve the restoration target.

In the context of natural resource damage, however, the first purpose of using CEA(minimising the costs) seems to be the most relevant since the restoration target is usuallypredetermined based on ecological requirements.

Further details on how to implement CEA are provided in Annex B. Here it suffices tooutline the process. The costs of each such restoration option are estimated over time and thediscounted costs are aggregated. The restoration option that achieves the restoration target atthe lowest cost is chosen.

As Figure 4.1 shows, the assessment of primary restoration options stops here unless the costsare deemed to be "excessive". If this is the case, further analysis is required, which isdiscussed in Section 4.5. Note that the USA legislation for damage assessment makesmeeting the primary restoration target a legal obligation, and hence does not allow adiscussion of the excessiveness of costs. Such discussion is possible for compensatoryrestoration only (see Chapter 5). However, depending on the structure of a liability regime,this procedure may be applicable to primary restoration options as well. Therefore, it isdiscussed in this report for completeness.

Box 4.4 overleaf summarises the selection of primary restoration options for the BlackbirdMine case.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Box 4-4: Blackbird Mine Case Study - Selection of Primary Restoration Projects

Restoration options aimed at restoring chinook salmon populations fall into two categories:re-introduction of naturally spawning salmon into Panther Creek; and smolt survivalactivities to increase the survival rate of smolts (young salmon) within the creek. The twoactivities are interrelated: either action (or set of actions) would not be as successfulperformed independently of the other. In other words, in-stream work to improve smoltsurvival increases the effectiveness of the hatchery, and vice versa. In-stream smolt survivalactivities alone would not be expected to restore baseline until 2150, due to the small strayrate of salmon into Panther Creek. Salmon, by instinct, return to the stream where they werereared to spawn. Salmon re-introduction alone could restore populations to their baselinelevels, but not within any reasonable time frame, and not as cost-effectively as whencombined with smolt survival activities. Thus, it was decided to combine actions from bothcategories, on both cost-effectiveness grounds, and the expected size of interim losses fromimplementation of one activity in isolation. Restoration actions selected were as follows:

1. Salmon re-introduction: Artificial propagation strategies were selected over natural re-introduction strategies in order to achieve a return to baseline within an acceptable timeframe. The plan to restore naturally spawning salmon included:

• trapping adults from selected donor drainage systems: for the first few years, naturalmigrating adults from a selected donor drainage would be trapped;

• an expansion of an existing hatchery: trapped adults would be transported to a hatcheryfor spawning, egg incubation, hatching and rearing to the pre-smolt life stage;

• construction of acclimation ponds on Panther Creek: pre-smolts would be transported tothe Panther Creek system and places in the acclimation pond for grow-out and smolting.Adult salmon are expected to return to Panther Creek 2 to 3 years after smolts arereleased; and

• construction of an adult fish trap on Panther Creek: after 2-3 years, half of the returningadult fish would be trapped in the Panther Creek fish weir and transported to the hatcheryfor spawning, egg incubation, hatching and rearing. The remaining half would beallowed to migrate upstream to spawn naturally.

• The process of trapping and transporting 50% of the adults would continue until thenumber of returning adults reaches baseline conditions, which is projected to occur in2021.

2. Smolt survival activities: This category of restoration entailed increasing the number ofhealthy smolts leaving Panther Creek. Final measures included:

• channel meander reconstruction: to increase available spawning and rearing habitat bydecreasing channel gradients and velocities, and increasing the length of the channel;

• riparian corridor fencing: to restore stream bank stability, riparian vegetation and fishhabitat in areas that are affected by livestock grazing; and

• construction of off-channel rearing habitat: these are designed to protect juveniles andmay be screened to keep out larger fish.

These measures represented those judged most biologically beneficial and cost-effective forrestoration.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


4.4 ESTIMATION OF INTERIM LOSSESInterim losses refer to the reduction in resources and the services they provide, relative tobaseline levels, which occur from the onset of an incident until complete recovery of theinjured resources. Note that if primary restoration is not possible, interim losses occur over aninfinite time period.

The magnitude of interim losses experienced depends on the primary restoration optionselected, and the time required for recovery to take place. Further discussion about interimlosses and restoration actions that need to be taken to compensate for them (compensatoryrestoration) can be found in Chapter 5 and Annex D.

4.5 ROLE OF COST-BENEFIT ANALYSISThe use of CBA is discussed here in the context of primary restoration options. However, itshould be noted that it is equally relevant to the selection of compensatory restorationoptions.

In some cases, the cost of the chosen primary restoration option may be deemed to be“excessive”. Without a benchmark against which the costs can be compared, it is not possibleto decide whether the costs are indeed excessive or not. This benchmark is the benefits ofprimary restoration.

The comparison of costs and benefits of a primary restoration option is referred to as cost-benefit analysis (CBA). CBA is a framework for measuring and discounting the costs andbenefits of an option and comparing the two. Strictly, if the (discounted) benefits of theoption are equal to or greater than its (discounted) costs, the option should be implemented.The opposite is also valid: if the (discounted) costs of the option exceed its (discounted)benefits, the option should not be implemented4. Explicit cost-benefit analysis is notnecessarily a requirement in project selection. However, it should be noted that in choosingto proceed with a given restoration option, one is making the implicit assumption that thebenefits of restoration exceed or equal the costs. For further details on how to undertake aCBA, see Annex B. However, CBA is only one input to the decision-making process. Theremay be other considerations, such as social or political imperatives that result in an optionbeing implemented even if the conclusion from CBA is to the contrary, or vice versa.

But what are the benefits of primary restoration? In short, the benefits are the restored abilityof the damaged resource to provide the services that benefit the public. These services couldbe related to the way the resource was used, such as sale in actual markets (e.g. commercialfisheries), recreation (e.g. angling), or for ecological services (e.g. watershed protection of aforest). The services that are related to the uses of the resource are said to generate usevalues. These ‘uses’ can be on-site as well as off-site (e.g. angling can take place in a wetlandbut also downstream in a river which is regulated by the wetland). The services could alsocome about independent of the way the resource was used, if used at all, since there could bepeople who do not use the resource but who nevertheless would benefit from knowing that itis restored. Such services that are not related to the uses of the resource are said to generatenon-use values. Further discussion on these values, how to identify them is presented inAnnex A.

4 In the language of the economics, the comparison of discounted benefits and costs generate, net discountedbenefits or what is referred to as the Net Present Value (NPV). If benefits are equal to costs, NPV is zero. If thebenefits are greater than costs, NPV is positive. And, if costs are greater than benefits, NPV is negative.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The fact that CBA compares costs and benefits requires both to be expressed in the same unit.Since costs are already expressed in monetary units, benefits must also be expressed inmonetary terms. If there is suitable existing economic valuation literature, benefits transfer5

can be implemented to express natural resource damage (or benefit of restoration) inmonetary units. Although the applicability of benefits transfer is likely to be limited to casesof minor damage and interim losses, it is practical and useful in providing ‘ball park’ figuresdepending on the availability of literature. In the absence of literature suitable for benefitstransfer, an original valuation study may need to be implemented. The choice of possiblemethods is between revealed preference or stated preference techniques. The former can onlyestimate use values and relies on data about consumer behaviour in actual markets. The lattercan estimate both use and non-use values through analysis of data generated by carefully-structured surveys. As Annex A points out, it may not always be possible to express benefitsin monetary units, in which case scoring systems like the ones referred to in Section 4.3above can be used (see Annex B). Annex B also discusses ways to implement CBA whensome benefits are expressed in monetary units and others in non-monetary units.

5 Benefits transfer is the process of borrowing monetary valuation results from existing studies and using themfor the valuation exercise in hand. Some adjustments to the original estimates may be required but not alwaysapplied.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


5 COMPENSATORY RESTORATION OPTIONS

This chapter presents the third step of a damage assessment, namely, the selection ofcompensatory measures. Figure 5.1 overleaf shows a flow-chart outlining the actions need tobe taken within this step of the assessment and refers to the sections of this chapter thatprovide more discussion about each action.

5.1 IDENTIFY THE OBJECTIVES OF COMPENSATORY MEASURESThe objective of compensatory measures is to compensate the public for interim losses,which are incurred during the recovery period. These losses occur for the simple reason thateven if full restoration of resources on-site is possible, restoration cannot happeninstantaneously. The concept of interim losses also applies where damage is irreparable, theonly difference being that in this case the time period over which losses are incurred isinfinite.

Figure 5.2 illustrates the concepts of damage and interim losses. Each graph in the figureplots the value of resource losses or gains on the vertical axis, and time on the horizontal axis.The baseline value of the resource is given by the horizontal line in each figure.

Panel 1 illustrates the case where full recovery of the resource is possible, and primaryrestoration options are available to accelerate recovery to the baseline. By contrast, Panel 2illustrates the case where full recovery is not possible, but primary restoration options areavailable to encourage recovery of the resource to a level above natural recovery. In eachcase, the total damage is given by the present value of areas A plus B in the relevant panel.In the case where primary restoration takes place to accelerate recovery, interim losses aregiven by area A. If, by contrast, primary restoration measures are not implemented, theninterim losses are equal to the total value of damage, i.e. the sum of areas A plus B.

The value of compensation must ideally equal the value of these interim losses. Area C inPanel 3 illustrates the magnitude of gains from a compensatory resource option. Themagnitude, or scale, of compensatory actions should be such that area C is equal to the sumof areas A+B if no primary restoration takes place, or area A if primary restoration does takeplace.

In order to determine the magnitude of compensation that must take place, these interimlosses must be estimated in terms of resource services or money. The benefits ofcompensation, whether compensation is made in monetary or resource terms, should be equalto the magnitude of interim losses. The following sections discuss resource and monetarycompensation, and estimation of trade-offs between resources lost and resources or moniesgained in compensation.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Figure 5-1: Choosing Compensatory Restoration Options

Only necessary if restoration targets are not mandatory and / orthere is an opportunity for cost discussions

Estimate Interim LossesSection 4.4

Identify Compensatory RestorationObjectivesSection 5.1

ResourceCompensation

Section 5.2

Assessment ofCompensatory

Restoration OptionsStops Here

Select and ScaleRestoration Option

MonetaryCompensation

Section 5.3

Mixed Resourceand MonetaryCompensation

Sections 5.2 & 5.3

Estimate the Valueof Damage

Cost of ChosenCompensatoryOption is not"Excessive"

��

Cost of ChosenCompensatory

Option is"Excessive"

��

Cost-BenefitAnalysis

Section 4.5

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Figure 5-2: Damage, Interim Losses and Restoration Options

A B

AB

C

time

valu

e of

reso

urce

time

time

valu

e of

reso

urce

valu

e of

reso

urce

Panel 3: Compensatory restoration option

Panel 2: Primary restoration option where full recovery is not possible

Panel 1: Primary restoration option where full recovery is possible

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


5.2 RESOURCE COMPENSATIONIf the money gained from liability is to be spent on restoration, then gains from restorationshould ideally be equal to losses from injury. The challenge in providing compensation ‘inkind’ through restoration projects is therefore to identify projects which fully compensate forlost resources and services.

Providing compensation, whether it be in monetary or resource terms, involves some degreeof substitution between resources. Compensatory restoration projects will, at the very least,involve trade-offs over time, i.e. resources are provided in the future to compensate forresources which are lost today. However, other types of substitutions may well also beinvolved: in space (projects may take place in a different geographical location); in the typesof services provided (e.g. ecological functions or recreational opportunities); in thepopulations who gain services from restoration projects compared to those who lose from thedamage. In practice it may be desirable to try to minimise these trade-offs, and to provideservices which closely match those which were lost, wherever possible. This has theadvantage of ensuring that those people who suffer from the damage to resources also gainfrom restoration projects, thereby minimising political and distributional issues. However, italso reduces the difficulties of estimating the trade-offs, and ensuring that the scale ofrestoration is appropriate.

The natural resource damage assessment process in the USA has a pronounced emphasis onproviding compensation ‘in kind’ through compensatory resource restoration projects, and onminimising the sorts of trade-offs outlined above as much as possible (NOAA, 1997). Thissection is based largely on the US guidance for selection of compensatory restorationprojects.

5.2.1 Identifying Compensatory Restoration ProjectsIncident-specific restoration targets may be developed by identifying the key characteristicsand quality attributes of the natural resources and services lost due to the incident. Thisinformation is generated in the damage assessment process. Providing compensation forthese lost resources and services is the prime consideration in identifying suitable restorationprojects. The projects may take place on-site or off-site, depending on the opportunitiesavailable and site-specific considerations. Other factors which may be considered include:• the cost of the option;• the extent to which each option is expected to compensate for interim losses;• the likelihood of success of each option;• the extent to which each option will prevent future injury as a result of the incident, and

avoid collateral injury as a result of implementing the option;• the extent to which each option benefits more than one natural resource and/or service;

and• the effect of each option on public health and safety.Longer term compensatory projects might involve more complex restoration of habitats andspecies populations of European importance (listed on appropriate annexes of Habitats andBirds Directives) but which have not been selected for inclusion within the Natura 2000network. This should be considered the least favourable option in most circumstances, asrestoration of such habitat should be undertaken in any event and not be dependent upon theneed to compensate for the loss of the original site.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


5.2.2 Classifying and Selecting Compensatory Restoration ProjectsOnce suitable, feasible restoration projects are identified, the next step in the restorationprocess is project classification and selection. Classification of compensatory restorationprojects involves an assessment of whether they provide services comparable to those lostdue to the injury, in order to minimise the trade-offs incurred. This assessment shouldconsider whether restoration options provide resources and services of the same type andquality and of comparable value to the services lost due to the injury. The classificationcriteria are useful in determining which restoration options are most suitable forimplementation.

The USA NRDA process classifies projects into four possible classes. Starting with the mostdesirable category, they are:Class I: Same type, same quality and comparable value;Class II: Same type, same or different quality and not of comparable value;Class III: Comparable type and quality; andClass IV: Not of comparable type and quality.

The aim of the classification process is to evaluate how well the damaged natural resourcesand services match the replacement natural resources and services on key characteristics andquality attributes. Even when a proposed action provides the same type of natural resourcesand services, a variety of substitutions (in time, space, species, etc) may be unavoidable. Theresult will be differences – in quality, economic value, and in populations who experience theservice losses and those who experience the gains provided by the restoration options.

Considerations in making these judgements include:

Types of resources and services: This involves making a judgement about the comparabilityof resources or services lost and restored. Both ecological services (e.g. hydrological, habitat,nutrient cycling, primary and secondary productivity) and human services (such as recreation,commercial opportunities, cultural/historic use and non-use services) should be considered.In determining whether resources and services are of the same type, consideration should begiven not only to the site capacity to provide these resources, but also whether theopportunity to provide the same type of services exists. For example, will the action increaseeconomic value by either increasing the quantity of uses (services) or enhancing the quality(or reducing the cost of access) of current uses?

If this is the case, then resources may be classified as of the same type. Where restorationwould not provide services of the same type but, for example, complementary to services,they might be considered of comparable type. An example might be restoration options toexpand the range of recreational activities available at the damaged site.

Quality of resources and services: Comparison of the quality of resources and servicesprovided by a restoration option may be done by direct comparison of each attribute orservice. However, in practice, it may be more practical to select a metric, or an index ofmetrics, to quantify services. For example, salmon populations may reflect the health of manyother aspects of an ecosystem. If the metric selected is judged to adequately reflect qualitydifferences, and restoration can be conducted such that quality of resources and services arethe same, then this can be used in the scaling process (see Section 5.2.3) to determine thenecessary amount of restoration which must take place. If, however, the metric does not fullyreflect the quality of resources at the sites, it may be possible to adjust it to reflectdifferences. For example, economic valuation methods may be used to calculate anadjustment factor to capture the greater relative value of the different services provided.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Value of resources and services: One final consideration, for resources/services of the sametype and quality, is whether they are of comparable value. This involves an assessment oftwo potential causes for non-comparable values: differences in the aggregate supply ordemand conditions. Evaluating the possible differences requires the judgement of thecompetent public authority, because the restored services and the future aggregate supply anddemand conditions are not observable when compensatory restoration actions are beingclassified. The smaller the damage and restoration action(s), the less likely it is that thechange in aggregate supply of natural resources is significant, and consequently the lesslikely that the value of the last available unit of natural resources and services will change.

Classification of projects is useful in ranking restoration options in order of desirability, i.e.potential to provide appropriate compensation. Projects of Class I are the most desirable, andshould be considered first, followed by Class II and III. Re-consideration of options may bedesirable if no options of Classes I-III are available. If several options of one class areavailable, then cost criteria may be used to select the most desirable project. A discussion ofcost-effectiveness is given in Section 4.3, with more details provided in Annex B.

5.2.3 Scaling Restoration OptionsThe classification of projects as Class I to IV above is useful in prioritising options in termsof which might be best-suited to providing compensating services of the same type and to thesame populations who incurred losses due to damage. ‘Scaling’ of restoration projectssimply refers to the determination of the appropriate ‘size’ of restoration activities. Thepresent value of benefits gained through restoration should be equal to the present value oflosses due to damage.

This process involves welfare economic considerations, and may or may not involve the useof economic valuation techniques. If restoration projects of Class I are available, then it maybe reasonable to assume that the public is willing to accept a one-to-one trade-off between aunit of lost services and a unit of services provided by the relevant restoration project.Scaling may then be implemented useing the ‘service-to-service’ approach, i.e. determiningthe size of the appropriate restoration actions such that the present discounted value ofservices gained from restoration is equal to the present discounted value of interim servicelosses. Note that this approach is only justifiable for restoration projects of Class I. If noClass I projects are available, then the service-to-service approach may not be used forscaling. Box 5.1 overleaf gives an example of scaling in the context of the Blackbird Minecase study.

Where no restoration projects of Class I are available, then more serious consideration of thetrade-offs between the value of services gained versus the value of services lost is merited.This may be the case where, for example, damage is significant, or where the environmentalresource impacted is of critical importance. The less similar and the more distant theresources identified for compensatory restoration, the harder it will be to be reasonably surethat restoration really does provide an appropriate level of compensation without conductingvaluation. Economic valuation techniques may be very useful for assessing acceptable trade-offs in this context. These techniques may be used to quantify the trade-offs, using eithermoney or resource trade-offs as a metric, to determine the appropriate scale of restorationactions. Details on the full range of techniques available, including their advantages anddisadvantages, are given in Annexes A and C.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


One final approach is the value-to-cost approach, where the restoration actions are scaled byequating the cost of restoration to the value (in monetary terms) of losses due to the injury.This approach is used in the USA where valuation is possible, but would imposeunreasonable time or cost requirements. This may occur, for example, where literature valuesfrom previous research are available to value lost services but are not available to value thegains from restoration actions. Where damages and the scale of restoration required arerelatively minor, this may be a reasonable approach, as it minimises estimation andassessment costs.

Further information, and practical experience in the USA with the use of these techniques isgiven in Annex D. The choice between the three different scaling approaches depends on,among other considerations, the magnitude of the likely damage to the resource, the criticalimportance of the resource impacted, and the range of restoration projects available. Wheredamage is relatively severe, then there is a strong case for a thorough investigation ofpreferences through implementation of either the service-to-service or value-to-valueapproach. If the resource concerned is of unique or critical importance, or if it is not possibleto identify compensatory restoration projects of the same type, quality and value, then theassumptions required to implement the service-to-service approach are not valid. In this case,preferences for trade-offs between damaged and replacement resources may only beconducted through the use of the value-to-value approach.

Other practical considerations, such as the time and cost of implementing each of theseapproaches, are discussed in Chapter 6.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Box 5-1: Blackbird Mine Case Study - Identifying Compensatory Restoration Projects

When identifying possible compensatory restoration projects, in the USA guidelines requiretrustees to first consider restoration actions that would provide services of the same type andquality, and of comparable value, to those lost. Compensatory restoration projects may beimplemented either on- or off-site.

While the trustees preferred restoration within the Panther Creek drainage system as a closerreplacement to the lost resources and services, a lack of available land limited the optionsavailable. Full compensation for interim losses through resource restoration options thereforerequired the trustees to consider projects outside the drainage system. Off-site options wereavailable to enhance the productivity of Panther Creek beyond the baseline level of services,to accelerate the rate of recovery to baseline, and to increase salmon populations in the rest ofthe basin system.

The trustees determined that such restoration projects would provide the same type ofresources and services as those lost. The occurrence of salmon captures the level of servicerestoration since the conditions necessary for salmon vitality (good water quality, adequatemigration, spawning and rearing habitat) are also necessary to support steelhead, residentfishes, streambed fauna and other services lost in the Panther Creek drainage. Therefore,spawning chinook salmon was identified as an appropriate metric to scale compensatoryrestoration projects.

The trustees also identified differences in the quality of services provided by the injury andreplacement resources. Qualitative research was undertaken, to gain information on thepublic’s preferences for wild relative to hatchery chinook salmon. The participants preferredwild salmon to hatchery reared salmon, given the viability and genetic diversity of wildstocks. However, faced with a reduction in wild stocks, participants considered a runrestoration scenario of the type proposed by the trustees, using a hatchery-assistedprogramme to rear wild donor stocks from an adjacent drainage, to be a close substitute towild stocks. As a result, the trustees determined the salmon to be restored by the selectedrestoration methods and wild salmon to be of comparable value.

Given comparability in type, quality and value of the lost and replacement resources, theproposed off-site compensatory restoration project was classified as a ‘Class I’ action. In thiscase use of the ‘service-to-service’ approach for scaling compensatory actions may bejustified.

Measuring Interim Losses / Scaling Restoration ProjectsThe appropriate mix and scale of restoration actions was estimated through a salmon lifecycle model that projects adult returns and smolt outward migrations in Panther Creek as afunction of the restoration actions (R2 Resource Consultants, 1995). The model tracks adultreturns to baseline and the cumulative losses from 1980 in order to estimate interim losses. Adiscount factor of 3% was applied to the calculation of interim losses and restoration gains.The trustees identified the most feasible and cost-effective restoration actions to return thesalmon population to baseline and to equate the present discounted value of restored salmonwith the present discounted value of salmon lost due to the injury. Figure 5-3 overleafillustrates the scaling concept based on the final settlement between trustees and theresponsible parties.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Figure 5-3: Primary and Compensatory Restoration Scaling Components of Panther Creek

0

50

100

150

200

250

300

1981

1991

2001

2011

2021

2031

2041

2051

2061

2071

2081

2091

2101

2111

2121

2131

2141

2151

Year

Adu

lt C

hino

ok S

alm

on

InterimLoss Natural Level of Services

Compensatory Restoration

Prim

ary

Res

tora

tion

Baseline

The metric for injured resources and services is the number of adult chinook salmon returning tospawn annually. Baseline is the level of salmon population given the current downstreamimpediments and current on-site conditions but for the discharge. It is assumed to be constantand equal to 200 adult spawners. Prior to the restoration, the level of services is zero. Servicesbegin to recover with the biological restoration activities, and the life cycle model predicts therecovery trajectory. Initiation of salmon recovery and return to baseline are expected to occur in2005 and 2021 respectively.

With restoration targeted for Panther Creek, compensatory actions were designed to enhance theproductivity of the site beyond the baseline level of services. These compensatory actions werealso intended to accelerate the rate of recovery to baseline. Thus, the compensatory servicesbegin to accumulate in 2021, the same time as baseline is restored.

The major components of the final salmon restoration plan included:• restoration of chinook salmon through rearing the progeny of a suitable donor stock in an

existing Idaho hatchery for release into Panther Creek;• construction of a fish barrier/trap and acclimation ponds to capture returning adults and to

imprint juveniles;• creation of 2 acres of off-channel habitat in Panther Creek to improve juvenile rearing

conditions (100 year project life);• realignment of 1.2 miles of Panther Creek that has been channelised and straightened to

conform to its natural meander pattern; and construction of riparian corridor fencing toexclude livestock (50 year project life);

• fencing 2 miles of private land along Panther Creek to exclude livestock and allowregeneration of riparian habitat, improving spawning and rearing conditions for anadramoussalmonids (50 year project life);

• fencing 8 miles of private lands along other Salmon River basin tributaries to excludelivestock and allow regeneration of riparian habitat, improving spawning and rearingconditions for anadramous salmonids.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


5.2.4 Implications for Monetary Value of LiabilityOnce appropriate restoration actions have been selected and scaled, liability for compensatoryrestoration is simply the cost of implementing these actions, plus the cost of the assessmentprocess. This should be added to the cost of primary restoration and assessment, and anyother relevant items (such as fines) which may be included in the regime.

Box 5-2 below presents the outcome of the settlement in the case of the Blackbird Mine.

Box 5-2: Blackbird Mine Case Study - Outcome of the Settlement

The Consent Decree requires the responsible parties to remediate the mine site and waterquality in accordance with the clean-up programme to be selected by the EPA, and toimplement a Biological Restoration and Compensation Plan (BRCP). The BRCP is designedto restore, enhance and create anadramous salmonid habitat on site-impacted and out-of-basinstreams; fund trustee supervision of the BRCP implementation, and; make cash payments fortrustees’ past damage assessment and response costs.

Under the terms of the settlement, the responsible party agreed to carry out the salmonrestoration plan with trustee oversight. Implementation will proceed over a period of years,with measures in Panther Creek timed to coincide with water quality remediation, which isexpected in 2005. It is estimated that the total cost will be $9 million, excluding damageassessment costs.

5.3 MONETARY COMPENSATIONEstimation of the ‘value of damage’ to natural resources for liability purposes can, inprinciple, be done in money terms, using economic valuation techniques. This processinvolves the estimation of trade-offs between resources lost and resources gained, however inthis case money is the metric used for both measurement and compensation.

An important point to note is that the value of damage, in monetary terms, is independent ofthe costs of cleaning up and restoration after an incident. While the value of damage isbased on public preferences for an environmental state, costs of clean-up and restoration arebased on the technical options available. It is therefore possible that the value of damage maybe greater or less than the costs of restoration. The term ‘value of damage’ here refers to themonetary value which would be estimated using economic valuation techniques (‘value-to-value’ techniques mentioned in Section 5.2). These are further elaborated in Annex A andAnnex C.

If the liable party has the option of either paying the monetary amount of damage orproviding restoration in kind, as outlined in Section 5.2, then the restoration costs will be theupper bound on the value of liability: if the value of damage exceeds primary andcompensatory restoration costs, then the polluter can opt to implement restoration.

In this same context, where restoration costs are considered ‘excessive’ compared to theexpected benefits, estimating the economic value of damages and benefits or of restorationmay be useful. This applies equally to primary and compensatory restoration options. This isimportant, since one objective of the proposed legislation is to avoid spending on restorationthat is disproportionate to the value of damage. Assessment of the costs and benefits ofrestoration options, as outlined in Section 4.5 and detailed in Annex B, may be used in thiscase.

Details on the full range of economic valuation techniques available, including theiradvantages and disadvantages, are given in Annexes A and C.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


6 CONCLUSIONS AND RECOMMENDATIONS

The framework recommended in Part A of this report has seven main steps:1. Assessment of pre-incident resource status (physical quantities of the resource, its

services, what use is made of these services and by whom);2. Assessment of damage (determining the damage and its significance and scaling of

damage);3. Design of primary restoration options (based on the assumption that primary restoration is

undertaken to help the damaged resource return to its pre-incident status)4. Selection of primary restoration option initially using cost-effectiveness analysis (or the

least cost analysis). Since the USA guidance leaves no room for negotiation about thetarget of the primary restoration, cost-effectiveness is the only analysis used for theselection of primary restoration options. However, in order to explore all possible optionsfor the design of a liability scheme in Europe, we also need to consider situations inwhich primary restoration costs are found to be ‘excessive’. If this is the case, costs needto be compared with the benefits of restoration, where benefits are defined as the avoideddamages. This is where cost-benefit analysis and economic valuation techniques can beused.

5. Estimation of interim losses that occur during the time period between the initial incidentand the recovery to pre-incident resource status (or baseline). Note that if the damage isirreversible, the interim losses occur over an infinite time period. Just as with the pre-incident resource status – but possibly to a more detailed extent – interim losses should beestimated not only in terms of the quantity and type of resource damaged, but also interms of what kind of services these resources were providing and to whom. Theexperience in the USA shows that this component requires the most economic input, interms of estimating the interim losses in monetary terms.

6. Design of compensatory restoration options in order to compensate for interim lossestaking into account the quantity and type of the damaged resource and the services itprovides.

7. Selection of the compensatory restoration option in the same way as that for primaryrestoration option: first by applying cost-effectiveness analysis (if interim losses are notestimated in monetary terms), and if costs are deemed excessive, by applying cost-benefitanalysis.

The case studies presented in Part B of this report serve to highlight these steps as much aspossible, given the background information available about them. The distinction is made ineach case study between what action, restoration and compensation was actually made andthat which could be made if the liability regime detailed in this report had been implementedinstead.

The main focus of this study is to discuss the potential role of economic valuation methodsand the potential role of CBA within a liability regime. It is important to note that the twohave related but separate roles. The principles of economic valuation are used to define thebaseline (the resource-service-value link mentioned above) and to estimate the value of theinterim losses, even if a CBA framework is not used for choosing between (primary orcompensatory) restoration options. On the other hand, if CBA is used, economic valuationmethods need to be implemented to estimate the benefits of restoration unless there is a goodjustification for using non-monetary expressions of benefits.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The economic valuation techniques include: (i) stated preference techniques which rely oncarefully structured surveys to elicit people’s preferences about natural resources; and (ii)revealed preference techniques which use data from selected actual markets (in this contextespecially recreational behaviour) to extrapolate people’s preferences for natural resourceswhich are assumed to be reflected in these actual markets. When it is not possible toimplement an original valuation study, estimates from the relevant literature can be borrowedto use in the context of the damage assessment in hand. This process is referred to as benefitstransfer and is another way to derive monetary expressions of damage to natural resources inthe current context. Finally, if it is not possible to estimate monetary expressions of naturaldamage, scoring and weighting techniques can be used.

The following criteria can be taken into account when deciding whether or not, and if so how,to use different types of valuation techniques, i.e. revealed preference, stated preference,benefits transfer and scoring/weighting techniques (see Annex A for further details):• Likely magnitude of the damage: The more severe the magnitude of the damage to a

natural resource, the more important it is that the valuation of damage is carried outthoroughly to ensure full compensation. Moreover, in the case of severe damage therequired assessment is likely to be more complex, and it may well be the case that non-use values are affected. Original studies, in particular stated preference techniques, aretherefore likely to be the most appropriate techniques for use in this context.

• Critical importance of the environmental resource impacted, the significance of theimpact and the type of value to be measured: the more important the resource and themore significant the impact, the greater the need for as comprehensive an analysis aspossible. For example, if non-use values need to be estimated, the only techniques ofrelevance are the stated preference techniques.

• Feasibility of compensatory restoration with resources of the same type, same quality andof comparable value: The less similar and the more distant the resources identified forcompensatory restoration, the harder it will be to be reasonably sure that restoration reallydoes provide an appropriate level of compensation without conducting valuation. Wheredamage is relatively severe and the resource concerned is unique or of critical importance,there may be a strong case for a thorough investigation of preferences to provide someassurance that the scale of restoration is appropriate to provide full compensation. Statedpreference techniques such as contingent valuation or choice modelling are likely toprovide the most accurate information for this purpose.

• Applicability: the purposes for which the above options are implemented for determinewhich option should be chosen.

• Time and data available for analysis: availability of data about the physical measure ofenvironmental impacts is a concern for all valuation options. The availability of economicvaluation data is typically not a concern for techniques that collect their own data.

• The cost of the scaling exercise (whether service-to-service or value-to-value) depends onthe complexity of the damage and restoration options which affect the complexity of thestudy design, the size of the sample and the complexity of the data analysis. However, thecrucial issue here is not the absolute cost of the exercise but its incremental cost in termsof additional information it provides and the increased accuracy and reliability of theresults produced at the end of the assessment process.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


• Whether the results of a valuation exercise are legally defensible depends on howstrongly a valuation approach is grounded in theory and how well it is implemented in theparticular study of concern. In general, the fewer assumptions required for the exercise,the more likely the results are to stand up to challenge.

• The fact that the valuation exercise is likely to take place after the incident causingdamage, complex designs would be necessary to account for possible strategic and protestbehaviour of the affected population.

• Differences in the estimates of people’s preferences (WTP and/or WTA) estimatesobtained by different studies have been cause for concern for some. However, in mostcases, such differences are to be expected as they result from different aspects ofeconomic value being estimated or different populations (such as users versus non-users)being covered by the studies. Although some of these differences could be symptomaticof inconsistencies with a study, there are guidelines to ensure that such inconsistencies areminimised (see, for example, NOAA, 1993 and EFTEC, 2001).

A similar list of considerations can also be presented for the choice between different levelsof analysis used to choose primary and/or compensatory restoration options (see Annex B forfurther details):• The choice between CEA and CBA is largely affected by whether the cost of the

restoration option identified by CEA is deemed to be ‘excessive’. If the cost is notdeemed excessive, then CEA is sufficient. If the cost is deemed excessive, however, thenCBA needs to be implemented.

• CEA does not require the measurement of the benefits of restoration so long as restorationtarget is identified and agreed.

• Strictly, CBA requires the benefits of restoration to be expressed in monetary units fordirect comparison with the costs of restoration. In the event, that monetary expression ofbenefits is not possible, CBA can include both monetary and non-monetary expressionsas discussed above. However, this should only be undertaken if it is proved that monetaryassessment is either not possible or feasible.

• Acknowledgement and incorporation of risk and uncertainty attached to differentrestoration options are necessary regardless of whether CEA or CBA is implemented.Some ways in which risk and uncertainty can be dealt with are complex and possibly notfeasible given their information requirements. However, others such as sensitivityanalysis have relatively less information requirements but can add significantly toexplaining the uncertainties and hence improving the quality of the resulting decision.

• The discount rate used for CEA and CBA has been the subject of ongoing debate.Currently, the European Member States use a range of discount rates, ranging from 3% to8%, while the European Commission employs a rate of 4%. There is some evidence thatthe ‘social rate of discount’ is towards the lower end of this range, while the opportunitycost of capital is somewhat higher. While choice of the discount rate to be used inanalysis is ultimately a political decision, for consistency in the implementation of thelegislation across the EU it may be desirable to ensure the rate chosen is consistent acrossMember States. The effect of the chosen rate on the final results may be tested throughsensitivity analysis.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


As can be seen from the above discussion, the choice about different levels of analysis is siteand event specific and depends on factors such as the scale of the damage, importance of thedamaged resource, the scale of the affected population and so on. Such factors affect thedesired level of accuracy and robustness, and information, time and resource requirements. Itis not possible at this stage to make recommendations that would apply to every possible casein the future.

Finally, the level of difficulty with any analysis depends on the analysts undertaking theanalysis. As with any other interdisciplinary work, assessment of damage, choice ofrestoration options and assessment of costs and benefits require experts from differentdisciplines to be involved in the process. A minimum requirement would be ecologists,economists and lawyers.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


PART BCASE STUDIES

1. Aznalcóllar Mine Toxic Spill2. Sea Empress Oil Spill3. Exxon Valdez Oil Spill

The following Chapters 7, 8 and 9 present case studies focusing on the above incidents. Theyare intended to outline the site, incident and scale of damage using factual informationavailable from current literature. They are then developed, using the liability regime set out inprevious Chapters, to illustrate how the regime could be practically implemented: damageassessments are drawn up on the basis of the limited information available to illustrate howthat process could work and the need for Primary and Compensatory Restoration activities isalso outlined. A comparison is made throughout between events and actions which didactually take place following the incidents; a distinction should therefore be made between thetheoretical application of the liability regime for the benefit of the case studies and eventswhich actually occurred.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


7 AZNALCÓLLAR MINE TOXIC SPILLAGE

7.1 SITE DESCRIPTIONThe Aznalcóllar mine, owned by Boliden Apirsa SL, is one of four operating in the Iberianpyrite (FeS2) belt and extracts zinc, silver, lead and copper from the pyrites. It is located inAndalusia, 35km to the north-west of the city of Seville and 45km north of the DoñanaNational Park.

The Doñana National Park, located between the right bank of the Guadalquivir River and theAtlantic Ocean in the provinces of Huelva and Sevilla, is noted by UNESCO for its greatdiversity of biotopes, particularly lagoons, marshlands, fixed and mobile dunes and scrubwoodland. The National Park and Ramsar site (designated wetland area of internationalimportance) cover an area of 50 720ha6. A peripheral buffer zone covering 26 540hacombines with the National Park to form the declared Biosphere Reserve (77 260ha). The sitehas also been designated as a zone for special protection of birds under EC legislation. Thelocation of the park is shown in Figure 7-1.

Figure 7-1: Location of Doñana National Park

The National Park is one of the biggest heronries in the Mediterranean region and is thewintering site for more than 0.5 million water fowl each year7. The area is home to a numberof important and varied species: 361 bird species recorded with 119 nesting regularly, 29mammals, 19 reptiles, 12 amphibians and 7 fish species, with a further 30 species recorded forthe Guadalquivir estuary. Endangered species include the Mediterranean lynx (Lynxpardelus), the Spanish imperial eagle (Aquila adalberti) and the spoon-bill (Platalealeucorodia). The National Park is managed by the State.

The Doñana Natural Park (112,000ha in total) surrounds the National Park and is managed bythe Junta de Andalucia. It also supports the migrant bird population and other important life-forms and habitats prevalent in the National Park.

6 Protected Areas Program. UNEP World Conservation Monitoring Centre. Doñana National Park.(www.wcmc.org.uk:80/protected_areas/data/wh/Doñana.html)7 Doñana National Park. Spain. Brief description. (www.unesco.org/whc/sites/685.htm)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


7.2 SITE SERVICESThe following services are provided by the Doñana National Park:

ECOLOGICAL SERVICES:

Geo-hydrological:• floodwater storage and conveyance• groundwater recharge and discharge• pollution assimilation• sediment trapping and control• nutrient cycling• shoreline stabilisation.

Production/Habitat:• fish and shellfish habitats• habitat for furbearers, waterfowl and other wildlife• food production• oxygen production• organic material• pollination• maintenance of gene pools• maintenance of plant populations


‘HUMAN’ RELATED SERVICES:

Recreational:• wildlife viewing

Entrance to the Doñana National Park is strictly controlled and zoned (special use, moderate use, restricted useand reserve zones). The Park is protected under law from hunting, drainage, forestry plantation and excessivetourist exploitation.

Commercial / public or private:• none

Cultural / historical:• historical

Health:• morbidity / mortality reductions due to provision of clean air, water and food

Scientific:• none


B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


7.3 INCIDENT DESCRIPTIONDuring the early morning of the 25 April 1998, the contention wall of the storage reservoir formining residues of the Aznalcóllar pyrite mines burst. The contents behind the dam comprisedpyritic sludge and water with metallic compounds (including arsenic, cadmium, zinc, iron,manganese and nickel) in solution and suspension.

As a result of the breach of the dam wall, both sludge and waters spilled into the River Agrio.These passed rapidly into the River Guadiamar, a tributary of the River Guadalquivir. Theavalanche of waste products overflowed the river channels of the Agrio and Guadiamar,spreading over adjacent land and affecting crops and marginal vegetation8. The quantity ofsludge deposited is estimated as 1.98 million m3.

The spill was diverted away from the National Park by a series of hastily-constructed barrierstowards the River Guadalquivir via the Canal de Aguas Mínimas and Brazo de la Torre. Thisescape route was closed five days later9. The contaminated waters remained in the Entremurosarea (a canalised zone of the River Guadiamar, 20km in length and 1km wide) within theDoñana Natural Park), which is one of the most important areas for aquatic birds in the wholearea. Between May and October 1998, quantities of toxic sludge were removed, and theremaining contaminated sediments neutralised by the addition of calcium carbonate andcalcium hydroxide.

7.4 SCALE OF DAMAGENo human lives were lost. The rivers suffered a great reduction in pH and an increase indissolved metals. Apart from wells which were covered by the spill, it seems that groundwaters were not affected10.

The surface area affected by the spill has been estimated at 4,286 ha, of which 1,054 ha areforest, grasslands and saltmarshes. The breakdown is as follows:

Cultivated herbaceous crops 999 haCultivated herbaceous crops under plastic 172 haCultivated fruit trees and olive groves 261 haRice paddies 491 haBrackish marsh grazing 315 haPastures 176 haUncultivated arable land 154 haOther 1,729 haTotal 4,286 ha

Of the 4,286 ha affected, 98 ha lie within the Doñana National Park (affecting 0.19% of theNational Park area) and in total 3.8% of the Doñana Natural and National Parks combinedarea was affected.

8 Informes de Coopers y Lybrand sobre el Seguimento del Accidente de Aznalcóllar. 1 Informe de la descripcióndel Accidente de Aznacóllar y Opinión sobre las Actuaciones y Medidas adoptas para paliar sus efectos: 30 June1998.9 La Garcilla (1998) Sociedad Española de Ornitología (SEO) www.seo.org/es/capanias/doniana.html10 Coopers & Lybrand (1998)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The fauna of Rivers Agrio and Guadiamar was significantly affected by the spill, particularlythe toxic sludge, which killed all aquatic life it came directly into contact with, owing tomechanical rather than toxic causes, i.e. being smothered or crushed.

Until 27 May 1998, when intensive retrieval of carcasses was halted, some 37.4 tonnes ofdead fish were collected (75-80% carp Cyprinus carpio, 10-16% thin-lipped grey mullet Lizaramada, 6-8% Barbus sclateri, 4% European eel Anguilla anguilla and other 5%). Alsocollected were 96 terrestrial vertebrates; one white stork, 40 marsh frogs (Rana perezi), 11mallards, 8 coot and 8 rabbits, though it is not clear that all these deaths directly related to theincident. Also 890 birds’ eggs were collected, plus 14 chicks and 9 live birds which were allsent to the recuperation centre of El Acebuche.

Direct losses to local economy in region of 40,000 million pesetas (€0.24 million) wereestimated after one year. Exports of Doñana strawberries were banned; cotton, cereals andpeaches were prohibited from being collected in the area; 4,700ha of arable, rice and pastureswere affected there was a prohibition on harvesting of seven species of mollusc fromGuadalquivir estuary; and hunting was banned in the three provinces with territory in Doñana- Cádiz, Sevilla and Huelva 11.

7.5 CASE STUDY ASSESSMENT OF DAMAGE TO THE DOÑANA NATIONAL PARKAND CORRIDOR ECOLOGICA DE RIO GUADIAMAR NATURA 2000 SITES

7.5.1 Scope of Damage AssessmentAt the time of the spill it is understood that the only Natura 2000 site to have been identifiedwas the Cotto Doñana National Park. The impact of the mine spill on the National Park waslimited by the swift intervention of the State authorities. This largely prevented pollution fromentering the National Park and Natura 2000 site. Despite this, a significant area of wetlandknown as the Entremuros, which lies adjacent to the Doñana National Park, was affected bythe spill. The Entremuros, a 20 km long canalised section of the Rio Guadiamar, covers anarea of 2,656ha. At the time of the spill, this extensive wetland was not part of a Natura 2000site, but has subsequently been classified as a Special Protection Area (SPA) in accordancewith the EU Birds Directive (Corredor Ecologico de Rio Guadiamar, 13,470 ha, ES6180005).For the purposes of this case study the Corredor Ecologico de Rio Guadiamar is consideredas if it were included in the Natura 2000 site series at the time of the accident.

7.5.2 Establishing Conservation ObjectivesConservation objectives for the National Park and the Corredor Ecologico de Rio Guadiamarhave been developed as if they were a single Natura 2000 site. In terms of administration thisis not the case. However in ecological terms the two sites are so closely related that there isunlikely to be any distortion of the assessment by making this assumption.

A total of 25 habitats listed on Annex I of the EU Habitats Directive are recorded as occurringwithin the Doñana National Park Natura 2000 site. These are listed in Table 7-1. In addition, atotal of 44 bird species listed on Annex I of the EU Birds Directive also occur in significantnumbers (recorded as more than present (p) on the Standard Data Form). These are listed inTable 7-2. A further 11 species of mammals, reptile,s amphibians and invertebrates have alsobeen recorded from the Natura 2000 site12.

11 SEO website (Spanish Ornithological Society: www.seo.org/es/campanias/doniana.html)12 Garcia Novo, F. “The Ecosystems of Doñana National Park”. Department of Ecology, University of Sevilla.http://www.enveng.ufl.edu/wetlands/Doñana.html

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 7-1: Annex I Habitats Present within the Doñana National Park pSCIHabitats(Annex I of Habitats Directive)

Wooded dunes with Pinus pinea and/or Pinus pinaster*

Dune scleorophyllous scrubs (Cisto-Lavenduletalia)

Dune Juniper thickets (Juniperous spp.)*

Eu-Atlantic decalcified fixed dunes (Calluno-Ulecetea)*

Salicorinia and other annuals colonising mud and sand

Spartina swards

Mediterranean and thermo-Atlantic halophilous scrubs (Arthrocnemetalia fruticosae)

Mediterranean salt-meadows (Juncetalia maritimi)

Shifting dunes along the shoreline with Ammophila arenaria (white dunes)

Fixed dunes with herbaceous vegetation (grey dunes)*

Malcolmietalia dune grasslands

Oligotrophic waters containing very few minerals of West Mediterranean sandy plains with Isoetes

Salix alba and Populus alba galleries

Thermo-Mediterranean riparian galleries (Nerio-Tamariceteae) and south-west Iberian Peninsula ripariangalleries (Securinegion tinctoriae)

Quercus suber forests

Hard oligo-mesotrophic waters with benthic vegetation of chara formations

Natural eutrophic lakes with Manopotamion or Hydrocharition-type vegetation

Dystrophic lakes

Mediterranean temporary ponds*

Southern Atlantic wet heaths with Erica ciliaris and Erica tetralix*

Thermo-Mediterranean and Pre-Steppe brush

Sclerophilous grazed forests (Dehesas) with Quercus suber and/or Quercus ilex

Mediterranean tall herb and rush meadows

Calcareous fens with Cladium mariscus and Carex davalliana*

Lagoons** = Priority habitats

SEO/Birdlife (Birdlife Partner in Spain) “Doñana disaster. Doñana: Preliminary Environmental Assessment”(1998) http://www.mme.hu/madar/Doñana2.htm

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 7-2: Annex I Bird Species Present in Significant Numbers (>p) Within the DoñanaNational Park pSCISpecies(Annex I of Birds Directive)Bar-tailed godwit Limosa lapponica Spoonbill Platalea leucorodiaCollard pratincole Glariola pratincola Purple heron Ardea purpureaBlack-winged stilt Himantopus himantopus Little egret Egretta garzettaSlender-billed gull Larus genei Squacco heron Ardeolla ralloidesWood sandpiper Tringa glareola Night heron Nycticorax nycticoraxGull-billed tern Gelochelidon nilotica Little bittern Ixobrychus minutusGolden plover Pluvialis apricaria Bittern Botaurus stellarisStone curlew Burhinus oedicnemus Glossy ibis Plegadis falcinellusAvocet Recurvirostra avocetta Greater flamingo Phoenicopterus ruberCrested coot Fulica cristata Marbled duck Marmaronetta angustirostrisPurple galinule Porphyrio porphyrio White-headed duck Oxyura leucocephalaRuff Philomachus pugnax Black-shouldered kite Elanus caeruleusAudouin’s gull Larus audouinii Montagu’s harrier Circus pygargusCormorant Phalacrocorax carbo sinensis Peregrine falcon Falco peregrinusPin-tailed sand grouse Pterocles alchata Spanish imperial eagle Aquila heliaca adalbertiBlack tern Chlidonias niger Hen harrier Circus cyaneusLittle tern Sterna albifrons Marsh harrier Circus aeruginosusCommon tern Sterna hirundo Red kite Milvus milvusSandwich tern Sterna sandvicensis Black kite Milvus mirgansKingfisher Alcedo atthis Egyptian vulture Neophren peronopterusBlack stork Ciconia nigra Griffon vulture Gyps fulvusWhite stork Ciconia ciconia Black vulture Aegypius monachus

Many of the habitats and their associated fauna and flora are confined to the drier parts of theNational Park, including a number of sand dune habitats. A review of the reported damagefrom the spill suggests that only a small number of these habitats were actually affected.

In addition to the large number of Annex I bird species recorded from the Natura 2000 site,many migrant birds also use the area. These are not recorded on the Standard Data Form forthe Doñana National Park Natura 2000 site. For the purposes of this assessment it has beenassumed that regularly occurring populations of migrant birds form an important conservationobjective for both the National Park and the Corredor Ecologico de Rio Guadiamar.

The case study damage assessment of the Aznalcollar Mine Spill on Annex I Habitats in theDoñana National Park and Corredor Ecologico de Rio Guadiamar Natura 2000 Sites ispresented in Table 7-3. The case study damage assessment of the accident on Annex IISpecies is presented in Table 7-4.

The impact of the spill on bird health and populations is difficult to determine from theevidence, however, there was clearly a significant impact on the habitat of a number ofwetland birds, resulting in elevated levels of metal contaminants in many species. Although itis difficult to make firm predictions of the effect of the spill on individual species, there issome information on a few species, whilst for others a prediction can be made on the basis oftheir habitat preferences. The likely impact on species for which some ecological impactinformation exists, or can be reasonably predicted, is assessed in Table 7-5.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 7-3: Case Study Damage Assessment of Aznalcollar Mine Spill on Annex I Habitats in the Doñana National Park and Corredor Ecologicode Rio Guadiamar Natura 2000 SitesConservation Objective Its natural range and areas it

covers is stable or increasingIts species structure and

functions exist and are likely tocontinue to exist


Habitats(Annex I of HabitatsDirective)

Impact Nature Signific-ance



Mediterranean saltmeadows(Juncetalia maritimi)

Adverse St, R Major Adverse St, R1 Major Adverse St, R1 Major

Salix alba and Populus albagalleries

? ? ? Adverse St, R1 Major Adverse St, R1 Major

Thermo-Mediterranean ripariangalleries (Nerio-Tamariceteae) andsouth-west Iberian Peninsulariparian galleries (Securinegiontinctoriae)


Sclerophilous grazed forests(Dehesas) with Quercus suberand/or Quercus ilex


Mediterranean tall herb and rushmeadows

Adverse St, R Major Adverse St, R1 Major Adverse St, R1 Major


1 = Presumes clean up limits duration of impact to Short Term

? = Information on changes in the extent of woodland habitat due to impact not available. Extent of other habitats (marshes and saltmarshes) clearly adversely affected as much wasploughed and replanted following the removal of contaminated material.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 7-4: Case Study Damage Assessment of Aznalcollar Mine Spill on Annex II Species in the Doñana National Park and Corredor Ecologicode Rio Guadiamar Natura 2000 Sites

Conservation Objective Its natural range and areas itcovers is stable or increasing

Its species structure andfunctions exist and are likely to

continue to exist


Species(Annex II of HabitatsDirective)

Population is maintaining itselfon a long-term viable basis

Natural range of the species isneither being reduced nor is

likely to be reduced


term basis

Otter Lutra lutra Adverse St, R Minor Adverse St, R Minor Adverse St, R Minor

Mauremys leprosa ? ? ? ? ? ?

Emys orbicularis ? ? ? ? ? ?

Testudo graeca ? ? ? ? ? ?

Cobitis taenia Adverse St, R2 Major Adverse St, R2 Major

Barbus comiza Adverse St, R2 Major Adverse St, R2 Major

Chondrostoma polylepis Adverse St, R2 Major Adverse St, R2 Major

Acipenser sturio Adverse St, R2 Major Adverse St, R2 Major

Aphanius iberus Adverse St, R2 Major Adverse St, R2 Major

Coenagrion mercuriale

Populations of regularly occurringmigratory birds

Adverse St, R Major None - - Adverse St, R Major


2 = Assumed that all fish species were adversely affected due to catastrophic impact of the spill on fish (37.4 tonnes killed). Further information on the distribution of Annex I fish inthe Rio Guadiamar is required to confirm this assumption.

? = Information on effects on reptiles and amphibians not available.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 7-5: Case Study Damage Assessment of Aznalcollar Mine Spill on Annex I Birds in the Doñana National Park and Corredor Ecologico deRio Guadiamar Natura 2000 Sites


Its species structure andfunctions exist and are likely to

continue to exist


Species(Annex I of Birds Directive)

Population is maintaining itselfon a long-term viable basis

Natural range of the species isneither being reduced nor is

likely to be reduced


term basisBlack-winged stilt Himantopus himantopus ? ? ? ? ? ? Adverse St, R MinorAvocet Recurvirostra avocetta ? ? ? ? ? ? Adverse St, R MinorCrested coot Fulica cristata ? ? ? ? ? ? Adverse St, R MinorPurple galinule Porphyrio porphyrio ? ? ? ? ? ? Adverse St, R MinorRuff Philomachus pugnax ? ? ? ? ? ? Adverse St, R MinorKingfisher Alcedo atthis ? ? ? ? ? ? Adverse St, R MinorWhite stork Ciconia ciconia ? ? ? ? ? ? Adverse St, R MinorSpoonbill Platalea leucorodia ? ? ? ? ? ? Adverse St, R MinorPurple heron Ardea purpurea ? ? ? ? ? ? Benefit? St, R MinorLittle egret Egretta garzetta ? ? ? ? ? ? Benefit? St, R MinorSquacco heron Ardeolla ralloides ? ? ? ? ? ? Benefit? St, R MinorNight heron Nycticorax nycticorax ? ? ? ? ? ? Benefit? St, R MinorLittle bittern Ixobrychus minutus ? ? ? ? ? ? Adverse St, R MinorBittern Botaurus stellaris ? ? ? ? ? ? Adverse St, R MinorKey: St = Short term, Lt = Long term, R = Reversible, IR = Irreversible, * = Priority habitats

Definitions of significance used:• Minor significance: the impact would have a significant adverse effect on the ecology of the feature, but the level of the effect is such that the resource would be capable of

absorbing this impact• Major significance: the impact would have a significant adverse effect on the ecology of the feature. Such an impact would present a measurable long term and permanent threat to

the viability of the resource within the Natura 2000 site

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


7.5.3 Conclusion of Significance AssessmentA considerably greater level of research and more detailed assessment of the impact of thespill is required before any firm conclusion can be drawn. However for the purposes of thecase study the above assessment tables do clearly demonstrate the method by which such anassessment can be undertaken.

In this example available information suggests that:

• Immediately following the accident there was a significant impact upon the woodland,saltmarsh and grazing land within the Entremuros region of the Rio Gaudiamar. Thisincluded at least three habitats listed on Annex I of the EU Habitats Directive.

• In addition, water from the river was prevented from entering the National Park, causingdrying of this area and creating a movement of birds from the National Park to thecontaminated wetland created by the retained polluted water within the Entremuros. Thisaction therefore resulted in a temporary loss of bird habitat from within the National Parkand the contamination of birds drawn to feed within Entremuros.

• Many fish were killed by the spill, and this is likely to have included ecologicallyimportant species listed on Annex II of the Habitats Directive. For some species of bird,the abundance of dead fish provided a positive benefit, although clearly this was onlyshort term.

The result of this assessment is that there was short-term significant damage to the Natura2000 site (Doñana National Park and Corridor Ecologica de Rio Guadiamar).

Restoration of the site was undertaken and continues today. This restoration is likely to resultin no long-term impact on the conservation status of the Natura 2000 sites affected. While thetarget of achieving baseline conditions may therefore be met, interim losses of resources andservices nevertheless exist.

7.6 ACTUAL RESTORATION OF DOÑANA NATIONAL PARK ANDESTABLISHMENT OF THE GREEN CORRIDOR

The main feature of primary restoration actually undertaken was the initial damming ofwaterways into the Doñana National Park to prevent contamination of the park with pollutedwater and sludge. After the spill the following actions were taken.• The removal of polluted sludge (5 – 7 million m3);• Removal of upper soil horizons and all organic material down to gravel in places;• Neutralisation of contaminated sediments with addition of calcium carbonate and calcium

hydroxide;• Removal of reed (Typha domingensis and Scirpus maritimus) from contaminated area of

Entremuros that grew following flooding within acid waters;• Replanting of contaminated soils with 18 species of plant to stop soil erosion (selected to

be un-palatable to herbivores and humans). These to be replaced later with more naturalvegetation; and

• Construction of small sediment traps at 400m intervals along the R. Guadiamar to preventmovement of contaminated sediment downstream. These will also be removed in time torestore natural river profile.

This appears to have been a relatively successful primary restoration programme.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Compensatory restoration work has been undertaken for the whole green corridor to linkSierra de Aracena with Doñana. This was formerly much degraded with only fragments ofPopulus alba (white popular) gallery woodland remaining (this is an Annex I habitat type).Much of the area seems to have been planted with Eucalyptus (non-native and of littleecological value apart from providing nest sites for herons) or to have been converted foragricultural use. The entire green corridor has also been declared a Special Protection Area(SPA) in accordance with the EU Birds Directive.

The need for these compensatory restoration projects was based upon the following interimlosses:• Loss of wetland habitat within the Doñana National Park during the time that polluted

water was retained in the Entremuros;• Loss of habitat within the remainder of the green corridor whilst it was smothered in

contaminated sludge – although it appears that this habitat was already significantlydegraded;

• Loss of habitat within the green corridor during and immediately after the removal ofcontaminated sludge;

• Damage to trees and woodland from machinery used in the sludge removal;• Loss of habitat whilst planted unpalatable species were used to remove contaminants from

the sediments; and• Impacts on herbivorous birds whilst eating contaminated vegetation within the

Entremuros in the 4 months after the spill. The long-term effects of this seem difficult togauge but appear to have caused deformity in white stork chicks (a species listed onAnnex I of the EU Birds Directive) and probably much reduced breeding success. OtherAnnex I species reported to have been directly effected by the spill include black kite,black-winged stilt and purple galinule.

7.7 COMPARISON OF ACTUAL AND POSSIBLE RESTORATION ACTIVITIESDecisions actually made Decisions that could be made

Damage significanceWas damage significant? Damage assessment was carried out by the

Park authorities and scientific communityfor individual species and habitats but noformal overall assessment appears to havebeen made.

The result of the case study assessment is thatthere was short-term significant damage to theNatura 2000 site (Doñana National Park andCorridor Ecologica de Rio Guadiamar).

Primary RestorationIs Primary Restorationdeemed possible?

Yes Yes

What was its objective? It appears to have been to restore theinjured resource to its baseline levels,though it is not stated formally orelaborated upon.

This is a satisfactory objective. In the contextof this study the objective should be to restorethe habitats and species populations ofimportance (as defined by their Natura 2000status) to levels achieving favourableconservation status within their natural range.

Chosen optionWhy was it chosen?

Primary restoration, following the initialdamning of the waterways into theNational Park, included significant clean-up activities, including elements of limitedintervention and full-scale reconstructionin the severely affected canalised toxicwater retention zone known asEntremuros.

The primary restoration method chosenappears to be realistic in that it is possible torestore the environmental damage caused inthe Entremuros area using the methodologyemployed. Non-intervention or limited-levelintervention on its own would not be sufficientto restore the damage in a reasonable amountof time.

Good choice? Alternatives? No alternatives appear to have beensuggested

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Decisions actually made Decisions that could be madeHow was it implemented? Toxic sludge was removed along with upper soil horizons and all organic material down to

gravel in places. Neutralisation of remaining contaminated sediments was undertaken.Removal of reed that grew following flooding within acid waters. Replanting ofcontaminated soils with plant species to stop soil erosion to be replaced later with morenatural vegetation. Construction of small sediment traps along the R. Guadiamar which willbe removed in time to restore natural river profile.

Were costs of this optioncalculated?

No information is available; it is assumedthat the costs were not formally calculated.

An assessment of the discounted costs of theproposed primary restoration option should bemade, particularly where a mixture ofintervention options could be suitable; the costof each option should be estimated. The use ofplanting and importation of genetic material,soil etc. reduces the overall natureconservation value of the option. Costs ofmonitoring and surveillance should beincluded along with costs of: damageassessment, legal costs, cleaning, habitatrestoration, and loss of income to thosedependent on the damaged site (probably notapplicable in this case)

Were costs considered tobe ‘excessive’ initially?

Presumably a subjective assessment wasmade that they were not ‘excessive’.

An objective assessment of option costs shouldbe made using CEA given that the costs areprobably not deemed excessive.

Interim lossesWhere there any? Yes, there were interim losses whilst the primary restoration activity was being undertaken

and whilst the regeneration process was occurring.Further losses were incurred as a direct result of the immediate post-spill and primaryrestoration activities.These losses included:• Loss of wetland habitat within the Doñana National Park during the time that polluted

water was retained in the Entremuros;• Loss of habitat within the remainder of the green corridor whilst it was smothered in

contaminated sludge;• Loss of habitat within the green corridor during and immediately after the removal of

contaminated sludge;• Damage to trees and woodland from machinery used in the sludge removal;• Loss of habitat whilst planted unpalatable species were used to remove contaminants

from the sediments;• Impacts on herbivorous birds whilst eating contaminated vegetation within the

Entremuros, long-term effects possibly including deformity in white stork chicks andprobably much reduced breeding success.

Was their value calculated? A value does not appear to have been calculatedCompensatory Restoration OptionChosen option Compensatory restoration work has been

undertaken for the whole green corridor tolink Sierra de Aracena with Doñana. Thiswas formerly much degraded with onlyfragments of Populus alba (white popular)gallery woodland remaining.

Why was it chosen? Detailed information as to why this optionwas chosen is not available.

The chosen option could be defined as a ClassIII option, i.e. its outcome will have acomparable type and quality as the damagedresource. Primary restoration of canalised zonewill actually be exceeded as CompensatoryRestoration will improve the habitat of thedamaged area beyond its pre-spill conditionand also extend restoration of less affectedareas of the corridor beyond their pre-spillcondition as well.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Decisions actually made Decisions that could be madeGood choice? Alternatives? No alternatives appear to have been

suggestedIf the full objectives of the intendedCompensatory Restoration are met then thisoption appears satisfactory as it is likely togenerate greater long-term benefits than thecurrent interim losses. However this is anopinion, not based on an objective assessment.Information is not available to determine theprocess by which it was decided to initiate thisCompensatory Restoration project. There doesnot appear to have been any assessment ofalternative options and so it is presumed thatCEA was not undertaken.

Cost of option The total cost of establishing the CorredorVerde is estimated as 3,746 millionpesetas (approx. €22,500)13.No information is available as to how thisvalue was derived, and indeed reliabilityof the estimated figure appearsquestionable.

Were costs considered tobe ‘excessive’ initially?

Unknown, presumably not.

N/a – not applicable

7.8 OUTCOME OF SETTLEMENTFor years previous to the spill, NGOs have been filing charges against Boliden-Apirsa SL forcontinuing filtration of contaminants from the container dam, but these cases have beensuccessively dismissed14. On 25 March 1998 (a month before the accident) another complaintto the EU by the regional NGO CEPA was dismissed15.

It was reported16 that following the accident Boliden-Apirsa SL said it would pay clean-upcosts, make an advance payment of around €6.5 million to reimburse local farmers for theirlosses and buy their poisoned harvest. Another estimate of the economic impact onagricultural property values the damage at only €0.14 million17. However the companyappears to have since decided not to accept responsibility for the accident18.

Legal action for criminal negligence was taken against 25 people19 considered responsible forthe Aznalcóllar mine spill20 however the case was dismissed by the judge of Sanlúcar laMayor as it was ruled that no-one was to blame. The Junta and MMA therefore paid for thecosts associated with the spillage, although an appeal against the decision is to be made.

13 ‘Medio Ambiente 1 1998 Corredor Verde del Río Guadiamar.’ published by Consejería de Medio Ambiente ofthe Junta de Andalucía)14 Greenpeace website (www.greenpeace.es/toxicos/Doñana/sevilla0.htm)15 SEO website (Spanish Ornithological Society: www.seo.org/es/campanias/doniana.html)16 Environmental News Network ‘Spain says toxic cleanup to cost $105 million’ (www.enn.com/enn-subscriber-news-archive/1998/05/052698/costs_22039.asp) (26th May 1998)17 ASAJA –in Rodriguez, J. C., ‘Technological hazards: The case of the Aznalcóllar mining incident and itsimpact on coastal activities’. Proceedings of the first workshop of the INDICCO project, Seville, Spain, 17th-20th Nov. 1999. EU FAIR Concerted Action: The INDICCO project: A European Database of Indicator CoastalCommunities (FAIR CT98-4399). Eds Diana Tingley, MacAlister Elliott & Partners & Dr. Ian Goulding,Megapesca Lda.18 SEO website (Spanish Ornithological Society: www.seo.org/es/campanias/doniana.html)19 13 worked for Geocisa who constructed the dam and 7 for Boliden-Apirsa; also 2 civil servants in Junta deAndalucía, 1 from Instituto Geominero Español and 3 from Intecsa, the associate of Dragados who prepared theconstruction project in 1978.20 Natuweb: news about dismissal of legal action against Boliden-Apirsa SL (www.natuweb.com/) (28.12.00)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8 SEA EMPRESS OIL SPILL8.1 SITE DESCRIPTIONThe Sea Empress oil spill occurred at the mouth of the Milford Haven waterway, located insouth-west Wales, UK in the region known as Pembrokeshire. The Sea Empress oil tankerbecame grounded on rocks whilst attempting to enter the Milford Haven waterway; it wasabout to offload its cargo of 130,000 metric tonnes of crude oil to the Texaco refinery. Thewaterway is home to several oil refineries and tanker movements are common in the area.

South-west Wales is an area of great natural beauty and ecological interest. It is home to thePembrokeshire Coast National Park which is the only UK National Park primarily designatedfor its coastal and estuarine features21. The coast is highly indented and there are many smallislands. Marine plants and animals characteristic of both the relatively warm Atlantic andcolder Arctic waters are found in the area; important bird populations exist of Manxsheerwater (Puffinus puffinus), gannets (Morus bassanus) and common scoter (Melanittanigra). The Milford Haven is a drowned valley with approximately 110km of coastline andCarmarthen Bay has extensive sandflats, dunes and four important estuaries (see Figure 8-1).

The affected area includes two proposed22 SACs (‘Pembrokeshire Marine’ and ‘CarmarthenBay and Estuaries’), 35 Sites of Special Scientific Interest (SSSIs), two National NatureReserves, one of the UK’s three Marine Nature Reserves, and much of the coastline itself hasbeen designated as Heritage Coast in recognition of its historical interest and importance.

Figure 8-1: Protected areas of south-west Wales

Source: SEEEC (1998), pg. Inside front cover

Tourism plays a key role in the local economy, and is centred on the coastal environments andheritage. A small fishing industry operates in the coastal waters and provides an attraction forthe tourism sector. Other sectors of importance are agriculture and the oil industry. 21 Sea Empress Environmental Evaluation Committee. ‘The Environmental Impact of the Sea Empress Oil Spill;Final report of the Sea Empress Environmental Evaluation Committee.’ The Stationary Office (1998)22 At the time of the spill there were three proposed SAC’s however following the moderation process in 2000these sites were merged and there are now two proposed SAC’s.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8.2 SITE SERVICESThe following services are provided in the area affected by the Sea Empress oil spill:

ECOLOGICAL SERVICES:

Geo-hydrological:• pollution assimilation• sediment trapping and control• nutrient cycling• shoreline stabilisation.

Production/Habitat:• fish and shellfish habitats• habitat for furbearers, waterfowl and other wildlife• food production• organic material• maintenance of gene pools• maintenance of plant populations


‘HUMAN’ RELATED SERVICES:

Recreational:• beach use/swimming• fishing, boating• wildlife viewing• coastal path walking

Commercial / public or private:• fishing• waterway navigation• property protection

Cultural / historical:• historical


B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8.3 INCIDENT DESCRIPTIONThe “Sea Empress” oil spill occurred 15th February 1996, at St. Anns Head, at the mouth ofthe Milford Haven waterway. Over a period of seven days, whilst rescue efforts exacerbatedthe situation, 72,000 tonnes of crude oil and 480 tonnes of heavy fuel oil were released intothe sea. This oil seriously affected 100 km of coastline. A Marine Exclusion Zone was put inplace along the affected coastline, extending out to sea and up inland waterways.

The worst affected areas were West Angle Bay to Linney Head, western Carmarthen Bay andthe southern shore of the Milford Haven waterway. The major clean-up operation lasted forseveral months, though the majority of main tourist beaches were open by the Easter holidays(5-8th April, 1996). Many shores continued to be affected by residual oil throughout thesummer of 1996 and autumnal storms caused previously-sunk oil to resurface. By the springof 1997 there was little visible evidence of oiling23.

Commercial and recreational fishing was banned in the area and restrictions lifted forparticular species as their tissue became free of contamination. Restrictions on the majority ofcommercial stocks were lifted at the end of August 1996. The tourism industry, dependent onthe natural beauty (particularly beaches and coast-line), heritage and coastal-related activitiesin the area, was also affected.

8.4 SCALE OF DAMAGEThe main reported environmental impacts of the spill24 were that:• large numbers of marine organisms were killed;• populations of amphipods (small crustaceans) either disappeared or were severely

depleted;• several thousand sea birds were killed, particularly common scoter, diver species,

guillemots, razorbills and divers, with significant impact on breeding guillemots; and• significant decrease in the population of rare cushion starfish Asterina phylacticaThere was also temporary damage to some algae, lichens and saltmarsh vegetation.

It is further concluded25 that there were:• no impacts on marine mammals;• whilst oil concentrations were found in some fish species, there was little lasting damage;• several important sea bird populations were not affected and there was no effect on

breeding success; and• rare plants in the area were not significantly affected.

23 Sea Empress Oil Spill. Overview of the Environmental Effects. (www.swan.ac.uk/biosci/empress/overview.htm/)24 Sea Empress Environmental Evaluation Committee. ‘The Environmental Impact of the Sea Empress Oil Spill;Final report of the Sea Empress Environmental Evaluation Committee.’ The Stationary Office (1998)25 Sea Empress Environmental Evaluation Committee. ‘The Environmental Impact of the Sea Empress Oil Spill;Final report of the Sea Empress Environmental Evaluation Committee.’ The Stationary Office (1998)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8.5 IMPACT ASSESSMENTTable 8-1: Case Study Damage Assessment of Sea Empress Oil Spill on Annex I Habitats in the Pembrokeshire Marine and CarmarthenBay and Estuaries Natura 2000 Sites

ConservationObjective

Its natural range and areas it coversis stable or increasing


exist


Habitats(Annex I)

Impact Nature Significance Impact Nature Significance Impact Nature Significance

Sandbanks which areslightly covered by seawater all the time

- - - Adverse St, R Minor Adverse St, R Minor

Estuaries - - - Adverse Lt, R Minor Adverse St, R MinorMudflats and sandflats notcovered by sea water at lowtide


Large shallow inlets andbays


Submerged or partiallysubmerged sea caves

- - - Adverse St, R Minor - - -

Reefs - - - Adverse St, R Minor Adverse St, R MinorLagoons* - - - - - - - - -Salicornia and other annualscolonising mud and sand


Spartina swards - - - ? ? ? ? ? ?Atlantic saltmeadows - - - ? ? ? ? ? ?Dunes with Hippophaerhamnoides

- - - - - - - - -


B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Table 8-2: Case Study Damage Assessment of Sea Empress Oil Spill on Annex II Species, SPA and Bird Populations in the PembrokeshireMarine and Carmarthen Bay and Estuaries Natura 2000 Sites



exist


Species(Annex II and SPA)



to be reduced


term basisOtter Lutra lutra - - - - - - - - -Grey seal Halichoerus grypus - - - - - - - - -Sea lamprey Pteromyzon marinus ? ? ? - - - - - -River lamprey Lampertafluviatilis

- - - - - - - - -

Allis shad Alosa alosa ? ? ? - - - - - -Twaite shad Alosa fallax ? ? ? - - - - - -Species(Birds populations)



to be reduced


term basisCommon scoter Melanita nigra Adverse Lt. R Minor - - - Adverse St, R MinorPopulations of regularlyoccurring migratory birds

Adverse Lt, R Minor - - - Adverse St, R Minor


Definitions of significance used

• Minor significance: the impact would have a significant adverse effect on the ecology of the feature, but the level of the effect is such that the resource would be capable ofabsorbing this impact

• Major significance: the impact would have a significant adverse effect on the ecology of the feature. Such an impact would present a measurable long term and permanentthreat to the viability of the resource within the Natura 2000 site.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8.6 CONCLUSION OF DAMAGE ASSESSMENTIt is difficult to make anything but a tentative assessment of the significance of the damage tothe Pembrokeshire coast from the Sea Empress oil spill. It is clear that a number of Annex Ihabitats were affected by the spill and that there was temporary damage to speciespopulations associated with many such habitats. However, official reports suggest that marineinvertebrate populations rapidly recovered following the spill and that there were fewdetectable effects within a year of the spill. The estuary habitat with softer sediments andreduced wave energy is likely to have suffered longer than the more exposed shores. This hastherefore been graded as a long term impact although in fact this is only relative whencompared with the other habitat impacts.

Information is not available to assess impacts on other Annex II species, in particular effectson migratory fish species such as the Allis and Twait Shad and Sea lamprey.

The grey seal population appeared little affected. To date, there appears to have been noevidence of pollution-related seal mortality. It is, however, quite possible that seals may haveexperienced short-term pollution symptoms such as eye and respiratory tract irritation.Longer-term health problems may manifest themselves in due course. One possibility mayinvolve effects arising from the concentration of toxins through the food chain - grey sealsconsume a variety of prey including crabs and fish.

The most significant impact of the spill appears to have been on populations of sea ducks anddivers. Large numbers of these were killed by the spill and there is some evidence of asuppressed population of common scoter in the winter following the spill. Again, this impacthas been graded long-term, but this is only relative to the shorter-term impacts on otherhabitats.

By the spring of 1999, the common scoter population within Carmarthen Bay appeared tohave recovered well.

In a press release from the Countryside Council for Wales on 23rd February 1999, Dr BillSanderson, CCW marine biologist said: "Little was known about the common scoter’sbehaviour in Carmarthen Bay, particularly its preferred food. There was a large drop in thenumber of these ducks spending the winter in Carmarthen Bay after the oil spill; this wasprobably due to the effects of the oil on the birds but may also have been due to effects on thesea bed creatures which are the scoter’s food source. Now, this survey indicates that thehealth of the environment in Carmarthen Bay has greatly improved.”

Although the impacts of the spill appear to have been short to medium-term in duration, theyaffected a wide range of habitats and species populations of European importance. For thisreason the damage caused by the spill would have been assessed as being significant in termsof the proposed liability regime.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8.7 ACTUAL RESTORATION OF PEMBROKESHIRE MARINE ANDCARMARTHEN BAY AND ESTUARIES NATURA 200 SITES

Primary restoration was confined to clean up and reliance on natural processes. This wouldbe termed limited intervention (although in practice it is really a combination of non-intervention and limited intervention depending on the substrate type and location of thecontamination).

This strategy seems to have been relatively successful but, as the Swansea University website states, there was some damage to beaches caused by the clean up, and in the inner partsof the estuary oil is likely to be trapped for a considerable time. There is therefore a case forsome compensatory restoration although this does not seem to have occurred. In additionthere are the interim losses due to loss of habitat whilst natural processes restored them. Thisclearly had an impact on sea ducks most notably the common scoter.

Quite what compensatory restoration options might be appropriate or available is difficult topredict. Fishing activity is cited as a threat to the area in the Standard Data Form and sosome further control of fishing activity within the area to ensure favourable condition of themarine habitats might be an appropriate strategy – although clearly this would be unpopularwith the fishermen and economically costly.

In sum, the primary restoration strategy seems to have been soundly based using acombination of non-intervention and limited intervention strategies. Despite this interimlosses have been incurred and there may be a need for further compensatory restorationprojects. Quite what these should be is difficult to define; perhaps some future control onfishing activity in the area might be ecologically appropriate, though economically costly andunpopular.

8.8 COMPARISON OF ACTUAL AND POSSIBLE RESTORATION ACTIVITIESDecisions actually made Decisions that could be made

Damage significanceWas damage significant? Damage assessment was carried out

through the Sea Empress EnvironmentalEvaluation Committee and local scientificcommunity for individual species andhabitats but no formal overall assessmentappears to have been made.

Although the impacts of the spill appear tohave been short to medium-term in durationthey affected a wide range of habitats andspecies populations of European importance.For this reason the damage caused by the spillwould have been assessed as being significantin terms of the proposed liability regime.

Primary RestorationIs Primary Restorationdeemed possible?

Yes Yes

What was its objective? It appears to have been to restore theinjured resource to it baseline levelsthough it is not formally stated orelaborated upon.

This is a satisfactory objective. In the contextof this study the objective should be to restorethe habitats and species populations ofimportance (as defined by their Natura 2000status) to levels achieving favourableconservation status within their natural range.In the UK ‘favourable condition’ tables areusually generated to provide the PrimaryRestoration objective for each habitat andspecies population within a Natural 2000 site.

Chosen optionWhy was it chosen?

Primary restoration was confined to cleanup and reliance on natural processes. Thiswould be termed limited intervention.

Due to the inaccessible and delicate nature ofmarine and shoreline habitats affected,expected resilience of the larger speciesaffected and potential for natural regenerationof smaller organisms affected the reliance onnatural processes of regeneration through thenatural degradation of oil spilt seems areasonable option to take.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Decisions actually made Decisions that could be madeGood choice?Alternatives?

No alternatives appear to have beensuggested.

How was it implemented In practice it was a combination of non-intervention and limited intervention depending onthe substrate type and location of the contamination. The clean-up process improved thespeed with which natural regeneration (non-intervention) could occur.

Were costs of this optioncalculated?

Some information is available in apiecemeal fashion; it is assumed that thecosts were not specifically calculated forthis purpose.The budget of the Sea EmpressEnvironmental Evaluation Committee(SEEEC) is estimated as being between £2million and £4 million and mainly devotedto research costs (the SEEEC was set up tomonitor and evaluate the environmentalimpact of the spill.)Compensation payments were made tofishermen for lost income following a banon fishing; their livelihood is dependenton a fully functioning marine ecosystem.Figures are similarly available for lossesto the tourist industry who livelihood is inpart dependent on the same.

An assessment of costs of the proposedprimary restoration option should be made. Inthis case, non-intervention includes costs of:damage assessment studies, legal costs,implementing monitoring and surveillance,loss of income to those dependent on thedamaged site of ecosystem.

Were they found to be‘excessive’ initially?

Presumably a subjective assessment wasmade that they were not ‘excessive’.

An objective assessment of the options shouldbe made using CEA.

Interim lossesWhere there any? Yes, whilst natural regeneration process

was progressing.Was their valuecalculated?

Not directly

Value of conservation/non-use costs

A value for conservation/non-useeconomic costs was calculated providingan assessment of the cost of the overalleconomic loss suffered as a result of thedamaged resource/service (in other wordsthe benefit that would be gained byrestoring the site to its pre-spill status).

How was it calculated? The conservation/non-use costs wereestimated by applying a replacement costto numbers of observed strandings of eachmarine mammal species. Non-use valueswere calculated using a benefits transfermethodology whereby a range of WTPvalues per household were extracted fromthree ‘appropriate’ studies and applied toall households in the Welsh Water regionto give the range of estimated non-usevalues.

Estimates of use value should have been basedon the uses of the area as defined in Section8.2. Benefits transfer could then be used.Such estimates would cover support speciesimplicitly, but would not provide a value peramphipod, etc.

What was the value ofinterim losses?

Economic conservation/non-use costs forthe spill were estimated to be between£22.5 million and 35.4 million.

Compensatory Restoration OptionChosen option Compensatory restoration did not occur.

Expected lack of funding may have been alimiting factor in developing such options.

Compensatory restoration options could bedeveloped to compensate for the interim lossessuffered. Such options could be on-site or off-site depending on opportunities available andsite-specific considerations.Options should be judged based on the: type,quality and value of resources and services tobe provided.

N/a – not applicable

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


8.9 OUTCOME OF SETTLEMENTThe onus for providing compensation for clean-up activities and losses to the tourism andfishing sectors, and other local economy impacts, lay in the first instance with Skuld P&IClub, insurers of the “Sea Empress” and in the second instance with the International OilPollution Compensation (IOPC) Fund which becomes activated when an oil spill occurs in amember country’s waters. This onus to provide compensation in the first instance does notmean the organisations admit liability for the incident; rather they are acting as a result of aninternational protocol established to ensure that those impacted by a spill receivecompensation as soon as possible, before liability has even been established which can takeyears. Liability for the Sea Empress incident was found to rest with the Milford Haven PortAuthority which was legally prosecuted in 1999.

The IOPC Fund 1971 Convention compensates for clean-up operations, property damage,consequential loss and pure economic loss. Environmental damage can be compensated forbut only where the economic loss can be quantified in monetary terms and further to this,claims can be submitted for reasonable costs for measures taken to reinstate the marineenvironment after an oil spill. Funds can also be provided for post-spill environmental studiescarried out to determine the precise nature and extent of the pollution damage and/or need forreinstatement measures. In the case of the Sea Empress incident no claims were made oraccepted for environmental damage.

Financial costs relate to the change in revenues to an operator resulting from the incidentwhilst economic costs relate to the lost returns, or profits, from an activity and thus presentstheir value from a national, as opposed to local or regional, perspective. All costs are valuedat their opportunity costs to the nation and hence transfer payments, such as taxes andsubsidies are removed as they represent neither a gain nor a loss to the nation as a whole.Table 8-3 summarises the findings of the Environment Agency ‘Sea Empress Cost-BenefitProject’26 which estimated the financial and economic losses associated with the spill at theend 1997.

Table 8-3: Summary of total costs resulting from Sea Empress oil spill (£ million)Financial costs Economic costsCategory Lower bound Upper bound Lower bound Upper bound

Direct costs 49.1 58.1 49.1 58.1Tourism 4.0 46.0 0.0 2.9Recreation - - 1.0 2.8Commercial fisheries 6.8 10.0 0.8 1.2Recreational fisheries 0.1 0.1 0.8 2.7Local industry 0.0 0.0 0.0 0.0Conservation/non-use - - 22.5 35.4Human health - - 1.2 3.0Total 60.0 114.3 75.3 106.1

Source: reproduced from the Environment Agency ‘Sea Empress Cost-Benefit Project’ (pg. xix)

The conservation/non-use costs were estimated by applying a replacement cost to numbers ofobserved strandings of each marine mammal species. This method was not extended toamphipods losses. Non-use values were calculated using a benefits transfer methodologywhereby a range of WTP values per household were extracted from three ‘appropriate’studies and applied to all households in the Welsh Water region to give the range of estimatednon-use values. 26 Environment Agency. ‘Sea Empress Cost-Benefit Project. Final Report. Research and Development TechnicalReport. P119’ (1998)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The Environment Agency prosecuted the Milford Haven Port Authority which pleaded guiltyto “causing polluting matter to enter controlled waters contrary to section 85(1) of the UKWater Resources Act 1991”27. On the 15th Jan 1999 the Port Authority was fined a total of £4million plus costs – the largest fine handed out for a pollution case in the UK at that time27.However an appeal was launched by the Milford Haven Port Authority and the Court ofAppeals reduced the fine to $0.75 million on 17th Mar 200028. It is not known to whatpurpose this money will be put. The Skuld P&I Club and IOPC Fund may also prosecute theMilford Haven Port Authority to recover compensation monies paid out before liability forthe incident had been established.

27 Environment Agency ‘Record fine over Sea Empress pollution’ 15th Jan 99 (http://www.environment-agency.gov.uk//modules/MOD44.506.html)28 Environment Agency ‘Agency disappointed at Sea Empress fine reduction’ 20th Mar 00 (http://www.environment-agency.gov.uk//modules/MOD44.1867.html)

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


9 EXXON VALDEZ OIL SPILL

9.1 SITE DESCRIPTIONThe Prince William Sound and Gulf of Alaska are remote and spectacular areas of pristinemarine and coastal environment. There is an abundance of wildlife and encompassesthousands of miles of indented, rugged coastline and natural marine environment.

The northern Gulf of Alaska is known for its rich marine life: millions of seabirds; abundantmarine mammals, including sea otters Enhydra luttis, killer whales Orcinus orca, humpbackwhales Megaptera novaeangliae, northern sea lions Eumetopias jubatus, harbour seals Phocavitulina, and porpoises; five species of Pacific salmon Oncorhynchus spp., Pacific herringClupea pallasi, and groundfish that support multimillion-dollar fisheries; and the intertidaland subtidal communities along its coasts that provide subsistence for coastal villages ofAlaskan natives. The nearly pristine conditions and the abundant wildlife in the Gulf ofAlaska are a magnet for tourism. Tourism and fisheries are key components of the Alaskancash economy, especially in coastal areas.

9.2 SITE SERVICESThe following services are provided in the area affected by the Exxon Valdez oil spill:

ECOLOGICAL SERVICES:Geo-hydrological:• pollution assimilation• sediment trapping and control• nutrient cycling• shoreline stabilisation

Production/Habitat:• fish and shellfish habitats• habitat for fur-bearers, waterfowl and

other wildlife• food production• organic material• maintenance of gene pools• maintenance of plant populations

Ecosystem Integrity:• natural open space• climate regulation• biodiversity storehouse• resistance and resilience

‘HUMAN’ RELATED SERVICES: Recreational:• beach use/swimming• fishing, boating• wildlife viewing

Commercial / public or private:• fishing• waterway navigation

Cultural / historical:• historical• spiritual• subsistence


9.3 INCIDENT DESCRIPTIONThe Exxon Valdez spill occurred on 23rd March 1989. Around 39,000 metric tonnes of crudeoil was released into the Prince William Sound, before spreading to the Gulf of Alaska,following the grounding of the Exxon Valdez on Bligh Reef in Prince William Sound.Approximately 1,300 miles of coastline were oiled, with 200 miles being heavily ormoderately oiled and 1,100 miles being lightly oiled. There is more than 9,000 miles ofshoreline in total in the spill region.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


The spill occurred in early spring just before the annual spawning by Pacific herring in PrinceWilliam Sound. Millions of salmon fry were also about to emerge from their gravel spawningbeds to be washed into the nearshore waters and sustained by the spring plankton bloom.Young seals and sea otter pups were still experiencing an unusually cold winter; sea ice hadnot melted in the backs of many bays. Thousands of wintering sea ducks remained along thecrenulated coasts of Prince William Sound and the outer Kenai Peninsula. Other seabirdswere converging on their breeding colonies in the Gulf of Alaska.

9.4 SCALE OF DAMAGEThe environmental effects of this spill are well documented and report an acute impact onseabirds, bald eagles Haliaeetus leucocephalus, marine mammals and intertidal communitiesin the Prince William Sound and parts of the northern Gulf of Alaska. Longer-term impactswere suffered by Pacific herring Clupea pallasi and pink salmon Oncorhynchus gorbuschaand the inter-tidal and sub-tidal environments. It is estimated that 250,000 sea birds, 2,800sea otters, 300 harbour seals, 250 bald eagles, up to 22 killer whales and billions of salmonand herring eggs were killed.

9.5 CLEAN-UP ACTIVITIESThe clean-up took four summers, though not all beaches were clean even after that length oftime and activity. It is estimated that wave action from winter storms did more to clean up thebeaches than all the human effort employed (at its peak there were 10,000 workers, 1,000boats and around 100 aeroplanes involved).

9.6 IMPACT ASSESSMENT FOR NATURAL RESOURCESTen years after the Exxon Valdez oil spill, assessments of its impact still vary.Environmental scientists with much experience of assessing the impacts of oil spills and whohave followed the process of recovery in the Prince William Sound and other impacted areashave concluded initial expectations of long-term damage do not concur with the reality tenyears on29. This conclusion is in part based on the assumption that the damaged environmentwas naturally in a changing state at the time of the spill and has continued to evolve overtime, partly as a result of the spill but also due to natural environmental changes and shock,e.g. the El Niño effect. They go on to report that over the last 10 years the natural process ofwave action and weathering has recovered the originally affected shoreline. Initialexpectations about the impact of the spill on long-term damage to harbour seal populationsmay have been unfounded, there have been no long-term detrimental effects on pink salmonpopulation runs, population density or habitat occupancy of half of the 23 seabird speciesexamined. Sea otter populations and the other seabird species appear to have recovered.

However the 2000 Status Report of the Exxon Valdez Oil Spill Trustee Council concludesthat eight species have not recovered (common loon, cormorant, harbour seal, harlequin duck,killer whale, and pigeon guillemot). Recovered species are bald eagles and river otters.Recovering resources, which have not yet met specific recovery objectives, are reported asblack oystercatchers, common murres, marbled murrelets, mussels, Pacific herring, pinksalmon, sea otters, sockeye salmon, clams and inter-tidal and sub-tidal communities.Recovery is unknown for cutthroat trout, Dolly Varden, Kittlitz’s murrelet and rockfish.

29 Wiens J, Brannon J, Burns J, Day R, Garshelis D, Hoover-Miller A, Johnson C, Murphy S. “10 Years afterthe Valdez Oil Spill: Fish and Wildlife Recovery. Following the Exxon Valdez Oil Spill.” Discussion andConclusions. Papers presented at the International Oil Spill Conference (IOSC) in Seattle, March 8-11, 1999.http://www.valdezscience.com/wiens/conclusion.html”

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


9.7 RECOVERY OBJECTIVES FOR HUMAN SERVICESRecovery objectives set following the oil spill are stated as follows:• Passive Use: Passive uses will have recovered when people perceive that aesthetic and

intrinsic values associated with the spill area are no longer diminished by the oil spill.• Commercial Fishing: Commercial fishing will have recovered when the commercially

important fish species have recovered and opportunities to catch these species are not lostof reduced because of the effects of the oil spill.

• Recreation and Tourism: Recreation and tourism will have recovered, in large part, whenthe fish and wildlife resources on which they depend have recovered and recreation use ofoiled beaches is no longer impaired.

• Subsistence: Subsistence activity recovery objectives were also set relating to the servicesprovided to communities, predominantly of Alaskan Natives, such as their reliance onharvests of subsistence resources, e.g. fish, shellfish, seals, deer and waterfowl, as part oftheir traditional lifestyle

The 2000 Status Report of the Exxon Valdez Oil Spill Trustee Council concludes that none ofthese levels of recovery has been achieved, and hence are classified as still recovering.

9.8 OUTCOME OF SETTLEMENTThe settlement between the State of Alaska, US government and Exxon was reached in 1991.The US is not a signatory to the IOPC Fund and therefore this fund was not activatedfollowing the incident. A mixture of large range Primary Restoration and CompensatoryRestoration projects was initiated after the spill, greatly aided by the quantum of thesettlement reached with Exxon.

The settlement following the spill included a number of elements:

• Criminal Plea Agreement. Exxon was fined US$ 150 million for the environmental crime.The court forgave $125 million of this fine in recognition of Exxon’s co-operation incleaning up and paying some private claims.

• Of the remaining $25 million of the Criminal Plea Agreement, $12 million went to theNorth American Wetlands Conservation Fund and $13 million to the national Victims ofCrime Fund.

• Criminal Restitution: As a result of injuries caused to the fish, wildlife and lands of thespill region $100 million was divided evenly between the federal and state government.

• Civil Settlement: US$ 900 million to be paid annually over a ten-year period forrestoration of resources (with provision for allowing a further claim to be made for $100million in the future if necessary). Of this amount, $213 million went towardsreimbursement of the federal and state governments for damage assessment and spillresponse, whilst the remaining $687 million went to the Exxon Valdez Oil Spill TrustCouncil.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


A breakdown of the use of payments from the Settlements is shown in Table 9-1 below.

Table 9-1: Use of Payments made by Exxon as part of Civil and Criminal SettlementsCivil Settlement* US $ millionReimbursements for Damage Assessment and Response Damage assessment, litigation and clean-up

213

Exxon Valdez Oil Spill Trust CouncilResearch, Monitoring and General Restoration Surveys and monitoring, general restoration included funding projects to protect archaeological resources, enhancesalmon streams, reduce marine pollution and restore damaged habitats

180

Habitat Protection Large Parcel and Small Parcel habitat protection programs

395

Restoration Reserve Savings account established to support long-term restoration activities beyond last payment from Exxon inSeptember 2001.

108

Science Management, Public Information and AdministrationManagement of annual work plan and habitat programmes, scientific oversight of research, monitoring and restorationprojects, agency co-ordination and overall administrative costs, costs of public meetings, newsletters and other meansof disseminating information to the public.

31

Criminal SettlementFederal government used most of its portion to help the Trustee Council with:Habitat Protection, Shoreline Monitoring, Oil Spill Research and General Restoration

50

State of Alaska divided its money between:Capital improvements benefiting fisheries and research, habitat improvements, subsistence and new recreationalfacilities.

50

* Note: the Civil Settlement totalled $927 million including interest payments

9.9 RISK AND UNCERTAINTY – IMPLICATIONS FOR LIABILITYThe Civil Settlement included a provision for allowing a further claim to be made for $100million in the future if necessary. It is not clear whether this claim has been made, howeverthe original 1994 Restoration Plan created a reserve account to fund restoration into the futureby setting aside money annually to ensure creation of a $140 million reserve. An additional$30 million of unspent funds was added to this reserve. Two funds were set up to fund along-term habitat restoration program and a multi-decadal research and community-basedrestoration program and their continuation was approved in 1999.

9.10 COMPARISON OF EXXON VALDEZ AND SEA EMPRESS OIL SPILLSSoon after the Exxon Valdez oil spill the United States and State of Alaska commissionedscientific studies to determine the significance of damage to the marine environment, affectedcoastline and habitats and species they support. Economic valuation studies were alsocommissioned to quantify various impacts and losses including a contingent valuation studyto determine the loss of passive use values30.

These assessments helped secure the Criminal and Civil Settlements with Exxon which weremade in 1991 and it was with money from the Civil Settlement that the Exxon Valdez OilSpill Trust Council was set up. This Council has undertaken the vast majority of Primary andCompensatory Restoration activities designed to restore the damaged natural resource and therelated-services it provides.

30 Carson, R. et. Al. ‘A Contingent Valuation Study of Lost Passive Use Values Resulting from the ExxonValdez Oil Spill. A report to the Attorney General of the State of Alaska.’ 10th November 1992.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Information is not available as to how the decision making process actually functioned, i.e.how choices were made between different Restoration activities, beyond the initial clean-upphase, and how their costs and benefits were assessed. However it could be assumed thatsuch a process was defined, or became more defined, with the establishment of the ExxonValdez Oil Spill Trust Council, not before, and therefore with the introduction of funds withwhich to pay for activities once the best course of action was been determined.

In the case of the Sea Empress oil spill, the Environment Agency, Countryside Commissionfor Wales and other local NGOs, Agencies and local government bodies were involved in theinitial clean-up stages following the spill. The Sea Empress Environmental EvaluationCommittee (SEEEC) was established 27th March 1996 by the UK government and its Termsof Reference included:

• Co-ordinate monitoring work carried out by government departments and other publicbodies to assess the environmental impact of the spill and subsequent clean-up activities;

• Ensure a comprehensive set of monitoring data on environmental distributions andimpacts is obtained;

• Assess the overall impact of the incident on environment resources in the area affected(including fisheries, agriculture, amenity and wildlife conservation) and assesssubsequent recovery of these resources;

• Publish principle findings and conclusions of these studiesEconomic impact assessments were commissioned separately by the local authority,Environment Agency and other public bodies.Information is not available to describe the process by which Primary or CompensatoryRestoration options were defined and decided upon; decisions were probably taken internallyat local and national government and government body level. However no clearly definedprocedure currently exists in the UK to assess whether such actions may be required and it isarguable that without the existence of a defined liability system, such a system will not bedeveloped. Claims to the IOPC Fund for environmental damage can theoretically be made,but were not in the case of the Sea Empress. It is unclear as to what type of body, i.e.government or NGO would be accepted as a suitable claimant.

Litigation for the costs of damage to the environment or natural resources are not common inthe UK although the situation is slowly changing. The Environment Agency did successfullyprosecute the Milford Haven Port Authority for causing polluted matter from the SeaEmpress to enter controlled waters and the initial fine handed down by the Courts was thelargest recorded at its time for a pollution related case, though its quantum was significantlyreduced on appeal by the defendant. If successful pollution-related litigation was morecommon in the UK, as it is in the US, then perhaps the UK government would haveestablished a more transparent system of damage assessment and an appropriate restoration-decision making structure.

The impact of the Exxon Valdez oil spill was much greater in magnitude than the SeaEmpress spill however it seems that early assessments of each spill were probablyoverestimated, partly as a result of media attention and an emotional response to thetragedies. The main difference in post-spill activities restoration and compensatory activitiesappears to result from the fact that the US government secured a hefty financial settlementfrom Exxon with which it was able to plan restoration activities through setting up the ExxonValdez Oil Spill Trust Council. SEEEC on the other hand was limited to mainly monitoringresearch and determining the impact of the Sea Empress spill.

B4-3040/2000/265781/MAR/B3 1488-REG/R/03/B


Given its lower impact on the marine environment, the scale of restoration activitiesexperienced in the Alaska would not have been at all appropriate, but a level ofCompensatory Restoration to compensate for Interim Losses may have been appropriate.However, this was probably not explored in the detail set out by this study of a suitableLiability Regime arguably because funds were unlikely, or perceived to be unlikely, to beavailable.

EUROPEAN COMMISSION DIRECTORATE-GENERAL...

Documents

Transcript of EUROPEAN COMMISSION DIRECTORATE-GENERAL...