RISK ASSESSMENT AND ABANDONMENT STRATEGYPETROLEUM PRODUCTION FIELD

J.T. Dance and R.D. Huddleston, EBA Engineering Consultants Ltd., Calgary, Alberta, L. Garrettand T. Robbie, Devon Canada Corporation, Calgary, Alberta

A multi-year strategy has been developed for reclamation and abandonment of a production field innortheastern Alberta. Production facilities within the field consist primarily of abandoned flare pitsand waste oil sumps, as well as historical crude oil and salt water releases. The strategy was basedupon qualitative and quantitative risk assessments to identify priority sites for remedial action andto decide appropriate clean up levels which would be protective of local surface and groundwaterresources.

The strategy focused on removal of as much original source material as practical and in situmanagement of residual hydrocarbons (primarily as a consequence of migration from the originalsource). Priorities were established based on the potential magnitude of contamination, an analysisof the potential environmental sensitivities in the field and appraisal of the valued ecosystemcomponents. Provision was made to update the priorities on a year to year basis as more waslearned about the site conditions in addition to regulatory changes and technical advances.

Quantitative risk assessments were used to assess the impact of residual contaminants on surfacewater and groundwater resources. Exposure pathway modelling was used to assess the impact ofnatural attenuation on reducing the concentrations before impacting aquatic resources. To simplifythe exposure pathway modelling compounds are characterised into four groups: C6 to C10, C11 toC16, C16 to C24 and the BTEX (benzene, toluene, ethylbenzene and xylene) compounds. Monitoringprograms were established to confirm the results of exposure pathways modelling.

Ultimately the cost of maintaining a monitoring program for the entire field will be minimisedusing predictive models to demonstrate natural attenuation effectiveness on reducing thehydrocarbon and metal concentrations in the groundwater.

Future risk management activities will focus on residual contaminant impacts to the environmentusing an evolving technique: Ecological Impact Monitoring (EIM). This technique focuses onsoils, aquatics and terrestrial indicator species.

INTRODUCTION

The Swan Hills Field is located about 200 km northwest of Edmonton Alberta. (Figure 1). Thefield has been in production since the mid 1950s. The field is operated by Devon CanadaCorporation (Devon) on behalf of a group of more than 20 production companies (the Partners).Devon is responsible for reclaiming and remediating the production leases.

Due to the environmental setting, the number, age and variety of production facilities throughoutthe field, it was recognised that effective management and site closure would be a multi-year effort(ten years or more). An overall plan to steer and direct the ongoing environmental managementand site closure planning was required to ensure consistency in the approach and communicationswith stakeholders throughout these ongoing efforts.

The overall plan uses a risk-based reclamation and abandonment strategy. That plan was prepared:

� to divide the active and inactive sites requiring environment management into manageableunits and decide the relative urgency for environmental management;

� to ensure all active and inactive sites are managed in a systematic manner (due diligence)and monies are spent on real environmental issues; and

� to reduce net treatment costs by identifying economies of scale throughout theenvironmental management process.

This paper describes the application of a qualitative risk analyses method to establish priorities forenvironmental management. Further, the overall approach to environmental management involvedremoval of as much of the source material (buried wastes and spill residues) as practical andmanagement of the residual contamination in place. �Source material removal� refers to theobviously contaminated soil or free product within the original pit or spill area but does not includeextensive lateral excavation of discoloured soils nor small lenses (30 cm or less thick) of visiblehydrocarbon contamination.

Source material removal leaves some hydrocarbon compounds [BTEX compounds and polycyclicaromatic hydrocarbons (PAHs)], trace and heavy metals and salinity parameters in the ground atconcentrations greater than the current regulatory criteria (1, 2, 3, 4). In place management of this�residual contamination� relies upon natural attenuation processes to reduce or eliminate anyharmful effects of off-site contaminant migration to sensitive environmental receptors. Ongoingenvironmental management of these residual contaminants involves characterisation of thepotential environmental risks, contingency planning and ongoing groundwater, soil and (possibly)biophysical monitoring.

THE SWAN HILLS FIELD

The Swan Hills field is located immediately to the northwest of the Town of Swan Hills. The fieldoccupies an area of 13 km by 17 km (421 km2) over Townships 66 to 68 and Ranges 9 to 11 W5M(Figure 2). The surrounding lands are primarily forested lands held in the public trust. There issignificant potential for sport fishing and hunting in the area. The field is located within the naturalzone and consequently the end land use is a natural area. However, as with many natural areas inAlberta, the potential use of the area as a recreational area has been considered in formulating thisstrategy.

Environmental Setting

The ground surface across the field is hilly and steep with local relief in the order of several tens ofmetres and grades of 5% to 10%. Several steeply incised valleys occur across the field.

There are two prominent surface water courses within the Swan Hills field, the Swan River, whichruns from west to east across the central portion and Edith Creek, which runs from north to south inthe eastern portion (Figure 2). Edith Creek joins with the Swan River in the northeast corner of thefield. Both occupy well-defined valleys. The valley of the Swan River is broader and flatter thanthe steeply incised valley for Edith Creek.

Only one named lake exists within the study area, Edith Lake. Several small ponds exist within thefield and are drained by small meandering creeks. Marshy areas and wetlands occur throughout thefield.

In the valleys of the major water courses, the surficial sediments are composed of alluvial sandsand gravels. In the upland areas, the surficial sediments consist of till occurring as a hummockymoraine. A thin (less than 1 m thick) ground moraine is the surficial sediment near the marshyareas. The ground moraine of silt, sand and cobbles also occurs across the south central portion ofthe field (Figure 3).

The bedrock underlying the valley of the Swan River and the northeast corner of the field consistsof the Cretaceous Wapiti Formation. This formation is composed of grey feldspathic, clayeysandstone, bentonitic mudstone and scattered coal beds.

The remainder of the field is underlain by the Tertiary Paskapoo Formation. This formationconsists of grey to greenish grey, thinly bedded chert and cherty sandstone, grey and green siltstoneand mudstone. It also contains thin limestone, coal and till beds. Both bedrock formations areweathered in the upper few metres and are highly fractured.

The field exists within an ecological region classified as the Upper Boreal Cordilleran Ecoregion.Lodgepole pine and black spruce dominate the upland areas while poplar forests and shrublanddominate the lower lying areas.

Overall, the wildlife diversity is low (4). The majority of birds are transitory species. Slopeswithin the field provide large grazing opportunities for deer, moose and black bear. No threatenedor endangered wildlife are known to occur in the area, although the great grey owl, lynx, moose,deer and grizzly bear are reported in the area (4). Rivers and streams in the area have somepotential for sport fishing. Aquatic habitats contain leopard frogs and turtles.

Type of Facilities and Production History

The field contains approximately 385 operational or suspended oil and gas sites. The originaldiscovery well in the field was drilled in 1956. Much of the initial development took place fromthe 1950s to early 1960s. Miscible injection began in the 1980s.

In 1999, the annual production for the field was about 1 million m3 oil, 403 million m3 gas and 14.8million m3 produced water. Total production to date amounts to 560 million m3 standard barrels.

The production facilities in the field consist of:

� individual well sites;� batteries;� satellites;� water flood plants; and� miscible flood plants.

Potential Environmental Liabilities

Potential environmental liabilities from historical operations in the Swan Hills field existthroughout the network of production facilities. The potential number of sites within the SwanHills Field was assessed by Devon in 1995. There were as many as 210 sites requiringenvironmental management.

Many of the satellites and batteries within the field contain multiple potential sources ofcontaminants. These sources include:

� only wastes and produced water placed in flare pits;� leakage from above ground and underground storage tanks at battery sites; and� pipeline breaks (oil and produced water or salt spills).

Subsurface contaminants associated with losses from these sources consist of:

� crude oil � soluble benzene, toluene, ethylbenzene and xylene (BTEX), C6 to C10

aliphatics, C10 to C16 aliphatics and aromatics and polycyclic aromatichydrocarbons (PAHs), C16 to C24 aliphatics and C24 to C60+ aliphatics;

� trace elements and heavy metals (principally aluminium, arsenic, barium, cadmium, copperand zinc); and

� produced water � salty groundwater (predominately chloride).

Leaching of the soluble elements to the groundwater and subsequent discharge of this groundwaterto local marshes, water bodies and water courses is a principal concern for environmentalmanagement over the longer term.

PRIORITISATION SCHEME

To establish the relative priorities for environmental management at individual facilities across thefield, a qualitative assessment of potential environmental risks was undertaken. Becauseprioritisation was considered an ongoing management tool, it involved an initial appraisal basedupon the environmental sensitivity and an annual reappraisal based upon the status of siteassessment, remedial action and monitoring programs.

The environmental sensitivities were mapped during a field program in the fall of 1993. The fieldvisit was performed by a geologist, vegetation ecologist and wildlife biologist and involvedmapping of terrain characteristics, vegetation types (particularly rare plants) and aquatic andterrestrial habitat within the field.

Terrain characteristics were used to assess the geologic and hydrogeologic conditions and consistedof the composition of the surficial sediments and bedrock, evidence of geologic instability, thepresence in a local recharge or discharge zone and the proximity to surface water. Nine terraintypes were identified:

� thinly mantled sandstone bedrock;� thinly mantled clayey sandstone bedrock;� thinly mantled bedrock outcropping within 25 m of a water course;� thin ground moraine;� hummocky moraine;� plateau uplands underlain by alluvia;� lowland alluvia; and� peat and bogs.

The ecological land classification segregated the area according to biodiversity, rare plant potential,sensitivity to disturbance and recreational potential. Four different ecological complexes wereidentified according to their physical settings as either:

� upland;� lowland;� floodplain; or� ravines.

Wildlife utilisation within the area was mapped according to:

� big game usage (bear, moose and deer);� species diversity; and� opportunities for exposure (grazing land or drinking water).

Mapping was facilitated by the extensive network of roads used to develop the field. A GeographicInformation System (GIS) was used to prepare the terrain characteristics and ecological landclassification into a series of overlays and develop the hazard potential map.

The results of the mapping identified the seven hazard units within the field are shown on Figure 4.The hazard units combine a unique combination of exposure pathways expressed by the terraincharacteristics and ecological receptors (Table 1).

The second ranking process involved a year to year evaluation of the individual site�s status andselection of the type of activities to be performed that year.

The activities in the remedial program at each individual site consist of either:

� site assessment;� remedial action; or� monitoring (to support the risk management plan).

A weighting scheme to rank each site�s environmental concerns and subsequently to assignpriorities for the next year�s work is shown on Table 2.

The factors used on Table 2 provide a qualitative re-appraisal of potential environmental risk byincorporation of site specific findings into the priority ranking scheme. The status of �SiteAssessment� activities was assigned the highest relative score because of its unknown nature. TheRemedial Action portion of the ranking scheme, however incorporated ranking factors includingthe nature and extent of contamination, contaminant velocity and distance to a local water body.The monitoring portion of the ranking scheme incorporated risk factors as statistical trends inchemical concentrations and evidence of natural attenuation.

QUANTITATIVE RISK ASSESSMENT PROCESS

Quantitative risk assessments are the primary tool used on a site by site basis to assess themanagement requirements of residual hydrocarbons, metals and salts (residuals) remaining afterremoval of the more heavily contaminated source materials. Ongoing management of the residualsincludes monitoring to confirm that the risk has been reduced or eliminated by removal of thesource materials and active remedial action if monitoring identifies an unacceptable risk.

At the outset of the reclamation and abandonment strategy in 1996, it was recognised that extensiverisk assessment for the number of individual sites in the Swan Hills Field would be costly and notsufficiently useful to ensure that Devon�s long term environmental management objectives wouldbe satisfied. As a consequence, a process for risk assessment involving a screening level problemformulation, characterising risk based upon the selection of most significant exposure pathwaysand keying on key environmental receptors was developed.

Regulatory Framework

The Province of Alberta has established a draft policy (3) to allow oil and gas operators to managerisk associated with the ongoing operation or abandonment of their facilities. The policy is basedupon a three tiered approach to establish site specific clean-up criteria. Tier 1 consists of genericcriteria representing concentrations of residuals in the soil selected to be safe for a variety of endland uses. Tier 2 involves a modification of Tier 1 Criteria for site specific circumstances and Tier3 involves a risk assessment process.

Generic criteria for Tier 1 application has been provided by the Province of Alberta (5) forhydrocarbons. These generic values closely follow those developed by TPHCWG, (6). CCME (1)provides criteria for a variety of other elements (including metals) which are used to complementthe AENV Tier 1 Criteria. Both sets of criteria are risk based, but do not consider unique sitefactors creating an opportunity to consider natural attenuation processes. To establish the need forfurther intervention to manage residuals, a three staged process was developed involving:

� Problem Formulation;� Risk Characterisation; and� Monitoring for Natural Attenuation.

This process is compatible with the regulatory risk management framework (1). Within thisregulatory framework, the management of residuals involves a screening process during theProblem Formulation to identify potential contaminants of concern. This screening processsubstantially reduces the analytical requirements of the risk assessment.

Problem Formulation

As applied to the individual sites in the Swan Hills Field, a Problem Formulation is used todocument decisions made concerning the contaminants of concern, the pathways linking thesources areas to the potential receptors and the environmental receptors. To providedocumentation, a screening level model developed for landfills (7) was adopted for application ofthe management of residual risks in the Swan Hills Field. This screening level model is more fullyexplained in Dance (8).

Contaminants of concern were identified from confirmatory testing on the floors and walls of theexcavation used to remove the source materials. A subsurface investigation using testpits,boreholes and monitoring wells is used to establish the nature and extent of contamination and toevaluate the potential for subsurface migration off the site. These results are also used to establishthe geometry of the source for selection of an applicable analytical model. Hydrocarboncompounds were separated into four groups: the C6 to C10 aliphatics, C11 to C16, C16to C24 and theBTEX compounds for exposure pathways modelling.

Pathways linking the source area to the receptors primarily are geological pathways characterisedfor their transport properties during a subsurface investigation. To date, only a groundwaterdischarge pathway and an inhalation pathway have been identified at the sites within the fieldassessed to date. End points consist of drinking water quality and fresh water aquatic lifeprotection criteria for the groundwater pathway and ambient air concentrations and thresholdconcentrations for the carcinogenic and non-carcinogenic compounds in breathing air. Plans are inplace to better define tolerable limits for terrestrial and aquatic wildlife which will allow risk basedperformance criteria to be relaxed. However, for the present, the published criteria have beenconsidered sufficiently conservative to protect Devon�s concerns.

Receptors are based upon the understanding that once the field is abandoned, the principal land usewill be to return the area to a natural habitat. As a consequence, the focus has been on preservingaquatic habitat and drinking water for terrestrial wildlife. However, given the physical setting ofthe field, it is likely that some recreational opportunities exist in the future. Consequently, vapourrelease to casual visitors in outdoor and indoor spaces has also been considered.

Figure 6 is a schematic illustrating the risk scenarios developed from the Problem Formulation fora specific facility in the Swan Hills Field.

Risk Characterisation

An exposure assessment was undertaken to characterise the potential risk to water resources andambient air quality represent by the residual contaminants in the soil after the source was removed.The exposure assessment was used to predict the potential concentrations of the contaminantreaching the exposure point. These exposure point predictors are compared to ambient air levels(risk thresholds of 10-5 and a hazard quotient of 0.2) for an inhalation pathway and to criteria forthe protection of fresh water aquatic life for the aquatic habitat pathway.

For the vapour inhalation pathway, two potential receptors were considered: A casual visitor to thesite (exposure point 1.5 m above ground level) and a recreational resident (basement air). For thepathway to the aquatic habitat, the exposure point was assumed to be the closest streambed ormarshy area downgradient from the site. This assumption was conservative since it neglected anymixing efforts within the streambed or due to seasonal recharge events.

Each potential exposure pathway was separated into three types of transport processes:

� release of contaminants from the residual contaminated soil either to the pore spaces (porewater) or the soil air;

� loading of the contaminant from the pore space to the transport media; and

� pathway modelling to assess the impact of natural attenuation in reducing theconcentrations in the environmental media during transport.

The three processes were segregated because it was recognised that each represented differentenvironmental compartment where remedial action could be implemented if necessary. Byseparating into compartments, the application of different engineering measures (e.g., closurecapping, barrier walls, dual phase extraction method or groundwater pump and treat systems) couldbe evaluated and costed separately to support future management decisions.

For the vapour inhalation and recreational residential portions of the risk characterisation, theapproach proposed by ASTM (9) (exposure to outdoor air) and the Johnson and Ettinger model(10) (exposure to indoor air) were applied. For the aquatic habitat pathway, the ASTM (11)analytical expressions were also used for each of the three types of transport processes. Details ofthe parameters used and their derivation are contained in Dance (8). Hydrocarbon transportproperties were derived from TPHCWG (12) and the compounds segregated into four groups:

� the C6 to C10 aliphatics;� the C10 to C16 aliphatics;� the C16 to C24 aliphatics; and� the BTEX compounds and naphthalene.

To assess the effects of groundwater transport on the attenuation of contaminants a variety ofaquifer configurations were used. These configurations were used to consider the impact ofvarying ages and transient versus steady-state loading to the groundwater. The analyticalexpression used to predict the concentrations at the exposure point followed ASTM (11).

Monitoring of Natural Attenuation Processes

May of the production faculties within the Swan Hills Field are associated with visible plumes offerric oxyhydroxide precipitates in neighbouring surface water bodies. The presence of theseoxyhdroxides is also common in sections of the field well removed from petroleum productionsites. Further, base upon hydraulic conductivity values, hydrogeologic assessments conductedthroughout the field and the results of monitoring programs conducted at several sites, it wasevident that zone of dissolved contamination in the groundwater were considerable smaller than therange predicted by advective transport. These observations suggest that considerably naturalprocesses influenced the contaminant concentrations and stabilised zones of groundwatercontaminated produced by leachate from the original source. Removal of a large mass ofcontamination by excavation of the source should intuitively suggest that groundwater quality inthe vicinity of the individual sites should improve with time. Counter to that was the understandingthat the zones of the residual contamination were created by a more mobile liquid phase. In somecircumstances the migration from the source had never been destroyed by incineration of the flarepit fluids. This understanding suggests that the residual contaminants might differ in chemical

composition and leaching characterisation from that of the �source materials� and therefore may bedisproportionately responsible for zone of groundwater contamination.

As part of the multi year strategy it was therefore proposed that a semi-annual monitoring progress(spring and fall sampling) be undertaken specifically directed at documenting the relative inversesin primary indicators of natural attenuation. Those factors included shrinking of the contaminantplane and declining concentrations within the monitoring wells. These monitoring programs weresupplemented by chemical analysis of the secondary condition of attenuation process. Thoseindicators include the consumption of oxygen, nitrogen and sulphate and the increase inconcentrations of iron, manganese and sulphide in downgradient monitoring wells. Predictivemodels are applied at selected sites to assess the predictive capabilities of available models(Bioplan II and Bioscreen) as a third means of illustrating natural attenuation.

It is anticipated that in the spring of 2002, sufficient information will be available at the selectedsites to modify the monitoring requirements for future sites. Understanding of the naturalprocesses and application of predictive models to predict when significant changes taken place is adesirable criteria of this monitoring program. Ultimately, monitoring at selected sites will enablethe predictive efforts to be confirmed and reduce the monitoring requirements resulting in aconsiderable cost reduction.

Future Action

Future actions as part of the risk assessment and abandonment strategy in the Swan Hills Field willinclude:

1. Revised monitoring schedules to reduce the number and frequency of wells to reduce thenet remediation and reclamation costs.

2 . More comprehensive characterisation of aquatic and terrestrial habitats to ensure thecriteria used for characterising risk are sensitive to the local environment and ensure thatthe environmental conditions can be sustained beyond the life time of the field. Bymeasuring the total environmental effect on the surrounding habitat, it is envisioned thatthe field can be safely abandoned for other beneficial future land uses.

REFERENCES CITED

1. CCME �Canadian Environmental Quality Guidelines,� published by the Canadian Council ofMinistries of the Environment (1999).

2 . CCME (Canadian Council of Ministers of the Environment), �Canada-Wide Standards forPetroleum Hydrocarbons (PHC) in Soil � User Guidance,� (under review) (December 2000)and CCME (Canadian Council of Ministers of the Environment), �Canada-Wide Standards forPetroleum Hydrocarbons (PHC) in Soil: Scientific Rationale,� Supporting Technical Document(December 2000).

3. AENV (Alberta Environment), �Policy for Management of Risks at Contaminated Sites inAlberta (draft), Environmental Sciences Division, Environmental Service (May 2000).

4 . Golder, �Swan Hills Site Prioritisation and Remedial Action Plan,� unpublished report byGolder Associates to Home Oil Limited (nee Devon Canada Corporation), (January 1993).

5. AENV �Volume 3: User Guide � Alberta Soil and Water Quality Guidelines for Hydrocarbonsat Upstream Oil and Gas Facilities,� Alberta Environment, prepared by Kormex InternationalLtd. (2001).

6 . TPHCWG (Total Petroleum Hydrocarbon Criteria Working Group), �Total PetroleumHydrocarbon Criteria Working Group Series,� 4 Volumes (1997).

7. Zeiss, C., �Identification and Screening of Waste Facility Impacts for Environmental HealthRisk Assessments,� Research report prepared for Alberta Health, Environmental HealthDivision (1993).

8. Dance, J.T., �Application of Risk Based Corrective Action. Planning to a Former Waste OilProcessing Pond,� these proceedings IPEC, (2001).

9. ASTM, �Standard Guide for Risk Based Corrective Action,� American Society for Testing andMaterials Designation E-2081-00, 2000).

10. Johnson, P.C., and R.A. Ettinger, �Heuristic Model for Predicting the Intrusion Rate ofContaminant Vapours into Buildings,� Environment SciTech, Vol. 25, No. 8 1145-1452(1991).

11. ASTM �Standard Guide for Remediation of Ground Water by Natural Attenuation atPetroleum Release Sites,� American Society of Testing and Materials Designation E-1943-98(1998).

12. TPHCWG (Total Petroleum Hydrocarbon Criteria Working Group), four volumes (1997).

Table 1. Hazard Unit Definition

Unit Exposure Pathway Ecological Receptor

1 sandstone bedrock upland forests

2 shale bedrock upland forests

3alluvial sand and gravels

over bedrockupland ravine

4 thin ground moraine terrace slope

5 hummocky moraine terrace slope

6 ground moraine upland plateau

7 alluvium � plateau upland marches

8 alluvium sand and silt marsh and streams

9 peat ponds

Table 2. Priority Ranking Weighting Factors

Activity FactorTotal

WeightWeighting

Site Assessment Stage 1 - Site Recommission 10(100) Stage 2 - Subsurface Investigation 25

Stage 3 - Assess Potential for Off-Site Impact 40Stage 4 - Reporting 15

Remedial Action residual in soil > Tier I 0(50) free product 5

dissolved in groundwater 10Contaminant Velocity immobile 0

< 1 m/year 5> 1 m/year 10>100 0100 to 20 m 520 m to 0 10landfill disposal 0on-site treatment 5barrier or containment walls 10

Remediation Planning proven technology 0pilot scale testing needed 5field trial needed 10

Monitoring RMP Prepared prepared 0(40) in preparation 10

Monitoring Progress > 5 years 0< 5 years 5initial 10

Statistical Trends decreasing 0no tread 5increasing 10O2, NO3, NO2 consumption 0Fe and Mn concentration 5SO4 - S 10

10

10

10

10

10

10

100

10

10

10

Status of Remedial Action

Evidence of Natural Attenuation

Progress of Site Assessment to Date

Nature and Extent of Contamination

Distance to Sensitive Area/Water Course

offsite