Development of a two stage liability allocation process in case of multiple party pollution

Abstract Liability allocation in pollution incidents involving multiple polluters is a challenging task even in cases where thorough forensic investigation has been carried out. A two-stage liability allocation process, i.e., technical liability allocation carried out by technical experts followed by legal liability allocation, is suggested for adoption in such cases. The liability allocation may be required for the impacts of pollution and/or for remediation of the polluted site. The first stage of liability allocation, i.e., technical liability allocation is discussed in the paper. The study identified and evaluated the most relevant technical factors that should be considered for responsibility allocation. These factors, quantified from the results of an environmental forensic investigation, can be aggregated as ‘Impact index’ and ’Remediation index’ for use in liability allocation. A method for the estimation of ’Impact index’ and ’Remediation index’ is suggested. The study argues that the overall liability allocation can be made more efficient and faster when the liability allocation is a two-stage process.


Introduction
Some of the fundamental purposes of the field of environmental forensics were understood to be 'identifying the source(s) of pollution and providing a technical framework for allocating the responsibility for cleaning up the pollution to the polluter(s) using scientific methods, in support of a legal regime assigning legal and financial responsibility for the cleanup of that pollution' (Varghese and Alappat 2012).However, the realization that pollution events form only a part of the wider spectrum of environmental issues necessitates a broader definition for environmental forensics.Consequently, Pivato, Gwinett and Varghese (2020a) defined environmental forensics as "a multidisciplinary science which aims to apply scientific methods and knowledge to the diverse range of environmental concerns in the context of regulatory and legal framework".In spite of the broader definition suggested, the investigation of pollution events remains the most frequent task of an environmental forensic expert (Pivato, Gwinett and Varghese 2020b).
Here, by pollution it is meant any uncontrolled release of chemicals into the environment which exceeds the criteria of risk defined by regulators or identified through technical analysis (such as risk analysis, Life Cycle Analysis, etc.).Polluting events are always accompanied by impact(s), acceptable or not, on the life and property of human beings and/or on his surrounding environment.The job of an environmental forensic expert is to evaluate the science and the relevant documentation surrounding the chemical releases, and derive a fair and equitable allocation based on the facts and supporting data.Some of the technical tools and methods available to an environmental forensic expert include chemical fingerprinting (Lin et al. 2015;Besser and Hamed 2019), mineralogical fingerprinting (Jeon et al. 2017;Kanbar et al. 2017), tree ring fingerprinting (Francov a et al. 2017;L opez-Vicente et al. 2018), remote sensing (Lu, Zhan and Hu 2016;Rahmati and Melesse 2016), use of biomarkers (Sol e et al. 2018), contaminant transport models (Varghese et al. 2015;Friberg et al. 2016;Samad et al. 2019), geospatial techniques (Hou 2017;Siddik 2018), age dating of chlorinated solvents (Morrison and Murphy 2015;Jakobsen et al. 2020), etc.The Pollution Crime Forensic Investigation Manual of INTERPOL (INTERPOL 2014) identifies different scenarios for pollution crimes and gives a step-by-step procedure for the environmental Crime Scene Investigation (CSI) that represents only the first step of a full environmental forensic investigation.Samad et al. (2019) have presented a general framework suitable for environmental forensic investigation and have applied it to a case of pollution from a cement manufacturing unit.Ashwini andVarghese (2019, 2021) have presented a framework for the investigation of sources and pathways of microplastics in the marine environment.Thus, there are many tools and methods available for fixing responsibility among the polluters in a pollution episode.It is the responsibility of the forensic expert to present the evidences collected using the tools and methods to various audiences (including, enforcement agencies, prosecutors, judges, and the court) to ensure the successful prosecution of an environmental case.
Pollution cases involving multiple polluters pose a severe challenge for forensic experts.Availability of guidelines on the procedure to be followed for investigation in such cases will not only make the job of forensic experts easier, but also help the courts in being more efficient with regards to liability decisions.Mudge (2008) and Samad, Varghese, and Alappat (2015), have identified the lack of proper guidelines for carrying out environmental forensic investigation as a reason for the courts not accepting the technical evidences presented before it by the experts.Priya, Varghese and Shah (2020), have undertaken a critical review of the factors the courts consider for responsibility allocation in pollution cases involving multiple polluters and have suggested segregating these factors into technical and non-technical.They have also suggested that liability allocation in cases involving multiple responsible parties may be a two-stage process: the first stage for technical liability allocation considering the technical factors and the second stage for final liability allocation (legal liability) where the already allocated fractional liability of a party gets modified based on nontechnical factors.The advantages of doing so as claimed by the authors are (i) it makes liability allocation easier for courts (ii) it may avoid costly litigation in certain situations, and (iii) the technical liability allocation can be delegated to environmental forensic experts who are better trained in technical matters.As the technical factors do not vary much from case to case, a general guideline may be possible for technical liability allocation based on the factors identified.Once the technical liability is fixed, the courts can perform the legal liability allocation considering additional non-technical factors, if any.Consequently, the overall allocation process becomes more transparent and fast, improving the efficiency of the environmental redress mechanism significantly.
In this study, we have identified the various technical factors that should be considered for responsibility allocation in a pollution event that involves multiple responsible parties and suggested ways to quantify these factors.We have also suggested a framework for using the identified factors for technical liability allocation.

Pollution liability
In its broad meaning, pollution liability includes liability for damage to natural resources in addition to the liability for damage to life and property (Abraham 1988).These two types of liabilities are treated differently by most legal systems.In this manuscript, we refer the 'liability for damage to life and property' as the 'impact liability' (as the liability arises from the impacts to life and property) and the 'liability for damage to natural resources' as the 'remediation liability' (as the liability arises from a requirement to clean up/remediate the damaged environment).The allocation of the liability of pollution, whether impact liability or remediation liability, is a complex and controversial process; especially, when there are multiple parties responsible for the pollution incident.Although Priya, Varghese and Shah (2020) had proposed liability allocation as a two-stage process of technical liability allocation followed by legal liability allocation, they have not clearly defined the terms 'technical liability' and 'legal liability'.For the purpose of the current study, we define technical liability as 'the liability that is proportional to the quantified share of each party in those attributes of pollution/ pollutant(s) which can be measured using scientific techniques'.Furthermore, we define legal liability as 'the liability of each party for pollution as decided by an authority authorized to carry out such responsibility allocation in the obeisance of any rules and regulations relevant to the case and applicable to the jurisdiction where the decision is made'.From these definitions, it follows that the legal liability of pollution may or may not depend on the technical liability of pollution.However, in many jurisdictions, technical liability is inherent in legal liability and the legal liability differs from technical liability to the extent that factors like agreement between parties and cooperation of parties with remediation agencies modify the liability allocation, as decided by the allocating authority.The factors influencing technical liability are measurable and hence not subject to personal bias.

Technical liability of pollution
Countries like the USA and the UK follow certain liability allocation guidelines.Gore factors developed by Albert Gore (H.R 7020, 96 th Cong., 2d Sess.(1980)) are used extensively in the U.S for liability allocation (ASARCO LLC v. Atlantic Richfield Company, LLC, No. 18-35934 (9th Cir. 2020), TDY HOLDINGS LLC TDY LLC V. UNITED States (372 F. Supp.3d 1091 (S.D. Cal.2019))).The Gore factors include technical as well as nontechnical factors.The factors that could be considered technical are the ability to distinguish the party's contribution to the nature and extent of the problem; the degree of parties' involvement in the activities that caused the problem; the quantity of waste involved and the toxicity of waste.
The guidelines for 'Non-Binding Preliminary Allocations of responsibility (NBAR)' developed by the U.S Environmental Protection Agency (USEPA (United States Environmental Protection Agency) 1987) act as a tool for preliminary liability allocation.The technical factors included in NBAR guidelines are the volume of waste, toxicity of waste and mobility of waste.NBAR guidelines also include non-technical factors like strength of evidence, ability to pay, litigation risks, public interest considerations, precedential value, inequalities and presence of aggravating factors.
The U.K follows the supplemental guidelines to the Environmental Protection Act 1990 (European Parliament 1990) named the Defra circular 01/2006 (Department for Environment, Food and Rural Affairs, United Kingdom 2006) to apportion liability for pollution episodes involving multiple responsible parties.The technical factors listed in the guidelines are: area contaminated by each party, quantity of pollutants, period of occupancy by each party, and the scale of operation.

Determinants of technical liability
The preceding discussions identified many technical factors for liability allocation.These factors mainly depend on properties of the contaminant and properties of the contaminated environmental medium, which can be quantified using scientific methods.Some of these factors are relevant only for impact liability allocation, while a few others are relevant only for remediation liability allocation.However, a few factors are relevant for both.A brief discussion on these factors is presented below:

Spread of chemicals in the environmental medium
The chemicals get dispersed in the environmental medium after their release.The spread of contaminants in the medium affects the impact liability as well as remediation liability.The wider the spread of a given quantity of chemicals in the medium, the more difficult it is to remediate the medium.The spread of chemicals in an environmental medium depends on the mobility of the chemical within that medium.Mobility of chemicals has been used as a criterion for liability allocation of superfund sites by U.S courts (Marryott 2000).In cases when different polluters release different chemicals, their liability differs with the mobility of chemicals.Chemicals having higher mobility are capable of producing larger plumes (Murphy 2000), though factors such as natural attenuation and matrix diffusion can result in plumes of similar lengths for chemicals of varying mobility potential (Kamath et al. 2012;Adamson et al. 2014).The mobility potential and consequently the spread of chemicals can be quantified using coefficients such as the octanol-water partition coefficient (K ow ) (USEPA (United States Environmental Protection Agency) 1996), the soil absorption coefficient (K oc ) (USEPA (United States Environmental Protection Agency) 2013), the Henry's law constant (H) and the solubility (S).These values are available from the databases of USEPA (EPA 2021), European Chemicals Agency (ECHA (European Chemicals Agency) 2021) and other agencies.
Another factor deciding the spread is the time/duration of discharge.In some instances, one polluter may be discharging small quantity of waste for a longer duration and another polluter may be discharging large quantity of waste for a smaller duration, both effectively contributing the same quantity of chemicals.In the former case, the spread of pollutants in the environmental medium is expected to be more.In certain situations, when the same mass of the same chemicals were released by different polluters at different times, it was found that the plume attributed to the chemical released earlier were larger than the plume attributed to the polluter who released the chemical at a later time (Marryott 2000).Thus the time and duration of release are also factors deciding the spread of contaminant.
For a given quantity of chemical contaminant, the easiest and direct way to quantify the spread is calculating the volume of the medium that is contaminated with the chemical; the volume of contaminated soil in case of soil pollution and the volume of aquifer in case of groundwater pollution.The USEPA which had earlier used the mass of the contaminant as the basis of the measure of its cleanup activities under Superfund and Resource Conservation and Recovery Act (RCRA) Corrective Action enforcement, later shifted to the volume of the contaminated medium as a measure to evaluate the effectiveness of cleanup activities (USEPA (United States Environmental Protection Agency) 2003).Although the volume of contaminated medium is a direct measure of the spread of the pollutant, in some cases the spread may have to be calculated using the coefficients described above, or the time/duration of release of the chemical pollutant.

Quantity of chemicals
The quantity of the pollutant emitted is the basis for regulatory enforcement by most environmental statutes namely, the Air (  (Marryott et al. 2000;Murphy 2000).Remediation cost is allocated among polluters primarily based on the quantity of chemicals in cases of commingled contamination caused by different polluters (Owete, 2007).For calculating the environmental risks from the discharge of chemicals into the environmental medium, quantity based calculations form the basis (Kaufman et al., 2005).Thus, the quantity of pollutants, often expressed as mass, is an important determinant of technical liability.
In deciding remediation liability, although the spread of contaminants, measured as the volume of the contaminated medium is the criteria recommended by the USEPA (USEPA (United States Environmental Protection Agency) 2003), it can be easily inferred from certain situations that the quantity of pollutants is also an important factor.In situations where the volume of the medium contaminated by two parties is more or less the same, the cleanup cost will differ based on the amount of pollutants in the medium.The environmental forensic techniques like transport modelling, chemical finger printing, oil finger printing, etc. can be used for estimating the volume/quantity of chemicals contributed by different polluters (Petrisor, 2014).

Toxicity of chemicals
In cases where chemicals released by different parties vary greatly in toxicity, the allocation based on the quantity of chemicals alone will not be reasonable (Reddy 2010).In such instances, the toxicity of chemicals released at a site also becomes important in apportioning impact liability among multiple polluters (Marryott 2000).Toxicity is an indication of the risk possessed by the release of chemicals into the environment (Beyer 2015).Although McGuire et al. (2020) relate risk magnitude, and thus toxicity of chemicals, directly with remediation costs, in the current procedure toxicity is considered as a determinant only of impact liability and not of remediation liability.Just as USEPA excluded toxicity from their remediation liability calculations (USEPA (United States Environmental Protection Agency) 1987), we also exclude toxicity as a determinant of remediation liability because toxicity of chemicals may not influence the clean-up efforts even when it influences the urgency for clean-up.
Chemicals can be toxic to the human being and/or to the ecosystems.In the first case, the term "toxicity" is generally accepted and the parameters that can be used to quantify the toxicity of chemicals to humans when exposure happens through various routes (inhalation, dermal contact, etc.) are the Slope Factor (SF) for carcinogenic compounds or the Reference Dose (RfD) for non-carcinogenic compounds (IRIS (Integrated Risk Information System) 2021).In the second case, the term "ecotoxicity" is preferred and the parameters used for quantification can be lethal concentrations (EC50, LC50) of the chemical to specific species under controlled conditions described in specific standardized protocols.The impact of toxicity on humans and ecosystems are both important for the allocation of impact liability.
The USEPA Integrated Risk Information Systems (IRIS) is a well-known reference that provides a methodology for estimating the toxicity of chemicals (IRIS (Integrated Risk Information System) 2021).The Agency for Toxic Substances and Disease Registry (ATSDR) provides the toxicity values of most of the chemicals (ATSDR (Agency for Toxic Substances and Disease Registry) 2021).Other databases that provide data on toxicity of chemicals include the European Chemicals Agency (ECHA) (2021), ECOTOXicology knowledgebase (ECOTOX (ECOTOXicology knowledgebase) 2021), The National Institute for Occupational Safety and Health (NIOSH (The National Institute for Occupational Safety and Health) 2021), etc.In the case where the polluters contribute the same chemical, toxicity need not be a factor, as the toxicity value will be the same for all polluters.

Remediability of the chemical
The term 'remediability' is used here as a measure of the ease with which a chemical contaminant can be separated from or rendered harmless in an environmental medium as part of the remediation efforts.In this sense, the remediability of the chemicals involved is one of the critical factors deciding the remediation expenses of a contaminated site.Remediability depends both on the properties of the contaminated medium and the chemical, however, for a given site it is mainly dependent on the properties of the chemical.
The chemical released combines with the environmental medium and forms a matrix.In some cases the chemical binds strongly with the environmental medium, making remediation difficult (Liu 2018;Li et al. 2019), where as in other cases, simple methods of remediation will decontaminate the environmental medium (Ossai et al. 2019;Teng et al. 2020).The remediability of chemicals is critical in selecting the remediation technologies to be adopted (ECETOC (European Centre for Ecotoxicology and Toxicology for chemicals) 2003; USEPA (United States Environmental Protection Agency) 2012) and consequently in deciding the cost of remediation (Li et al. 2019).
Another study that is seen to address the concept of remediability is by McGuire et al. (2020).They quantified remediability using a concept named 'remediation difficulty' which is the difference in risk magnitude before and after remediation.Although such a calculation can be very useful for an already performed remediation exercise, it fails to perform a predictive role.We propose determination of a 'remediability score' on a scale of 0-100 for each chemical -environment matrix pair, which will give a measure of the ease with which the site can be remediated.The lower the score, the more is the remediability of the chemical in the medium contaminated by it.The score is proposed to be estimated based on the factors which the USEPA has identified as important in screening soil and groundwater remediation technologies (FRTR (Federal Remediation Technologies Roundtable) 2021).These factors include the different attributes of the technology available for remediation for the case, namely, complexity and resource requirement of the technology, stage of development and confidence in the technology, efficiency of the technology in terms of contaminant reduction in the environmental matrix, efficiency of the technology in terms of long-term effectiveness, disturbance caused to the soil and the time requirement.

Persistence of chemicals
The persistence of chemicals gives the duration for which a chemical remains in the environment before degradation, and can be a factor in deciding both the impact liability and remediation liability.

Suggestions on quantifying technical liability
The factors identified above can be used to allocate liability among the polluters.We suggest that these factors can be aggregated apportioning weights to each factor to form 'Impact Index' (II) and 'Remediation Index' (RI) as illustrated in Figure 1.Each polluted site/pollution incident shall have an 'II' indicating the potential of the site/incident to harm life and property and 'RI' indicating the effort required to remediate the affected site.With the contribution of each party to the technical factors listed in the previous section obtained through an environmental forensic investigation, the party's 'II' and 'RI' can be calculated.The 'II' and 'RI' of the site are, respectively, the sum of the impact indices and remediation indices calculated for individual polluters.Finally, the technical liability of the party shall be allotted proportional to the party's share in the site's indices.The indices calculated for the site will help the authorities in making a comparison between different polluted sites with regard to the severity of pollution.

Illustration of the calculation of RI and II
The procedure for the calculation of 'II' and 'RI' for a typical scenario involving two polluters (Figure 2) is illustrated here as an example.It is worth to mention that the calculation can be applied similarly to more than two polluters (see supplementary materials).
Two polluters 'A' and 'B' were discharging chemicals 'P' and 'Q' respectively to the site 'X '.From forensic investigations, it was found that the quantity of chemicals discharged by polluter A and polluter B were W1 and W2, respectively.The toxicity of chemicals released by A and B were T1 and T2 respectively.The persistence of chemicals released by polluters A and B were found to be P1 and P2 respectively.Investigations show that Polluter A started discharging the chemicals about three years before polluter B. The spread of chemicals in the environmental medium depends on the quantity and duration of release in addition to the properties of the chemical and the environmental matrix.Since the quantity and duration are more for polluter A, the volume of medium contaminated by polluter A (V1) is larger than that by polluter B (V2).Remediability of chemicals will be R1 and R2 for the chemicals released by polluters A and B.
II A and II B , the impact indices of polluters A and B, respectively, are calculated as Where w iv , w iw , w it and w ip are weights attached to volume of contaminated medium, quantity of chemicals, toxicity of chemicals and persistence of chemicals, respectively, relevant for II.
Similarly, the remediation indices RI A and RI B of polluters A and B, respectively, are Where w rv , w rw, w rr and w rp are the weights corresponding to volume of contaminated medium, quantity of chemicals, remediability of chemicals and persistence of chemicals, respectively, relevant for the RI.The total II and total RI of the polluted site is given by (5) Technical liability of individual polluters A and B for impacts is given by, Technical liability of individual polluters A and B for remediation is given by, In the above illustration, instead of using the actual values of V, W, T, P and R, their normalized values may be used to get the values of the indices in a particular range, for example 1-1000.The weights attached to each factor shall be calculated taking expert opinion using any of the accepted structured procedures like analytical hierarchy process (AHP) (Souissi et al., 2019), Fuzzy AHP (Khashei-Siuki,  (Seifi, Dehghani, and Singh 2020), social vulnerability index (Tate 2012), etc. Probability distributions (PDs) of parameters are usually found from literature studies or measurements.Where no information is available, PD may be decided based on expert opinion (Verbeeck et.al 2006).The PDs of qualitative data are normally modelled as triangular or uniform distributions since the degree of uncertainty is high in such cases (USEPA (United States Environmental Protection Agency) 1996).

Final liability allocation
After the technical liability allocation is carried out by a forensic expert adhering to the guidelines, the results are submitted to the court for final allocation (Figure 1).Though the technical liability allocation by the forensic expert is subject to the scrutiny of the courts if there are complaints by parties, the scrutiny may be limited to the verification of adherence to the guidelines for allocation.
Although the CERCLA allocation may also be considered as a two stage process, where US EPA (a regulatory body of technical experts) optionally carry out a preliminary allocation following NBAR guidelines to avoid costly litigation, the procedure we suggest is different in many ways.NBAR guidelines are not objective in nature providing scope for discretionary allocation by EPA.NBAR guidelines are optional and is no longer common for liability allocation of superfund sites (Graves et al 2000).However, the technical liability allocation stage proposed by us is not optional.It is a mandatory step where fractional liability of PRPs are estimated based on factors that can be quantified using scientific methods.The second stage in our procedure will be required if any party can show the existence of non-technical factors (for example, agreement between parties, cooperation with regulatory agencies, parties becoming insolvent, etc.) that may distort the allocation arrived at using technical factors.Even in the absence of non-technical factors, the PRPs may opt for second stage if they are not convinced with the technical liability allocation.In the latter case, the scrutiny of the courts can be limited to the verification of adherence to the procedure of liability allocation.
The two stage liability allocation method proposed by us also intends to make use of the latest advancements in the field of environmental forensics.The modern tools of environmental forensics that are widely recogonized by scientific community allows the characterization of the sources of pollution and consequently the polluters, with better accuracy.Thus, a higher accuracy is expected in technical liability allocation step.The final allocation can be made considering only those non-technical factors that are relevant for the case making allocation more accurate and efficient.Thus, the liability allocation process with clearly defined guidelines makes the method meet the legal standard of "fair and equitable" allocation.
When it comes to allocation of costs, it can be done proportional to the fractional liability of each PRP.The costs incurred for remediation or that awarded as damages will vary significantly from country to country depending on the economic status and other factors.However, with a robust procedure in place to determine the proportion in which the costs are to be allocated, all countries can adopt it irrespective of their economic status.

Conclusion
The two-stage liability allocation strategy of technical liability followed by legal liability that can be adopted in pollution episodes involving multiple polluters is discussed in the paper.The definitions for technical and legal liability are suggested from the perspective of a forensic expert.The technical factors identified to affect technical liability are the spread of chemicals in the environmental medium, the quantity of chemicals, the toxicity of chemicals, the remediability of chemicals and the persistence of chemicals.The concept of remediability score is also discussed, which gives the ease with which the site can be remediated.The mathematical calculation of the technical liability to be borne by each polluter in a multiple party pollution episode is explained in the paper.The quantification of technical liability with well-defined guidelines makes the overall allocation process more efficient and faster.The proposed procedure is designed to suit all legal jurisdictions.

Table 1 .
Factors affecting the technical liability of pollutionnear here.