department of natural resources and environmental control

Division of Air and Waste Management

Site Investigation and Restoration Branch

General Notice

Arsenic Risk Management Proposal Draft Background Document

June 22, 2005

Summary

This document reflects a proposal for public comment by Department of Natural Resources and Environmental Control (DNREC) staff, working with other state agency staff, to establish a cleanup goal for arsenic in soil at residential settings to background levels–using 11 parts per million (ppm) as a default background concentration, and to undertake a significant public participation process to solicit and use public input before finalizing this cleanup goal. It has been prepared in response to Governor Ruth Ann Minner’s June 9, 2005 directive to DNREC Secretary John A. Hughes to “review…and propose appropriate standards and policies” for arsenic in soil, focusing on ensuring the health and safety of Delawareans, and to solicit public input on the draft arsenic standard prior to adoption. The DNREC Division of Air and Waste Management staff has prepared this background document, in collaboration with the Division of Water Resources (DWR), the Division of Public Health (DPH), the Department of Agriculture (DDA) and the Department of Justice (DOJ) as part of DNREC’s response to Governor Minner’s directive.

This document is intended to support the public participation process by providing the supporting information and rationale for DNREC’s proposal. To help facilitate this public involvement process, it has been drafted for a general, not a specialized technical audience. As such it contains background information on scientific issues (e.g., chemistry, geology and toxicology of arsenic) and describes the basis for setting cleanup standards and goals. After this background information, the document considers several policy options for alternative standards, and summarizes the implications for each option. The document is intended to be used in combination with discussions between DNREC staff and interested public to allow for a constructive interaction, rather than presuming to be a stand-alone document. Section 5 provides more discussion of the intended public participation process.

The basis for DNREC’s proposal (See Option C, Section 3.3) to establish a cleanup goal of background concentration, with a default background of 11 ppm, is the need to reduce public exposure to Arsenic as low as possible. This proposal will be effective in draft immediately upon release and in final form upon adoption by the Secretary of DNREC. For carcinogens like Arsenic, Delaware law mandates that standards be set at levels associated with an incremental lifetime cancer risk increase of one in one hundred thousand (1/100,000 or 1 x 10-5) or at background. Using standard assumptions about exposure and dose, a background cleanup goal (11 ppm) could result in a slightly elevated lifetime cancer risk. DNREC staff and other participating state agencies do not believe this elevated risk is realistic because of the number of conservative assumptions on which this risk-to-dose relationship is based. These assumption include: all individuals eat the same amount of soil during for their entire life, that all of the Arsenic to which a person is exposed is in the most toxic form, and that all the Arsenic is bioavailable (See Section 1.4 and Section 1.5). Although these assumptions, adopted from U.S. EPA guidance, may appear extremely conservative, DNREC believes it is prudent to consider these assumptions to ensure protection of human health.

The background default concentration of 11 ppm should be possible to achieve during cleanups in most situations because it appears to reflect a midrange of the background concentrations of Arsenic found in Delaware soils. A cleanup goal lower than 11 ppm would not be technically feasible because the background concentrations in Delaware soils are higher in many situations. The cost implication of an 11 ppm default background concentration is not yet clear. It is possible it could result in fewer cleanups being completed because of the cost to complete each cleanup to this standard. DNREC staff will continue to collect information on this issue, but will be implementing cleanups at residential sites to the goals established, regardless of the cost.

This proposal also includes a schedule to review this standard annually to determine whether there is new information about the toxicity of Arsenic to ensure the standard remains adequately protective and compliant with the legal requirements in Delaware. This review will include public participation.

In recent months concerns have arisen from staff in the Department of Natural Resources and Environmental Control (DNREC), and the general public, about the recent Arsenic standard adopted in June 2004. As a result, Governor Minner on June 9, 2005, (See Attachment A) directed DNREC Secretary John Hughes as follows:

Immediately after the Governor’s tasking, DNREC Division of Air and Waste Management (DAWM) Director, James D. Werner convened a meeting of technical and management staff from DNREC DAWM and the Division of Water Resources (DWR), the Division of Public Health (DPH), the Delaware Department of Agriculture (DDA) and the Department of Justice (DOJ) (see Attachment B for list of attendees) to scope out the task and develop a cooperative approach to respond. The group agreed to work together to review the current standard and policies, per the Governor’s request. The staff work group also committed to obtaining and using the best technical and legal input in developing a proposal, and to subject that proposal to public comment.

In assessing the June 2004 Interim Standard, the Delaware state agency staff who met agreed that the DNREC was remiss in not including public involvement in the development and adoption of the current standard. From a technical perspective, the group agreed that, because the assumptions used by EPA and scientific bodies in the development of the risk assessment calculations are sufficiently conservative¹ that the interim standard did not pose an imminent risk or substantial lack of protectiveness. Nonetheless, the staff work group agreed that rigorous technical and legal analysis, with public input was appropriate to ensure developing a cleanup standard that is adequately protective of human health and has earned public confidence. Finally, the staff work group agreed that whatever standards and policies were developed, it should be reviewed on a regular basis and include public input. The June 2004 DNREC Air and Waste Management Division memorandum articulating an interim policy on Arsenic included among other things, a timetable for reviewing the interim policy every six months. Until now, DNREC had not conducted such a follow-up review.

In addition, the staff level working group quickly reached consensus on three fundamental principles regarding the establishment of a new cleanup goal for Arsenic in soil. The principles must be:

1. Based on adequate public comment and involvement because of the fundamental policy nature of the risk choices to be made and, accordingly, the need for public involvement in decisions that affect them;

2. Informed by the best available scientific information and the best available risk assessment advice from DPH and other scientific sources; and

3. In addition to the Governor’s directive to base a new standard on the evaluation of the best scientific information available, public involvement, and consideration of standards and policies used by EPA and other States, the agency staff agreed that a new standard must also be consistent with DNREC’s legal mandate for risk management.

The staff working group developed a draft outline of the topics to be covered in the proposed risk management revision and divided up task assignments. The group did not address in detail or resolve the question of what is the scope of the applicability of whatever standard is proposed and established (i.e., conversions of industrial sites to residential, agricultural site conversion, golf courses, etc.). The group agreed to seek advice from legal counsel and public input on the legal framework governing the question of scope and applicability of the standards and policies.

The general scope of this draft background document is focused on the issue of establishing a protective and legally mandated goal for surface soil (approximately 0-6 inches deep) concentration for Arsenic in residential or unrestricted land use situations. The proposed action level can be found in Section 3.6. Other land use situations (e.g., industrial or commercial) are addressed using risk-based corrective action guidance, which may include considerations and assumption about current and future land use. To ensure a manageable scope for this background document, it does not consider other potential exposures to Arsenic in food, drinking water or occupational settings such as deep mining, metal smelting, etc. The Department sought to write this draft background document in plain English to maximize its accessibility to a wide and general audience and help facilitate an informed public discussion of the topic.

1.1 Arsenic

The study of toxicology is dominated by the ancient adage,

Arsenic is perhaps the best example of this adage in practice. A recent National Geographic article referred to Arsenic as “…the poison of kings and king of poisons.”³Arsenic is colorless, tasteless and odorless.

Only a small amount of Arsenic can be harmful. Approximately 70-200 milligrams of Arsenic trioxide can be fatal. Based on this assumption, an amount of Arsenic weighing the same as one U.S. nickel (five grams) would contain 16 to 71 fatal doses. In the United States, the average daily dietary intake of Arsenic is approximately 30 micrograms, or 0.03 milligrams – approximately 0.04 percent of a fatal dose. Arsenic may enter the food chain through a variety of natural or unnatural mechanisms, such as eating shell fish, which have eaten algae in which naturally-occurring arsenic has accumulated. Hence, at least 2,000 times the average daily dietary dose in the U.S. would be required to receive a fatal dose.⁴

In the 5th Century B.C., Hippocrates used Arsenic to treat ulcers. In the 18th Century, Arsenic was an ingredient in Fowler’s solution, created in 1786, for treatment of psoriasis. In the early 20th Century, Paul Ehrlich considered the father of modern chemotherapy, promoted a form of Arsenic (“Salvarsan”) as the first treatment for syphilis. These therapies continued in use until the 1940s when they were replaced by modern antibiotics⁵. The FDA recently approved Trisenox (Arsenic trioxide) for the treatment of patients with acute promyelocytic leukemia.⁶

Traces of Arsenic found in the French Emperor Napoleon’s hair have led to speculation that he died of Arsenic poisoning, either intentionally or accidentally. Arsenic was found in the wall paper at his Longwood estate, which was painted with Scheele’s Green, which contains copper arsenide.⁷

The recent history of Arsenic in Delaware is one of clear public concern and need for effective public health protection. Delawareans have expressed concerns about the incidence of cancer and the potential link to environmental causes.⁸ The potential link between Arsenic and cancer has been raised in the news media and public meetings.⁹

Arsenic has not been produced in the United States since 1985, although the U.S. continues to be the world’s largest consumer (21,000 tons in 2003) with most Arsenic being imported from China.¹⁰ Arsenic has been used and distributed in the environment in a variety of ways. Here are some examples of Arsenic sources:

Although the selection of a standard for Arsenic is ultimately a policy issue decided with public involvement (See Sections 2 and 3), there are a variety of scientific disciplines that can help form this decision, including:

Table 1: Arsenic Content of Some Commercially Available Fertilizer Products and Potting Soils¹⁴

1.2 Arsenic Chemistry

Arsenic, a naturally occurring element, is found throughout the environment. Arsenic is found in two forms: organic and inorganic.

Generally, Arsenic combined with elements such as oxygen, chlorine and sulfur forms inorganic Arsenic compounds. Examples of inorganic Arsenic compounds include, but are not limited to, Arsenic pentoxide and trioxide. The pentoxide form of inorganic Arsenic is referred to as arsenate [As (V)] and is relatively immobile in soil. The trioxide form of inorganic Arsenic is referred to as arsenite [As (III)] and is relatively mobile in soil. A United States Geological Survey (USGS) and New Jersey Department of Environmental Protection (NJDEP) study found arsenite concentrations of more than 100 ppm in soil at depths of 100 cm (approximately 3 feet). Generally, Arsenic combined with carbon and hydrogen forms organic Arsenic compounds. Examples of organic Arsenic include, but are not limited to, arsanilic acid, arsenobetaine and dimethylarsinic acid.

Environmental Arsenic testing is conducted in the field or in a fixed laboratory as a measurement of total Arsenic. There are many test procedures used to calculate the concentration of total Arsenic in the environment. The most common test procedures utilized to test for Arsenic are: Atomic Absorption (AA), Inductive Coupled Plasma (ICP) and X-ray Fluorescence (XRF). XRF and AA/ICP are US EPA approved methods and the results received are equally accurate and precise. Only a qualified professional can interpret and compare the variations and similarities between the two methods. Arsenic risk values (see Table 2) are based on receiving AA/ICP results. Therefore, the AA/ICP results are generally used in performing risk assessments. The cost for performing AA/ICP analysis is much higher than XRF. Therefore, XRF data is used as a screening analysis for conducting site evaluations because of cost considerations.

The Standard Operating Procedure for Chemical Analytical Programs (SOPCAP) under the Hazardous Substance Cleanup Act provides detailed policy for the analysis of environmental samples and the use of XRF as a screening tool.

Table 2: Variations in Arsenic Risk Management State Reference Residential Use Non-Residential Use

1.3 Geology of Arsenic

Two regional geologic provinces are present in Delaware. The northern part of the state, north of a line extending between Newark and Wilmington, is underlain by the igneous and metamorphic bedrock of the Appalachian Piedmont Province. South of the Newark-Wilmington line, the remainder of the surface deposits of the state lie within the Atlantic Coastal Plain Province. The two provinces are very different in age, in the way they formed, and in the types of rocks they contain.¹⁵

The Piedmont Province consists of very old (480 million years of age) “hard rock” or crystalline metamorphic bedrock which is generally of two distinct types; (1) the Wilmington Complex and (2) the Glenarm Series. These rocks formed from shallow coastal sediments behind a volcanic island chain that was subjected to plate deformation causing high-grade heat and pressure.¹⁶,¹⁷

The Coastal Plain Province consists of a seaward-thickening wedge of sedimentary rocks and unconsolidated sediments from 120 million years of age to recently formed marsh deposits. This wedge of gravel, sand, silt, and clay thickens from 0 feet thick to nearly 8,000 feet thick at the southern border of Delaware. These sediments came from two sources; (1) non-marine sediments eroded from the Piedmont and the Appalachian Mountains and carried by streams to the coast, and (2) marine sediments that were deposited from the ocean at times when global sea level were high.¹⁸

There has been no comprehensive study of the relationship between the rocks of the Delaware Piedmont and the naturally-occurring Arsenic in the soil. There are a number of methods used to help determine “background” (natural and other) concentrations of arsenic, which are described in Attachment D and summarized here. One method for determining natural background is to analyze arsenic concentrations in soil deep below the ground surface, which can be obtained from “borrow pits” where clean soil is excavated for use as clean fill elsewhere. The minerals that predominate in the Piedmont rocks are not generally considered Arsenic bearing but the composition of in-place piedmont soils indicate that they do contain trace amounts of Arsenic. The borrow pit samples were obtained from an area where there was no evidence of prior excavation or disposal. The use of deep soil for determining natural background concentrations of arsenic is a valid method if there is no mechanism for arsenic to have migrated to the deep soil. Also, the arsenic is not likely to have migrated down into deep soils because the science indicates that arsenic does not migrate through soil except for very short distances and usually at only high source soil concentrations.

Soil samples collected from borrow pits in New Castle County near Concord Pike (Rt. 202 and I-95) approximately 30-50 feet below ground surface have been found to have Arsenic concentrations ranging from 3 to 18 parts per million (ppm) and a median of approximately 10 ppm (See Table 3). The source of Arsenic is probably from a Glauconite green sand layer formed millions of years ago. Arsenic was part of the combination of the original elements contained on Earth when the planet was formed approximately 4.5 billion years ago. Over the billions of years, arsenic has subsequently been redistributed by a variety of normal geologic events including volcanic eruptions.

Table 3: Borrow Pit Soil New Castle County¹⁹

There is evidence for a relationship between the type of geologic unit and the Arsenic soil concentration in the soil found in the Coastal Plain. The geologic creation of arsenic-bearing glauconite is the result of depositions of algae that take up, metabolize, and retain Arsenic in surface waters. When the algae dies and settles to the bottom, the Arsenic in the dead algae becomes part of the deposit. When this process happens over thousands of years, the Arsenic will accumulate in estuaries, which are now geologically mapped as Glauconite or greensands. Glauconite is dull green iron-silicate mica mineral found in shallow marine sediments. Bands of Glauconite are found in some of the Coastal Plain of New Jersey and southern New Castle County, Delaware, and in the deeper units in southern Delaware. The soils that form over these glauconite-containing units, called "greensands", are highly productive agriculturally and have been determined to contain naturally-occurring Arsenic.²⁰

A study by Dooley analyzed the composition of naturally occurring New Jersey greensands at seven sites, reporting a range of 7 to 31 mg/Kg (ppm) of Arsenic. Therefore, it has been shown that greensand soils contain Arsenic as "natural background". Greensands soils have been used as a soil amendment to improve productivity in other less productive areas (areas not mapped as having Glauconite deposits). As a result, it is likely that Arsenic containing Glauconite Greensands are now more widespread than the geologic mapping would indicate.²¹

The USGS reported Arsenic levels of 4.8 to 23 mg/Kg (ppm) in samples of stream sediments from the Lower Susquehanna River. Similar amounts of Arsenic were found in another USGS study in which the average arsenic soil concentrations were reported as 8.3 mg/Kg for New Castle County, 4.6 mg/Kg for Kent County, and 4.9 mg/Kg (ppm) for Sussex County.²²

A 1993 DNREC study found a similar distribution with average background concentrations of 10.6 mg/Kg in New Castle County, 7.8 mg/Kg in Kent County, and 8.7 mg/Kg in Sussex County. The maximum background reported was 48mg/Kg (ppm).²³

1.4 Toxicology of Arsenic

The toxicology of Arsenic mainly depends on its chemical and physical form, exposure route and dose. The most toxic form of Arsenic is the inorganic form not its organic form. Among the forms of inorganic Arsenic the trioxide [As (III)] form, with a valence state (electronic charge) of 3, is more toxic than the pentaoxide form [As (V)], which has a valence state of 5. Some research has demonstrated circumstances in which the organic form of Arsenic is converted to the inorganic form.²⁴ The type Arsenic, As (III) and/or As (V) in soil depends on the exposed environmental conditions (whether in an oxygen rich or depleted environment, the amount total organic carbon, and the ph of soil) that effect the Arsenic form. Generally, the Arsenate [As (V)] is the most common oxidized form of Arsenic in soil. Arsenic strongly sorbs to iron and hydroxides on soil particles. In addition, sandy particles have a lower capacity to sorb Arsenic than clay and silt and therefore usually have less Arsenic content.

The Food and Drug Administration established standards for the concentrations of Arsenic in chickens, turkeys and swine, which are often fed pharmaceutical feed containing roxarsone, of 2 ppm in uncooked edible byproducts and 0.5 ppm in uncooked muscle tissue and eggs.²⁵

The International Agency for Research on Cancer (IARC) review has identified Arsenic as a carcinogen with “Human Sufficient Evidence.”²⁶ A common toxicology reference refers to Arsenic as a “Confirmed human carcinogen producing liver tumors.”²⁷ Arsenic in drinking water is associated with bladder cancer.²⁸

Symptoms from Arsenic exposure include hyperkeratosis, blackfoot disease, myocardial ischemia, liver dysfunction, epithelioma, and several cancers. The major risk from exposure to Arsenic is not from inhalation, but rather by ingestion.²⁹

1.5 Human Health Risk Assessment for Arsenic in Soil

Human Health Risk Assessment is a process to estimate the chance that a person will be harmed now or in the future if that person comes in contact with chemicals (for example Arsenic) present in a property (site). This process produces numbers that show how great (or small) the risks may be. For example, a conservative risk assessment would indicate that an individual living on a residential property for 30 years with 4 ppm (mg/Kg) of Arsenic in soil and ingesting (eating) 200 mg of soil as a child and 100 mg of soil as an adult each day for 350 days per year for 30 years will have a chance of developing cancer somewhere between zero and one in 100,000. For non-cancer health risk the acceptable standard is expressed as Hazard Index (HI) of one (1). The hazard quotient of 1 is the ratio of a representative site concentration in soil to that of a reference dose concentration determined to have non-cancer health effects. The HI of 1 for arsenic is 23 ppm, which is based on the assumption that a child age 1-6 years in a residential setting will ingest 200 mg/day of soil.

There is significant conservatism built into the risk assessment process. One of the assumptions is that 100% of the Arsenic in the soil ingested by an individual will be absorbed by the human body (bioavailability). The USEPA Risk Assessment Guidance for Superfund (RAGS) Volume 1 Human Health Evaluation Manual (1989) includes a discussion of determining the relative bioavailability of a chemical in the media of interest such as soil.³⁰ The bioavailability of a chemical in a soil matrix is influenced by many factors including the physical and chemical interaction with the matrix, in addition to the solubility and biological factors. For arsenic the oral toxicity factor that predicts the potential for causing cancer is based on humans ingesting arsenic in drinking water over long periods of time in Taiwan. These people experienced an increase in non-fatal skin cancers as compared to a group of people that did not ingest arsenic in drinking water. In this study, the arsenic is assumed to be 100% bioavailable for uptake into the gastrointestinal tract and the bloodstream. Arsenic in drinking water is present in a soluble form and therefore the bioavailability is high.

The arsenic present in soil is non water-soluble and binds to the soil matrix so less arsenic is available to be absorbed in human body. A recent study by Roberts et. al.³¹ showed that only about 10% to 25% of the Arsenic in soil was absorbed by monkeys when they were fed arsenic contaminated soil. The State of Florida Department of Environmental Protection (FDEP) has also published a study³² with extensive research by the University of Florida to determine the relative bioavailability for arsenic in soil as compared to the high bioavailability of arsenic in drinking water. FDEP used this information to recently make a protective science-based policy decision to adopt the default relative oral bioavailability factor of 33%, as a worst case risk assessment where there is potential for individuals to ingest arsenic in soil. This was implemented in February 2005.³³

If DNREC were to utilize 33% as the factor to reflect the worst case bioavailability for arsenic in Delaware soil, the result would be a three-fold increase in the soil standard for arsenic to 12 ppm for 1 in 100,000 cancer risk level (1 x 10-5).

Other conservative assumptions:

Exposure Assumptions used by DNREC for Arsenic in soil is the worst case scenario which assumes that a child resident will ingest 200 mg of soil and an adult resident will ingest 100 mg of soil 350 days/year for a total of thirty years. This does not include a pica child.

Toxicity Assessment looks at how much of a substance causes what kind of harm to humans. Toxicity to humans is not usually measured directly by intentionally exposing people, for obvious ethical reasons. Rather it is determined indirectly, usually by extrapolation of animal studies to humans. Many conservative assumptions are made which include:

It is apparent from the discussion that uncertainties are inherent in the toxicity assumptions.

2.1 Arsenic Standard Setting Process

The process for selecting a soil concentration action level is ultimately a policy choice – a decision to be made with public involvement. The public has a fundamental right to be involved in decisions that could affect them. The decisions can and should be informed by good science and engineering, and much of this background document seeks to provide that information in a clearly accessible manner. Nonetheless, despite this sea of scientific information, however deep and carefully plumbed, the decision remains a policy choice. Hence, the process used by DNREC’s Air and Waste Management Division will be focused on a constructive public involvement process, using the following steps:

1. Develop a background document and proposed revision for internal and interagency review.

2. Develop a public involvement plan, ensuring adequate opportunities for meaningful input from a variety of stakeholders, including various Department Advisory Committees and representatives of all interested parties.

3. Conduct public workshops to present and accept comments on the draft revision to the Arsenic action level and the supporting documentation.

4. Based on the workshop comments and the advisory committee inputs, develop a policy for signature by the Division Director, after review by other cooperating DNREC divisions and state agencies (DHHS/DPH, DDA, DELDOT). 5. Implement the policy with adequate opportunities for public oversight and involvement to ensure it is being implemented fairly.

6. Review and revise as necessary the action level to ensure it is keeping pace with new scientific research developments and to determine whether there are any implementation issues arising.

2.2 What Dose of Arsenic Causes Health Impacts?

The dose of arsenic³⁴ causing a health impact depends on a variety of factors, including:

The prudent and most conservative (i.e., most protective) basis for evaluating health impacts from exposure to toxic substances are to focus on the health impacts that occur at the lowest dose. In the case of Arsenic, no lower threshold is recognized by EPA’s Science Advisory Board³⁵ – i.e., some risk of cancer exists for virtually any Arsenic exposure. Hence, exposure standards for carcinogens like Arsenic are expressed in terms of the lifetime probability of dying of cancer, such as one chance in a million or one in ten thousand.

The underlying statute governing the Department mandate to protect human health and the environment regarding hazardous substances such as Arsenic is the Hazardous Substance Cleanup Act (HSCA), 7 Del.C. Chapter 91 law.³⁶ The HSCA regulations require that the Department perform cleanup activities to achieve standards that are protective of cancer risks using a lifetime cancer risk probability of 1/100,000 or, in loose scientific notation, 1 x 10-5 . In the case of Arsenic, however, the background concentrations of Arsenic averages approximately 11 ppm (with a range of up to 25-30 ppm), which is more than double the soil concentration (4 ppm) associated with a risk level of 10-5 for residential properties. The law and regulations directs that, in cases where the background concentration is higher than the concentrations associated with a risk level of 10-5, then the background concentration shall be the cleanup goal for residential properties.

In addition to cancer, Arsenic is known to cause a variety of non-cancer effects, including neurological dysfunction. For non-cancer health risk, the acceptable standard is expressed as Hazard Index (HI) of 1. The hazard index of 1 is the ratio of a representative site concentration in soil to that of a reference dose concentration determined to have non-cancer health effects. The HI of 1 for arsenic is 23 ppm, which is based on the assumption that a child age 1-6 years in a residential setting will ingest 200 mg/day of soil.

2.3 Legal Mandate for Protecting Public Health

DNREC’s legal mandate and authority for establishing action levels for soil cleanup (including Arsenic) to protect public health is based on the HSCA,³⁷ including the authority to promulgate regulations.³⁸ (See Attachment C). The law directs DNREC to establish procedures “for identifying cleanup levels based on site specific risks.”³⁹ The HSCA regulations provide a limited basis for considering cost as well as “background” concentrations, which allows for some consideration of technical practicability (i.e., can it be achieved realistically using available technology) in the law.

The Secretary of DNREC has promulgated regulations under this HSCA authority to establish procedures for determining cleanup levels for releases of hazardous substances, which includes:

Accordingly, control of future land use is essential to the protectiveness of the remedy in cases where a cleanup assumes a future land use other than “unrestricted.”⁴¹

The statute governing the Department mandate to protect human health and the environment regarding hazardous substances such as Arsenic is the HSCA 7 Del.C. Chapter 91. Section 9.4 of the HSCA regulations sets out a risk-based approach for establishing soil cleanup levels, and it identifies two types of risk to human health which must be protected: cancer risks and non cancer risks. The cancer risk level established in Subsection 9.4(2) is “10-05” (also expressed as 10-5) which is defined in Subsection 2.1 to mean “the potential risk for one additional cancer death caused by exposure to a carcinogen in a human population of 100,000 in a lifetime.”

The HSCA regulations require that the Department perform cleanup to achieve levels that are protective of cancer risks using a lifetime cancer risk probability of 1/100,000 or, in loose scientific notation, 10-5 . However, when the background concentration of a hazardous substance is higher than the 10-5 risk level, the regulations require that the background concentrations be used as the cleanup level. In the case of Arsenic, the background concentrations of Arsenic average approximately 11 ppm (with a range of up to 25-30 ppm) as discussed in Section 1.3 and Attachment D. Consequently, because the observed background concentration is higher than the concentrations associated (using standard dose and exposure assumptions) with a risk level of 10-5 (or a 4 ppm concentration for residential properties), the law directs DNREC to use background concentrations as a cleanup goal.

The law appears clear on this issue, and this is the direction DNREC staff will follow in conducting and overseeing cleanups: remove contaminated soil until a residual concentration is achieved that is equivalent to the local natural background for the area where possible. A “default background” concentration of 11 ppm should be used in cases where the guidance or historic contamination does not allow for a meaningful determination of background that is adequately protective.

2.4 Standards used by USEPA and other states

Several methods are used to determine cleanup standards. Some states call these standards screening values, other use default/site-specific background and/or site specific risk assessments and/or some combination of the above. Table 2 illustrates the various risk management strategies (Arsenic standard application) by the USEPA and states throughout the region.

Some examples of cleanup goals in certain western sites where industrial contamination has occurred (e.g., ASARCO smelter in Tacoma Washington) include residential land use cleanup of 100 ppm. In Anaconda, Montana, EPA uses residential cleanup concentrations of 250 pm Arsenic.

In the District of Columbia, a panel recommends the adoption of the 20 ppm remediation level as proposed by the USEPA. The Panel believes that the 20 ppm remediation level should not pose a health hazard to the community.⁴²

Many states, however, seek to reduce Arsenic concentrations to default or site specific background levels. Table 4 illustrates the default or site specific background value used by various states throughout the region. It is important to note that federal law and Delaware regulations do not require site cleanup below the background level.

Determining an Arsenic cleanup standard or goal is a difficult task, the flow chart below outlines Delaware’s Arsenic risk management strategy for the selection of a cleanup standard.

Table 4: State Specific Default Arsenic Background Level/Range⁴³

Arsenic Management Strategy

(Flowchart graphic not supported. Contact either DNREC or the Registrar’s office for a copy)

2.5 Environmental Justice Considerations

The USEPA defines environmental justice (EJ) as the fair treatment for people of all races, cultures, and incomes, regarding the development of environmental laws, regulations and policies.⁴⁴DNREC has adopted this definition and policy and strives to implement cleanup strategies equitably in all communities throughout the State with explicit attention to ensuring adequate protection and public participation in situations where disadvantaged populations may bear a disproportionate amount of adverse environmental health effects from pollution. The Department ensures that facilities and environmental sites are held to the same standards throughout the State.

In determining safe cleanup standards for Arsenic, EJ considerations are of paramount importance. Diverse communities are often located in older neighborhood once home to early Delaware industry, including tanneries. For this reasons, it is extremely important that the process of determining “background” concentrations do not result in a less protective cleanup goal being used, simply because of a generally higher level of Arsenic not of natural origin. In other words, citizens should not be given any less protection simply because they live in old industrial neighborhood. This would not only be intolerably unjust, but would defeat the very purpose for which the environmental cleanup program is focused – to cleanup contaminated sites and make communities safe and livable.

To assure that the legacy of higher concentrations of arsenic that might occur in the soil of these communities does not result in less protective cleanup standards, DNREC staff will not use the “background” determination process prescribed in guidance for a local area unless an area can be found that is not impacted by historic contamination. If no un-impacted area can be found, a cleanup goal of 11 ppm, the average background concentration of Delaware soil, will be utilized. It is a Department priority to make environmental justice considerations paramount in these communities.

There is no objectively “correct” soil concentration standard based solely on scientific and engineering information, although this information can certainly be useful. The selection of a soil concentrations action level and cleanup goal is ultimately a policy choice – a decision to be made with public involvement with as much public transparency as possible. The public has a fundamental right to be involved in decisions that could affect them. The decisions can and should be informed by good science and engineering, and much of this background document seeks to provide that information is a clearly accessible manner. Nonetheless, despite this sea of scientific information, however deep and carefully plumbed, the decision remains a policy choice. Hence, this section presents a series of possible concentrations and the implications to help inform a policy dialogue and allow for meaningful public involvement.

3.1 Option A - 0.4 ppm

Using standard risk assessment exposure assumptions (e.g., how much soil a child or adult eats,⁴⁵) a soil concentration of 0.4 ppm would result in an incremental⁴⁶ lifetime⁴⁷ cancer risk increase of one in a million. This risk probability (1/1,000,000) is on the conservative end of the spectrum for “acceptable risks” that have been selected in the United Sates during the past 35 years of environmental policy making. Generally, environmental policy decisions in the U.S. have selected risks ranging from a high end risk of one in ten thousand (1/10,000) to one in a million (1/1,000,000). This risk-to-dose relationship (i.e., 0.4 ppm equals 1/1,000,000) also assumes that all of the Arsenic to which a person is exposed is the Arsenic in the most toxic form (i.e., it is inorganic Arsenic not organic Arsenic), and that the inorganic Arsenic is all trioxide [As (III) with a valence state (electronic charge) of 3], although some Arsenic exists in organic form and some inorganic Arsenic has a valence state of 5 [As(V)] or pentaoxide), rather than 3 (See Section 1.2 on Arsenic Chemistry). It also assumes that 100% of the Arsenic in soil will be absorbed by the individual. A recent study with monkeys showed that only up to 25% is absorbed (bioavailability).

Perhaps the most obvious consideration to evaluating this 0.4 ppm option is technical feasibility. This 0.4 ppm concentration standard cannot be achieved realistically in the field during cleanups because the background concentrations of Arsenic found in the soil in Delaware are significantly higher than 0.4 ppm (See Table 4) Also, 0.4 ppm falsely implies a level of precision to the 1/10th of a ppm that is unattainable using typical analytical instrumentation.

3.2 Option B - 4 ppm

Using similar risk assessment exposure assumptions, a soil concentration of 4.0 ppm would result in an incremental lifetime cancer risk increase of one in a one hundred thousand. This risk probability (1/100,000) is in the middle of the spectrum for “acceptable risks” that have been selected in the United Sates during the past 35 years of environmental policy making.⁴⁸

This risk-to-dose relationship (i.e., 4.0 ppm equals 1/100,000) is also based on the conservative assumption that all of the Arsenic to which a person is exposed is the Arsenic in the most toxic form (i.e. it is inorganic Arsenic not organic Arsenic), and that the inorganic Arsenic is all trioxide (As (III) with a valence state (electronic charge) of 3), although some Arsenic exists in organic form and some inorganic Arsenic has a valence state of 5 [As(V) or pentaoxide], rather than 3 (See Section 1.2 and 1.3 on the Chemistry and Geology of Arsenic). It also assumes that 100% of the Arsenic in soil will be absorbed by the individual. A recent study with monkeys showed that only up to 25% is absorbed (bioavailability).

This 4.0 ppm concentration would be difficult to achieve during cleanups in most situations because the background concentrations of Arsenic found in the soil in Delaware are, on average, higher than 4.0 ppm (See Table 4). This option could create a very significant budget shortfall to address many of the historic fill sites in the State that may not pose a risk to human health or the environment. Paradoxically, setting a more strict cleanup goal may result in less health protection because fewer sites may be cleaned up with the available funding, leaving some sites completely untouched by remediation, while other sites are cleaned up to more stringent cleanup goals.

3.3 Option C- 11 ppm

Using similar risk assessment exposure assumptions, a soil concentration of 11 ppm would result in an incremental lifetime cancer risk increase of approximately three in a one hundred thousand. This risk probability (3/100,000) is roughly in the middle of the spectrum for “acceptable risks” that have been selected in the United Sates during the past 35 years of environmental policy making. It is important to note that risk probabilities are generally not intended to be interpreted as anything more precise that order of magnitude (i.e., 1/10th or 1/100th NOT 2/100 or 3/100) estimates, so a risk extrapolation of 3/100,000 may imply greater precisions than is technically possible.

This risk-to-dose relationship (i.e., 11 ppm equals 3/100,000) is also based on the conservative assumption that all of the Arsenic to which a person is exposed is the Arsenic in the most toxic form (i.e., it is inorganic Arsenic not organic Arsenic), and that the inorganic Arsenic is all trioxide (As (III) with a valence state (electronic charge) of 3), although some Arsenic exists in organic form and some inorganic Arsenic has a valence state of 5 [As(V) or pentaoxide], rather than 3 (See Section 1.2 on the Chemistry of Arsenic). It also assumes that 100% of the Arsenic in soil will be absorbed by the individual. A recent study with monkeys showed that only up to 25% is absorbed (bioavailability).

This 11 ppm concentration would be possible to achieve during cleanups in most situations because the background concentrations of Arsenic found in the soil in Delaware are, on average, higher than 4.0 ppm (See Table 4). In some cases, the wide area background concentrations of Arsenic are higher than 11 ppm, and significant resources could be used in seeking to attain a cleanup goal of 11 ppm, which would be a relatively small decrease in the Arsenic concentration and accordingly a small incremental decrease in risk.

This option could create a significant budget shortfall to address many of the historic fill sites in the State that may not pose a risk to human health or the environment. Paradoxically, setting a more strict cleanup goal may result in less health protection because fewer sites may be cleaned up with the available funding, leaving some sites completely untouched by remediation, while other sites are cleaned up to more stringent cleanup goals.

3.4 Option D – 23 ppm

Using similar risk assessment exposure assumptions, a soil concentration of 23 ppm would result in an incremental lifetime cancer risk increase of one in twenty thousand. This risk probability (approximately 1/20,000) is toward the high end of the spectrum for “acceptable risks” that have been selected in the United Sates during the past 35 years of environmental policy making.⁴⁹

This risk-to-dose relationship (i.e., 23 ppm equals 1/20,000) is also based on the conservative assumption that all of the Arsenic to which a person is exposed is the Arsenic in the most toxic form (i.e., it is inorganic Arsenic not organic Arsenic), and that the inorganic Arsenic is all trioxide [As (III)] with a valence state (electronic charge of 3), although some Arsenic exists in organic form and some inorganic Arsenic has a valence state of 5 [As(V) or pentaoxide] rather than 3 (See Section 1.2 on the Chemistry of Arsenic).

The then DNREC-DAWM Director signed a memorandum in June 2004 designating 23 ppm as the interim action level and later referred to it as a standard. The basis for this number appears to be the use of a hazard index of 1.0. Because the carcinogen risk is lower (i.e., 4 ppm associated with a risk of 1/100,000) then the hazard index number is not controlling. Moreover, because the background level is lower than this hazard index number, then it does not appear to have a role in carrying out DNREC’s mandate to protect human health and the environment to the extent feasible in HSCA.

3.5 Option E – 40 ppm

Using similar risk assessment exposure assumptions, a soil concentration of 40 ppm would result in an incremental⁵⁰ lifetime⁵¹ cancer risk increase of one in a ten thousand. This risk probability (1/10,000) is on the high end of the spectrum for “acceptable risks” that have been selected in the United Sates during the past 35 years of environmental policy making.⁵²

This risk-to-dose relationship (i.e., 40 ppm equals 1/10,000) is also based on the conservative assumption that all of the Arsenic to which a person is exposed is the Arsenic in the most toxic form (i.e., it is inorganic Arsenic not organic Arsenic), and that the inorganic Arsenic is all trioxide [As (III) with a valence state (electronic charge) of 3], although some Arsenic exists in organic form and some inorganic Arsenic has a valence state of 5 [As(V) or pentaoxide], rather than 3 (See Section 1.2 on the Chemistry of Arsenic).

This arsenic concentration limit has historically been used for industrial sites as an action level and cleanup goal. This limit should not therefore be considered adequately protective of individuals in residential settings, using common exposure and dose assumptions.

3.6 DNREC Option Analysis

Based on the above options, and the information currently available, the Department proposes the continuation of a risk-based approach to cleanup in accordance with 7 Del.C. Chapter 91 and the HSCA regulations. The HSCA and regulations allow for the use of background levels. In cases where a local background level cannot be developed, the Department will utilize the default background level of 11 ppm.

The revised Arsenic action level will be for all cleanup sites for which future land use is reasonably anticipated to be residential or unrestricted and under the regulatory authorities of the Division of Air and Waste Management. The sites include old industrial properties being redeveloped, active facilities conducting a cleanup and old industrial site undergoing a cleanup. The revised action level will be effective in draft immediately upon release and as a final action level upon adoption by the Secretary of the Department.

4.1 Process for Determining Background Concentrations

The “background concentration” approach relies on empirical correlation between bulk soil concentrations and presumed least impacted, NOT “natural”, locations in the State. Establishing a natural (i.e., uncontaminated by any anthropogenic, or human produced or moved, Arsenic) soil location and concentrations is very difficult. Ultimately, it is impossible to prove a negative – in the case, “Prove that this site has never been contaminated by human beings.” It is important to recognize that background standards merely provide some context for risk-based standards. As a technical analysis, background standards cannot be considered a substitute for consideration of fate, transport, exposure and risk, but a full site-specific risk assessment is not always feasible, and cleaning up to background is as much as is technically feasible in most cases. The procedure for determining site specific background levels is found in the Delaware Hazardous Substance Cleanup Act Remediation Standards Guidance, http://www.dnrec.state.de.us/dnrec2000/Divisions/AWM/sirb/DOCS/PDFS/Misc/RemStnd.pdf. This document will be used by DNREC staff to establish site specific background levels.

4.2 Annual Review Process

The Department will compile all relevant Arsenic data for sites regulated by the Department and sampled after the effective date of the revised action level. The data will be reviewed annually by the Department in consultation with staff from DDA and DHHS/DPH. Each year by September 1 the Department will report on the results of its review and invite public comment, providing public workshops and following the public participation plan process to review and revise the policy. Each year the public review process will involve interested Department advisory committees and the following questions: What should the Arsenic action level be? Should it apply to golf courses, orchards or other sites with properly applied pesticides and fertilizers? Is the public involvement plan adequate? If not, how should it be changed?

5.0 Public Participation Plan

In order to ensure the public has ample opportunity to participate in the development and adoption of an Arsenic action level for the State, a public participation plan is being developed, (See Table 5). DNREC encourages and welcomes comments from the general public and all interested parties. In an effort to do so, DNREC will involve all interested advisory committees, host a minimum of one public workshop in each county, offer to attend any civic or other organization meeting, notice the proposed draft policy in the Delaware Register of Regulations and develop press announcements on the draft policy proposal. DNREC will staff all meetings and take notes on the issues and concerns raised. The notes will be made available to the attendees and all interested parties both in hard copy and via the DNREC web site. In addition, DNREC will take written comments either in hard copy or email format. All comments will be reviewed, when possible categorized and responded to as a part of the action level development. Once all comments have been reviewed and a draft response prepared, the DNREC will hold additional public advisory committee meetings to review and discuss the responses and any proposed changes to the draft document. Once the issues raised from the comments have been resolved to the satisfaction of the majority of the advisory committees, by vote, the final policy will be noticed in the Delaware Register of Regulations, the News Journal and press releases will be sent out. In addition, the DNREC will be prepared to provide a presentation to any organization or governmental entity that requests such.

The DNREC will provide staff support to the advisory committees for both the administrative and technical needs. The DNREC will coordinate with other State and Federal agencies and when necessary, seek outside contractual support for technical issues. This will require significant staff resources and may require the reassignment or delay of other duties to complete the project on the anticipated timeline. The resources will be required from both the Site Investigation and Restoration Branch and the Public Affairs Office.

Table 5: Implementation Schedule

ATTACHMENT A: Governor’s Directive

State of Delaware

Office of the Governor

Ruth Ann Minner

Governor

To: Secretary John Hughes

From: Governor Ruth Ann Minner

Date: June 9, 2005

Cc: David Small

James Werner

Re: Protecting Delaware Citizens from Arsenic

Arsenic is among a number of toxic substances known to cause cancer and to which our citizens may be exposed in their communities and workplaces. Whether it is derived from industrial or natural sources, we have a duty to protect our citizens from harmful exposure to toxic substances, such as arsenic.

It has become clear that there is a lack of understanding regarding the Department's rationale and method for setting an acceptable limit for arsenic in soil and whether the current level is protective of public health. This situation has been caused in part by numerous changes to the limit over the years and in part by a lack of public input in the recent change in the limit.

Accordingly, I am directing you to lead an immediate and expedited review of standards and policies related to arsenic cleanup. This review should include:

My request is that you begin this review immediately and take action to propose appropriate standards and policies within two weeks.

ATTACHMENT B: Staff Work Group Meeting

Attendance at June 7, 2005 Staff Work Group Meeting

To Discuss Changes in DNREC Risk Management for Arsenic

(Lukens Drive Building, New Castle, DE)

DNREC/DAWM:

DNREC/DWR:

Division of Public Health:

Department of Agriculture:

AG:

ATTACHMENT C: Legal Basis For Arsenic Risk Management Standards

The Hazardous Substance Cleanup Act, 7 Del.C. Chapter 91 (“HSCA”), gives the Secretary of the Department of Natural Resources and Environmental Control (“DNREC”) certain powers and authorities to investigate and remediate releases of hazardous substances. See, generally, 7 Del.C. Section 9104. Section 9104(a)(2) grants the Secretary the authority to draft regulations:

In drafting regulations, Section 9104(b)(2)(g) directs the Secretary to establish procedures “for identifying cleanup levels based on site specific risks.”

The Secretary has used this authority granted by HSCA to promulgate the Regulations Governing Hazardous Substance Cleanup (“Regulations”). Section 9 of the Regulations establishes procedures for determining cleanup levels for releases of hazardous substances. Subsection 9.1(1), which establishes a risk based approach to cleanup levels, states:

Section 9.4 sets out a risk-based approach for establishing soil cleanup levels, and it identifies two types of risk to human health which must be protected: cancer risks and non cancer risks. The cancer risk level established in Subsection 9.4(2) is “10E-05” which is defined in Subsection

2.1 to mean “the potential risk for one additional cancer death caused by exposure to a carcinogen in a human population of 100,000 in a lifetime.” The non-cancer risk level established in Subsection 9.4(2) is a “hazard index value of one.” A “hazard index” is defined in Subsection 2.1 to mean “the numerical value obtained by dividing a person’s expected daily intake of a non-carcinogen by a level which is not expected to produce toxic effects.” By establishing a “hazard index of one” (expressed as “HI-1”) the Regulations require a soil cleanup level where the daily intake of a hazardous substance is not expected to produce toxic (i.e. non-cancer) effects.

The Regulations also provide that the risk-based cleanup level for a hazardous substance will vary depending on the type of use (such as residential, commercial or industrial) which will occur on the site or property. Subsection 9.1(5) sets out two types of cleanup levels where the risks from hazardous substances are quantifiable: compliance cleanup levels (also referred to as “unrestricted” or “residential” cleanup levels), and conditional cleanup levels (also referred to as “restricted” or commercial or industrial” cleanup levels). These are defined in Subsection 9.1(5) as follows:

(a) Compliance cleanup levels: These will be established at concentrations which are protective of public health, welfare, and the environment and, which require no restrictions on the use of the facility. Compliance cleanup levels shall be established in accordance with Subsections 9.2 – 9.4 and as directed by the Department.

(b) Conditional cleanup levels: These represent concentrations which are protective of public health, welfare, and the environment under restricted facility use conditions. Conditional cleanup levels may be established where the person undertaking the remedy can demonstrate that such levels are consistent with state and federal laws, that all practicable methods of treatment are utilized, and that institutional controls are implemented in accordance with conditions as determined to be appropriate by the Department.

Generally, compliance (i.e. unrestricted/residential) cleanup levels are lower than conditional (i.e. restricted/commercial or industrial) cleanup levels for a given hazardous substance because a residential setting will result in more prolonged exposure to the hazardous substance (in terms of hours of exposure over a lifetime) than will a commercial or industrial setting. Therefore, it will represent a higher risk, and the risk-based cleanup level will be lower in a residential setting than in a commercial or industrial one. When the risks from a hazardous substance are not quantifiable, Subsection 9.1(5)(c) states:

When there are multiple contaminants at a facility, the cleanup level of each contaminant shall be such that sum of the risks posed by the contaminants shall not exceed 10E-05 cancer risk or a hazard index value of one.

Many hazardous substances, like Arsenic, are naturally occurring. Once a risk-based cleanup level is established for a hazardous substance, it must be compared to the natural background level for that hazardous substance. Section 2.1 of the Regulations defines “Background” or “Natural Background” as “the level of contamination present in an area from naturally occurring substances, excluding contaminants and other contributions resulting from human activity.” Essentially, this represents the level of a naturally occurring hazardous substance that would have existed in the environment before human activity.

The Regulations do not permit cleanup levels to be set below the natural background level of a hazardous substance. Subsection 9.4(2)(a) states:

When the natural background level exceeds the 10E-05 cancer risk level or a hazard index value of one level, for direct exposure or inadvertent ingestion, then the background level will be the cleanup level.

Thus, where the natural background level is higher than the risk-based cleanup level for a hazardous substance, the background level becomes the cleanup level.

The Regulations require that the cleanup levels established in Section 9 be used in determining an appropriate remedial alternative (i.e. type of cleanup or “remedial action”) for a particular site.

The procedure for developing remedial alternatives is described in Subsection 8.5(4)(a), which states:

After the initial screening is performed, an evaluation shall be conducted of the remaining alternatives considering the following factors:

Subsection 8.4(4)(c) describes how the remedial alternatives are ranked in order of preference:

For remedial action alternatives which comply with Subsection 8.5(4)(b)(i) and (ii), and satisfy the remaining evaluation criteria of subsection 8.5 (4)(b), preference shall be given to the remedial action which is most cost effective, and cost shall include present and future direct and indirect capital costs, operation and maintenance costs, compliance monitoring costs, and other foreseeable costs.

The method of selection of a remedial action from the remedial alternatives is set out in Subsection 8.6 which states:

The Department shall select a remedial action from the alternatives developed for the facility based on the determination of which remedial action complies with Subsection 8.5(4)(b)(i) and (ii) and best complies with the remaining criteria in Subsection 8.5(4)(b), and complies with Subsection 8.5(4)(c).

Therefore, the selected remedial action must be one that: 1 complies with the cleanup levels established in Section 9 Subsection 8.5(4)(b)(i);

2 complies with all applicable local, state and federal laws and regulations (Subsection 8.5(4)(b)(ii);

3 best complies with the remaining criteria in Subsection 8.5(4)(b)(iii)-(ix); and

4 is the most cost effective (Subsection 8.5(4)(c)).

ATTACHMENT D: Natural Background Concentrations of Arsenic in Delaware Soils

Determining background concentrations of Arsenic is important to risk management and standard setting because the background information typically provides a lower boundary below which cleanup concentrations cannot normally be achieved. Accordingly, it is equally important that DNREC exercise great care in it evaluation of background concentrations to seek to distinguish between natural background concentrations and background concentrations of Arsenic that may result from widespread distribution as a result of human activity (i.e., “anthropogenic” Arsenic). For this reason, DNREC staff analyzed multiple sources of data on arsenic concentrations to better understand background concentrations and seek to determine what a reasonable “natural” concentration of arsenic would be.

There are a number of methods used to help determine “background” (natural and other) concentrations of arsenic: average, the 95% upper confidence level to name a few. The following is a brief summary in determining background: One method for determining natural background is to analyze Arsenic soil concentrations below the ground surface, which can be obtained from “borrow pits” where clean soil is excavated for use as clean soil elsewhere such at new construction sites. Soil samples collected from borrow pits in New Castle County near Concord Pike (Rt. 202 and I-95) approximately 30-50 ft below ground surface have been found to have Arsenic concentrations ranging from 3 to 18 ppm and a median of approximately 10 ppm (See Table 3). The source of Arsenic is probably from the formation of Glauconite over many years. The minerals that predominate in the Piedmont rocks are not generally considered Arsenic bearing but they do contain trace amounts of Arsenic. The borrow pit samples were obtained from an area where there was no evidence of prior excavation or disposal. The use of deep soil for determining natural background concentrations of Arsenic is a valid method if there is no mechanism for Arsenic to have migrated to the deep soil. Also, the Arsenic is not likely to have migrated down into deep soils because the science indicates that Arsenic does not migrate through soil except for very short distances and at high source soil concentrations. Although analysis of deep borrow pit data may not be representative of surface soils to which the public may be exposed (i.e. surface soils may have more organic matter), it is useful as a basis for comparison. For example, if deep borrow pit soils have comparable levels of arsenic to surface soils in undisturbed areas, it suggests that Arsenic from air sources is an insignificant source of contamination. In addition, scientific data suggests that arsenic is not very mobile except in extremely high concentrations also supports this hypothesis.

A second method for determining background concentrations of arsenic is direct analysis of Delaware surface soil from location not believed to have been contaminated from industrial sources. DNREC staff analyzed soil samples from various locations throughout the state (See Figure 1) to better understand soil concentrations of Arsenic and to contribute, along with other sources of information, to a determination of background concentrations, natural and otherwise in Delaware. This analysis is described in more detail in a technical background memorandum from Rick Greene, DNREC/DWR to James D. Werner, Director, DNREC/DAWM http://www.dnrec.state.de.us/dnrec2000/Divisions/AWM/sirb/clnupnum.asp.

The DNREC soil assessment considered two primary datasets with a total of 55 samples analyzed: 20 soil samples collected from various parks in the Wilmington area at locations unaffected by any known direct industrial input; and 35 soil samples collected through Delaware at background locations as part of analysis to determine area background concentrations during waste site assessments. The samples from this second data set were collected at locations similar in soil and other geological characteristics, but where there was no evidence of being affected by the waste disposal or contamination at the subject waste site.

The results of this analysis provide useful insight into understanding background concentrations of Arsenic in Delaware soils. First, the background concentration of Arsenic in Delaware soils is not a single, constant value. Rather 'background' is a range of values, which can be described as data distributions. Based on the available data, the range of background Arsenic concentrations in Delaware soils falls between 0.58 and 31 ug/g dw (micrograms Arsenic per gram of soil on a dry weight basis, which is approximately the same as milligrams per kilograms or parts per million or ppm).

A second observation regarding this soil concentration data is that the soil concentrations are not distributed with a central tendency of average concentration (i.e. clumped in the middle of the range like a bell-shaped curve of the largest number of sample concentrations in the middle and fewer data points at the low and high extremes, often referred to as a “normal” distribution). Instead, the data appears to be distributed in a “logarithmic” pattern. The implication of this data distribution is that selecting an average concentration to reflect the observed background concentrations would be relatively arbitrary and would not reflect the actual concentrations of Arsenic concentrations found in soil in Delaware. Moreover, characterizing an average concentration as reflecting “background” would result in half of observed soil concentrations therefore being “above background” when in fact they are actually legitimately within the range of observed background concentrations, and simply above average.

A third observation from the soil Arsenic concentration assessment was the analysis of the upper end of the observed soil concentrations. Using all available background Arsenic data statewide, the 95th percentile concentration (the concentration below which 95 percent of all soil samples are expected to be found) is 29.1 ppm. The 95th percentile for all New Castle County (including Wilmington data) is 21.6 ppm. The 95th percentile concentrations for Kent and Sussex Counties are 24.8 and 14.9 ppm, respectively. This latter observation indicates that, despite the widespread application of chicken litter containing arsenic residues, no widespread elevated concentrations of arsenic in Sussex County were observed in this data analysis.

Finally, a third method of evaluating background soil arsenic concentrations is to compare Delaware data to national data. The U.S. Geological Survey in 1984 published a comprehensive analysis of thousands of soil sample form around the country, including analysis of Arsenic.⁵³ The results of the analyses show that the range of Arsenic concentrations is up to nearly 100 ppm, with an average concentration of approximately 10 ppm. The higher concentrations tend to be present in western alkaline soils. Although this range does not necessarily reflect the pattern DNREC believe to be present in Delaware, it indicates that the other data sources are in the same order of magnitude range of observed concentrations.

Because of the similarity of the findings for multiple studies, using independent data sets and independent analyses, the results appear to be reproducible. In scientific studies, when an analyst observes such reproducibility of results, it tends to provide additional support for the robustness (i.e., reliability and confidence) of the results.

ATTACHMENT E: University of Delaware Study: Scope, Summary and Schedule

http://www.dnrec.state.de.us/dnrec2000/Divisions/AWM/sirb/clnupnum.asp

Arsenic Study by University of Delaware

The University of Delaware (Dr. D.L. Sparks and Dr. J. Thomas Sims) prepared a scope of work to address Arsenic (As) in Delaware soils in conjunction with the Department of Natural Resources and Environmental Control under a Collaborative Agreement.

Background: Arsenic commonly occurs in soil and water due to natural geological processes and due to human activities. Agricultural sources of Arsenic to Delaware soils include Poultry Litter (PL), historic use of pesticides, and municipal sewage sludge’s used as soil amendments. Industrial sources of Arsenic include wastes from tanneries, wood treating facilities and coal combustion for electric power. Long-term inputs from human activities (e.g., inorganic and organic arsenical pesticides, defoliants, wood preservatives, manures, and biosolids) to agricultural fields have increased total Arsenic levels up to as high as 165 ppm (mg/Kg) in soil. The Delmarva Peninsula is one of the most concentrated poultry production areas in the US. Poultry litter is generally applied on agricultural lands. Total Arsenic concentrations in PL vary. Limited data have shown ground water from agricultural fields of the Pocomoke River Basin in Maryland and Delaware having total dissolved Arsenic concentrations as high as 23 ppb (µg/ L). A majority of the Delaware soils are highly susceptible to Arsenic leaching to ground waters due to their sandy texture, low organic matter, clay, and metal oxide contents.

Purpose: The purpose of the study is to characterize the amount, chemical forms, speciation, and solubility of Arsenic in agricultural and industrial soils and forested soils in Delaware (as natural background).

Objectives: The four objectives of the study are as follows:

1) To characterize the type of Arsenic (speciation) and its distribution in Delaware soils, as impacted by long term applications of all potential Arsenic sources, such as poultry litter and biosolids.

2) To determine the controlling factors for the retention, release and potential mobility to groundwater of Arsenic in Delaware soils.

3) To quantify the potential for Arsenic leaching from Delaware soils, the Arsenic type leaching to groundwater and the potential for best management practices to mitigate Arsenic leaching to groundwater.

4) To study Arsenic in soils and determine the associations and distributions of Arsenic and other co-contaminating metals in soil contaminated by tannery.

Study Methods:

Objective 1: The type and amounts of Arsenic will be characterized in the profiles of (i) benchmark soils of Delaware, as affected by past land use; and (ii) soils contaminated from industrial activities, particularly tannery wastes. Agricultural soils to be sampled will include those that can be documented to have: (i) received regular application of PL for 10-20 years; (ii) received no PL, or any other organic by-product for 10-20 years (e.g., commercial fertilizer use only); (iii) soils that have been regularly amended with municipal biosolids (and no PL) for 10-20 years. At each site soil samples will also be collected from nearby forested areas mapped as the same soil series; these soils will represent background concentrations of soil Arsenic. Soils will be characterized for properties relevant to Arsenic retention and release.

Objective 2: is to determine the Arsenic retention capacity of Delaware soils. Adsorption of arsenate [As (V)] by selected surface and subsoil horizons will be monitored as a function of time and at pH range of these soils. The As (V) concentration (as Na2HASO4.7H20) reacted with the soils will be based on typical application rates of As (V) in PL-amended soils, and determined from isotherm studies where equilibrium As (V) in solution will be assessed versus As (V) adsorption. Arsenate will be studied since the recent research has shown that the solid state speciation of Arsenic in aged PL is largely As (V). Desorption studies (in conjunction with other parameters) will be performed. This information will provide an index to the mobility of Arsenic in soils and thus its potential to leach to groundwater.

Objective 3: is to study the mobility of Arsenic in soils, and thus the potential for Arsenic contamination of ground waters by leaching. This will be completed in laboratory studies using large, undisturbed soil columns collected from selected soils. These analyses will provide a mass balance for Arsenic in the soils.

Objective 4: is to characterize the soils for total and water soluble Arsenic. This data will provide an index of the mobility and bioavailability of Arsenic present in the tannery contaminated soils and the impact of long-term residence time effects on Arsenic fate in the environment. These studies will provide information on the form in which Arsenic is present in the soils.

Timeline:

This study is scheduled for completion by January 2006.

Figure 1: ARSENIC CONCENTRATION MAP 1

(Contact either DNREC or the Registrar’s office for a copy of the Map)

³ Newman, Cathy, “The Poison Paradox: Too Much can Kill, A Little Can Cure,” National Geographic, May 2005 page 8.