Polychlorinated Dioxins and Furans: Sources, Emissions, and Levels

 

Dr. Bruce D. Rodan
Dr. David H. Cleverly

 

Slides 1 & 2: Introduction

Thank you for allowing us the opportunity to present an outline of work performed by the United States Environmental Protection Agency (EPA) on sources, emissions and levels of polychlorinated dioxins and furans. We trust that this presentation will serve as an introduction to the scientific information and terminology used in the risk assessment of dioxins and dioxin-like chemicals. The text of the presentation is formatted to be read in conjunction with the accompanying slides, which are indicated by sequence number and title. The presentation covers the chemical structure of these chemicals; the concept of toxicity equivalence; emission sources in the United States; environmental levels and human exposures in the United States; plus a brief note on dioxin toxicity. Should additional information be desired on any of these topics, the EPA internet site (http://www.epa.gov/) offers a valuable resource where work products from the dioxin reassessment and other United States regulatory activities are posted for public dissemination.

 

The structure of dioxins, furans, and PCBs warrants special consideration, as it is the specific shape of the molecules and attached chlorine atoms that governs their toxicity. Dioxin is the abbreviated, common, name for the 75 chemicals comprising the family of polychlorinated dibenzopara-dioxins (PCDDs). The most toxic of the dioxin molecules, 2,3,7,8-tetrachlorodibenzo-p-dioxin, is depicted in the upper left of slide 3. All dioxin molecules have in common two benzene rings joined by two oxygen molecules (di-oxin). Varying numbers of chlorine atoms (up to eight) can be located at different positions around this structure, resulting in 75 different structural configurations known as congeners. Note the flat, planar structure of this molecule and the symmetrical location of the chlorine atoms at the far ends. Seven of the PCDD congeners are considered to exhibit 'dioxin-like toxicity', resulting from the structural configuration of the chlorine atoms.

 

Furans are depicted in the upper right corner of slide 3, the specific example being 2,3,7,8-tetrachlorodibenzofuran. Polychlorinated dibenzofurans (PCDFs) differ structurally from PCDDs only by a carbon-carbon bond substituting for one of the oxygen bonds. Differing possible arrangements of chlorine atoms around the dibenzofuran molecule give rise to 135 congeners, of which 10 are considered to exhibit toxic, 'dioxin-like', properties.

 

Polychlorinated biphenyls (PCBs) are depicted at the bottom of slide 3. A critical difference between PCBs and the PCDD/PCDFs is the ability of the PCB to rotate around the single bond connecting the benzene rings. This rotation occurs principally when there are one or more chlorine atoms located near the connecting bond, in what is known as the 'ortho' position. Rotation within the PCB molecule causes a loss of the flat, planar, shape, thereby reducing its potential to exhibit 'dioxin-like' properties. Of the 209 PCB congeners, 13 are thought to exhibit 'dioxin-like' toxicity. Other PCB congeners also exhibit toxicity, although their mechanism of action is different from the dioxin-like congeners.

 

You will have noted from the above discussion that there are a total of thirty PCDDs, PCDFs, and PCBs that are currently considered to exhibit dioxin-like toxicity. This raises a problem for toxicity assessments where measurements detect various levels of the different PCDD/PCDF/PCB congeners, each of which has a different potential to elicit dioxin-like effects. Rather than perform thirty individual assessments, scientists have developed the concept of toxicity equivalence to sum the effects of dioxin-like chemicals. Each congener is given a toxicity equivalence factor (TEF) based on its specific ability to elicit dioxin-like effects. The congener 2,3,7,8-tetrachlorodibenzo-p-dioxin (slide 3, upper left) is the most toxic congener and is given a TEF of one. Other congeners are given TEFs that are fractions of one. The total toxic equivalent quantity (TEQ) is the sum of all the individual PCDD/PCDF/PCB concentrations multiplied by their specific toxicity equivalence factors (TEF).

 

A word of caution is necessary here, as there are a number of different units for reporting dioxin measurements. It is important to be aware of the specific unit in use when comparing studies and between exposures and toxic effects. Measurement units can include: 

-- levels of individual congeners, particularly 2,3,7,8-TCDD;
-- TEQs based on dioxins only;
-- TEQs based on dioxins and furans;
-- TEQs based on dioxins, furans and dioxin-like PCBs;
-- total dioxins, not adjusted to TEQs; and
-- calculations based on different TEF values.

This table provides the list of the current TEF values in use by the U.S. EPA for human health assessments.

 

This figure graphs the draft inventory of dioxin releases from known sources in the United States in 1987. The inventory was developed using a matrix of measured dioxin/furan TEQ emissions for each source category, multiplied the estimated national activity/emissions from that source category. The high, central and low range estimates reflect a judgment by the researchers of their confidence in the accuracy of the estimate. Approximately 16.2 kg of dioxin TEQ were released into the environment from known United States sources in 1987. The two principal sources were municipal solid waste incineration and medical waste incineration. These two sources accounted for a large majority of known United States dioxin TEQ emissions. Pulp and paper mills, secondary smelters, and hazardous waste facilities, etc., contributed much lower overall levels of dioxin TEQ to the environment.

 

Please note that this graph provides overall estimates of dioxin TEQ releases into the United States environment. The estimates do not reflect:

 

-- unconfirmed dioxin emission sources where U.S. data is currently unavailable, such as metal sintering;
-- dioxin reservoirs where the potential for human exposure has not been quantified, such as from dioxin contamination of pentachlorophenol treated utility poles;
-- local source effects, such as levels close to industrial facilities; or,
-- sources that may lead to high individual human exposures, such as previously seen with chlorophenol manufacture, use (Agent Orange) and disposal (Times Beach contamination episode).

This is a continuation of the previous graph of the 1987 United States dioxin emission inventory. Please note the 250-fold scaling reduction of the y-axis, indicating that these are very minor sources of dioxin emissions on a national scale.

 

These graphs represents the draft update of the United States dioxin emission inventory for 1995. The total United States dioxin TEQ emission estimate of 2.5 kg represents a marked reduction from the 16.2 kg reported in the 1987 inventory. This reduction has occurred predominantly due to dioxin emission reductions from municipal solid waste and medical waste incinerators.

 

Ozette Lake lies on the Olympic Peninsula of Washington State in the northwest United States, remote from any local sources of dioxin TEQ emissions. The dioxin and furan levels measured in Ozette Lake sediment cores represent long range contamination from sources in the United States and elsewhere. This graph records the deposition rates of dioxin and furans over time from approximately 1720 until the late 1970s based on these sediment cores. Dioxin and furan deposition rates began to increase in the 1930s, peaked in the 1950s, and then began declining.

 

The pathways from dioxin emission source to human consumption are illustrated in this figure. Dioxins emitted from combustion and industrial sources, or re-entrained from environmental reservoirs, are transported to distant locations through atmospheric or aquatic pathways. The dioxins are deposited on agricultural crops, taken up in the food supply, and then bioaccumulated and biomagnified through the food chain. This is the predominant pathway for human exposure, excluding isolated exposure episodes resulting from industrial or waste disposal accidents.

 

Average dioxin TEQ levels in United States media are displayed in this table. The term ppt is the abbreviation for parts per trillion, ppq for parts per quadrillion, and pg/m3 for picograms (1 trillionth of a gram; 10-12g) per cubic meter. Rural environments have substantially lower dioxin and furan levels than urban environments.

 

Average dioxin TEQ levels in the United States food supply are displayed in this table. The designation 'ND' in the title section of this slide refers to non-detect, and 'DL' abbreviates detection limit. Because dioxins are found at very low levels, it can be difficult for a laboratory to determine if a measurement is real or an artifact of the instrumentation. The level at which this uncertainty occurs is called the detection limit. Measured results above the detection limit are considered valid. Measurement results below the detection limit cannot be reliably distinguished from a zero measurement, and are termed non-detects. A non-detect is not the same as a zero value, as some dioxin could be present yet not quantifiable.

 

The treatment of non-detects can substantially impact final estimates of the dioxin level. Non-detects are treated in two ways in this table. In column three, non-detects are assigned a value of half the detection limit. For example, if the detection limit was 1 part per trillion, the associated non-detect result would be assigned the value 0.5 parts per trillion. In column four, non-detects are assigned a zero value. Thus, column four reports lower results than column three due to different ways of assigning values to the non-detect results.

 

The average serum levels of dioxin-like substances in United States citizens are presented in slide 14. Please note the different ways that dioxin levels are reported here, either as parts per trillion (ppt) of the congener 2,3,7,8-TCDD, ppt of the PCDD/PCDF toxic equivalent quantity, or by calculating a TEQ value for the combination of PCDDs, PCDFs and co-planar PCBs. It is useful to keep these average human body burden values and exposures in mind as they provide a means of comparing dioxin levels between peoples in different regions and different occupational and environmental circumstances. Note, too, that these are average measurements and that doses can be substantially higher for certain sub-populations, particularly breast feeding infants and the families of subsistence fishermen.

The principal route of exposure to dioxin-like chemicals for the general population is through ingestion of food. As previously illustrated in slide 11, dioxins are bioaccumulated through the food chain and ultimately ingested by humans. It is estimated for the United States that approximately 96% of dioxin intake is through food ingestion, totaling approximately 60 picograms TEQ/day or 110 picograms TEQ with co-planar PCBs added. A considerably smaller dose is inhaled or obtained through soil or water ingestion.

 

As noted, certain human populations are more highly exposed to dioxin-like chemicals than others. These populations include: 

-- workers at chlorophenol production plants or following accidents at these facilities, such as the 1976 explosion in Seveso, Italy;
-- fishermen and their families, who may consume contaminated fish from local sources; and,
-- nursing infants, where dioxin-like substances bioconcentrate in mothers milk and are ingested by the infant during susceptible stages of growth and development.

Integral to an understanding of the toxicity of dioxin-like agents is their ability to bind to the aryl hydrocarbon (Ah) hydroxylase receptor, transport to the cell nucleus, and directly increase the transcription of dioxin responsive genes. The strength of binding by the respective PCDDs, PCDFs and co-planar PCBs to the Ah receptor correlates with their ability to induce DNA transcription. This is the foundation for the toxicity equivalence factors (TEFs; slides 4 & 5). The receptor-based mechanism and ability to disrupt cellular processes is the basis for the exquisite toxicity of dioxin-like substances to certain species.

 

Other proteins noted on this diagram are the 90 kiloDalton heat-shock protein (Hsp90), an accompanying 50 kiloDalton protein (p50), and the Ah-receptor nuclear transferase (ARNT) protein.

 

In February, 1997, the International Agency for Research on Cancer (IARC) classified 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) as a category 1, 'known' human carcinogen. The conclusion that 2,3,7,8-TCDD is likely to be carcinogenic to humans at some dose was based on: unequivocal animal bioassay evidence of carcinogenicity; limited human, epidemiological, information from occupational and other exposures; plus mechanistic plausibility.

 

The following dioxin-related effects have been observed at or near general U.S. population body burden levels: 

-- enzyme induction;
-- immune system changes;
-- development milestone changes;
-- glucose tolerance changes;
-- altered hormone levels, etc.

These effects are generally considered to be adaptive changes, but have the potential to be adverse.

 

Combustion and incineration sources account for more than 90% of known environmental dioxin releases in the United States. Other sources of exposure, such as chemical manufacture and processing, contribute smaller amounts of dioxin-like substances into the environment. Subsequent human exposure occurs through the food chain, where the ingestion of soil, meats, dairy products and fish accounts for more than 95% of dioxin intake.

 

It is important to note that the United States dioxin emission inventory is a work in progress. The contribution to the emission inventory from reservoir sources of dioxin, such as pentachlorophenol-treated utility poles, and other industrial practices, such as metals sintering, remain to be determined. Recent U.S. studies have also indicated that the open burning of household waste in barrels may produce substantial dioxin/furan emissions per kilogram of waste burned. This finding has considerable relevance to developing countries, where the open burning of municipal and other wastes is more widespread.

The 1987 and 1995 inventories support sediment core findings that dioxin emissions in the United States have been declining over recent years. The inventory findings suggest that the replacement of older technologies, equipment and practices with less polluting alternatives has contributed to this reduction. Data generated throughout this period provides a valuable record of the dioxin emissions resulting from various types and standards of industrial and incineration facilities. Many of these dioxin emission factors will be of relevance to current activities in developing countries. We anticipate that data from this record will be of assistance to other countries should they decide to proceed with national dioxin inventories.