EPA Findings

Children’s Blood Lead Investigation

Historically, inorganic lead has been released to the environment by many human activities such as mining, smelting, use of leaded gasoline, and manufacturing of batteries, plastics, and chemicals. Lead is not volatile, so it usually moves through the air as fine dust which deposits and contaminates soil within a few miles of its source. People can be exposed to lead in air, food, drinking water (and beverages), soil and dust, and across the placenta before birth.

Important toxic effects of lead include anemia, hypertension, and damage to the kidneys, testicles, and nervous system. Small children are most sensitive to toxic effects of lead because they suffer significant losses in motor skills and cognitive ability at lead doses which do not affect adults. EPA considers children with blood lead levels of 10 or more micrograms of lead per deciliter of blood to be at risk of irreversible damage to the nervous system.

Chester officials provided records of over 10,000 blood lead measurements for children, which EPA entered into a computer database. Age and gender were not reported (although all were reported to be seven years or younger at the time of the test), nor was information available about how the children were chosen for blood lead sampling. Lead concentration data for air, tap water, soil, dust, and food were not available. This limited database allowed EPA to compare blood lead levels in Chester with those in similar Eastern cities but did not support conclusions about sources of lead exposure.

Average blood lead levels in Chester between 1989 and 1993 (Figure 4-16) were higher than 1990 averages in Boston, Baltimore, or Cincinnati. However, blood lead in Chester decreased significantly during this five-year period, so that in 1992 and 1993 Chester blood lead levels were similar to those in Baltimore. With the limited database it was not possible to tell if the decline in blood lead was real or artificial (caused by sampling different groups of children or by medically treating children with high blood lead levels).

EPA compared the Chester blood lead observations with predictions from a computer model that predicts blood lead. Because lead levels in Chester’s air, water, soil, and food were not available, EPA used national averages to make the predictions. To match the Chester blood lead data it was necessary to add 130 micrograms of lead intake per day to the national averages.

EPA determined the average blood lead level for each residence by combining multiple measurements from the same child and from siblings. A map of blood lead levels in Chester was prepared. The map showed no noticeable patterns of blood lead; there appears to be no part of Chester where blood lead is higher or lower than the others.

Overall, EPA’s analysis of blood lead suggests that:

1. Recent measurements of Chester children blood lead levels are similar to those in similar Eastern U.S cities.

2. Children in Chester receive lead exposures which are substantially higher than the U.S. average.

3. It is not possible with the limited data available to tell the source of the children’s excess lead exposure.

4. The problem of high blood lead appears to be city-wide rather than confined to specific neighborhoods.


Modeled Air Concentrations

As was previously noted, no new data was gathered for this study. The recent years air data that existed was often developed for specific purposes, e.g. compliance monitoring of permitted emission parameters, or was presented in format which was not compatible for risk calculation purposes. This presented a pattern of data gaps in an important medium of concern, air.

It was decided that sufficient information existed regarding the industry types, geographical locations, and production capabilities, and that meteorologic data combined with actual or generic emission levels could be utilized in a computer modeled simulation of speciated ambient air quality.

Estimated air concentrations for 699 chemicals were provided for approximately 1400 locations in Chester City. Of the pollutants assessed, 640 are gaseous in nature, while 59 exist as particulate matter2.

Although emission contributions from many sources were modeled, only the total concentration of each pollutant at each location was considered in risk calculations. Of the 699 chemicals evaluated, 122 have toxicity values in the form of reference dose (RfDs) or cancer slope factors (CSFs). Five of the modeled chemicals are criteria pollutants and are regulated under the authority of the Clean Air Act via the National Ambient Air Quality Standards (NAAQS).

For chemicals with reference doses (RfDs) or cancer slope factors (CSFs), modeling results were screened using RBCs as described above to identify chemicals of potential concern (COPCs). Accordingly, inhalation under a standard residential exposure scenario was considered. In instances where both an RfD and a CSF exist for a given COPC, only the most sensitive endpoint (cancer or non-cancer) was evaluated.

Estimated criteria pollutant concentrations were compared to the NAAQS. (This approach for evaluating potential threats is similar to the methodology employed for assessing non-cancer threats posed by chemicals with RfDs.)

For gasoline and diesel, carcinogenic risks were assessed based upon respective unit risks for these compounds, as determined by a recent USEPA investigation (USEPA, 1993c).

For the criteria pollutants, predicted concentrations at each grid location were compared to NAAQSs.

Individual Risks

At various locations in Chester, several chemicals were predicted to exist in air at concentrations of potential concern. Chromium VI was determined to contribute the most to carcinogenic3 risk at any given location, while hydrogen chloride presents the greatest non-cancer threat. A summary of the highest individual risks in Chester City is presented in Table 4-32 for carcinogenic COPCs, and in Table 4-33 for COPCs with non-cancer endpoints.

None of the predicted concentrations of criteria pollutants in Chester exceeded NAAQSs, as illustrated in Table 4-34.

Cumulative Risks

Cumulative carcinogenic risks and non-cancer threats are predicted to exceed levels considered safe at several locations in Chester City. The range of aggregate carcinogenic risks in Chester as a result of inhalation is estimated to be 1.1E-5 to 6.6E-54. For non-cancer endpoints, the range of Hazard indices (HI) is predicted to be 1.0 to 3.8. The risks are also displayed on Figures 4-29, 4-30, 4-31, 4-32, 4-33, and 4-34.

Cumulative values for the criteria pollutants were estimated to range from 0.6 to 1.6. This is illustrated on Fig. 4-35.

It is possible to discuss the culpability of various sources of air pollution to these risks. As outlined in the section on air quality modeling, a large number of sources was modeled, the sources vary dramatically in their contribution to both carcinogenic risk and noncarcinogenic hazards.

Point sources accounted for roughly 40 percent of environmental carcinogenic risk in Chester and more than half of the sub-chronic risk. Delcora and Sun each contribute roughly one quarter of the long-term cancer risk. Delcora and P.Q. Inc. emit chromium and arsenic, Delcora emits those and other heavy metals, and Sun emits many organic species. DuPont and Westinghouse account for approximately 80 percent of the non- cancer risk.

Area Source Emissions

County-wide estimated emissions were available for area sources of air contaminants. These data were not conducive to the performance of a quantitative risk assessment because of the difficulty in identifying individual chemicals and separating the Chester area out from the county. However, a qualitative/semi- quantitative assessment follows.

Sources of toxic air releases which are small when evaluated individually but are significant when combined with other facilities of similar type in a given geographic area are termed area sources. Volatile organic compounds (VOCs) are of particular concern because some are classified by USEPA as probable or possible human carcinogens. Also, they photochemically combine with oxides of nitrogen (N0x) and carbon monoxide (C0) in the presence of sunlight to form ozone, which causes respiratory problems and plant damage.

Information about area sources comes from two sources of data. Information about the location, industry type, and number of employees is available through Dun and Bradstreet. Information about the amount of VOCs released per employee per year is available in USEPA, l991d. Combining these two databases gives an estimate of VOC emissions per facility per year.

A list of facilities with Standard Industrial Classification (SIC) codes between 4000 and 9999 (which include businesses such as transportation services, gasoline service stations, automobile repair shops, and dry cleaners), and within the study area was retrieved from the Dun and Bradstreet (D&B) data base. [Facilities with SIC codes between 2000 and 3999 (manufacturing) are reported in the TRI data base and are evaluated in the Air Toxics Modeling portion of the study].

A grid system was established for the study area, with each grid square approximately one square kilometer (or about 1/2 mile by 1/2 mile), and the sum of the estimated emissions for each facility within a given grid square was calculated. The values for the grid system were assigned colors from red to green, with grey indicating no facilities.

Fig. 4-36 shows the estimated emissions for all the grid squares in the study area. Fig. 4-37 highlights the top 9 (15%) grid squares, which represent estimated annual releases of VOCs of over 40,000 pounds. Fig. 4-38 shows the minority distribution of the study area with the 9 high squares indicated in cross- hatching. This indicates that grid squares 6, 7, and 8 are in an area with a very high percentage of minority population, indicating that the potential for impact to the minority community is greatest in these areas.

There are several limitations to the approach used to estimate the VOC emissions for the area sources. First, the D&B data base does not contain every facility in the study area that releases VOCs. In addition, the estimates of VOC releases are based on studies of “typical” facilities and are not actual measures of the releases from the facilities in the study area. The actual type and amount of VOC releases is not available. The estimates are not identified for the specific SIC codes that were identified in the D&B database, so that approximate values were used instead of SIC code-specific ones.


A study of the existing public health status of the community and a specific epidemiological study to try to establish cause-and-effect links between environmental risks and health effects were beyond the scope of the environmental risk project. However, the state health department, as a preliminary exercise, looked at the mortality rate for certain diseases in the city as compared to the state and county. This exercise may be found in Appendix III. This may give useful information regarding the existing health of the community, although it cannot be used to establish causes of the health conditions.

Surface Water, Sediment, Fish Tissue

Three main data sources were used for surface water, sediment, and fish tissue data: the STORET database, CERCLIS files, and the National Study of Chemical Residues in Fish.

The CERCLIS database was described previously. Five CERCLIS sites in the Chester study area had surface water and/or sediment data. These sites underwent data quality review in accordance with the Quality Assurance Plans under which the work was authorized.

The National Study of Chemical Residues in Fish was performed by USEPA to study fish tissue contamination nationwide (USEPA, 1992b). This study began as an outgrowth of the National Dioxin Study, which found notable concentrations of dioxins in fish tissue. It involved the collection of fish tissue from over 300 stations nationwide.

One station from this study was located within the Chester study area, and these fish tissue results were used for the Chester risk assessment. Analytical data were obtained in accordance with the analytical procedures and quality assurance plans cited in the national study.

Table 4-23 presents the risks associated with direct contact with surface water at each location. It can be seen that the Hazard Indices for each location are less than l, indicating that significant adverse non-cancer health effects due to contact with surface water at the reported concentrations are not expected. Estimated cancer risks are at or below lE-6 for all locations except the Delaware County Incinerator Landfill #l (3.9E-5). The cancer risk at this site was based on arsenic and beryllium in a drainage ditch water sample taken adjacent to the landfills. The water sample was reported as “greenish brown” and is likely to have contained high amounts of suspended solids. The feasibility of people actually swimming in a drainage ditch depends upon its depth and width, seasons of flow, and may also depend upon its aesthetic appeal.

Table 4-24 presents the risks associated with direct contact with sediment at each location. It can be seen that the Hazard Indices for each location are less than l, indicating that significant adverse non-cancer health effects due to contact with sediment at the reported concentrations are not expected. Estimated cancer risks were all below lE-5.

It is likely that most of the general population of Chester does not consume locally-caught fish. However, subpopulations may exist consisting of occasional fishers or possibly even subsistence fishers. Subsistence fishers could have risks higher than those quantitated herein.

Drinking Water

This study investigated the drinking water quality of both private and public well users in the City of Chester and surrounding municipalities including Marcus Hook Borough, Trainer Borough, Chester City, Chester Township, Linwood, Upland Borough and Eddystone Borough. The potability of the groundwater in the study area and potential risk to private well users was evaluated by qualitative assessment of the existing monitoring well data from Comprehensive Environmental Response, Compensation, and Liabilities Information System (CERCLIS) and Resource Conservation and Recovery Act (RCRA) sites. Environmental equity issues that would require further study were identified where appropriate with respect to the data obtained to date.

Private Well Investiqation

The U.S.Department of Census data obtained in 1990 involved a random door-to-door survey of the housing units (both vacant and occupied) in the study area (see Table 4-1). An assessment of the data indicated that less than 1% of the housing units in the study area may obtain their drinking water source from private wells. The Chester Water Authority and Health Departments are not aware of any residential properties using local groundwater for drinking or bathing purposes. The local health department indicated that the entire population of Chester is connected to a public water supply (PWS). However, the health department did acknowledge that verification that none existed would be quite difficult. Based on U.S. Census data there are an estimated 61 private wells in the study area, of which approximately 31 are believed to be dug wells and approximately 30 are believed to be drilled wells. The data are extrapolations, from a smaller sample size, of the actual figures that would have been obtained from a complete count (USDOC, 1990). Therefore, the exact number of private wells in the study area is largely unknown.

Efforts to obtain locational information for any of the 61 private wells identified on the census tract (Figure 4-2) have been hampered primarily because of those regulations which protect census participants individual rights to privacy. It should be noted that information retrieval from the census tract is limited to a scale of census blocks which are a geographic area of about 200 people.

Public Water Supply

Drinking water quality from public water sources in the study area was investigated because greater than 99% of the population is expected to obtain their drinking water from a public supply. The study area is served by the Chester Water Authority except for Eddystone, which is served by the Philadelphia Suburban Water Company. It should be noted that Philadelphia Suburban Water Company purchases water for Eddystone from the Chester Water Authority. This water undergoes no additional treatment; therefore, the actual source of drinking water for Eddystone is the Chester Water Authority.

Tables 4-3, 4-4, and 4-5 summarize risks for the 1-year and 30-year exposure scenarios for the PWSs.


The TRI database contains information about chemical releases from industrial manufacturers and processors (primary Standard Industrial Classification (SIC) codes 20-39) to the environment. Since 1987, facilities meeting established thresholds have been required to report release data according to section 313 of the Emergency Planning and Community Right-to-Know Act of 1986 (EPCRA).

Region III has developed a method for evaluating these releases in terms of their relative toxicity. This method is documented in the Chemical Indexing System for the Toxic Chemical Release Inventory Part I: Chronic Index (USEPA, 1993d). The Chemical Indexing analysis provided in the present report displays the 1992 TRI data in terms of the Chronic Index (toxicity-weighted releases) and Residual Mass (non-weighted releases) for Region III, highlighting TRI facilities in Delaware County, Pennsylvania.

The Regional maps (Figures 4-26, 4-27, and 4-28) show TRI releases in terms of the Chronic Index, including non- carcinogenic and/or carcinogenic index dose. Those releases which do not have an associated toxicity factor are combined according to the amount of the release and are termed Residual Mass. The resultant Chronic Indices and Residual Mass values are summed for each facility and for each 8 x 8 mile geographic grid area in Region III. Combining the facility Chronic Indices within a geographic grid gives an indication of the potential for cumulative hazard from TRI facilities within a given geographic area.

In Delaware County, 28 facilities were subject to TRI reporting under EPCRA for the reporting year (RY) 1992. A summarized priority listing of these facilities is included in Table 4-27 and a complete listing is provided in Tables 4-28 and 4-29. Table 4-27 shows a quantitative summary of the facilities which ranked in the top 90th percentile – 95% confidence of the 28 facilities subject to reporting under EPCRA. Table 4-27 shows the top six TRI facilities in the Chronic Index and Residual Mass ranking.

It has not been determined whether these releases were continuous for the entire year or if they reflect one-time accidental releases or spills. In addition, the proximity of these releases relative to potentially exposed populations has not been established. The determination of a potential health threat of the volumes released depends on the proximity of the stack to residential areas, the surrounding terrain and the meteorological conditions. Furthermore, should it be determined that additional analysis is required at any site listed in this report, documentation which identifies these release as continuous or intermittent should be obtained prior to the analysis.


One of the study objectives was to be responsive to environmental concerns raised by the citizens in the study area. Some of these were issues for which USEPA had no available database and could therefore not assess with quantitative risk assessment. These issues included odors and noise and are addressed below.


Odor is a very difficult sensory phenomenon to describe objectively. Many attempts and subsequently many descriptors have been utilized in trying to describe the human olfactory system and especially its variability, thresholds and the time duration aspect of the sensation.

It is key to understand that many odors may be perceived at concentrations as low as 1 part per billion (e.g. ammonia ethylacrylate, isopropylmercaptan), while still others can be detected as low as 1 part per trillion (e.g. n-butyric acid). The mere ability to sense an odor does-not necessarily mean that it is harmful at threshold levels. On the other hand, some chemicals which are potentially harmful at low concentrations may not be perceived by most humans at levels which are significantly harmful. This certainly exacerbates individual fears and adds to stress associated with the perceived odors which people encounter.

A major source of concern in the Chester neighborhoods are the odors which seem to emanate from the large industries along the Delaware River coastline. It may be that individual small industrial or commercial operations could be sources of these emissions.

Although the incidence of odor complaints has been one of the greatest concerns in Chester, the pervasiveness of odor could not be addressed quantitatively in the environmental risk assessment. This does not diminish the importance of odors to residents, nor is it meant to ignore or screen them out of the assessment. There were virtually no data available at the onset of the study related to odors.

For purposes of this report, odors are being considered only as a source of further investigation. They are a nuisance which may add to the overall stress of residing in an urbanized environment.


Many residents of Chester have complained that environmental noise diminishes the quality of life they experience in a home setting. They cite numerous sources of the noise and have requested help from the industrial community and the environmental agencies in reducing noise to acceptable, non- intrusive levels. Some of the sources identified include:

  • truck traffic passing through residential areas
  • industrial operating equipment
  • aircraft over-flights
  • music sources, such as car radios, home hi-fi
  • train pass-by

As part of the Chester Risk Project, USEPA staff reviewed applicable environmental noise studies performed in the Chester area and performed a literature search for any applicable mitigation measures. This limited search found a Pre-Operational Noise Monitoring Study (Westinghouse, 1991) and a subsequent Noise Report Summary (Westinghouse, 1993).

In the study, environmental noise monitoring was performed at seven locations. This was considered to be background noise monitoring, at facility site locations, prior to final construction and operation of the Delaware County Resource Recovery facility. A total of three continuous 24-hour time periods were sampled including one weekend day and two weekdays. An additional four locations were sampled in the residential community in February 1991 in areas adjacent to the Resource Recovery facility.

Although there was some variability in the measured noise data due to short-duration transient events, the levels measured in and around the facility and in the residential neighborhoods are typical of urban residential settings and would be considered generally acceptable.

A noise control ordinance for the City of Chester, Pennsylvania was passed on January 14, 1993. This ordinance applies to vehicles, appliances and equipment, and includes many of the “nuisance” type of unwanted sounds. The ordinance includes subjective aspects of noise as well as objective criteria limits for motorized vehicles and property line limits depending on land use zoning. ____________________________________

2 small solid particles like dust which move with air currents  

3 cancer causing

4 1. lE-05 is a scientific notation used in risk characterization to express an excess cancer risk in the general population of 1.1 persons out of 100,000 would be expected to incur (not die from cancer but incur a cancer) a cancer above and beyond the normal incidence of cancer.

 Last modified: 11 November 1997