Perchlorate Contamination in the Arden Cordova Water Service Area
RANCHO CORDOVA, SACRAMENTO COUNTY, CALIFORNIA
In late January and early February 1997, Aerojet, as a part of their ongoing monitoring of certain off-site public drinking water wells, detected perchlorate in five off-site public drinking water wells west of Aerojet (11). To analyze these water samples, Aerojet used a refined or improved analytical method such that instead of a detection level of 400 ppb, they were able to obtain a detection limit of 35 ppb.
Of these five wells that were contaminated with perchlorate, three wells (Wells 13, 15, 16) are a part of the Cordova System (Figure 3). In February 1997, the concentrations in Cordova wells 13, 15, and 16 were reported as 220, 95, and 210 ppb perchlorate, respectively (11). Subsequent re-testing of the wells showed comparable levels. These detectable levels of perchlorate exceed the concentration (4 to 18 ppb) suggested by the USEPA provisional reference dose (1 to 5E-4 mg/kg/day) based on a 70 kg individual consuming 2 liters of water a day (12).
Southern California Water Company immediately shut the three Cordova System wells off after being notified by Aerojet on February 11, 1997 of the perchlorate levels (8). Because the Cordova Water Service Area wells are interconnected, water from other wells was used to supply those who had previously gotten water from wells 13, 15, and 16.
In March 1997, the Sacramento District field staff of the CDHS Division of Drinking Water (DDW) began sampling public water supply wells in the area of the known perchlorate contaminated wells. The well samples are processed by the CDHS's Radiation and Sanitation Laboratory with a detection limit of 4 ppb. In March, DDW staff sampled 41 public supply wells, including 18 Cordova System wells (13). On April 8, 1997, DDW staff notified the Southern California Water Company of the perchlorate levels found in wells 11 (4.4 ppb), 14 (4.4 ppb), and 19 (6.8 ppb). Since these wells exceeded 4 ppb, the low end of the acceptable range based on USEPA's provisional reference dose range, Southern California Water Company discontinued using these wells on April 8, 1997 (8).
In May, DDW headquarter staff reviewed the USEPA's provisional reference dose range and decided to adopt the drinking water concentration estimated from the upper end of USEPA's acceptable provisional reference dose range as the concentration (18 ppb perchlorate) at which the water purveyor would have to notify their water customers if they were to deliver the water to them (see Attachment A). After an independent review of perchlorate toxicity when using the use of the upper limit of the acceptable range (18 ppb), Arden Cordova placed wells 11, 14, and 19 back into service on June 19, 1997 (8). All three wells have levels of perchlorate in the well water between 4 and 12 ppb perchlorate.
The DDW field staff have continued to play the lead role in monitoring the perchlorate contamination in drinking water sources. In April, DDW staff sampled 22 wells, including 8 Cordova System wells (13). In May, DDW staff sampled 43 locations, including 18 Cordova System wells and the Cordova Systems raw influent from the Folsom South Canal (13). In June, DDW staff sampled 47 locations, including 18 Cordova System wells and the untreated influent from the Folsom South Canal (13). In July, DDW staff analyzed water from 40 locations, including 16 wells Cordova System wells and the raw influent from the Folsom South Canal (14). In August, DDW staff analyzed water from 42 locations, including 19 wells Cordova System wells and the raw influent from the Folsom South Canal (15). The Folsom South Canal influent does not have detectable levels of perchlorate (data not shown). No other Cordova System wells have been identified that have perchlorate levels exceeding 18 ppb perchlorate (Table 1). As stated previously, three Arden Cordova wells, which have levels between 4 and 18 ppb, are again in use. Perchlorate has been detected but not quantified (means that it is less than 4 ppb), in Cordova System wells #1, 3, 4, 6, 7, and 21.
In March 1997, soon after they became aware of the potential perchlorate contamination, Southern California Water Company sent a letter to all of their Cordova System customers notifying them of the problem and announcing a public meeting on March 18, 1997 at which they would disseminate more information about the contamination (see Attachment B).
At the meeting, a panel of experts, invited by Southern California Water Company, presented and responded to the origin of the perchlorate contamination, perchlorate toxicity, and water quality and service issues. Approximately 100 people attended the meeting on March 18, 1997. The meeting was well covered by the written and television press. The audience had a number of questions and statements concerning water quality, health concerns, water supply, and what was being done to make Aerojet fix the problem. At one point, a woman who has a thyroid problem asked those people in the audience to raise their hand if they had a thyroid problem, and it seemed that a significant portion of the audience responded.
Aerojet sent letters to everyone that attended the March 18th meeting and to people on their mailing list in which they invited interested persons to attend a public meeting on April 17, 1997. The focus of the April meeting organized by Aerojet was thyroid function and perchlorate toxicity. At this meeting, CDHS cooperative agreement staff responded to requests for health studies raised by the community by announcing that we were pursuing a review of available health statistics. After the meeting, CDHS cooperative agreement staff were approached by several concerned people, including a medical director of a company with a large number of employees working within the Cordova Water Service Area. The medical director was interested in getting as much information about perchlorate toxicity as possible and requested a fact sheet about perchlorate toxicity that could be shared with the employees.
The Central Valley Regional Water Quality Control Board (RWQCB) conducted a public workshop on the perchlorate issue on April 22, 1997. The RWQCB staff presented to the board an overview of the perchlorate contamination emanating from Aerojet General Corporation, and perhaps Purity Oil and McDonnell Douglas properties. During the comment period, CDHS cooperative agreement staff spoke briefly about the known and unknowns of perchlorate toxicity. The Air Force informed the board of their commitment to fund several studies to more thoroughly investigate perchlorate toxicity. Several water purveyor representatives asked the board to consider restricting any further reinjection of perchlorate-contaminated water from Aerojet water treatment facilities. The RWQCB staff responded that stopping reinjection at this time would not affect the movement of perchlorate-contaminated groundwater in the near future; however, continued treatment of the groundwater for trichloroethylene and other volatile organic chemicals is critical to stopping their continued movement. Quite a few members of the public gave comments regarding health concerns about the perchlorate. The board encouraged everyone involved to communicate often with the public about the perchlorate issues.
In April, CDHS cooperative agreement staff prepared a draft of a fact sheet focusing on perchlorate and health issues (see Attachment C). CDHS cooperative agreement staff asked for comments on a draft fact sheet DDW staff, RWQCB staff, and all water purveyors including Southern California Water Company. CDHS made the final perchlorate fact sheet available in hard copy and electronic format to Southern California Water Company and to the medical director who was interested in receiving such information.
Since the first letter sent to each customer in the Cordova System, the Southern California Water Company has put two notices in the local newspapers (see Attachments D and E). On April 29, 1997, Southern California Water Company issued a public notice reiterating the closure of the three wells with higher amounts of perchlorate and also notifying the public that three additional wells had been taken out of service when the March DDW water sampling results had shown these three wells to be contaminated with lower levels of perchlorate. In mid July, the Southern California Water Company notified their customers through a public notice that based on DDW's reevaluation of the provisional action level for perchlorate in drinking water and raising of the level from 4 to 18 ppb, they had restored the use of the three wells that had levels between 4 and 18 ppb. In the notice, Southern California Water Company referred interested people to their web site for water quality information and the fact sheet on perchlorate and health created by CDHS cooperative agreement staff. They also made the CDHS perchlorate fact sheet available at their offices or by mail.
It is not clear when the perchlorate contamination reached the Cordova System wells, because Aerojet had previously been using an analytical method to monitor for perchlorate that was not sensitive enough to adequately assess the migration of perchlorate. In fact, until recently, Aerojet had a perchlorate reporting level to RWQCB of 400 ppb, based on the fact that the older method had a practical quantitation limit for perchlorate of 400 ppb (16). It was not until Aerojet improved upon the analytical method they had been using and were able to obtain lower detection limits, that the perchlorate contamination could be adequately addressed.
Though we do not have good monitoring information, we do know that Aerojet began reinjecting water from their treatment plants on the west boundary of the site in 1984 and 1985, which continues to this day (17). Thus, assuming that it took a couple of years for the perchlorate to move from the reinjection wells to the Cordova System wells, perchlorate has probably been a contaminant in the Cordova System wells since 1987.
The exposure to the perchlorate contamination in Cordova System wells #13, #15, and #16 ceased on February 11, 1997 when Aerojet notified Southern California Water Company of the perchlorate contamination and Southern California Water Company immediately took them off-line (8).
Three other Cordova System wells (Cordova System wells #11, #14, and #19) have had levels of perchlorate between 4 and 11 ppb measured in the well water. Southern California Water Company took these wells off-line on April 8, 1997 because the levels exceeded 4 ppb, but placed them back on-line on June 19, 1997 because the levels did not exceed 18 ppb (8). Several other wells have had detectable but not quantifiable levels (<4 ppb) of perchlorate (Table 1).
The Cordova System wells provide water to 11,650 connections, approximately 36,500 customers, mostly residents (7). However, residential and commercial users closest to the wells with the highest exposure (Cordova System wells #13, 15, or 16) are likely to have had the highest exposure (8). Similarly, exposure to lower levels of perchlorate occurred and still occurs to the residents and commercial users closest to Cordova System wells #11, 14, and 19.
For a target population to be exposed to environmental contamination, there must be a mechanism by which that contamination comes into direct contact with the target population. An exposure pathway is the description of this mechanism (18). A completed exposure pathway consists of five parts: a source of contamination, an environmental medium and transport mechanism, a point of exposure, a route of exposure, and a receptor population. For a population to be exposed to an environmental contamination, a completed exposure pathway (all five elements) must be present.
In the next few paragraphs, CDHS will describe how we evaluated the completed exposure pathway related to the perchlorate contamination of the Cordova System for three receptor populations: residential, worker, and frequent customer/visitor exposure to Cordova System well water (Table 2), assuming that these receptor populations live/work/visit a house or business that is located directly next to the perchlorate contaminated wells (Cordova System wells #11, 13, 14, 15, 16, and 19). In essence, we are examining only those scenarios in which the house or business is receiving 100% of their water from one well. As described in the background section, most Cordova System customers get water that comes from more than one source, and thus the exposure scenarios that we will evaluate may not reflect the exposure for a resident/employee/visitor who lives/works/visits a house or building that is not located right next to a perchlorate-contaminated well. We hope to more adequately address the complexity of exposure through exposure dose reconstruction (see Recommendations section), but in this health consultation we will only consider the exposure pathway scenarios described in this paragraph.
When evaluating the potential health impact from exposure to contaminated potable water, CDHS considered all routes of exposure to perchlorate in the water. The most important route of exposure is through ingestion of the water. We did not evaluate exposure from eating homegrown fruits and vegetables that were irrigated with perchlorate-contaminated water, because we were not aware of bioconcentration parameters related to perchlorate (there are investigations into this issue, see Public Health Recommendations and Actions Section). We did not evaluate inhalation exposure to perchlorate in the potable water because perchlorate is not volatile (does not become a gas).
For certain chemicals, skin contact with contaminated water can be an important route of exposure. Generally speaking, skin absorption of a chemical is based on how much that chemical likes to be in fat-like surroundings. Inorganic ions like perchlorate do not like being in fat-like surroundings and thus their uptake by the skin, a fat-like environment, are typically less than 10% and frequently less than 1%. Since the permeability characteristic for perchlorate is not known, we used the permeability characteristic of another anion, chloride (1 x 10-10 cm/sec) to evaluate skin exposure to perchlorate (19). We found that skin contact would result in an exposure dose estimate that is less than 0.0005% of the dose estimate that would be received by ingesting the water. Therefore, CDHS focused on ingestion in calculating dose estimates.
The amount of Cordova System perchlorate-contaminated water that is ingested will be determined for each exposure pathway. In this analysis of exposure through ingestion, it will be assumed that there is 100% absorption of perchlorate into the body from the gut from the amount water that is ingested.
This health consultation focuses on perchlorate exposure and thus the toxicological evaluation will focus on perchlorate. CDHS acknowledges that there low levels (below the drinking water standard) nitrates and nitrite, naturally-occurring and agriculturally-related, in the well water; however, the affect of nitrates/nitrites in combination with perchlorate will not be evaluated due to lack of toxicological information that would allow such an evaluation.
Most of the information about the toxicity of perchlorate comes from studies of potassium perchlorate as a treatment for hyperthyroidism, resulting from Graves' Disease. Perchlorate inhibits the secretion of thyroid hormones (and can thus relieve the symptoms of Graves' Disease) by competitively inhibiting the accumulation of iodide in the thyroid (20). Discontinued administration of the ammonium perchlorate to Graves' Disease patients does result in a return to their hyperthyroid condition (21). People who have been treated with perchlorate have reported gastrointestinal irritation, skin rash, and hematological effects including agranulocytosis, aplastic anemia, and lymphadenopathy (20). The severe hematological effects seem to be more likely to occur when large doses of more than 1,000 mg/day (approximately 14 mg/kg/day for a 154 pound man) are used (22).
Potassium perchlorate was extensively used for treatment of Graves' Disease patients in the late 1950s and 1960s. After the reports of the severe hematological effects, potassium perchlorate was not used for many years (23). In the early 1980s, physicians in Europe began using it again for the treatment of Graves Disease, and reporting no serious side effects occurring as long as the dose was kept below 1,000 mg/day (approximately 14mg/kg/day for a 154 pound man)(22). In addition, potassium perchlorate has also been found helpful in treating thyrotoxicosis resulting as a side effect from other drug therapies (24-28).
There are only a few studies of the short-term exposure in persons without Graves Disease (29). The animal studies that have been conducted have also involved short-term exposures and the doses were too high to see a level where there was no effect on the thyroid. Both human and animal studies have primarily examined the effects of perchlorate on the thyroid, interference with the production of thyroid hormones resulting in a below normal level of thyroid hormone in circulation (hypothyroidism). The effect of perchlorate on systems other than the thyroid needs to be explored, especially, effects on the blood system (described above) and developmental effects (described below).
Children are not little adults, their bodies are not fully developed, and may not respond to a perchlorate in the same manner as an adult. For instance, thyroid hormone is critical to normal brain and physical development, and the critical period for this dependency on thyroid hormone begins in the uterus and extends up until three years of age. After the age of 3, thyroid hormone continues to play a primary role in physical development until puberty. Thus, a low level or absence of thyroid hormone in utero or in childhood may lead to irreversible mental retardation and retarded physical growth.
Perchlorate can cross the placenta and thus could affect the developing fetus, though these effects have not been studied in humans. It is known, however, that drugs currently being used to treat Graves' Disease such as propylthiouracil do cross the placenta and can produce neonatal hypothyroidism (30, 31) and fetal in utero goiter (enlargement of the thyroid)(32-34). In fact, because the developing fetus's thyroid is immature, propylthiouracil is a more potent suppressor of thyroid function in the fetus than in the mother (35).
In a study of the effects of potassium perchlorate (740mg/kg/day for the mother) fed to pregnant guinea pigs during pregnancy, a 15-fold enlargement of thyroid of the newborns was noted, even though no increase in size of the mother's thyroids occurred (36). Thyroid hormone levels of the newborn guinea pig were not measured in this study. Another animal study in which the mother was given fairly high levels of perchlorate, also resulted in increased thyroid weight in the offspring and the mother (37). At this time, it is unclear whether lower doses of perchlorate would affect the thyroid of the developing fetus and young child and thus affect thyroid function at a time when normal thyroid hormone production is important to brain development.
There are animal studies underway which are exploring the toxicity of perchlorate, including effects on the immune system and developmental effects (see the Recommendations section at the end of the text for more information).
In 1992 and 1995, USEPA staff reviewed the perchlorate toxicology studies and derived a provisional reference dose (RfD)(12, 29). An RfD is a dose to which a person could be exposed over long-term period without having any appreciable risk of a noncancer health effect. The USEPA applied an uncertainty factor of 300 or 1000 to the No Observable Adverse Effect Level of 0.14 mg/kg/day (NOAEL)(29, 38) to derive an RfD of 1 to 5 x 10-4 mg/kg/day (12). (If one assumes that a person drinks 2 liters/day of water and weighs 70 kilograms, the reference dose range corresponds to an acceptable range of perchlorate in drinking water of 4 to 18 ppb).
The uncertainty factor of 300 or 1000 is derived from multiplying the following (12):
|*||An uncertainty factor of 10 to account for extrapolation from the acute exposure in the NOAEL study to chronic exposure of an RfD;
|*||An uncertainty factor for database deficiencies (3 or 10) to account for data limitations including limited data on subchronic and chronic exposure to low doses of perchlorate, limited data on other organ system effects, limited data on the effects on the hematopoietic system, and a lack of reproductive and multigenerational data;
|*||An uncertainty factor of 10 to protect sensitive subpopulations which would include groups such as hypothyroid patients and individuals with low iodine diets or with genetically impaired iodine accumulation.|
The only information about the possible carcinogenicity of perchlorate has to do with cancers of the follicular thyroid cells (12). Interference with the normal thyroid-pituitary feedback mechanism, such as that caused by perchlorate, can theoretically lead to thyroid follicular cell neoplasia. Several animal studies found that thyroid tumors were induced in both rats and mice by long-term administration of high doses of perchlorate. However, humans are not supposed to be as sensitive as the rat to thyroid cancer (39, 40). Since perchlorate's possible carcinogenic effects on the thyroid are based on the same mechanism (interfering with the thyroid-pituitary homeostasis) that determines its noncarcinogenic effects, it may be appropriate to consider the RfD as a dose which does not pose a significant risk of thyroid cancer (29).
It is even harder to determine whether or not perchlorate exposure can cause any other type of cancer. If a link is discovered, it will probably be based on perchlorate acting not as a mutagen (causing genetic changes) but rather as a growth promoter, an effect associated with a threshold. In other words, below a certain threshold, perchlorate would not have cancer-causing effects. More toxicological information is needed to ascertain whether perchlorate can cause cancer and if it can, at what dose this effect may start occurring.
Using USEPA's provisional reference dose (0.0001 to 0.0005 mg/kg/day) based on perchlorate's effect on the thyroid, CDHS evaluated the noncancer (thyroid) health impact of the completed exposure pathway, drinking water containing perchlorate from Cordova System wells, for three receptor populations: adult resident, worker, and frequent adult customer/visitor (Table 2). For wells no longer in service, the last concentration of perchlorate measured in the well before the well was taken off-line was used. If a well is in service, the highest concentration of perchlorate measured in the well when it was on-line was used in the dose calculations.
Though it is possible to estimate a dose for a child drinking the Cordova System water, CDHS did not calculate this dose because we are not confident about how to interpret the dose estimate. To compare the estimate of a child's dose with toxicological information based on adult exposure ignores the fact that a child is not a small adult, especially when it comes to the importance of the thyroid in normal brain development (see above). Thus, until there is more information about perchlorate's effect on children, CDHS is not able to evaluate past and current exposures to a young child drinking the Cordova System water.
Residential exposure in the Cordova System: CDHS estimated the exposure for a adult resident who lives 24 hours per day, seven days a week, for 52 weeks of the year near Cordova System wells #11, 13, and 14 (Table 3 is a list of the exposure parameters used in the toxicological evaluation). Concentrations in wells #11, 13, and 14 were used for the dose calculations of the residential exposure because these wells primarily serve residences. CDHS will assume that the adult resident drinks 2 liters/day. CDHS estimated doses for a adult resident exposed to water coming from Cordova System wells #11, 13, and 14 with contaminant levels of 4.4, 220, 4.4 ppb, respectively.
The estimated dose for adult residential exposure to water from Cordova System wells #13 (0.0063 mg/kg/day) exceeds the provisional reference dose range (0.0001 to 0.0005 mg/kg/day) which means that noncancer (thyroid depression) health effects may have occurred when adult residents who lived close to Cordova System wells #13 were exposed to the perchlorate-contaminated water from these wells. However, because there is a very large uncertainty factor associated with the provisional reference dose and the estimated dose does not approach the NOAEL (0.14 mg/kg/day), it is unlikely that the adult residential exposure to well #13 did cause any noncancer health effects.
The Southern California Water Company took well #13 out of service on February 11, 1997, so there is no current or future exposure to well #13 water. Since it is thought that the effect of perchlorate on thyroid function returns to normal after exposure to perchlorate ceases, any effects that may have occurred to adult residents receiving water from well #13, should no longer be occurring.
The estimated dose for the adult resident exposed to water from well #11 (0.0001 mg/kg/day) or 14 (0.0002 mg/kg/day) does not exceed the provisional reference dose range (0.0001 to 0.0005 mg/kg/day). This means that noncancer (thyroid depression) health effects would not have occurred or should not be occurring to the adult resident drinking or washing with water from well #11 or 14. Wells #11 and 14 are currently in service.
Worker exposure in the Cordova System: CDHS will estimate the exposure for a worker who works eight hours a day, five days a week, for 50 weeks of the year (assumes a two week vacation) at a business that is served by Cordova System wells #15, 16, and 19 (Table 2 is a list of the exposure parameters used in the toxicological evaluation). Concentrations in wells #15, 16, and 19 were used for the dose calculations of the worker exposure because these wells primarily served commercial connections. CDHS will assume that the worker is involved in manual labor and thus drinks a relatively large quantity of water each day (3.7 liters (15.6 cups)(41). CDHS will estimate the dose if the worker is exposed to water coming from Cordova System wells #15, 16, and 19, with contamination levels of 95, 220, and 6.8 ppb, respectively.
The estimated dose for worker exposure to water from Cordova System wells #15 (0.0011 mg/kg/day) or #16 (0.0027 mg/kg/day) exceeds the provisional reference dose range (0.0001 to 0.0005 mg/kg/day) which means that noncancer (thyroid depression) health effects may have occurred when workers of businesses located close to Cordova System wells #15 or #16 were exposed to the perchlorate-contaminated water from these wells. However, because there is a very large uncertainty factor associated with the provisional reference dose and the estimated doses do not approach the NOAEL (0.14 mg/kg/day), it is unlikely that these exposures did cause any noncancer health effects.
The Southern California Water Company took wells #15 and 16 out of service on February 11, 1997, so there is no current or future exposure to well #15 or 16 water. Since it is thought that the effect of perchlorate on thyroid function returns to normal after exposure to perchlorate ceases, any effects that may have occurred to workers receiving water from well #15 or 16, should no longer be occurring.
The estimated dose for the worker exposed to water from well #19 (0.0001 mg/kg/day) does not exceed the provisional reference dose range (0.0001 to 0.0005 mg/kg/day). This means that noncancer (thyroid depression) health effects would not have occurred to the worker drinking or washing with water from well #19. Well #19 is currently in service.
Frequent customer or visitor exposure to Cordova System businesses: CDHS will estimate the exposure for an adult visitor or customer who goes once a day, five days a week, for 50 weeks of the year (assumes a two week vacation) to a business that is served by Cordova System wells #15, 16, and 19 (Table 2 is a list of the exposure parameters used in the toxicological evaluation). Concentrations in wells #15, 16, and 19 were used for the dose calculations of the frequent adult customer/visitor exposure because these wells primarily serve commercial connections. CDHS will assume that the adult customer/visitor drinks one cup of water (0.24 liters) per trip to the business. CDHS will estimate the dose if the frequent adult customer/visitor is exposed to water coming from Cordova System wells #15, 16, and 19, with contamination levels of 95, 220, and 6.8 ppb, respectively.
The estimated dose for the frequent adult customer/visitor exposed to water from well #15 (0.0002 mg/kg/day), #16 (0.0005 mg/kg/day), or #19 (0.000016 mg/kg/day) does not exceed the provisional reference dose range (0.0001 to 0.0005 mg/kg/day). This means that noncancer (thyroid depression) health effects would not have occurred to the frequent adult customer/visitor drinking or washing with water from well #15, 16, and 19. Wells #15 and 16 have been taken out of service. Additionally, drinking the water from Cordova System well #19, which is being used, should not cause noncancer health effects to occur.
The estimated dose for the frequent adult customer/visitor exposed to water from well #19 (0.000016 mg/kg/day) does not exceed the provisional reference dose range (0.0001 to 0.0005 mg/kg/day). Thus, noncancer (thyroid depression) health effects would not have occurred or should not be occurring to the adult customer/visitor drinking or washing with water from well #19.