Water
Authors
The information in this chapter was prepared by the following individuals:
Description of drinking water:
o Lawrence Baker, Ph.D., Department of Civil and Environmental Engineering, Arizona State University
Description of untreated surface and ground water:
o Mason Bolitho, Arizona Department of Water Resources
Drinking Water
Background
Human health risks associated with drinking water fall into three categories:
o Pathogens: Bacteria, viruses, and protozoa that can cause disease
o Carcinogens: Chemicals that can cause cancer
o Non-carcinogenic chemicals: Chemicals that cause health problems other than cancer.
This assessment is based upon an extensive water quality data base for 1,800 community water supplies. The data base is maintained by the DEQ and this assessment was completed with assistance from Terry Fields and Dean Landis of DEQ.
As defined by the EPA, community water supplies are those water systems that have more than 25 service connections.
Private domestic wells are not regulated or systematically monitored. Health risks associated with private wells are treated in the next section of this chapter.
Indicators of Contamination
Chemical Indicators
Because a relatively small number of chemicals are likely to be present in drinking water, the Committee developed a short list of indicator chemicals for the assessment. Criteria for an indicator chemical were:
o It must be known or strongly suspected agent of human disease.
o It should have some likelihood for appearing in water supplies, based on widespread use, solubility in water, and other factors.
o It must be monitored for a substantial fraction of the population. Normally, this would mean that it is regulated.
The following indicator chemicals were selected: Chlorinated solvents, metals, pesticides, gasoline constituents, nitrates, trihalomethanes, and radon.
Chlorinated solvents
o Trichloroethylene (TCE)
TCE was used extensively in the 1950s, 1960s, and 1970s as an industrial solvent. TCE is found in groundwater in many parts of the U.S., primarily due to disposal of spent solvents in dry wells, lagoons, and leach fields.
TCE is found in numerous locations in the Phoenix and Tucson metropolitan areas. TCE was chosen as an indicator chemical due to its widespread occurrence in groundwater and soils in Arizona.
EPA classifies TCE as a probable human carcinogen and has set the maximum contaminant level (MCL) for drinking water at 5.0 microns/L
o Perchloroethylene(PCE)
PCE was widely used as a dry cleaning agent, degreaser, and chemical intermediate. PCE is found in groundwater throughout the U.S., primarily due to disposal of solvents and through unintentional leaks and spills.
In Arizona, PCE is most commonly found in the Phoenix and Tucson metropolitan areas. However, because PCE was so widely used as a dry cleaning agent, it has also been found in the groundwater of smaller communities like Payson.
PCE was chosen as an indicator chemical due to its widespread occurrence in groundwater and soils in Arizona.
PCE is a probable human carcinogen with an MCL of 5.0 microns/L.
o Vinyl Chloride
Vinyl chloride is a chemical intermediate for manufacture of polyvinyl chloride and other resins and is a degradation product TCE and PCE. It is most commonly found in groundwater where there was disposal of these solvents and where partial degradation has occurred.
Vinyl chloride has been found in groundwater at several sites in Arizona. It is a known human carcinogen with an MCL of 2.0 microns/L.
Metals
o Lead
Human exposure to lead occurs from lead-based paint found on older houses, emissions from vehicles using leaded gasoline as fuel, and ingestion of drinking water contaminated by lead solder and fixtures.
At elevated levels, lead causes neurological impairment in children.
Because lead exposure comes from several sources (soil, air, and water), the health risk assessment for lead was developed separately and can be found in Chapter 8.
o Arsenic
Arsenic and its compounds are widely used in industry in the manufacture of pesticides, herbicides, pharmaceuticals, and other products and is a common contaminant in mine drainage.
Unlike most chemicals discussed in this report, arsenic occurs naturally at relatively high concentrations in both surface waters and groundwaters in parts of Arizona (Owens-Joyce et al., 1984; Welch et al., 1988; Baker et al., 1994).
Arsenic is a known human carcinogen. Arsenic also causes hyperpigmentation, keratosis and possible vascular complications and a disease called blackfoot disease.
The reference (safe) dose for non-carcinogenic effects is 9 microns/L. The lowest observed effect level for non-carcinogenic effects is 170 microns/L. The MCL for arsenic is 50 microns/l.
o Chromium
Chromium and its compounds are used in metal plating, photography, as a catalyst, and in other applications.
Chromium occurs naturally in groundwater in some places in Arizona but is also found as a contaminant caused by disposal activities. In groundwater, chromium occurs in two valence states, +3 and +6. Hexavalent chromium (+6) is considered the most toxic form and is considered a Class A carcinogen. Some organic chromium compounds are also considered carcinogenic.
The MCL for chromium is 0.05 mg/L.
o Mercury
Elemental mercury is a silvery, liquid metal at room temperature and is used in a variety of industrial applications. Mercury salts were formerly used as rodenticides and for other purposes.
Chronic exposure to mercury or mercury salts affects the central nervous system, causing loss of motor function, brain damage, blindness and deafness.
Because of its serious effects on humans and its presence in some wastewater discharges, mercury has been chosen as an indicator chemical.
The MCL for mercury is 2 microns/L.
Pesticides
Of the many pesticides used in Arizona, only two are likely to be found in groundwater. They are dibromochloropropane (DBCP) and ethylene dibromide (EDB).
o Dibromochloropropane (DBCP)
DBCP was widely used as a nematocide in Arizona from 1955 to 1979, when its use was suspended because of groundwater contamination. In Arizona DBCP was used primarily on citrus and cotton crops in Maricopa and Yuma counties.
As a result of its widespread use prior to 1979, large areas of groundwater are contaminated with DBCP, sometimes to depths of 500 feet. Numerous municipal drinking water wells have been closed due to DBCP contamination.
DBCP is a suspected carcinogen. The MCL for DBCP is 0.2 microns/L.
o Ethylene dibromide (EDB)
EDB was once widely used as a fungicide on agricultural crops. Its use was discontinued when it was found to be a groundwater contaminant.
Gasoline Constituents
Gasoline is comprised largely of four compounds: benzene, toluene, ethylbenzene, and xylenes. Collectively these compounds are known as "BTEX" contaminants.
Of these four contaminants, benzene has the highest mobility and the lowest MCL. Because of these, the Committee considered benzene as the best indicator chemical for gasoline contamination.
Benzene is a suspected carcinogen with an MCL of 5 microns/L.
Nitrates
Nitrate is one of the most common groundwater contaminants in the U.S.
Nitrate contamination occurs largely from nitrogen fertilization of crops, but can also occur from septic tank contamination, wastewater treatment plant effluents, and industrial contamination.
Nitrate contamination of groundwater is very prevalent in parts of Arizona. Nitrates are known to cause methemoglobinemia ("blue baby syndrome") and may be precursors to carcinogenic nitrosamines.
The MCL for nitrate is 10 mg/L.
Trihalomethanes
The main source of trihalomethanes is chlorination of drinking water containing natural organic matter. There are four main trihalomethanes: chloroform, dichlorobromoform, dibromochloroform, and bromoform.
The MCL for total trihalomethanes was 100 microns/L but has recently been lowered to 80 microns/L.
Radon
Radon is a colorless, odorless, radioactive gas that is found naturally in air, soils, and in groundwater. It is an inert gas which is a by-product of decay of other radioactive elements, such as uranium and thorium, in soils and rocks. Radon is known to cause lung cancer in humans.
Water-Borne Diseases
A variety of water-borne diseases are known to contaminate water supplies and cause illness or death.
Prior to the advent of modern water treatment involving disinfection and filtration, water-borne diseases killed thousands of people in the U.S. each year. In areas of the world that lack modern sanitation and water treatment, water-borne diseases are a major contributor to mortality.
Three types of microbes cause water-borne diseases:
o Viruses: Water-borne viral diseases include polio, hepatitis, and various gastrointestinal diseases.
o Bacteria: Diseases caused by water-borne bacteria include cholera, typhoid, and salmonella. Recently, a major cholera epidemic in South American had infected nearly 400,000 individuals (Glass et al., 1992).
o Protozoa: Water-borne protozoans that cause disease include Giardia and Cryptosporidium. Both of these cause severe diarrhea, but are generally not fatal to otherwise healthy individuals.
Major outbreaks of water-borne disease have become relatively rare in the U.S. over the past century. Reasons for this include:
o Adoption of reliable drinking water treatment including chlorination and sand filtration
o Collection and treatment of sewage
o Vaccination for some diseases like polio
o A competent regulatory bureaucracy that sets and enforces standards for public water supplies
o Development of state and national public health services that track disease outbreaks and takes actions to limit their spread.
Tests for Microbial Quality of Drinking Water
Coliform Test
Currently, the main test for microbial quality of drinking water is the coliform test. This test measures the presence of coliform bacteria. A positive coliform test does not necessarily mean that the water contains disease-causing organisms (pathogens). It means only that there is a potential that pathogens, if present, could survive.
Coliform bacteria are therefore considered an indicator of the potential presence of pathogens.
o To make this point, note that for 88% of the water-borne disease outbreaks that occurred in 1991-1992 caused by bacteria, viruses, or unknown agents, the coliform bacteria test was positive (Moore et al., 1994).
Coliform are a less reliable indicator of protozoan infections. Only 33% of the protozoan disease outbreaks during the same period of 1991-1992 were accompanied by positive coliform tests.
Nevertheless, for the vast majority of cases in which a water system "flunks" its coliform test, there is no outbreak of water-borne disease.
Several studies have shown that the presence of coliforms is correlated with gastrointestinal illness (Peterson and Hines, 1960).
Turbidity Test
A second indicator of microbial water quality is turbidity. Turbidity is one measure of the amount of suspended particles in water. It is thought to be a better measure of the potential for some protozoan diseases than are coliforms (Bellamy et al., 1993).
Results
About 95% of all Arizonans (3.96 million people) drink water from 1,800 regulated community water supplies.
All community water supplies with more than 25 service connections are required to monitor the bacterial and chemical quality of their supply water.
These data, which are maintained in a computerized data base at (ADEQ), provide the basis for a reasonable risk assessment.
Methods of data analysis
Assessment of microbial risks
There are two types of violations of the bacterial standards:
o "MCL violations," in which coliform bacteria are detected for a specific number of samples. The number of samples and the number of detections depends upon the size of the system.
o Sampling status violations in which samples were simply not collected. Three frequency-of-violation classes were developed:
No violations
Violations for one to three months
Violations for more than three months.
Assessment of Chemical Risks
Chemical risks were broken down into cancer risks and non-cancer risks.
o Cancer Risks: A reasonably well-accepted, quantitative risk assessment method is available for cancer risk assessment.
In this approach, an individual's risk of cancer due to exposure from a known or suspected carcinogen is calculated based on his or her exposure (amount of water drunk times the concentration of contaminant in the water), and a cancer potency or slope factor, normalized to body weight.
After this calculation is completed, the number of cancers within a population within a single year can be calculated by dividing the total number of lifetime cancers in the population by 70 years, an average lifetime.
Assessment of Non-Cancer Risks
As discussed earlier, chemical contaminants in drinking water can cause a variety of health problems other than cancer.
For most of these chemicals there is no quantitative model to assess the risk on the basis of dose. The Committee therefore simply compared measured concentrations in water supplies with EPA's maximum contaminant levels (MCLs).
This comparison provides an estimate of the number of individuals who are exposed to unhealthy levels of pollutants in drinking water, but does not give us a quantitative estimate of the illnesses or deaths that result.
Refer to Appendix C for a comparison of these risks.
Results of Data Analysis
Disease Risks from Community Water Supplies
o Coliform testing violations
Table 18.1 summarizes coliform violations for Arizona community water supplies in 1993. A violation is a failure to sample according to a prescribed schedule.
Table 18.1 Coliform Testing Violations in Arizona, 1993
-------------------------------------------------------------------------------------------- | Status | Number | Total | Percent of | Average | | | of | Population | Total | Service | | | Systems | | Population | Population | ============================================================================================ | No violations | 683 | 2,160,093 | 54.6 | 3,162 | -------------------------------------------------------------------------------------------- | Violations for 1 to 3 months | 820 | 1,656,957 | 41.9 | 2,020 | -------------------------------------------------------------------------------------------- | Violations for more than 3 months | 295 | 139,086 | 3.5 | 471 | --------------------------------------------------------------------------------------------
o Coliform MCL violations
Table 18.2 summarizes coliform MCL violations for Arizona community water supplies in 1993.
Table 18.2 Coliform MCL Violations in Arizona, 1993
-------------------------------------------------------------------------------------------- | Status | Number | Total | Percent of | Average | | | of | Population | Total | Service | | | Systems | | Population | Population | ============================================================================================ | Number of samples during 1993 | 107 | 19,469 | 0.5 | 181 | -------------------------------------------------------------------------------------------- | No violations | 1,470 | 3,563,207 | 90 | 2,423 | -------------------------------------------------------------------------------------------- | Violations for 1 to 3 months | 217 | 308,709 | 7.8 | 1,422 | -------------------------------------------------------------------------------------------- | Violations for more than 3 months | 4 | 64,751 | 1.6 | 16,187 | --------------------------------------------------------------------------------------------
An MCL violation means that the number of coliforms measured in samples were detected more than a specified number of times. The allowable number of positive (coliform-containing) samples per month varies with the size of the community water supply.
Table 18.2 shows that 81% of the community water supplies had no bacterial MCL violations during 1993. This represents about 90% of the 3.96 million people served by community water supplies.
Another 217 systems, serving 8% of the population served by community drinking water systems, had MCL violations for one to three months.
Only four water systems had bacterial MCL violations more than three months. This chronically-exposed population includes 64,751 people, about 1.6% of the population.
The majority of community water systems sampled their water regularly.
54% of the systems had no sampling violations.
42% of the systems had three or fewer missing samples.
Cumulatively, these two groups comprise 96% of the population served by community water supply systems.
Chronic sampling violations occurred in 295 community water systems that served 3.5% of the population.
No samples were collected at all for 107 systems serving 0.5% of the population. The latter are mostly small water systems serving less than 500 people.
In summary, community water supplies for 392,929 Arizonans, about 10% of the state's population, either did not meet the coliform MCL on one or more occasions or simply did not conduct the required coliform testing at all.
Many more community water supplies experience occasional sample violations.
o Turbidity
Arizona only requires turbidity testing for water systems that use surface water (about 60 systems).
Because sampling requirements changed in 1993, the Committee used 1992 data for this analysis. The data show:
17% of the community water supplies that require turbidity testing failed to sample on all required occasions.
Of the systems that were sampled, only two MCL violations were incurred.
All of the sampling and MCL violations for turbidity were seen in systems serving fewer than 10,000 people.
No community water supply experienced both turbidity and coliform MCL violations in 1992.
Taken together, the coliform and turbidity data indicate that a sizable fraction of Arizona's population is drinking water that does not meet standards intended to indicate the potential for microbial disease.
Reporting Water-Borne Disease Outbreaks
A second tool that the Committee used in assessing microbial risk assessment is the actual reporting of water-borne disease outbreaks.
The Centers for Disease Control (CDC) tracks outbreaks of water-borne diseases. Ideally, the Committee could use reported outbreaks as a basis for risk assessment.
In recent years (1986-1994) there was only one reported water-borne disease outbreak in Arizona. This was an outbreak of a viral infection that affect 900 people in Sedona in April of 1989.
The CDC reported 34 outbreaks of water-borne diseases affecting 17,464 people in the U.S. during 1991 and 1992 (Moore et al., 1994).
Seven outbreaks were caused by protozoan parasites (Giardia and Cryptosporidium).
o One outbreak was a viral infection (hepatitis).
o One outbreak was a bacterial disease (shigella).
o Two outbreaks were caused by chemical contamination.
o Twenty-three outbreaks were acute gastrointestinal illness ("the runs") of unknown origin.
o One person died, from fluoride poisoning (Moore et al., 1994).
If all microbial diseases were reported, the CDC reports would provide an accurate assessment of risks from water-borne diseases. These numbers would yield an annual risk to water-borne disease of around one out of 15,000 people.
It is well understood, however, that reported outbreaks under-represent the actual number of incidences of disease. Estimates of under-reporting range from a factor of ten to a factor of 100 (Gerba, U. of A.).
There are at least three reasons for substantial under-reporting:
o Reporting of suspected water-borne diseases is largely voluntary. Because of this, doctors are unlikely to report suspected water-borne diseases unless a substantial fraction (estimated at 1%) of the population suddenly becomes ill.
o Because many diseases that can be transmitted by water are also transmitted by food and contact with other diseased individuals, it is difficult to establish the etiology (cause) of a specific outbreak.
o Many incidences of mild gastroenteritis (stomach cramps and diarrhea) are simply not brought to the attention of doctors but are treated at home.
One recent study compared incidence of common diarrhea between homes receiving normal community water supply and those using water that had been further purified by in-home reverse osmosis systems. The study concluded that about one-third of all cases of diarrhea are the result of contaminated drinking water (Payment et al., 1991). This may represent a worst-case estimate, since this study was based on a water treatment plant that was downstream from a sewage treatment plant.
Outbreaks of protozoan diseases are generally not associated with coliform violations, but are more likely associated with filtration problems at water treatment plants that use surface waters. Because some surface water treatment plants serve large populations, outbreaks of protozoan diseases, particularly cryptosporidiosis, are often large.
Four major outbreaks of cryptosporidiosis in the U.S. and Britain in recent years have caused 15,000 to 370,000 illnesses each (NEJ, 1994). The most famous of these was the outbreak in Milwaukee, Wisconsin, in which 370,000 people became ill and around 140 died. Most of the deaths occurred in people who had weakened immune systems due to AIDS, cancer therapy, or other conditions.
Water-borne diseases pose the greatest risk for people whose immune systems are not functioning normally. This includes the elderly, those with AIDS, organ transplant recipients, and patients receiving cancer treatment.
Risks from Chemicals
Violations of chemical MCLs in drinking water are presented in Table 18.3.
Table 18.3 Select Chemical Contaminants that Exceeded Selected MCLs in 1993
-------------------------------------------------------- | Chemical | Number of | Number of | Population | | | Systems | Samples | Affected | ======================================================== | Arsenic | 24 | 57 | 12,705 | -------------------------------------------------------- | Benzene | 0 | 0 | 0 | -------------------------------------------------------- | Chromium | 4 | 7 | 60,000 | -------------------------------------------------------- | DBCP | 0 | 0 | 0 | -------------------------------------------------------- | EDB | 0 | 0 | 0 | -------------------------------------------------------- | Mercury | 0 | 0 | 0 | -------------------------------------------------------- | Nitrate | 63 | 251 | 251,023 | -------------------------------------------------------- | PCE | 0 | 0 | 0 | -------------------------------------------------------- | TCE | 1 | 3 | 174,170 | -------------------------------------------------------- | TCA | 0 | 0 | 0 | -------------------------------------------------------- | TTHM | 11 | 64 | 198,467 | --------------------------------------------------------
Modeled cancer risks are shown in Table 18.4.
Table 18.4 Estimated Cancer Risks from Select Contaminants in Drinking Water
-------------------------------------------------------------------- | Chemical | Lifetime Cancers in | Number of Cancers | | | Sampled Population | Per Year in Sampled | | | | Population | ==================================================================== | Arsenic | 1,200 | 17 | -------------------------------------------------------------------- | Trihalomethanes | 175 | 2.5 | -------------------------------------------------------------------- | Benzene | 0.5 | 0.007 | --------------------------------------------------------------------
Lifetime cancers in the sampled population were calculated using the formula adopted for this study.
Sampling for trihalomethanes is only required for surface water supplies, which serve about 1.37 million people.
The Committee assumed that all systems supplied by surface water supplies were sampled at least once.
Chemical-by-Chemical Risks
Arsenic
Among the chemicals evaluated, arsenic appears to be the most likely to cause cancer.
Using the cancer risk assessment approach adopted for this study, the Committee estimated that arsenic causes 24 cancers per year throughout the state.
On the average, the odds of an Arizonan getting cancer by consuming arsenic in drinking water is about one in 2,500.
o This risk is about twice as high for Yavapai County. The high levels of arsenic in many wells in Yavapai County are mostly natural (Owens-Joyce, 1984).
Concentrations of 200 microns/L are have been reported for several wells in the Camp Verde area.
Around 12,000 people drink water that exceeds the MCL, around 0.3% of the population.
In addition to cancer risks, arsenic is known to cause hyperpigmentation of skin and a disease called blackfoot disease at levels less than 170 microns/L.
Benzene
Benzene, a component of gasoline, was never detected at levels above the MCL of 0.005 mg/L.
The Committee infers a cancer risk of less than one cancer per year from benzene in community water supplies.
Chromium
Chromium exceeded the MCL in four community supply systems that serve 60,000 people:
o Paradise Valley
o Lake Havasu City (one violation)
o Show Low (one violation)
o Kingman (four violations)
DBCP
Although the use of this agricultural pesticide was halted in 1979, many wells still have measurable concentrations of DBCP.
A risk assessment study for DBCP (ADWR, 1989) found that 77 wells in Maricopa and Yuma counties contained measurable levels of DBCP, and that 10,994 individuals were exposed to DBCP in drinking water. However, no wells exceeded the MCL of 0.025 microns/L.
ADWR's report indicated that DBCP caused only 0.05 total cancers among this exposed population.
This translates to an lifetime risk of about one in 230,000 for the exposed population.
There were no violations of the DBCP standard in 1993.
EDB
Like DBCP, EDB has been reported at low levels in Arizona's groundwater (Daniel et al., 1988).
However, there were no violations of the EDB standard in 1993.
Mercury
There were no violations of the mercury standard in 1993.
Nitrate
Elevated nitrate levels in groundwater are fairly common in Arizona, particularly in Maricopa County.
Water supplies affecting a quarter of a million individuals (6% of the population) exceeded the 10 mg/L MCL at least once. About 12,700 people were drinking water that had an average nitrate level of 10 mg/L or more.
The large discrepancy between these numbers results from the fact that several very large water supply systems that violated the standard one or a few times had average nitrate levels below 10 mg/L.
o For example, Glendale (population = 160,000) had one sample that exceeded the nitrate MCL. The average nitrate concentration for Glendale for 15 samples was only 4 mg/L.
Unlike many contaminants for which the MCL is based on chronic exposure, short- term (acute) exposure to elevated nitrate levels could cause methemoglobinemia in infants. Thus, every MCL violation is a potential health risk.
The main reason for the 10 mg/L NO3-N standard is that nitrate can cause methemoglobinemia in infants. Although methemoglobinemia can occur at the 10 mg/L MCL level, this disease is very uncommon even when drinking water in which nitrate concentrations are very high.
When methemoglobinemia has been observed in infants drinking water with nitrate in the 10-20 mg/l range, it has generally been association with microbially- contaminated water or nitrate consumption from other sources.
A second reason that methemoglobinemia is not more common is that community water supplies which exceed the nitrate standard are required to send consumers a notice that the violation has occurred. This notice would warn consumers of the potential hazard. Presumably most people with infants would use alternative water supplies.
No cases of methanoglobanemia have been reported in Arizona in many years.
PCE
There were no violations of the PCE MCL for community water supplies in 1993.
Radon
There were no violations of the radon MCL for community water supplies in 1993.
TCA
There were no violations of the TCA standard for community water supplies in 1993.
TCE
One large water system violated the MCL for TCE in 1993.
Statewide, the cancer TCE contamination probably causes less than one cancer per year.
Trihalomethanes
THMs are sampled only for surface water supplies because it is unlikely that groundwater used for water supply will have significant THM levels when chlorinated.
The cancer risk for trihalomethanes was computed by summing up the risks from the individual components of THMs (chloroform, dichlorobromomethane, dibromochloromethane, and bromoform).
On the average, the average lifetime cancer risk to THM exposure is about one in 10,000. The Committee estimates that exposure to THMs in drinking water causes less than three cancers per year.
Uncertainties
Water-Borne Disease
The most severe limitation for assessing the impact of water-borne disease is that MCL violations for coliforms or turbidity are imperfect predictors of disease.
Although bacterial and viral disease outbreaks are usually accompanied by positive coliform tests, outbreaks from protozoan diseases are not.
Because of widespread concern about cryptosporidiosis resulting from the Milwaukee outbreak and from subsequent observations that over 80% of surface waters contain Cryptosporium, the EPA has proposed routine monitoring for this organism in surface waters.
Cancer Risks
The Committee used the zero threshold linear model to assess cancer risks. This model is widely used by EPA and other agencies but is not without criticism. The main criticism is that it assumes no threshold level of exposure below which there is absolutely no risk of cancer (tolerance thresholds are widely accepted for non-cancer effects).
Because there is considerable uncertainty regarding low-level thresholds, and no knowledge of what these thresholds limits might be, the Committee adopted this approach knowing that it might eventually prove to be overly conservative. Actual cancer risks for assessed chemicals are probably lower than estimated.
This uncertainty arises in large measure from the fact that controlled studies of carcinogens on humans would be unethical. Because of this, the Committee relied on epidemiological studies, generally based on high-level exposures (for example, industrial exposures prior to regulation), or laboratory animal experiments. Extrapolation from these types of studies to low-level cancer risks in human is accompanied by a high level of uncertainty.
Non-Carcinogens
For noncarcinogens, the Committee simply estimated the size of the populations that consume water that exceeds the MCLs.
The significance of MCL with respect to actual human health risks depends upon the safety factor that is incorporated into the development of the MCL. When the risks are well understood, the safety factors are low.
For example, the MCL for nitrate (10 mg/L) is exactly the same as the "no observed effect level." In other words, slightly higher concentrations are known to cause health problems, at least in some circumstances.
The same is true for fluoride. Fluoride levels twice as high as the MCL will result in mottling of teeth.
For chemicals whose toxicity is poorly understood, safety factors up to 1000 are used in the development of MCLs.
Chemical and Bacteria Standards
For both chemical and bacterial standards, there is some ambiguity regarding the point of sampling for compliance sampling. Therefore, in some cases, the water quality in the data base may not be exactly the same as the quality of water that consumers are drinking.
There is a considerable fraction of the population drinking water from community water supplies that has not been regularly tested.
Conclusion
o By far the most pervasive problem is a failure for many communities to meet bacterial water quality standards.
About 2% of the community water supplies regularly fail the coliform test.
About 4% of the community water supplies frequently violate the sampling regulations.
Only about half of Arizona's population drinks water from community water supplies that fully meet established bacterial standards.
Although no water-borne disease outbreaks have been reported in Arizona in the past few years, there are undoubtedly many thousands of mild gastrointestinal infections each year caused by contaminated drinking water. Beyond this, there is always the possibility of a much larger outbreak of more serious disease.
o The sum of estimated cancer risks from four contaminants--arsenic, trihalomethanes, TCE, and benzene--was about 20 cancers per year throughout the state.
Because this estimate is based on very conservative assumptions, it is probably higher than the actual incidence of cancers caused by drinking water.
The chemical that poses the greatest cancer risk is arsenic, which is a natural contaminant in some areas of Arizona.
o Nitrate contamination of groundwater is fairly widespread. A large population, about 6% of the state, is exposed to drinking water that occasionally exceeds the MCL.
A smaller population (12,000 individuals) drinks water from community water supplies that has an average nitrate concentration over 10 mg/L, indicating chronic exposure.
So far, there have been no reported cases of methemoglonemia in Arizona, but the prevalence of nitrate-contaminated public water supplies suggests that the potential for this disease is fairly high.
o Individual community supply systems have various contamination problems, but in the aggregate, the actual health risks from these problems is probably very small.
Untreated Surface Water And Groundwater
Untreated Surface Water
The Arizona Water Quality Assessment for 1994 indicates that two waterbodies in Arizona are contaminated to an extent that may threaten public health or aquatic life. They are:
o Painted Rock Borrow Pit Lake/Gila River/Salt River downstream from Phoenix from pesticides
o Lake Havasu from fecal coliform.
Anecdotal reports indicate that raw surface water use for potable purposes is very limited in Arizona, and is probably confined to isolated uses in mountainous areas.
The primary risks from potable use of untreated surface water are from bacteria and possibly from acid mine drainage.
The contamination at Lake Havasu appears to be the only surface water quality problem to which significant numbers of people are exposed. This contamination has resulted in closure of beaches along Lake Havasu several times in recent years.
Potable use of untreated surface water is not known to be practiced at dwellings such as farms, ranches and houses. However, potable use during recreation such as hiking, camping and backpacking is known to occur. Such potable use carries the risk of exposure to pathogens such as Giardia and hepatitis A.
According to ADHS, there were 333 cases of giardiasis reported in Arizona in 1994. This figure represents only a portion of all giardiasis cases in Arizona, since the disease is under-reported. Most giardiasis cases in adults can probably be traced to drinking untreated water. Thus, use of untreated surface water for drinking poses a risk to significant numbers of persons who camp and hike in Arizona.
Private Domestic Wells
Potable use of groundwater from private domestic wells occurs in many parts of Arizona, particularly in rural areas not served by municipal providers or private water systems.
The number of persons drinking water from private domestic wells was computed by multiplying the total number of wells used for domestic purposes by the number of persons per household in each county. The results are shown in Table 18.5.
Table 18.5 Domestic Wells for Potable Purposes in Arizona
------------------------------------------------------------------- | County | Number of | Persons Per | Number of | | | Private | Household | Persons | | | Domestic | | Using Private | | | Wells | | Domestic | | | | | Wells | =================================================================== | Apache | 6,944 | 3.80 | 26,387 | ------------------------------------------------------------------- | Cochise | 6,783 | 2.68 | 18,178 | ------------------------------------------------------------------- | Coconino | 3,426 | 2.99 | 10,244 | ------------------------------------------------------------------- | Gila | 3,759 | 2.56 | 9,623 | ------------------------------------------------------------------- | Graham | 1,197 | 3.09 | 3,699 | ------------------------------------------------------------------- | Greenlee | 823 | 2.85 | 2,346 | ------------------------------------------------------------------- | La Paz | 3,674 | 2.56 | 9,405 | ------------------------------------------------------------------- | Maricopa | 12,247 | 2.59 | 31,720 | ------------------------------------------------------------------- | Mohave | 2,971 | 2.47 | 7,356 | ------------------------------------------------------------------- | Navajo | 4,982 | 3.44 | 17,138 | ------------------------------------------------------------------- | Pima | 8,813 | 2.49 | 21,944 | ------------------------------------------------------------------- | Pinal | 5,075 | 2.83 | 14,362 | ------------------------------------------------------------------- | Santa Cruz | 1,499 | 3.35 | 5,022 | ------------------------------------------------------------------- | Yavapai | 11,085 | 2.35 | 26,050 | ------------------------------------------------------------------- | Yuma | 3,951 | 2.87 | 11,339 | ------------------------------------------------------------------- | Arizona Totals | 77,229 | NA | 215,083 | -------------------------------------------------------------------
The figures in Table 18.5 are based on 1990 U.S. Census information. The Table indicates that about 215,000 persons, or about 5% of Arizona residents, rely on groundwater from private domestic wells for drinking water supply.
The following text describes risks from private domestic wells not affected by hazardous waste sites such as Superfund sites or underground storage tanks, which are discussed in Chapter Seven.
Arsenic
Of the contaminants commonly found in rural domestic water supplies, naturally- occurring arsenic is the most common.
Arsenic is classified in Cancer Class A (known human carcinogen) by EPA. Most exceedences of the maximum contaminant level (MCL) for arsenic occur in the Verde Valley, the adjacent Agua Fria basin, and the New River area. These areas are growing rapidly in population and rural residents rely almost exclusively on private domestic wells that are not treated or blended with other supplies.
Water quality data indicate that arsenic concentrations ranging up to 240,000 microns/l are present in groundwater in 30% of wells sampled in the Verde River basin, which includes Camp Verde, Cottonwood, Jerome, Sedona, and part of Payson.
Elevated arsenic levels are also found in the Black Canyon City and New River areas.
The population of the Verde River basin in 1990 was 57,800 persons and the combined populations of the Agua Fria basin and New River area were about 6,300. The 1990 U.S. Census data indicate that approximately 28%, or 16,200 persons in the Verde River basin used unregulated private domestic water for potable use.
The same percentage is used for the Agua Fria basin and New River, indicating a total of about 18,000 persons using private domestic water from wells.
Using the risk assessment adopted by the Committee for this study, the risk for the annual number of new cancers from this risk ranges from 0.27 to 10.4.
Chromium
Chromium contamination of groundwater is found only in a few areas in the state. Most known elevated chromium concentrations occur in the Ranegras Plain area.
The Ranegras Plain area of southwestern Arizona is sparsely populated but almost completely dependent on groundwater supplies and private domestic wells. Approximately 300 people are exposed to naturally high levels of chromium in groundwater.
Water quality data indicate that groundwater in Vicksburg Junction-Hope area contains up to 240 microns/l chromium. The only water supply available in the area is groundwater from private domestic wells, so the entire population of this area is exposed to high levels of chromium, the effects of which are discussed in the first section of this chapter.
Lead
Historically, lead has not been a major problem in groundwater in Arizona. The only area identified by ADEQ in Arizona as having lead in groundwater is at Pinal Creek near Globe. The contamination is due to acid mine drainage. No human exposures to lead and other contaminants in groundwater are known in this area.
Pinal Creek flows into Roosevelt Lake, a primary source of drinking water for the Phoenix area, but no elevated lead levels in Phoenix-area water have been observed as a result.
Dissolved Solids
Dissolved solids, including nitrate, sulfate, fluoride and other constituents, occur throughout the state but are most abundant in agricultural areas and areas downgradient from major population centers and wastewater treatment plants.
The Buckeye and Gila Bend areas downgradient from Phoenix commonly exhibit elevated levels of dissolved solids.
Both active and former agricultural areas tend to exhibit elevated levels of nitrates.
Some nitrates in groundwater may be naturally occurring.
The MCL of 10 mg/l nitrates is exceeded in many wells in the Phoenix area. Other locations with nitrate problems are Bullhead City, Quartzite, and Green Valley.
Quartzite, Buckeye and Gila Bend probably contain the most persons exposed to elevated levels of nitrates from private domestic wells. The number of persons exposed in this manner probably number several thousand.
Sulfates are found in groundwater in historical mining areas such as Bisbee, Naco, San Manuel, and Globe/Miami.
Sulfates also occur in groundwater in Maricopa, Coolidge, Casa Grande, and Gila Bend.
Fluorides occur in wells in the Harquahala Valley, Ranegras Plain basin, Parker, Tombstone, St. David, and Mammoth.
In Ranegras Plain, 37 of 48 wells exceed health-based guidance levels for fluoride. As stated previously, almost the entire population in the Ranegras Plain area is dependent on groundwater from private wells.
Radiochemicals
Radon
Health risks from ingestion of radon are considered to be much lower than risks from inhalation.
Duncan et al. (1993) performed a reconnaissance of radon 222 (222Rn) in groundwater in the areas shown in Table 18.6. The mean values found, measured in picocuries (pCi) per liter are shown in the Table.
Table 18.6 Groundwater Reconnaissance for Radon 222 (222Rn)
-------------------------------------------------- | Location | 222Rn Concentration (pCi/l) | ================================================== | Payson | 3,644 | -------------------------------------------------- | Sierra Vista | 978 | -------------------------------------------------- | Paulden Area | 695 | -------------------------------------------------- | Kingman Area | 675 | -------------------------------------------------- | Yuma Area | 661 | -------------------------------------------------- | Verde Valley | 630 | -------------------------------------------------- | Safford | 560 | -------------------------------------------------- | New River Area | 596 | --------------------------------------------------
If the Payson-area concentration of 3,644 pCi/l (the largest in the reconnaissance), and an ingestion cancer potency factor of 1.4 x 10-12 risk/pCi are combined, a lifetime risk of 1.5 cases in 100 million persons could be expected.
Uranium
Uranium is found in groundwater in the Colorado Plateau area of northeastern Arizona.
Duncan et al. (1993) reports median uranium concentrations of 2,400 microns/l in a well near mine tailings near Tuba City, and concentrations of from 19 microns/l (median of fourteen wells) to 29 microns/l along the Puerco River.
Of 18 samples taken in the Casa Grande, Gila Bend, and Stanfield areas, six samples exhibited high uranium levels. In one sample, the uranium concentration was 52 microns/l.
In Bisbee, two copper mine monitor wells had uranium concentrations of 104 microns/l.
Several wells in the Prescott area exceed the proposed uranium MCL of 20 microns/l.
Other areas that exceed the proposed standard include:
o Carefree (24 microns/l)
o Casa Grande (33 microns/l)
o Dewey (20 microns/l)
o Gila Bend (21 microns/l)
o Rough Rock on the Navajo Indian Reservation (22 microns/l)
It is reasonable to say that most wells in the northwestern part of Arizona, where uranium mineralization is common in aquifers, and in areas dependent on granitic aquifers, some uranium may be found in groundwater.
The primary risk from ingestion of uranium in water is not cancer, but renal failure or diminished renal function. According to ADHS, a lifetime of ingestion of water containing 20-30 microns/l of uranium would present negligible risk of renal damage or cancer.
Pesticides
Pesticides occur in groundwater in limited areas in Arizona. Most contamination is due to dibromochloropropane (DBCP) and ethylene dibromide (EDB) in historic citrus areas in Maricopa and Yuma counties. Table 18.7 indicates locations of known contamination and concentrations in groundwater in those areas.
Table 18.7 Occurrence of EDB and DBCP in Groundwater in Arizona
----------------------------------------------------------------------------------- | Location | EDB | DBCP | | | Concentration | Concentration | | | (Microns g/l) | (Microns g/l) | =================================================================================== | Yuma (Yuma County) | 0.019 | up to 137 | ----------------------------------------------------------------------------------- | Chandler Heights (Maricopa County) | - | up to 1.5 | ----------------------------------------------------------------------------------- | Deer Valley (Maricopa County) | 8.0 | 0.8 | ----------------------------------------------------------------------------------- | Goodyear/Litchfield Park (Maricopa County) | - | 0.4 | ----------------------------------------------------------------------------------- | Laveen/South Phoenix (Maricopa County) | - | 4.5 | ----------------------------------------------------------------------------------- | Mesa Falcon Field (Maricopa County) | - | up to 4.74 | ----------------------------------------------------------------------------------- | Phoenix (Maricopa County) | up to 10,000 | - | -----------------------------------------------------------------------------------
In 1989, the ADHS performed a risk assessment for exposures to DBCP in these counties. Concentrations of DBCP ranging from 0.002 to 137.0 microns/l in groundwater from wells were noted.
The ADHS risk assessment calculated that 0.047 excess lifetime cancers may be due to DBCP exposure in the Maricopa and Yuma counties study area and that exposure is probably continuing via private domestic wells.
EDB is also considered a probable human carcinogen, and is found in groundwater in the same areas in which DBCP is found.
DDT and its metabolites DDD and DDE are found in the food chain and in soils in present and former agricultural areas in the state but are not detected in groundwater. Any exposure via surface water would be very limited.
Conclusion
o Because of extremely limited use of raw surface water for potable use, exposure to surface water via ingestion is very low and thus risk is very low. Some dermal contact and incidental ingestion of contaminated surface water may conceivably occur at sites such as Lake Havasu, the Middle Gila River, and Oak Creek.
o Exposure to naturally occurring arsenic in private wells in the Verde Valley and New River areas is the most serious problem affecting private domestic wells statewide. Lifetime risks of cancer from exposure to arsenic in these areas are estimated to range from 3.86 x 10-3 (low) to 5.79 x 10-2 (high).
o A limited number of persons are exposed to naturally-occurring chromium in the Ranegras Plain area in western Arizona.
o Ingestion of naturally-occurring radiochemicals such as radon is not believed to be a substantial health risk.
o Limited exposure to pesticides in Yuma and Maricopa counties is estimated to result in 0.047 excess lifetime cancers. No other pesticide occurrences in Arizona groundwater are known.
o Thousands of persons are exposed to noncarcinogenic dissolved solids such as nitrate, sulfate, and fluoride, but the cumulative health effects of these exposures are not known.
o While potable use of untreated surface water is not known to occur at dwellings in Arizona, exposure to pathogens such as Giardia poses a risk to persons who drink untreated surface water.
Acknowledgments
Terry Fields and Dean Landis, both of DEQ, contributed extensively to the data analysis in this section. Detailed reviews were provided by Chuck Graff and Cathy Nair, both with DEQ, Gerry Hiatt (EPA), and Jane Orient (ACERP Human Health Committee). Conversations with Norm Peterson (ADHS) and Chuck Gerba (University of Arizona) provided many useful insights.