Quality f Grund Water frm Private Dmestic Wells Find ut details abut the recently released USGS reprt n grund water wells. By Leslie A. DeSimne, Pixie A. Hamiltn, and Rbert J. Gillim This article highlights majr findings frm tw USGS reprts: DeSimne (2009) and DeSimne and thers (2009). These reprts can be accessed at http://water.usgs.gv/nawqa. This article is fllwed by a summary f treatment cnsideratins and ptins fr wners f private dmestic wells, written by Cliff Treyens f the Natinal Grund Water Assciatin. The Natinal Water-Quality Assessment (NAWQA) Prgram f the U.S. Gelgical Survey sampled grund water frm abut 2100 private dmestic wells acrss the United States frm 1991 t 2004. The sampled wells were lcated in 48 states and in 30 f the natin s 62 principal aquifers used fr water supply (Figure 1). Dmestic wells are an imprtant water supply t characterize because mre than 43 millin peple abut 15% f the natin s ppulatin rely n these wells fr drinking water. The quality and safety f water frm dmestic wells are nt regulated by the Safe Drinking Water Act r, in many cases, by state laws. Rather, individual hmewners are respnsible fr maintaining their dmestic well systems and fr any rutine mnitring. In this study, samples were cllected frm wells prir t mst in-hme plumbing r treatment and thus d nt necessarily represent water frm the tap. As many as 219 prperties (such as ph) and majr ins, nutrients, radinuclides, trace elements, pesticides, vlatile rganic cmpunds, and micrbial cntaminants were measured. Measured prperties and cnstituents are referred t in this article and in the USGS study as cntaminants, regardless f the measured cncentratin r ptential fr adverse health effects. This terminlgy fllws the definitins f the Safe Drinking Water Act, which states that a cntaminant is any physical, chemical, bilgical, r radilgical substance r matter in water. Hwever, the presence f a cntaminant in water des nt necessarily mean that there is a health cncern. Cntaminant cncentratins are placed in a human-health cntext thrugh cmparisns t human-health benchmarks, including the U.S. Envirnmental Prtectin Agency s maximum cntaminant levels (MCLs) fr cntaminants regulated in drinking water (which includes prpsed MCLs fr radn), and USGS healthbased screening levels (HBSLs) fr unregulated cntaminants. The EPA MCLs and mst state standards are enfrceable regulatins fr public water systems, but are nt used t regulate cntaminant cncentratins in dmestic wells. HBSLs are nnenfrceable guidelines fr unregulated cntaminants develped by the USGS in - 1 -
Figure 1. Dmestic wells sampled in this study are lcated in 48 states and parts f 30 f the 62 principal aquifers f the United States. Dmestic wells are categrized by principal aquifer rck type: blue basin-fill and ther nn-glacial sand and gravel aquifers; grey glacial aquifers; yellw castal plain aquifers; green sandstne aquifers; purple sandstne and carbnate aquifers; brwn carbnate aquifers; range basalt aquifers; red crystalline-rck aquifers. - 2 -
cllabratin with the EPA and thers, using EPA methds and the mst current EPA peer-reviewed, publicly available health-risk assessments and txicity infrmatin. Because HBSLs are calculated using EPA txicity infrmatin and methds, HBSLs are equivalent t existing EPA lifetime health advisry and cancer risk cncentratins (when they exist), except fr unregulated cntaminants fr which mre recent txicity infrmatin has becme available. Cmparisns f measured cncentratins t MCLs and HBSLs serve as a screening-level assessment that prvides an initial perspective n the ptential significance f cntaminant ccurrences t human health and n pririties fr future mnitring and research. Such assessments are nt designed t evaluate specific effects f cntaminants n human health, nr are they a substitute fr cmprehensive risk assessment that generally includes multiple expsure pathways and many additinal factrs. Recent advances in labratry analytical methds have given USGS scientists the tls t detect a variety f cntaminants at lw cncentratins, ften 100 t 1000 times lwer than drinking water standards and ther human-health benchmarks. Lw-level detectins d nt necessarily indicate a cncern fr human health, but rather help identify the envirnmental presence f cntaminants (sme f which may nt be cmmnly mnitred) and help track changes in cntaminant ccurrence and cncentratins ver time. Results f this study generally are cnsistent with previus studies f dmestic wells but als mre fully characterize a cmprehensive suite f individual cntaminants and cntaminant mixtures than previusly reprted. This study als describes cntaminant ccurrence fr reginal aquifers, prviding a framewrk fr understanding waterquality cnditins within units f similar hydrgelgy althugh cntaminant ccurrence can vary ver shrt distances and with depth within these aquifers because f variability in cntaminant surces, aquifer characteristics, and gechemical cnditins. Selected prtins f the principal aquifers were included in the study design, which fcused n NAWQA grund water assessment areas that were sampled during 1991-2004, and the findings are nt statistically representative f all dmestic wells acrss the U.S. Majr Findings Mre than ne in five f sampled dmestic wells cntained ne r mre cntaminants at a cncentratin greater than an EPA MCL r ther human-health benchmark. Benchmarks used are EPA MCLs fr cntaminants regulated in public water systems and USGS HBSLs fr unregulated cntaminants. Althugh n individual cntaminant exceeded a human-health benchmark in mre than 7% f sampled wells, 23% f wells cntained ne r mre different cntaminant greater than either an MCL r HBSL. Of the sampled wells, 13% had ne r mre different cntaminant greater than an MCL. Cntaminants mst frequently fund at cncentratins greater than human-health benchmarks were inrganic chemicals, mstly derived frm natural surces. These included radn, arsenic and several ther trace elements, nitrate, and fluride (Table 1). Each f these cntaminants was individually greater than its benchmark in abut 1% t 7% f sampled wells (using the higher f tw prpsed MCLs fr radn 4000 piccuries per liter, pci/l). Except fr nitrate, these cntaminants riginate in grund water primarily frm gelgic surces. Marked reginal patterns amng principal aquifers were nted in the ccurrence f many naturally ccurring cntaminants. Radn, fr example, was fund at relatively high cncentratins in crystalline-rck aquifers in the Nrtheast, in the central and suthern Appalachians, and in central Clrad. The ccurrence f radn in grund water is cntrlled in large part by the presence f uranium-bearing rcks. Nitrate was the mst cmmn cntaminant derived frm man-made surces that was fund at cncentratins greater than human-health benchmarks. Nitrate was measured at cncentratins greater than the EPA MCL f 10 milligrams per liter as nitrgen in abut 4% f sampled wells. Nitrate ccurs naturally in grund water, but elevated cncentratins usually riginate frm man-made surces such as fertilizer and septic system effluent. A separate analysis f 436 additinal dmestic wells in areas f relatively intense agriculture shwed that nitrate cncentratins were greater than the MCL in nearly 25% f the wells. - 3 -
Table 1. Eight inrganic cntaminants were fund at cncentratins greater than human-health benchmarks in 1% r mre f dmestic wells sampled in this study. Except fr nitrate, these cntaminants are derived primarily frm natural surces. [Human-health benchmarks: MCLs are U.S. Envirnmental Prtectin Agency Maximum Cntaminant Levels fr public water supplies and HBSLs are U.S. Gelgical Survey health-based screening levels. Surces f cntaminant in drinking water: mdified frm U.S. Envirnmental Prtectin Agency (2008) 1 and frm references cited in DeSimne and thers (2009). pci/l, piccurie per liter; ug/l, micrgram per liter; mg/l, milligram per liter.] Cntaminant Human-health benchmark Value Type Surces f cntaminant in drinking water Number f wells sampled Frequency f cncentratins greater than benchmark (percent f wells) Principal aquifers with cncentratins mst frequently greater than humanhealth benchmarks Arsenic 10 ug/l MCL Aquifer materials; pesticide applicatin Brn 1,000 ug/l HBSL Aquifer materials; sewage r septic-system effluent; fertilizer applicatin; sme industrial wastes Fluride 4 mg/l MCL Water additive which prmtes strng teeth; aquifer materials; discharge frm fertilizer and aluminum factries 1774 6.8% Crystalline-rck aquifers in New England; basin-fill aquifers in the western and suthcentral U.S., and basaltic-rck aquifer in Idah (mre than 10% f wells > 10 ug/l). 535 1.3% Sme basin-fill and sandstne aquifers in the western and suth-central U.S. 2157 1.2% Sme basin-fill and sandstne aquifers in the western and suth-central U.S. (abut 4% f wells > 4 mg/l). Manganese 300 ug/l HBSL Aquifer materials; sme industrial wastes Nitrate 10 mg/l as N MCL Fertilizer use; manure; sewage and septic-system effluent; aquifer materials Radn 4,000 pci/l 300 pci/l Prpsed MCLs 2 Radiactive decay f uranium in aquifer materials 2159 5.2% Glacial aquifers; sme sandstne aquifers in central Appalachian regin; sme castal plain aquifers in Sutheast (abut 8% t 60% f wells > 300 ug/l). In all aquifers, in wells with lw disslved xygen. 2132 4.4% Sme basin-fill aquifers in the Suthwest and Califrnia; glacial aquifers in the upper Midwest; sme castal-plain and crystallinerck aquifers in the central Appalachian regin (mre than 10% f wells > 10 mg/l as N). Generally, in areas f agricultural land use. 1958 4.4% Crystalline rck aquifers in the Nrtheast, central and suthern Appalachians, and central Clrad (abut 30% f wells 65% > 4,000 pci/l). Strntium 4,000 ug/l HBSL Aquifer materials 488 7.3% Sme basin-fill and sandstne/carbnaterck aquifers in the Suthwest and suthcentral U.S. Uranium 30 ug/l MCL Aquifer materials 1725 1.7% Basin-fill aquifers in the West; crystalline-rck aquifers in the Rcky Muntains and in the Nrtheast (abut 3% t 26% f wells > 30 ug/l). 1 U.S. Envirnmental Prtectin Agency, 2008, Drinking water cntaminants: accessed Nvember 25, 2008 at www.epa.gv/safewater/cntaminants/index.html. 2 U.S. Envirnmental Prtectin Agency, 1999, Prpsed radn in drinking water rule: Washingtn, DC, U.S. Envirnmental Prtectin Agency, Office f Water, EPA 815- F-99-006, 6 p., accessed Octber 26, 2005 at www.epa.gv/safewater/radn/prpsal.html - 4 -
Few man-made rganic cntaminants exceeded human-health benchmarks. Only seven f 168 rganic cmpunds analyzed tw insecticides, ne herbicide, tw slvents, and tw fumigants were fund at cncentratins greater than human-health benchmarks, each in less than 0.5% f the sampled wells. Diverse rganic cmpunds (including herbicides, insecticides, slvents, disinfectin byprducts, gasline hydrcarbns and xygenates, refrigerants, and fumigants) were detected in 60% f the sampled wells at lwer cncentratins. This indicates that a variety f cntaminant surces (including agricultural, dmestic, and industrial) can affect the quality f water frm dmestic wells. Abut half f the sampled wells had at least ne prperty r cntaminant at a level utside the range f values recmmended by the EPA fr nn-health-related, csmetic, r aesthetic purpses. Levels f ph and cncentratins f disslved slids, irn, and manganese were cmpared t EPA secndary maximum cntaminant levels (SMCLs), which are nn-enfrceable guidelines regarding aesthetic quality r ther nnhealth effects, and were utside SMCL ranges in 15% t 21% f wells. Fluride cncentratins were greater than the fluride SMCL in 4% f wells. Reginal patterns were apparent. Fr example, lw ph values ccurred in wells that tap crystalline-rck and castal-plain sand aquifers, mstly in the East, that have relatively little capacity t neutralize acids. High disslved slids were fund in basin-fill, sandstne/carbnate, and sandstne aquifers underlying parts f the western and suth-central United States, which can result frm easily weathered rcks and sediments, dry climate, and heavy irrigatin. Micrbial cntaminants were detected in as many as ne-third f sampled wells. Ttal clifrm bacteria, a brad grup f bacteria frm sil, water, and animal feces, were detected in abut 34% f sampled wells. Escherichia cli was detected in abut 8% f sampled wells. E. cli is typically nt harmful itself, but is an indicatr f fecal cntaminatin and, therefre, the pssible presence f pathgens. The apprximately 400 wells sampled fr micrbial cntaminants were a subset f the 2100 study wells and were clustered in a few gegraphic areas. Cntaminants usually c-ccurred with ther cntaminants as mixtures. Of the sampled wells, 4% cntained mixtures f tw r mre cntaminants at individual cncentratins greater than health benchmarks (using the higher f tw prpsed MCLs fr radn 4000 pci/l). Nearly 75% f sampled wells cntained mixtures f tw r mre cntaminants with cncentratins greater than ne-tenth f their individual benchmarks. Cncentratins greater than ne-tenth f benchmarks were used t identify cntaminants that were appraching levels f ptential health cncern. The mst cmmn mixtures f cntaminants at these levels were cmpsed f frequently detected inrganic cntaminants including nitrate, arsenic, radn, and uranium. Nitrate and (r) radn at cncentratins greater than ne-tenth f their benchmarks cccurred with the rganic cmpunds atrazine, deethylatrazine, r chlrfrm in 10% t 15% f sampled wells. The cmbinatin f nitrate and atrazine may have greater ptential fr health effects than either cntaminant alne, but there are n health benchmarks fr this mixture. Implicatins and Future Needs Overall, mre than ne in five dmestic wells had ne r mre cntaminants at a cncentratin greater than a human-health benchmark, which suggests a substantial ptential fr adverse effects n human health. Hwever, findings shw that the ptential fr effects is nt evenly distributed acrss the natin. Many cntaminants in dmestic wells fllw distinct gegraphic patterns related t gelgy, gechemical cnditins, land use, r ther influences. The greatest ptential fr successfully addressing water-quality cncerns fr dmestic wells is with targeted appraches in specific areas where cncentratins are highest in relatin t human-health benchmarks, and high prprtins f the ppulatin depend n dmestic wells. Ntably, traditinal wellhead prtectin appraches t preventing cntaminatin generally are nt effective fr cntaminants that ccur naturally in an aquifer. Findings frm this study underscre the cntinuing need t imprve assessment and scientific understanding f the water quality f dmestic wells. Several specific needs are highlighted as fllws. - 5 -
Imprved infrmatin is needed n the number f peple cnsuming water frm dmestic wells in specific regins and aquifers, and n water quality cnditins in the particular aquifer znes that are tapped by wells. Such infrmatin is essential fr evaluating the ptential human health implicatins and pssible mitigatin appraches. Cntinued public educatin and testing f dmestic wells are needed because hmewners are nt required t mnitr private wells and may nt be aware f cntaminants in their surce f drinking water. In particular, hmewners may nt be aware that many cntaminants knwn t be f cncern are primarily frm natural surces in sme areas, r that cntaminants frm man-made surces may be present in grund water frm current r previus land uses r activities. Cntinued water-quality assessment and research are needed t mre fully understand natural and man-made factrs and transprt mechanisms assciated with the mvement f cntaminants t dmestic wells. This is particularly imprtant because f the cmplex and ften lcalized nature f grund water flw and quality. Finally, refined infrmatin n the water quality f dmestic wells, as with ther drinking water surces, will need t be integrated with mre cmprehensive risk assessment t place the ptential cncerns abut specific cntaminants and mixtures f cntaminants in a brader human-health cntext. The ttal cmbined txicity f cntaminant mixtures can be greater than that f any single cntaminant. Cntinued research is needed because mst humanhealth benchmarks are based n txicity data fr individual cmpunds and the ptential fr additive r synergistic effects f mixtures f cntaminants at lw levels is nt well understd. This article is based n technical findings presented in USGS Scientific Investigatins Reprt 2008-5227 Quality f Water frm Dmestic Wells in Principal Aquifers f the United States, 1991-2004 and in a cmpanin USGS reprt, Circular 1332. These reprts, alng with data and dwnladable graphics, can be btained nline at http://water.usgs.gv/nawqa. Questins regarding the reprts shuld be directed t Leslie DeSimne in the USGS Massachusetts Rhde Island Water Science Center at (508) 490-5023 r ldesimn@usgs.gv. WaterWell NGWA. - 6 -