ro" or ARS^ITC SOURCES
IN GROUMDJ-'ATSR
i, OrinV.ir.g ^.'Ater Rranch
U.S. Cnvironne.ntal Protection r.qency
:i?.y, 19Q1
-------
V'r>c occurrence of elevated arsenic concentrations in water supplies in vew
hanpshire is an issue of current public concern. Questions have boen
rdLS^'. regar-ling the health effects of varyinq arsenic concentrations, its
source, and alternative solutions to the situation. The Tnvironrcental
Protection Agency (EPM in P.oston ha = worked in conjunction with the Centers
For Disease Control (CDC) in Atlanta and the State of New Hampshire Water
S ipply and Pollution Control Commission to attempt to answer the«=e and
•Jt'.ior questions.
This report will address EPA's involvement with the arsenic issue in Vew
Hampshire and present our findings to date. It is important to note,
ho-vev-r, that ZP\' 3 authority under the Safe Drinking Water Act of 1074
extesi Is only co public water systems (serving 25 or more people). Because
the elevated arsenic levels are found nredonninantly in donostic (orivate)
•:atar supplies which arc not under federal jurisdiction, "PA was greatly
Ir-ited by the availability of funds -ind personnel to undertake a large
scale stu-Jy. However, from a public health aspect, CPA agreed to provide
support in three areas of major oublic concern: to provide analytical
assistance to CDC for their epidoniologiccil study; to conduct an investiaa-
tion to tier.2r.line the source of arsenic contamination; ard finally, to
provi-ie analytical support for a pilot project to evaluate the effectiveness
of various available hor.e treatment devices in reducing arsenic concentra-
tions. This report will focus prirarily on our findings regarding the
source investigation since both the CDC enii'eniolocjical study and the ho">e
troatiient device study are currently in progress. Findings of those
will oe Made available upon thsir conpletior..
OACKGBQTJUP
Up-3-=>r the Safe Drinking Water Act (SWA) of 1974, national standards for the
quality of public drinking water supplies were established to protect public
health. Routine chemical analyses of public drinking vat^r supplies in
Hudson, i.'ew Hampshire, as required by the SDWA, indicated that several wells
had arsenic concentrations exceeding the national standard of 0.05 nilli-
grans per liter (mg/1) recor.mended for drinking water. As a result of these
finding*! the Gtate of Mew Hampshire cond\icted additional arsenic testing
of approximately 200 domestic (private) water systems in the Hudson vicinity
to determine the scope of the contamination between April 1930 - January
1931. Approximately 10% of those 200 samples contained arsenic in excess
of the national standard, ranging from 0.05 mg/1 - 0.37 mg/1.
The Town of Hudson and several bordering towns made arrangements with
private laboratories for the testing of water samples fron privata wells.
The to^ns coordinated the collection of water samples and the compilation
of the results. The analyses were paid for by the well owners. These
results were a part of the data base available during this study. (Table
1)
Because of the increasing public interest in this matter and the State's
resource limitations, officials from the New Hampshire Hater Supply and
Pollution Control Commission requested assistance from both the Environ-
mental Protection Agency (EPA) and the Centers for Disease Control (CDC)
-------
for wat«r analyses as well as epider.iological investigations. L'PA anreed
to Tovide analytical -support and guidance to the State to the naxinun
extent possible and requested CDC's assistance in conducting an epidenio-
ioi'i.-ral 3tudy.
In r>eci-:~iV:fc:r 1930, ŁP"\ assisted the Stats in collecting water sarnies in
to\..~i.« adjoining Hudson to try to determine the extent of the arsenic con-
tamratior. in that vicinity and al^o provided analytical services to the
State f'jr .-ru.ila.ty control determinations.
In r"eV>ruary, 1980 Dr. Peter Protean of CDC conducted an eoidenxoloqical
nvestiqation of infants born in 1^30 in the Hudson area. Out of 17ft Hudson
l-jsbies 'norn in that year, inn \.-sre on nrlvate drinkina watar supplies; thesn
ire to be the Jata points in the study. The study orconpasses results fron
frse infor~Mticn sources: questionnaires, physical ex^-imat ions, and
urine and -Mter analyses. The ouestionnairas were in the Form of a personal
iptervi.2*1 with the mothers. Tafornation gathered v/as diet specific; i.p.
the notiers* pr3-ar.d ; ostr.an.il mating habits as v/ell as the infants' diet.
f-ft^r these vitervie-'s/ Or. LTot.ian ex.anin^d the j'cin and ^'^i".o"i»?T3 of the
nfants foe oiavic.is synr.t-.'T'^s oF arsenic poisoning. Urinn sanoles fron t>ie
infanti and water "r-.vvoles from the inoivii-'i.-'al hones -./ere al-so coll»ctd
witn. CDC's Piolonic=»l Laboratory perfor-nn7 the uriie analyses and "TPA's
:iu-iicij"-?.l ~:n"iror.incntal .".ssoarch Laboratory ClSRL) in Cincinnati perf-.»r.-iirn
the vvrtc-?r analyses. *-.ssi-ilating all of thn in-jicated data, COC's finding*:
vi 11 '>e ";as^d uiop c<^rr-ilatio-is dr=»-«n fror arsenir levels found in urine
ar.'-l -y^ter of the exposed versus the non-exposed population. Conpletion
of the study 13 expected within t/e noxt few nonths.
In -addition to aosistng CDC with their study, EP7V recognized the need to
evaluate various w*ter treatment r-ethods for their effectiveness in arsenic
removal and to investigate and determine the source of arsenic contamina-
tion. Since r:rzv r'L^T. is currently conducting an evaluation of point of
\ice treatment systems for arsenic renoval, they agreed to provide sar.ple
analysis for our regional evaluation. This study has been initiated, and
the results will be presented in the near future. The findings of the
source study are presented herein.
Contaminant tioyeirent
The arsenic occurrence is dependent on nany hydrogeologic paraneters.
Understanding these p^raneters will aid in the comprehension of the problem.
The following oaragraphs are a brief description of the typical hydroc-*>o-
logical conditions of the Hudson area.
Groundwater flows fron areas of higher elevation to areas of lower elevation,
jast as surface water does. However, groundwater moves at a nnuch slower
rate than surface water, (feet or fractions of feet per day versus feet per
second) because groundwater moves through soil and rock and not over the
ground surface. This slower rate of movement allows the formation of a
discrete plume of contaminated groundwater downgradient of a contamination
source. This plume can be traced back to the source of contamination.
The movement of contaminants in the groundwater system is dependent not only
on the topography but also on the unconsolidated and consolidated materials
of the area.
-------
-3-
Svou;ulvrf3ter occurs between grains of sand and gravel in the soil, and along
joints d-id fractures in the bedroc'--. Joints are narrow, planar onenings,
•>one of ,ihich extend for long distances. Th<=y are rap.donly spaced and can
• if =»t ill angles from honrontal to vertical. In general, joints intersect
oach other and ~iay be interconnected over a considerable area. Joints
occupy only a very snail ->art of the total part of the bedrock volune.
7'.i.?rjCor<;, the porosity, permeability, and specific yield of the bedrock
ar-: low. '"hen a well is Drilled into t'-e bedrock, water is obtained fron
one -or .nany of the joints which the veil intersects. The quality and
quantity of the *at.ir varies deoending on the joints which are intercepted
an..; the naterials contained in them.
* f
Investigated gources
ila?ar-loi3s %ast.a, pesticide's and naturally occurring arsenic-bearing minerals
in the soil or roc:: ./ere cited as possible sources for the elevated
levels in Lhe tooted water supplies. The information studied to
in.inratos a natural, bedrock relate:! source for the -jrsenic.
I. l-a.-ar-Jous \;aste
iiii.:ar.lous vciste is not the most nrobable source of arsenic in southern !:ew
lia-ip'shire since thare appears to be no likely waste sources or generators
of arsenic in t\e Hudson area. Sulfide mineral nining and smelting are
the largest p-an caused contribution of arsenic to the natural cycle
(National Acadeiy r>f Sci-ence*?, 1976). T'nesfi industries are not condor, to
tl->.^ area, «nd r.o other --/aste sources of arsenic -/ere found.
A hazardous waste dunp v;ould be a localized source of pollutant's. Leach-
ate, rfhich forms with rain «/ater , percolates through dump material and
roaches groundv/ater to for-n. a contanination olume that moves down ground-
water gradient from the source. In-oared groundwater plumes are lirited in
areal extent. .'laps which locate the donestic water supplies tested for
arsenic have been prepared by the Town of Hudson and other nearby towns.
rhe-se Maps indicate the occurrence of groundwater containing elevated
arsenic levels is random and widespread, and not centralized in a defined
plune.
As sunnarized in the preceding paragraphs, the absence of arsenic waste
sites and qenerators, as well as the random presence in groundv/ater, appears
to eliminate hazardous waste as a source for the contamination.
II. Pesticides
Arsenic-bearing pesticides, another suspected source, do not appear to be
the primary source of arsenic in groundwater. However, in sone cases,
pesticides nay contribute to the arsenic content of the groundwater.
Arsenic-bearing pesticides were used in New Hampshire from the late 1300 's
until the early 1900 's primarily in apple, pear, and potato farming. The
FDA imposed residue tolerance levels on arsenic found on fruit in 1926,
(Xazmaier, 1931) and, consequently, the use of arsenic pesticides was
reduced. Arsenic pesticide use by Hudson farmers from 1966 was provided
by the New Hampshire Pesticide Commission. Table II indicates that only
small amounts of arsenic bearing pesticides have been used recently.
-------
-4-
A«, '.'ito leachat-.os from a Hazardous waste site, arsenic from pesticide use
/ould represent a localized source (an orchard, for example ) and should
rov C.jdson, 1931) do not suqoest a discrete plume,
but suirri-st a random occurrence independent of land ase.
Finally, arsenic is readily bonded in top soil by natural processes (\'oolson,
1951, rational V:.vie'r.y of Sciences, 1^75), and, therefore, would not be
reloa-jp.-1. in nassi'«» quantities, hut over a period of years. Slow release
of arsenic should niniir.ize the concentrations of arsenic in groundvater,
and not cause the elevated levels which have been found.
These F
-------
-5-
•jreater percentage of bedrock wells contain detectable limits of arsenic
t'nan the shallower, soil wells (-12% of Hudson's tested bedrock veils contain
nore than .01 milligrams per liter versus 2% of tested soil veils).
Th-> luu'.son area is underlain by rocks of the Oakdale a.nd Paxton Formations
(Harosh et al, 1978). The rock sequence is composed of thin-bedded calcare-
ous 'i.c T'.er.y of Sciences, 197G). "ven
At the low end of this range there would be enough arsenic in pyrite to
causo elevated levels in the grour. 1r.ot«r.
During cbis investigation, several minerals have been suggested by residents
of the area as h-sing possible sources of arsenic. Among thp«5-» are lollinrj-
ite and arsenopyrite. Lollingite is a rare mineral, and is found only in
a limited tyne of geologic regine ("eyers & Stewart, 1959) which is not
to occar in the Hudson area. Arsenopyrite is not corwnon in the
.iic fornations vwhich underlie Hudson and has not been found Curing
this study. Therefore, these two minerals are not considered to contribute
to the elevated arsenic levels in Huson.
i-s previously discussed, groundv/ater occurs in joints in the bedrock. If
a jOLnt contains mi-.erals which are a source of arsenic, th«>n groun-'-vater
in that joint could contain elevated levels of arsenic which would he
reflected in wells penetrating the joint. neighboring joints ^ay be free
of arsenic-bearing minerals so wells tapning water from these joints would
be free from arsenic. This could explain the random distribution of
elevaceJ levels.
In an attenpt to sunnort the occurence of arsenic in the bedrock, both =)PA
and Northern Analytical Laboratories tested water sanples for elements which
are known to occur in nature with arsenic. Northern Analytical Laboratory
provided the EPA with an Inductively Coupled Plasma (ICT?) multielement
scan of 25 water samples from Hudson. Concentrations for the following
elements were determined: Zn, Pb, DC, Ni, B, Mn, Fe, Cr, Kg, Mo Al, V,
Be, Ca, Cu, Ag, Se, Sr, Ba, Ti, Y, and Co. The samples tested contained
arsenic levels varying from undetectable to 0.367 milligrams per liter.
The results show no predictable patterns or ratios of various elements,
therefore no conclusions could be made from the data. An additional 11
samples were tested by EPA for concentrations of As, Ca, Ba, Pb, Cu, Ni,
iln, Ilg, Se, and Fe. Again the results were inconclusive. Although this
multielement scan does not support the theory that arsenic is present in
the rock, it also does not disprove this theory since arsenic is not always
associated with a particular mineral or group of elements.
Several homes in the Hudson area are using paper filters to remove sediment
from their drinking water supply. A sediment sanple obtained from one of
-------
-fi-
t'-.ese filters was obtained for determining the ojcdence of arsenic in the
se.li.i3nt, IT! the source naterial for the arsenic. The analysis indicated
that the sample conained 3100 ppm (.31%) arsenic. However, the exact
•ineralogy was -"-asked by the abundant quartz contained in the sample.
Additional sanple preparation indicated copper arsenosulphides and silver
arsaiosulphides or a r^i
-------
-7-
arsenic, and the anticipated length of tine before regeneration or replace-
.wr.t of unit components is necessary. We will also have a weekly record of
the arsenic levels fro.* too wells over a six month period to measure fluc-
t -.atior.s in arsenic concentrations.
wiples from sediment filter treatment units in use in the area
..•ill be obtained in an attempt to determine a parent material for the
ar.,e-uc. ::-ray Ji- fraction of filtered mat-rial from one wall has indicated
a .-iin«ral source but preparation of the sample seems to have destroved its
-
?os;j.hle C5JS.5S fcsr the -iobi Upturn of arsenic (for example, iron ba-.-
will be researched and tests will h-s riade to determine if these 'caus-s
could he oli-anated.
As part of -?.',.'s support of CDC's epide.nological study, drinkincj -.;at-r
samples fro.n four wells *eru sent to several laboratories for arsenic
speciation. Speciation studies rleternine the valence of the arsenic in the
-.aiple, either arsenite ( ^3 ) or ar senate ( + 5). CDC requested this data to
det-2r.--.ine whether exosoure to As (+3) or As (+5) had rade a difference in
the risk factor to those infants in the study.
rPA's ;ie-/ Cnqland Re?ion«il Laboratory in Lexincrton, Massachusetts prepared
split sa-nples and sent the.~s to laboratories having a current interest in
the detection of low levels of arsenic (+3 and +5 species).
Table IV presents the results from the five laboratories that participated
in the st-.idy and lists the analytical nethods used. As can be seen on
Table IV, both As ( + 3) and As (+5) are present. Due to the difficulty of
spvjciation neas-irenents and the range of results, al] of the investigators
agreed that averaging the data and listing the ranges would be the best way
to present the results. CDC has received copies of these results and will
include them in their report.
-------
References
•3aros'i, P.J., Fahey, R.J. and F-^ase, X.H., The 3edrock Geology of the Land
Area of t!ie Boston 2° L'neet; Massachusetts, Connecticut, Rhode Island, and
;;c-». :i:i.\r>fehire, U.-J. Gcoloqic dirvey, Open File Renort, 1973.
Parosh, P.J., and Dease, II.H., Correlation of the OdV.-iale and Paxton Torm-
atLO".^ ••/it'i their Equivalents froni Eastern Connecticut to Southern 'laine,
:.n Abstracts with PrograT.s, Ceol. Coc. AM. Vol:13 No. 3, 1981.
'Hiil^oi (Tnvn of), sr-^onic in ;rat'2r (".ap), 1Q31.
"v^zmaier, -i., T'eno to A.C. Tjincoln, 'Hrief Review of ''ajor TJses of Arsenical
Pesticic'e^, "?.rch 24, 1931.
Neynrs, T.R., and St-jwart, C.T7., The f~-eology of "ew '-'annshire, Part III,
••1'ies eind Minerals, '.3e-.; I-crpshire Department of Resources and Economic
Development, Toncord, 7T-? ^ "an^sliire, 10|5°.
L-'ational Ac.ii"Je:ny of Sciences, Arsenic, '?ashington, D.C., 1976^
Woolson, "Z. , Personal Ccv-ununication, April 15, 1981.
-------
TABLE I
AKiJEIilC CO'JCSNI icc ?oi <.co|j - .37
e jna'.yons3 1000 <.002 - .620
il-3 to CP--»
20 ql tern at .2 wat-sr sources vere sanplef? nil <.005 10/1
'.l=ic.a ,i
-------
TAPLE II
TOTAL AflOUCiT OF -^buMICAL CO'TOUMDS USCD IN HUDSON, NH
19r,6
320 lo
1 Ji-7
u:n iii
106:'
2^3 lh
1-»59
100 lh
1
1970
1oO Ib
1971
0
1972
0
1-173
0
1074
0
I
1975
0
1976
0
1377
0
1--'78
2 A lh
-------
": in
oF ShAT.LOW "FLT., P.'JSTS
T1"^ . /* '
Hudson
Lit.~hfi.elcl
Londonderry
TO."1;,:,
%
i DUG
25
2?
13
31
'•A:;K --,p
CR-V.TL "AC'TD
1
4
3
3
UNSPECIFIED
13
13
I Veils Sho-.-inq
TOT'"L | dS <.002 P-.o/l
1
45 | 37
1
25 | 23
1
16 | 0
I
I
37 | 60
I 59%
"ells shoving
as > .050 ng/1
0
1{wash?)
1Mu«)
2
2*
:JOTS: A sa'-.-'ary of d;
results on soil
in this TaMe.
fron uitchfield
cnat these -/ells
T9ble, 6?^ of the
ninety percent
ta fro.n the Ll.ree towns fron which -VR have r«ceiverl
w-^lls (dug, -vas'-i, oravel packsd, point) xs nccse-itrst'.
It should be r:otecl that public water surcoly wells
ami f;ud«;on have not been in c lade''.. Tosts indicate
are belov the nCL for arsenic.- As shown on the
wells are below che detection liiit. In addition,
of the -./ells renortec1 are less than .01 irillj.i-,rrirs/
liter and only 21 exceed the .05 ng/1 MCL.
-------
,6: IV
s-.nle No.
SL>ECI
------- |