Groundwater chemistry during test-pumping at Sundby, Verdal, Mid-Norway

(1)

Groundwater chemistry during test-pumping at Sundby, Verdal, Mid-Norway

BERNTOlAV Hll MO,OlAM.SJETHER&SISSEl TVEDTEN

Hilmo, B.O..Soother, O.M.&Tvedten,S. 1992:Groundwater chemistry duringtest-pumping at Sundby,Verdal,Mid-Norway.Nor.geol.unaers.Bull.422,27-35

TheSundby aquifer,Verd al, Mid-Norwayhasbeen evaluated foryield andgroundwate rquality as a potential water resource for thelocalmunicipality.Test-pumping ofthe aquife rcommencedin 1989,after detailed geological investigations.The aquifer's potential yield satisfies the demand, but the observedincreaseinsalinity asfunction ofwithdrawalposesaproblem.

The geological setting of theaquifer,thedistan ce tothe fjord and the ionic ratios observed in groundwatersamplesindicatethat the salinityisnotthe result ofintrusionofpresent fjordwater,but ismostlikely duetotheleaching ofsalts fromrelict connatewaterin the surrounding glaciomarine or marine sediments.

Bemt O/av Hi/mo &a/a M.seetner,Norges Geotoqiss» Unaerseketse,Postboks3006-Lade,N- 7002 Trononeim,Norway.

Sisse/Tvedten,CH Knudsen AlS,Buskerudveien70, N-3024Drammen,Norway.

Introduction

The Sundby aquifer lies by the River Verdalsel- va in the Verdalen valley,80 km northeast of Trondheim and approximately 8 km from Trondheimsfjord (Fig ) ).The aquifer has been the subject of intensive studies during the last decade.The main purpose of the investigations was to study the quantity and quality of the available groundwater in order to evaluate the aquifer as a water resource for Verdal municipal ity.

Test-pumping of the aquifer was initiated in 1989, after detailed geological investigations. The results of these studies are described by Tvedten (1989). The yield of the established well-field satisfied the demand,but the observed increase in salinity as a function of with- draw al posed aquality problem.As a consequ- ence of this,the Geological Survay of Norway (NGU) decided to extend the test-pumping programme and estab lished a new well site in anattempt to lowe rthe salinity of the pumped ground water (Hilmo 1990).

Geological setting

Ice movements during the Quaternary have been reconstructed by Sveian (1989). During the ceqlaciation, which occurred during the latest Younger Dryas and early Preborealperi- ods (c. 10,500-9,800 years S.P.), halts or

Fig. 1.location of theSundby aquifer in Verdalen, Mid- Norway.

small advances of the ice-front resulted in several ice-marginal gravel deposits in Verda- len (Fig. 2).The locations of the deposits are strongly influenced by the underlying bedrock topography. The main drainage during degla- ciation followed the south side of the Verdalen valley.

(2)

28 Berntotev^Hi/mo^,^O/aM. Seettier & Sisse/Tvedten

j/

^'^km

I: ::~ ~ARINEDEPOSITS

PTI1 ^FLUV^IAl.^DEPOSITS

..~~.:~.~7f GLACIO FlUVIAlDEPOSITS

c:=J EXPOSED BEDROCKOR THINCOVEROF HUMUSON BEDROCK

D

ÂRÊAÔF^{Fig 3}

The ice-marginal sand and gravel deposi t whichconstitutes the Sundby aquifer does not extend up to the marine limit during deglaciati- on (c. 180 m above prese nt sea level). Later river terraces are cut into the depositatdiffe- rent levels,and the present level of the River Verdalselva is 5-8 m above sea level (Fig.3).

Itwasnotuntil c.1500 years ago that the sea had regressed to this level (Sveian 1989),but itis conceivablethat freshw ater under hydro- static pressure may have flowed through the aquifer earlier than 1500yrs. S.P.,displacing saline waters.

The structure of the Sundby aquifer is well know n,and is based on geolog icaldata from 36 bore holes,measurements of groundwater level in 15 observa tion wells, three seismic refraction profiles and electrical resist ivity measurements(lateral profilingand depth pro- filing). The aquifer is dista lly (Le.the western side) covered by clay, and in this part of the aquifer sand and gravel beds interfinger with marin e and glaciomarineclays (Fig.3).Upstre- am of the aquifer , one can observe exposed bedrock in the riverbank. In the investigated area,the aquifer is largely overlainby mudsli- de sediments. In a small area in the eastern part of the aquifer,as a result of postglacial eros ion,hydraulic communicationbetweenthe river and aquifer is possible. The top of the gravel deposit dips towards the north. How

NGU-BULL.422.1992

Fig.2.Quaternarygeolog ical map.showing Ice-marginal and glaclfluvial depos us Inthe lowerpart of Verdaten(afterSveian 1989).

far the aquifer extendsina northerly direction is uncertain.

The aquifer appears to be rechar ged by run-off from the hillside (from the south),and discharges into the river under non -pump ing conditions. Infiltrat ion from the river to the aquifermay be poss ible during floo dperiods, but only in a limited area by the southern river bank in the vicinity of observ ation well nO.3 (Fig. 3).

Thebedroc kin Verdalen consists ofPalaeo- zoic phyllites ,greywac ke,limestones and gre- enstones (Wolff 1979), and these litho logies arealso common in the overlying Quart ern ary sediment s. Occasionally one can find pebbles and clastsof biotite-ri chmicagneiss andgrani- te in the Quatern ary gravel deposits. These are derived from basement rock s located furt- her to the east. X-ray diffraction analyses of sediment samples show that the main mine- ralsinthegraveldepositsarequartz,plagioc la- se,mica,amphibo le,calciteand chlorite(Tved- ten 1989).

Test-pum ping procedure

Test pumping at Sundby was carried out during two periods, the first period (Aug.-Dec.

1989) utilizing the wells in well area Wl and the second period (Mar.-Jul. 1990)utilizing the wells in w~1I area W2 (see below). The total

(3)

Groundwater chemistryduringtest-pumping 29

NGU - BULL.422,1992

16m 6m

-t.m

-llom -c <,

-2lom

<:»

o

~I..:...<J

[O""'"YI

~

• •

MARINEDEPOSITS (SILTAND CLAYI FLUVIALDEPOSITS(SANDAND GRAVELI GLACIOFLUVIALDEPOSITS (SANDAND GRAVEL!

OBSERVATION WELL

WELLSITEAREA(TWO WELLSAT EACH SITE)

Fig.3,Block -diagram showing the stratigrap hyof the aquifer andthe locationof the well sites atSundby.

pumping rate varied between 25 and 50 LIs.

Before test-pumping commenced ,fifteen observation wells (diameter 5/4") were established in order to monitor changes in groundwater level during pumping.

Four groundwater abstraction boreholes were drilled.Two 4"diameter bore holes were esta blished at well area W1, with well- screen at 10-19 m depth. From one to four months after thestart of thetest-p um ping at well area W1, a consistent increase in the conte nt of totaldissolvedsolidsof thepumped groundwa - ter was noted (Tab. 1).

Later, in March 1990, two 3" boreholes were drilled furt her upstream atwell area W2 (Fig. 3), in order to increase the contribution from river-bed infiltration and thus to reduce the content of total dissolved solids .The well screen s here were place d between 8-18 m depth.

Groundwater quality monitoring

Groundwater samples for chemical analysis were collected from the observationwells pri- or to test -pumping. During the two periods

oftest-p ump ing,water was sampled from the abstraction boreholes initially once per day and later once every week. Electrical conductivity,alkalinity,pH and major ions in the water samples were analysed by standard procedu- res at the Geological Survey of Norway (0de-

gard & Andreassen 1987).

At 5 m below ground level in the vicin ity of well area W1,a marine clay was sampled and later subjected to pore-wate ranalysisand dete rmin ation of the adsorbed catio n content.

This was done to evaluate whether the rather high content of total dissolved solids in the grou ndwater could be dueto leachingof relict salts from marine clays. The content of ex- changable cations was determ ined by shaking 10 g clay in 50 ml 0.2 M BaCI,solution buffe- red with NH,CI to pH = 8.5, for 3 periods of 15 min. before final rins ing in 50 ml deionized wate r (Grim 1953 an d Hilmo 1989 ).

In addition to inorganic chemical analys is, water samp les were collected from W1 for determin ation of microbiological paramete rs. These analyses indicated that the biolog ica l quality of the wate r satisfies Nor wegian drin- king water stan dar ds (SIFF 1987).

(4)

30 BerntO/avHilmo,

ots

^M^.

ssnner

& Sisse /Tvedte n Afte r two months of pumping at W2 a 240 litre sample of groundwater was collected on May 23^r^{d ,} 1990,and analysed for its content

of "C and "C. The CO, gas generated from

the sample by adding acid was stored for six week sprior to analysis in order toavoid conta- mination from radon.

Results and discussion

Aquifer hydrogeology

Data from the test-pumping at W1 gave the the followin g hydrog eolo gical parameters:

Transmissitivity (T):0.058m'/s(5000m'/day) Specific storage(S) :0.15 - 0.2 in the eastern part of the aquifer and

<

0.1in the western part.

Hydraulic conductivity c. 3.8 x 1A·' rn/s (330 m/day)

Measure ments during tes t-pumping indicate that the groundwater level in the aquifer ad- justs quickly to water-l evel variation s in the river. During autumn 1989 net variation in the in river stage was measured to be 1.5m(Tved- ten 1989).The groundwater levelin observation well no. 6 (Fig. 3) was drawn down by only 0.55 m, after correction for river level variation.during three months of test- pumping from W1.

Ground water tempe ra ture

Thegroundwate rtemperature measur edinthe well area W1 was found to be stable at5.6°C during the four -mon th test-pumping period.

while the temp eratur e in the river varied from 16°C to 6°C.The annualmean airtemperatu- reis 4.6°C .Thegroundwatertemperaturewas measur edinseveralobservation wellsat different depths and at various distances (8, 38, 54 and 130m) from W1.There wasno sign ifi- cant change in temperature with depth in the aquife r,nor with distance from the mostlikely area ofriver water infiltration,during the pumping period.

NGU •BULL. 422.1992

Ground water chemistry

Groundwater in Norway'contains. in genera l.

low amounts of total dissolved solids .This is to be expected in a temperate climatewith a long winter seasonand a short residence time for most natural waters including groundwa- ters. Processes such as hydrol ysis. redo x- reactions . ion-exchange.dissolutionandpreci- pitat ion can however cause large variations in the chemistry of the groundwate r.The effe ct of the various reactions on the chemistry of thewaterintheaquifer dependsontemperat u- re. pH. Eh. residence time and mineralogical composition of the surrounding bedrock and sedime nts.

Ion concentrations during test-pumping of well areaW2 are considerably lower than tho- seof waters pumped from wellareaW1(Tab- le 1). The lower amounts of total dissolved solids in W2 may be due to a higher degree of infiltration from the river. pH-values of around 8 and alkalinities varying from 3.5 to 4.1mmolesrL in W1 and 2.0.to 3.0mrnoles/t, in W2 were recorded.

The electr ical conductivity and ion-content of the analysed samples are show n versus time of pumping in Fig. 4 and Table 1. The first samples of water collected at W1 and W2 are best charact erised as a calcium-bicarbonate type with moderate amo unts of to tal disso lved solids. After almost two months of pumpingat W1 thesalinityincreased drastical- ly.The increasesin chloride andsodium concentrations were most significant (Table 1).

After an interr uption of the pumping.the electrical conductivity is reduced (Fig. 4). This suggests that the aquifer receives a signifi- cantinflux of freshwaterwhenpumping stops.

Sources of salinity in groundwater

The low concentrations of Cl' and Na+at the start of the test- pum ping indicate that the ini- tialgroundwater composition isonlymarginal- ly influenced by marine waters. The drastic increase in salinity after a period of pumping . is presumably an indication of marineinfluen- ce.Topography.distancefromprese ntshoreli- ne (c. 8 km) and the Ghyben - Herzberg- equation,however, indicate thatdirect infiltration of saline fjord water is not likely to occur.

Possible sourcesfor the salinityin theground- water are thus assumed to be:

(5)

NGU"BULL.422, 1992 Groundwater chemistryduring test-pumping 31

Tables1: Chemica lana lyses ofgrou ndwa tersamples from (Table 1a)wellareaWl and(Table 1b) wellareaW2.Table 1c show sanalysesof por e wate r sq ueezed froma claysam pleand adsorbed catio nsin claysamples.",,"= not analysed. Tabl e 1a:Well area Wl

Date Electr ical pH Alkalinity F" C1- a- ^SO,= Na" K+ Mg'+ Ca'+ Fe Mn Cond.

pS/cm mmo llL rnq/], rnq/l, mg/L rnq/L mg/L mg/L mq/L rnq/l, mg/L rnqrl,

10/8 426 7.8 3.5 0.3 13 0.03 45 8 7 8 69 0.1 0.1

11/8 459 7.8 3.7 0.3 14 0.06 53 9 8 9 75 0.2 0.1

12/8 472 7.8 3.8 0.4 15 0.04 56 9 8 9 78 0.3 0.1

13/8 487 7.8 3.8 0.6 17 0.37 57 10 8 10 79 0.3 0.1

17/8 502 7.8 4.1 11 8 lQ 80 0.4 0.1

24/8 525 8.0 4.1 0.4 16 0.01 60

sto p

12/9 543 7.9 4.1 0.4 22 0.07 69 12 8 10 83 0.0 0.1

20/9 819 80 4.0 0.3 111 0.48 85 36 8 14 105 0.6 0.2

26/9 1114 7.7 3.9 <0.5 195 0.53 97 65 9 17 115 0.7 0.2

5/10 2220 8.0 3.7 <0.5 657 1.82 158 197 13 33 151 1.1 0.2

stop

12/10 1880 7.6 3.8 <0.5 526 1.31 149 167 11 27 129 0.9 0.2

27/11 4080 8.0 3.8 <2.5 1200 268 529 15 78 153 1.4 0.3

19/12 4500 7.6 3.7 <2.5 1300 300 603 17 85 136 0.0 0.2

Table 1b:Well area W2

Date Electri cal pH Alkalinity F" Cl"

ar

^SO,= ^Na+ ^K+ Mg'+ Ca'+ Fe Mn Cond.

pS/cm mmol/L mg/L mg/L mg/L mg/L mg/L rnq/L mg/L rnq/l, mg/L mg/L

713 255 8.0 2.3 0.3 4 0.02 19 8 2 5 40 0.2 0.1

9/3 270 8.0 2.4 0.4 5 0.02 23 10 2 5 40 0.1 0.1

22/3 394 8.0 2.6 0.5 27 0.10 34 31 3 7 42 0.3 0.1

28/3 518 7.9 2.8 0.6 30 0.17 46 44 3 9 46 0.2 0.1

6/4 619 7.9 .3.0 0.7 68 0.33 56 66 4 11 50 0.4 0.1

17/4 801 7.8 2.9 0.9 116 0.43 62 82 4 14 51 0.5 0.1

stop

8/6 896 8.3 2.9 0.6 276 0.74 90 91 4 18 55 0.0 0.1

14/6 1149 8.0 3.1 5.7 244 0.86 97 120 5 24 63 0.0 0.1

28/6 1260 7.8 3.0 271 1.05 111 137 4 27 63 0.3 0.1

sto p

3/7 871 7.7 3.0 0.3 158 0.55 75 84 4 18 55 0.1 0.1

sto p

17/7A 160 8.1 1.3 0.1 5 0.01 14 4 1 5 22 0.0 0.1

17/7B 300 8.0 2.0 0.2 17 0.07 39 16 2 8 34 0.0 0.1

27/7 488 7.9 2.8 0.2 40 0.16 61 33 3 12 48 0.1 0.1

1/8 667 7.8 2.9 0.2 102 0.34 68 58 3 16 50 0.1 0.1

Table lc:Marine Clay Sample

Sample F- CI- Br- SO,= Na+ K+ Mg'+ Ca'+ Fe Mn

Pore water squeezedfrom clay

(mg/L) 7.7 900 448 743 29 43 36 0.1

(rneq/t.) 0.4 25 9.3 32 0.7 3.5 1.8 0

Ads orbed cations

in meq/l00 9clay 1.2 0.9 1.8 6.9

(6)

- - - --'

/~

32 Bern t O/avHi/mo ,O/aM.Srether& Sisse /Tvedten

el. cond.(m S/cm)

5 ,--- - - - -

4

I

/

i /' ·/

p ump fa ilure

~

NGU-BULL.422.1992

.i>:

»>

p u mp fail ure

.- ~ I

, /

- -

.>

140

o '--- - - - --- --- ---- --- - -- -'--- -

o

²⁰ ⁴⁰ ⁶⁰ ⁸⁰ ¹⁰⁰ ^{12 0}

Days afte r st ar t o f pu mp ing

160

- . - We l l 1 - :- Well 2

Fig.4.Electricalconductivityplotted as afunction ofpumping-time(Piediagrams:ion-distribution atthe start of test-pump ing and at time of maximum totaldissolved solids ).

(a) relict ocean water within isolated sandl gravel layers.

(b) leaching of fossil salts from marine sedi- merits.primarily clays.

If we considerthe increase in the concentrati- onof the different ions andtheir ratios during test -pumpin g, it is poss ible to evaluate the causes of the increasing salinity. Fig. 5 & 6 show the ratios of Br-/GI-,SO,=ICI' Ca'+/CI' , Na+/CI-and Mg'+ICI-asafunction ofthe chlo- rideconcentrat ion.The resultsfrom bothwell- areas are presented.

The Br-ions and CI--ions both have amarine source.They are assumed to be products of weatherin g to a negligible degree. They are unlikely to be precipitated or adsorbed onto clay minerals, and one would thus expect a constant ratio between Br- and Cl' indepen- dent of salinity and Ct-concen trat ion (Fig. 5).

Despite the inaccuracy in the determination atlow Br-concentrations.this is in fact found to be the case;the ratio isappro ximate lycons- tant and similar to that in Mean Standard Ocean Water.

Sulphat e in groundwater originates

(a)from weatheringof sulphidesandsulphate- minerals

(b)from the marine environment

(c) from infiltration of anthrop oge nically conta- minated precipitation.

Sou rces (a)and (c)appearto be evidentfrom the high concentrationsof sulphate,andthere - forehigh SO,=/CI'-ratios,atlow concentrations of CI- (Fig. 5). In grou ndwat er with a higher salinity,the SO,=/CI--ratioapproac hes thevalue fou nd in Mean Standard Ocean Water.

The dissolution of aquifercalcite(Paleozoic limestone clasts and marine molluscs) is pro- bably the main source of the Ca> in the groundwate r.However,theCar-co ncentration increases with increasingsalinity (Table 1 and Fig.6)suggesting anadditionalmarinecontri- bution. All the groundwater samples that are analysed appear to be saturated or slightl y supers aturated with respect to calcite (appa- rent supe rsatu rat ion of the waters with respect to calcite may, however, be an artifact ofusing concentrations instead of ion activiti- es in the calculation).

(7)

NGU -BULL. 422, 1992 Groundwaterchemistry duringtest-pump ing 33

B(I Cr

10-3,--- - - - - - - - - - - - - - - - - ,

Na7CI-

3,5, -- - - - - - - - - - - - - - - - - --,

8-3

2,5

+

6- 3 +

8 1 0 12 14 18 18

ClIrneqv/I)

40

+ + + +

10

o^~-_ _---L. _ L . _'____ _.J

o +

1,5++

t

⁺ ⁺ ^Na'^{Cn n}^{standa rd}^sea^water

1

%- +~+ :;:--,- ++ -,---I '---1

0,5 + + +

20 + +

Br/ennstandard sea wat er

L

+ ++

8 + + ,+ + +

Fig,5,Ratiosof Br-/CI- andSO,=/CI-as afunction ofthe chlori deconcentration,Theresults frombothW1&W2 are presented,

In addition to their marine origin,sodium- and magnesium-ions are derived from weather ing of rocks and minerals. Tiius it might be expected that the Na+/CI-- and Mg'+ICI--ratios in the sampled groundwater would always remainhigher than theratios present in ocean water. In spite of this,the Na+ICI'- and Mg'+1 CI- - ratios in groundwater became less than the same ratios in ocean water as pumping progressed and salinity increased (Fig. 6).

This isinterpreted as being due to preferenti- al adsorption and ion- exchange of Na+and Mg'+on clay minerals.

Clay minerals have a net negative charge.

Cationic attraction to the basal planes of clay particlesincreases with increasingvalence and decreasing hydrated ionic radii.This leads to an enric hment of polyvalent cations in adsorbed positions relativeto the ionic composi tion in theliquid phase(VanOlphen,1977),Becau- se of repulsion from the clays' basal planes and weak or no attraction to theedges,mono-

40 40

30 20

Cnmeqvll)

+ +

r :

^sta^ndard^sea₊ ^water₊

+

Fig,6,Ratiosof Na+/CI-,Mg"/CI-andCa'+/CI-asafuncti- onofthe chlorideconce ntra tion,Theresultsfromboth W1

&W2arepresented,

valentanions will leachmore easilyfrom marine clay than catio ns and especially polyvalent cations.Water leached from marin e clays will therefo re have lower Na - ICI-and Mg'+ICI- ratios than ocean water.The pore waterche- mistryandthemeasured compositionof adsorbed cationsforthe marineclay sampledin the vicinity of W1 supports this hypoth esis (Ta b.

1),An elevated concentra tio nofNa-inrelation to CI- is found in the pore wate r squeezed from the clay samples , This is also foun d in

Mg'7CI- 4

3;-

J

...

' /

+"!j.

4

0

0 10

Ca'!CI- 4

+ 3

t-

+ + +

+ ++

+ +-jc

0

0 10

40

+ + 30

,--

S04/C1instandardseawater

l

20 Cl"(meqvll) +

10 1

...

+

1,5

(8)

34 BerntOtev Hllmo,O/aM.senner& Sisse/Tvedten the pore waterof marinesensitiveclays sampled in other localities in Trondelag and North Norwa y (Hilmo 1989).This pres sure -extracted pore water of the marine clay sample isbelie- vedto reflect the remnantconcentrationof the ions after at least 2000 year s of fresh water leaching. In contrast to the ionic comp o sition of adsorbed cations, the concentr ation of Mg'+ relative to Na+ and Cl' is low er in the por e water, compared to ocean water. The net eff ect appears to be that water leached from a marine clay with a fossil salt content isimpoverishedin cationsin comparisonto Cl'.

When abstraction of qroundwxter exce eds the amount which can readily be infiltrated from the surface,anincreasing proportion of the groundwater will originate in sand- and gravel deposits which are in close contact with the marine clays.Mapping of the aquifer revealed several thin sand- and gravel deposits which inter finger with marine clays dista l- Iy (Fig.3).These sediments are pres umed to act asdrainageavenues for theleachedpore- waters.

Carbon isotopes

Analyses of the car bon isotopesinwater from well W2 gave the following result s:

o"C = -12.2'/",

"C = 12.72

=

^0.10 ^cp^m, ^correspond^ing ^to

60.2 z; 0.5% activity,andanapparent(assu- ming all carbon is biogenic) age of 4290 :t 55 years (Srether 1990).

The water sample used for the dating had analkalinity of 2.84rnrnoles/l. ,It was supe rsa - turate d with respect to calcite. The alkalinity was higher than in seawate r (2.3 mmoI/L).By comparing the conce ntrations ofchlor ide and bromide in the groundwater to those of seawa-

ter, a 'dilution factor' of about one hundred

times is estimate d.The contribution of bicarbonate from seawater canno t thus be more than about 0.02 mrnoles/L.i.e. less than one perce nt of the total bicarbonate.

Typically, meteoric water has a o"C of around -7"1" ,.As it percolate s into the gro und, however, it acq uires biogenica lly derived CO, from the soil zone,with a O"C of -25"/,,,.Mari- ne carbonates typically have a o"C value of +1

't:

(L1oyd&Heathcote 1985).Theo"C value obtained for the groundwater is thus typical

NGU·BULL.422.1992

forthestandard modelof bicarbonategenerat i- on in groundwater,where 50% ofbicarbo nate carbonis bioge nic,and 50% is fro m dissolved carb onate, leading to a typical OIlC value of -12 to -13'/00'

COl+HlO+CaCO,=Cal++2HCO,' OIlC -25°/ +1°/ • -12'/, If one also assumes that 50% of the "C is bioge nic,and 50% is inorganic (wi th "C

=

0,) the age of the biogenic carbo n (the age of recharge) is calculated as -1540 year s. One can also apply a mor ecomplicated correction metho d such as that of Wigley (1971),which yields an age of between -1800 and -2200 years.Such negativ e ages areplainlymeaning- less,and this may be due to:

(a) The assumption that the "Ccontent ofthe aquifer carbonateiszerobeing incorrect.This is the mo st likely reason. If the aquifer carbonate contains young carbo nate material ("C>O),the above assumpt ion would lead to over co rrection of the age. In fact, Tved ten (1989) foundthat theaquifermaterialcon tains up to 10% calcite. This calcite is composed both of c1asls of Paleozoic bedrock and of recent marine molluscs (Le. finite "c).

(b)Carbo n from humicacidsinthe groundwater affecting the results.

(c) Exchange of groundwa ter and aquifer carbonate,a pheno meneo n noted in saline envi- ronm ents by Wendt (1971).

Insumma ry,thecarbonisotopes give no conc- lusive information on the water's age. The obtained 'negative' dates are not. however, incon sistent with a modern rech arg e water which has been overco rrected for its "C age (due to the finite "C content of the aquifer carbonate).

Conc lusion

Test-pumping of the SundbyaquiferinVerda- len, Mid-Norway, yielded groundwater with a highe r con tent of total dissolved solids than is acceptable for domestic water(SIFF,1987).

Dur ing the first week s of the pumping period the groundwater could be characterised as a

(9)

NGU-BULL.422.1992

calcium bicarbonatewater with rnocerate total dissolvedsolids.After more than threemonths of test-pump ing at W1,thesalinityhadincrea- sed to approximately 2.5 g/L. The maximum salinity during test-pumping at W2 was 0.7 g/L. The ratios of Br-/CI- and SO,=/CI' in the saline water samples are close to those found in Standard Mean Ocean Water,and indicate a marine origin for the salinity.

Although the well screens are situated under sea level, it is regarded as unlikely that ocean water can intrude the 8 km from the fjord through thickclay sediments during pumping.The high content of dissolvedsolids may therefore be due to relict saline water from isolated sand and gravel layers or be leached from marine clays.

The Mg"/CI- and particularly Na'/CI' ratios in the saline groundw ater are lower than the corresponding ratios in ocean water.This was rather unexpected, as additional Na- and Mg" ions might be expected to originate from weathering. It is, however, believed that cations such that Na- and Mg'. are being adsorbed onto clay particles. Water leached from marineclay willthus haveahigher concentrati- on of CI- compared..to Na+and.Mg" than ocean water.Analysis has confirmed that the pore water of a clay" sampled in'the vicinity of W1 is enriched in Na- ions compared to CI- ions, and that adsorbed cations are enriched in Mg" and Ca- compa red with Na-.

Thin layers of sand and gravel interfingering with clay deposits have been mapped in the distal part of the aquifer.These layers act as drainage avenues for the leached pore water . which becomes anincreasingly important com- ponent of the abstractedgroundwater withtime of pumping.

Groundwater chemistryduring test-pumping 35

References

Gnrn,R.E. 1953:Clay minera logyMcGraw-Hill.New Yorkl London/Toronto.368pp.

Hlimo.B.O.1989:Msrinesensiti ve leirers mineralsammen- setning.kolloidk jemiogmeksniske eqenskep er.Thesis (Doctorof Eng ineering)no.1989:2 1. Institutl forGeolo- giogBergteknikk.NTH.Trondheim .Norway.

L1oyd.JW.& Heathcot e.J.A.1985:Naturalinorg anic hydro- chemistryin relat iontoqroundweter.CtarendonPress.

Oxfo rd.U.K..296 pp.

SIFF 1987:Kvalitetsnorm erfor drikkevann.steteosInstitutt for Fotkenetse (SIFF). report 'Drikkevenn G2·. ISBN 82-7364-013-2.72 pp.'

Sveian. H. 1989: Stikles tad. Quatern ar y geological map 1722 IV 1:50 000 withdescription.Nor.q eot.unde rs.

Sret her.O.M .1990:Aldersdatering og kjemisk analyse av grunnvann fra Sundby.Verdal kommune.NGUintern al report90.015.

Tvedten.S.1989:Provepumpingavgrunnvannsbronner ved Sun~byiverdst.Thesis (hovedoppgavejin engineer ing geology. NorgesTekniskeHogskole.Trondheirn,75 pp.

Van Olphe n.H.1977:Anint ro ductio ntoclaycolloidchemis- try.2nd edition.Wiley-Inter science.318pp.

Wendt. I.197 1:Carbon and oxygen isotopeexchange betwe- enHCO.in saline solution and solidCaCO..Earth and plan.sci.le tt.. 12.439-442.

W'gley.T.M.L. 1976:Effectof mineralprecipitationonisoto- piccompo sition and"C datingof groundwater.Nature.

263.16/9176.219-221.

Wolff.F.Chr.1979:Beskrivelse tilde berggrunnsgeologis- ke kart Trondheim og 0stersund. Scale 1:250 000.

Nor.geol.unaers.353(Skri fter31).

0degard. M.&Andreassen,B. 1987:Metho ds for water analys isat the GeologicalSurvey of Norway.In:Geome- dicstConsequencesof Chemica lCompositio n ofFresh- water.The Norweg ian Academy of Science and Let- ters.133-150.

Manuscript received February 1992;revised typescript accepted June 1992.