Hydrogeology in the Service of Man, Mémoires of the 18th Congress of the International Association of Hydrogeologists, Cambridge, 1985.
SOME CONSIDERATIONS ON GROUND WATER DATING USING ENVIRONMENTAL ISOTOPES.
J.Ch. FONTES
Laboratoire d"HYDROLOGIE et de 6E0CHIMIE ISOTOPIQUE Université de Paris-Sud, E-9I405 ORSAY Cedex.
ABSTRACT : Some aspects of the use of tritium, carbon 14 and chlorine 36 for ground water dating are discussed with special reference to possible future development of these methods. Main principles of tritium modelling and various processes of estimation of ,4C initial activity are reviewed. Some examples of "*C1 variations in old and recent ground waters illustrate the considerable potentialities of this new technique.
1. INTRODUCTION: Time, or velocity are the least easy to estimate among hydrogeological variables or parameters. Requirements of Darcy's law applicability and especially the condition of steady state flow, put severe limitations on any transit time evaluation. This is especially the case in confined aquifers where successive inputs have varied according to long-term climatic fluctuations. This and further weaknesses of Darcy's law (extrapolation of local measurements of permeability, estimation or "guesstimation" of effective porosity and tortuosity, requirement of laminar flow) have induced considerable efforts to develop independent methods of time evaluation. Environmental isotopes provide many potential chronometers for ground waters studies (cf. table 1). However, because of either practical or theoretical limitations, suitable hydrochronometers are essentially : - tritium - carbon 14 - chlorine 36 whereas other method suffer problems of : sampling difficulties (krypton 85, argon 39), non conservative tracers (tritium-helium 3, silicon 32, uranium 234 - uranium 238, isotope 2 equilibration rate S04 ~-H20, helium) development of proper analytical techniques ( krypton 81 ).
2.TRIT1UM : Tritium has been for many years the most popular environmental isotope in ground water studies. This was certainly due to the simplicity of the interpretations in some favourable cases.
2.1. Sources : Tritium is produced by the action of neutrons on nitrogen atoms: ,4N + 'n = 3H + 12C. It decays into ^He with a half-life of 12.43 a.
2.1.1. Natural production : The production is induced by secondary neutrons from the cosmic radiations. A production
- 118 - Isotopes Half-life (a) Origin Time range Qualities Limitations or process (a)
85Kr 10.8 Nuclear reactors since 1960 No interactions Sampling, counting
3H 12.43 Cosmic rays since 1952 Ideal behaviour Various sources Thermonuclear tests commonly suitable Reactors
3H-3He 12.43 Cosmic rays since 1952 Direct determination Analytical - Crustal Thermonuclear tests and possibly of turnover time production of ^He Reactors 100
32 Si « 100 Cosmic rays « 100 (?) Few, link between A0 (?) - Samples Nuclear tests 3H, 39Ar, 14C (?) 10m , counting time Crustal (?)
39Ar 269 Cosmic rays « 2000 BP No interactions, sample 10m3 Crustal comparison with '\ counting time
'\ 5730 Cosmic rays, «3.104 In all ground Complicated chemical Thermonuclear tests *7.104 waters and isotopic systems Crustal
*kr 2.!xt05 Cosmic rays 5x105(?) No interactions, Analytical ideal for very old waters
36c, 5 6 3.06x10 Cosmic rays 2x10 Hydrophilous, AQ (?), various Nuclear tests time range sources, acess to Crustal accelerator
234u_ 5 5 2.5x10 Decay of uranides 5x10 (?) Time range of very AQ (?) chemical 238u chain old waters interactions
!80 Reaction Stable 105 (?) Additional tool Calibration (?), 2 (so4 " rate commonly suitable redox and pH
-H20) dependent
4He Accumu- Stable « 106 Additional tool Not conservative mutation rate
Tableau 1 : Theoretically available environmental isotopes for ground water dating.
- 119 - rate of 0.5 ± 0.3 nuclei per cm2 of earth surface per second (Nir et al, 1966 in Gat, 1980) or 1.5 x !017Bq.a~' (Eisenbud et al, 1978) was estimated.
3 After formation, H is oxidized into H20 and is removed from the troposphere by precipitation. At steady state (production rate equals removal rate) the total amount of natural tritium in the atmosphere correspond to about 2.6 x 1018 Bq (Eisenbud et al, 1978).
2.1.2. Thermonuclear detonations : Since the first test (31st October 1952) large amounts of 3H were introduced into the stratosphere. Global inventory of tritium from weapon tests reached 1.15 x 1020 Bq in 1963. The moratorium on aerial thermonuclear tests in 1963 stopped almost completely (further aerial tests by China, France and India did not contribute significantly to the tritium budget), the supply of weapon bomb in the atmospheric reservoir. The radioactive decay will decrease this figure to the natural level approximately by the year 2030 (Eisenbud et al, 1979)..
2.1.3. Nuclear power reactors : This supply is becoming the major source of 3H. Tritium is produced by ternary fission and by activation of some light elements (B, Li) which are used in reactors, it has been estimated that 3H from nuclear power reactors will become prédominent over bomb tritium in the atmosphere in 1985 (Eisenbud et al, 1978). Losses from tritium industry (watches, signals..) contribute also significantly to the increase in atmospheric 3H.
2.2. Units : Radioactivity units are not frequently used in environmental tritium studies. The very small amounts of 3H present in natural waters led to the definition of a tiny reference : the "Tritium Unit" (TU). Its corresponds to an atomic ratio of one atom of 3H for 1018 atoms of ^H. Strictly speaking the Tritium Unit should be referred to as the Tritium Ratio (TR). The corresponding radioactivity in water is0.118 Bq.kg"' (or 3.19 pCi.kg~',or 7.2dpm.kg~Mn the previously used units).
2.3. Distribution patterns : Both natural and artificial tritium are inequally distributed at earth's surface.
2.3.1. Cosmic tritium : Natural tritium is produced in major quantities at high latitudes. This is due to the minor magnetic deviation of charged cosmic particles (alpha and protons) which produce secondary neutrons by interaction with the first air molecules in the high atmosphere. Natural fall-out in precipitation is evaluated between 25 to 20 TU at high latitudes and A TU in equatorial regions (see 6at, 1980).
2.3.2. Bomb tritium .- 2.3.2.1. Global variations : The thermal convection which occurs during aerial tests brings a large fraction or radioactive nuclides and debris up to the stratosphere. There, the laminar structure of air masses and the monodirectional winds induce a general motion of rotation at the latitude of injection. The largest aerial tests were performed at mean latitudes in the Northern hemisphere. This explains the heterogeneity of 3H distribution in latitudes and the rather low tritium activity of rains in the austral hemisphere (Fig. 1 ).
- 120 - Northern 2000 j- HemisDhere [TU] [
1500!- !-
t-
1000^ a tf 500^ U/'^v' ww V VVsyHl
r 200^ Tropics au] -
150^
100- S i •« 50h •A JWsAib ^ Southern 100 Hemisphere [TU] r 751- I h f 50r iii r s 25r- ! it M Nmutiw ^
1955 60 65 70 1975
Fig. 1. Time and space distribution of bomb tritium. From Groeneveld ( 1977) in Gat ( 1980).
- 121 - 2.3.2.2. Seasonal variations : The mechanisms of 3H fall-out from the stratosphere into the troposhere, where its residence time is short (21-40 days according to Eisenbud et al, 1978 ; 5 to 20 days according to Gat, 1980) is not completely understood. Several processes are possible (diffusion through the tropopause, sowing of convective clouds which penetrate into the stratosphere), but the main transfer probably occurs through tropopause discontinuities or extrusions of the tropopause. These discontinuity zones appear seasonnaly when polar and equatorial tropopauses begin their opposite migration toward pole and equator respectively, at the end of winter. The result is a "spring peak" which is approximately 3 times the average annual weighted activity. Then tropopauses move back at the end of fall and give rise to a "winter valley" with half of the annual average activity.
2.4 Sinks : Main sinks for tritium are residence time and related radioactive decay in soils, in aquifers and mainly in the "mixed layers" of ocean waters which corresponds to a thickness of about 100 m of surface waters. The relatively low renewal of these waters involves old (tritium free) bottom waters as well as supplies from the continents. The result is a very low tritium content of ocean surface waters. Through molecular exchange at the interface and also by mixing with oceanic vapour, the tritium content of the atmospheric moisture decreases as long as the depressions travel over ocean surfaces. Precipitations over oceanic islands and coastal regions are thus depleted in tritium as compared to continental rains. During its transit over the continents the initial marine vapour mixes gradually with increasing amounts of vapour generated by evaporation from soils and lakes and by évapotranspiration. This continental vapour corresponds to rapidly recycled precipitations which thus contain tritium. A continental effect occurs which relates the increase in 3H of meteoric water to the distance to the coastline along pathes of atmospheric moisture masses ( Fig.2).
2.5. Reference stations : Despite of the numerous, above discussed, causes of variations tritium distribution in precipitation is not erratic at the scale of the hemisphere. Weighted monthy averages are closely correlated (Fig.3). Therefore, any partial record at a given location may be reconstructed with a reasonable confidence by comparison with a reference stations where a long record is available. Available reference station and related results are found in the AIEA-WMOnetwork (AIEA, 1969, 1970, 1971, 1973, 1975, 1979). A special reference station is Ottawa because it provides the only records starting from the beginning of the fifties i.e. before the first thermonuclear tests.
2.6. The use of tritium for ground water studies : 2.6. /. Qualitative : Tritium is often used as a guide to evaluate ground water residence times. Last pre-bomb precipitation occured in 1951-52 with a 3H content of 4 to 20-25 TU. In 1985 this activity range has decayed down to 0.6 to 3.7 TU. The practical rule to identify ground waters older than 40 years at the end of the eighties will be: - high and medium continental latitudes : < 3 TU - mid coastal and low latitudes : < 0.5 TU.
Until the end of the 60's it was possible to distinguish recent (tritium rich) waters and mixtures between recent and old ground waters ( intermediate tritium content). An empirical limit of about 40 TU was admitted below which the presence of an old component was highly - 122 - ro o i o o C\J O
E
CO o r- o to m m *" 00 ro 7~
o CD
O tf> m [/-• / o : - - - \ O rO O J o m _J •• ~""L °;- > "" < or ,J '-/ .;--;• / rO,' o ^ '•,?'-; ? Ft- •* \ 5 • '- •-/ C\J o •^ /^ 5 _i. O «i ro *3> - 123 - probable (at mid and low latitudes from the Northern hemisphere). This is not so simple in the 80"s. Due to the decrease in 3H which was marked until early 70's, recent recharge may contain less 3H than waters precipitated some years before. Furthermore, mixing even including waters without 3H, may have a higher tritium content than present day recharge, provided it includes a significant contribute from the 60's. For instance, in the Republic of Djibouti (East Africa) recent recharge from 1974-1976 had a 3H content of about 9 TU .waters older by about 10 years contained more than 40 TU and mixtures between pre-bomb and recent waters could exhibit any 3H content below 40 TU ( Fontes et al, 1980). In 1984, the recent recharge had a 3H content of about 4 to 5 TU and the criteria were almost reversed since mixtures including waters from the 60's had 3H contents of about 20 TU. Care should thus be taken in the interpretation, even qualitative, of 3H data. An accurate estimation of the local 3H content in the recharge is necessary with special reference to the past few years. In industrialized regions the picture is still more complex owing to the increase in 3H supply from nuclear reactors and tritium industries (Weiss et al, 1978). 2.6.2. Quantitative : 2.6.2.1. Basic treatments : The extreme possible behaviours of the successive tracer inputs in an aquifer are described by two models of flow which are of general validity for any kind of tracer, environmental or artificial (Yurtsever, 1983) : - Pure piston flow model (PFM) : Successive contributes are stratified upon the previous ones and do not mix between each other. Each input undergoes a radioactive decay of 5.5 % per year. In this case, the turn-over time t0 equal to the transit time T or the residence time t of the tracer in the system with : Xt CD(t) = CR(t-t) e~ M where C stands for concentration, R and D refer to discharge and recharge respectively, X is the decay constant. - Pure mixing model (PMM) : Each episode of recharge with a tracer concentration CR mixes instantaneous and completely in the reservoir (volume V) already perfectly homogeneized. The discharge has a tracer content CD which is at any time an aliquot of the reservoir. However, since in the case of bomb tritium input concen- trations have varied in time, this model must be considered as a recursive equation : k t t (CD)l = q(CR)t+(l-q)(CD)l.,e- I -( -')] <2) where q is the fraction of the reservoir which is recharged ( and d i scharged) bet ween t-i and t, q = V/tQ with tQ : means transit time = turn over time, k is the constant expressed in appropriate units (reciprocal of years or months) which accounts for the radioactive decay for time interval t -1 to t (k equals X and e~^ = 0.005 if the time interval is the year). - 124 - TU * T " T i" ' 'i1 t r'rrt - t i i T *""? r"ry % f'"r ' ' y~f. : i s y, • / y / HDF y // '/,' eee. //'/• X ' • X / p '/ / • /•/ / • s/:/ • • tee. /// . .y// \ \ V/J/ • ' / • VALENT IB -> t * 1 à * t t t 1 A 1 1 1 1 1 1 1 \ A 1 1 o-i- lee leee TU Fig.3. Monthly tritium contents of precipitation at two stations (Hof. PRO, 50°N. !2°E and Yalentia Ireland, 52°N, lO'W). Slope of the correlation a 2.74 standard deviation (dotted lines.) a ±40it From Weiss et al ( 1978). 125 Solutions are of the form : (3) t,Cn(t)= I CR(t_ ^e-^V^ t = 1 with t = residence time i.e. the time elapsed since an input of tracer has entered the system. However, most of the actual flows are probably obeying to models which are in between Piston Flow Model and Pure Mixing Model. These indermediate models are the Dispersive Models. The following differential form which accounts for tracer balance between stages n-l and n: can be applied to dispersive systems assuming steady-state flow conditions through a cascade of n cells of volume V with X = nV/Q and a unique input of tracer in the first cell. Solutions are of the form : 1 (CD)n = {(CR)0 [(Q/vnF e-^/^Mn-1 )!]-' (5) th where (CD) is the discharge concentration of cell n , (CR)0 is the initial recharge concentration at the 13* entry of cell. Normalized response-time curves can be analytically derived as (Przewlocki & Yurtsever, 1974 ;Yurtsever, 1983): C(t) = [nn.tn-1.e-nt][(n-l)!]-1 (6) where t=t/tQand tQ = nV/Q. Various C( t) curves for different values of n are shown on figure 4. It is remarkable that if n = 1 (one cell) the model becomes a Pure Mixing Model and it can be demonstrated that if n -» co the model becomes a Piston Flow Model. 2.6.2.2. Case studies .- The Pure Mixing Model and the Dispersive Model have been used for the interpretation of tritium data from shallow ground water systems. In all the applications the best possible fit is searched out by adjustment of the value of X, However two critical points must be kept in mind. First, the input concentration is not the weighted mean tritium concentration in precipitation but in the water which infiltrates i.e. the monthly balance between precipitation and évapotranspiration, and second, it is assumed in all the models that no delay and no mixing are induced by the transit of water in the unsaturated zone, which is obviously a crude assumption. - 126 - > Q AD E o in > < c f*> CM U) 0 : U)d - 127 - PMM gave satisfactory results ( Fig.5) for the evaluation of turn-over time ( 15 to 25 a) of karstic springs in Southern Turkey ( Yurtsever, 1979). In the small basin (0.4 km2) of Grafendorf in Austria tritium data for the period 1969-1973 were processed with an input function calibrated by comparison with Ottawa record (Przewlocki & Yurtsever, 1974). Computed évapotranspiration (Thornwaite) allowed to weight the monthly input function. A six cells model was used with a larger cell at the input (2.5 relative volume unit : r.v.u.) followed by 5 identical (0.7 relative volume unit) cells for a total volume of 6 r.v.u. The best fit was obtained with an annual discharge of 0.8 r.v.u. The turn-over time is then : t0 = (S Vj)/Q = 6/0.8 = 7.5 a. The size of the aquifer 3 3 5 ! is thus V = Qt0 deduced from the average discharge 0 = 2.75 10~ m. s~' =8.7x 10' m°a" which multiplied by 7.5 gives a total volume of 6.5 x 106m3. The discharge of the aquifer of the Oloma River in Norway was treated by Yurtsever & Payne ( 1978) with a multi-cell model and adding the simple condition that discharge from each cell is at any time proportional to the volume of the cell. Best fitting curve is shown of Figure 6. The adjustment, leads to a value of 10.2 a for the mean transit time (turn-over time). 2.7. Conclusion on the use of tritium in ground water studies : Bomb tritium concentrations in the high atmosphere are continuously decreasing and become hardly distinguishable from natural levels especially at low latitudes. At mid latitudes, the level of bomb tritium in precipitation becomes lower than that of tritium released by tritium industries and reactors However, because injections from this latter source are restricted to the troposphere, their distributions remain local or regional. It may thus be advanced that tritium pictures could lose most of its value as a global tracer at mid latitudes whilst it could become harder to measure at low latitudes where the most acute problems of ground waters resources have to be tackled. These somewhat pessimistic conclusions on the future of tritium must be attenuated by the development of comparisons with HC data, by the appearance of new potential indicators of short residence time like 85Kr or 36C1 and by the perspectives of the use of 3H-3He chronometer for shallow ground waters studies. However, before tritium patterns will become more confused through stratospheric losses, decay and mixing with local sources, it may be time to conceive large banks of monitored samples for some selected shallow aquifers of high economical or theoretical importance. - 128 - 100(^ 800 600 Computed tritium 400" output curves - 200 Z 100 Observed o tritium output O 80 curves 5 K = 50 years => 60 O o_ cc 40 O Dumanli spring 20 A M—1 spring a M-15 spring » Monavçat river (Baseflow) 10 1967 1968 1969 1970 1971 1972 YEARS Fig.5. Tritium content and outpout curves calculated from equation (3) for various values of turn-over time in some karstic systems from Turkey. Prom Yurtsever (1979) in Yurtsever (1983). - 129 - CO CE < LU > CM CO Z et LU 2 LU CD OC c '=> ^_ 03 CO r-. < O LU S _i - .ti co < •»••' 0"> 3 h- O "S3 CO t_ CD 19 7 > < CD CO >_ CD ^ XD c 'o ^ Ci r- fa — 3: CD c_ è- 00 CD "O (O •S- co 05 «- t_ CO CD P > S S s r CD => CO O) xz >- o o 8 E Ê = (O •^ E o> cm) 1N31N03 WnillUl 130 5. CARBON 14. 3 I. Production .- 3.1. I. Cosmic : Carbon 14, the only radioactive isotope of carbon is produced by cosmic rays according to the reaction ,4N (n, p) !4C. Through (i" emission, radiocarbon decreases back into !4N with a period of 57301 30 a. 3.1.2. Nuclear tests : Radiocarbon was.also produced through (n,p) reactions during aerial tests of nuclear devices. As for tritium the excess of l4C in the atmosphere reached a maximum in 1963 when it was approximately equal to the natural steady-state level, and decreased thereafter. Carbon dioxide is more readily homogeneized than water vapour in the atmosphere. However, ,4 latitudinal variations are still readible in bomb C02 distribution but until the end of the 60's (Fig.7). 3.2. Standard for reporting activities : The reference for ]AC measurement is the so-called "modern carbon" which is supposed to 14 account for the C activity of atmospheric C02 at the end of the XIX century (i.e. before any significant dilution by "dead" C02 from fossil fuel oxidation and before contamination by radiocarbon produced during nuclear tests). The "modern carbon" was defined, by reference to a wood grown in the years 1880, as 95^ of the sepcific activity of an oxalic acid prepared by the "National Bureau of Standards", Washington D.C., USA. The activity of the modern carbon was evaluated at 13.56 ± 0.07 dpm g-' of carbon (0.226 Bq.g-1). Radiocarbon is measured either in gaseous form (C02, CH4 and sometimes C2 H2, C2 H6) on proportional counters or in liquid form (C$ Hg) in liquid scientillation spectrometry, or in solid form (graphite) in Tandem Accelerator Mass Spectrometry. Accuracies reach ± 0.1, 0.5 and 1 % respectively. 3.3. Penetration of l4C into ground water cycles and origin of the dissolveo carbon : Carbon 14 is introduced from the atmosphere where it is formed into ground water mainly through respiration of terrestrial végétais and decay of organic matter in soils. This C02 is then uptaken by the chain of carbonate equilibria : C02 (gas) + H20 = C02 ( aqueous) ( 7 ) C02 (aqueous) = H2C03 (8) + (9) H2C03 =HC03- + H 2 + (,o) HCO3- =C03 " + H 2 2+ C03 + Me = MeC03 (solid) (li) where Me2+ is generally Ca2+. Seepage from lakes where the same chain of equilibria takes place under the influence of environmental C02 (atmospheric, biogenic) represent another way for the introduction of radiocarbon into ground waters. The Total Dissolved Inorganic Carbon (TDIC) from ground water may also be generated through more complex reactions of rocks dissolution involving organic acid or C02 derived from vegetal activity : - 131 - <—ai e m r- n <+• C • ^ • -— r- C""-) f l'O V - 132 - 2+ R - COOH + Me C03 = H C03" + Me + R COO" (12) 2+ CaAl2Si208 + 6H20 + 2H2C03 = Ca + 2A1(0H)3 + 2H4Si04 + 2HC03~ (13) + + Na (or K) AlSi308 + ?H2 + H2C03 = Na (or K )+Ai(0H)3 + 3H4Si04 +HC03" (14) All these processes must be divided into those in which all the resulting TDIC (mainly HCO3" at pH between 7 and 8) is derived from atmospheric carbon via biogenic processes and those in which an extra source of carbon (solid carbonate from soil or aquifer matrix) is involved. In the first case the ,4C content of the TDIC will reflect that of the atmosphere, in the second case the 14C content of the TDIC is diluted by a possibly old carbon. Incongruent dissolution of dolomite in a ground water saturated with respect to calcite and under saturated with respect to dolomite, will also produce a chemical dilution of the 14C in the TDIC despite alkalinity remains constant. A slight dissolution of gypsum would favour this incongruent dissolution of dolomite because the increase in Ca2"1" will lead to calcite saturation. Base exchange on clay minerals is another source of chemical dilution since Ca2+ and Mg2+ are entrapped on the solid phase. The solution thus becomes under saturated with respect to calcite and dolomite and can dissolve a further amount of these minerals. The chemical dilution of 14C which occurs in that case, can be detected by the increase in alkalinity. Furthermore, it must be noted that these processes of chemical dilution are not necessarily 2 linked to the local occurence of carbonate layers. Reaction ( 13) shows that Ca * and HC03" and thus that calcite precipitation may occur in regions of crystalline rocks. Therefore, a chemical dilution of ,4C in the TDIC may occur if the calcite is sufficiently old for its own ,4C to be decayed. Such a case was observed in shallow (recent) waters from the Canadian shield (Fritz et al, 1978). J. 4 Isotope effects of the TDIC : 2- Each of the carbon bearing species (C&j (gas) H2C03 HCOj", COj carbonate (solid), may isotopically react with any of the other one/These reactions are easily evidenced by the variation in stable isotope content ( ,3C) which can be measured with a much higher accuracy (at least 10 times better) than ,4C variations. Stable isotope contents of carbon species are expressed in delta (&) units or per mil versus the usual PDB standard : b = [R(sample)/R(standard)] - 1, generally written as '/.. i.e. multiplied by 103, with R=,3C/,2C. The stable isotope fractionation between two compounds 1 and 2 is defined as : a = R,/R2 = (1 + &,)/(! + b2) (15) or as: S =(ct - 1)« In ct (16) generally expressed in %. i.e. multiplied by 103. The fractionation obeys to an similar definition for 14C with : H 14 a C = A,/A2and e c = (A,/A2-l) (17) where A is homologous to 14C/( 12C + 13C). - 133 - The most Isotope partionning reactions are : C02(gas) - H2C03 , HC03" - calcite, C02 (gas) - HC03~ with the following isotope effects for carbon 13 ( !iook, 1980) : 3 E[H2C03 - C02(gas)]£ag « o(H2C03) - Ô(C02 gas) « [-(0.73/T) + 0.9110 ~ (18) 3 S[caicite-HC03"]= Ss « è(cakite) - 5(HC03~) * [-(4.232/T) + 15.1110 " (19) 3 £[C02gas-HC03"]= S » Ô(C02gas) - Ô(HC03") * [-(9483/T) + 23.89J10 ~ (20) and where T is the temperature in Kelvin which are also generally expressed in */.». For thermodynamics! reasons the isotope effect (or ' isotope fractionation) is approximately 2.3 times greater for 14C than for 13C (Saliege & Fontes, 1983) : £14C»2.3£13C (21) 14 For instance : A (C02gas)/(A(HC03~) = ct C introducing eq.( 17) gives : A (HC03~) = ,4 ,4 A (C02 gas)/a C « A (C02gas)/( 1 + £ C) with the approximation : ( ! + £)" ' ~ 1 - £ it comes : ,4 13 A(HC03~) »A(C02gas)(l-£ C) *A(C02gas)(1 -2.3 £ 0 ,3 For A (C02gas) = 1,00 (= 100!? of modern carbon) at 25°C eq.(20) gives e C = -0.00792 and finally it comes that the activity of the bicarbonate A(HC03~) in equilibrium with atmospheric C02 (before nuclear tests) is enriched with respect to it by 0.018 i.e. ].%% . 5.5. Estimation of the initial MC content of the TDIC : The basic problem in radiocarbon "age" determinations is the evaluation of the initial activity. This initial activity is defined as the 14C content of the TDIC after every processes of chemical dilution and isotopic exchanges have taken place and before any decay process has started. Any radiocarbon study of ground water would thus require a complete chemical study of ground water as well as a good knowledge of soil and aquifer mineralogy. The discussion will be limited to some basic principles regarding the kind of isotopic system involving carbon bearing species. - all the TDIC is of biogenic origin or completely equilibrated with recent biogenic C02 - the TDIC if of mixed origin (organic and mineral) uncompletely equilibrated with recent biogenic C02 or with solid carbonate. In the former case the initial activity of the TDIC will be close to 100 % of modern carbon, in the latter an initial activity lower than 100 % will be expected. - 134 - J.S./. TDIC with a biogenic isotope signature : 3.5.1.1. Total derivation of TDIC from organic matter : This is the case of non carbonate soils where the C02 production gives rise to an active weathering of silicates according to reactions ( 13) and ( ! 4). All the bicarbonate is of recent biogenic origin and pH tends to rise towards the transition 2 HC03" - CO3 " if most of the C02 is converted into HCO3". in that case the stable isotope ,4 content and the C activity of the TDIC are close to those of soil C02 (about -20 %» in most of high and mid latitude regions, see Dever et ai, 1982 and 100!? of modern carbon respectively. Such a process is invoked by Akiti and Fontes ( 1985) to account for the isotopic !3C and ,4C contents found in northern Ghana (Ô'3C = -17 to -19%.,A14C = 36 to 1 ! \%) with !3 a soil C02 with a C content of about -20$. The release of alkaline and earth-alkaline ions during weathering tends to increase the pH which is buffered by the transformation of soil C02 into HCO3". 3.5.1.2. Total equilibrium with biogenic C02 • When in any kind of soil, C02 partial pressure porosity and the ratio gaseous to aqueous carbon are high a complete reequilibration of all the isotope contents may occur under the control of the gas phase. The stable isotope contents and the activities are defined by : ô(TDIC) = {Znrij x [ô(C02 gas) - s^/Im, (22) and A(TD1C) = [2mjxA(C02gas)(l - 2,3 B-ïïllïï^ (23) 2 where m is the modality, i refers to the carbon bearing species : HCO3", HCO3" or C03 ", ef refers to the stable isotope fractionation between C02 gas and species i. Detailed calculations on this system were proposed by Deines et al ( 1974) and Wigley 2 ( 1976). As an example at pH 7, pCC^ = 1 Q~ and t = 25°C for à (C02 gas) = -25%. and A(C02gas)= 100* one obtains : ô(TDIC) T-17%. and A(TDIC) T 1021. Such is probably the mode! which applies to ground waters from Oahu Islands in Hawaï (Hufen et al, 1974) where w(TDIC) = -17 to -19%., A(TDIC) =90 to 108!?. Soil COj is circulating through fractured lavas and equilibrates with TDIC. A somewhat similar situation is recorded from a basaltic aquifer in Southern Brazil (Gallo, 1979) and from crystalline aquifer in the Sahellan 2one ( Dinger et al, 1984 ;0usmane it is interesting to note that in the case of a total equilibrium controlled by C02 (gas) (open system on the gas phase), the final activity of the TDIC is independent of the process of mineralization which takes place before or during the isotope exchange. Then, any kind of process of mixing, even including dead carbonate (as discussed below) which should have produced an activity lower than 100$ will lead to an activity of about 100S if the successive C0o exchange is complete. Radiocarbon activities of TDIC from ground waters from karst in Southern France (Fleyfel, 1980) or from soil water from Champagne chalk (Dever et ai, 1982) indicate that isotope exchange in an open system may be the prédominent process. 3.5.2. Mixing of two sources of carbon within the TDIC : The basic chemical scheme is the reaction of carbonic acid on soil carbonate i.e. calcite : C02 + H20 + CaC03 = ( HC03)2 Ca (24) - 135 - Several models are available (see Fontes, 1983 for a detailed discussion) to account for this dilution in a closed system, more or less modified by further isotope exchange. 3.5.2.1. Empirical approach : Some authors adopt a value of 85SE to account for the chemical dilution of the ,4C content of the TDIC. This empirical approach (Yogel and Ehhalt, 1963 ; Vogel, 1970) does not correspond to any chemical or isotopic model but is still used perhaps because it could hardly lead to unrealistic results (Frolich et al., 1984 ; Rudolph et al., 1984). 3.5.2.2. Chemical mixing model : The dilution of the biogenic carbon (aqueous CC^ + 0.5 HC03~) in the TDIC (aqueous COj + HC03~) is equal to the initial activity : A0 = Ag{[(mC02(aq.) + 0.5m HC03")]/[m(C02(aq.) + mHC03~)]} (25) 14 where m refers to molality and Ag is the C content of the gaseous phase (Aq is taken as 100$ tor prebomD period). I his chemical balance must include an additive term ( +0.5mHC03".Ac) if the activity AçOf the solid carbonate is significantly greater than zero. This approach proposed by Vogel & Ehhalt ( 1963) and ingerson & Pearson ( 1964) was extensively used by Tamers ( ! 975). The method gives reasonable results in carbonate soils where the partial pressure of C02 is low. Obviously it does not allow any isotope exchange and may thus give misleading results for instance in the case of an open system (see 3.5.1.2). 3.5.2.3. Isotope mixing model : This model also considers the isotope balance in the mixing between two sources of carbon : the gas phase ( hQ, Ag) and the solid phase ( ôc, Aç). The initial activity Ag of the TDIC is thus given by : Ao = l^TDIC-ôc)/(V MV^'Ac (26) This model, known as the Pearson's model (Ingerson & Pearson, 1964 ; Pearson & Hanshaw, 1970) is useful when the isotope exchange between TDIC and the gas or the solid phase remains limited (Talma et al, 1984). 3.5.2.3. tlixing with exchange : This model consists in mixing the solid phase with a bicarbonate in equilibrium with soil C02: Ao = K *TDIC " 6c)/( bg - sg " M[Ag " Ac^ *c (27> This treatment was proposed by Oonfiantini (1972) and tested in several studies performed by the IAEA (Salem et ai, 1980) and was adopted by others research groups (Calf & Habermehl, 1984 ; Blavoux & Olive, 1981) and gave results consistent with the geological context. It may be of help if no informations are available on the hydrochemistry 3.5.2.4. Isotope mixing model with exchange in closed system .• This model considers the isotope mass balances for the following processes : carbon dioxide dissolution, isotope exchange with C02 gas in the unsaturated zone and carbonate dissolution ((look, 1972, 1976, 1980). The final derivation is : - 136 - A0 = (a/DAgQ + 0.5 ( 1 - a/ZXAgj, + A]o) + k (28) which is equivalent to a pure chemical balance (3.5.2.2) corrected by an additive term of isotope exchange k. k = Ago(l -28g)-0.5(Aao + A]o)m with ôj-ta/Zjôgp- 0.5(1 - a/Z)(ôM+ ô]o) m = (2g) ôgo " eg(1 +V_0'5(Ôao+ôlo> where b and A stand for 13C contents, S is given by eq(20), all in decimal fractions, indices ao, !o and go refer to initial aqueous CO2, solid carbonate and gas respectively, H is the sum of the chemical activities of the various forms of dissolved carbon and a is the chemical activity of the aqueous CO2 and where a complete isotope exchange takes place between the initial gas (go) and the bicarbonate (bo) which is obtained by mixing of aqueous C02 and solid carbonate: Egaôgo- ôboand ôbo = a5(ôao+ôlo)- This model is gives satisfactory adjustements (hook, 1976) when the partial pressure of C02 in soils is high and when all parameters are precisely defined. It displays a very high sensitivity to the corrective term. 3.5.2.5. Global exchange-mixing mode/ in closed system : The approach is similar to the previous one but allows for an exchange process with the solid phase. The TDIC is considered as the mixing of three parts of carbon : a gas phase, a solid phase and a fraction of one them which is completely exchanged with the other. From 13C and 14C isotope mass balances it comes as.in the previous approach : A,-, = ( 1 - (WOT)A., + (CM/CT)AM + k i-vi\ The corrective term k writes as follows : [Rg(i - 2.3e)-nrl]{ô - (Ch/Cr)ôn - [1 - (Cn/cT]ôg} k - - <31 > (ôg - E - Ô„) where C refers to molar concentrations and where subscripts T, M and g refer to the TDIC, the carbon of mineral origin and the gas phase, respectively and where e is the isotope enrichment factor between C02 gas and the solid carbonate approximately the sum £_ + es, eqs(20)and(19). It can be demonstrated that C^ is identical to half of the total alkalinity for any pH values. This model may be refined by the use of e if the exchange is in favour of the gas phase ( k Consistent results were obtained with this model in the aquifer of the "Calcaire carbonifère" in the North of France (Fontes & Gamier, 1979) in the South of Spain (Plata el ai, 1984) or in Romania (Tenu & Davidescu, 1984). However, as the previous one, the model is highly sensitive to the correction term. 3.5.2.6. Mass tranfer models : They consists in step by step chemical mass-balances which allow to define the reactions of dissolution-precipitation which take place along flow paths. Then an isotope mass-balance gives the modification of !3C content of each carbon bearing species between inital and final situations. When calculations allow simultaneous prediction of the hydrochemistry and of the '3C content of the TDIC, the initial '3C (and 14C) contents are evaluated and the age is computed. In the Floridian aquifer (Plummer, 1977) the mass-transfer model gave estimations of transit times from 3200 to 36000 a over a distance of about 300km. Corresponding ground water velocities appear rather high. A somewhat similar and simpler approach was proposed by Reardon & Fritz ( 1978). Temperature pC02, pH and b C of soil C02 in the recharge zone are adjusted acceptable values in an open system in order to account for final conditions after dissolution of solid carbonate. The value of Ap is then calculated for initial conditions i.e. before any decay and after ^every dilution processes. The supply of carbon of mineral origin is given by the arnont of Ca2+ and Mg2+ corrected for sulphate contribution and ion exchange. Satisfactory results were obtained in Austria and UK (Andrews et al., ! 984). Mass-transfer and isotope mass-balance models are conceptually satisfying because they put chemical and isotopic constraints on the boundary conditions (partial pressure of C02, !3 pH, temperature C content of soil C02 in the recharge zone). However, they do not take into account the possibility of isotope exchange with the carbonate matrix of the aquifer. 3.6. Further causes of uncertainty in MC interpretations : Many problems are still to be solved for a "safe" use of radiocarbon in ground water dating especially in the case of carbonate aquifers or even in aquifers in which carbonates occur as fracture minerals: a) Retardation through chemical and isotope exchange with carbonate matrix (Mozeto et ai., 1984 ; Gamier, 1985). For instance, it should be kept in mind that carbonate saturation means that dissolution proceeds at the same rate as precipitation Thus, isotopes are still exchanging between liquid and solid phase while the chemical composition remains constant. Carbon 13 will reach equilibrium at the interface but !4C equilibrium will be continuously counteracted by the radioactive decay within the solid phase. These exchange processes are probably responsible for the reverse correlation which is frequently observed between ,3C and 14C content of the TDIC. This would suggest that chemical effects are often prédominent over age effects and that many waters are actually younger than estimated from nhem inal nr isntnne mortels b) Diffusion through dead-end pores or stagnant fractures. Such a process may represent another sink for HC (Neretnieks, 1981 ; Neretnieks et ai., 1982). However no specific field-study is available on this process which would result in additional delay for the 14C of the TDIC compare with the water. - 138 - c) Mixing : it should be remembered that time estimation deai with carbon 14 atoms rather than with residence time, transit time or turn-over time of ground water. In the case of mixing between different supplies the resulting ,4C activity is given by: A= iQ^Aje-^V^jQi^f^i (32) where Q is the flow, M is the TDIC concentation, A the MC activity of the TDIC and t the residence time of each single supply i into the resulting mixing. d) Deep production. Very low activities (close to 1 or 2$ of modern carbon) are frequently measured in deep confined aquifers, where "dead" TDIC would be expected. These small amounts of 14C are generally attributed to mixing. However, it was shown (Zito et al, 1979) that a small deep production of ,4C under the influence of rock radioactivity was theoretically possible. Field investigations are needed to estimate the real importance of such a deep supply. e) Changes in the global water budget. Paleogeographical and paleoclimatological evidences suggest a major episode of general drainage during the low marine level induced by the last glacial epoch (between 25 and 12000 BP). The beginning of this discharge period coïncided with the end of a period of high lake level especially at tropical latitudes (see Street et al., 1983). Therefore, ground water flow conditions have undergone major changes at about 25000 BP. f) Changes in vegetal cover. Climatic or man-induced variations in the vegetal cover may ,3 have produced changes in the C content of soil C02 which is a basic parameter for all the models of reconstruction of the initial activity of the TDIC. g) Deep supply of CC^. in volcanic or tectonically active regions, some uprise of deep C02 may modify. For all these reasons, estimations of ground water "ages" greater than 25.10^ a are risky. 3.7. Conclusions : The application of the radiocarbon method to ground water studies suffers numerous and severe limitations. However this isotope also displays incomparable basic qualities : it is present in all ground waters and its half-live is the only one well-suited for the investigation of old ground waters in the time range 102 to 104 a. Age evaluations is site specific and requests a detailed knowledge of water chemistry but also of soil and aquifer characteristics. In the case of carbonate aquifers time estimates may even become a question of art and personal feeling. Some hydrogeological situations and the corresponding applicable models for A0 reconstructions are represented on Figure 8. In many situations where TDIC is in an open system over soil C02, the chronometer is reliable. Such situations must be explored in shallow aquifers where thermo-nuclear 14C may provide an additional labelling for the evaluation of Ag. Furthermore, 14C offers in that case a complementary tool for 3H studies with the advantage of an easy link to the prebomb situation. Further extrapolation could then be made more safely to old ground waters in similar situations. 139 High C02 production OPEN SYSTEM vs C02 Porous soil 6 (TDIC) = «n?Ot)-e Non carbonate A(TDIC) = A(COz) (1-2.3e) aquifer No correction :mhÉÈd^ÉkÉMm High C02 production OPEN SYSTEM vs C02 Porous soil S (TDIC) = S (C02)-£ Carbonate aquifer A (TDIC) = A(C02) (1-2.3e) (fractured) No correction i^kii^Vj ïj/{i^M. .im2\--/.vl>S'-j'y •-,, Medium C02 production Complete transfer of Low porosity (clay C02 into TDIC fraction) 6 (TDIC) = & (C02) Non carbonate aquifer A (TDIC) s A(C02) ~^jwpm No correction JJIUUijj.-UiLiCJlJI*; Low COj. production Closed System with Carbonated soil few isotopic exchange Low porosity Correction with all Non carbonate aquifer available models : T,P,G,M,F&G,P1,R&F MijWjbliMt High C02 production Closed system with C02 Carbonate soil exchange Models P,G,M, Aquif.miscelleaneous F&G.P1.R&F rocks Medium (or high) C02 Closed system production Partial exchange with Carbonate soil, low carbonate porosity Models P.F&G Carbonate rock. Fig.8. Various hydrogeological and pedological situations and corresponding models for the evaluation of the ,4C initial activity of the TDIC. Models : T = Tamers, P = Pearson, G = Gonfiantini, M = Mook, F et 6 = Fontes & Gamier, PI = Plummer, R et F = Reardon & Fritz. Modified from Fontes ( 1983). - 140 - 4. CHLORINE 36: This environmental isotope is also produced by cosmic rays and secondary neutrons. For several reasons it is potentially one of the most powerful tools in ground water dating. Its half-like of about 301 x lC^a allows to explore theoretically about 3. lO^a i.e. all the time range of the Quaternary. The very simple geochemistry of the element precludes most of the reactions of exchange, absorption, redox, precipitation, which can affect the behaviour of other time indicators and especially of ,4C. 4.1. Sources of 36 CI : 411 Atmospheric production • Main atmospheric sources are spallation reactions on argon nuclei in the high atmosphere. An additional supply is provided by neutron irradiation of 35C1 atoms from aerosols. As expected from variations of cosmic ray fluxes at earth's surface, the total fall-out depends on the latitude (Fig.9). Estimates of the total fall-out of 3&C1 range between 15 and 30 at m~2.s~' (see Bentley et al, 1985). After production 36Ci nuclei is probably entirely transformed into 36CF ions. It is then introduced as meteoric chloride into soil waters and ground waters and follows the geochemical cycle of chloride. 4.1.2. Sur fax and subsurface production . At earth surface and in the first horizons of soils, cosmic rays may still produce 36C1 through spallation reactions mainly on potassium from rock minerals. Therefore, this production depends on soil lithoiogy. However, after a long time the production of chlorine 36 atoms is equilibrated by their radioactive decay. The corresponding steady state value of 36C1 concentration is called secular equilibrium. It takes approximately I06a to be reached. If the r8te of chemical erosion is lower than the rate of decay a high 36C! concentration may be obtained in the rock. This 36C1 is then leached and added to meteoric 36C1 into soil water. Furthermore, soil solutions may not be homogenous. Under evaporative climates salts tend to concentrate in surface horizons where vapour transport is dominant (Barnes & Allison, 1983). During heavy rains, infiltration takes place through the largest pores. Chloride rich solutions then remain in the microporosity where their residence time may be long enough for neutron irradiation and 36C1 production to become significant. (An example of long residence time of CI" in soil microporosity is provided by regions where Holocene marine transgression left, behind salty solutions which are not completely washed out after about 5000 a of successive emergence e.g. lands invaded by the Tyrrel sea in Northern Quebec or by the Yoldia sea in Scandinavia). Activated chloride is then transferred into circulating soil solutions through diffusion processes. 4.1.3. Deep production : Neutrons are produced in rocks through the interaction of alpha particles (decay of uranium and thorium chains) on some light nuclei (Li, B, rig, Al..). Resulting neutrons are absorbed in the rock itself, chiefly by many reactions of activation. One of them 35C1 (n,Y ) 36C1, produces 36C1- In that case also ^Cl content of the rock reaches a secular equilibrium value which will be high in uranium and thorium rich rocks like granite and other crystalline rocks (see Fontes, 1985, Andrews et al, 1985 for a detailed discussion). 4.1.4. Nuclear weapons : Nuclear detonations which occured in marine environment from 1952 to 1958 generated a large amount of 36C1 through neutron irradiation of 35C1. - 141 - 60" 70" SO" GEOMAGNETIC LATITUDE Fig.9. Distribution of cosmic rays produced 36C1 in function of the latitude. From Bentley et al. (1985). 142 A fairly defined peak of 36Ci occured in precipitation as recorded in ice stratigraphy (Elmore et ai, 1932, fig. 10). This peak is several orders of magnitude higher than natural 36C1 activities in precipitations. It may be used as a time indicator in ground water and soil waters studies. Furthermore because marine nuclear tests occured nearby the equator, the resulting 36C1 was well distributed at the global scale. 4.2. Measurements, notations and ranges of variât tons : Measurements are currently performed by TANDEM ACCELERATOR. MASS SPECTROMETRY at the University of Rochester ( D. Elmore). Contents in o6Cl can be expressed in atomic ratios 36C: (35C1 + 37Ci). For simplicity this ratio is multiplied by 1015. Typical values are about 1 in sea water, 50 to 200 in prebomb precipitation and more than 2000 for rains of the 57-67 period. This notation is dependent on the total chloride content of the sample and can be affected by processes of salt dissolution along ground water flow paths. Another possible expression of 36C1 contents is in concentration per volume unit (atoms r1) generally multiplied by 10""7. This notation is conservative through leaching of salts, but depends strongly on the concentration which occur during evaporation and évapotranspiration. Values for 36 concentration range between 5 and 50 x 10~7at.!~! for recent prebomb ground waters ( Bentley et al., 1985). 4.5. Case studies : Some applications of the 36C1 technique are rewieved by Bentley et al., 1985. Time dependence is given by the general equation : R^Roe-^ + Reqd-e"^1) (33) where t = time, R = atomic ratio, X = decay constant of 36C1. The first term of the second member accounts for radioactive decay of an initial input (symbolo) whereas the second one expressed the build-up due to the deep production and the concentration distance equilibrium value (symbol eq). Transit time to secular investigations may deal with: - pure cosmic fall out - cosmic input (fall out + surface production) - deep production - nuclear supply. •4.3.1. Pure cosmic fall-out : The most significant study is that of the Great Artesian Basin in central Australia. The aquifer, one of the largest in the world with that of the Continental Intercalaire in Northern Sahara, is confined within sandstones of Jurassic and Cretaceous age. These rocks are not favourable to subsurface and deep production of 36C1. Thus no further significant input than cosmic fall-out is expected. Along flow paths of about 700km, the atomic ratios R0 decrease from about 110 to about 5. lO"15 (Fig. 1 ! ). Corresponding time of transit are of the order of 1.5 Ma provided the following conditions are satisfied : a) the decrease in atomic ratio is only due to radioactive decay ( i.e. the system is closed with respect to 35CÎ and to total chloride) ; b) the values of R observed in the vicinity of the recharge zone are representative of the initial values R0. - 143 - 36C1 at. m"2 s l { h } Dye 3 u 103 102 195(1 1955 196(1 19*5 197(1 1975 I97X years Fig. 10. Thermonuclear peak of 36C1 In an antarctic ice core. Modified from Elmore et al. (1982). - 144 - £ S o i o O 145 c) no significant fluctuations of R0 occur over the calculated time range. Calculated values of permeabilities from pumping tests and ]4C measurements are in agreement with those given by 36C1 measurements (Calf & Habermelh, ! 984). 4.3.2. Cosmic input (fall-out * surface production) : This mechanism was proposed to explain the rather high chlorine 36 contents from Saharian deep ground waters. The study is still preliminary and deals with the eastern part of the aquifer of the Continental Intercalaire in Norh eastern Algerian Sahara. The aquifer is mainly made of sandstones inter bedded with marls. Chlorine 36 was measured on 5 boreholes roughly distributed along a flow direction from the recharge zone on the western part of the system. The concentrations ( in atoms 1~ ' ) are higher by approximately two orders of magnitude, than expected for rain waters at these latitudes (Fontes et al, in prep.). The évapotranspiration should have had removed about 99 % of the rainfall. This appears rather incompatible with a major recharge process at the scale of the whole Sahara. A significant deep production within the aquifer can be ruled out because 36C1 contents decrease along the flow path. An explanation for the high 36C1 content is an extra source in the recharge zone. This additional supply is attributed to the spallation on potassium and perhaps also on calcium atoms and to the activation of 35C1 accumulated in the unsatured zone. Calculations assuming a constant initial value suggest that deep Saharian ground waters are much more aged than previously assumed on the base of 14C measurements (Sonntag e, a/., 1978). They would have implied between 105 and 5 x 105 a for a transit of about 280km. Estimated values of permeability coefficients l(T6m.s~1 are about one order of magnitude lower than those deduced from pumpings tests. 4.3.3. Deep production : This phenomenon was considered as weak or negligible (Bentley & Davis, ! 982 ; Bentley et al., 1985) until a clear cut evidence ot ibCl deep generation was found in tne deep grouno water system of Stripa (site of the "Stripa Project". The scope of the Stripa Project ( run by OECD) is a detailed evaluation of the environmental conditions (rock mechanics, ground water flow,.., in a geological site which could be used for nuclear waste isolation). A batholith of granite is intrusive in metamorphosed rocks (leptite) all of Precambrian age. Galleries of a previous iron mine in the leptite were overdug within the granite. Artesian 2 saline ( up to 700 ppm for CI" and 200 ppm of S04 ~) ground water are out flowing from deep bore holes drilled in the granite. The origin of these low flows ( between 0.5 and 5 1 mirf1 ) and of their salt content are under discussion. Extensive hydrogeological studies including hydrochemistry and environmental isotope studies are still in progress. Preliminary 36CT measurements (Michelot et al., 1984) show that deep ground waters contained more than two orders of magnitude of 36C1 than estimated on infiltrating meteoric waters. These waters were tritium free and displayed TDIC with very low !4C1 contents. Surface and subsurface production may be excluded because of the low rate of chemical erosion and because of the absence of salt accumulation in the unsaturated zone. Deep production of 36C1 is taking place in the system. Rock chemistry is well known and the production of neutrons and of 36C1 may be estimated. The very high uranium and thorium contents of the rocks would induce a neutron flux of about 12.1CPn.kg~!.a~' (Andrews;?/ al, 1985) which is approximately 10 times higher than for the Standard granite (Feiqe et al, -1 1968). The calculated secular equilibrium value Req would be close to 300 x 10 ^ (Fontes et al, 1984). Further in situ measurements of the neutron flux lead to values of P.eq (granite) and R (leptite) respectively close to 200 and 40 x 10"'5(Adrews et al, !985). - 146 - From these values it can be inferred that dissolved CI" is not in equilibrium with the granite. A simple derivation of ground water salinity through leaching of fluid inclusions or chloride bearing minerals is thus excluded. In a diagram ^6C1/(C1^) vs C)~ most of the points lie between the representative lines of secular equilibrium for granite and leptite ( Fig. 12). This may be due to a: a) Mixing between CI" derived from leptite and from granite without any age effect (fast circulation) ; b) Long residence time in granite for chloride (and water) which was previously in secular equilibrium with the leptite and in that case estimations of residence times in the granite may be done ( Fig. 12) through eq.( 33) which reduces to its build up term. In both cases the chloride in equilibrium with the leptite may be either produced by rock dissolution or generated by the long (>106a) residence time in the leptite of an allochtonous chloride of sedimentary origin. Hypothesis b has presently our preference because preliminary measurements indicate that rock contents in soluble salts are low, especially in 2 S04 ~ in the granite and in CI" in the leptite. 4. J 4. Nuclear origin : The recharge rate of a shallow unconfined aquifer in Canada (Bentley et al, ! 982) was estimated by the location of 1958 peak within the profile (Fig. 13). Another elegant study on recharge rate under semi-arid conditions was performed on a soil profile near Tucson (Arizona). The nuclear peak is found at a depth of 17cm. Chlorine 36 balance within the profile is in agreement with an average infiltration rate of about 42cm a"1. 4.4. Conclusion on 56 CI availability in hydrogeological studies : Sedimentary rock poor in uranium and thorium represent favourable cases for age determination. The decay function can be applied assuming a constant input of 36C1. This initial supply is both of meteoric origin and induced by cosmic rays in the shallowest horizons of soils. When the lithology includes several types of rocks : sedimentary (high and low chloride matrices) metamorphic and crystalline, several cases must be distinguished (Fig. H). If circulations are fast dissolution and mixing processes play the major role, and no age estimate can be inferred from 36C1 contents, if circulations are low radioactive decay and build-up are prédominent and residence times may be evaluated. In the special case of recent water, 35C1 offers several advantages over tritium and other short-lived isotopes like krypton 85 : - Negligible decay - Peak very well defined - No losses or gain through isotope exchange reaction - Good distribution in both hemisphere. However for the time being 36C1 measurements are still severely restricted by the availability of accelerators. But the possibilities of the method are such, chiefly in hydrogeology but also in glaciology, paleoclimatology and astrophysics, that a large effort of equipment is predictible in the near future. - 147 - 200 Cl" mgl" Fig. 12. Evolution of the 36C1 contents vs total C\~ at Stripa. Points fall in between the representative lines for secular equilibrium with granite and leptite respectively. Intermediate lines correspond to various times of build-up of 36C1 through close contact with the granite of a chloride previously in equilibrium with the leptite. From Andrews et al (1985). - 148 - Î6C1 and JH A Cl ai. I ' x If) A = tritium units r Ml IT DEPTH, m. Fig. 13. Trititum and 36C1 contents in a shallow aquifer in Ontario (Canada). Modified from Bentley*?/ 149 - co CD CO ce <= !=. CD o en CL >• co E ^—« 3 O O- CO > CD CD _£Q.- CD qC ST3 © c S S O) E .- CO ce =5 " 8 HZ (O i_ =3 -i^ >- •° 8 c_ , L. CD fo ~ s— c; CD o co *" rri CD fJ' V> "O e K co ê '5> c5 ce x: • o Cil oco. Q. F c_ • CO Z3 8o ^—^ CE_ CD co CO ^ . •>- © E - CC et « & oc ce - 150 - REFERENCES. Airey, P.L., Bentley, H.W., Calf, 6.E., Davis, S.N., Elmore, D., Gove, H.E., Habermehl, MA, Phillips, F.M., Smith, J.W., Torgensen, T. 1983. Isotope Hydrology of the Créât Artesian Basin, in "Proc. Int. Conf, Sydney, ! 983, Aust. Govt. Publ. Serv. Canberra. Andrews, J.N., Balderer, W., Bath, A.H., Clausen, H.B., Evans, 6.V., Florkowski, T., Goldbrunner, J.E., Ivanovich, M., Loosll, H. and Zojer, H. 1984. Environmental isotope. studies in two aquifer systems : A comparison of groundwater dating methods. In Isotope Hydrology ' 983, IAEA, Vienna, 535-576. Andrews, J.N., Fontes, J-Ch., Michelot, J-L. and Elmore, D. 1985. In situ 36C1 production and groundwater evolution in crystalline rocks at Str ipa, Sweden ( in press). Barnes, C.J. and Allison, 6.B. 1983. The distribution of deuterium and 180 in dry soils, i. Theory. J. of Hydro!., 60, 141 - 156. Bentley, H.W. and Davis, S.N. 1982. Application of AMS to hydrology, in 2nd Annuat Symposium on Acceleration Mass Spectrometry, Argonne National Laboratories. Bentley, H.W., Phillips, F.M., Davis, S.N., Gifford, $., Elmore, D., Tubbs, L.E. and Gove, H.V. 1982. Thermonuclear 36C1 pulse in natural water. Nature , 300, n°5894, 737-740. Bentley, H.W., Phillips, F.M. and Davis, S.N. 1985. Chlorine-36 in the terrestrial environment. Handbook of Environmental Isotope Geochemistry, P. Fritz and J-Ch. Fontes edrs. vol.!, The Terrestrial Environment, B, Elsevier Publ. Co., (in press). Blavoux,B. and Olive, P. 1981. Radiocarbon dating of groundwater of the aquifer confined in the Lower-Triassic Sandstones of the Lorraine Region, France, J. Hydre/54, 167. Calf, G.E. and Habermehl, MA 1984. Isotope hydrology and hydrochemistry of the Great Artesian Basin, Australia. In Isotope Hydrology \98ô, IAEA, Vienna, 397-4!3. Deines, P., Langmuir, D. and Harmon,R.S. 1974. Stable carbon isotope ratio and the existence of a gas phase in the evolution of carbonate groundwaters, Geochim. Cosmochim. Actt 38,1147-1164. Dever, L., Durand, R., Fontes, J.Ch. et Vachier, P. 1982. Géochimie et teneurs isotopiques de systèmes saisonniers de dissolution de la calcite dans un sol sur craie. Geochim. Cosmochim. Acta, Vol. 46, 1947-1956. Dinger, T., Dray, M., Zuppi.G.M., Guerre, A., Tazioli, 6.S., Traore, S. 1984. L'alimentation des eaux souterraines de la zone Kolokani-Nara au Mali. In Isotope Hydrology] 983, IAEA, Vienna, 341-365. Downing, R.A., Smith, D.B., Pearson, F.J., Monkhouse, R.A. and Otlet, R.L. 1977. The age of groundwater in the Lincolnshire limestone, England and its relevance to the flow mechanism. J. of Hydro!., 33,201 -206. Drost, W. 1983. Isotope Techniques in civil engineering. In Guidebook on Nuclear Techniques in Hydrology, IAEA, Technical Reports Series, No.91, Vienna, 403-409 Eisenbud, M., Bennett, B., Blanco, R.E., Compere, E.L., Goldberg, E., Jacobs, D.G., Koranda, J., Moghissi, AA, Rust, J.H. and Soldat, J.K. 1979. In Behaviour of Tritium in the Environment . IAEA, Vienna, 585-588. Feige, Y., Oltman, B.G. and Kastner, J. 1968. Production Rates of Neutrons in Soils due to Natural Radioactivity. J. Geophys. Res., 73,3135-3142. Fleyfel, M. 1980. Etude hydrologique, géochimique et isotopique des modalités de minéralisation et rie transfert riu carbone rians la zone ri'infiltration d'un anuifère karstique : Le Baget, Thèse, Univ. Paris VI. Fontes, J.Ch. 1983. Dating of Groundwater. In Guidebook on Nuclear Techniques in Hydrology , IAEA, Technical Reports Series, N° 91, Vienna, 285-317. - 151 - Fontes, J.Ch. 1985. Méthode au Chlore 36 - Datation des eaux. In Méthodes de datation par les phénomènes nucléaires naturels Coll Scientif. CEA E. Roth, B. Poty Coord., Chap. XIV ( in press). Fontes, J-Ch. and Gamier, J-li. 1979. Determination of the initial 14C activity of the total dissolved carbon -, a review of the existing models and a new approach, Water Resour. Res. 15,2,399-413. Fontes, J.Ch., Pouchan, P., Saliège, J.F. andZuppi, G.M. 1980. Environmental isotope study of groundwater systems in the Republic of Djibouti. In Arid Zone Hydrology Investigations with Isotopes Techniques, IAEA, 237-262. Fontes, J.Ch., Brissaud, I. and Michelot, J.L. 1984. Hydrological implications of deep production of chlorine- 36. Nuclear Instr. and Methods, vol 233, n'2, 303-307. Fritz, P., Reardon, E.J., Barker, J., Brown, R.M., Cherry, JA, Kill»/, R.W.O. and Mcnaughton, D. 1978. The carbon isotope geochemistry of a small groundwater system in northern Ontario, Water Resour. Res H, I059-1067. Frohlich, K.,Gellermann, R. and Hébert, D. 1984. Uranium isotopes in a sandstone aquifer ; Interpretation of data and implications for groundwater dating. In Isotope Hydrology 1983, Al EA,Vienna, 1984,447-466. Gallo, G. 1979. Utilisation complémentaire des isotopes du milieu et de l'hydrochimie pour l'étude hydrogéologique des eaux souterraines de la régio de Ribeiro Preto (Brésil), Isotope Hydrology \979 (Proc. Symp. Neuherberg, 1978) Vol. 1, IAEA, Vienna, 1979, 367-410. Gamier, J.M. 1985. Retardation of dissolved radiocarbon through a carbonated matrix. Geochim. Cosmochim. Acta, vol. 49, 683-693. Gat, J.R. 1980. The isotopes of hydrogen and oxygen in precipitation. In Handbook o, Environmental Isotope Geochemistry, P. Fritz and J-Ch. Fontes edrs., vol. 1 The Terrestrial Environment, A, Elsevier 21-47. Gonfiantinl, R. 1972. Notes on isotope hydrology, internal publication, IAEA, Vienna. Hufen, T.H., Lau, L.S. and Buddemeir, R.W. 1974. Radiocarbone Î3C and tritium in water samples from basaltic aquifers on the island of Oahu-Hawaii, in Isotope Techniques in Groundwater Hydrology, (Proc. Symp. Vienna, 1974) vol. 2, IAEA ( 1974) 111. I.A.E.A., 1969. Environmental Isotope Data No. 1 : World Survey of Isotope Concentration in Precipitation ( 1953-1963), Technical Reports Series, N" 96, IAEA, Vienna. I.A.E.A., 1970. Environmental Isotope Data n°.2 : World Survey of Isotope Concentration in Precipitation ( 1964-1965), Technical Reports Series, N° 117, IAEA, Vienna. I.A.E.A., 1971. Environmental Isotope Data n°3 : World Survey of Isotope Concentration in Precipitation ( 1966-1967), Technical Reports Series, N° 120, IAEA, Vienna. IAEA, 1973. Environmental Isotope Data n°4 : World Survey of Isotope Concentration in Precipitation ( 1968- 11969), Technical Reports Series, N° 147, IAEA, Vienna. I.A.E.A., 1975. Environmental Isotope Data n°5 : World Survey of Isotope Concentration in Precipitation ( 1970-1971) Technical Reports Series, N° 165, IAEA, Vienna. I.A.E.A. 1979. Environmental Isotope Data n°6 : World Survey of Isotope Concentration in Precipitation ( 1972- ! 975), Technical Reports Series, N° 192, IAEA, Vienna. Ingerson, E. and Pearson, F.J. Jr. 1964. Estimation of age and rate of motion of groundwater by the 14C-method, Recent Researches in the Field of Hydrosphere. Atmosphere and Nuclear Geochemistry, Maruzen, Tokyo, 263-283. Levin, I., 1978. Régionale Modellierung des atmospharischen C02 auf Grund von C- 13 and C-14 Messungen, Diplomarbeit, Institut fiir Umweltphysik, Univ. Heidelberg. Michelot, J.L., Bentley, H.W., Brissaud, !., Elmore, D. and Fontes, J.Ch. 1984. Progress in environmental isotope studies (36C1, 34S, 180) at the Stripa site. In Isotope Hydrology 1983",, IAEA, Vienna, 207-229. - 152 - Mook, W.6. 1976. The dissolution-exchange model for dating groundwater with 14C, in Interpretation of Environmental Isotope and Hydrochernical Data ir, Groundn'ater HydrologyiProc. Panel Vienna, 1975), IAEA, Vienna ,213-225. hook, W.G. 1980. Carbon-14 in hydrogeological studies. In Handbook of Environmenta, Isotope Geochemistry, Vol.1 (edrs. P. Fritz and J-Ch. Fontes, vol 1, The Terrestrial Environment, A) Elsevier, 49-71. Mozeto, A.A., Fritz, P. and Reardon, E.J. 1984. Experimental observations on carbon isotope exchange in carbonate-water systems, Geochim. Cosmochim. Actald, 495-504. Neretnieks, I. 1981. Age dating of groundwater in fissured rock : influence of water in micropores. Water Resour. Res 17 (2), 421-422. Neretnieks, I., Eriksen, T. and Tahtinen, P. 1982. Tracer movement in a single fissure in granitic rock : some experimental results and their interpretation. Water Resour. Res 18 (4), 849-858. Nir, A., Kruger, S.J., Ligenfelder, R.E. and Flamm, E.J. 1966. Natural tritium. Rev. Geophys. ,4,441-456. Ousmane, B., Fontes, J-Ch., Aranyossy, J-F. et Joseph, A. 1984. Hydrologie isotopique et hydrochimie des aquiféres discontinus de la bande sahélienne et de l'Aïr (Niger). In Isotope Hydrology1983, IAEA, Vienna, 367-395. Pearson, F.J. Jr. and Hanshaw, B.B. 1970. Sources of dissolved carbonate species in groundwater and their effects on carbone-14 dating, in Isotope Hydrology(Proc. Symp. Vienna, 1970) IAEA, Vienna, 271 -286. Plata, A., Baonza, E., Sildago, A. 1984. Hidrologia isotopica de las aguas subterraneas del Parque Nacional Donana y zona de influencia. In Isotope Hydrology î 983, IAEA, Vienna, 321-340. Plumer, L.N. 1977. Defining reactions and mass transfer in part of the Floridian aquifer, Wat. Resour. /fe? 13, 801 - 8 ! 2. Przewlocki, K. and Yurtsever, Y. 1974. Some conceptual mathematical models and digital simulation approach in the use of tracers in hydrological systems. In Isotope Techniques in Groundwater Hydrology', vol. II, IAEA, Vienne, 425-450. Reardon, E.J. and Fritz, P. 1978. Computer modelling of groundwater !3C and !4C isotope compositions,^ Hydro! 36,201-224. Rudolph, J., Rath, H.K. and Sonntag, C. 1984. Noble gases and stable isotopes in ,4C-dated palaeowaters from central Europe and the Sahara. In Isotope Hydrology'1983, IAEA, Vienna, 467-477. Salem, 0., Visser, J.H., Dray, M. and Gonfiantini, R. 1980. Groundwater flow patterns in the western Lybian Arab Jamahiriya, in Arid-lone Hydrlogy: Investigations with Isotope Techniques, IAEA Vienne, 165-179. Saliège, J.F. et Fontes, J-Ch. ! 984. Essai de détermination expérimentale du fractionnement des isotopes 13C et 14C du carbone au cours de processus naturels, i ntern. Journal o, Applied Radiation and Isotopes,vol.35,n"l,55-62. Sonntag, C, Klitzsch, E., Lôhnert, E.P., El-Shazly, E.M., Munnich, K.O, Junghans, Ch., Thorweihe, U., Weistroffer, K., Swailen, F.M. 1979. Paiaeoclimatic information from deuterium and oxygen- 18 in carbon- ! 4 dated North Saharian groundwaters : Groundwater formation in the past, in Isotope Hydrology (Proc. Symp. Neuherberg, 1978) vol.2, IAEA, Vienna, 569-581. Talma, A.S., Vogel, J.C. and Heaton, T.H.E. 1983. The geochemistry of the Uitenhage artesian aquifer : Carbone solution in a closed system. In Isotope Hydrology, 1983, IAEA, Vienna, 481-497. Tamers, MA. 1975. Validity of radiocarbon dates on groundwater. Geophys. Surv. ,2 ,217. - 153 - Tenu, A. and Davidescu, F.D. 1984. Recherches isotopiques sur les eaux souterraines des formations poreuses de la partie orientale de !a dépression Pannonienne. In Isotope Hydrology !983, IAEA, Vienna, 249-269. Vogel, J.C. ! 970 . Carbon- 14 dating of groundwater, in Isotope Hydrology 1970 ( Proc. Symp. Vienna, 1970) IAEA, Vienna, 225-237. Vogel, J.C. and Ehhalt, D. 1963. The use of carbon isotopes in groundwater studies, Radioisotopes in Hydrologyi.Proc. Symp. Tokyo, 1963), IAEA Vienna, 383-395. Weiss, W., Bullacher, J. and Roether, W. '979. Evidence of pulsed discharges of tritium from nuclear energy installations in central european precipitation. In Behaviour of Tritium in the Environment IAEA, Vienna, 17-30. W;gley, T.M.L 1975. Carbon dating of groundwater from closed and open systems, Wat. Resour. Res 11,2, 324-328. Wigley, T.M.L. 1976. Effect of mineral precipitation on isotopic composition and !4C dating of groundwaters, Nature (London), 263. 219. Yurtsever, Y. 1979. Environmental isotopes as a tool in hydrogeological investigations of southern karst regions of Turkey, Proc. Int. Sem. on Karst Hydregeelogy Oymapinar-Antalya, Turkey. Yurtsever, Y. 1983. Models for tracer data analysis, in Guidebook on Nuclear Techniques in Hydrology, Technical Reports Series n°91, IAEA, Vienna, 381 -402. Yurtsever, Y. and Payne, B.R. 1978. A digital simultation for a tracer case in hydrological systems. Multi compartment mathematical model. Proc. Sacond Int. Conf. on Finite Elements in Water Resources, London. - 154 -