Several factors need to be considered when planning a field investigation involving environmental tracers. For groundwater dating, there is often a choice of tracer, and those that are selected will depend upon the required precision of analyses, the site conditions and the project budget. A summary of the advantages and limitations of the different tracers is provided in Table 3. The factors that should be considered in the selection of tracer include:
Availability of Piezometers
Are groundwater samples to be obtained from existing wells, or will piezometers be installed specifically for the project? The length of the well screen of existing wells will impact on the precision that can be obtained with different residence time tracers and may therefore affect the choice of tracer. If piezometers are to be specifically installed, then the length of the well screen is usually a trade-off between the desire to sample a discrete interval of the aquifer and the need for sufficient well-yield for purging and sampling. Piezometer diameter may be influenced by pump availability and minimum diameter requirements of downhole geophysical or sampling equipment. Drilling methods may be influenced by a need to avoid introducing air into the aquifer.
The Expected Age of the Groundwater and Required Precision
Although the purpose of measuring environmental tracers is often to estimate the age of the groundwater, it is sometimes necessary to know the approximate age so that the most appropriate tracers can be chosen for the study. Information related to choosing tracers is provided in Figure 2 and Table 3. The expected age of the groundwater and the required precision of groundwater age measurements will affect the choice of tracers. Only event markers with monotonically increasing atmospheric concentrations can uniquely provide groundwater ages over the full age range of the tracer. The sensitivity of different tracers for estimating groundwater age is largely dependent on shape of their input function and measurement precision. Radon is the most reliable tracer for ages of days to weeks; 3H/3He for ages of months to a few years; SF6, 85Kr, 3H/3He have good precision for ages from a few years to about 20 years; SF6, 85Kr, 3H/3He and CFCs have good precision from about 20 – 40 years; and CFCs, SF6 and 85Kr have good precision for ages from about 40 – 60 years (Figure 6). Fewer tracers are available for older groundwater.
Local Geology and Geochemical Environment
The use of environmental tracers for estimating groundwater flow usually relies on conservative tracer behavior, and this can be affected by the geological and geochemical environment. CFC-11 is likely to degrade in groundwater with a low dissolved-oxygen concentration, and all CFCs can degrade in highly anoxic groundwater (Hinsby et al., 2007). CFCs can also sorb to aquifer materials with a high organic carbon content (Choung and Allen-King, 2010), and so other age indicators may be preferred in this environment. Subsurface production of SF6 has been reported in some aquifers, particularly in granites and rocks containing fluorite (Chambers et al., 2019). Interpretation of 14C is more difficult in carbonate aquifers due to chemical reactions that can modify the 14C activity of TDIC, although there are few alternative tracers over the time range represented by 14C (Figure 2).
Sampling Requirements and Analytical Facilities
Some environmental tracers are straightforward to sample, whereas others require more care and/or specialized sampling equipment. For others (e.g., radon), the short half-life means that samples need to be analyzed by the laboratory within a short time (usually a few days) after samples are collected. This requires careful planning. Although most environmental tracers only require collection of relatively small sample volumes (< 1 litre), this can vary depending on analytical method, and some environmental tracers require much larger volumes (e.g., 39Ar, 81Kr). In some aquifers, piezometers will not yield enough water for these analyses. Where specialized sampling equipment is required, it can usually be obtained from the laboratory where the analyses are to be performed.
For groundwater dating, analytical costs are similar for most tracers (around US$300-400 per sample in 2020). Exceptions are 36Cl and the isotopes of noble gases (39Ar, 81Kr, 85Kr), which are more costly, in part because of the scarcity of analytical facilities. However, these tracers offer some advantages over other tracers, as the noble gases are non-reactive in all environments, and 36Cl and 81Kr are the only available tracers for ages beyond 50,000 years. 222Rn is cheaper than the other age tracers (US$50-$150), but only provides ages for groundwater less than several weeks old. This limits its application in most environments. Field-portable analytical units are available for 222Rn, 2H, 18O and some of the noble gases.
Ancillary Information Requirements
Interpretation of some tracers is assisted by the availability of ancillary information. For example, dissolved gases such as CFCs and SF6 require information on recharge temperature, the amount of excess air and salinity, to calculate tracer solubility in water. Temperature and excess air can be estimated with concurrent N2 and Ar measurements, or more precisely, with measurements of other noble gases. For relatively high solubility tracers such as CFCs, recharge temperature can sometimes be estimated from air temperature data, although more precise ages can be determined using noble gas data. This information is not needed for 81Kr, 85Kr or 39Ar, because their ages are based on isotope ratios, rather than chemical activity. Local measurements of atmospheric levels of 85Kr and 3H concentrations in precipitation can improve the accuracy of these methods. 2H and 18O studies will benefit from collection of rainfall data, particularly if the site is remote from an existing rainfall measurement station. Information on carbonate chemistry of the aquifer solids and soil gas (including 13C) will greatly improve the ability of 14C data to accurately constrain groundwater residence times. Dissolved oxygen measurements can provide useful information regarding the potential for degradation of CFC-11. Information on production rate is needed for accumulating tracers.
The Potential for Site Contamination
Several of the environmental tracers used as event markers for estimation of groundwater age are atmospheric contaminants. Their use as groundwater age indicators relies on either their atmospheric concentrations being relatively uniform (e.g., CFCs, SF6, 85Kr), or the time of maximum atmospheric concentration being the same over large areas (e.g., 3H). There is the potential for these tracers to be present as local contaminants, either in the groundwater or through locally elevated atmospheric concentrations. Elevated atmospheric concentrations of CFCs can occur in large urban areas or close to industrial facilities (Ho et al., 1988), and groundwater contamination can also occur at urban and industrial sites. Low solubility gases can also be contaminated by well development using compressed air, particularly in low permeability formations (Figure 50). There is also potential for contamination of 85Kr at radioactive waste disposal sites or at sites close to nuclear reactors or nuclear test facilities, and for contamination of 3H at landfills, nuclear facilities and disposal sites. The use of CFCs as groundwater dating tools is severely limited where groundwater is contaminated by urban or industrial sources.
Given the potential limitations of some tracers, and the difficulty of predicting their limitations at specific sites, it is useful to analyze for more than a one tracer. Measuring multiple age tracers can also provide information on mixing processes, that would not be apparent from measurement of a single tracer.
Table 3 – Advantages and limitations of different environmental tracers for groundwater dating.
|Tracer||Approximate Age Range||Advantages||Limitations|
|222Rn||1 – 10 days||Completely unreactive||Sample must be delivered to the laboratory rapidly, due to the short half life.|
|3H||Provides indication of presence of 1950s 1960s precipitation.||Inert, and easy to sample. Reliable indicator of the presence of young (<70 year) precipitation.||Cannot provide precise age estimates without parallel 3He measurements.|
|3H/3He||0.1 – 70 years||In suitable environments, can have high precision over a large age range.||Highly sensitive to mixing, and so precise age estimates require short well screens.|
|CFC-12||1950 – present. Non-unique ages since mid 1990s.||Most inert of three common CFCs.||Contamination is commonly observed in urban and industrial environments.|
|CFC-11||1950 – present. Non-unique ages since late 1980s.||Usually included with analysis of CFC 12 and can provide useful supporting information.||Contamination is commonly observed in urban and industrial environments; Degrades in anaerobic environments.|
|CFC-113||1970 – present. Non-unique ages since early 1990s.||Sometimes included with analysis of CFC 12 and can provide useful supporting information.||Contamination is commonly observed in urban and industrial environments; Can sorb onto aquifer materials.|
|SF6||1970 – present.||Monotonically increasing atmospheric concentrations. Largely unreactive.||Highly sensitive to excess air; Subsurface production can be locally important.|
|85Kr||1950 – present.||Monotonically increasing atmospheric concentrations; Completely unreactive; Insensitive to recharge temperature and excess air.||Analysis is relatively costly.|
|39Ar||50 – 1000 years||Completely unreactive.||Requires large sample volume; Subsurface production can sometimes occur.|
|14C||200 – 30,000 years||Most widely available tracer for this time range.||Interaction with carbon in aquifer materials can complicate interpretation.|
|36Cl||50,000 – 1,000,000 years; Can also provide indication of presence of 1950s – 1960s precipitation.||50,000 – 1,000,000 years; Can also provide indication of presence of 1950s – 1960s precipitation.||50,000 – 1,000,000 years; Can also provide indication of presence of 1950s – 1960s precipitation.|
|81Kr||50,000 – 1,000,000 years||Completely unreactive.||Requires very large sample volumes.|