Development of New Tracers to Determine Groundwater Travel Time near Managed Recharge Operations
Type: Scientific Investigation
Year Released: 2015
Funding Partners: Bureau of Reclamation, Orange County Water District
Total Investment: $150,875 (Cash: $130,875, In-Kind cash and service: $20,000)
Principal Investigator: Jordan F. Clark, University of California, Santa Barbara
Water quality studies near MAR operations have shown that subsurface residence time is one of the most important parameters for potential contaminant removal. Current SWRCB DDW regulations require underground retention times of 6 to 24 months from infiltration to extraction for potable use. The development of new tracer methods that require minimal field and laboratory work, and that can resolve subsurface travel times on timescales of <1 year, is paramount for protecting public and environmental health.
Goals and Objectives
Study objectives were to assess three alternative tracers to replace sulfur hexafluoride (SF6), a synthetic gas and the principle deliberate tracer for work near MAR sites. SF6 is a strong greenhouse gas and its emission is being regulated in California. Two natural (intrinsic) tracers were evaluated: heat (with recharging water ~10° C warmer than native groundwater) and radio-sulfur (35S). Also evaluated were the added (deliberate) tracers10B, as 10B-enriched boric acid, and conservative bromide (Br–).
35S was evaluated by measuring natural levels of 35S activity in dissolved sulfate in MAR source water and groundwater. Travel times were estimated using the radioactive decay equation and compared to travel times quantified by previous deliberate tracer studies at the Rio Hondo and Orange County recharge sites.
10B, Br–, and heat transport were assessed during a single experiment at the USGS research test basin at San Gabriel Spreading Grounds. Sampling occurred at nine monitoring wells in a line down gradient from the basin. Heat, requiring no artificial input, was interpreted from temperature changes recorded hourly at high-sensitivity well loggers.
Findings and Conclusions
The 35S method was successfully applied at RHSG due to high 35S activity in the recharge source waters; however, its application was limited at OCWD MAR facilities due to consistently low 35S activity in source waters. Compared to the travel times determined by a SF6 deliberate tracer experiment, 35S travel times at the RHSG were within the range of six weeks at four of the six monitoring wells. Production wells travel times were also in agreement for both tracer experiments. Overall, the application of the intrinsic 35S tracer at this site provided reasonable estimates of subsurface travel times.
The usefulness of this technique is highly dependent on the 35S activity in MAR source waters, especially for MAR facilities using recycled water since recycled water is likely to have high concentrations of SO4 and low 35S activity. In order to quantify 35S subsurface travel times of 9 month (3 half-lives) for transport of recharge water to nearby wells, surface water 35S activity should ideally be 8 times above the minimal detectable activity (MDA). Based on the minimal detection limits determined in this study (0.5 to 3.4 mBq/L), an input end-member of >20 mBq/L is recommended for application of the 35S tracer method at other MAR facilities. Careful characterization of the 35S activity in MAR source waters is especially important when determining the feasibility of using 35S as a natural (intrinsic) tracer at these sites.
10B and Br– tracer breakthroughs were observed at six and at seven of nine monitoring wells, respectively. Subsurface residence times ranged from 0.7 to 40 days. 10B arrived 25% later than Br– on average, indicating retardation most likely through exchange with clay surfaces. As a result, travel times may overestimate actual groundwater flow time, especially in wells far (in depth and lateral distance) from the injection point. This is a weakness of the boron tracer, but may be a relict of sampling bias caused by a relatively poor resolution. Regulations could adapt to our finding by, for instance, requiring a longer travel time (1 year instead of 6 months) to ensure the public potable supply. The advantage of isotopic tracers is that significantly less mass is needed for the same volume of water and detection is easier because of non-linear mixing. 10B measurements were also made on the more-affordable ICP-MS system, with analytical uncertainty better than ±15‰ δ11B. Deliberate tracer studies are commonly hindered by (1) mass needed, with large projects approaching unrealistic costs, raising mixing concerns in the tagged basin and the potential to affect flow, and (2) poor sampling resolution, limited largely by manpower hours and analytical expenses.
Recycled water is warmed at sewage treatment plants and after being transferred to a spreading pond, it can acquire the local diurnal heating/cooling trends. Native groundwater not influenced by MAR operations averaged between 17.9 and 25.0 °C depending on well depth and rarely fluctuated more than 0.5 °C. Infiltrating recycled water averaged 28.9° during early September and 25.6° by mid-November. Reliable residence times to all nine wells were successfully determined from temperature changes. The longest flow path was around 6 months. Diurnal peak matching is ideal as it allows for precise travel times that can be calculated many times during the recharge event, but was only observed at one well in the near-field. Most wells recorded a warming period that spanned days to weeks as the warm recharge plume arrived at each well. The warming period was defined here to begin when a well logger recorded temperatures 0.5 °C above that of background. The period ends as temperatures plateau, indicating steady-state conditions between the continuously recharging recycled plume and native groundwater. Even without a diurnal signal, heat flow interpretations yielded travel times to all wells within the same range as the geochemical tracers. Moreover, whereas the added tracers became too dilute to detect above background at two wells, temperature changes were successfully measured. This study showed that heat has great potential as an intrinsic tracer at MAR facilities.