Investigation of Groundwater Flow Variations Near Spreading Ponds with Repeat Deliberate Tracer Experiments

Transcription

1 ABSTRACT & POWERPOINT PRESENTATION Investigation of Groundwater Flow Variations Near Spreading Ponds with Repeat Deliberate Tracer Experiments Jordan F. Clark, Ph.D. Professor, Dept. of Earth Science and Program of Environmental Studies University of California, Santa Barbara Santa Barbara, California Managed Aquifer Recharge Symposium January 25-26, 2011 Irvine, California Symposium Organizers: National Water Research Institute Orange County Water District Water Research Foundation

2 Investigation of groundwater flow variations near a spreading pond with repeat deliberate tracer experiments Jordan F Clark Dept of Earth Science, University of California, Santa Barbara, CA At ISMAR-06 symposium (Phoenix 2007), Fox et al. presented a numerical study of flow beneath the Mesa spreading ponds (AZ) to set criteria for the installation of a groundwater mound monitoring well. Using a geostatistical technique to characterize the subsurface heterogeneity, they tested more than 50 possible hydrogeological fields. Their results demonstrated that the combination of subsurface heterogeneity and wetting cycles (when and where recharge occurred) greatly influenced subsurface travel times to potential well locations directly below the ponds. Fox et al. study has significant implications for results from deliberate tracer studies that are generally conducted during one hydraulic condition. This paper compares the results of repeat tracer experiments at Kraemer Basin (Orange County, CA) and Rio Hondo Spreading Ponds (Los Angeles County, CA). Results from the first experiment at Kraemer Basin, which was conducted in Oct 1998, showed a region of highly transmissive material extending down gradient from the basin for more than 3 kilometers. Mean groundwater velocities were determined to be approximately 2 km/yr in this region. A second experiment was initiated in Jan 2008 to determine if the travel times from this basin to monitoring and productions wells changed during the past decade in response to new boundary conditions. Results indicate that flow near Kraemer was stable and travel times to most wells (6 out of 9) determined during both experiments agree within 3 weeks, the experimental uncertainty. Very similar results were observed near the Rio Hondo Spreading grounds where experiments were conducted in Winter 2003, Spring 2005, and Winter Travels times based on the first detections to unaltered wells agree within a few weeks during the different experiments. It is likely that the tracer experiment results differ from Fox et al. because of the larger spatial scales examined at Californian MAR locations compared to their modeling study.

3 Investigation of groundwater flow variations near spreading ponds with repeat deliberate tracer experiments Jordan F. Clark Dept. of Earth Science University of California, Santa Barbara

4 Where is the recharge water? The need to understand groundwater flow. Hydraulic connections, Flow paths, GW velocities, and mixing between native and recharge water Monitor water quality changes Biogeochemical reaction rates Permitting of sites in California with a reuse component 6 month retention

5 Geologic Effects on Travel Time Spreading Pond K 1 K 3 > K 2 Well will mix the different Production well withcomponents multiple long screens How can these direct fast flow paths be quantified? K 1 Weeks/Months K 2 K 3 Months/Years Years/Decades/Centuries After T. Johnson, WRD

6 Investigating Travel Time Three independent methods used Hydrogeologic interpretations / calculations Use known aquifer characteristics to estimate water velocities and travel times Transient/intrinsic tracers Geochemical dating techniques Deliberate tracers Physically add tracer to the surface water followed by sample collection at wells Best method for mapping direct fast flow paths

7 Deliberate tracers--experimental Design Inject sulfur hexafluoride (SF 6 ) or noble gas isotope into recharge water Bubble with metering valve Difficult to maintain steady flow Inefficient, only 1%-20% will dissolve Inject pre-saturated solution Monitor tracer concentration in surface water Gas loss via exchange with atmosphere is important Typically, air-water transfer gas rate is <4 cm/hr Empirically defines tracer input to groundwater Periodically sample wells Breakthrough Curves

8 Laboratory Column Experiments 3 H 2 O Kr (H=18) SF 6 (H=170) C/C 0 C/C Pore Volumes Pore Volumes Trapped air acts to retard the transport of gases Vulava et al (ES&T, 2002)

9 SF 6 Tracer Experiment Methods

10 SF 6 Breakthrough Curves-WRD First Arrival Tail Local Peaks First Arrival - fastest flow paths; usually defined by first detection Tail & local peaks - Result of trapped air or multiple flowpaths?

11 Need for Repeat Experiments? Travel times and flow paths should reflect hydraulic conditions at the time of the experiment Hydraulic conditions at recharge basin should vary Seasonal changes in recharge and pumping Interannual variations due to availability of water, etc. Long term changes in basin management So, how robust are the results?

12 Results Repeat Experiments Orange County Kraemer Basin Long term transport through high hydraulic conductivity zone Oct 1998 Noble Gas vs Jan 2008 SF 6 Experiments Montebello Forebay (LA County) Near field transport and the importance of depth rather distance as the primary factor controlling travel time Feb 2003, June 2005, and Jan 2009 SF 6 Experiments

13 OCWD Northern Basins Kraemer Basin Oct 1998 (LLNL) Exp 136 Xe for ~ 7 days δ 18 O for ~ 28 days Recharge Rate = ~80 cfs SAR January 2008 Exp SF 6 for ~14 days Recharge Rate = ~70 cfs [SF 6 ] = 66 pmol/l Detect. Limit = 0.5 pmol/l

14 Am7 Am KB1 1.0 Results from 1998 Tracer Experiments 136 Xe & δ 18 O SF Kilometers 0 2 Travel time from Kraemer Basin Travel time from the Santa River SAR experiment wells K.B. experiment wells SAR and K.B. Experiment wells Mean Arrival -12 δ 18 O ( ) δ 18 O 136 Xe KB Time (yr) Am7 Clark et al., Ground Water, Xe (10-14 mol/mol)

15 1998 vs 2008 Results

16 Montebello 2003 Experiment Travel Times to Nearby (<150 m) Wells 18 production wells Screen depth: m Majority > 47 m Screen lengths: m 10 monitoring wells Screen depth: m Majority < 15 m Screen length: 6-23 m Challenge: Spike every basin at nearly the same time for ~2 weeks Basin Volume = 5 x 10 6 m, recharge rate ~1 m/day

17 Depth below Nearest Pond (m) Experimental Results: SF 6 All wells within 150 m Production Monitoring y = x R 2 = SF 6 Arrival Time (yr) McDermott et al., J Hydrol. Eng., 2008

18 WRD 2003 Tracer Experiment Findings SF 6 and hydrogeologic arrival times (Bookman report) are a function of depth, not distance (within 150 m). SF 6 arrival times are shorter then hydrogeologic travel times at some production wells: SF 6 arriving from preferential flow paths or discontinuities in the clay layers Possible leaky confining units A few (4 of 18) production wells have travel times <6 months and were not in compliance with reuse regulations Would the travel time increase by simply deepening the well?

19 Conceptual Model of Flow near the San Gabriel Spreading Grounds Spreading Pond K 1 Can we block this shallow flow path? > 6 Months based on on SF 6 results Production Well With Long Screen Install packer K 2 K 3 Months/Years Years/Decades/Centuries After T. Johnson, WRD

20 2005 SF 6 Experiment Packer placed in Well # and therefore increased the depth to screen from 41 m to 60 m Travel time = 8 weeks in 2003

21 Results of Well Modification 0 Depth below Nearest Pond (m) # # Production Wells Monitoring Wells Monitoring Wells SF 6 Arrival Time (yr) McDermott et al., J Hydrol. Eng., 2008

22 2009 SF 6 Experiment Packer placed in Well # and therefore increased the depth to screen from 14 m to 60 m Travel time = 16 weeks in 2003 & 20 weeks in 2005

23 Rio Hondo Basin Tracer Results Arrival time of first detect (weeks) Well Feb 2003 June 2005 Jan AP 8 * BP AP 16 * 20 * BP >42 *Unmodified well Agreement: Good Poor

24 Rio Hondo Basin Tracer Results Arrival time of first detect (weeks) Well Feb 2003 June 2005 Jan AP 8 * BP AP 16 * 20 * BP >42 *Unmodified well

25 Rio Hondo Basin Tracer Results Arrival time of first detect (weeks) Well Feb 2003 June 2005 Jan Basic inference/observation that depth 2 is the most important 12 factor controlling arrival time was confirmed Increasing APdepth of screen 8 * brought 28 the production 16wells in BPcompliance with the reuse 26regulations AP 16 * 20 * BP >42 *Unmodified well Agreement: Good Poor

26 Conclusions At California spreading basins with long history of use and high recharge rates, gas tracer experiments can be used to determine travel times o Gas retardation due to trapped air is small Repeat tracer experiments yield similar but not the exact travel time information Close to spreading ponds, depth is the most important factor controlling arrival times

27 References 1. Clark, J. F., G. B. Hudson, M. L. Davisson, G. Woodside, and R. Herndon (2004) Geochemical imaging of flow near an artificial recharge facility, Orange County, CA. Ground Water, 42, Clark, J. F., G. B. Hudson, and D. Avisar (2005) Gas transport below artificial recharge ponds: Insights from dissolved noble gases and a dual gas (SF 6 and 3 He) tracer experiment. Environmental Science and Technology, 39, Avisar, D. and J. F. Clark (2005) Evaluating travel times beneath an artificial recharge pond using sulfur hexafluoride. Environmental and Engineering Geoscience, 11, Clark, J. F. (2006) Managing aquifer recharge: How can isotope hydrology help? Water & Environment News, Newsletter of the Isotope Hydrology Section, International Atomic Energy Agency, 21, Clark, J. F. and G. B. Hudson (2006) Excess air: A new tracer for artificially recharged surface water. In: Recharge Systems for Protecting and Enhancing Groundwater Resources, IHP-VI, Series on Groundwater No. 13, UNESCO, p Available at 6. Clark, J. F. (2007) Tracing Recharge Water from Spreading Ponds: A Decade of Studies. In: Management of Aquifer Recharge for Sustainability, ed. Fox, P., Acacia, Phoenix, p McDermott, J. A., D. Avisar, T. Johnson, and J. F. Clark (2008) Groundwater travel times near spreading ponds: Inferences from geochemical and physical approaches. Journal of Hydrologic Engineering, ASCE, 13,