- Top of page
- Results and Discussion
Starting in 2008, a 4-year tracer study was conducted to evaluate ambient changes in groundwater concentrations of a 1,3,6-naphthalene trisulfonate tracer that was added to drill water. Samples were collected under open borehole conditions and after installing a multilevel groundwater monitoring system completed with 11 discrete monitoring zones within dense and fractured basalt and sediment layers in the eastern Snake River aquifer. The study was done in cooperation with the U.S. Department of Energy to test whether ambient fracture flow conditions were sufficient to remove the effects of injected drill water prior to sample collection. Results from thief samples indicated that the tracer was present in minor concentrations 28 days after coring, but was not present 6 months after coring or 7 days after reaming the borehole. Results from sampling the multilevel monitoring system indicated that small concentrations of the tracer remained in 5 of 10 zones during some period after installation. All concentrations were several orders of magnitude lower than the initial concentrations in the drill water. The ports that had remnant concentrations of the tracer were either located near sediment layers or were located in dense basalt, which suggests limited groundwater flow near these ports. The ports completed in well-fractured and vesicular basalt had no detectable concentrations.
- Top of page
- Results and Discussion
A history of more than 60 years of waste disposal at the Idaho National Laboratory (INL) has resulted in measureable concentrations of contaminants in the eastern Snake River Plain (ESRP) aquifer. The aquifer underlies the INL and consists primarily of layered basalt (fractured and dense) intermixed with sediment layers of varying thickness and composition. The U.S. Geological Survey (USGS), in cooperation with the U.S. Department of Energy (DOE), uses multilevel monitoring systems (MLMSs) in select INL boreholes to examine the vertical extent of waste water within the ESRP aquifer (Figure 1). Other studies have highlighted the usefulness of MLMSs in fractured rock aquifer systems (Einarson and Cherry 2002; Einarson 2006; Parker et al. 2006; Cherry et al. 2007; Meyer et al. 2008).
The ESRP aquifer at the INL consists of about 85% basaltic lava flows that contain vesicular zones and cooling fractures on the top sides of the flows and massive cores toward the interiors of the flows, and 15% fluvial and eolian sediment, andesite flows, and rhyolite domes (Ackerman et al. 2010). Groundwater moves horizontally through basalt interflow zones and contact between basalt flows, and vertically through joints and fracture zones in the basalt. Horizontal flow velocities of 2 to 20 ft/d (0.6 to 6 m/d) have been calculated at the INL based on the movement of various constituents in different areas of the aquifer at the INL (Robertson et al. 1974; Mann and Beasley 1994; Cecil et al. 2000; Busenberg et al. 2001). Localized tracer tests have revealed vertical and horizontal transport rates as high at 60 to 150 ft/d (18 to 46 m/d) (Nimmo et al. 2002; Duke et al. 2007). The transmissivity of 94 wells at the INL ranges from about 1.1 to 760,000 ft2/d (0.1 to 70,600 m2/d) with an average transmissivity of about 93,000 ft2/d (8,640 m2/d) (Ackerman 1991, Table 3). Generally, monitoring wells completed at the INL penetrate fractured basalt layers and aquifer test data reflect high transmissivities (Ackerman 1991; Bartholomay et al. 1997); however, low transmissivities have been observed where monitor wells do not penetrate into basalt fractures or where fine-grained sediment layers fill fractured media. In general, the transmissivity of the basalts that comprise the aquifer in this study area are much higher relative to other aquifers in the United States (USGS 2000).
Between 2005 and 2007, six wells were equipped with MLMS; waste water contaminants were found in some level of the aquifer in all six MLMS-equipped wells (Bartholomay and Twining 2010). However, it is uncertain whether the results from four of the wells may have been influenced by drill water that was filled from a potable production well (CFA-1) in which some waste water constituents were previously reported (Davis 2010). To examine residence time for drill water within well USGS 108 after drilling (coring and reaming), a 1,3,6-naphthalene trisulfonate (136NTS) tracer was mixed into drill water that was continuously injected during recompletion of USGS 108. Injected tracer concentrations were measured, and water samples were collected from the open borehole (2008 to 2010) and from discrete intervals of the well (2010 to 2011) after installing the MLMS. Prior to installation of a MLMS in USGS 108, no well purging was performed and tracer detections and non-detections are the result of ambient groundwater movement through the ESRP aquifer.
The multilevel monitoring technology used in this study is the Westbay SystemTM, manufactured by Schlumberger Water Services (2012). The MLMSs installed at the INL are composed of two parts: (1) a modular borehole completion system consisting of packers, measurement ports, and variable length sections of polyvinylchloride casing (Figure 2) and (2) a portable sampling probe (MOSDAXTM) and data acquisition system (MAGITM). The MLMSs are engineered to measure hydraulic head and water temperature and to collect water samples at discrete depths as described in more detail by Fisher and Twining (2011) and Twining and Fisher (2012). Table 1 gives the location of measurement zones, ports, interval depths, and hydraulic head values for the MLMS installed in USGS 108.
Table 1. USGS 108 Multilevel Well Completions Used for 136NTS Sampling, Idaho National Laboratory
|Zone Number||Zone Depth Interval||Hydaulic Head (ft amsl)||Port Number||Port Coupling Depth (ft BLS)||Site Number|
|Bottom (ft BLS)||Top (ft BLS)||Length (ft)|
Of the five sample ports installed in this well, sample ports 5, 9, 12, and 15 (sample zones 4, 7, 9, and 11; Table 1) are in a location within the borehole where basalt fractures are prevalent which may suggest increased transmissivity and groundwater movement; however, sample port 1 (zone 1) is located close to a sediment layer and dense basalt where groundwater flow may be restricted, even though there is some fractured basalt located at the bottom of zone 1 (Figure 3). The groundwater residence time within zone 1 is expected to be much longer than the groundwater residence time for zones completed in the more fractured media with higher groundwater movement. Tracer data will be used to examine the residence time of drill water in discrete zones completed in various media within the MLMS and under open borehole conditions.
Figure 3. Geophysical and lithologic logs, multilevel port and zones locations, and tracer concentrations for USGS 108.
Download figure to PowerPoint
This paper describes the results of a field experiment in well USGS 108 at the INL that involved mixing a 136NTS tracer with drill water that was continuously used while USGS 108 was being cored and reamed in preparation for installing a MLMS. The experiment was designed to simulate drilling conditions used to complete other MLMS wells and to test the length of time drill water could remain detectable, and at what concentrations, under open hole conditions and after installing a MLMS. The study was needed to determine if contaminated drilling water likely affected water chemistry concentrations at four other MLMS wells at the INL. In theory, ambient flow in the high-transmissive fractured basalt should move all drill water away from the well within days or months; however, a MLMS can effectively seal off select discrete zones of much lower transmissivities resulting in longer groundwater residence times.
Open borehole thief samples were collected after coring at two selected depths, near 660 ft (201 m) and 780 ft (238 m) below land surface (BLS), during four sampling events (October 29, 2008 to May 10, 2010); one sample was collected after reaming on June 1, 2010 and before installation of a MLMS in USGS 108. The MLMS was installed during August 2010, and groundwater samples were collected during four sampling events from up to 10 discrete monitoring zones that included sampling 14 of 16 measurement ports (Figure 3). Forty-seven samples were collected over the course of 4 years; 11 quality control samples were also collected and analyzed.
Geophysical logs and core collected from USGS 108 were used to examine basalt and sediment properties and to verify zones of groundwater movement. Detailed core descriptions are documented in Hodges et al. (2012); cores were used to verify areas of dense basalt, fractured basalt, and sediment interpreted from geophysical data. The geophysical data that were evaluated included:
- Natural gamma is a measure of the gamma radiation emitted by the naturally occurring radioisotopes within the rock material composing the borehole wall. Elevated natural gamma readings at the INL typically indicate the presence of a sedimentary layer.
- Neutron is a measure of the hydrogen content of the rock, which, when saturated, is directly related to the porosity of the porous medium. A high neutron porosity indicates the presence of highly fractured basalt or sediment; whereas, a low neutron porosity indicates an area of dense basalt.
- Left and right caliper uses three extendable, spring-loaded arms to measure drill-hole diameter with an accuracy of ±0.15 in. Changes in the regular drill-hole diameter may be due to collapse of the loose or highly fractured rock formations—areas unsuitable for packer placement.
- Short-spaced and long-spaced gamma-gamma dual density, also known as the induced gamma-density, is a measure of the bulk density of a rock material near a borehole wall. The bulk density of a rock material is inversely related to its porosity.
Results and Discussion
- Top of page
- Results and Discussion
Results from open borehole thief samples collected after coring indicate that small amounts of tracer (about 6 and 8 µg/L at 660 ft [201 m] and 780 ft [238 m] BLS, respectively) remained in the aquifer 28 days after coring was completed (Table 2). These initial results were not expected within the upper 150 ft (46 m) of the aquifer (top of the water table is about 612 ft [187 m] BLS) where transmissivity estimates were 150,000 ft2/d (13,935 m2/d) (Ackerman 1991, Table 3). It was believed that the high transmissivity in the upper part of the aquifer would allow sufficient ambient flow to remove the tracer from the open hole. A second set of samples was collected about 7 months later (June 1, 2009); the tracer was not detected in these samples or in subsequent samples collected on December 2, 2009 and May 10, 2010 prior to reaming the well (Table 2). The USGS began reaming the well to prepare it for MLMS installation in May 2010. The tracer mixed in drill water used in May 2010 was not detected in the aquifer in two thief samples collected on June 1, 2010 after reaming was completed (Table 2). Thief samples were not collected from deeper in the aquifer after coring because of a rock collapse in the hole near 860 ft (262 m) BLS. The rock collapse also prevented the collection of flowmeter readings. The initial presence of the tracer after coring but not after reaming could have resulted because about three times more water was used over twice as many days in the coring process versus the reaming process.
A MLMS was installed in USGS 108 during late August 2010. Temporary protective casing was placed in USGS 108 after the sample was collected after reaming to keep the borehole open for MLMS installation. Where the sections of aquifer penetrated by USGS 108 are comprised of fractured basalt (Figure 3), it is assumed that groundwater flows horizontally through the fractured basalt. Hydraulic head information presented by Twining and Fisher (2012) for USGS 108 suggests there is the potential for upward flow between zone(s) 3 and 11 then shifting to downward flow near the top of zone 2 (Table 1). The hydraulic head range for USGS 108 is 0.80 ft (0.24 m) over 550 ft (168 m), suggesting that there is also some vertical flow within the borehole.
Samples collected during September and November 2010 indicate concentrations of tracer in zone 1 (port 1) were 15.6 and 17.4 µg/L, respectively; whereas, zone 7 (port 9) concentrations were 0.72 and 6.3 µg/L, respectively (Figure 3). The drill water had an average 136NTS concentration of about 10,200 µg/L (Table 2) and samples collected within zones 1 and 7 show significant dilution; however, tracer concentrations within these two zones did increase slightly with time where other zones were less than the detection limit during both sample events (Figure 3). The primary difference for zones where tracer was detected was the occurrence of fine grained sediment near the sampling ports (Figure 3).
Examination of the core and geophysical data for USGS 108 indicates that sediment layers were present in both zones 1 and 7, but that no sediment layers were present in the other three zones sampled for the first two rounds (Figure 3). The presence of sediment probably slows horizontal groundwater movement because of filling and/or plugging factures with fine material resulting in lower transmissivity and lengthened groundwater residence times under ambient conditions. The tracer data collected during 2010 (Figure 3) suggest the sediment layers may retain the tracer longer than zones completed in fractured basalts but it was unclear whether this was related to low transmissivity within discrete zones that contain sediment or if groundwater containing tracer was the result of sorption/desorption in fine grained sediment layers. Testing for sediment tracer sorption was not conducted as part of this study.
To test whether sediment layers retain tracer within the matrix and slowly release detectable concentrations, two additional rounds of samples were collected in 2011. These samples included every zone within USGS 108 MLMS, except zone 10 where casing restricts groundwater flow (Figure 3). During both sample periods, concentrations were detected in zone 1 (ports 1 and 2), zone 2 (port 3), zone 4 (port 6), and zone 6 (port 8), but were not detected in zone 7 (port 9) where concentrations were detected during the first two rounds of sampling (Figure 3). The results from samples collected in 2011 were not as expected because tracer was detected within three zones with no sediment (Figure 3); however, the ports where tracer was detected were located adjacent to dense basalt layers and include zone 2 (port 3), zone 4 (port 6), and zone 6 (port 8). In addition, no tracer was detected in zone 3 (port 4), zone 5 (port 7), and zone 8 (port 11) where sediment layers were present. The zones with sediment, but no detections, all had sediment interspersed with fractured lava and the ports were not located directly adjacent to the sediment layers, so ambient flow may be moving the tracer away from these zones. The presence of tracer in the first two rounds in zone 7 (port 9) but not in the subsequent rounds may indicate that some sorption/desorption may be occurring. The results from the 2011 samples suggest that sediment layers and areas of dense basalt should be considered when predicting the groundwater residence time for discrete zones completed with MLMS.
Assuming dense basalt restricts horizontal and vertical flow and fractured basalt allows horizontal flow movement away from the well, it would be expected that ports responding to the annular space adjacent to fractured basalt would not have tracer present while ports near dense basalt might have tracer present. Sediment layers may also restrict horizontal flow depending on the interbed composition and have also been shown to restrict vertical flow in many of the MLMS at the INL (Fisher and Twining 2010; Twining and Fisher 2012). Geophysical data and core descriptions show that every port that had tracer present was completed either in dense basalt (ports 2, 3, 6, and 8) or in immediate proximity to sediment layers (ports 1 and 9) (Figure 3). Zone 4 was completed with paired ports, two ports within one discrete zone, where port 5 was located in fractured basalt and did not show tracer and port 6 was located directly below a packer bladder and in dense basalt that did show tracer. Where the tracer detection show opposite results within the same zone, one detect and one nondetect, we consider that port 5 is located in fractures with horizontal groundwater flow probably occurring that removes the tracer whereas, slow diffuse flow from the matrix of the dense basalt may account for the tracer observed from port 6. Slow diffuse flow may also account for the presence of tracer in ports 2, 3, and 8; but several ports completed in dense basalt (ports 4, 7, 11, and 13) showed no tracer present, so water flow characteristics of the various dense basalt materials along with possible vertical water movement within zones are needed to fully assess the matrix effects of the basalt in relation to tracer concentrations.
The two deepest zones (sample ports 1 to 3) show similar tracer concentration with each sample period. Zone 1 (ports 1 and 2) and zone 2 are located at the base of the borehole where core and geophysical data indicate mostly dense basalt and sediment. There is a fractured interval at the bottom of zone 1 (Figure 3); however, sluff material from installation may be limiting groundwater movement within this fractured media. The dense basalt layer, starting near 1120 ft (341 m) BLS, restricts hydraulic connection with zones above and limits horizontal flow within zone 2. Furthermore, a sediment layer located within zone 1 is considered as a possible source of tracer detection within zone 1; however, the tracer detections are likely the combination of both dense basalt and sediment present in zone 1. Because of low transmissivity and limited groundwater flow in both zones 1 and 2, the tracer probably hangs around longer in these two zones and possible absorption/desorption is greater than the advection away from the ports which may account for the similar concentrations.
Tracer concentrations were also detected within zones 4, 6, and 7. Measurement ports where tracer was detected within zones 4 and 6 are located less than 0.6 ft (0.2 m) below the packer bladders and within dense basalt (Figure 3). Concentrations in water samples collected on November 10, 2011 from zones 4 and 6 were approximately 33 and 63% less, respectively, than concentrations found in samples collected on June 22, 2011 (Figure 3). Where a measurement port is located adjacent to a packer in an area of dense basalt, it is likely that groundwater moves very slowly directly below the packer. The detection of decreasing tracer concentrations within zone 4 (port 6) and zone 6 (port 8) suggest that slow diffuse flow from the matrix of dense basalt is occurring and tracer is slowly being removed. Tracer samples within zone 7 (port 9) show tracer detections during 2010; however, follow-up sampling in 2011 show no detectable tracer concentrations (Figure 3). Ambient groundwater flow within zone 7 (port 9) likely was slowed by the presence of two sediment layers, described as calcite cemented silt and clay in composition (Hodges et al. 2012, Appendix C), which probably slowed the initial movement of the tracer out of the formation.
In order to assess if contaminated drill water used at four wells at the INL could be affecting post-drilling water chemistry concentrations (Bartholomay and Twining 2010), 136NTS tracer concentrations from this study were compared with tritium results from the four wells. For example, the initial drill water from CFA-1 used to drill four wells (USGS 103, 132, 133, 134, Figure 1) between 2006 and 2007 at the INL had tritium concentrations that averaged about 7450 pCi/L (USGS data accessed on June 11, 2013 from the web at: http://nwis.waterdata.usgs.gov/usa/nwis/qwdata/?site_no=433204112562001); whereas the tracer used here had an average concentration of about 10,200 µg/L (Table 2). Tritium concentrations above the 200 pCi/L detection level in the four wells had concentrations that ranged from 210±70 to 530±70 pCi/L for samples collected from 1 to 3 years after drilling (Bartholomay and Twining 2010) while the tracer detected in sample ports in this study ranged from 0.72 to 117.5 µg/L. The sample ports in the four MLMS wells at the INL with tritium present were all installed in zones with fractured basalts except for the zone 644 ft BLS in USGS 134 (Twining and Fisher 2010) which was completed in massive basalt. Given that the tracer used in this study probably does not move as readily in groundwater as tritium and that aquifer characteristics are similar between the MLMS wells (contain fractured basalt, dense basalt, and sediment layers), the tritium concentrations are not affected by drill water in the other 4 wells except possibly in the zone 644 ft BLS in USGS 134.