5
Monitoring and Data Quality

Implementing a groundwater monitoring plan includes four elements: drilling wells, completing the wells, obtaining groundwater samples from the wells, and analyzing the samples. Monitoring is done to measure the extent of contaminant migration along expected pathways or to determine that the water is free of contamination. Monitoring is the only direct means to confirm models and predictions about subsurface contaminant transport and to provide early warning of potential contamination in drinking water supplies.

This chapter deals with the actual practices of conducting monitoring—in particular ensuring that the Los Alamos National Laboratory’s (LANL’s) groundwater monitoring data are reliable. The first part of the chapter deals with LANL’s well construction work. The second deals with the specific data-quality questions presented in the committee’s task statement. LANL’s understanding of contaminant pathways, which is essential for developing a monitoring plan, is discussed in Chapter 4.

The data-quality questions raised in the committee’s task statement are:

  1. Is the laboratory following established scientific practices in assessing the quality of its groundwater monitoring data?

  2. Are the data (including qualifiers that describe data precision, accuracy, detection limits, and other items that aid correct interpretation and use of the data) being used appropriately in the laboratory’s remediation decision making?

The short answer to the first item is a qualified yes. LANL is using good practices in terms of having the proper quality assurance and quality control (QA/QC) plans and documentation in place, but falls short of consistently carrying out all the procedures cited in the plans. Well drilling and completion methods are continuing to evolve, and the site is only beginning to implement its groundwater monitoring program under the Consent Order. Many if not all of the wells drilled into the regional aquifer under the Hydrogeologic Workplan appear to be compromised in their ability to produce water samples that are representative of ambient groundwater for the purpose of monitoring.

The short answer to the second question, as it is written, is no. Although LANL appears to be generating sound analytical data, the results reported in databases and LANL reports often do not carry the proper qualifiers according to good QA/QC practices. This especially applies to analytical results near or below the limits of practical quantitation and detection, near the natural background, or both. The difficulty here is that reported detection of contamination that is not statistically significant may be taken as real by regulators and other stakeholders—with concomitant concerns and calls for remedial actions.

WELL CONSTRUCTION

LANL will continue to construct water wells for at least three purposes. Each purpose has implications for the drilling and completion methods selected, as follows:

  1. Characterization: Characterization of the site’s hydrogeology and subsurface contamination in soil and groundwater at LANL is far from complete. Drilling for characterization can be relatively quick and inexpensive to survey hydrogeologic conditions over large areas. However, characterization can also become slower and more expensive if data needs include, for example, detailed identification of perched water zones, collecting core or cuttings for chemical analyses, and performing geophysics. The latter was more generally the case for characterization wells under the Hydrogeologic Workplan (LANL, 1998a).

  2. Monitoring: Monitoring wells are designed and constructed to minimize their own effects on the groundwater that they are intended to monitor, and



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report 5 Monitoring and Data Quality Implementing a groundwater monitoring plan includes four elements: drilling wells, completing the wells, obtaining groundwater samples from the wells, and analyzing the samples. Monitoring is done to measure the extent of contaminant migration along expected pathways or to determine that the water is free of contamination. Monitoring is the only direct means to confirm models and predictions about subsurface contaminant transport and to provide early warning of potential contamination in drinking water supplies. This chapter deals with the actual practices of conducting monitoring—in particular ensuring that the Los Alamos National Laboratory’s (LANL’s) groundwater monitoring data are reliable. The first part of the chapter deals with LANL’s well construction work. The second deals with the specific data-quality questions presented in the committee’s task statement. LANL’s understanding of contaminant pathways, which is essential for developing a monitoring plan, is discussed in Chapter 4. The data-quality questions raised in the committee’s task statement are: Is the laboratory following established scientific practices in assessing the quality of its groundwater monitoring data? Are the data (including qualifiers that describe data precision, accuracy, detection limits, and other items that aid correct interpretation and use of the data) being used appropriately in the laboratory’s remediation decision making? The short answer to the first item is a qualified yes. LANL is using good practices in terms of having the proper quality assurance and quality control (QA/QC) plans and documentation in place, but falls short of consistently carrying out all the procedures cited in the plans. Well drilling and completion methods are continuing to evolve, and the site is only beginning to implement its groundwater monitoring program under the Consent Order. Many if not all of the wells drilled into the regional aquifer under the Hydrogeologic Workplan appear to be compromised in their ability to produce water samples that are representative of ambient groundwater for the purpose of monitoring. The short answer to the second question, as it is written, is no. Although LANL appears to be generating sound analytical data, the results reported in databases and LANL reports often do not carry the proper qualifiers according to good QA/QC practices. This especially applies to analytical results near or below the limits of practical quantitation and detection, near the natural background, or both. The difficulty here is that reported detection of contamination that is not statistically significant may be taken as real by regulators and other stakeholders—with concomitant concerns and calls for remedial actions. WELL CONSTRUCTION LANL will continue to construct water wells for at least three purposes. Each purpose has implications for the drilling and completion methods selected, as follows: Characterization: Characterization of the site’s hydrogeology and subsurface contamination in soil and groundwater at LANL is far from complete. Drilling for characterization can be relatively quick and inexpensive to survey hydrogeologic conditions over large areas. However, characterization can also become slower and more expensive if data needs include, for example, detailed identification of perched water zones, collecting core or cuttings for chemical analyses, and performing geophysics. The latter was more generally the case for characterization wells under the Hydrogeologic Workplan (LANL, 1998a). Monitoring: Monitoring wells are designed and constructed to minimize their own effects on the groundwater that they are intended to monitor, and

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report hence to provide samples that are truly representative of the actual groundwater. Monitoring wells include wells upgradient of disposal locations to establish a baseline composition of the natural groundwater, downgradient wells for early detection of migration toward receptors or compliance boundaries, and near-source wells to monitor known contaminant movement or demonstrate the effects of remediation strategies. Remediation—for example, wells to pump contaminated groundwater out of an aquifer so that the water can be treated and, often, returned to the aquifer: This application was not discussed by LANL during the committee’s study period. In meetings with the committee, LANL emphasized that well design, drilling methods, and well development—particularly for the approximately 1000-foot-deep wells that reach the regional aquifer—are evolving (Broxton, 2006). The committee considers this evolution an important and essential part of the program. Drilling Methods Drilling is the means of penetrating into the Earth’s surface to access the underlying geological formations for study and/or to physically sample groundwater. LANL’s drilling program under the Hydrogeologic Workplan considerably expanded LANL’s ability to sample and characterize the Pajarito Plateau; see Color Plates 9 and 10. The drilling work itself, however, had a long and difficult evolution, including technical problems, unexpected high cost, and inconsistent objectives. Further, the drilling efforts shifted from primarily characterization toward a multiple use approach that included using a single borehole for both characterization and monitoring (Nylander, 2006). LANL sought and received external review and advice during the course of this work as noted in Chapter 2. The very act of drilling always damages to varying degrees the geologic formation penetrated by the borehole. This can lead to temporary or sometimes permanent changes in the hydrogeologic and geochemical properties of the formation and the nearby groundwater. Drilling a groundwater monitoring well to 1000 feet while inflicting as little permanent damage to the formation as possible is a technical challenge. Successful drilling is very site specific with heavy reliance on the expertise of the drilling personnel. Although there are many drilling methods, see Table 5.1, the use of rotary drilling (i.e., drilling with a rotating drill bit) is the most common. All rotary drilling methods require the use of a fluid to clear the drill bit of cuttings, to cool the bit to prolong its usefulness, and sometimes to keep the formation around the hole from collapsing before the well is completed. There are many types of drilling fluids—including air, water, and “muds,” which may be clays and/or synthetic materials—and additives to improve properties of some fluids. Depending on the formation, purpose of the well, and available drilling equipment, drillers may use a variety of fluids and additives. Broxton (2006) and Nylander (2006) describe efforts by LANL, the Department of Energy (DOE), and their drilling contractors to install the deep wells into the regional aquifer (R-wells) required by the Hydrogeologic Workplan. Air and/ or water were found to be inadequate as drilling fluids due largely to the depth to be drilled and the instability of some formations to be drilled through, although procedural errors have also been cited (Gilkeson, 2007). Lack of lubrication and the tendency of the boreholes to collapse resulted in slow progress and instances of stuck drill pipe and bits. Broxton (2006) lists a total of over 2600 feet of stuck drill pipe abandoned in place in 8 R-wells. As a result of these experiences, more traditional fluids—municipal water with organic chemical additives (EZ-Mud® and Quik-Foam®)—were used in most of the R-wells.1 In eight of the R-wells, bentonite mud was used as the drilling fluid for at least part of the well depth (LANL, 2005d). Completion Methods Completion refers to steps that convert a borehole to a well. Once the borehole is drilled to its planned depth, the drilling tools are removed, and the screen and casing are lowered into the hole. If an outer casing has been used to keep the borehole open as the drill bit advanced, that outer casing is carefully removed as the screen and well casing are installed. The screen allows groundwater to enter the well from the saturated aquifer material around the screen (see Figure 5.1). The length of the screen and the depth at which it is placed are selected to best fulfill the intended purpose of the well, given existing knowledge of the site’s hydrogeology and borehole information collected during drilling. Placement of the screen can be considered part of the three-dimensional challenge of locating the well on the surface and then placing the screen at an appropriate depth to sample the groundwater of interest. Screening There is no universal technically correct length or position in an aquifer for placing the well screen, although guidelines can be agreed to beforehand. For example, the Consent Order suggests placing a single, relatively short (5- to 10-foot) screen in zones of relatively high hydraulic conductivity to monitor so-called fast paths for lateral flow. The Environmental Protection Agency (EPA) (Aller et al., 1991) and the American Society for Testing and Materials (ASTM, 1995) recommend screened intervals of 2 to 10 feet. 1 EZ-Mud and Quik-Foam are registered trademarks of the Baroid company.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report TABLE 5.1 Drilling Methods That Are Potentially Applicable to Well Construction at LANL Method Advantages Disadvantages for Monitoring Rotary with air as the drilling fluid to bring cutting to the surface Relatively fast, moderately expensive, no added liquids or additives. Best for hard rock formations. Air injection may strip volatile organic compounds (VOCs), change redox potential, and induce biodegradation. Well development is critical. Difficulty with sloughing unconsolidated sediments. Rotary with air as the drilling fluid plus outer casing advance (lowered by its own weight or percussion-hammered). Also known as Dual Wall Reverse Circulation with air Over-reaming bit allows casing to follow bit downhole to prevent unconsolidated materials from sloughing. Casing can be removed slowly during well construction to facilitate screen and sand/ gravel pack location. With sufficient air pressure, may avoid additives. Expensive, relatively slow. Air injection may strip VOCs, change redox potential, and induce biodegradation. Well development is critical. Rotary with water as the drilling fluid to bring cuttings to the surface Fast, relatively inexpensive. Can also employ dual wall reverse circulation equipment. Water in borehole complicates identification of water-bearing layers and can change hydrologic and geochemical properties near borehole. May lose circulation in unconsolidated materials. Well development is critical. Rotary with drilling muds made from slurry of water and mud additives to bring cuttings to the surface and to keep the borehole open in unconsolidated zones Same as water rotary except with mud additives to prevent lost circulation and stabilize borehole wall. Fast and moderately expensive. Established practice for potable water production wells. Additives may be reactive with chemicals of potential concern (COPC). Requires aggressive well development to reduce mudcake on borehole walls. Typically inappropriate for monitor wells for reactive COPCs. Rotary with cold nitrogen rather than air as drilling fluid (cryogenic rotary) Cold nitrogen gas in standard air rotary process can freeze borehole wall in wet unconsolidated zone. Non-reactive nitrogen gas cannot change geochemistry. Like air rotary, gas injection at high pressure can affect local hydrologic characteristics near borehole. Tested at DOE facilities but not readily available. Likely expensive. Boring into the Earth with a hollow-stemed auger bit Fast, inexpensive, good geologic samples, no added fluids required. Limited to shallow depths, cohesive sediments. Cable-tool drilling—raising and dropping a heavy bit into the borehole, and removing cuttings with bailers Can be done without added fluids if unconsolidated materials in saturated layers do not slough into borehole. Geologic samples are relatively undisturbed. Samples can be collected ahead of the hole with conventional geotechnical samplers. Usually requires stepped-down borehole diameter as hole deepens. Slow, moderately expensive. Few vendors for environmental applications. Drilling with resonant high frequency vibration to drive drill pipe into the subsurface (sonic drilling) No drilling fluids required, can penetrate all formations at any angle, no cuttings. Provides continuous core in drill pipe. High cost, few vendors. Geologic sampling could require additional equipment. SOURCE: Committee. List based on Consent Order Section X.B. EPA (1992) acknowledges the need to design the screen length to meet the objectives of the well. As part of the Hydrogeologic Workplan, LANL contractors did geophysical testing in both open and cased hole conditions in order to determine the high-conductivity fast-pathway zones in the formations around the borehole; see Sidebar 5.1. This geophysical testing provided information to establish locations of the higher-permeability zones by characterizing the subsurface lithologic units in terms of their moisture content (including perched groundwater), capacity for flow, and stratigraphy and mineralogy. Previous problems in installing well screens at LANL have been reported to include excessively long screens, screens installed at the wrong depths to intercept contaminants, too many screens per well, and screen materials that corrode in groundwater (Gilkeson, 2006b). The use of overly long screens can cause dilution of sampled contaminants. Multiple screens, on occasion as many as nine screens in some LANL wells, can cause dilution or possibly cross-contamination of samples if there is leakage between screens. Nylander (2006) reported differing technical views on screen length throughout the period of the Hydrogeologic Workplan. Development After the screens are in place, the well is developed (ASTM, 1994, 1995). This final step of the well construction process is intended to remove drilling fluids and repair damage done to the formation adjacent to the borehole wall by the well drilling. For monitoring wells, the goal is to restore the properties of the original formation around the screened

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report FIGURE 5.1 Components of a water sampling well. This illustration shows a well with three screened intervals (near top, center, and bottom of figure) for sampling water at three different elevations. As depicted in the illustration, each screen is surrounded by permeable sand or gravel pack to allow water to enter the well. Installing multiple screens, ensuring that each is hydraulically isolated by the use of mechanical devices called “packers,” and developing multiscreen wells are difficult. SOURCE: Broxton, 2006. SIDEBAR 5.1 Geophysical Testing to Position Well Screens Downhole geophysical tools are often applied in hydrogeologic characterization programs to identify changes in lithology indicated bymineralogical, permeability, and porosity variations. The extensive suite of geophysical testing done on most R-wells included nearly continuous measurements along the length of the borehole to measure the following:a Total and effective water-filled porosity and pore size distribution, for estimation of hydraulic conductivity, Bulk density considering both water- and air-filled porosity, Bulk electrical resistivity at multiple depths, Bulk concentrations of selected mineral-forming elements, Spectral natural gamma-ray emissions, Bedding orientation and geologic texture, Acoustic compressional wave velocity, Borehole azimuth and inclination, and Borehole diameter. In addition to helping establish higher-permeability zones for the purpose of well screening, the geophysical testing provided data to correlate variations in seismic velocity versus depth in order to calibrate surface seismic surveys and to evaluate borehole conditions including borehole diameter, vertical deviation, and degree of drilling fluid invasion.    aSchlumberger (2003), which was compiled for well R-20, is an example of a typical geophysical report.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report interval, especially with respect to chemical conditions, porosity, and permeability in order that water samples taken from the well are actually representative of the native aquifer (Broxton, 2006). There is general agreement that the use of bentonite clay and organic additives has compromised the ability of at least some R-wells to yield water samples that are truly representative of the ambient, undisturbed groundwater conditions (LANL, 2005d; Ford et al., 2006; Ford and Acree, 2006; NMED, 2006). Robert Gilkeson, a registered geologist and former advisor to LANL, stated that bentonite clay and/or organic drilling additives had invaded the screened intervals in all of the LANL characterization wells (Gilkeson, 2006a,b). He illustrated a conceptual model of how these materials can set up a “reactive capture barrier” that would tend to remove contaminants from sampled groundwater; see Figure 5.2 (also see Chapter 3, Figure 3.2). LANL’s groundwater analyses typically show the presence of naturally occurring cations and anions indicating that, if it is occurring, reactive capture probably does not function as an absolute barrier. However, the degree to which contaminants might be attenuated is uncertain. LANL has acknowledged that residual drilling fluids have affected the multiscreen R-wells. In terms of providing samples representative of regional groundwater, LANL found that “single-screen wells generally provide the most defensible data” (LANL, 2005d, p. v). Because the construction of these wells was expensive, some $1 million to $2 million for each well (Broxton, 2006), LANL began work in 2006 to try to recover some of the compromised screened intervals (LANL, 2005d, 2006e,f). This rehabilitation effort is itself controversial (Gilkeson, 2006a,b; LANL, 2004d). The New Mexico Environment Department’s (NMED’s) notice of disapproval of the Well Screen Analysis Report (letter dated September 18, 2006) indicated continued disagreement on a number of important issues regarding the rehabilitation work. After this report entered review, NMED accepted LANL’s approach to identifying the impacts of drilling fluids (NMED, 2007b) via the Well Screen Analysis Report, Revision 2 (LANL, 2007c). According to LANL, a key component of the accepted methodology is the acknowledgment that a well screen at a particular location needs to provide reliable data only for potential chemicals of concern at that location. In addition, NMED responded to the Laboratory’s report on preliminary results of the pilot well rehabilitation study at three of the impacted characterization wells (LANL, 2007d) by requesting a revised well rehabilitation plan (NMED, 2007c). NMED has also requested assessments of the current groundwater monitoring network by area (e.g., TA-21, TA-54, Mortandad Canyon). According to the request, these network area assessments will evaluate the location of wells, the reliability of data from the wells, and well construction in relationship to the contaminants of concern at these areas. The area assessments will make recommendations on the specific wells to be rehabilitated or replaced. The revised well rehabilitation plan describes approaches to redeveloping wells that are determined by area assessments to be critical for monitoring. The area assessment is to be completed by December 2007, while well rehabilitation and/or replacement is expected to be completed by the end of FY09. LANL’s Plans for Well R-35 LANL will drill new monitoring wells under the Consent Order (see Table 5.2). R-35 is the first regional well being drilled during 2007. This well has the primary objective of monitoring for chromium in the upper portion of the regional aquifer, particularly relative to the PM-3 water supply well; see Color Plate 10. Plans for drilling R-35 evolved during the committee’s study period. The June 2006 workplan for drilling this well described a graded approach of using air as the drilling fluid for the first tens of feet, then water, foam, and finally muds as necessary to reach the target depth. In a March 2007 letter to NMED, LANL amended this approach and announced its intention to drill R-35 to depth using air as the only drilling fluid: The revised approach is to drill using casing-advance air-rotary with intent to maximize the potential for success of the air rotary method to accomplish the objectives. Each borehole will initially be drilled open hole with air-rotary foam-assist through the vadose zone to a depth above the regional aquifer. Casing will be set to hold back any perched water encountered and to prevent caving of the borehole wall. Casing will then be advanced while drilling the remainder of the borehole using conventional air-rotary to total depth (Mangeng and Rael, 2007). Well R-35 will actually consist of two adjacent boreholes. The shallower, R-35b, with a target depth of about 900 feet, will be screened in the most transmissive zone about 50 feet below the top of the regional aquifer (the water table). The deeper, R-35a, will be screened about 300 feet below the top of the regional aquifer. This will be in the most transmissive zone that corresponds to the upper portion of the screen in well PM-3. R-35 will thus consist of two single-screen wells. The Mangeng and Rael (2007) letter noted that the amended approach is consistent with input from the Northern New Mexico Citizens’ Advisory Board and other knowledgeable stakeholders. However, it is a significant change from LANL’s presentations to the committee, which emphasized problems with air-rotary casing advance drilling encountered with the equipment and procedures used during the Hydrogeologic Workplan.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report FIGURE 5.2 Reactive contaminant capture barrier. Geologist Robert Gilkeson described concepts of how drilling fluids could form a zone that removes contaminants from sampled groundwater. This would invalidate affected well screens as sampling points. SOURCE: Adapted from Gilkeson, 2006a.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report TABLE 5.2 Current LANL Estimate of Numbers of Monitoring Wells to Be Drilled   Groundwater Sampled Location Alluvial Intermediate Regional Los Alamos/Pueblo Canyons Watersheda 1 2 3a Mortandad Canyon Watersheda       Water Canyon/Cañon de Valle Watersheda 2 3 2a Pajarito Canyon Watershedb 11 1 2 Sandia Canyon Watersheda   2 2a MDA G, L, Ha     4-6a MDA A, B, T, U, V (TA-21)a     4a MDA Ca     1a MDA AB     1 Totalsa 14 8 19-21 aNMED approval of area-specific monitoring network assessments letter will finalize the number of wells required (NMED, 2007a). bPer NMED approved Investigation Workplan for Pajarito Canyon • Water Canyon/Cañon de Valle assessment submitted to NMED 4/30/07 • Mortandad Canyon and Area C assessment due to NMED 6/28/07 • TA-54 assessment due to NMED 7/31/07 • Sandia Canyon assessment due to NMED 9/14/07 • TA-21 and Los Alamos/Pueblo Canyon assessment due to NMED 12/30/07 SOURCE: Adapted by LANL from NMED, 2007a. Mat Johansen, National Nuclear Security Administration liaison to the committee, informed the committee that the key to the expected success of using air-rotary with casing advance as needed for R-35 was agreement with NMED on the target zones for the two wells (Johansen, 2007). According to Johansen, with the target zones identified, the objectives of the drilling are much more focused than for the wells drilled from 1998 through 2004 under the Hydrogeologic Workplan. Those wells included objectives such as detailed geologic and hydrologic characterization of the approximately 800 to 1000 feet of vadose zone, and characterization at greater depths within the regional aquifer. Those general characterization objectives influenced the choice of drilling approaches used in past wells. Johansen noted that most of the characterization data needed to plan R-35 were available from three nearby R-wells that were drilled under the workplan. Committee Observations on Well Construction LANL’s well construction practices (drilling, screening, development) changed significantly during the Hydrogeologic Workplan to meet changing objectives and constraints (time, money). Plans for constructing new wells continued to change during the committee’s study. Changes will continue to be driven by technology, project objectives, and constraints. For example, the plans being made for R-35 seem appropriate given its objectives, but the objectives are narrow and the hydrogeological environment the well will penetrate has already been characterized by previous drilling. Future drilling under the Consent Order may encounter challenges similar to those of the Hydrogeologic Workplan. Table 5.1 provides a description of standard drilling techniques, along with their probable advantages and disadvantages for application at LANL. It is unlikely that any single one of these techniques will satisfy all of the site’s future needs for characterization, monitoring, and eventually remediation. Recognizing that decisions made over the course of the Hydrogeologic Workplan cannot be changed, it is important to incorporate lessons learned into future drilling. In this context, the committee made general observations that may be useful to LANL in constructing new wells during the remainder of its groundwater protection program. Drilling Test holes are often used in water well drilling programs to help identify the most productive zones and locations in heterogeneous aquifers prior to drilling and construction of the intended well. When drilled primarily for geologic information through collection of cuttings and occasional core samples, test holes can be relatively inexpensive and fast. Additives can be used to expedite the drilling because the hole will not be used for quantitative water or soil analyses. In complex conditions such as the LANL subsurface, test holes can allow identification of multiple water-bearing zones and application of downhole geophysical tools. The information from test holes can then be used to plan the drilling procedures and develop construction specifications for the desired monitoring or production well(s). Considering the very high cost of constructing wells to meet multiple objectives under the Hydrogeologic Workplan and the clear need to have characterization information available before installing a monitoring well, it would appear that drilling one or more

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report simple test holes near a planned monitoring well location could help ensure successful installation of the well. For monitoring wells, given the uncertainties about effects of drilling muds and additives and the importance of minimizing alterations in the groundwater environment around screened intervals, the portion of the borehole to be sampled should be drilled to the extent possible with air or water as the circulating fluid. Advancing an outer casing to keep the borehole open can reduce or prevent the need for more complex drilling fluids. Mud or other additives are a last resort, but it may not be possible to completely avoid them, for example to keep boreholes from collapsing during drilling and well construction. The Consent Order allows mud rotary drilling while providing cautions about changes in the near-borehole environment that can be caused by bentonite and ionic or organic polymer fluids. In addition, the Consent Order recognizes that a polyacrylamide mud, such as EZ-Mud®, can be used appropriately if it is followed with a dispersant, such as BARAFOS®, to facilitate the breakdown and removal of the polymer. If the appropriate dispersant is applied, there should be reasonable success in recovering the dispersed and degraded EZ-Mud®. There are other options for drilling fluids. Xanthan gum, also used in enhanced oil recovery, is far less anionic that EZ-Mud® and should offer fewer sorption sites. Starch is an option also. Combinations of bentonite and organic polymer to form a “low fluid loss” mud that reduces the amount of drilling fluid that is pushed into the formation offer another approach. Most of these options are not new (Nylander, 2006), but there is no evidence that their potential to alter the geochemical environment around LANL well screens has been evaluated. Screening and Purging the Screens In some instances, multiple screens in one borehole are desirable for measuring vertical gradients in pressure (“head”) and groundwater composition. However, EPA (1991) and field experiences indicate that multiple screens in deep wells are prone to problems. LANL’s experiences during the Hydrogeologic Workplan indicated that construction of multiscreen wells is difficult and problematic. Disadvantages of multiple screens for well construction at LANL usually outweigh their possible advantages. Hydraulic separation of multiple screens is difficult under the simplest geologic conditions. Multiple screens, such as used in most of the compromised wells at LANL, are hard to develop individually, requiring “packers” to isolate each screen from its neighbors; see Figure 5.1. The relatively thin saturated zones contacted by each screen may not sustain great enough pressure changes (induced by pumping or “surging”) to move water in and out of the screened areas to clear out the drilling fluids. The only way to completely avoid the possibility of cross-contamination between zones is to use single-screened monitoring wells. If sampling pumps are installed in each screen, the combination of materials used in the casing, screen, pump, and discharge piping must be selected to prevent galvanic corrosion, which can result in spurious detections of metal corrosion products. Construction requires careful selection of casing and screen materials to have required strength for deep holes. Material failures have occurred at LANL, e.g., at R-25 (Nylander, 2006). Generally screens are placed in the most permeable zone of the aquifer they are intended to sample. Geophysical logs, even as complete a suite as those used by LANL, infer permeability, but they do not of themselves measure it. The practice of inferring permeability from geophysical measurements is, nevertheless, widespread and accepted. Absent a nearby test hole, taking a side wall core during drilling of the monitoring well could be a partial solution. This core could also be used to evaluate the correspondence between geophysical measurements and hydrologic properties. Borehole flow meters to sense flow directions and velocities within the saturated zone offer another possibility. This type of data can be useful to establish flow directions that are affected by local heterogeneity or anisotropy, and may not be discernable by inferred flow lines from head contour maps. Given that drilling and well construction inevitably causes disturbance of the subsurface formation, industry experience is that typically the native geochemical and hydrological conditions tend to re-establish as groundwater flows around and through the well screen. To help ensure this re-equilibration, application of proper purging techniques in both well development and groundwater sampling is necessary for collection of representative groundwater samples, especially in the regional aquifer. The most trustworthy sampling technique includes purging three or more well volumes from the monitoring well before sample collection (ASTM, 1992). While this method requires containment and potential treatment of much more water that the minimum-purge techniques, it better ensures that samples from the developed wells represent the conditions in the nearby aquifer. Purging is much easier to control and complete with single-screened monitoring wells, as noted earlier. The uncertainty in determining the elevation of the more permeable zones is part of the larger issue of sampling along the contaminant pathways that were described in Chapter 4. The screen is intended to sample a particular pathway, which requires having a good estimate of that pathway in three dimensions. If the pathway is different than that presumed, a migrating contaminant would be missed. The issue here is one of robustness of the sampling and monitoring plan since knowledge of the pathways is always uncertain. Concluding Comments on Well Construction The changes and evolution of LANL’s drilling program are in keeping with the development of any major scientific undertaking; indeed such evolution is essential. One cannot

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report know all the answers at the outset and learns as the program progresses. However in following the drilling program, the committee concluded that the program has evolved more from an operational approach—try and see what works—rather than from using careful analysis of past results to inform future planning. LANL scientists expressed concerns with drilling muds early in the Hydrogeologic Workplan, but their concerns were essentially laid aside when initial efforts to use air-rotary drilling failed to meet programmatic requirements, and the use of bentonite mud and additives was deemed the only way to proceed. Should air-rotary prove unsatisfactory for R-35 or any future well, the committee is concerned that LANL could not present a scientific rationale for switching to another drilling fluid or additive. Without a scientific basis to underpin such a change of plans, the concerns and issues raised with the existing R-wells could be repeated. SAMPLING PROTOCOL As noted at the beginning of this chapter, the committee answered the question: “Is the laboratory following established scientific practices in assessing the quality of its groundwater monitoring data?” with a qualified yes. The committee found that LANL has in place the proper data quality procedures to generate sound data from groundwater monitoring—with the caveat that water samples are indeed representative of the actual groundwater. However, it is not clear how such procedures are actually carried through in LANL’s use and reporting of sampling data and its uncertainties, as will be discussed in this section. In reviewing LANL’s data quality program, the committee used the following working definitions: Quality: The totality of features and characteristics of a product or service that bear on its ability to meet the stated or implied needs and expectations of the user. Quality assurance (QA): An integrated system of management activities involving planning, implementation, assessment, reporting, and quality improvement to ensure that a process, item, or service is of the type and quality needed and expected by the customer. Quality control (QC): The overall system of technical activities that measure the attributes and performance of a process, item, or service against defined standards to verify that they meet stated requirements established by the customer; operational techniques that are used to fulfill requirements for quality. Quality Assurance Project Plan (QAPP): A formal document describing in comprehensive detail the necessary QA, QC, and other technical activities that must be implemented to ensure that the results of the work performed will satisfy the stated performance criteria. As defined for Superfund in the Code of Federal Regulations (40 CFR 300.430), the QAPP describes policy, organization, and functional activities, along with the data quality objectives and measures necessary to achieve adequate data for use in selecting the appropriate remedy. The QAPP is a plan that provides a process for obtaining data of sufficient quality and quantity to satisfy data needs. Table 5.3 lists documents reviewed by the committee to better understand LANL’s sampling and analytical methods, data review and compilation, data documentation, and record keeping. Section 10 of LANL’s QAPP requires independent assessment of how all data are generated, reviewed, statistically compiled, and made public with specific focus on and how specific QA/QC procedures are used. Committee members compared data from analyses of groundwater samples posted on LANL’s Water Quality Data- TABLE 5.3 Quality Assurance Documents Reviewed Subject Area Plans Reviewed   Quality Assurance and Quality Control (QA/QC) procedures Quality Management Plan for Los Alamos National Laboratory Risk Reduction and Environmental Stewardship-Remediation Services Project (RRES-QMP, R3); ER2004-012; April 15, 2004 Quality Assurance Project Plan (QAPP) for the Groundwater and Persistent Surface Water Monitoring Project (ENV-WQH-QAPP-GWSW, R0), Controlled Document signed May 8, 2006 Specific sampling and analytical procedures 2006 Integrated Groundwater Monitoring Plan for Los Alamos National Laboratory (LANL, 2006a) Interim Measures Work Plan for Chromium Contamination in Groundwater (LANL, 2006d) Sampling and analytical procedures, along with data review and statistical compilation approaches LANL Groundwater Background Investigation Report, Rev. 1 (LANL, 2006b)  

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report base (WQDB) website2 to these data quality procedures. The WQDB is a public website, which provides real-time access to the results of chemical analyses of LANL’s groundwater samples. Compilations of these data support LANL’s annual Environmental Surveillance Reports and other compliance and decision-making documents. The website notes that its data are in various stages of review and are flagged to give an indication of their current status. As one example of the results of the committee’s comparison, it is unclear how QA/QC procedures were used in the sample analyses and what the specific criteria for acceptance or rejection of analytical results were. Chromium was reported in well R-32 at concentrations between 0.5 and 3 µg/L, but these are at or below the Method Detection Limits (MDLs) cited: <0.503 to <7.4 µg/L. In other cases, sampling results fall within the cited MDL-Practical Quantitation Level (PQL) range, yet they are not identified as J-values, as described in Sidebar 5.2. The committee encountered instances of inconsistency in data reporting. Table C-4 (Appendix C) in the Integrated Groundwater Monitoring Plan (LANL, 2006a) gives the MDL for total chromium as 1 µg/L and the PQL as 5 µg/L. The indicates a more precise knowledge of the MDL than the range of <0.503 to <7.4 µg/L reported on the WQDB. While the Integrated Plan reports both total chromium (Cr) and hexavalent chromium (Cr6+), it gives the analytical method only for total Cr. One does not know the analytical method used for Cr6+ or the MDL and PQL values for the method. Explaining how data are obtained is as important as reporting the data themselves. In addition, LANL reports MDL and PQL values that are not appropriately rounded, and thus give an impression of accuracy and precision that do not truly exist. For example, the MDL for Cr of 0.503 µg/L on the WQDB should be rounded to 0.5 µg/L. In the Integrated Plan (Table 4.2-4a) the background chromium concentration in regional groundwater reported as 4.083 µg/L should be rounded to 4.0 or 4.1 µg/L. While the above discussion assumes that representative groundwater samples are collected for subsequent analysis, it is essential to remember that there is debate regarding this assumption, especially related to multi-screen wells. Thus, as part of a sound QAPP, results from these suspect wells should be flagged as such. A good deal of misinformation can result if publicly available databases or compilations of LANL monitoring data do not identify the soundness of all data reported according to the data quality objectives that are clearly spelled out in the QAPP. DATA QUALITY FOR REMEDIATION DECISIONS The committee was asked, “Are the data (including qualifiers that describe data precision, accuracy, detection limits, and other items that aid correct interpretation and SIDEBAR 5.2 Limits of Contaminant Detection and Quantitation To be able to clearly differentiate waters impacted by LANL site activities from non-impacted waters (i.e., background), as well as to determine when an impacted water exceeds a regulatory guideline and/or standard and may require active remediation, it must be established that such determinations are based on statistically sound analytical data. In this regard, the Method Detection Limit (MDL) and the Practical Quantitation Level (PQL) are the two measures of analytical capability used for this purpose. The MDL is a measure of method sensitivity. It is defined in 40 CFR Part 136 Appendix B, pp. 554-555, as “the minimum concentration of a substance that can be reported with 99% confidence that the analyte concentration is greater than zero.” MDLs can be operator, method, laboratory, and matrix specific. Due to normal day-to-day and run-to-run analytical variability, MDLs may not be reproducible within a laboratory or between laboratories. The regulatory significance of the MDL is that EPA uses the MDL to determine when a contaminant is deemed to be detected and it can be used to calculate a PQL for that contaminant. In the preamble to a November 13, 1985 rulemaking (50 FR 46906), the PQL was defined as “the lowest concentration of an analyte that can be reliably measured within specified limits of precision and accuracy during routine laboratory operating conditions.” The EPA has used the PQL to estimate or evaluate the minimum concentration at which most laboratories can be expected to reliably measure a specific chemical contaminant during day-to-day analyses of drinking water samples. A PQL is determined either through the use of inter-laboratory study data or, in the absence of sufficient information, through the use of a multiplier of 5 to 10 times the MDL. In practical terms, ASTM (ASTM Standard D 596-01, Standard Guide forReporting Results of Analysis of Water) defines the MDL as the concentration below which a chemical cannot be said to be present with any confidence. Furthermore, an analytical result between the MDL and PQL implies that the respective chemical is present but cannot be quantified. Concentrations of chemicals below an MDL are generally identified as “<#” or “#U” values with the # being the chemical-specific MDL. A chemical concentration between the MDL and PQL is estimated with the indicator “J” and is referred to as a “J-value.” use of the data) being used appropriately in the laboratory’s remediation decision making?” The committee’s short answer is no, for several reasons. Formally, LANL had not begun remediation activities during the committee’s study period (Dewart, 2006) and the committee heard no 2 See http://wqdbworld.lanl.gov.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report presentations about this aspect of its remediation decision making. More to the point, however, the committee became concerned about LANL’s use of results from measurements near or below the limits of practical quantitation and detection, near the natural background, or both, in some of its key documents; see Sidebar 5.3. In terms of supporting future remediation decision making, data in LANL’s Groundwater Background Investigation Report (LANL, 2006b) appear to be derived from sound sampling and analysis. The report clearly lays out how data were collected and also pays adaquate attention to QA/QC procedures as well as how MDL concentration levels were handled. By setting up this background information for all three groundwater regimes (i.e., alluvial groundwater, intermediate-perched groundwater, and the regional aquifer groundwater), LANL is in a good position to statistically determine any future increases above background concentrations. While the Background Investigation Report shows good statistical data compilation focused on well-documented QA/QC approaches, gaps remain. The report is not clear on how the QAPP procedures were actually followed and implemented, and in fact it does not reference the QAPP. The report also contains discrepancies in terms of documenting the actual analytical methods used and the respective MDL and PQL for the analyses. One example is for Cs-137. The background investigation report (Table 4.2-4a) gives a Cs-137 concentration of 1.1 pCi/L without specifying the MDL or PQL. Notably, 1.1 pCi/L is below the PQL for Cs-137 that LANL cites elsewhere—8 pCi/L in the Integrated Groundwater Monitoring Plan (Table C-4). In another important example, the mean Cr concentration in a filtered sample representative of the background in the regional aquifer is given as 4.083 µg/L with a standard deviation of 5.948 µg/L (Table 4.2-4a). The same report (Table 4.1-2) cites the MDL as being either 2 or 10 µg/L depending on the particular analytical method used. Thus the actual mean Cr background concentration is not established. All that can be inferred is that the true background level is somewhere in the 1-10 µg/L range. On this basis, it appears that the majority of the Cr concentrations cited in Figure 3-3 of the Interim Measures Work Plan for Chromium Contamination in Groundwater (LANL, 2006d) are background levels and that only the Cr concentrations cited for wells R-28 and R-11 can be attributed to LANL operations. Yet without this clarification, one can infer that all the levels cited in that figure are significant (i.e., greater than background). The Consent Order specifies that remediation meet State of New Mexico water quality standards as well as any other applicable regulations (Table B.2 of the Integrated Groundwater Monitoring Plan). For some contaminants, however, current analytical methods appear to be inadequate to ensure compliance with these requirements. That is, some MDL and PQL concentrations cited in Table C-4 of the Integrated Monitoring Plan are above the regulatory limits cited in Table B.2. For example, the cleanup requirement (Table B.2) for arsenic is 0.45 µg/L, but the analytical MDL is 6 µg/L and the PQL is 15 µg/L (Table C-4). Likewise, for different Aroclors the cleanup criterion is 0.00064 µg/L while the MDL range is 0.0875-0.4165 µg/L and the PQL is 0.5 µg/L. SIDEBAR 5.3 Citizens’ Concern for Radionuclides Reported in Drinking Water Near the end of this study, the non-governmental organization Concerned Citizens for Nuclear Safety (CCNS) and Robert H. Gilkeson, a registered geologist, brought to the committee’s attention data in LANL’s Draft Site-Wide Environmental Impact Statement (SWEIS; DOE, 2006) that indicated contamination of drinking water supply wells by neptunium and other radionuclides, including plutonium, americium, strontium, and cesium. CCNS and Gilkeson pointed out that data tables in the draft SWEIS showed, for example, that neptunium (Np-237) was detected in 4 of 13 samples from Los Alamos County supply wells and in 2 of 3 samples from the Buckman well field that supplies over 40 percent of the drinking water for residents of the city of Santa Fe. Mean concentrations of Np-237 were 10.6 and 10.3 pCi/L, respectively. These reported concentrations approach the EPA limit of 15 pCi/L for alpha-particle emitting nuclides in drinking water. In its memorandum to the committee, CCNS and Gilkeson stated: “We are surprised at the high levels of neptunium. This contamination may be because of the poor precision of the gamma spectroscopy analytical method. The LANL scientists claim the neptunium contamination doesn’t exist and the detects are ‘false positives.’ Nevertheless, the contamination is presented as valid detections in the data tables in the draft LANL SWEIS” (Gilkeson and Arends, 2007, p. 5). In responding to CCNS, LANL did in in fact attribute the reported data to “false positives,” stating: “Detections of LANL-derived contaminants, such as plutonium, americium, and strontium, have occurred sporadically in water supply wells…. Because the overall frequency of detection is low, we believe that these sporadic detections are false positives or caused by problems at the analytical laboratory. This conclusion is supported by numerous reanalyses of these samples and by lack of consistent detections in paired samples” (Phelps, 2007, p. 2). This exchange between CCNS and LANL is a good example of why the committee is concerned about LANL’s representations of groundwater sampling data. Whether or not the data were statistically significant, and the committee takes no position on this, the data were reported by LANL in its draft SWEIS and, reasonably, taken as real concerns by public stakeholders.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report FINDINGS AND RECOMMENDATIONS ON MONITORING AND DATA QUALITY General Findings Any monitoring activity faces a conundrum: If little or no contamination is found, does it mean that there is in fact little or no contamination, or that the monitoring itself is flawed? During this study the committee was presented a good deal of information suggesting that most or all wells into the regional aquifer at LANL (R-wells) are flawed for the purpose of monitoring. The committee did not disagree, but rather found a lack of basic scientific knowledge that could help ensure future success. Evidence about the conditions prevalent around the screens in the compromised wells is indirect—relying on plausible but unproven3 chemical interactions, general literature data, analyses of surrogates, and apparent trends in sampling data that may not be statistically valid. LANL is using good practices in terms of having the proper QA/QC plans and documentation in place, but falls short of consistently carrying out all the procedures cited in the plans. Although LANL appears to be generating sound analytical data, the results reported in databases and LANL reports often do not carry the proper qualifiers according to good QA/QC practices. This especially applies to analytical results near or below the limits of practical quantitation and detection, near the natural background, or both. Detailed Findings and Recommendations Data from scientifically vetted (peer-reviewed) studies are necessary to authoritatively address concerns and uncertainties about how drilling and well completion processes might alter the native conditions around well screens and to ensure reliable monitoring activities in the future. The committee received little scientific information—for example, on a par with LANL’s publications about vadose zone pathways (VZJ, 2005)—regarding the geochemical behavior of contaminants in the subsurface or effects of non-native materials (drilling fluids, additives, construction materials) on the geologic media to be sampled. Recommendation: LANL should plan and carry out geochemical research to ascertain the interactive behavior of contaminants, materials introduced in drilling and well completion, and the geologic media. As a part of LANL’s future plans for sitewide monitoring, this work would include: Determining the nature of interactions among materials proposed for use in constructing monitoring wells and the types of geological media that LANL intends to monitor; Quantitative measurement of sorption of contaminants onto the natural, added, and possibly altered constituents that constitute the sampling environment of a monitoring well; and Publication of results in peer-reviewed literature. The committee is not recommending open-ended research. Rather, targeted investigations would underpin plans for future monitoring of specific areas of the site: contaminants of greatest concern in the area, geologic media expected to be sampled (known from previous site characterization), and drilling fluids, additives, and other materials intended to be used in constructing the monitoring well(s). Screening tests envisioned by the committee would include simple batch equilibrium tests to measure solubilities and sorption coefficients (Kd) and to determine what, if any, interactions actually occur among drilling materials and the geologic media—and whether alterations are permanent or temporary. More detailed column tests can simulate and measure effects of flow rate and surface area (mass transfer) around the well screens. Planning, conducting, and interpreting the results will require the high quality of science one would expect of a national laboratory. LANL’s work under the Hydrogeologic Workplan significantly enhanced understanding of the hydrological characteristics of the site, and lessons learned during the program can improve future drilling efforts. Wells constructed under the Hydrogeologic Workplan were intended for characterization. LANL later attempted to use the characterization wells that reached the regional aquifer for monitoring. As noted earlier, their use for monitoring was evidently compromised by drilling and well development procedures. Recommendation: LANL should plan and conduct future characterization drilling and monitoring well drilling as separate tasks. For monitoring locations where characterization data are unavailable, LANL should consider drilling simple test holes to obtain these data before attempting to install the monitoring well(s). With the more complete hydrogeologic characterization that is now available (see Chapter 4), LANL can design and construct future monitoring wells more confidently. LANL’s plans to obtain geologic and geophysical logs during drilling further increase confidence that well screens can be installed to intercept a contaminant pathway. Recommendation: LANL should design and install new monitoring wells with the following attributes: A borehole drilled through the monitoring zone without the introduction of drilling muds or additives (i.e., use air or water), 3 Not directly observed and measured under LANL site conditions.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report One screened interval that targets a single saturated zone, and A carefully planned design (length and depth) of the well screen, which is confirmed with information collected in the drilling process. Drilling under specific conditions and sampling requirements can lead to exceptions to the above, and adapting to circumstances will be necessary. With regard to LANL’s practices in assessing the quality of its groundwater sampling data, the committee found that good data quality procedures are in place, but there is a lack of follow-through in how the data are reported. Recommendation: LANL should ensure that there is consistency and clarity of all related sampling and analytical procedures with documented follow-through and appropriate action. This especially relates to: Having clear data quality objectives; Documenting how samples are to be collected; Documenting how data are handled, statistically compiled, and reported; Clear documentation of the quality of the data; and Identification of all suspect data. Interpreting data at or near analytical detection limits is an area of growing scientific interest. LANL can benefit from scientific exchanges with other groups and organizations that are actively working in this area (e.g., EPA, American Society for Testing and Materials). Lack of agreement between LANL, regulators, and concerned citizens as to what constitutes the appropriate representation of groundwater contamination data is a source of confusion and distrust. Recommendation: LANL should ensure that measurements at or near background levels or near analytical detection limits (i.e., MDLs and PQLs) are scientifically and statistically sound and are reported appropriately. The LANL site office of DOE should take steps to ensure that LANL and site regulators agree on how all such data are to be handled, compiled, and reported. LANL should make more effort to ensure that data uncertainties are made clear to public stakeholders. LANL’s Groundwater Background Investigation Report (LANL, 2006b) is an important step in establishing levels of naturally occurring contamination in the regional aquifer, although some gaps were identified by the committee. The Integrated Groundwater Monitoring Plan (LANL, 2006c) lists non-LANL sources of groundwater contamination. Such data are important to support future remediation decision making. Recommendation: LANL should continue to track regional groundwater monitoring wells and water supply wells routinely to improve the statistical basis for reporting any increases above background. LANL’s Quality Assurance Project Plan should enforce the documentation of any and all instances where it is believed that chemicals or radionuclides detected in groundwater are not the result of LANL operations, for example, naturally occurring or anthropogenic contaminants or the result of sampling artifacts.

OCR for page 49
Plans and Practices for Groundwater Protection at the Los Alamos National Laboratory: Final Report This page intentionally left blank.