Date: Fri, 26 Feb 1999 14:56:57 -0700 Subject: Trace Element News v.3 From: Dennis Helsel , "Robert J Gilliom, Supv. Hydrologist, Sacramento, CA" , "Jeffrey D Stoner, Supervisory Hydrologist, Denver, CO" , "Carol A Couch, Hydrologist, Norcross, GA" , "Alan H Welch, Hydrologist, Carson City, NV" , "Peter C VanMetre, Project Chief, Hydrol(RGEG), Austin, TX" , "Lawrence R Deweese, Fish and Wildlife Biologist, Denver, CO" , "David P Krabbenhoft, Research Hydrologist (GeoChem), Middleton, WI" , "Edward Callender, Hydrologist, Reston, VA" , "Terry M Short, Physical Scientist, Menlo Park, CA" , "Michael T Koterba, Hydrologist, Baltimore, MD" , "Dennis A Wentz, Willamette NAWQA Chief, Portland, OR" , "Timothy L Miller, Chief, NAWQA, Reston, VA" , "William G Wilber, National Synthesis Coordinator, Reston, VA" , "Jeffrey T Armbruster, Asst. Reg. Hydr.-NAWQA, Norcross, GA" , "Michael C Yurewicz, Asst. Reg. Hydrol.-NAWQA, Reston, VA" , "Ivan C James II, Asst. Reg. Hydrologist-NAWQA, Denver, CO" , "Marc A Sylvester, Asst. Reg. Hydrologist-NAWQA, Menlo Park, CA" , reg.wqspecs@srv1rvares.er.usgs.gov, "Wayne W Lapham, Hydrol(Geol) GW Spec, Reston, VA" , "Larry R Shelton, Hydro Tech, Sacramento, CA" Well Selection for Subunit Surveys Trace Element Synthesis News v.3 This is the third in a series of info emails from NAWQA Trace Element Synthesis. By request, more people have received this than the previous two. If you would like to be removed from the list, just reply and let me know. If you would like a copy of one or both of the previous two, also let me know. v.1 Trace Element Synthesis Staffing and plans 2/10/99 v.2 Changes to TE lab methods for bed sediments 2/19/99 This email is long enough that I've added an "executive summary", which should be enough for many of you. Cheers. --------------------------------------------------------- To: 97 Study Unit Chiefs From: Chief, Trace Element Synthesis Subject: Well Selection for Subunit Surveys Several 97 study units have asked about whether there are additional well selection criteria for SUS wells, since schedule 2703 for trace elements is now part of the standard NAWQA protocol. Guidance for well selection with regards to trace element contamination is given below. For more information than the discussion below, contact Michael Koterba [mkoterba, MD District, (410) 238-4240] or Alan Welch [ahwelch, NV District, (775) 887-7609]. SUMMARY * Suitability of casing and screen materials: PTFE, PVC stainless galvanized (iron is not acceptable) The degree to which these materials contribute iron or other trace elements, or adsorb trace elements, generally increases with (1) the age of the material (since installation, in years) (2) the acidity of the environment, and (3) with the oxidation potential of the environment (the more anaerobic or anoxic the environment) * Pumps: wells with water-lubricated submersible pumps are preferred over those with suction-lift or jet pumps. The latter could result in loss of radon and CFCs (age-dating). * If most or all wells are of doubtful quality, wells with higher yields may reduce contamination from bad well construction. A few study units have installed wells, with installation costs paid by others who wanted to know the quality of water in the aquifer, not just the quality of what was being pumped. * All wells sampled for trace elements must include measurement of turbidity in the field. Turbidity data are invaluable for evaluating sources of high concentrations of iron and other metals. --------------------------------------------------------- THE DETAILS Guidance on Well Selection for Study(Sub)-Unit Surveys: Well Characteristics and Materials When Sampling for Trace Elements Guidance on the selection of wells for Study(Sub)-Unit Surveys (SUS) was previously provided to the Study Units. This guidance relates to the collection of data for a variety of inorganic and organic constituents, and should be followed for the selection of wells for trace element sampling, beginning in FY 99. The purpose of this discussion is to better enable Study-Unit teams to select appropriate wells for all types of samples required, including trace elements. Well Casing and Screen Material Lapham and others (USGS WRIR 96-4233) provide excellent discussions of the relative suitability of different casing and screen materials (Table 6, pp. 25-26), and their relative leaching or sorption and desorption properties (Table 7, p. 27) with respect to organic and inorganic constituents. This information leads to the following overall hierarchy of most suitable to least suitable well casing and screen materials for SUS: Relative ranking (Most to least preferred) PTFE PVC Stainless steel Galvanized or carbon Steel PTFE: polytetrafluoroethylene, rigid. PVC: polyvinylchloride, rigid (Schedule 40 or Schedule 80, with no cutting oil or cleaning solvent residues). Stainless steel: (316 better than 304 in relation to corrosion) In particular, PVC contains tin or antimony used in its manufacture. The least corrosive of the steels, SS316, contains iron alloyed with chromium (16-18%), nickel (11-14%), manganese (2%), molybdenum (2-3%), and small amounts (less than 1%) of silicon, carbon, phosphorus, and sulfur. Formula SS304 contains iron alloyed with chromium (18-20%), nickel (8-11%), and similar amounts of manganese, silicon, carbon, phosphorous, and sulfur. "Galvanized" steel, because of its protective coating, can be a source of zinc and cadmium in addition to iron and manganese. Carbon steel comes in a number of grades that in addition to iron and manganese can contain different metal sulfides, and dis- solved metal species. Galvanized and carbon steel can also potentially provide active adsorption sites. The degree to which these materials contribute iron or their trace elements, or adsorb trace elements generally increases with (1) the age of the material (since installation, in years) (2) the acidity of the environment, and (3) with the oxidation potential of the environment (the more anaerobic or anoxic the environment) Thus, if the environment is highly acidic and anaerobic, and one selects wells with carbon casings that are a decade or two in age, it is likely that the water in that well, and in its immediate vicinity, is not going to reflect the trace-element chemistry of the aquifer. Seals and Pumps Wells with good annular and surface seals (bentonite and cement, respec- tively) are preferred. If these seals are in close proximity (within a meter or two) to the open or screened interval of the well, however, they could affect the trace-element chemistry of the sample (WRIR 96- 4233, pp. 67-70). For most wells selected for a SUS, the open (screened) interval generally will lie well below the annular and surface seals. Therefore, the integrity of these seals in the prevention of direct surface-to-groundwater flow is the main concern. Overall, wells with water-lubricated submersible pumps are preferred over those with suction-lift or jet pumps. In addition to VOCs, these wells also will be sampled for the volatile trace element, radon. Regardless of pump type, most selected wells will have discharge rates that generally are less than ideal for trace-element sample collection. Study-Unit teams should attempt to select wells that will allow them to use a portable submersible pump. This allows low flow rates (half-a- liter or less per minute) to be used (which create less turbulence) to obtain samples. This low flow rate should be set during the last half-hour of purging, to allow at least three to five sets of field- parameter measurements to be taken and coincide with the removal of the last of the third casing volume of the required purge. Installation Methods Ideally, the best drilling methods for wells used for water-quality sampling--trace element, or otherwise--are those that introduce no foreign substance (air, water or mud) into the borehole (particularly below the water level), nor allow water to move from one part of the borehole to another (because of advanced casing). The same can be said for development methods. To the extent that this information can be found, and wells can even be drilled or developed in the manner above in the SUS for which wells are being selected this should guide well selection. When the Perfect Wells Cannot be Found One might think that purging can be used to alleviate problems associated with wells that are relatively old, consist of steel casing, or lie in acidic and/or anoxic environments. But cor- roded materials could dislodge at any time, and become entrained in pump discharge, which for most SUS wells generally cannot be controlled (they're often supply wells with fixed rate submerged pumps). Don't overpurge wells! Wells pumped for too long create a drawdown that changes the typical flow pattern of water to the well, and isn't sampling the targeted water. There remains some controversy about overpurging for VOC sampling. So stay with the recommended purging protocol for NAWQA: three times the casing volume when water found standing in well, with stable field parameters; but no more than four times the casing volume. One might also think that artifacts from sampling can be removed by filtering, but that is not necessarily the case. Dislodged particles probably are colloidal, with a significant fraction smaller than 0.45 microns. Eventually, methods such as tangential-flow filtration (say at 0.005 microns) could be used to remove these, but this is not our current practice. So, it is conceivable that you could end up in the situation where the only wells you can select are ones likely to compromise the data collected. If that is the case, the only real option is to have a well installed. In some cases, this is exactly what the Study-Unit team has done. For example, if the aquifer being sampled is an extremely important one, such as a sole-source aquifer, it has been possible for some Study-Unit Teams to convince water regulators and purveyors of the need to install a well or in some cases even an entire SUS network. Perhaps the best example of the latter is the Edwards Aquifer SUS of the SCTX. The SCTX team were able to convince the City of San Antonio to pay for most installation costs. Contact Lynne Fahlquist (lfahlqst) for their negotiating strategy. When the wells selected could compromise TE data, how can they be sampled to reduce, or at least to assess, possible TE contamination? Wells selected for the SUS could fall into the questionable category because of their age, casing material, environmental settings, or limitations of sampling (using pump in well). One possible help with these is to select wells with high yields, hoping that dilution is the solution to contamination. A likely better solution for the future is to use tangential filtration. The TE Synthesis is examining this with several SUs this year. Since few study units are using TF, and the data are therefore not comparable to past and most existing SUS data, we don't recommend using it in place of 0.45 filters at this time. You could certainly do both, at twice the cost, but that's impractical for most of us. Contact Michael Koterba or Alan Welch with any questions on this. To better assess contamination, all SUS wells sampled for trace elements must include field turbidity (in NTU's only -- see Koterba and others, OFR 95- 399, p. 57, Table 19, Item 3). If NTU values do not fall below 1 or 2 NTU, this signifies the potential inclusion of colloids in the sample. While this doesn't prove TE data are compromised, it is a strong potential indicator. Contact Michael Koterba for questions on suitable (NTU) turbidimeters. --------------------------------------------------------------- Dennis R. Helsel Ph: (303) 236-2101 x227 Trace Element National Synthesis Fax: (303) 236-4912 U.S. Geological Survey PO Box 25046, MS-415 dhelsel@usgs.gov Denver, CO 80225