Www.epa.Gov Superfund Remedytech Tsp Download Gw Sampling Guide

Douglas Yeskis* and Bernard Zavala**

GROUND WATER FORUM ISSUE PAPER

BACKGROUNDThe Ground Water, Federal Facilities and EngineeringForums were established by professionals from theUnited States Environmental Protection Agency(USEPA) in the ten Regional Offices. The Forums arecommitted to the identification and resolution ofscientific, technical, and engineering issues impactingthe remediation of Superfund and RCRA sites. TheForums are supported by and advise OSWER’sTechnical Support Project, which has establishedTechnical Support Centers in laboratories operated bythe Office of Research and Development (ORD),Office of Radiation Programs, and the EnvironmentalResponse Team. The Centers work closely with theForums providing state-of-the-science technicalassistance to USEPA project managers.

This document provides sampling guidelines primarilyfor ground-water monitoring wells that have a screenor open interval with a length of ten feet or less andwhich can accept a sampling device. Procedures thatminimize disturbance to the aquifer will yield the mostrepresentative ground-water samples. This documentprovides a summary of current and/or recommendedground-water sampling procedures. This documentwas developed by the Superfund/RCRA Ground WaterForum and incorporates comments from ORD,Regional Superfund hydrogeologists and others.These guidelines are applicable to the majority ofsites, but are not intended to replace or supersederegional and/or project-specific sampling plans. These

guidelines are intended to assist in developing sam-pling plans using the project-specific goals and objec-tives. However, unusual and/or site-specific circum-stances may require approaches other than thosespecified in this document. In these instances, theappropriate Regional hydrologists/geologists shouldbe contacted to establish alternative protocols.

ACKNOWLEDGMENTSA document of this scope involved significant partici-pation from a number of people, such that any omis-sion in these acknowledgments is purely uninten-tional. We thank all of the participants involved in thedevelopment of this document! The authors acknowl-edge the active participation and valuable input fromthe committee from the Ground Water Forum of DickWilley, Region 1; Ruth Izraeli and Kevin Willis, Region2; Kathy Davies, Region 3; Robert Puls, ORD-NRMRL; and Steve Gardner, ORD-NERL. In addition,valuable input from former members of the committeeare gratefully acknowledged. And finally, the peerreviews of the document completed by FranceskaWilde of the Water Division of the U.S. GeologicalSurvey, Reston, VA; Richard Duwelius and RandyBayless of the Indiana District of the U.S. GeologicalSurvey, Indianapolis, IN; Steve White of the OmahaDistrict of the U.S. Army Corps of Engineers, Omaha,NE and Karl Pohlmann of the Desert ResearchInstitute, Las Vegas, NV are gratefully acknowledged.

Technology Innovative officeOffice of Solid Waste and EmergencyResponse, US EPA, Washington, DC

Walter W. Kovalick, Jr., Ph.D.Director

* U.S. Environmental Protection Agency, Region 5 77 West Jackson Boulevard Chicago, Illinois 60604

**U.S. Environmental Protection Agency, Region 10 1200 Sixth Avenue Seattle, Washington 98101

Ground-Water SamplingGuidelines for Superfund andRCRA Project Managers

2

Office of Solid Waste EPA 542-S-02-001and Emergency May 2002Response www.epa.gov/tio

www.clu-in.org/tio/tsp

INTRODUCTIONThe goal of ground-water sampling is to collectsamples that are “representative” of in-situ ground-water conditions and to minimize changes in ground-water chemistry during sample collection and han-dling. Experience has shown that ground-watersample collection and handling procedures can be asource of variability in water-quality concentrationsdue to differences in sampling personnel, samplingprocedures, and equipment (U.S. EnvironmentalProtection Agency, 1995).

Several different ground-water sampling procedurescan be used, which vary primarily through the criteriaused to determine when a sample is representative ofground-water conditions. No single method or proce-dure is universally applicable to all types of ground-water-sampling programs; therefore, considerationshould be given to a variety of factors when

determining which method is best suited to site-specific conditions. These site-specific conditionsinclude sampling objectives, equipment availability,site location, and physical constraints. This paper willdiscuss each of these conditions and how they maycontribute to the decision in choosing the appropriatesampling methodology and equipment to be usedduring ground-water sampling.

This paper focuses on ground-water sampling proce-dures for monitoring wells only where separate, free-phase, Non-Aqueous Phase Liquids (NAPLs) are notpresent in the monitoring well. Residential and/ormunicipal-production wells where special samplingprocedures and considerations need to be imple-mented are not discussed in this document. Therecommendations made in this paper are based onfindings presented in the current literature, and will besubject to revision as the understanding of ground-water-sampling procedures increases.

TABLE OF CONTENTS

INTRODUCTION........................................................2SAMPLING OBJECTIVES...........................................3INFORMATION NEEDED PRIOR TO SAMPLING........4

BACKGROUND DATA....................................4REFERENCE POINT......................................4TOTAL WELL DEPTH....................................5DEPTH TO WATER.......................................5

GROUND-WATER SAMPLING METHODS..................5PURGING AND SAMPLING DEVICES............6POSITION OF SAMPLE INTAKE....................7PURGE CRITERIA.........................................8

Low-Stress Approach........................8Well-Volume Approach......................8

LOW-PERMEABILITY FORMATIONS.............9DECISION PROCESS FOR DETERMININGAPPLICABLE SAMPLING METHODOLOGY.............10POTENTIAL PROBLEMS.........................................10

SAMPLE CONTAINERS................................10FIELD FILTRATION OF TURBID SAMPLES.10

SAMPLER DECONTAMINATION...............................11POST-SAMPLING ACTIVITIES.................................11

CONCLUSION............................................................11REFERENCES...........................................................11

TABLES1. Stabilization Criteria with References for Water-Quality-Indicator Parameters...............................................172. Applicability of Different Approaches for Purging and Sampling Monitoring Wells...........................................................18

ATTACHMENTS1. Example Sampling Checklist..................... 212. Example Ground-Water Sampling Field Sheets.......................................................253. Example Standard Operating Procedure:

Standard Operating Procedure forLow-Stress (Low Flow)/MinimalDrawdown Ground-Water SampleCollection..........................................29

4. Example Standard Operating Procedure:Standard Operating Procedure for theStandard/Well-Volume Method forCollecting a Ground-Water Sample...41

3

SAMPLING OBJECTIVESThe objective of a good sampling program should bethe collection of a “representative” sample of thecurrent ground-water conditions over a known orspecified volume of aquifer. Ideally to meet thisobjective, sampling equipment, sampling method,monitoring well construction, monitoring welloperation and maintenance, and sample handlingprocedures should not alter the chemistry of thesample. A sample that is obtained from a poorlyconstructed well, or using improper sampling equip-ment, or using poor sampling techniques, or whichhas been preserved improperly, can bias the samplingresults. Unrepresentative samples can lead tomisinterpretations of ground-water-quality data.Generally, the costs of obtaining representativeground-water samples are insignificant whencompared to potential remedial responses that maybe implemented based on erroneous data or whenconsidering the overall monitoring program costs overthe life of the program (Nielson, 1991).

The data quality objectives (DQOs) of the samplingprogram should be thoroughly developed, presentedand understood by all parties involved. To develop theDQOs, the purpose of the sampling effort and datause(s) should be clearly defined. The samplingguidelines presented here can be used for a variety ofmonitoring programs, these include site assessment,contaminant detection, site characterization,remediation, corrective action and compliancemonitoring.

For example DQOs for a site characterizationsampling effort might vary from those of a remediationmonitoring sampling effort. This difference could be inhow much of the screen interval should be sampled. Asite characterization objective may be to collect asample that represents a composite of the entire (oras close as is possible) screened interval of themonitoring well. On the other hand, the monitoringobjective of a remediation monitoring program may beto obtain a sample that represents a specific portion ofthe screened interval.

Additionally, the site characterization may requireanalyses for a broad suite of contaminants, whereas,the remediation monitoring program may requirefewer contaminants to be sampled. These differences

may dictate the type of sampling equipment used, thetype of information collected, and the samplingprotocol.

In order to develop applicable DQOs, a site concep-tual model should be developed. The site conceptualmodel should be a dynamic model which is constantlyrevised as new information is collected and pro-cessed. The conceptual model, as it applies to theDQOs, should focus on contaminant fate and trans-port processes, such as contaminant pathways, howthe geologic materials control the contaminant path-ways (depositional environments, geologic structure,lithology, etc.), types of contaminants present (i.e.,hydrophobic versus hydrophilic), and the processesthat influence concentrations of the contaminantspresent such as dilution, biodegradation, and disper-sion. The detail of the conceptual model will dependgreatly on the availability of information, such as thenumber of borings and monitoring wells and theamount of existing analytical data. Clearly, a site thatis being investigated for the first time will have a muchsimpler conceptual model compared to a site that hashad a Remedial Investigation, Feasibility Study, andRemedial Design, (or, within the RCRA Program, aRCRA Facility Assessment, a RCRA Facility Investiga-tion, and a Corrective Measures Study), and is cur-rently in remediation/corrective action monitoring.Specific parameters that a conceptual model shoulddescribe that may impact the design of a ground-water-sampling program include:

a) The thickness, lateral extent, vertical andhorizontal flow direction, and hydraulic con-ductivity contrasts of the geologic materialscontrolling contaminant transport from the site(thick units versus thin beds versus fractures,etc.)

b) The types of contaminants to be sampled(volatile organic compounds, semi-volatileorganic compounds, metals, etc.) and factorsthat could bias sampling results (turbidity formetals, co-solvation effects on PCBs, etc.)

c) Lateral and vertical distribution of contami-nation (contaminants distributed throughout anentire unit being monitored versus localizeddistribution controlled by small scale features,etc.)

4

Vertical aquifer characterization is strongly recom-mended prior to the completion of a ground-watermonitoring well installation program. A detailed verticalaquifer characterization program should include fieldcharacterization of hydraulic conductivities, determi-nation of vertical and horizontal flow directions, as-sessment of lithologic and geologic variations, anddetermination of vertical and horizontal contaminantdistributions. The successful aquifer characterizationprogram provides detailed information to guide thetechnical and cost-effective placement, vertically andareally, of monitoring wells.

INFORMATION NEEDED PRIOR TO SAMPLINGTo ensure appropriate methodology and expedientcollection of water-quality samples, information isneeded before a sample is collected. Someinformation should be obtained prior to the start offield activities such as well condition, construction,water-level information, contaminant types and con-centrations, and direction(s) of ground-water flow.Field measurements, such as depth to water and totalwell depth will be needed prior to purging. Beforecommencement of all field activities, the field healthand safety plan should be consulted under thedirection of the site health and safety officer.

BACKGROUND DATAWell construction and maintenance information areneeded to better plan the sampling program, optimizepersonnel, and obtain more representative samples.Prior to field activities, personnel should have specificinformation including well casing diameter, boreholediameter, casing material, lock number and keys,physical access to wells, and length of and depth towell screen. The diameter of each well casing is usedto select the correct equipment and technique forpurging and sampling the well. A site map with pos-sible physical barriers and description of access isnecessary to allow for the selection of proper equip-ment based on several factors, such as portability,ease of repair, power sources, containment of purgewater, and well accessibility. The length and depth ofeach well screen and depth to water is importantwhen placing a sampling device’s intake at the properdepth for purging and sampling and for choosing asampling device. Well development information isneeded to ensure that purging and sampling rates willnot exceed well development extraction rates. Previ-ous sampling information should be provided and

evaluated to determine the nature and concentrationsof expected contaminants. This will be useful indetermining the appropriate sampling method andquality assurance/quality control (QA/QC) samples(for example, field duplicates, equipment blanks, tripblanks). Attachment 1 is an example of a samplingchecklist for field personnel. This information shouldbe kept in the field for easy access during samplingactivities.

When evaluating previous sampling information,consideration should be given to the amount of timethat has expired between the last sampling effort andthe planned sampling effort. If this time exceeds oneyear, the need for redevelopment of the monitoringwells should be evaluated. The necessity of redevel-opment can be evaluated by measuring constructeddepth compared to the measured depth. If the depthmeasurement indicates siltation of the monitoring wellscreen, or evidence exists that the well screen isclogged, the well should be redeveloped prior tosampling. The assessment of the condition of themonitoring wells should be completed several weeksprior to sampling activities in order to allow the properrecovery of the developed wells. This is especiallyimportant in wells where prior sampling has indicatedhigh turbidity. The time for a well to re-stabilize afterdevelopment is dependent on site-specific geologyand should be specified in the site sampling plan. Thedevelopment method, if necessary, should be consis-tent with the sampling objectives, best technicalcriteria and USEPA guidelines (Aller et al., 1991;Izraeli et al., 1992; Lapham et al., 1997).

REFERENCE POINTEach well should be clearly marked with a well identi-fier on the outside and inside of the well casing.Additionally, each well should have a permanent,easily identified reference point from which all depthmeasurements are taken. The reference point (the topof the inner casing, outer casing, or security/protec-tive casing) should remain constant through all mea-surements, should be clearly marked on the casingand its description recorded. Whenever possible, theinner casing is recommended as a reference point,because of the general instability of outer casings dueto frost heaving, vehicular damage, and other phe-nomena which could cause movement of casings.The elevation of this reference point should be knownand clearly marked at the well site (Nielson, 1991).

5

This reference point should also have a known latitudeand longitude that are consistent with the Regionaland National Minimum Data Elements requirements.The elevation of the reference point should be sur-veyed relative to Mean Sea Level (MSL) using theNAVD 88 datum.

TOTAL WELL DEPTHThe depth of the well is required to calculate thevolume of standing water in the well and to documentthe amount of siltation that may have occurred.Moreover, measuring the depth to the bottom of a wellprovides checks for casing integrity and for siltation ofthe well screen. Corrosion can cause leaking orcollapse of the well casing, which could lead to erro-neous or misleading water-level measurements.Corrosion, silting, and biofouling can clog wellscreens and result in a sluggish response or noresponse to water-level changes, as well as changesin ground-water chemistry. Well redevelopment orreplacement may be needed to ensure accuratecollection of a representative water-quality sample.

Total well depths should be measured and properlyrecorded to the nearest one-tenth of a foot using asteel tape with a weight attached. The steel tapeshould be decontaminated before use in another wellaccording to the site specific protocols. A concern isthat when the steel tape and weight hit the bottom ofthe well, sediment present on the bottom of a wellmay be stirred up, thus increasing turbidity which willaffect the sampling results. The frequency of total welldepth measurements varies, with no consensus for allhydrogeologic conditions. The United States Geologi-cal Survey (USGS) recommends a minimum of oncea year (Lapham et al., 1997). USEPA also recom-mended one measurement per year (Barcelona et al.,1985) but later recommended a total well depth betaken every time a water-quality is collected or awater-level reading taken (Aller et al., 1991). There-fore, when possible, the total depth measurementsshould be taken following the completion of sampling(Puls and Barcelona, 1996). When total-well-depthmeasurements are needed prior to sampling, asmuch time as possible should be allowed prior tosampling, such as a minimum of 24 hours. The weightof electric tapes are generally too light to determineaccurate total well depth. If the total well depth isgreater than 200 feet, stretching of the tape must betaken into consideration.

DEPTH TO WATERAll water levels should be measured from the refer-ence point by the use of a weighted steel tape andchalk or an electric tape (a detailed discussion of thepros and cons of the different water level devices isprovided in Thornhill, 1989). The steel tape is a moreaccurate method to take water levels, and is recom-mended where shallow flow gradients (less than 0.05foot/feet or 0.015 meter/meters) or deep wells areencountered. However, in those cases where largeflow gradients or large fluctuations in water levels areexpected, a calibrated electric tape is acceptable. Thewater level is calculated using the well’s referencepoint minus the measured depth to water. At depthsapproximately greater than 200 feet, the water-level-measuring device should be chosen carefully, assome devices may have measurable stretching.

The depth-to-water measurement must be made in allwells to be sampled prior to activities in any singlewell which may change the water level, such asbailing, pumping, and hydraulic testing. All readingsare to be recorded to the nearest one-hundredth of afoot.

The time and date of the measurement, point ofreference, measurement method, depth-to-water levelmeasurement, and any calculations should be prop-erly recorded. In addition, any known, outside influ-ences (such as tidal cycles, nearby pumping effects,major barometric changes) that may affect waterlevels should be noted.

GROUND-WATER SAMPLING METHODSThe ground-water sampling methods to be employedshould be dependent on site-specific conditions andrequirements, such as data-quality objectives and wellaccessibility. Ground-water sampling methods varybased on the type of device used, the position of thesampler intake, the purge criteria used, and thecomposition of the ground water to be sampled (e.g.,turbid, containing high volatile organics, etc.). Allsampling methods and equipment should be clearlydocumented, including purge criteria, field readings,etc. Examples of appropriate documentation areprovided in Attachment 2 of this document and Ap-pendix E of the U.S. Environmental Protection Agency,1995 document.

6

The water in the screen and filter pack is generally in aconstant state of natural flux as ground water passesin and out of the well. However, water above thescreened section remains relatively isolated andbecome stagnant. Stagnant water is subject to physio-chemical changes and may contain foreign material,which can be introduced from the surface or duringwell construction, resulting in non-representativesample data. To safeguard against collecting asample biased by stagnant water, specific well-purging guidelines and techniques should be fol-lowed.

A non-representative sample also can result fromexcessive pumping of the monitoring well. Stratifica-tion of the contaminant concentrations in the aquifermay occur, or heavier-than-water compounds maysink to the lower portions of the aquifer. Excessivepumping can dilute or increase the contaminantconcentrations from what is representative of thesampling point.

PURGING AND SAMPLING DEVICESThe device used to purge and sample a well dependson the inner casing diameter, depth to water, volumeof water in the well, accessibility of the well, and typesof contaminants to be sampled. The types of equip-ment available for ground-water sampling includehand-operated or motor-driven suction pumps, peri-staltic pumps, positive displacement pumps, sub-mersible pumps, various in-situ devices and bailersmade of various materials, such as PVC, stainlesssteel and Teflon®. Some of these devices may causevolatilization and produce high pressure differentials,which could result in variability in the results of pH,dissolved oxygen concentrations, oxidation-reductionpotential, specific electrical conductance, and concen-trations of metals, volatile organics and dissolvedgases. Therefore, the device chosen for well purgingand sampling should be evaluated for the possibleeffects it may have on the chemical and physicalanalyses. In addition, the types of contaminants,detection levels, and levels of concern as describedby the site DQOs should be consulted prior to theselection of a sampling device. The same device usedfor purging the monitoring well should be used forsampling to minimize agitation of the water column(which can increase turbidity, increase volatilization,and increase oxygen in the water).

In general, the device used for purging and samplingshould not change geochemical and physical param-eters and/or should not increase turbidity. For thisreason, low-flow submersible or positive-displacementpumps that can control flow rates are recommendedfor purging wells. Dedicated sampling systems aregreatly preferred since they avoid the need for decon-tamination of equipment and minimize turbulence inthe well. If a sampling pump is used, the pump shouldbe lowered into the well as slowly as possible andallowed to sit as long as possible, before pumpingcommences. This will minimize turbidity and volatiliza-tion within the well.

Sampling devices (bladders, pumps, bailers, andtubing) should be constructed of stainless steel,Teflon®, glass, and other inert materials to reduce thechance of these materials altering the ground water inareas where concentrations of the site contaminantsare expected to be near detection limits. The sampletubing thickness should be maximized and the tubinglength should be minimized so that the loss of con-taminants through the tubing walls may be reducedand the rate of stabilization of ground-water param-eters is maximized. The tendency of organics to sorbinto and out of many materials makes the appropriateselection of sample tubing materials critical for thesetrace analyses (Pohlmann and Alduino, 1992; Parkerand Ranney, 1998). Existing Superfund and RCRAguidance suggest appropriate compatible materials(U.S. Environmental Protection Agency, 1992). Spe-cial material considerations are important whensampling for non-routine analyses, such as age-dating and biological constituents.

Preferably, wells should be purged and sampled usinga positive-displacement pump or a low-flow submers-ible pump with variable controlled flow rates andconstructed of chemically inert materials. If a pumpcannot be used because the recovery rate is so slow(less than 0.03 to 0.05 gallons per minute or 100 to200 milliliters per minute) and the volume of the waterto be removed is minimal (less than 5 feet (1.6meters) of water), then a bailer with a double checkvalve and bottom-emptying device with a control-flowcheck valve may be used to obtain the samples.Otherwise, a bailer should not be used when samplingfor volatile organics because of the potential biasintroduced during sampling (Pohlmann, et al., 1990;Yeskis, et al., 1988; Tai, et al., 1991). A peristaltic

7

pump also may be used under these conditions,unless the bias by a negative pressure may impactthe contaminant concentrations of concern (generallyat depths greater than 15 to 20 feet (4.5 to 6 meters)of lift). Bailers should also be avoided when samplingfor metals due to increased turbidity that occurs duringthe deployment of the bailer, which may bias inorganicand strongly hydrophobic parameters. Dedicatedsampling pumps are recommended for metals sam-pling because the pumps avoid the generation ofturbidity from frequent sampler deployment (Puls etal., 1992). A number of alternate sampling devices arebecoming available, including passive diffusion sam-plers (Vroblesky and Hyde, 1997; Vroblesky, 2001aand b) and other in-situ sampling devices. Thesedevices may be particularly useful to sampling low-permeability geologic materials, assuming the deviceis made of materials compatible with the analyticalparameters, meet DQOs, and have been properlyevaluated. However, the site investigator shouldensure the diffusion membrane materials are selectedfor the contaminants of concern (COCs) present atthe site. Comparison tests with an approved samplingmethod and diffusion samplers should be completedto confirm that the method is suitable for the site.

POSITION OF SAMPLE INTAKEEssentially there are two positions for placement ofthe sample pump intake, within the screen and abovethe screen. Each of the positions offers advantagesand disadvantages with respect to the portion of thewell screen sampled, data reproducibility and potentialpurge volumes.

When the sampling pump intake is set above the wellscreen, the pump generally is set just below the waterlevel in the well. The sampling pump then is pumpeduntil a purge criterion is reached (commonly eitherstabilization of purge parameters or a set number ofwell volumes). If the distance between the water leveland the top of the screen is long, there is concern thatthe water will be altered geochemically as it flowsalong the riser pipe, as water flows between the wellscreen and the sampling pump intake. This is espe-cially a concern if the riser pipe is made of similarmaterial as the COC (such as a stainless steel riserwith nickel as a COC, or PVC with organics as aCOC). Keely and Boateng (1987) suggested that tominimize this potential influence, the sample pump belowered gradually while purging, so that at the time of

the sampling the pump intake is just above the screen.This would minimize contact time between the groundwater and the well construction materials while sam-pling, as well as ensure the evacuation of the stagnantwater above the screen.

With the final location of the sampling pump intakejust above the well screen, the sample results may bemore reproducible than those collected by positioningthe pump intake within the well screen. Results maybe more reproducible because the sampler canensure that the ground water is moving into the wellwith the same portions of the aquifer being sampledeach time assuming the same pump rate. If the pumpis placed into different portions of the screen eachtime, different portions of the aquifer may be sampled.Of course, this can be avoided by the use of dedi-cated, permanently installed equipment. Additionally,the placement of the pump at the same verticalposition within the screen can be ensured by the useof calibrated sampling pump hose, sounding with aweighted tape, or using a pre-measured hose.

The placement of the pump above the screen doesnot guarantee the water-quality sample represents theentire well screen length. Any bias in the pump place-ment will be consistently towards the top of the wellscreen and/or to the zone of highest hydraulic conduc-tivity. Another possible disadvantage, or advantage,depending on the DQOs, of the placement of thepump above the well screen is that the sample mayrepresent a composite of water quality over the wellscreen. This may result in dilution of a portion of thescreen that is in a contaminated portion of an aquiferwith another portion that is in an uncontaminatedportion of the aquifer. However, shorter well screenswould minimize this concern.

When the pump intake is positioned within the wellscreen, its location is recommended to be oppositethe most contaminated zone in the well screen inter-val. This method is known as the low-flow, low-stress,micropurge, millipurge, or minimal drawdown method.The well is then purged with a minimal drawdown(usually 0.33 feet (0.1 meters) based on Puls andBarcelona, 1996) until selected water-quality-indicatorparameters have stabilized. Use of this method mayresult in the vertical portion of the sampled aquiferbeing smaller than the well screen length. Thismethod is applicable primarily for short well-screen

8

lengths (less than 5 feet (1.6 meters)) to better char-acterize the vertical distribution of contaminants (Pulsand Barcelona, 1996). This method should not beused with well-screen lengths greater than 10 feet (3meters). By using this method, the volume of purgewater can be reduced, sometimes significantly, overother purging methods.

However, two potential disadvantages of this methodexist. The first potential disadvantage may involve thelower reproducibility of the sampling results. Theposition of the sampling pump intake may vary be-tween sampling rounds (unless adequate precautionsare taken to lower the pump into the exact position inprevious sampling rounds, or a dedicated system isused), which can result in potentially different zoneswithin the aquifer being sampled. This potentialproblem can be overcome by using dedicated sam-pling pumps and the problem may be minimized bythe use of short well screens. The second potentialdisadvantage, or advantage, depending on the DQOs,may be that the sample which is collected may betaken from a small portion of the aquifer volume.

PURGE CRITERIA“Low-Stress Approach”The first method for purging a well, known as the low-stress approach, requires the use of a variable-speed,low-flow sampling pump. This method offers theadvantage that the amount of water to be container-ized, treated, or stored will be minimized. Thelow-stress method is based on the assumption thatpumping at a low rate within the screened zone willnot draw stagnant water down, as long as drawdownis minimized during pumping. Drawdown should notexceed 0.33 feet (0.1 meters) (Puls and Barcelona,1996). The pump is turned on at a low flow rateapproximating the estimated recovery rate (based onthe drawdown within the monitoring well during sam-pling). This method requires the location of the pumpintake to be within the saturated-screened intervalduring purging and sampling. The water-quality-indicator parameters (purge parameters), pH, specificelectrical conductance, dissolved oxygen concentra-tion, oxidation-reduction potential, temperature andturbidity, are monitored at specific intervals. Thespecific intervals will depend on the volume within thetubing (include pump and flow-through cell volumes),pump rate and drawdown; commonly every three to

five minutes. These parameters should be recordedafter a minimum of one tubing volume (include pumpand flow-through-cell volumes) has been purged fromthe well. These water-quality-indicator parametersshould be collected by a method or device whichprevents air from contacting the sample prior to thereading, such as a flow-through cell (Barcelona et al.,1985; Garske and Schock, 1986; Wilde et al., 1998).Once three successive readings of the water-quality-indicator parameters provided in Table 1 have stabi-lized, the sampling may begin. The water-quality-indicator parameters that are recommended includepH and temperature, but these are generally insensi-tive to indicate completion of purging since they tendto stabilize rapidly (Puls and Barcelona, 1996).Oxidation-reduction potential may not always be anappropriate stabilization parameter, and will dependon site-specific conditions. However, readings shouldbe recorded because of its value as a double checkfor oxidizing conditions, and for some fate and trans-port issues. When possible, especially when samplingfor contaminants that may be biased by the presenceof turbidity, the turbidity reading is desired to stabilizeat a value below 10 Nephelometric Turbidity Units(NTUs). For final dissolved oxygen measurements, ifthe readings are less than 1 milligram per liter, theyshould be collected with the spectrophotometricmethod (Wilde et al., 1998, Wilkin et al., 2001),colorimetric or Winkler titration (Wilkin et al., 2001).All of these water-quality-indicator parameters shouldbe evaluated against the specifications of theaccuracy and resolution of the instruments used.

During purging, water-level measurements must betaken regularly at 30-second to five-minute intervals(depending on the hydraulic conductivity of theaquifer, diameter of the well, and pumping rate) todocument the amount of drawdown during purging.The water-level measurements will allow the samplerto control pumping rates to minimize drawdown inthe well.

“Well-Volume Approach”The second method for purging wells is based onproper purging of the stagnant water above thescreened interval and the stabilization of water-quality-indicator parameters prior to sampling. Severalconsiderations in this method need to be evaluatedbefore purging. For monitoring wells where the waterlevel is above the screens, the pump should be set

9

near the top of the water column, and slowly loweredduring the purging process. For watercolumns within the well screen, the pump should beset at a sufficient depth below the water level wheredrawdown during pumping does not allow air to enterthe pump. The pump should not be allowed to touchor draw sediments from the bottom of the well, espe-cially when sampling for parameters that may beimpacted by turbidity. The well-purging rate should notbe great enough to produce excessive turbulence inthe well, commonly no greater than one gallon perminute (3.8 liters per minute) in a 2-inch well. Thepump rate during sampling should produce a smooth,constant (laminar) flow rate, and should not produceturbulence during the filling of bottles. As a result, theexpected flow rate for most wells will be less than onegallon per minute (3.8 liter per minute), with expectedflow rates of about one-quarter gallon per minute (500milliliter per minute).

The stabilization criteria for a “well-volume approach”may be based on the stabilization of water-quality-indicator parameters or on a pre-determined wellvolume. Various research indicates that purgingcriteria based on water-quality-indicator parameterstabilization may not always correlate to stabilizationof other parameters, such as volatile organic com-pounds (Gibs and Imbrigiotta, 1990; Puls et al., 1990).A more technically rigorous sampling approach thatwould yield more consistent results over time wouldbe a time-sequential sampling program at regular well-volume intervals while measuring water-quality-indicator parameters. However, the cost would beprohibitive for most sites. For comparison of water-quality results, by sampling under the same conditions(same purge volume and rate, same equipment,same wells, etc.) temporal evaluations of trends maybe considered.

The stabilization requirements of the water-quality-indicator parameters are consistent with thosedescribed above for the low-stress approach. Theparameters should be recorded approximately everywell volume; when three successive readings havereached stabilization, the sample(s) are taken(Barcelona et al., 1985). If a ground-water monitoringwell has been sufficiently sampled and characterized(at least several rounds of water-quality samplesobtained, including the field parameters, during severalseasonal variations), and if water-quality-indicator

parameters are no longer needed as a part of sitecharacterization and/or monitoring, then samplescould be obtained based on a specific number of wellvolumes at the previous pumping rates.

LOW-PERMEABILITY FORMATIONSDifferent procedures must be followed in the case ofslow-recovery wells installed in low hydraulic conduc-tivity aquifers. The following procedures are notoptimum, but may be used to obtain a ground-watersample under less than ideal conditions. Onesuggested procedure is to remove the stagnant waterin the casing to just above the top of the screenedinterval, in a well screened below the water table, toprevent the exposure of the gravel pack or formationto atmospheric conditions (McAlary and Barker,1987). At no point should the pump be lowered intothe screened interval. The pumping rate should be aslow as possible for purging to minimize the drawdownin the well. However, if a well has an open intervalacross the water table in a low permeability zone,there may be no way to avoid pumping and/or bailinga well dry (especially in those cases with four feet ofwater or less in the well and at a depth to watergreater than 20 to 25 feet (which is the practical limitof a peristaltic pump)). In these cases, the well maybe purged dry. The sample should be taken no soonerthan two hours after purging and after a sufficientvolume for a water-quality sample, or sufficient recov-ery (commonly 90%) is present (Herzog et al., 1988).In these cases, a bailer with a double check valve witha flow-control, bottom-emptying device may be used,since many sampling pumps may have tubing capaci-ties greater than the volume present within the well. Ifthe depth of well and water column are shallowenough, consideration of a very low-flow device, suchas a peristaltic pump, should be considered, espe-cially if constituents are present that are not sensitiveto negative pressures that may be created with theuse of the peristaltic pump. If such constituents arepresent and sampled with a peristaltic pump, a nega-tive bias may be introduced into the sampling results.To minimize the bias, thick-walled, non-porous tubingshould be used, except for a small section in thepump heads, which require a greater degree offlexibility. As stated earlier in this paper, the DQOs forthe sampling should be consulted to consider thepotential impact of the sampling device on the poten-tial bias versus the desired detection levels.

10

Another method to be considered for low-permeabilityconditions is the use of alternative sampling methods,such as passive diffusion samplers and other in-situsamplers. As more sites are characterized with thesealternative sampling methods and devices, the poten-tial bias, if any, can be evaluated with regard to thesampling DQOs. Regional hydrologists/geologists andRegional quality-assurance specialists should beconsulted on the applicability of these methods for thesite-specific conditions.

DECISION PROCESS FOR DETERMININGAPPLICABLE SAMPLING METHODOLOGYOnce the project team has determined the samplingobjectives and DQOs, reviewed the existing data, anddetermined the possible sampling devices that can beused, the team must decide the appropriate samplingmethodology to be used. Table 2 provides a summaryof considerations and rationale to be used in estab-lishing the proper ground-water-sampling programusing site-specific conditions and objectives.

POTENTIAL PROBLEMSThe primary objective is to obtain a sample represen-tative of the ground water moving naturally (includingboth dissolved and particulate species) through thesubsurface. A ground-water sample can be compro-mised by field personnel in two primary ways: takingan unrepresentative sample and handling the (repre-sentative) sample incorrectly. There are numerousways of introducing foreign contaminants into asample. These must be avoided by following strictsampling protocols and transportation procedures,and utilizing trained personnel. Common problemswith sampling include the use of inappropriate samplecontainers and field composites, and the filtration ofturbid samples.

SAMPLE CONTAINERSField samples must be transferred from the samplingequipment to the container that has been specificallyprepared for that given parameter. Samples must notbe composited in a common container in the field andthen split in the lab. The USEPA Regional policy onsample containers should be consulted to determinethe appropriate containers for the specified analysis.

FIELD FILTRATION OF TURBID SAMPLESThe USEPA recognizes that in some hydrogeologicenvironments, even with proper well design, installa-tion, and development, in combination with the low-flow purging and sampling techniques, sample turbid-ity cannot be reduced to ambient levels. The wellconstruction, development, and sampling informationshould be reviewed by the Regional geologists orhydrologists to see if the source of the turbidity prob-lems can be resolved or if alternative sampling meth-odologies should be employed. If the water sample isexcessively turbid, the collection of both filtered andunfiltered samples, in combination with turbidity, TotalSuspended Solids (TSS), Total Dissolved Solids(TDS), pumping rate, and drawdown data is recom-mended. The filter size used to determine TSS andTDS should be the same as used in the field filtration.An in-line filter should be used to minimize contactwith air to avoid precipitation of metals. The typicalfilter media size used is 0.45 µm because this iscommonly accepted as the demarcation betweendissolved and non-dissolved species. Other filtersizes may be appropriate but their use should bedetermined based on site-specific criteria (examplesinclude grain-size distribution, ground-water-flowvelocities, mineralogy) and project DQOs. Filter sizesup to 10.0 µm may be warranted because larger sizefilters may allow particulates that are mobile in groundwater to pass through (Puls and Powell, 1992). Thechanging of filter media size may limit the comparabil-ity of the data obtained with other data sets and mayaffect their use in some geochemical models. Filtermedia size used on previous data sets from a site,region or aquifer and the DQOs should be taken intoconsideration. The filter media used during theground-water sampling program should be collected ina suitable container and archived because potentialanalysis of the media may be helpful for the determi-nation of particulate size, mineralogy, etc.

The first 500 to 1000 milliliters of a ground-watersample (depending on sample turbidity) taken throughthe in-line filter will not be collected for a sample inorder to ensure that the filter media has equilibratedto the sample (manufacturer’s recommendations alsoshould be consulted). Because bailers have beenshown to increase turbidity while purging and sam-pling, bailers should be avoided when sampling fortrace element, metal, PCB, and pesticideconstituents. If portable sampling pumps are used, the

11

pumps should be gently lowered to the sampling depthdesired, carefully avoiding lowering it to the bottom ofthe well, and allowed to sit in order to allow any par-ticles mobilized by pump placement to settle. Dedi-cated sampling equipment installed in the well prior tothe commencement of the sampling activities is oneof the recommended methods to reduce turbidityartifacts (Puls and Powell, 1992; Kearl et al., 1992;Puls et al., 1992; Puls and Barcelona, 1996).

SAMPLER DECONTAMINATIONThe specific decontamination protocol for samplingdevices is dependent on site-specific conditions, typesof equipment used and the types of contaminantsencountered. Once removed from the well, non-dedicated sampling equipment should be decontami-nated to help ensure that there will be no cross-contamination between wells. Disposable items suchas rope and low-grade tubing should be properlydisposed between wells. Cleaning thoroughly thatportion of the equipment that is going to come intocontact with well water is especially important. Inaddition, a clean plastic sheet should be placedadjacent to or around the well to prevent surface soilsfrom coming in contact with the purging and samplingequipment. The effects of cross-contamination can beminimized by sampling the least contaminated wellfirst and progressing to the more contaminated ones.Equipment blanks should be collected on a regularbasis from non-dedicated equipment, the frequencydepending on the sampling plan and regional proto-cols, to document the effectiveness of the decontami-nation procedures.

The preferred method is to use dedicated samplingequipment whenever possible. Dedicated equipmentshould still be cleaned on a regular basis to reducebiofouling, and to minimize adsorption effects. Dedi-cated equipment should have equipment blanks takenafter every cleaning.

POST-SAMPLING ACTIVITIESSpecific activities should be completed at monitoringwells at regular intervals to ensure the acquisition ofrepresentative ground-water samples. Activitiesinclude hydraulic conductivity testing to determine if amonitoring well needs redeveloping and/or replacing.Another activity that needs to be completed is regularsurveying of well measuring points impacted by frost

heaving and site activities. The schedules of theseactivities are to be determined on a site-by-site basisin consultation with regional geologists or hydrologists,but at a minimum, should be every five years.

CONCLUSIONThis document provides a brief summary of the state-of-the-science to be used for Superfund and RCRAground-water studies. As additional research iscompleted, additional sampling experience with othersampling devices and methods and/or additionalcontaminants are identified, this paper may be revisedto include the new information/concerns. Clearly thereis no one sampling method that is applicable for allsampling objectives. As new methods and/or equip-ment are developed, additional standard operatingprocedures (SOPs) should be developed and at-tached to this document. These SOPs for ground-water sampling should include, at a minimum: intro-duction, scope and application, equipment, purgingand sampling procedures, field quality control, decon-tamination procedures and references. ExampleSOP’s for the low-stress/minimal-drawdown and well-volume sampling procedures have been included asAttachments 3 and 4. These example SOPs are to beconsidered a pattern or starting point for site-specificground-water-sampling plans. A more detailed discus-sion of sampling procedures, devices, techniques,etc. is provided in various publications by the USEPA(Barcelona et al., 1985; U.S. Environmental ProtectionAgency, 1993) and the U.S. Geological Survey (Wildeet al., 1998).

REFERENCESAller, L., T.W. Bennett, G. Hackett, R.J. Petty, J.H.Lehr, H. Sedoris, D.M. Nielson and J.E. Denne, 1991,Handbook of Suggested Practices for the Design andInstallation of Ground-Water Monitoring Wells; U.S.Environmental Protection Agency, EPA/600/4-89/034,221 pp.

Barcelona, M.J., J.P. Gibb, J.A. Hellfrich, and E.E.Garske, 1985, Practical Guide for Ground-WaterSampling; U.S. Environmental Protection Agency,EPA/600/2-85/104, 169 pp.

12

Garske, E.E., and M.R. Schock, 1986, An InexpensiveFlow-Through Cell and Measurement System forMonitoring Selected Chemical Parameters in GroundWater; Ground Water Monitoring Review, Vol. 6, No. 3,pp. 79-84.

Gibs, J. and T.E. Imbrigiotta, 1990, Well-PurgingCriteria for Sampling Purgeable Organic Compounds;Ground Water, Vol. 28, No. 1, pp.68-78.

Herzog, B.L., S.J. Chou, J.R. Valkenburg and R.A.Griffin, 1988, Changes in Volatile Organic ChemicalConcentrations After Purging Slowly RecoveringWells; Ground Water Monitoring Review, Vol. 8, No.4, pp. 93-99.

Izraeli, R., D. Yeskis, M. Collins, K. Davies and B.Zavala, 1992, GROUND WATER ISSUE PAPER:Monitoring Well Development Guidelines forSuperfund Project Managers; U.S. EnvironmentalProtection Agency, 4 pp.

Kearl, P.M., N.E. Korte, and T.A. Cronk, 1992, Sug-gested Modifications to Ground Water SamplingProcedures Based on Observations from the ColloidBorescope; Ground Water Monitoring Review, Vol. 12,No. 2, pp. 155-161.

Keely, J.F. and K. Boateng, 1987, Monitoring wellInstallation, Purging, and Sampling Techniques - Part1: Conceptualizations; Ground Water, Vol. 25, No. 4,pp. 427-439.

Lapham, W.W., F.D. Wilde and M.T. Koterba, 1997,Guidelines and Standard Procedures for Studies ofGround-Water Quality: Selection and Installation ofWells, and Supporting Documentation; U.S. Geologi-cal Survey Water-Resources Investigations Report96-4233, 110 pp.

McAlary, T.A. and J.F. Barker, 1987, VolatilizationLosses of Organics During Ground Water Samplingfrom Low Permeability Materials; Ground WaterMonitoring Review, Vol. 7, No. 4, pp. 63-68.

Nielson, D.M., 1991, Practical Handbook of Ground-Water Monitoring; Lewis Publishers, 717 pp.

Parker, L.V. and T.A. Ranney, 1998, Sampling Trace-Level Organic Solutes with Polymeric Tubing: Part 2,Dynamic Studies; Ground Water Monitoring andRemediation, Vol. 18, No. 1, pp. 148-155.

Pohlmann, K.F., R.P. Blegen, and J.W. Hess, 1990,Field Comparison of Ground-Water Sampling Devicesfor Hazardous Waste Sites: An Evaluation usingVolatile Organic Compounds; U.S. EnvironmentalProtection Agency, EPA/600/4-90/028, 102 pp.

Pohlmann, K.F. and A.J. Alduino, 1992, GROUND-WATER ISSUE PAPER: Potential Sources of Error inGround-Water Sampling at Hazardous Waste Sites;U.S. Environmental Protection Agency, EPA/540/S-92/019.

Puls, R.W., J.H. Eychaner, and R.M. Powell, 1990,ENVIRONMENTAL RESEARCH BRIEF: Colloidal-Facilitated Transport of Inorganic Contaminants inGround Water: Part I. Sampling Considerations; U.S.Environmental Protection Agency, EPA/600/M-90/023,12 pp.

Puls, R.W. and R.M. Powell, 1992, Acquisition ofRepresentative Ground Water Quality Samples forMetals; Ground Water Monitoring Review, Vol. 12, No.3, pp. 167-176.

Puls, R.W., D.A. Clark, B. Bledsoe, R.M. Powell andC.J. Paul, 1992, Metals in Ground Water: SamplingArtifacts and Reproducibility; Hazardous Waste andHazardous Materials, Vol. 9, No. 2, pp. 149-162.

Puls, R.W. and M.J. Barcelona, 1996, GROUND-WATER ISSUE PAPER: Low-Flow (Minimal Draw-down) Ground-Water Sampling Procedures; U.S.Environmental Protection Agency, EPA/540/S-95/504,12 pp.

Tai, D.Y., K.S. Turner, and L.A. Garcia, 1991, The Useof a Standpipe to Evaluate Ground Water Samples;Ground Water Monitoring Review, Vol. 11, No. 1, pp.125-132.

Thornhill, J.T., 1989, GROUND-WATER ISSUEPAPER: Accuracy of Depth to Water Measurements;U.S. Environmental Protection Agency, EPA/540/4-89/002, 3 pp.

13

U.S. Environmental Protection Agency, 1992, RCRAGround-Water Monitoring: Draft Technical Guidance;EPA/530-R-93-001.

U.S. Environmental Protection Agency, 1993, Subsur-face Characterization and Monitoring Techniques: ADesk Reference Guide: Volume I: Solids and GroundWater Appendices A and B; EPA/625/R-93/003a.

U.S. Environmental Protection Agency, 1995, GroundWater Sampling - A Workshop Summary, Dallas,Texas, November 30-December 2, 1993; EPA/600/R-94/205, 146 pp.

Vroblesky, D.A., 2001a, User’s Guide for Polyethylene-Based Passive Diffusion Bag Samplers to ObtainVolatile Organic Compound Concentrations in Wells,Part 1: Deployment, Recovery, Data Interpretation,and Quality Control and Assurance; U.S. GeologicalSurvey Water-Resources Investigations Report 01-4060,18 pp.

Vroblesky, D.A. ed., 2001b, User’s Guide for Polyethyl-ene-Based Passive Diffusion Bag Samplers to ObtainVolatile Organic Compound Concentrations in Wells,Part 2: Field Tests; U.S. Geological Survey Water-Resources Investigations Report 01-4061, variouslypaginated.

Vroblesky, D.A. and Hyde, W.T., 1997, DiffusionSamplers as an Inexpensive Approach to MonitoringVOCs in Ground Water; Ground Water Monitoring andRemediation, Vol. 17, No. 3, pp. 177-184.

Wilde, F.D., D.B. Radtke, J.Gibs and R.T. Iwatsubo,eds., 1998, National Field Manual for the Collection ofWater-Quality Data; U.S. Geological Survey Tech-niques of Water-Resources Investigations, Book 9,Handbooks for Water-Resources Investigations,variously paginated.

Wilkin, R.T., M.S. McNeil, C.J. Adair and J.T. Wilson,2001, Field Measurement of Dissolved Oxygen: AComparison of Methods, Ground Water Monitoringand Remediation, Vol. 21, No. 4, pp. 124-132.

Yeskis, D., K. Chiu, S. Meyers, J. Weiss, and T.Bloom, 1988, A Field Study of Various SamplingDevices and Their Effects on Volatile Organic Con-taminants; Proceedings of the Second NationalOutdoor Action Conference on Aquifer Restoration,Ground Water Monitoring and Geophysical Methods,National Water Well Association, May 1988.

14

This page is intentionally blank.

15

TABLES:

Stablization Criteria with References forWater-Quality-Indicator Parameters

andApplicability of Different Approaches for Purging

and Sample Monitoring Wells

16


17

TABLE 1: Stabilization Criteria with References for Water-Quality-Indicator Parameters

Parameter Stabilization Criteria Reference

pH +/- 0.1 Puls and Barcelona, 1996; Wilde et al., 1998

turbidity +/- 10% (when turbidity is Puls and Barcelona, 1996; greater than 10 NTUs) Wilde et al., 1998

dissolved oxygen (DO) +/- 0.3 milligrams per liter Wilde et al., 1998

specific electrical +/- 3% Puls and Barcelona, 1996conductance (SEC)

oxidation-reduction +/- 10 millivolts Puls and Barcelona, 1996potential (ORP)

18

1 Hyd

raul

ic c

ondu

ctiv

ities

of a

quife

r mat

eria

ls v

ary

from

low

hyd

raul

ic c

ondu

ctiv

ities

(cla

ys, s

ilts,

ver

y fin

e sa

nds)

to h

igh

cond

uctiv

ities

(gra

vels

, san

ds, w

eath

ered

bedr

ock

zone

s). T

his

term

for t

he u

se o

n th

is ta

ble

is s

ubje

ctiv

e, a

nd is

mor

e de

pend

ent o

n th

e dr

awdo

wn

indu

ced

in a

mon

itorin

g w

ell w

hen

sam

pled

with

agr

ound

-wat

er s

ampl

ing

pum

p. F

or in

stan

ce, i

n a

wel

l bei

ng p

umpe

d at

4 li

ters

per

min

ute

(l/m

in) w

ith le

ss th

an 0

.1 fe

et o

f dra

wdo

wn,

can

be

cons

ider

ed to

hav

ehi

gh h

ydra

ulic

con

duct

ivity

. A w

ell t

hat c

an s

usta

in a

0.2

to 0

.4 l/

min

pum

ping

rate

, but

has

mor

e th

an 0

.5 fe

et o

f dra

wdo

wn

can

be c

onsi

dere

d to

hav

e lo

why

drau

lic c

ondu

ctiv

ity. T

o as

sign

abs

olut

e va

lues

of h

ydra

ulic

con

duct

iviti

es to

wel

l per

form

ance

and

sus

tain

able

pum

ping

rate

s ca

nnot

be

com

plet

ed b

ecau

se o

fth

e m

any

fact

ors

in m

onito

ring

wel

l con

stru

ctio

n, s

uch

as w

ell d

iam

eter

, scr

een

open

are

a, a

nd le

ngth

of s

cree

n.

2 See

last

par

agra

ph u

nder

the

SAM

PLIN

G O

BJEC

TIVE

S se

ctio

n.

Aqu

ifer/P

lum

eC

hara

cter

izat

ion

Dat

aN

eeds

prio

r to

Cho

osin

g Sa

mpl

ing

Met

hod

2

Low

-Str

ess

App

roac

hW

ell-V

olum

e A

ppro

ach

Oth

ers

(suc

h as

pas

sive

diff

usio

nsa

mpl

ers,

in-s

itu s

ampl

ers,

and

oth

erno

n-tra

ditio

nal g

roun

d-w

ater

sam

plin

gpu

mps

)

App

licab

le G

eolo

gic

Mat

eria

ls1

Con

stitu

ent T

ypes

Met

hod

is A

pplic

able

Dat

a Q

ualit

yO

bjec

tives

Mat

eria

ls w

ith m

oder

ate

tohi

gh h

ydra

ulic

con

duct

iviti

es.

May

be

appl

icab

le to

som

e lo

why

drau

lic c

ondu

ctiv

ities

, if c

anm

eet m

inim

al d

raw

dow

ncr

iteria

.

Hig

h de

finiti

on o

f ver

tical

hydr

aulic

con

duct

ivity

dis

tribu

-tio

n an

d ve

rtica

l con

tam

inan

tdi

strib

utio

n

Mai

nly

reco

mm

ende

d fo

rco

nstit

uent

s w

hich

can

be

bias

ed b

y tu

rbid

ity in

wel

ls.

Appl

icab

le fo

r mos

t oth

erco

ntam

inan

ts.

1) H

igh

reso

lutio

n of

plu

me

defin

ition

bot

h ve

rtica

lly a

ndho

rizon

tally

.2)

Red

uce

bias

from

oth

ersa

mpl

ing

met

hods

if tu

rbid

ity is

of c

once

rn.

3) T

arge

t nar

row

sec

tions

of

aqui

fer.

Mat

eria

ls w

ith lo

w to

hig

hhy

drau

lic c

ondu

ctiv

ities

Plum

e an

d hy

drau

lic c

ondu

ctiv

itydi

strib

utio

ns a

re le

ss c

ritic

al

Appl

icab

le fo

r all

sam

plin

gpa

ram

eter

s. H

owev

er, i

f tur

bidi

tyva

lues

are

ele

vate

d, lo

w-s

tress

appr

oach

may

be

mor

e ap

pli-

cabl

e if

cons

titue

nts

of c

once

rnar

e tu

rbid

ity s

ensi

tive.

1) B

asic

site

cha

ract

eriz

atio

n2)

Mod

erat

e to

hig

h re

solu

tion

ofpl

ume

defin

ition

(will

be d

epen

-de

nt o

n sc

reen

leng

th).

3) T

arge

t sam

ple

com

posi

tion

tore

pres

ent e

ntire

scr

eene

d/op

enin

terv

al

Mat

eria

ls w

ith v

ery

low

to h

igh

hydr

aulic

con

duct

iviti

es

May

nee

d to

con

side

r the

deg

ree

of h

ydra

ulic

and

con

tam

inan

tve

rtica

l dis

tribu

tion

defin

ition

depe

nden

t on

Dat

a Q

ualit

yO

bjec

tives

and

sam

pler

type

.

Con

stitu

ents

of c

once

rn w

ill be

depe

nden

t on

the

type

of

sam

pler

.

1) C

an b

e ap

plic

able

to b

asic

site

cha

ract

eriz

atio

n, d

epen

ding

on s

ampl

er a

nd m

etho

dolo

gyus

ed.

2) C

an re

duce

bia

s fro

m o

ther

sam

plin

g m

etho

ds.

3) M

ay y

ield

hig

h re

solu

tion

ofpl

ume

defin

ition

.

TAB

LE 2

: App

licab

ility

of D

iffer

ent A

ppro

ache

s fo

r Pur

ging

and

Sam

plin

g M

onito

ring

Wel

ls

19

ATTACHMENT 1Example Sampling Checklist

20


21

SAMPLING CHECKLIST

Well Identification:________________________

Map of Site Included: Y or NWells Clearly Identified with Roads: Y or NWell Construction Diagram Attached: Y or N

Well Construction:

Diameter of Borehole:________ Diameter of Casing:__________Casing Material:____________ Screen Material:______________Screen Length:_____________ Total Depth:______________

Approximate Depth to Water:_____________Maximum Well Development Pumping Rate:_________________Date of Last Well Development:_____________

Previous Sampling Information:

Was the Well Sampled Previously: Y or N(If Sampled, Fill Out Table Below)

Table of Previous Sampling Information

ParameterPreviouslySampled

Number ofTimes Sampled

MaximumConcentration Notes (include previous purge rates)

22


23

ATTACHMENT 2Example Ground-Water Sampling Field Sheets

24


25

GROUND-WATER SAMPLING RECORD Well ID:_______________

Station #:______________Facility Name: Date:____/____/____

Well Depth:__________ Depth to Water:__________ Well Diameter:___________

Casing Material.:__________ Volume Of Water per Well Volume:______________

Sampling Crew:__________________,____________________,___________________,______________________

Type of Pump:_________ ___________ Tubing Material:__________________ Pump set at _________________ ft.

Weather Conditions:_________________________________ NOTES:_________ ________________________

______________________________________________________________________________________________

Other Parameters: ___________________Sampled at:_______________ Parameters taken with :_________________________________________Sample delivered to ______________________________ by ____________________________ at___________.Sample CRL #:______________ OTR #:______________ ITR #:______________ SAS #:__________________

Parameters Collected Number of Bottles Bottle Lot Number

_____________________________________ _________ _______________

_____________________________________ _________ _______________

_____________________________________ _________ _______________

_____________________________________ _________ _______________

_____________________________________ _________ _______________

Temp.(0C)Time

WaterLevel

VolumePumped

PumpingRate

DO(mg/l)

SEC(µS/cm) pH

ORP(mV)

Turbidity(NTU)

GROUND-WATER SAMPLING PARAMETERS

26

Ground Water Sampling Log

Site Name: Well #: Date:Well Depth( Ft-BTOC1): Screen Interval(Ft):

Well Dia.: Casing Material: Sampling Device:

Pump placement(Ft from TOC2):

Measuring Point: Water level (static)(Ft):

Water level (pumping)(Ft): Pump rate(Liter/min):

Sampling Personnel:

Other info: (such as sample numbers, weather conditions and field notes)

Water Quality Indicator Parameters

Type of Samples collected:

1 casing volume was:Total volume purged prior to sample collection:1BTOC-Below Top of Casing2TOC-Top of Casing3Specific Electrical Conductance

Stabilization Criteria

D.O. +/- 0.3 mg/lTurb. +/- 10%S.C. +/- 3%ORP +/- 10 mVpH +/- 0.1 unit

ORP(mv)

DO(mg/L)

Waterlevel(ft)

Pumpingrates

(L/Min)

Time Volumepumped

(L)

Temp.(C0)

pHTurb.(NTU)

SEC3

27

ATTACHMENT 3Example Standard Operating Procedure:

Standard Operating Procedure forLow-Stress (Low Flow)/Minimal Drawdow

Ground-Water Sample Collection

28


29

INTRODUCTIONThe collection of “representative” water samples fromwells is neither straightforward nor easily accom-plished. Ground-water sample collection can be asource of variability through differences in samplepersonnel and their individual sampling procedures,the equipment used, and ambient temporal variabilityin subsurface and environmental conditions. Manysite inspections and remedial investigations requirethe sampling at ground-water monitoring wells withina defined criterion of data confidence or data quality,which necessitates that the personnel collecting thesamples are trained and aware of proper sample-collection procedures.

The purpose of this standard operating procedure(SOP) is to provide a method that minimizes theimpact the purging process has on the ground-waterchemistry and the volume of water that is beingpurged and disposed of during sample collection. Thiswill take place by placing the pump intake within thescreen interval and by keeping the drawdown at aminimal level (0.33 feet) (Puls and Barcelona, 1996)until the water quality parameters have stabilized andsample collection is complete. The flow rate at whichthe pump will be operating will depend upon bothhydraulic conductivity of the aquifer and the drawdownwith the goal of minimizing the drawdown. The flowrate from the pump during purging and sampling willbe at a rate that will not compromise the integrity ofthe analyte that is being sampled. This samplingprocedure may or may not provide a discrete ground-water sample at the location of the pump intake. Theflow of ground-water to the pump intake will be depen-dent on the distribution of the hydraulic conductivity (K)of the aquifer within the screen interval. In order tominimize the drawdown in the monitoring well, a low-flow rate must be used. “Low-Flow” refers to thevelocity with which water enters the pump intake fromthe surrounding formation in the immediate vicinity ofthe well screen. It does not necessarily refer to theflow rate of water discharged at the surface, whichcan be affected by flow regulators or restrictions (Pulsand Barcelona, 1996). This SOP was developed bythe Superfund/RCRA Ground Water Forum and drawsfrom an USEPA’s Ground Water Issue Paper, Low-Flow (Minimal Drawdown) Ground-Water SamplingProcedure, by Robert W. Puls and Michael J.Barcelona. Also, available USEPA Regional SOPs

regarding Low-Stress (Low-Flow) Purging and Sam-pling were used for this SOP.

SCOPE AND APPLICATIONThis SOP should be used primarily at monitoring wellsthat have a screen or an open interval with a length often feet or less and can accept a sampling device thatminimizes the disturbance to the aquifer or the watercolumn in the well casing. The screen or open intervalshould have been optimally located to intercept anexisting contaminant plume(s) or along flowpaths ofpotential contaminant releases. Knowledge of thecontaminant distribution within the screen interval ishighly recommended and is essential for the successof this sampling procedure. The ground-watersamples that are collected using this procedure areacceptable for the analyses of ground-water contami-nants that may be found at Superfund and RCRAcontamination sites. The analytes may be volatile,semi-volatile organic compounds, pesticides, PCBs,metals, and other inorganic compounds. Thescreened interval should be located within the con-taminant plume(s) and the pump intake should beplaced at or near the known source of the contamina-tion within the screened interval. It is critical to placethe pump intake in the exact location or depth foreach sampling event. This argues for the use ofdedicated, permanently installed, sampling deviceswhenever possible. If this is not possible, then theplacement of the pump intake should be positionedwith a calibrated sampling pump hose sounded with aweighted-tape or using a pre-measured hose. Thepump intake should not be placed near the bottom ofthe screened interval to avoid disturbing any sedimentthat may have settled at the bottom of the well.

Water-quality-indicator parameters and water levelsmust be measured during purging, prior to samplecollection. Stabilization of the water-quality-indicatorparameters as well as monitoring water levels are aprerequisite to sample collection. The water-quality-indicator parameters that are recommended includethe following: specific electrical conductance, dis-solved oxygen, turbidity, oxidation-reduction potential,pH, and temperature. The latter two parameters areuseful data, but are generally insensitive as purgingparameters. Oxidation-reduction potential may notalways be appropriate stabilization parameter, and willdepend on site-specific conditions. However, readings

Standard Operating Procedure for Low-Stress (Low-Flow)/Minimal Drawdown Ground-Water Sample Collection

30

should be recorded because of its value as a doublecheck for oxidation conditions and for fate and trans-port issues.

Also, when samples are collected for metals, semi-volatile organic compounds, and pesticides, everyeffort must be made to reduce turbidity to 10 NTUs orless (not just the stabilization of turbidity) prior to thecollection of the water sample. In addition to themeasurement of the above parameters, depth towater must be measured during purging (U.S. Envi-ronmental Protection Agency, 1995).

Proper well construction, development, and mainte-nance are essential for any ground-water samplingprocedure. Prior to conducting the field work, informa-tion on the construction of the well and well develop-ment should be obtained and that information factoredinto the site specific sampling procedure. The Sam-pling Checklist at the end of this attachment is anexample of the type of information that is useful.

Stabilization of the water-quality-indicator parametersis the criterion for sample collection. But if stabilizationis not occurring and the procedure has been strictlyfollowed, then sample collection can take place oncethree (minimum) to six (maximum) casing volumeshave been removed (Schuller et al., 1981 and U.S.Environmental Protection Agency., 1986; Wilde et al.,1998; Gibs and Imbrigiotta., 1990). The specificinformation on what took place during purging mustbe recorded in the field notebook or in the ground-water sampling log.

This SOP is not to be used where non-aqueousphase liquids (NAPL) (immiscible fluids) are present inthe monitoring well.

EQUIPMENT! Depth-to-water measuring device - An electronic

water-level indicator or steel tape and chalk, withmarked intervals of 0.01 foot. Interface probe fordetermination of liquid products (NAPL) presence,if needed.

! Steel tape and weight - Used for measuring totaldepth of well. Lead weight should not be used.

! Sampling pump - Submersible or bladder pumpswith adjustable rate controls are preferred. Pumpsare to be constructed of inert materials, such as

stainless steel and Teflon®. Pump types that areacceptable include gear and helical driven, cen-trifugal (low-flow type), and air-activated piston. Anadjustable rate, peristaltic pump can be usedwhen the depth to water is 20 feet or less.

! Tubing - Teflon® or Teflon®-lined polyethylenetubing is preferred when sampling for organiccompounds. Polyethylene tubing can be usedwhen sampling inorganics.

! Power source - If a combustion type (gasoline ordiesel-driven) generator is used, it must be placeddownwind of the sampling area.

! Flow measurement supplies - flow meter, gradu-ated cylinder, and a stop watch.

! Multi-parameter meter with flow-through cell - Thiscan be one instrument or more contained in aflow-through cell. The water-quality-indicatorparameters that are monitored are pH, ORP/Eh,(ORP) dissolved oxygen (DO), turbidity, specificconductance, and temperature. Turbidity readingsmust be collected before the flow cell because ofthe potential for sediment buildup, which can biasthe turbidity measurements. Calibration fluids forall instruments should be NIST-traceable and thereshould be enough for daily calibration throughoutthe sampling event. The inlet of the flow cell mustbe located near the bottom of the flow cell and theoutlet near the top. The size of the flow cell shouldbe kept to a minimum and a closed cell is pre-ferred. The flow cell must not contain any air orgas bubbles when monitoring for the water-quality-indicator parameters.

! Decontamination supplies - Including a reliable anddocumented source of distilled water and anysolvents (if used). Pressure sprayers, buckets ordecontamination tubes for pumps, brushes andnon-phosphate soap will also be needed.

! Sample bottles, sample preservation supplies,sample tags or labels, and chain-of-custodyforms.

! Approved Field Sampling and Quality AssuranceProject Plan.

! Well construction, field, and water quality datafrom the previous sampling event.

! Well keys and map of well locations.! Field notebook, ground-water sampling logs, and

calculator. A suggested field data sheet (ground-water sampling record or ground-water samplinglog) are provided at the end of this attachment.

31

! Filtration equipment, if needed. An in-line dispos-able filter is recommended.

! Polyethylene sheeting placed on ground aroundthe well head.

! Personal protective equipment as specified in thesite Health and Safety Plan.

! Air monitoring equipment as specified in the SiteHealth and Safety Plan.

! Tool box - All needed tools for all site equipmentused.

! A 55-gallon drum or container to contain thepurged water.

Construction materials of the sampling equipment(bladders, pumps, tubing, and other equipment thatcomes in contact with the sample) should be limited tostainless steel, Teflon®, glass, and other inert mate-rial. This will reduce the chance that sampling materi-als alter the ground-water where concentrations of thesite contaminants are expected to be near the detec-tion limits. The sample tubing diameter should bemaximized and the tubing length should be minimizedso that the loss of contaminants into and through thetubing walls may be reduced and the rate of stabiliza-tion of ground-water parameters is maximized. Thetendency of organics to sorb into and out of materialmakes the appropriate selection of sample tubingmaterial critical for trace analyses (Pohlmann andAlduino, 1992; Parker and Ranney, 1998).

PURGING AND SAMPLING PROCEDURESThe following describes the purging and samplingprocedures for the Low-Stress (Low-Flow)/ MinimalDrawdown method for the collection of ground-watersamples. These procedures also describe steps fordedicated and non-dedicated systems.

Pre-Sampling Activities (Non-dedicated and dedicatedsystem)

1. Sampling must begin at the monitoring well with theleast contamination, generally up-gradient or farthestfrom the site or suspected source. Then proceedsystematically to the monitoring wells with the mostcontaminated ground water.

2. Check and record the condition of the monitoringwell for damage or evidence of tampering. Lay outpolyethylene sheeting around the well to minimize the

likelihood of contamination of sampling/purging equip-ment from the soil. Place monitoring, purging andsampling equipment on the sheeting.

3. Unlock well head. Record location, time, date, andappropriate information in a field logbook or on theground-water sampling log (See attached ground-water sampling record and ground-water sampling logas examples).

4. Remove inner casing cap.

5. Monitor the headspace of the monitoring well at therim of the casing for volatile organic compounds(VOC) with a photo-ionization detector (PID) or flameionization detector (FID) and record in the logbook. Ifthe existing monitoring well has a history of positivereadings of the headspace, then the sampling mustbe conducted in accordance with the Health andSafety Plan.

6. Measure the depth to water (water level must bemeasured to nearest 0.01 feet) relative to a referencemeasuring point on the well casing with an electronicwater level indicator or steel tape and record in log-book or ground-water sampling log. If no referencepoint is found, measure relative to the top of the innercasing, then mark that reference point and note thatlocation in the field logbook. Record information ondepth to ground water in the field logbook or ground-water sampling log. Measure the depth to water asecond time to confirm initial measurement; measure-ment should agree within 0.01 feet or re-measure.

7. Check the available well information or field infor-mation for the total depth of the monitoring well. Usethe information from the depth of water in step six andthe total depth of the monitoring well to calculate thevolume of the water in the monitoring well or thevolume of one casing. Record information in fieldlogbook or ground-water sampling log.

Purging and Sampling Activities8A. Non-dedicated system - Place the pump andsupport equipment at the wellhead and slowly lowerthe pump and tubing down into the monitoring welluntil the location of the pump intake is set at a pre-determined location within the screen interval. Theplacement of the pump intake should be positioned

32

with a calibrated sampling pump hose, sounded with aweighted-tape, or using a pre-measured hose. Referto the available monitoring well information to deter-mine the depth and length of the screen interval.Measure the depth of the pump intake while loweringthe pump into location. Record pump location in fieldlogbook or ground-water sampling log.

8B. Dedicated system - Pump has already beeninstalled, refer to the available monitoring well informa-tion and record the depth of the pump intake in thefield logbook or ground-water sampling log.

9. Non-dedicated system and dedicated systems -Measure the water level (water level must be mea-sured to nearest 0.01 feet) and record information onthe ground-water sampling log, leave water levelindicator probe in the monitoring well.

10. Non-dedicated and dedicated systems - Connectthe discharge line from the pump to a flow-throughcell. A “T” connection is needed prior to the flow-through cell to allow for the collection of water for theturbidity measurements. The discharge line from theflow-through cell must be directed to a container tocontain the purge water during the purging and sam-pling of the monitoring well.

11. Non-dedicated and dedicated systems - Startpumping the well at a low flow rate (0.2 to 0.5 liter perminute) and slowly increase the speed. Check waterlevel. Maintain a steady flow ratewhile maintaining a drawdown ofless than 0.33 feet (Puls andBarcelona, 1996). If drawdown isgreater than 0.33 feet, lower theflow rate. 0.33 feet is a goal to helpguide with the flow rate adjust-ment. It should be noted that thisgoal may be difficult to achieveunder some circumstances due togeologic heterogeneities within thescreened interval, and may requireadjustment based on site-specificconditions and personal experi-ence (Puls and Barcelona, 1996).

12. Non-dedicated and dedicatedsystems - Measure the discharge

rate of the pump with a graduated cylinder and a stopwatch. Also, measure the water level and record bothflow rate and water level on the ground-water sam-pling log. Continue purging, monitor and record waterlevel and pump rate every three to five minutes duringpurging. Pumping rates should be kept at minimal flowto ensure minimal drawdown in the monitoring well.

13. Non-dedicated and dedicated systems - Duringthe purging, a minimum of one tubing volume (includ-ing the volume of water in the pump and flow cell)must be purged prior to recording the water-qualityindicator parameters. Then monitor and record thewater-qualityindicator parameters every three to fiveminutes. The water-quality indicator field parametersare turbidity, dissolved oxygen, specific electricalconductance, pH, redox potential, and temperature.Oxidation-reduction potential may not always be anappropriate stabilization parameter, and will depend onsite-specific conditions. However, readings should berecorded because of its value as a double check foroxidizing conditions. Also, for the final dissolvedoxygen measurement, if the readings are less than 1milligram per liter, it should be collected and analyzewith the spectrophotometric method (Wilde et al.,1998 Wilkin et al., 2001), colorimetric or Winklertitration (Wilkin et al., 2001). The stabilization criterionis based on three successive readings of the waterquality field parameters; the following are the criteriawhich must be used:

Parameter Stabilization Criteria Reference

pH +/- 0.1 pH units Puls and Barcelona, 1996;

Wilde et al., 1998

specific electrical +/- 3% S/cm Puls and Barcelona, 1996

conductance (SEC)

turbidity +/- 10% NTUs (when turbidity Puls and Barcelona, 1996;

is greater than 10 NTUs) Wilde et al., 1998

dissolved oxygen +/- 0.3 milligrams per liter Wilde et al., 1998

oxidation-reduction +/- 10 millivolts Puls and Barcelona, 1996

potential (ORP)

33

Once the criteria have been successfully met indicat-ing that the water quality indicator parameters havestabilized, then sample collection can take place.

14. If a stabilized drawdown in the well can’t be main-tained at 0.33 feet and the water level is approachingthe top of the screened interval, reduce the flow rate orturn the pump off (for 15 minutes) and allow for recov-ery. It should be noted whether or not the pump has acheck valve. A check valve is required if the pump isshut off. Under no circumstances should the well bepumped dry. Begin pumping at a lower flow rate, if thewater draws down to the top of the screened intervalagain, turn pump off and allow for recovery. If twotubing volumes (including the volume of water in thepump and flow cell) have been removed during purg-ing, then sampling can proceed next time the pump isturned on. This information should be noted in the fieldnotebook or ground-water sampling log with a recom-mendation for a different purging and sampling proce-dure.

15. Non-dedicated and dedicated systems - Maintainthe same pumping rate or reduce slightly for sampling(0.2 to 0.5 liter per minute) in order to minimizedisturbance of the water column. Samples should becollected directly from the discharge port of the pumptubing prior to passing through the flow-through cell.Disconnect the pump’s tubing from the flow-throughcell so that the samples are collected from the pump’sdischarge tubing. For samples collected for dissolvedgases or VOC analyses, the pump tubing needs to becompletely full of ground water to prevent the groundwater from being aerated as it flows through thetubing. The sequence of the samples is immaterialunless filtered (dissolved) samples are collected andthey must be collected last (Puls and Barcelona,1996). All sample containers should be filled withminimal turbulence by allowing the ground water toflow from the tubing gently down the inside of thecontainer. When filling the VOC samples, a meniscusmust be formed over the mouth of the vial to eliminatethe formation of air bubbles and head space prior tocapping. In the event that the ground water is turbid,(greater then 10 NTUs), a filtered metal (dissolved)sample also should be collected.

If filtered metal sample is to be collected, then an in-line filter is fitted at the end of the discharge tubingand the sample is collected after the filter. The in-line

filter must be pre-rinsed following manufacturer’srecommendations and if there are no recommenda-tions for rinsing, a minimum of 0.5 to 1 liter of groundwater from the monitoring well must pass through thefilter prior to sampling.

16A. Non-dedicated system - Remove the pump fromthe monitoring well. Decontaminate the pump anddispose of the tubing if it is non-dedicated.

16B. Dedicated system - Disconnect the tubing thatextends from the plate at the wellhead (or cap) anddiscard after use.

17. Non-dedicated system - Before locking the moni-toring well, measure and record the well depth (to 0.1feet).

Measure the total depth a second time to confirminitial measurement; measurement should agreewithin 0.01 feet or re-measure.

18. Non-dedicated and dedicated systems - Closeand lock the well.

DECONTAMINATION PROCEDURESDecontamination procedures for the water level meterand the water quality field parameter sensors.The electronic water level indicator probe/steel tapeand the water-quality field parameter sensors will bedecontaminated by the following procedures:

1. The water level meter will be hand washed withphosphate-free detergent and a scrubber, then thor-oughly rinsed with distilled water.

2. Water quality field parameter sensors and flow-through cell will be rinsed with distilled water betweensampling locations. No other decontamination proce-dures are necessary or recommended for theseprobes since they are sensitive. After the samplingevent, the flow cell and sensors must be cleaned andmaintained per the manufacturer’s requirements.

Decontamination Procedure for the Sampling PumpUpon completion of the ground water sample collec-tion the sampling pump must be properly decontami-nated between monitoring wells. The pump anddischarge line including support cable and electrical

34

wires which were in contact with the ground water inthe well casing must be decontaminated by thefollowing procedure:

1. The outside of the pump, tubing, support cable andelectrical wires must be pressure-sprayed withsoapy water, tap water, and distilled water. Sprayoutside of tubing and pump until water is flowing offof tubing after each rinse. Use bristle brush to helpremove visible dirt and contaminants.

2. Place the sampling pump in a bucket or in a shortPVC casing (4-in. diameter) with one end capped.The pump placed in this device must be completelysubmerged in the water. A small amount of phos-phate-free detergent must be added to the potablewater (tap water).

3. Remove the pump from the bucket or 4-in. casingand scrub the outside of the pump housing andcable.

4. Place pump and discharge line back in the 4-in.casing or bucket, start pump and recirculate thissoapy water for 2 minutes (wash).

5. Re-direct discharge line to a 55-gallon drum. Con-tinue to add 5 gallons of potable water (tap water) oruntil soapy water is no longer visible.

6. Turn pump off and place pump into a second bucketor 4-in. casing that contains tap water. Continue toadd 5 gallons of tap water (rinse).

7. Turn pump off and place pump into a third bucket or4-in. casing which contains distilled/deionizedwater, continue to add 3 to 5 gallons of distilled/deionized water (final rinse).

8. If a hydrophobic contaminant is present (such asseparate phase, high levels of PCBs, etc.), anadditional decontamination step, or steps, may beadded. For example, an organic solvent, such asreagent-grade isopropanol alcohol may be added asa first spraying/bucket prior to the soapy waterrinse/bucket.

FIELD QUALITY CONTROLQuality control (QC) samples must be collected toverify that sample collection and handling procedureswere performed adequately and that they have notcompromised the quality of the ground-watersamples. The appropriate EPA program guidancemust be consulted in preparing the field QC samplerequirements for the site-specific Quality AssuranceProject Plan (QAPP).

There are five primary areas of concern for qualityassurance (QA) in the collection of representativeground-water samples:

1. Obtaining a ground-water sample that isrepresentative of the aquifer or zone of interest inthe aquifer. Verification is based on the field logdocumenting that the field water-qualityparameters stabilized during the purging of thewell, prior to sample collection.

2. Ensuring that the purging and sampling devicesare made of materials, and utilized in a mannerthat will not interact with or alter the analyses.

3. Ensuring that results generated by theseprocedures are reproducible; therefore, thesampling scheme should incorporate co-locatedsamples (duplicates).

4. Preventing cross-contamination. Sampling shouldproceed from least to most contaminated wells, ifknown. Field equipment blanks should beincorporated for all sampling and purgingequipment, and decontamination of the equipmentis therefore required.

5. Properly preserving, packaging, and shippingsamples.

All field QC samples must be prepared the same asregular investigation samples with regard to samplevolume, containers, and preservation. The chain-of-custody procedures for the QC samples will beidentical to the field ground-water samples. Thefollowing are QC samples that must be collectedduring the sampling event:

Sample Type Frequency! Field duplicates 1 per 20 samples! Matrix spike 1 per 20 samples! Matrix spike duplicate 1 per 20 samples! Equipment blank per Regional

requirements or policy! Trip blank (VOCs) 1 per sample cooler! Temperature blank 1 per sample cooler

35

HEALTH AND SAFETY CONSIDERATIONSDepending on the site-specific contaminants, variousprotective programs must be implemented prior tosampling the first well. The site Health and Safety Planshould be reviewed with specific emphasis placed onthe protection program planned for the samplingtasks. Standard safe operating practices should befollowed, such as minimizing contact with potentialcontaminants in both the liquid and vapor phasethrough the use of appropriate personal protectiveequipment.

Depending on the type of contaminants expected ordetermined in previous sampling efforts, the followingsafe work practices will be employed:

Particulate or metals contaminants1. Avoid skin contact with, and incidental ingestion of,

purge water.2. Use protective gloves and splash protection.

Volatile organic contaminants1. Avoid breathing constituents venting from well.2. Pre-survey the well head space with an appropri-

ate device as specified in the site Health andSafety Plan.

3. If monitoring results indicate elevated organicconstituents, sampling activities may be con-ducted in level C protection. At a minimum, skinprotection will be afforded by disposable protectiveclothing, such as Tyvek®.

General practices should include avoiding skin contactwith water from preserved sample bottles, as thiswater will have pH less than 2 or greater than 10. Also,when filling pre-acidified VOA bottles, hydrochloricacid fumes may be released and should not be in-haled.

POST-SAMPLING ACTIVITIESSeveral activities need to be completed and docu-mented once ground-water sampling has been com-pleted. These activities include, but are not limited tothe following:

1. Ensuring that all field equipment has been decon-taminated and returned to proper storage location.

Once the individual field equipment has beendecontaminated, tag it with date of cleaning, sitename, and name of individual responsible.

2. Processing all sample paperwork, including copiesprovided to the Regional Laboratory, SampleManagement Office, or other appropriate samplehandling and tracking facility.

3. Compiling all field data for site records.4. Verifying all analytical data processed by the

analytical laboratory against field sheets to ensureall data has been returned to sampler.

REFERENCESGibs, J. and T.E. Imbrigiotta, 1990, Well-PurgingCriteria for Sampling Purgeable Organic Compounds;Ground Water, Vol. 28, No. 1, pp 68-78.

Pohlmann, K.F. and A.J. Alduino, 1992, GROUND-WATER ISSUE PAPER: Potential Sources of Error inGround-Water Sampling at Hazardous Waste Sites,US Environmental Protection Agency. EPA/540/S-92/019.

Puls, R.W. and M.J. Barcelona, 1996, GROUND-WATER ISSUE PAPER: Low-Flow (Minimal Draw-down) Ground-Water Sampling Procedure, US Envi-ronmental Protection Agency. EPA/540/S-95/504, 12pp.

Schuller, R.M., J.P. Gibb and R.A Griffin, 1981, Rec-ommended Sampling Procedures for MonitoringWells; Ground Water Monitoring Review, Spring 1981,pp. 42-46.


U.S. Environmental Protection Agency, 1986, RCRAGround-Water Monitoring Technical EnforcementGuidance Document; OSWER-9950.1, U.S. Govern-ment Printing Office, Washington, D.C., 208 pp.,appendices.

U.S. Environmental Protection Agency, 1995, GroundWater Sampling - A Workshop Summary, Texas,November 30-December 2, 1993, EPA/600/R-94/205,146 pp.

36

U.S. Environmental Protection Agency Region 1,1996, Low Stress (low flow) Purging and SamplingProdedure for the Collection of Ground water SamplesFrom Monitoring Wells, SOP#: GW 0001, July 30,1996.

U.S. Environmental Protection Agency Region 2,1998, Ground Water Sampling Procedure Low Stress(Low Flow) Purging and Sampling, GW SamplingSOP Final, March 16, 1998.

Wilde, F.D., D.B. Radtke, J.Gibs and R.T. Iwatsubo,eds., 1998, National Field Manual for the Collection ofWater-Quality Data; U.S. Geological Survey Tech-niques of Water-Resources Investigations, Book 9,Handbooks for Water-Resources Investigations,variously paginated.

Wilkin, R.T., M.S. McNeil, C.J. Adair and J.T. Wilson,2001, Field Measurement of Dissolved Oxygen: AComparison of Methods, Ground Water Monitoring andRemediation, Vol. 21, No. 4, pp. 124-132.

37

SAMPLING CHECKLIST

Well Identification:________________________

Map of Site Included: Y or NWells Clearly Identified with Roads: Y or NWell Construction Diagram Attached: Y or N

Well Construction:

Diameter of Borehole:________ Diameter of Casing:__________Casing Material:____________ Screen Material:______________Screen Length:_____________ Total Depth:______________

Approximate Depth to Water:_____________Maximum Well Development Pumping Rate:_________________Date of Last Well Development:_____________

Previous Sampling Information:

Was the Well Sampled Previously: Y or N(If Sampled, Fill Out Table Below)

Table of Previous Sampling Information

ParameterPreviouslySampled

Number ofTimes Sampled

MaximumConcentration Notes (include previous purge rates)

38

Ground Water Sampling Log

Site Name: Well #: Date:Well Depth( Ft-BTOC1): Screen Interval(Ft):

Well Dia.: Casing Material: Sampling Device:

Pump placement(Ft from TOC2):

Measuring Point: Water level (static)(Ft):

Water level (pumping)(Ft): Pump rate(Liter/min):

Sampling Personnel:

Other info: (such as sample numbers, weather conditions and field notes)

Water Quality Indicator Parameters

Type of Samples collected:

1 casing volume was:

Total volume purged priorto sample collection:

1BTOC-Below Top of Casing2TOC-Top of Casing3Specific Electrical Conductance

Stabilization Criteria

D.O. +/- 0.3 mg/lTurb. +/- 10%S.C. +/- 3%ORP +/- 10 mVpH +/- 0.1 unit

ORP(mv)

DO(mg/L)

Waterlevel(ft)

Pumpingrates

(L/Min)

Time Volumepumped

(L)

Temp.(C0)

pHTurb.(NTU)

SEC3

(S/cm)

39

ATTACHMENT 4Example Standard Operating Procedure:

Standard Operating Procedure forthe Standard/Well-Volume Method for

Collecting a Ground-Water Sample

40


41

INTRODUCTIONThe collection of “representative” water samples fromwells is neither straightforward nor easily accom-plished. Ground-water sample collection can be asource of variability through differences in samplingpersonnel and their individual sampling procedures,the equipment used, and ambient temporal variabilityin subsurface and environmental conditions. Manysite inspections and remedial investigations requirethe sampling at ground-water monitoring wells withina defined criterion of data confidence or data quality,which necessitates that the personnel collecting thesamples are trained and aware of proper sample-collection procedures.

The objectives of the sampling procedures describedin this document are to minimize changes in ground-water chemistry during sample collection and trans-port to the laboratory and to maximize the probabilityof obtaining a representative, reproducible ground-water sample. Sampling personnel may benefit from aworking knowledge of the chemical processes thatcan influence the concentration of dissolved chemicalspecies.

The well-volume method described in this standardoperating procedure (SOP) provides a reproduciblesampling technique with the goal that the samplesobtained will represent water quality over an entireopen interval of a short-screened (ten feet or less)well. This technique is appropriate for long-term anddetection monitoring of formation water quality. Theresulting sample generally represents a composite ofthe screened interval, and thus integrates small-scalevertical heterogeneities of ground-water chemistry.This sampling technique also is useful for screeningpurposes for detection monitoring of contaminants inthe subsurface. However, the detection of a lowconcentration of contaminant in a thin contaminatedzone or with long well screens may be difficult andshould be determined using detailed vertical profilingtechniques.

This method may not be applicable for all ground-water-sampling wells, such as wells with very lowyields, fractured rock, and some wells with turbidityproblems. As always, site-specific conditions andobjectives should be considered prior to the selectionof this method for sampling.

SCOPE AND APPLICATIONThe objective of a good sampling program should bethe collection of a representative sample of the cur-rent ground-water conditions over a known or speci-fied volume of aquifer. To meet this objective, thesampling equipment, the sampling method, themonitoring well construction, monitoring well opera-tion and maintenance, and sample-handling proce-dures should not alter the chemistry of the sample.

An example of how a site’s Data Quality Objectives(DQOs) for a characterization sampling effort mightvary from those of a remediation monitoring samplingeffort could be a difference of how much of thescreened interval or aquifer should be sampled. A sitecharacterization objective may be to collect a samplethat represents a composite of the entire (or as closeas is possible) screened interval of the monitoringwell.

Additionally, the site characterization may require alarge suite of contaminants to be sampled and ana-lyzed, whereas, the remediation monitoring programmay require fewer contaminants sampled and ana-lyzed. These differences may dictate the type ofsampling equipment used, the type of informationcollected, and the sampling protocol.

This sampling method described is for monitoringwells. However, this method should not be used forwater-supply wells with a water-supply pump, withlong-screened wells in complex hydrogeologic envi-ronments (such as fractured rock), or wells withseparate phases of liquids (such as a Dense or LightNon-Aqueous Phase Liquids) present within thescreened interval.

EQUIPMENT! Depth-to-water measuring device - An electronic

water-level indicator or steel tape and chalk, withmarked intervals of 0.01 foot. Interface probe formeasuring separate phase liquids, if needed.Pressure transducer and data logger optional forfrequent depth-to-water measuring in same well.

! Steel tape and weight - Used for measuringtotal depth of well. Lead weights should not beused.

! Sampling pump - Submersible or bladder pumpswith adjustable rate controls are preferred. Pumps

Standard Operating Procedure for the Well-VolumeMethod for Collecting a Ground-Water Sample

42

are to be constructed of inert materials, such asstainless steel and Teflon®. Pump types that areacceptable include gear and helical driven,centrifugal (low-flow type), and air-activated piston.Adjustable rate, peristaltic pumps can be usedwhen the depth to water is 20 feet or less.

! Tubing - Inert tubing should be chosen based onthe types and concentrations of contaminantspresent, or expected to be present in themonitoring well. Generally, Teflon®-based tubing isrecommended when sampling for organiccompounds. Polyethylene or Teflon® tubing can beused when sampling for inorganic constituents.

! Power source - If a combustion type (gasoline ordiesel-driven) device is used, it must be locateddownwind of the point of sample collection. Ifpossible, it should also be transported to the siteand sampling location in a different vehicle fromthe sampling equipment.

! Flow-measurement equipment - Graduatedcylinder or bucket and a stop watch, or a flowmeter that can be disconnected prior to sampling.

! Multi-parameter meter with flow-through cell - Thiscan be one instrument or multiple probes/instru-ments contained in a flow-through cell. The water-quality-indicator parameters that are measured inthe field are pH, oxidation/reduction potential (ORP,redox, or Eh), dissolved oxygen (DO), turbidity,specific electrical conductance (SEC), andtemperature. Calibration standards for allinstruments should be NIST-traceable, withinexpiration dates of the solutions, and sufficient fordaily calibration throughout the sampling collection.

! Decontamination supplies - A reliable anddocumented source of distilled water and anysolvents (if used). Pressure sprayers, buckets ordecontamination tubes for pumps, brushes andnon-phosphate soap also will be needed.

! Sample bottles, sample preservation supplies andlaboratory paperwork. Also, several coolers, andsample packing supplies (absorbing packingmaterial, plastic baggies, etc.).

! Approved plans and background documents -Approved Field Sampling Plan, Quality AssuranceProject Plan, well construction data, field andwater-quality data from the previous samplingcollection.

! Site Access/Permission documentation for siteentry.

! Well keys and map showing locations of wells.! Field notebook, field data sheets and calculator. A

suggested field data sheet is provided at the end ofthis attachment.

! Filtration equipment - If needed, this equipmentshould be an in-line disposable filter used for thecollection of samples for analysis of dissolvedconstituents.

! Polyethylene sheeting - Used for decontaminationstations and during sampling to keep equipmentclean.

! Site Health and Safety Plan and requiredequipment - The health and safety plan along withsite sign-in sheet should be on site and bepresented by the site health and safety officer.Personnel-protective and air-monitoring equipmentspecified in the Site Health and Safety Plan shouldbe demonstrated, present and in good workingorder on site at all times.

! Tool box - All needed tools for all site equipmentused.

! A 55-gallon drum or container to contain thepurged water.

Construction materials of the sampling equipment(bladders, pump, bailers, tubing, etc.) should belimited to stainless steel, Teflon®, glass, and otherinert materials when concentrations of the site con-taminants are expected within the detection limitrange. The sample tubing thickness and diametershould be maximized and the tubing length should beminimized so that the loss of contaminants absorbedto and through the tubing walls may be reduced andthe rate of stabilization of ground-water parameters ismaximized. The tendency of organics to sorb into andout of many materials makes the appropriateselection of sample tubing materials critical for thesetrace analyses (Pohlmann and Alduino, 1992; Parkerand Ranney, 1998).

Generally, wells should be purged and sampled usingthe same positive-displacement pump and/or a low-flow submersible pump with variable controlled flowrates and constructed of chemically inert materials. Ifa pump cannot be used because the recovery rate ofthe well is so low (less than 100 to 200 ml/min) andthe volume of the water to be removed is minimal(less than 5 feet of water in a small-diameter well),then a Teflon® bailer, with a double check valve andbottom-emptying device with a control-flow check

43

valve may be used to obtain the samples. Otherwise,a bailer should not be used when sampling for volatileorganics because of the potential bias introducedduring sampling (Yeskis et al., 1988; Pohlmann et al.,1990; Tai et al., 1991). Bailers also should be avoidedwhen sampling for metals because repeated bailerdeployment has the potential to increase turbidity,which biases concentrations of inorganic constituents.Dedicated sampling pumps are recommended formetals sampling (Puls et al., 1992).

In addition, for wells with long riser pipes above thewell screen, the purge volumes may be reduced byusing packers above the pumps. The packer materi-als should be compatible with the parameters to beanalyzed. These packers should be used only onwells screened in highly permeable materials, be-cause of the lack of ability to monitor water levels inthe packed interval. Otherwise, if pumping ratesexceed the natural aquifer recovery rates into thepacked zone, a vacuum or negative pressure zonemay develop. This may result in a failure of the sealby the packer and/or a gaseous phase may develop,that may bias any sample taken.

PURGING AND SAMPLING PROCEDUREWATER-LEVEL MEASUREMENTSThe field measurements should include total welldepth and depth to water from a permanently markedreference point.

TOTAL WELL DEPTHThe depth of each well should be measured to thenearest one-tenth of a foot when using a steel tapewith a weight attached and should be properly re-corded. The steel tape should be decontaminatedbefore use in another well according to the site spe-cific protocols. A concern is that when the steel tapeand weight hit the bottom of the well, sedimentpresent on the bottom of a well is stirred up, thusincreasing turbidity, which will affect the samplingresults. In these cases, as much time as possibleshould be allowed prior to sampling, such as a mini-mum of 24 hours. If possible, total well depth mea-surements can be completed after sampling (Puls andBarcelona, 1996). The weight of electric tapes isgenerally too light to determine accurate total welldepth. If the total well depth is greater than 200 feet,stretching of the tape must be taken intoconsideration.

DEPTH TO WATERAll water levels should be measured from thereference point by use of a weighted steel tape andchalk or an electronic water-level indicator (a detaileddiscussion of the pros and cons of the different waterlevel devices is provided in Thornhill, 1989). The steeltape is a more accurate method to take water levels,and is recommended where shallow flow gradients(less than 0.05 feet/feet) or deep wells areencountered. However, in those cases where largeflow gradients or large fluctuations in water levels areexpected, a calibrated electric tape is acceptable. Thewater level is calculated using the well’s surveyedreference point minus the measured depth-to-waterand should be measured to the nearest onehundredth of a foot.

The depth-to-water measurement must be made ineach well to be sampled prior to any other activities atthe well (such as bailing, pumping, and hydraulictesting) to avoid bias to the measurement. Allreadings are to be recorded to the nearest onehundredth of a foot. When possible, depth-to-waterand total well depth measurements should becompleted at the beginning of a ground-watersampling program, which will allow any turbidity tosettle and allow a more synoptic water-levelevaluation. However, if outside influences (such astidal cycles, nearby pumping effects, or majorbarometric changes) may result in significant water-level changes in the time between measurement andsampling, a water-level measurement should becompleted immediately prior to sampling. In addition,the depth-to-water measurement during purgingshould be recorded, with the use of a pressuretransducer and data logger sometimes more efficient(Barcelona et al., 1985, Wilde et al., 1998).

The time and date of the measurement, point ofreference, measurement method, depth-to-watermeasurement, and any calculations should beproperly recorded in field notebook or sampling sheet.

STATIC WATER VOLUMEFrom the information obtained for casing diameter,total well depth and depth-to-water measurements,the volume of water in the well is calculated. Thisvalue is one criteria that may be used to determine thevolume of water to be purged from the well before thesample is collected.

44

The static water volume may be calculated using thefollowing formula:

V = r2h(0.163)Where:

V = static volume of water in well(in gallons)

r = inner radius of well casing(in inches)

h = length of water column (in feet)which is equal to the total welldepth minus depth to water.

0.163 = a constant conversion factorthat compensates for theconversion of the casing radiusfrom inches to feet for 2-inchdiameter wells and the conver-sion of cubic feet to gallons,and pi (π). This factor wouldchange for different diameterwells.

Static water volumes also may be obtained fromvarious sources, such as Appendix 11.L in Driscoll(1986).

WELL PURGINGPURGE VOLUMESIn most cases, the standing water in the well casingcan be of a different chemical composition than thatcontained in the aquifer to be sampled. Solutes maybe adsorbed or desorbed from the casing material,oxidation may occur, and biological activity is pos-sible. Therefore, the stagnant water within the wellmust be purged so that water that is representative ofthe aquifer may enter the well.

The removal of at least three well volumes is sug-gested (USEPA, 1986; Wilde et al., 1998). Theamount of water removed may be determined bycollecting it in a graduated pail of known volume todetermine pumping rate and time of pumping. A flowmeter may also be used, as well as capturing allpurged water in a container of known volume.

The actual number of well volumes to be removed isbased on the stabilization of water-quality-indicatorparameters of pH, ORP, SEC, DO, and turbidity. The

water initially pumped is commonly turbid. In order tokeep the turbidity and other probes from being cloggedwith the sediment from the turbid water, the flow-through cell should be bypassed initially for the firstwell volume. These measurements should be takenand recorded every ½ well volume after the removal of1 to 1 ½ well volume(s). Once three successivereadings of the water-quality-indicator parametersprovided in the table have stabilized, sampling maybegin. The water-quality-indicator parameters that arerecommended include pH and temperature, but theseare generally insensitive to indicate completion ofpurging since they tend to stabilize rapidly (Puls andBarcelona, 1996). ORP may not always be an appro-priate stabilization parameter, and will depend on site-specific conditions. However, readings should berecorded because of its value as a double check foroxidizing conditions, and for some fate and transportissues. When possible, especially when sampling forcontaminants that may be biased by the presence ofturbidity, the turbidity reading is desired to stabilize at avalue below 10 Nephelometric Turbidity Units (NTUs).For final DO measurements, if the readings are lessthan 1 milligram per liter, they should be collected withthe spectrophotometric method (Wilde et al., 1998,Wilkin et al., 2001), colorimetric or Winkler titration(Wilkin et al., 2001). All of these water-quality-indicatorparameters should be evaluated against the specifica-tions of the accuracy and resolution of the instrumentsused. No more than six well volumes should bepurged, to minimize the over pumping effects de-scribed by Gibs and Imbrigiotta (1990).

Purging MethodsIn a well that is not being pumped, there will be littleor no vertical mixing in the water column betweensampling events, and stratification may occur. Thewater in the screened section may mix with theground water due to normal flow patterns, but thewater above the screened section will remain isolatedand become stagnant. Persons sampling shouldrealize that stagnant water may contain foreign mate-rial inadvertently or deliberately introduced from thesurface, resulting in unrepresentative water quality. Tosafeguard against collecting nonrepresentative stag-nant water in a sample, the following guidelines andtechniques should be adhered to during samplecollection:

45

1. As a general rule, monitoring wells should bepumped or bailed (although bailing is to be stronglyavoided) prior to collecting a sample. Evacuation of aminimum of three volumes of water in the well casingis recommended for a representative sample. In ahigh-yielding ground-water formation where there isno stagnant water in the well above the screenedsection (commonly referred to as a water-table well),evacuation prior to sample withdrawal is not as criticalbut serves to field rinse and condition samplingequipment. The purge criteria has been describedpreviously and will be again in the SAMPLING PRO-CEDURES section on the following page. The rate ofpurging should be at a rate and by a method that doesnot cause aeration of the water column and shouldnot exceed the rate at which well development wascompleted.

2. For wells that can be pumped or bailed to drynesswith the sampling equipment being used, the wellshould be evacuated to just above the well screeninterval and allowed to recover prior to sample with-drawal. (Note: It is important not to completely de-water the zone being sampled, as this may allow airinto that zone which could result in negative bias inorganic and metal constituents.) If the recovery rate isfairly rapid and time allows, evacuation of more thanone volume of water is preferred.

3. A non-representative sample also can result fromexcessive prepumping of the monitoring well. Stratifi-cation of the contaminant concentrations in theground-water formation may occur or heavier-than-water compounds may sink to the lower portions of

the aquifer. Excessive pumping can decrease orincrease the contaminant concentrations from what isrepresentative of the sampling point of interest, aswell as increase turbidity and create large quantitiesof waste water.

The method used to purge a well depends on theinner diameter, depth-to-water level, volume of waterin the well, recovery rate of the aquifer, and accessi-bility of the well to be sampled. The types of equip-ment available for well evacuation include hand-operated or motor-driven suction pumps, peristalticpumps, submersible pumps, and bailers made ofvarious materials, such as stainless steel andTeflon®. Whenever possible, the same device usedfor purging the well should be left in the well and usedfor sampling, generally in a continual manner frompurging directly to sampling without altering positionof the sampling device or turning off the device.

When purging/sampling equipment must be reused inother wells, it should be decontaminated consistentwith the decontamination procedures outlined in thisdocument. Purged water should be collected andscreened with air-monitoring equipment as outlined inthe site health and safety plan, as well as water-quality field instruments. If these parameters and/orthe facility background data suggest that the water ishazardous, it should be contained and disposed ofproperly as determined on a site-specific basis.

During purging, water-level measurements should berecorded regularly for shallow wells, typically at 15- to30-second intervals. These data may be useful in

dissolved oxygen (DO) +/- 0.3 milligrams per liter Wilde et al., 1998

Table of Stabilization Criteria with References for Water-Quality-Indicator Parameters Parameter Stabilization Criteria Reference

pH +/- 0.1 Puls and Barcelona, 1996;

Wilde et al., 1998

turbidity +/- 10% (when turbidity is Puls and Barcelona, 1996;

greater than 10 NTUs) Wilde et al., 1998

specific electrical +/- 3% Puls and Barcelona, 1996

conductance (SEC)oxidation-reduction +/- 10 millivolts Puls and Barcelona, 1996

potential (ORP)

46

computing aquifer transmissivity and other hydrauliccharacteristics, and for adjusting purging rates. Inaddition, these data will assure that the water leveldoesn’t fall below the pump intake level

SAMPLING PROCEDURESGround-water sample collection should take placeimmediately following well purging. Preferably, thesame device should be used for sample collection aswas used for well purging, minimize further distur-bance of the water column, and reduce volatilizationand turbidity. In addition, this will save time and avoidpossible contamination from the introduction of addi-tional equipment into the well, as well as using equip-ment materials already equilibrated to the groundwater. Sampling should occur in a progression fromthe least to most contaminated well, if known, whenthe same sampling device is used.

The sampling procedure is as follows:

1) Remove locking well cap. Note location, timeof day, and date in field notebook or on anappropriate log form.

2) Note wind direction. Stand upwind from thewell to avoid contact with gases/vapors ema-nating from the well.

3) Remove well casing cap.4) If required by site-specific conditions, monitor

headspace of well with appropriate air-moni-toring equipment to determine presence ofvolatile organic compounds or other com-pounds of concern and record in field logbook.

5) If not already completed, measure the waterlevel from the reference measuring point onthe well casing or protective outer casing (ifinner casing not installed or inaccessible) andrecord it in the field notebook. Alternatively, if noreference point exists, note that the water levelmeasurement is from the top of the outerprotective casing, top of inside riser pipe,ground surface, or some other position on thewell head. Have a permanent reference pointestablished as soon as possible after sam-pling. Measure at least twice to confirm mea-surement; the measurement should agreewithin 0.01 feet or re-measure. Decontaminatethe water-level-measuring device.

6) If not already completed, measure the totaldepth of the well (at least twice to confirmmeasurement; the measurement should agreewithin 0.01 feet or re-measure) and record it inthe field notebook or on log form. Decontami-nate the device used to measure total depth. Ifthe total well depth has been measured re-cently (in the past year), then measure it at theconclusion of sampling.

7) Calculate the volume of water in the well andthe volume to be purged using the formulapreviously provided.

8) Lay plastic sheeting around the well to mini-mize the likelihood of contamination of equip-ment from soil adjacent to the well.

9) Rinse the outside of sampling pump withdistilled water and then, while lowering thepump, dry it with disposable paper towels.

10) Lower the pump (or bailer) and tubing downthe well. The sampling equipment shouldnever be dropped into the well because thiswill cause degassing of the water upon impact.This may also increase turbidity, which maybias the metals analysis. The lowering of theequipment should be slow and smooth!

11) The pump should be lowered to a point justbelow the water level. If the water level isabove the screened interval, the pump shouldbe above the screened interval for the reasonsprovided in the purging section.

12) Turn the pump on. The submersible pumpsshould be operated in a continuous, low-flowmanner so that they do not produce pulsatingflows, which cause aeration in the dischargetubing, aeration upon discharge, orresuspension of sediments at the bottom ofthe well. The sampling pump flow rates shouldbe lower than or the same as the purgingrates. The purging and sampling rates shouldnot be any greater than well developmentrates.

13) Water levels should be monitored duringpumping to ensure that air does not enter thepump and to help determine an appropriatepurging rate.

14) After approximately one to two well volumesare removed, a flow-through cell will be hookedup to the discharge tubing of the pump. If the

47

well discharge water is not expected to behighly turbid, contain separate liquid phases, orminimal bacterial activitiy that may coat or clogthe electrodes within the flow-through cell, thenthe cell can be immediately hooked up to thedischarge tubing. This cell will allow measure-ments of water-quality-indicator parameterswithout allowing contact with the atmosphereprior to recording the readings for temperature,pH, ORP, SEC, DO and turbidity.

15) Measurements for temperature, pH, ORP,SEC, DO, and turbidity will be made at eachone-half well volume removed. Purging maycease when measurements for all five param-eters have stabilized (provided in the earliertable) for three consecutive readings.

16) If the water level is lowered to the pump levelbefore three volumes have been removed, thewater level will be allowed to recover for 15minutes, and then pumping can begin at alower flow rate. If the pump again lowers thewater level to below the pump intake, thepump will be turned off and the water levelallowed to recover for a longer period of time.This will continue until a minimum of two wellvolumes are removed prior to taking theground-water sample.

17) If the water-quality-indicator parameters havestabilized, sample the well. Samples will becollected by lowering the flow rate to a ratethat minimizes aeration of the sample whilefilling the bottles (approximately 300 ml/min).Then a final set of water-quality-indicatorparameters is recorded. The pump dischargeline is rapidly disconnected from the flow-through cell to allow filling of bottles from thepump discharge line. The bottles should befilled in the order of volatile organic com-pounds bottles first, followed by semi-volatileorganic compound’s/pesticides, inorganics,and other unfiltered samples. Once the last setof samples is taken, if filtering is necessary, anin-line disposable filter (with appropriatelychosen filter size) will be added to the dis-charge hose of the pump. Then the filteredsamples will be taken. If a bailer is used forobtaining the samples, filtering occurs at thesampling location immediately after the sampleis obtained from the bailer by using a suction

filter. The first one-half to one liter of sampletaken through the filter will not be collected, inorder to assure the filter media is acclimated tothe sample. If filtered samples are collected,WITHOUT EXCEPTION, filtering should beperformed in the field as soon as possible aftercollection, and not later in a laboratory.

18) All appropriate samples that are to be cooled,are put into a cooler with ice immediately. All ofthe samples should not be exposed to sunlightafter collection. Keep the samples from freez-ing in the winter when outside temperaturesare below freezing. The samples, especiallyorganics, cyanide, nutrients, and otheranalytes with short holding times, are recom-mended to be shipped or delivered to thelaboratory daily. Ensure that the appropriatesamples that are to be cooled remain at 4oC,but do not allow any of the samples to freeze.

19) If a pump cannot be used because the recov-ery rate is slow and the volume of the water tobe removed is minimal (less than 5 feet ofwater), then a Teflon® bailer, with a doublecheck valve and bottom-emptying device witha control-flow check valve will be used toobtain the samples. The polypropylene ropeused with the bailer will be disposed of follow-ing the completion of sampling at each well.

20) The pump is removed from the well anddecontaminated for the next sampling location.

Additional precautions to ensure accurate and repre-sentative sample collection are as follows:

! Check valves on bailers, if bailers are used, shouldbe designed and inspected to ensure that foulingproblems do not reduce delivery capabilities orresult in aeration of the sample.

! The water should be transferred to a samplecontainer in a way that will minimize agitation andaeration.

! If the sample bottle contains no preservatives, thebottle should be rinsed with sample water, which isdiscarded before sampling. Bottles for sampleanalyses that require preservation should beprepared before they are taken to the well. Careshould be taken to avoid overfilling bottles so thatthe preservative is not lost. The pH should bechecked and more preservatives added to inor-

48

ganic sample bottles, if needed. VOA bottles thatdo not meet the ph requirements need to bediscarded and new sample bottles with morepreservative added should be prepared immedi-ately.

! Clean sampling equipment should not be placeddirectly on the ground or other contaminatedsurfaces either prior to sampling or during storageand transport.

Special Consideration for Volatile Organic CompoundSamplingThe proper collection of a sample for dissolved volatileorganics requires minimal disturbance of the sampleto limit volatilization and therefore a loss of volatilesfrom the samples. Preferred retrieval systems for thecollection of un-biased volatile organic samplesinclude positive displacement pumps, low-flow cen-trifugal pumps, and some in-situ sampling devices.Field conditions and other constraints will limit thechoice of appropriate systems. The principal objectiveis to provide a valid sample for analysis, one that hasbeen subjected to the least amount of turbulencepossible.

1) Fill each vial to just overflowing. Do not rinsethe vial, nor excessively overflow it, as this willeffect the pH by diluting the acid preservativepreviously placed in the bottle. Another optionis to add the acid at the well, after the samplehas been collected. There should be a convexmeniscus on the top of the vial.

2) Do not over tighten and break the cap.3) Invert the vial and tap gently. Observe the vial

closely. If an air bubble appears, discard thesample and collect another. It is imperativethat no entrapped air remains in the samplevial. Bottles with bubbles should be discarded,unless a new sample cannot be collected, andthen the presence of the bubble should benoted in the field notes or field data sheet. Ifan open sample bottle is dropped, the bottleshould be discarded.

4) Orient the VOC vial in the cooler so that it islying on its side, not straight up.

5) The holding time for VOCs is 14 days. It isrecommended that samples be shipped ordelivered to the laboratory daily. Ensure that

the samples remain at 4oC, but do not allowthe samples to freeze.

Field Filtration of Turbid SamplesThe USEPA recognizes that in some hydrogeologicenvironments, even with proper well design, installa-tion, and development, in combination with the low-flow rate purging and sampling techniques, sampleturbidity cannot be reduced to ambient levels. The wellconstruction, development, and sampling informationshould be reviewed by the Regional geologists orhydrologists to see if the source of the turbidity prob-lems can be resolved or if alternative sampling meth-ods should be employed. If the water sample isexcessively turbid, the collection of both filtered andunfiltered samples, in combination with turbidity, TotalSuspended Solids (TSS), Total Dissolved Solids(TDS), pumping rate, and drawdown data is recom-mended. The filter size used to determine TSS andTDS should be the same as used in the field filtration.An in-line filter should be used to minimize contactwith air to avoid precipitation of metals. The typicalfilter media size used is 0.45 µm because this iscommonly accepted as the demarcation betweendissolved and non-dissolved species. Other filtersizes may be appropriate, but their use should bedetermined based on site-specific criteria (examplesinclude grain-size distribution, ground-water flowvelocities, mineralogy) and project DQOs. Filter sizesup to 10.0 µm may be warranted because larger sizefilters may allow particulates that are mobile in groundwater to pass through (Puls and Powell, 1992). Thechanging of filter media size may limit the comparabil-ity of the data obtained with other data sets and mayaffect their use in some geochemical models. Filtermedia size used on previous data sets from a site,region, or aquifer and the DQOs should be taken intoconsideration. The filter media used during theground-water sampling program should be collected ina suitable container and archived because potentialanalysis of the media may be helpful for the determi-nation of particulate size, mineralogy, etc.

The first 500 to 1000 milliliters of sample takenthrough the filter, depending on sample turbidity, willnot be collected for a sample, in order to ensure thatthe filter media has equilibrated to the sample. Manu-facturers’ recommendations also should be consulted.Because bailers have been shown to increase

49

turbidity while purging and sampling, they should beavoided when sampling for trace element, metal,PCB, and pesticide constituents. If portable samplingpumps are used, the pumps should be gently loweredto the sampling depth desired, carefully avoiding beinglowered to the bottom of the well. The pumps, onceplaced in the well, should not be moved to allow anyparticles mobilized by pump placement to settle.Dedicated sampling equipment installed in the wellprior to the commencement of the sampling activitiesis one of the recommended methods to reduceturbidity artifacts (Puls and Powell, 1992; Kearl et al.,1992; Puls et al., 1992; Puls and Barcelona, 1996).

DECONTAMINATION PROCEDURESOnce removed from the well, the purging and sam-pling pumps should be decontaminated by scrubbingwith a brush and a non-phosphate soapy-water wash,rinsed with water, and rinsed with distilled water tohelp ensure that there is no cross-contaminationbetween wells. The step-by-step procedure is:

1) Pull pump out of previously sampled well (orout of vehicle) and use three pressure spray-ers filled with soapy water, tap water, anddistilled water. Spray outside of tubing andpump until water is flowing off of tubing aftereach rinse. Use bristle brush to help removevisible dirt, contaminants, etc.

2) Have three long PVC tubes with caps orbuckets filled with soapy water, tap water anddistilled water. Run pump in each until approxi-mately 2 to 3 gallons of each decon solution ispumped through tubing. Pump at low rate toincrease contact time between the deconsolutions and the tubing.

3) Try to pump decon solutions out of tubing priorto next well. If this cannot be done, com-pressed air may be used to purge lines.Another option is to install a check valve in thepump line (usually just above the pump head)so that the decon solutions do not run backdown the well as the pump is lowered downthe next well.

4) Prior to lowering the pump down the next well,spray the outside of the pump and tubing withdistilled water. Use disposable paper towels todry the pump and tubing.

5) If a hydrophobic contaminant is present (suchas separate phase, high levels of PCBs, etc.),an additional decon step, or steps, may beadded. For example, an organic solvent suchas reagent-grade isopropanol alcohol may beadded as a first rinse prior to the soapy waterrinse.

If the well has been sampled with a bailer that is notdisposable, the bailer should be cleaned by washingwith soapy water, rinsing with tap water, and finallyrinsing with distilled water. Bailers are most easilycleaned using a long-handled bottle brush.

It is especially important to clean thoroughly theportion of the equipment that will be in contact withsample water. In addition, a clean plastic sheet shouldbe placed adjacent to or around the well to preventsurface soils from coming in contact with the purgingequipment. The effects of cross-contamination alsocan be minimized by sampling the least contaminatedwell first and progressing to the more contaminatedones. The bailer cable/rope (if a bailer is used) andplastic sheet should be properly discarded, as pro-vided in the site health and safety plan, and newmaterials provided for the next well.

FIELD QUALITY CONTROLThe quality assurance (QA) targets for precision andaccuracy of sampling programs are based on accu-racy and precision guidelines established by theUSEPA. When setting targets, keep in mind that allmeasurements must be made so that the results arerepresentative of the sample water and site-specificconditions. Various types of blanks are used to checkthe cleanliness of the field-handling methods. Theseare known as field blanks, and include field equipmentblanks and transport blanks. Other QA samplesinclude spike samples and duplicates.

There are five primary areas of concern for QA in thecollection of representative ground-water samples:

1. Obtaining a sample that is representative ofwater in the aquifer or targeted zone of theaquifer. Verify log documentation that the wellwas purged of the required volume or that thetemperature, pH, ORP, SEC, DO and turbiditystabilized before samples were extracted.

50

2. Ensuring that the purging and sampling de-vices are made of materials and utilized in amanner that will not interact with or alter theanalyses.

3. Generating results that are reproducible.Therefore, the sampling scheme shouldincorporate co-located samples (duplicates).

4. Preventing cross-contamination. Samplingshould proceed from least to most contami-nated wells, if known. Field equipment blanksshould be incorporated for all sampling andpurging equipment; decontamination of theequipment is therefore required.

5. Ensuring that samples are properly preserved,packaged, and shipped.

FIELD EQUIPMENT BLANKSTo ensure QA and quality control, a field equipmentblank must be included in each sampling run, or forevery twenty samples taken with the sampling device.Equiptment blanks allow for a cross check and, insome cases, quantitative correction for imprecisionthat could arise due to handling, preservation, orimproper cleaning procedures.

Equipment blanks should be taken for each samplebottle type that is filled. Distilled water is run throughthe sampling equipment and placed in a sample bottle(the blank), and the contents are analyzed in the lablike any other sample. Following the collection of eachset of twenty samples, a field equipment blank will beobtained. It is generally desirable to collect this fieldequipment blank after sampling a relatively highlycontaminated well. These blanks may be obtainedthrough the following procedure:

a) Following the sampling event, decontaminateall sampling equipment according to the sitedecontamination procedures and beforecollecting the blank.

b) VOA field blanks should be collected first, priorto water collected for other TAL/TCL analyses.A field blank must be taken for all analyses.

c) Be sure that there is enough distilled water inthe pump so that the field equipment blank canbe collected for each analysis.

d) The water used for the field equipment blankshould be from a reliable source, documented

in the field notebooks, and analyzed as aseparate water-quality sample.

TRIP BLANKSA trip blank should be included in each sample ship-ment and, at a minimum, one per 20 samples. Bottles,identical to those used in the field, are filled withreagent-grade water. The source of the reagent-gradewater should be documented in the field notebooks,including lot number and manufacture. This sample islabeled and stored as though it is a sample. Thesample is shipped back to the laboratory with the othersamples and analysis is carried out for all the sameconstituents.

DUPLICATE SAMPLESDuplicate samples are collected by taking separatesamples as close to each other in time and space aspractical, and should be taken for every 20 samplescollected. Duplicate samples are used to developcriteria for acceptable variations in the physical andchemical composition of samples that could resultfrom the sampling procedure. Duplicate results areutilized by the QA officer and the project manager togive an indication of the precision of the sampling andanalytical methods.

HEALTH AND SAFETY CONSIDERATIONSDepending on the site-specific contaminants, variousprotective programs must be implemented prior tosampling the first well. The site health and safety planshould be reviewed with specific emphasis placed onthe protection program planned for the samplingtasks. Standard safe operating practices should befollowed, such as minimizing contact with potentialcontaminants in both the liquid and vapor phasesthrough the use of appropriate personal protectiveequipment.

Depending on the type of contaminant expected ordetermined in previous sampling efforts, the followingsafe work practices will be employed:

Particulate or metals contaminants1. Avoid skin contact with, and accidental inges-

tion of, purge water.2. Wear protective gloves and splash protection.

51

Volatile organic contaminants1. Avoid breathing constituents venting from well.2. Pre-survey the well head space with an appro-

priate device as specified in the Site Healthand Safety Plan.

3. If air monitoring results indicate elevatedorganic constituents, sampling activities maybe conducted in Level C protection. At aminimum, skin protection will be afforded bydisposable protective clothing, such asTyvek®.

General practices should include avoiding skin con-tact with water from preserved sample bottles, as thiswater will have pH less than 2 or greater than 10.Also, when filling, pre-preserved VOA bottles, hydro-chloric acid fumes may be released and should not beinhaled.

POST-SAMPLING ACTIVITIESSeveral activities need to be completed and docu-mented once ground-water sampling has been com-pleted. These activities include, but are not limited to:

! Ensuring that all field equipment has been decon-taminated and returned to proper storage location.Once the individual field equipment has beendecontaminated, tag it with date of cleaning, sitename, and name of individual responsible.

! Processing all sample paperwork, including copiesprovided to Central Regional Laboratory, SampleManagement Office, or other appropriate samplehandling and tracking facility.

! Compiling all field data for site records.! Verifying all analytical data processed by the

analytical laboratory against field sheets to ensureall data has been returned to sampler.

REFERENCESBarcelona, M.J., J.P. Gibb, J.A. Hellfrich and E.E.Garske, 1985, Practical Guide for Ground-WaterSampling; U.S. Environmental Protection Agency,EPA/600/2-85/104, 169 pp.

Driscoll, F.G., 1986, Groundwater and Wells, 2nd Ed.;Johnson Division, St. Paul, Minnesota, 1089 pp.

Gibs, J. and T.E. Imbrigiotta, 1990, Well-PurgingCriteria for Sampling Purgeable Organic Compounds;Ground Water, Vol. 28, No. 1, pp 68-78.

Herzog, B.L., S.J. Chou, J.R. Valkenburg and R.A.Griffin, 1988, Changes in Volatile Organic ChemicalConcentrations After Purging Slowly RecoveringWells; Ground Water Monitoring Review, Vol. 8, No. 4,pp. 93-99.

Kearl, P.M., N.E. Korte, and T.A. Cronk, 1992, Sug-gested Modifications to Ground Water SamplingProcedures Based on Observations from the ColloidBorescope; Ground Water Monitoring Review, Vol. 12,No. 2, pp. 155-161.

Keely, J.F. and K. Boateng, 1987, Monitoring WellInstallation, Purging, and Sampling Techniques - Part1: Conceptualizations; Ground Water, Vol. 25, No. 4pp. 427-439.

McAlary, T.A. and J.F. Barker, 1987, VolatilizationLosses of Organics During Ground Water Samplingfrom Low Permeability Materials; Ground WaterMonitoring Review, Vol. 7, No. 4, pp. 63-68.

Nielson, D.M., 1991, Practical Handbook of Ground-Water Monitoring; Lewis Publishers, 717 pp.


Pohlmann, K.F., R.P. Blegen, and J.W. Hess, 1990,Field Comparison of Ground-Water Sampling Devicesfor Hazardous Waste Sites: An Evaluation usingVolatile Organic Compounds; EPA/600/4-90/028,102 pp.

Pohlmann, K.F. and A.J. Alduino, 1992, GROUND-WATER ISSUE PAPER: Potential Sources of Error inGround-Water Sampling at Hazardous Waste Sites;US Environmental Protection Agency. EPA/540/S-92/019.

Puls, R.W. and R.M. Powell, 1992, Acquisition ofRepresentative Ground Water Quality Samples forMetals; Ground Water Monitoring Review, Vol. 12, No.3, pp. 167-176.

52

Puls, R.W., D.A. Clark, B. Bledsoe, R.M. Powell andC.J. Paul, 1992, Metals in Ground Water: SamplingArtifacts and Reproducibility; Hazardous Waste andHazardous Materials, Vol. 9, No. 2, pp. 149-162.

Puls, R.W. and M.J. Barcelona, 1996, GROUND-WATER ISSUE PAPER: Low-Flow (Minimal Draw-down) Ground-Water Sampling Procedures; U.S.Environmental Protection Agency, EPA/540/S-95/504,12 pp.

Tai, D.Y., K.S. Turner, and L.A. Garcia, 1991, The Useof a Standpipe to Evaluate Ground Water Samples;Ground Water Monitoring Review, Vol. 11, No. 1, pp.125-132.

Thornhill, J.T., 1989, SUPERFUND GROUND WATERISSUE: Accuracy of Depth to Water Measurements;US Environmental Protection Agency. EPA/540/4-89/002, 3 pp.

U.S. Environmental Protection Agency, 1986, RCRAGround-Water Monitoring Technical EnforcementGuidance Document; OSWER-9950.1, U.S. Govern-ment Printing Office, Washington, D.C., 208 pp.,appendices.

U.S. Environmental Protection Agency, 1995, GroundWater Sampling-A Workshop Summary, Dallas,Texas, November 30-December 2, 1993, EPA/600/R-94/025, 146 pp.

Wilde, F.D., D.B. Radtke, J.Gibs and R.T. Iwatsubo,eds., 1998, National Field Manual for the Collection ofWater-Quality Data; U.S. Geological SurveyTechniques of Water-Resources Investigations, Book9, Handbooks for Water-Resources Investigations,variously paginated.

Wilkin, R.T., M.S. McNeil, C.J. Adair and J.T. Wilson,2001, Field Measurement of Dissolved Oxygen: AComparison of Methods, Ground Water Monitoringand Remediation, Vol. 21, No. 4, pp. 124-132.

Yeskis, D., K. Chiu, S. Meyers, J. Weiss and T. Bloom,1988, A Field Study of Various Sampling Devices andTheir Effects on Volatile Organic Contaminants;Proceedings of the Second National Outdoor ActionConference on Aquifer Restoration, Ground WaterMonitoring and Geophysical Methods, National WaterWell Association, May, 1988.

53

GROUND-WATER SAMPLING RECORD Well ID:_______________

Station #:______________Facility Name: Date:____/____/____

Well Depth:__________ Depth to Water:__________ Well Diameter:___________

Casing Material.:__________ Volume Of Water per Well Volume:______________

Sampling Crew:__________________,____________________,___________________,______________________

Type of Pump:_________ ___________ Tubing Material:__________________ Pump set at _________________ ft.

Weather Conditions:_________________________________ NOTES:_________ ________________________

______________________________________________________________________________________________

Other Parameters: ___________________Sampled at:_______________ Parameters taken with :_________________________________________Sample delivered to ______________________________ by ____________________________ at___________.Sample CRL #:______________ OTR #:______________ ITR #:______________ SAS #:__________________

Parameters Collected Number of Bottles Bottle Lot Number

_____________________________________ _________ _______________

_____________________________________ _________ _______________

_____________________________________ _________ _______________

_____________________________________ _________ _______________

_____________________________________ _________ _______________

Temp.(0C)Time

WaterLevel

VolumePumped

PumpingRate

DO(mg/l)

SEC(µS/cm) pH

ORP(mV)

Turbidity(NTU)

GROUND-WATER SAMPLING PARAMETERS

Date post:	29-Dec-2015
Category:	Documents
Upload:	subhradeb-pramanik
View:	16 times
Download:	3 times

Www.epa.Gov Superfund Remedytech Tsp Download Gw Sampling Guide

Documents