Guidance for Collecting Discharge-Weighted Samples in Surface Water Using an Isokinetic Sampler. For the entire report see sw99.01.pdfDate: Tue, 03 Nov 1998 08:55:39 -0500 From: Nana FryeTo: "A - Division Chief and Staff" , "B - Branch Chiefs and Offices" , "DC - All District Chiefs" , "S - Special Distribution for Research" , "FO - State, District, Subdistrict and other Field Offices" , "PO - Project Offices" , wqspecs@usgs.gov, owq@usgs.gov Cc: " WRD Archive File, " , Nana Snow Subject: OWQ 99.02/OSW 99.01--Guidance for Collecting Discharge-Weighted Samples in Surface Water Using an Isokinetic Sampler In Reply Refer To: October 28, 1998 Mail Stop 412 or Mail Stop 415 OFFICE OF WATER QUALITY TECHNICAL MEMORANDUM 99.02 OFFICE OF SURFACE WATER TECHNICAL MEMORANDUM 99.01 Subject: Guidance for Collecting Discharge-Weighted Samples in Surface Water Using an Isokinetic Sampler PURPOSE AND SCOPE The purpose of this memorandum is to provide guidance for collecting discharge-weighted, depth-integrated samples in surface water using isokinetic samplers. Tables 4-15 and 17-24 in Appendix 4 quantify acceptable ranges of reeling and transit rates for rigid-bottle and bag isokinetic samplers when used with standard reels. This memorandum also reviews common terminology to provide a better understanding of surface-water sampling procedures. This memorandum does not provide guidance on other sampling techniques such as point sampling, area-weighted sampling, or non-isokinetic sampling. The techniques may be useful and/or desirable depending on the sampling design and objectives of a project. BACKGROUND Under most field conditions, isokinetic, depth-integrated sampling techniques must be used to collect discharge-weighted samples. Constituent concentrations determined from discharge-weighted samples are used to compute the discharge of any constituent. The discharge of any constituent is the product of the stream discharge and the discharge-weighted concentration of the constituent. The Office of Surface Water and the Office of Water Quality recognize that the uses and limitations of depth-integrating samplers are not well documented. Consequently, samples that must be collected using discharge-weighted, depth-integrated, isokinetic sampling techniques are sometimes being collected under depth and velocity conditions that are outside the range where isokinetic samples are obtainable with available samplers. The following information is presented to better define and document the operational ranges of the most commonly used depth-integrating samplers used by the U.S. Geological Survey. Similar information is presented in Chapters A2, A4, and A6 of Techniques of Water-Resources Investigations book 9, "National Field Manual for the Collection of Water-Quality Data," by Wilde and others, eds., in press. GUIDANCE FOR COLLECTING DISCHARGE-WEIGHTED SAMPLES IN SURFACE WATER USING AN ISOKINETIC SAMPLER Many factors can affect whether the concentration of a constituent (property) in a discharge-weighted sample adequately represents the discharge-weighted concentration of that constituent in the stream at the time of sampling. This memo primarily discusses the inherent physical limitations of commonly used depth-integrating samplers. These samplers collect isokinetic samples under a relatively narrow set of conditions that need to be understood by those collecting the sample. The operational ranges of commonly used samplers are presented in tables 4 through 15 and 17 through 24 in Appendix 4. If water samples are obtained within these operational ranges, the sample can be reasonably assumed to be representative of the stream at the time of sampling. Equal-Discharge-Increment and Equal-Width-Increment Sampling Methods: Uses and Limitations Isokinetic sampling is necessary to discharge-weight (velocity-weight) samples and to accurately collect the sand fraction of suspended sediment. Equal-discharge-increment (EDI) and equal-width-increment (EWI) sample-collection methods are specifically designed to result in the collection of discharge-weighted, depth-integrated, isokinetic samples (Appendixes 2 and 3). If used correctly, and samples are taken within the limitations of the sampler used, both methods result in samples that have the same concentration of constituents. EDI is the most universally applicable discharge-weighted sampling method. This method can be used to collect a single composite sample or a series of samples representing each increment of discharge. The basic assumption that must be made for the EDI method to be properly used is that the concentration of any constituent collected at the centroid of the equal increment of discharge represents the mean concentraton in that entire increment of discharge. When using the EDI method and compositing the sample, the total composite sample volume can be estimated on-site before sampling begins because an approximately equal volume (at least the minimum volume shown for the deepest vertical) of water is collected at each increment of discharge. The total composite volume can be estimated by multiplying the volume collected at the deepest vertical by the number of increments of equal discharge sampled. When using the EDI method and not compositing, the samples at each vertical are analyzed separately. The volume collected at each vertical can be any volume from within the isokinetic range of the sampler for that vertical. The total constituent discharge is the sum of the products at the individual increment stream discharge and the constituent concentration from that increment. The EDI method can be used to collect discharge-weighted samples at water velocities less than about 1.5 feet per second in nonstratified streams. Although the samplers do not collect true isokinetic samples at flows less than about 1.5 feet per second, a lack of suspended sand makes it unnecessary to collect fully isokinetic samples under these conditions (Office of Water Quality Technical Memorandum 76.17, "Water Quality--Sampling Mixtures of Water and Suspended Sediment in Streams," May 12, 1976, states that a velocity of 2 feet per second is required to transport sand). The EWI method cannot be used under these low velocity conditions since this method assumes isokinetic sampling in each vertical, which is not possible at velocities less than about about 1.5 feet per second. The EWI method is broadly applicable to streams in which the cross section has a relatively uniform depth and water velocity. EWI is more limited in application than is the EDI method, primarily because of the requirement to use only one transit rate and because of sampler limitations. All EWI samples must be collected within the isokinetic range of the sampler because EWI samples are by definition discharge-weighted samples and the isokinetic collection ability of the sampler is used to discharge weight the sample. All EWI water-quality samples must be composited. The EWI method cannot be used if a significant number of verticals in the cross section require transit rates slower than the transit rate used at the deepest, fastest vertical because of the one-transit-rate requirement. Tables 7, 11, 15, 20, and 24 in Appendix 4 provide transit rates for a range of stream depths and velocities for several bottle and nozzle combinations. To determine if the slower velocity verticals can be sampled at the same transit rate as the faster velocity verticals, compare the slowest transit rate that will fill the bottle at the deepest (highest velocity) vertical with the maximum rate allowable at the slowest vertical. When using a bottle sampler, the full reeling or transit rate at the deepest, fastest vertical will usually exceed the fastest allowed rate at one or two verticals near the streambank. The difference in constituent concentration in a composite sample caused by this error may be insignificant because (a) the cumulative discharge associated with slow and shallow sections is usually negligible with respect to the total discharge, and (b) the sample volume collected isokinetically from these sections is negligible with respect to the total sample volume. Also, there may be compensating errors of excessive transit rates and oversampling in slow water velocities. Currently available bottle samplers generally are not designed to collect samples isokinetically at water velocities of less than about 1.5 feet per second. Currently available bag samplers generally do not collect samples isokinetically at water velocities of less than about 3 feet per second. Thus, the EWI method cannot be used at cross sections at which all, or large parts, of the sampling cross section have velocities of less than about 1.5 feet per second when using a bottle sampler, or less than 3 feet per second when using a bag sampler. Usable Range of Bottle Samplers Generally, bottle samplers (see Appendix 1, "Definitions") can collect isokinetic samples in streams up to 15 feet deep, at water velocities greater than about 1.5 feet per second, as long as the sampler does not fill above the outlet of the nozzle, or the transit rate does not exceed 0.4 times the mean stream velocity at the sampling vertical (see Appendix 3). Common errors observed in the use of a 3-liter bottle sampler include excessively fast transit rates and its use in streams that are too shallow. A clear indication that the transit rate is too fast is the absence of bubbles from the exhaust port when the sampler is lowered, or an insufficient volume of water in the sampler after a round-trip transit has been completed (see Appendix 4, tables 4 through 24. Usable Range of Bag Samplers Currently available bag samplers may collect samples isokinetically to any depth that the bag capacity is not exceeded by the minimum round-trip sample volume (Appendix 4, table 16) if (1) the temperature is greater than about 8 degrees Celsius, (2) the mean velocity at verticals is more than about 3 feet per second, and (3) the transit rate is less than 0.4 times the mean stream velocity at the sampling vertical (see Appendix 4, tables 17-24). Because several factors can affect the sampling efficiency of bag samplers, it is recommended that a field calibration of the bag samplers hydraulic efficiency be done on-site before each set of samples is collected. Sampling Streams Less Than 15 Feet Deep Container selection There is no substantial difference in the range of depths and velocities that can be sampled with different 1- and 3-liter sample bottle and nozzle combinations. However, transit rates can differ substantially for different 1- and 3-liter bottle and nozzle combinations. The 1-liter bottle sampler is the best choice for isokinetic sampling for water chemistry in streams less than 15 feet deep. The 1-liter sampler has a smaller unsampled zone and requires much smaller minimum volumes for each vertical than the 3-liter bottle sampler. The 3-liter bottle sampler has an unsampled zone of at least 7 inches and should not be used in streams less than about 2 to 3 feet deep when (1) the EWI method is being used or (2) sand is to be analyzed as part of the sample and the stream velocity is sufficient to transport sand. The 3-liter bottle sampler requires very slow transit rates in slow to moderate stream velocities. Currently available bag samplers can be used for depth-integrated, isokinetic water-quality sampling of streams less than 15 feet deep and provide a much wider isokinetic range in depth and velocity than do bottle samplers. Bag samplers require water temperatures above about 8 degrees Celsius, velocities greater than 3 feet per second, strict attention to transferring all the sand out of the bag, and clean sampling techniques when appropriate. The D-77 bag sampler can be used in streams as shallow as 2 to 3 feet deep. Frame-type bag and bottle samplers require deeper streams in order to minimize the effect of the unsampled zone. For deep, swift streams (greater than about 7 feet per second) a heavily weighted frame-type bag or bottle sampler would be a reasonable choice for water-quality sampling. Sampling Streams More Than 15 Feet Deep Point samplers, as described in Edwards and Glysson (1998), are the preferred samplers to collect isokinetic, depth-integrated samples in streams deeper than about 15 feet. Point samplers are known to contaminate trace-element samples and cannot be easily sterilized so that if samples are to be analyzed for trace elements or bacteria, bag samplers must be used. Nozzle selection For 1-liter bottles, the 5/16-inch nozzle for shallow depths and the 1/4-inch nozzle for greater depths provide adequate ranges in transit and reeling rates. For 3-liter bottles, even the 5/16-inch nozzle requires excessively slow transit rates at shallow depths. For a bottle sampler, larger nozzles provide greater range between the slowest and fastest isokinetic transit rates (in feet per second). Smaller nozzles provide a larger difference between the slowest and fastest, total round-trip transit time. These statements may seem counter intuitive but examination of the tables for bottle transit rates and reeling rates will clarify the statement. Smaller nozzles require slower transit rates. Nozzle size has little effect on minimum sample volumes. For a pint bottle, the 3/16-inch nozzle increases the isokinetic depth capabilities from 9 feet for the 1/4-inch nozzle to 15 feet for the 3/16-inch nozzle. No substantial increases in depth capabilities are provided by reducing the nozzle size for any other bottle. Nozzles 3/16 inch and larger are recommended for sampling suspended sediment. For bag samplers, smaller nozzles may be preferred because they provide isokinetic sampling in greater depth and velocity ranges and smaller minimum volumes than do larger nozzles. Smaller nozzles also provide a greater range between the slowest and fastest, total round-trip transit time. And, as opposed to bottle samplers, smaller nozzles also provide greater range between the slowest and fastest isokinetic transit rates. LOCATION AND DESCRIPTION OF OTHER INFORMATION In the public ftp depot on srv3rvares.er.usgs.gov/, the directory contains Excel workbook files that include tables 3 through 24 of Appendix 4 and additional workbook files for different bottle, bag, and nozzle sizes. The workbooks can be printed as is or can be modified to meet user needs. SELECTED REFERENCES Edwards, T.K., and Glysson, G.D., 1998, Field methods for measurement of fluvial sediment: U.S. Geological Survey Techniques of Water-Resources Investigations, book 3, chap. C2, 80 p. Federal Interagency Sedimentation Project, 1952, The design of improved types of suspended-sediment samplers--Interagency Report 6: Minneapolis, Minnesota, St. Anthony Falls Hydraulic Laboratory, 103 p. Wilde, F.D., Radtke, D.B., Gibs, Jacob, and Iwatsubo, R.T., eds., in press, Selection of equipment for water sampling, chapter A2 of National Field Manual for the Collection of Water-Quality Data: U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chap. A2. Wilde, F.D., Radtke, D.B., Gibs, Jacob, and Iwatsubo, R.T., eds., in press, Collection of water samples, chapter A4 of National Field Manual for the Collection of Water-Quality Data: U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chap. A4. Wilde, F.D., and Radtke, D.B., eds., in press, Field Measurements, chapter A6 of National Field Manual for the Collection of Water-Quality Data: U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chap. A6. Thomas H. Yorke, Jr. /s/ Janice R. Ward /s/ Chief Acting Chief Office of Surface Water Office of Water Quality 4 attachments Keywords: Isokinetic, EDI, EWI, sampler, depth-integrated sample, discharge-weighted sample, area-weighted sample, surface-water quality, transit rate, reeling rate, suspended sediment. Distribution: A, B, DC, S, FO, PO District Water-Quality Specialists Regional Water-Quality Specialists OWQ Staff Appendix 1. DEFINITIONS Isokinetic sampling: "To sample in such a way that the water-sediment mixture moves with no change in velocity as it leaves the ambient flow and enters the sampler intake." (ASTM) Discharge-weighted sample: A sample that contains an equal volume from each unit of discharge sampled. Depth-integrated sample: A sample that is collected so that each vertical portion of the stream depth is represented in the sample in proportion to the desired sampling scheme. Depth integration (for a discharge-weighted sample as defined by ASTM): "A method of sampling at every point throughout a given depth (the sampled depth) whereby the water-sediment mixture is collected isokinetically so that the contribution from each point is proportional to the stream velocity at the point. This process yields a sample with properties that are discharge weighted over the sampled depth." (ASTM) Depth integration to collect a discharge-weighted sample: "Depth-integrated sample--a discharge-weighted (velocity-weighted) sample of water-sediment mixture collected at one or more verticals in accordance with the technique of depth integration; the discharge of any property of the sample expressible as a concentration can be obtained as the product of the concentration and the water discharge represented by the sample." (ASTM) Equal-width-increment (EWI) and equal-discharge-increment (EDI) sample-collection methods: Sampling methods that are specifically designed to result in the collection of discharge-weighted, depth-integrated, isokinetic samples. The procedures for collecting EWI and EDI samples are described in Edwards and Glysson (1998). When either method is used properly, the resulting samples contain the same property concentrations. Bottle samplers: Samplers that have rigid sample containers. Because these bottles are rigid, they do not instantly transmit the ambient pressure to the interior of the sample container and have neither pressure compensation nor opening and closing valves. Point samplers described in Edwards and Glysson (1998) use rigid bottles but have pressure compensation and opening and closing valves and are not considered bottle samplers for the purposes of this document. The tables in Appendix 4 were not designed for use with point samplers. Point samplers should perform as bottle samplers if held open from before the sampler enters the water to until the sampler leaves the water. Bag samplers: Samplers that have sample containers (bags) that instantly transmit the ambient pressure to the interior of the sample container and do not have opening or closing valves. Transit rate: The rate at which the sampler is passed through the water from the stream surface to the streambed or from the streambed to the surface. Unsampled zone: The part of the sampling vertical, usually assumed to be the zone from the streambed to the sampler intake. Sampler intakes are generally 4 to 7 inches above the streambed, depending on the type of sampler used. Increment: Refers to the subdivisions of the stream cross section made based on equal widths (using EWI) or equal discharge (EDI). Vertical: Refers to that location within the increment at which the sampler is lowered and raised through the water column. Centroid: The vertical within the increment at which discharge is equal on both sides. Appendix 2. Some uses and advantages of the equal-width-increment (EWI) and equal-discharge-increment (EDI) sampling methods EWI method USE EWI WHEN: · Information required to determine locations of sampling verticals for the EDI method is not available. OR · The stream cross section has relatively uniform depth and velocity. AND ESPECIALLY WHEN: · The location of EDI sampling verticals changes significantly at the same discharge from one sampling time to another. This situation occurs frequently in sand bed streams. Advantages of the EWI method · The EWI method is easily learned and used for small streams. · Generally, less time is required on site if the EWI method can be used and information required to determine locations of sampling verticals for the EDI method is not available. EDI method USE EDI WHEN: · Information required to determine locations of sampling verticals for the EDI method is available. AND ESPECIALLY WHEN: · Small, non-homogeneous increments need to be sampled separately from the rest of the cross section. The samples from those verticals can be analyzed separately or appropriately composited with the rest of the cross-sectional sample. (Have your sampling scheme approved.) OR · Flow velocities are less than the isokinetic transit-rate range requirement. A discharge-weighted sample can be obtained, but the sample will not be isokinetic. OR · The EWI sampling method cannot be used. For example, isokinetic samples cannot be collected because stream velocities and depths vary so much that the isokinetic requirements of the sampler are not met at several sampling verticals. OR · Stage is changing rapidly. (EDI requires less sampling time than EWI, provided the locations of sampling verticals can be determined quickly.) Advantages of the EDI method · Fewer increments are necessary, resulting in a shortened collection time (provided the locations of sampling verticals can be determined quickly and constituents are adequately mixed in the increment). · Sampling during rapidly changing stages is facilitated by the shorter sampling time. · Subsamples making up a sample set may be analyzed separately or may be appropriately composited with the rest of the cross-sectional sample. · The cross-sectional variation in constituent discharge can be determined if sample bottles are analyzed individually. · A greater range in velocity and depth can be sampled isokinetically at a cross section. · The total composite volume of the sample is known and can be adjusted before sampling begins. Appendix 3. Isokinetic, depth-integrating water-quality samplers and sampler characteristics This table could not be converted to text. The table is in Framemaker and is in the attached Framemaker file. Appendix 4. TABLES Tables 3 through 24 provide guidelines for using bag and bottle samplers to collect discharge-weighted, depth-integrated, isokinetic samples. Tables of reeling rates and transit rates (tables 4-15 and 17-24) list the theoretically defined minimum and maximum reeling rates (in seconds per turn) and transit rates (in feet per second) for various stream depths and velocities for commonly used nozzles and bottle or bag combinations when using an A, B, or E reel. In the tables, the minimum values are defined as "full" to indicate that when using a listed rate, the bottle will be full after one round-trip transit; the maximum values are defined as "fastest" to indicate the fastest reeling rate or transit rate that can be used for the isokinetic range of the sampler. The tables also list the volumes that should be in the samplers after one complete round-trip transit. The sample volumes, reeling rates, and transit rates assume one complete round-trip vertical transit of a sampler that, starting empty, goes from the stream surface to the streambed and returns to the surface at a sampling vertical of specified depth and mean velocity for a given bottle and nozzle combination. All depths shown in tables 3 through 24 are water depth minus the unsampled zone. A key assumption used here and in previously published work is that the velocity distribution at each vertical is that described in Edwards and Glysson (1998) in which the water velocity at the deepest point in the transit is 0.5 of the mean stream velocity in the vertical. The information provided in the tables is not new, but rather is a tabular representation of the information presented in the following references: Edwards and Glysson, 1998; Federal Interagency Sedimentation Project (FISP), 1952; and a written communication (distributed with each US D-77 sampler) from Hydrologist-in-Charge, Federal Inter-Agency Sedimentation Project, 2/21/79, Operating Instructions D-77 Suspended Sediment Sampler or similar identically computed information for newer samplers. The values in these tables were computed at each depth and velocity from the minumum and maximum transit rate ratios shown in figures similar to 39, 40, and 41 of Edwards and Glysson (1998) for the applicable nozzle and bag or bottle combination. The utility of the tables of reeling and transit rates may be enhanced if used with a vertical transit pacer VTP 74 (available from FISP). The mean velocity and depth of a sampling vertical must be known to use the tables and assure that 1- and 3-liter bottle samplers are used within their isokinetic range. The mean velocity of a vertical can be estimated adequately for sampling purposes by dividing 10 by the seconds required for a floating object to travel 11.6 feet at the sampling vertical. (Timing a peanut passing an 11.6-foot length of flagging trailing from a suspension cable works quite well.) Tables For Bottle Samplers Tables 3 through 15 in Appendix 4 apply to specific bottle, cap, and nozzle combinations and apply to samplers when that bottle, cap, and nozzle combination is used with the sampler. For example, the table for a 1-liter bottle and 5/16-inch nozzle applies to any of the approved samplers (such as US DH-81, US DH-95, US D-95) when that bottle, cap, and nozzle are used in the sampler. The range of velocities on the tables may exceed the velocity of a stream in which some samplers are stable. (An aluminum D-77 sampler is unstable in stream velocities greater than 3.5 feet per second, but the 3-liter table shows reeling and/or transit rates for 9 feet per second.) Table 3 lists the minimum volume that must be in the sample bottle after the first transit of the sampler from the stream surface to the streambed and return to the surface, at a sampling vertical of specified depth for a given bottle and nozzle combination. If the volume of sample in the bottle is less than that listed in table 3, the sample was not collected isokinetically. A volume equal to or greater than that listed and less than the maximum volume indicates, but does not guarantee, that the sample was collected isokinetically. Further indication that a sample was collected isokinetically is obtained by comparing the volume in the sampler with the volume computed from the product of nozzle area, mean stream velocity, and total transit time at the vertical. The volumes in table 3 were calculated for each size sample bottle using the minimum allowable transit rate for that bottle, nozzle, and depth combination. The minimum required volume depends only on the stream depth, bottle size, and atmospheric pressure and is independent of stream velocity and transit rate. The volumes listed in table 3 are for sea level and should be increased by about 4 percent for each 1,000 feet of elevation. When a sampler filled to the maximum (full) volume is tipped down from the horizontal, water will spill out of the nozzle; this spillage might increase the concentration of sand in the sample. When using the EWI method, sample spillage would result in underrepresentation of that vertical in the composited sample. In some conditions the maximum depth of sampling should be limited because the "full" volume of the sampler needs to be limited to a volume such that water will not be spilled when the sampler is used. For bottle samplers, the tables provide reeling and transit rates designated as "-10 tip." When these or faster rates are used, the sampler will not spill if tipped 10 degrees down from horizontal. A 10-degree-down tip reduces the operational depth of 1- and 3-liter bottle samplers about 3 feet because of the reduced maximum sample volume. When a sampler is filled to a volume exceeding the -10 tip volume, watch carefully to assure that the sampler has not overfilled. When a sampler is filled to the maximum (full) volume it is difficult to determine that it has not overfilled and spilled back to the maximum (full) volume. Tables 4, 5, and 6 list the minimum (full), -10 tip, and maximum (fastest) reeling rates (in seconds per turn) for various depths and velocities for a 1-liter bottle, cap, and 1/4-inch nozzle combination when using an A, B, or E reel. Table 7 lists the full, -10 tip, and fastest transit rates (in feet per second) for various depths and velocities for a 1-liter bottle, cap, and 1/4-inch nozzle combination. Tables 8,9, and 10 list the full, -10 tip, and fastest reeling rates (in seconds per turn) for various depths and velocities for a 1-liter bottle, cap, and 5/16-inch nozzle combination when using an A, B, or E reel. Table 11 lists the full, -10 tip, and fastest transit rates (in feet per second) for various depths and velocities for a 1-liter bottle, cap, and 5/16-inch nozzle combination. Tables 12, 13, and 14 list the full, -10 tip, and fastest reeling rates (in seconds per turn) for various depths and velocities for a 3-liter bottle, cap, and 5/16-inch nozzle combination when using an A, B, or E reel. Table 15 lists the full, -10 tip, and fastest transit rates (in feet per second) for various depths and velocities for a 3-liter bottle, cap, and 5/16-inch nozzle combination. Appendix 4.--Table 1. List of tables for bottle samplers Table Type Bottle Nozzle Reel Units 4 Reeling 1 L 1/4 A seconds/turn 5 Reeling 1 L 1/4 B seconds/turn 6 Reeling 1 L 1/4 E seconds/turn 7 Transit 1 L 1/4 any feet/second 8 Reeling 1 L 5/16 A seconds/turn 9 Reeling 1 L 5/16 B seconds/turn 10 Reeling 1 L 5/16 E seconds/turn 11 Transit 1 L 5/16 any feet/second 12 Reeling 3 L 5/16 A seconds/turn 13 Reeling 3 L 5/16 B seconds/turn 14 Reeling 3 L 5/16 E seconds/turn 15 Transit 3 L 5/16 any feet/second Tables for Bag Samplers Table 16 lists the minimum (full) volume that must be in a bag sampler after the first complete transit from the surface of the stream to the streambed and return to the surface, at any sampling vertical of specified depth for the specified nozzle. If there is less sample in the sampler than listed in table 16, the sample was not collected isokinetically, possibly because the transit rate exceeded four-tenths the mean stream velocity at that vertical. (Four tenths the mean stream velocity at a vertical is the maximum (fastest) transit rate allowed for isokinetic sampling.) The depths and velocities in the tables are arbitrary but focus on typical conditions that may frequently be encountered. There are many configurations for bag samplers and only the field personnel will know the stable range of their bag-sampler configuration. There is no single, exact volume for a bag sampler because each bag installation results in a slightly different volume. The full volumes used to develop tables for bag samplers assume the sampler is not allowed to spill and the nozzle is not tipped below horizontal. The maximum usable volume of a 3-liter bag sampler is estimated to be 2.6 liters based on USGS field experience. Tables 17, 18, and 19 list the minimum (full) and maximum (fastest) reeling rates (in seconds per turn) for various depths and velocities for a 3-liter bag, cap, and 1/4-inch nozzle combination when using an A, B, or E reel. Table 20 lists the minimum and maximum transit rates (in feet per second) for various depths and velocities for a 3-liter bag, cap, and 1/4-inch nozzle combination. Tables 21, 22, and 23 list the minimum and maximum reeling rates (in seconds per turn) for various depths and velocities for a 3-liter bag, cap, and 5/16-inch nozzle combination when using an A, B, or E reel. Table 24 lists the minimum and maximum transit rates (in feet per second) for various depths and velocities for a 3-liter bag, cap, and 5/16-inch nozzle combination. Appendix 4.--Table 2. List of tables for bag samplers Table Type Bag Nozzle Reel Units 17 Reeling 3 L 1/4 A seconds/turn 18 Reeling 3 L 1/4 B seconds/turn 19 Reeling 3 L 1/4 E seconds/turn 20 Transit 3 L 1/4 any feet/second 21 Reeling 3 L 5/16 A seconds/turn 22 Reeling 3 L 5/16 B seconds/turn 23 Reeling 3 L 5/16 E seconds/turn 24 Transit 3 L 5/16 any feet/second Appendix 4.--Table 3. Minimum volumes for bottle samplers This table could not be converted to text. The table is in Framemaker and is in the attached Framemaker file.