August 16, 1976 QUALITY OF WATER BRANCH TECHNICAL MEMORANDUM N0. 76.24-T Subject: EQUIPMENT & SUPPLIES: Sample Splitter for Water- Sediment Samples Within the next few months, all offices that are involved in the collection oŁ water-quality samples under the NASQAN program will be receiving a new type of sample splitter to be used on site to split samples into the various containers required by the Central Laboratories system. The attachment to this memorandum describes the USGS Churn Splitter and presents procedures to be followed in splitting composited samples. Research and development of this splitter was carried out for the Quality of Water Branch by Jack Dewey, assisted by Bruce Delaney and Kim Ong of the New Mexico District. Experiments have shown that if proper procedures are followed splits of samples containing up to 5000 mg/l sand (coarser than 0.062 mm) can be obtained to within +/- 10% of the true concentration. Splitters are being furnished at this time for use at all NASQAN sites; however, we recommend that they also be used whenever water sediment sample splitting of large composit samples is required in the field or laboratory. We plan to have smaller versions (2 gallon or smaller) available within the next few months. Those who receive these splitters will find a copy of the instructions inside. After delivery of the instructions to the manufacturer, however, we revised the 2nd paragraph of page 2 and the spelling of "liter." Because this is the first model, we will welcome comments regarding its use and/or improvement. R. J. Pickering Chief, Quality of Water Branch Attachment WRD Distribution: A, B, FO, PO Procedures for Subsampling Water-Sediment Mixtures (14-Liter Churn Splitter) General - The water-quality laboratory may require 4 to 16 subsamples of a representative cross-section sample of the water- sediment mixture (streamflow) for water-quality analyses. The cross-section sample is collected in,l-pint-or-l quart bottles using suspended-sediment samplers at no fewer than three and preferably eight to ten verticals (ETR or EDI techniques).These samples are composited into one single representative cross- section sample of the streamflow. This composited sample can then be split, using the churn splitter, into the required 4 to 16 representative subsamples as explained under Procedure. Samples collected for organic analyses (e.g. organic carbon, pesticides) should not be composited in this container because of the possibility of contamination from the plastic. 14-Liter churn splitter - The churn splitter is a 1/4 inch thick polyethylene cylinder, 10 inches in diameter and 12 inches deep with a lid. It has been manufactured for the Survey by a commercial manufacturer. The valve and spout are polypropylene. The stirring disc is a 3/8 inch thick polyethylene disk, 9-15/16 inches in diameter with 16 holes, 8 as scallops in the outer edge and 8 in an inner circle. The handle, a l-inch diameter by 18-inch long polyethylene rod, is welded perpendicular to the center of the disc and supported by four ribs. A small "lip" on the disc aligns with the valve, and a guide notch and rib are provided to maintain the correct alignment. Replacement valves and spouts are available from the Quality of Water Branch. Procedure - This procedure requires a total sample volume of 8 to 14 liters, of which 4 to 10 liters are suitable for water- sediment mixture subsamples. The remaining 4 or more liters may be used for filtered subsamples if required by the analytical schedule. If not, they may be discarded. This size churn splitter does not reliably produce representative water-sediment mixture subsamples when it contains less than about 4 liters. Before starting to collect the representative sample of the streamflow, label all the subsample containers to be used and determine the total sample volume needed. Add to this sample volume at least 10% to cover filter losses and spillage. It is less frustrating to throw away a small.amount of sample than to have to go back and collect another cross-section sample. Collect approximately one liter of water and thoroughly rinse the churn splitter. Representative samples of the streamflow are collected by using standard EDI or ETR sampling techniques as described in "Field Methods for Measurement of Fluvial Sediment" TRI Book 3, Chapter C2. Specific sample volumes cannot be obtained with sediment samplers, but properly-collected pint bottles (approximately two-thirds full) will yield about 1/3 liter each. Only one sediment sample bottle is used over and over again in collecting the cross-section samples in order to minimize the amount of sediment lost in transferring samples from the bottles to the churn splitter. Each time the bottle is filled, the sample is poured into the splitter and the bottle is used again so that each succeeding sample washes the sediment left from the previous one into the splitter. Remember that the volume to be used for water-sediment mixture subsamples must be "on top of the 4 liters of sample in the tank from which representative water-sediment mixture subsamples cannot be obtained. Suspended-sediment concentration should always be determined whenever a sample is analyzed for total concentrations of chemical constituents. The sample for determination of suspended-sediment concentration can be collected (l) as a separate cross-section sample as if no other sampling were required, or (2) it may be obtained as a single-bottle subsample from the churn splitter if the amount of sediment in a single bottle appears to be sufficient for the lab to obtain accurate weights of both the fine and sand fractions. The fieldman can decide whether (1) or (2) will be used by looking at the first bottle collected to see whether an appreciable amount of sand settles to the bottom of the bottle within 20 to 30 seconds; if so, (2) can be used. If in doubt, always use alternative (1). When the required volume plus 10% for waste is in the churn splitter, place all water-sediment mixture subsample containers within easy reach, so that once started the stirring can be continuous. The sample should be stirred at a uniform rate of approximately nine (9) inches per second. As the volume of sample in the tank decreases the round-trip frequency should increase so that the churning disc velocity remains the same. The disc should touch the bottom of the tank on every Stroke, and the stroke length should be as long as possible without breaking the water surface. Before using the sample splitter for first time, practice this stroke using tap water. Observe that, as the stroke length and/or disc velocity is increased beyond the recommended rate, there is a sudden change of sound and churning effort which is accompanied by the introduction of excessive air into the mixture. The introduction of excessive air into the sample is undesirable because it may tend to change the dissolved gases, bicarbonate, pH, and other characteristics. On the other hand, inadequate stirring may result in non-representative subsamples. The sample in the splitter should be stirred at the uniform churning rate for about 10 strokes prior to the first withdrawal to establish the desired stirring rate of nine (9) inches per second and to assure uniform dispersion of the suspended matter. The churning must be continuous during the withdrawals; therefore, if a break in withdrawals is necessary, the stirring rate must be reestablished before continuing the withdrawals. When all of the required water-sediment mixture subsamples have been obtained, the remaining portion of the sample is used as necessary for the filtered samples. It will be advantageous to allow the sediment to settle out in the mixing tank for a.few minutes before pouring the sample into the filter apparatus. When all of the necessary filtered subsamples have been obtained, the mixing tank, churning disc and filter apparatus should be cleaned thoroughly with deionized water. If deionized water will not remove all of the residue, clean by using a small amount of a detergent such as Alconox, rinse with a weak acid solution (4 ml of nitric acid per liter of water), rinse repeatedly with tap water, and then rinse with deionized water. Equipment not furnished - 1. A stand to support the mixing tank that will allow the subsample containers to be placed under the spout. 2. Small supports to hold various sizes of subsample containers such that the top or opening is at or near the mixing tank spout. NOTE: When used in compositing samples collected from a bridge or roadway the lid should be kept on at all times except when pouring sample, in order to protect sample from dust contamination.