By Henry Nowicki, Barbara Nowicki and Charles Kieda
Summary: This article presents general considerations for sampling drinking water supplies for subsequent determination of organic and inorganic chemicals at laboratories using regulatory testing methods. The goal is to reduce errors and lost time in the process that can hurt business.
When it comes to sampling, amajor mistake likely will bemade. The project manager often assigns this task to relatively new and inexperienced staff members. While there’s a myth that anyone can get field samples and resources need to be concentrated at the laboratory where measurements are made, it has been our experience that the opposite is true. Clients often tell us they encounter inexperienced samplers and would like to work with more knowledgeable individuals.
It makes sense that highly qualified personnel should do the sampling and face-to-face client interface functions. Keeping existing clients is much easier than having to find new ones. Conversely, it’s only logical to keep inexperienced employees out of field sampling and direct client communication until they are properly trained. Samplers need to coordinate with labs. Having samples arrive at the lab when they’re expected is a good strategy. Prepare your lab by having good communications. Multiple measurements on-site also ensure a good sample.
Take the lab on the road
We recommend environmental project managers reconsider their allocation of resources between sampling and laboratory testing. Also, much of the laboratory work can now be taken to the job site in the field, allowing an individual who can make and explain the field testing results directly to the client if needed. Projects are decided because of the sampler’s knowledge and communication skills. The sampler’s role shouldn’t be trivial. The client doesn’t want to spend large sums of money to find out the samples weren’t taken correctly, the laboratory data isn’t useful and work must be repeated because of improper sampling.
Do a site visit and make a sampling plan (proper containers, preservatives, representative and completeness of sampling, holding times, transportation, blanks, duplicates, etc.) before starting the project. The sampling plan should be a written document, essential for large and complex projects. Consult with the client on what needs to be measured and the concentrations, parts per million/billion. Find out how the data will be used and if prior data on the measurements are available; this information is useful for the quality assurance plan. Sampling considerations for selected inorganic and organic determinations for drinking water supplies follow.
Types of metal water fractions
Before collecting a sample, decide what fraction is to be analyzed (dissolved, suspended, total or acid-extractable). This will partly determine whether the sample is acidified with or without filtration and the type of digestion required.
- Dissolved metals—Those constituents (metals) of a non-acidified sample that pass through a 0.45 µm membrane filter
- Suspended metals—Those constituents (metals) of a non-acidified sample that are retained by a 0.45 µm membrane filter
- Total metals—The concentration of metals determined on an unfiltered sample after vigorous digestion, or the sum of the concentrations of metals in both dissolved and suspended fractions
- Acid-extractable metals—The concentration of metals in solution after treatment of an unfiltered sample with hot dilute mineral acid
To determine dissolved and suspended metals separately, filter immediately after sample collection. Don’t preserve with acid until after filtration. Serious errors may be introduced during sampling and storage because of contamination from the sampling device, failure to remove residues from sample container, and loss of metals by adsorption on and/or precipitation in sample container caused by failure to acidify the sample properly.
Sample containers for metals
The best sample containers are made of quartz or TFE. Because these containers are expensive, the preferred sample container is made of polypropylene or linear polyethylene with a polyethylene cap. Borosilicate glass containers may also be used, but avoid soft glass containers for samples containing metals in the microgram per liter (µg/L) range. Store samples for determination of silver in light-absorbing containers. Use only containers and filters that have been acid rinsed.
Preserve samples immediately after sampling by acidifying with concentrated nitric acid (HNO3) to pH < 2. Filter samples for dissolved metals before preserving. Usually 1.5 ml concentration HNO3 per liter sample is sufficient for short-term preservation. For samples with high buffer capacity, increase amount of acid (5 ml may be required for some alkaline or highly buffered samples). Use commercially available high-purity acid or prepare high-purity acid by sub-boiling the distillation of acid.
After acidifying the sample, store it ideally in a refrigerator at approximately 4oC to prevent change in volume due to evaporation. Under these conditions, samples with metal concentration of several µg/L are stable for up to six months (except mercury, for which the limit is five weeks). For µg/L levels, analyze samples as soon as possible after collection.
Alternatively, preserve samples for mercury analysis by adding 2 ml/L, 20 percent (weight by volume), K2 Cr2O7 solution (prepared in 1 + 1 HNO3). Store in a refrigerator not contaminated with mercury. Be aware that mercury concentrations may increase in samples stored in plastic bottles in mercury-contaminated laboratories.
Special precautions are necessary for samples containing organic compounds and trace metals. Because many constituents may be present at µg/L concentrations, they may be totally or partially lost even if proper samples and preservation procedures are used. This is known as a false negative result (see Sampling terms).
Types of organic water fractions
Two major types of organic water fractions are semi-volatile organic compounds (SVs) and volatile organic compounds (VOCs). VOCs can be transferred from the water phase to the vapor phase by passing a gas stream through water. A rule of thumb is that a compound with water solubility of less than 1 percent and molecular weight of 350 or less will be efficiently (90 + percent) sparged from water by an applied gas stream. This technique is called purge-and-trap analysis. SVs are compounds that can be transferred from the aqueous phase to methylene chloride at acidic and basic pH. The SVs include neutral, acidic and basic organic compounds, which can be analyzed by conventional gas chromatography interfaced to a mass spectrometer. Pesticides and polychlorinated biphenyls (PCBs) are classified as SVs.
Unlike metal sampling, which predominately uses plastic containers, organic samples must be collected in glass. Plastic removes some organics from water. Glass containers must be scrupulously washed to provide containers to make trace determinations. Several firms sell pre-cleaned glass containers to U.S. Environmental Protection Agency (USEPA) standards. The cleaning process can be done in a laboratory, and the cleaning steps can be found at http://www.epa.gov/Region4/sesd/eisopqam /eisopqam.pdf on pages 339-341. Also, the USEPA site and Standard Methods1 have a lot of information about sampling. Sample containers must be demonstrated to not provide any detectable contaminants for volatile organic analysis (VOA) and SV determinations. A supply of sample containers for short-term notification needs to be available.
SV containers have brown glass to protect light-sensitive organic compounds. A common name for SV containers is a Boston brown quart bottle. The lid contains a Teflon pad insert with no glues on the lid. Lids with glues would contaminate the water sample. Each sample must be labeled with waterproof ink to maintain its unique identification.
VOA containers are clear glass with a septa and a threaded plastic cap for closure when filled. The septa has a Teflon pad on at least one side to protect the sample from loss by absorption into the septa’s rubbery part. When filling the VOA container, the sample volume needs to eliminate all vapor head-space above the water. An air space the size of a pea would result in significant loss of VOA from the water phase to the air phase—a drastically low VOA lab result. After the container has been filled with the water sample to a concave meniscus—water level curve—above the top of the container, the septa is added with the teflon side toward the water and the lid is screwed on. Field testing is recommended to assure there’s no head-space at a 180° inversion of the filled container. If there’s an air bubble in the sample, open the vial and add a little water and then close for another inversion test. Don’t empty the sample vial and start again. If you do this, the preservative will be lost and the sampler will use a dirty container on the second refill.
Preservatives for organics
After collection of VOA and/or SV samples, they must be stored on ice at 4oC until delivered to the lab. Cold storage slows sample deterioration, but it doesn’t stop it. Glass containers should be wrapped in bubble-wrap, or other actions taken to avoid glass breakage in transportation. Packing smart is important. Taking duplicate samples to assure one survivor is expensive but sometimes necessary.
VOA samples are preserved with mineral acid to extend the shelf life of aromatic compounds. Acid destroys the microorganisms in a sample and extends the holding time from seven to 14 days for benzene, toluene and xylenes. Labs need to validate the received sample is at pH 2.
Chlorine sampling & storage
Chlorine in aqueous solution isn’t stable and the chlorine content of samples or solutions, particularly weak solutions, will decrease rapidly. Exposure to sunlight or other strong light or agitation will accelerate the reduction of chlorine. Therefore, start chlorine determinations immediately after sampling, avoiding excessive light and agitation. Don’t store samples to be analyzed for chlorine. Samples kept for short periods before testing should be in completely filled sample containers that minimize the head-space above the water, and immediately be placed on ice in a dark cooler.
You need to evaluate how you’re presently sampling. Do you have the correct sampling containers, preservatives and paper work? Does the lab or client do the sampling? If the client does the sampling, are they using and following good quality assurance practices? Remember most of the errors and time consumption in environmental chemistry measurements are due to sampling—and not the instrumental methods.
Accuracy: The degree of agreement of a measured value with the true or expected value of the quantity of concern.
Blank: The measured value obtained when a specified component of a sample isn’t present during the measurement. The measured value/signal for the component is believed to be due to artifacts; hence, it should be deducted from a measured value to give a net value due to the component contained in a sample. The blank measurement must be made so that the correction process is valid.
Bias: A systematic error inherent in a method or caused by some artifact or idiosyncrasy of the measurement system. Temperature effects and extraction inefficiencies are examples. Blanks, contamination, mechanical losses and calibration errors are examples of the latter kind. Bias may be both positive and negative, and several kinds can exist concurrently so net bias is all that can be evaluated except under special conditions.
Bulk sampling: Sampling of a material that doesn’t consist of discrete, identifiable, constant units, but rather of arbitrary, irregular units.
Composite sample: A sample composed of two or more increments selected to represent a population of interest.
Cross sensitivity: A quantitative measure of the response obtained for an undesired constituent (interferant) as compared to that for a constituent of interest.
Detection limit: The smallest concentration/amount of some component of interest that can be measured by a single measurement with a stated level of confidence.
Duplicate measurement: A second measurement made on the same (or identical) sample of material to assist in the evaluation of measurement variance.
Duplicate sample: A second sample randomly selected from a population of interest to assist in the evaluation of sample variance.
False negative result: The test result indicates the target analyte isn’t present in the sample when it truly is present.
Gross sample: One or more increments of material taken from a larger quantity of material for assay or record purposes.
Protocol: A procedure specified to be used when performing a measurement or related operation as a condition to obtain results that could be acceptable to the specifier.
Random sample: A sample selected from a population that uses a randomization process.
Split sample: A replicate portion or sub-sample of a total sample obtained that’s not believed to differ significantly from other portions of the same sample.
- Greenberg, A., L. Clesceri, and A. Eaton, ed., “Standard Methods: for Examination of Water and Wastewater,” Washington, D.C., American Public Health Association, 2000.
About the authors
Henry Nowicki, Ph.D., directs laboratory testing and consulting services for Professional Analytical and Consulting Services (PACS) Inc. Dr. Nowicki has 25 years of practical experiences with activated carbon and other sorbents applied to environmental water, air and soil projects. He’s also a member of the WC&P Technical Review Committee.
Barbara Nowicki directs the PACS Short Courses and Conferences program. Fifty-seven courses and four focused conferences are provided annually including the International Activated Carbon Conference in September and the Environmental Sampling, On-Site Analysis and Sample Preparation Conference in October, both in Pittsburgh.
Charles Kieda works on special projects at PACS Inc., including technical writing, methods development and comparison, R&D, organic compound measurement, field sampling, environmental data validation and technical litigation support for lawyers.