Chapter 2      Sampling

                      Sampling is a vital part of studying the quality of water in a water treatment process, distribution system, or source of water supply.  The major source of error in the whole process of obtaining water quality information often occurs during sampling.  While this fact is well recognized, it must be strongly and repeatedly emphasized.

                 In any type of testing program where only small samples (a liter or two) are withdrawn from perhaps millions of gallons of water under examination, there is potential uncertainty because of possible sampling errors.  Water treatment decisions based upon incorrect data may be made if sampling is performed in a careless and thoughtless manner.  Obtaining good results will depend to a great extent upon the following factors:

1.  Ensuring that the sample taken is truly representative of the water under consideration,

2.  Using proper sampling techniques, and

3.  Protecting and preserving the samples until they are analyzed (in a timely manner).

                 Collection of representative samples within the plant is really no different from sample collection in a stream or river.  The operator simply wants to be sure the water sampled is representative of the water passing that sample point.  In many instances, much money and effort is spent buying and installing pumps and piping to sample from a point that turns out to be not representative of the stream.  A sample tap in a dead area of a reservoir or on the floor of a process basin serves no purpose in helping the plant operator control the water quality.  The operator is urged to find each and every sample point and ensure it is located to provide a useful and representative sample.  If the sampling point is not properly located, plan to move the piping to a better location.

                 Over 50 % of the faulty data which occur in laboratory test results actually stem from the sampling process, rather than during laboratory activities.  The objective of sampling is to obtain a portion of material small enough to be transported to and handled conveniently in the laboratory while still providing test results which accurately represent the characteristics of the bulk material.  Sampling problems appear in a variety of areas such as:

                 Suitability of the container

                 Cleanliness of the container

                 Type of sample taken

                 Representativeness of the sample

                 Handling and storage of the sample prior to analysis

                 Length of time between sampling and analysis

                 For most purposes, the collection of a suitable sample for testing requires a considerable amount of forethought and preparation.  The major considerations consist of what test is to be performed and how the test results are supposed to be related to the tested material.  A sample taken for fecal coliform count will probably not be suitable for residual chlorine determination.  Obtaining a sample to determine BOD (Biochemical Oxygen Demand) and TSS (Total Suspended Solids) which fluctuates with influent daily flow, will use a completely different technique than taking a sample to determine the volatile acids/alkalinity ratio in an anaerobic digester.

                 The two types of samples most commonly obtained are the GRAB  sample and the COMPOSITE sample.

Grab Sample:  Grab samples are samples collected over a period of time not exceeding 15 minutes.  They reflect the source material conditions at a particular instant of time.  A grab sample is required for test parameters which must be measured immediately after collection and for tests which require the entire contents of a single sample container for analysis or for tests where the parameter will change over time.  Examples of test parameters which require grab samples are temperature, pH, residual chlorine, and dissolved oxygen.

Composite Samples:

  Composite samples are obtained by taking an appropriate number of grab samples collected at equal intervals or proportional to flow.  Flow proportional composite samples are collected when the flow and waste characteristics are continually changing. This single sample reflects the average conditions of a point source during an interval.  The nature of the composite sample requires the tested parameters to be stable in the container for the duration of sampling, often 24 hours. [This is usually meant that the samples are refridgerated to 4 °C.]

                 The sampling of a point source that is homogenous is a simple task.  Unfortunately most bodies of water are not well mixed and obtaining samples which are truly representative of the whole body depends to a great degree upon the sampling technique:

          Take samples where there is good mixing action in the wastestream.

          Sample each wastestream consistently in the same location (day after day).

representative of the bulk material and not the stagnant material in the line. AND, I recently observed, the valve should be as full open as possible, else it can happen that the valve can do some filtering.

          Clearly mark and identify the separate sample containers from each location.

          For most parameters, rinse the sampling device and the bottle  thoroughly with deionized water or with the sample (discarding the rinse into the effluent). Open the container carefully just  before the collection, holding the lid, close immediately after collection. Check the tightness of closure once more. Use a separate collecting device at each station, if possible. Check with the lab concerning whether the bottle should be Full (no air at all, a practiced tecnique of getting the top on without trapping an air bubble) or leave a small air gap.

                 Large solids (greater the ¼ inch in diameter) should be excluded from a sample.

                 Influent samples should be taken at points of high turbulence to ensure good mixing.  Sampling points should always be above plant return lines and sampling equipment should be placed to not interfere with flow measuring devices.  The preferred sampling points for raw wastewater are:

                 Waste flowing from the last process in the manufacturing operation.

                 Pump wet well ( if turbulent).

                 Upflow collection lines.

                 Aerated grit chamber.

                  Effluent  samples should be collected at the most representative site downstream from collection points of all effluent waste streams but prior to the entry to receiving waters.  For some municipal plants, samples should be collected after chlorination.  This will require dechlorination of some samples for  parameters such as BOD, toxicity testing and ammonia at the time of sample collection. (Testing for residual chlorine is a whole other story).

                 The act of dechlorination is one of a larger group of steps, called PRESERVATION, taken to ensure that the sample that is analyzed is reflective of the concentrations of the target analyte in the bulk material at the moment of sampling.  The three processes which act upon target analytes in samples while in the sample bottle are biological, chemical and physical.  Almost all wastewater contains microscopic plants and animals which can feed upon analytes, changing their concentration in the sample.  Some analytes such as nitrate and acid are byproducts of biological activity and the organisms can increase the concentration.  The organisms themselves can multiply and give falsely high counts, or poison themselves with their waste products and die off giving lower than actual counts.  Chemical interactions are varied and largely unpredictable because of the complexity of waste streams.  Known interferents can be controlled such as sulfide in a sample reacting with cyanide to form thiocyanate, removing both analytes from solution.  Nitrite can be oxidized by hexavalent chromium or other chemicals to nitrate and both target analyte concentrations changed.  Physical processes include precipitation and loss of trace heavy metals from solution and volatilization of dissolved gases from liquid and sludge samples. 

The EPA regulates the manner in which samples are taken, the preservatives used in the samples, the containers appropriate for sampling and the maximum length of time which can pass before the samples are analyzed (HOLDING TIME).  This information is published in Table II, 40 CFR 136.  It is a listing of the containers, preservation and maximum holding time for each parameter and is reproduced in Appendix I.   Once the sample is taken, steps must be taken to reduce any possibility that the sample will change on its own before analysis.

                 By far, the most common preservative is Cooling to 4 °C.  This can be achieved by making a slurry of ice and water and placing the collected sample in the slurry.  This serves to slow biological activity in the sample and keep dissolved gases in solution, however it can hasten certain physical processes such as precipitation of metals.  Other common preservatives include use of sodium thiosulfate at a final concentration of 0.008 % (80 mg/L of sample) to remove chlorine, addition of copper sulfate or mercuric chloride to halt biological activity, and addition of zinc acetate to trap sulfides.  Appropriate pH adjustment to either: over 12 SU (Standard Units) with sodium hydroxide, or less than 2 SU with hydrochloric, sulfuric or nitric acid (depending on the intended analyte) can inactivate biological processes, prevent known chemical reactions and maintain a target analyte in solution.  The analyst must be aware that acidification of nitrate-nitrite samples with nitric acid or preservation of sulfate samples with sulfuric acid will invalidate the analyses, so the sampler must make adequate notation.  Samples which are not preserved must be analyzed immediately.  Samples which are taken at the wastewater plant for analysis at the laboratory on-site are normally analyzed immediately and so preservation is not a big deal.  However, for samples which are analyzed by a commercial laboratory off-site, the improper use of preservatives can invalidate the sample for regulatory purposes and misinform plant operators.

                 The same consideration applies to holding times.  On-site laboratory analysis is done within, at most, hours of the sampling time and so operators and analysts rarely give any thought to holding times.  The collection of a sample for off-site analysis and then letting it sit in the trunk of a car until someone can run the sample over to the lab generally results in a missed holding time and an invalid sample.

                 The containers used for sampling can also make or break an analysis.  An operator obtained a water sample for chromium, nickel, iron, lead, copper and zinc trace metals analysis in a used paint can and took it to a laboratory.  Much to his surprise he was above his permit level on all the analytes.  The container had contributed metals to his sample and he was in violation.  The EPA requires use of either high density polyethylene or glass containers with Teflon® lid liners for most analyses.  Metals samples should always be collected in high density polyethylene containers since exchange phenomena can occur in glass; however, mercury can diffuse through the walls of plastic bottles and be lost.  The container must also be of the appropriate size.  Collecting a 100 mL sample for an analysis which requires 1000 mL is a waste of effort, while collecting a gallon of sample for a test which only need 5 mL simply contributes to laboratory waste disposal problems.  The cleanliness of the containers is of utmost importance.  For most purposes containers can be reused if they are cleaned with a non-phosphate detergent, rinsed with hot tap water, rinsed with dilute acid, then re-rinsed with tap water followed by reagent water and air drying.  For certain target analytes there are preferred cleaning procedures.  For example, for nitrate analysis the container should not be rinsed with nitric acid and for chloride analysis you would not use hydrochloric acid.  The suitability of the cleaning procedure should be checked at regular intervals by filling a cleaned container with reagent water and then analyzing the water for the target analytes.  If none are detected, then the cleaning procedure is shown to be suitable.  More information can be found in the Standard Methods for the Examination of Water and Wastewater.

 

 

Non-representative sampling:

I.               If an operator took a sample from a fire hydrant, the laboratory data would be representative of the contents of the fire hydrant.  However, fire hydrants, because of their construction and infrequent use, can give very erratic results when one is trying to monitor the water main.  An ideal water sampling point should be one that has a short, direct connection to the main, is flushed regularly or continuously and is made of corrosion-resistant material.

 

II.             If an operator takes a digester sample from a line (pipe) which is incompletely flushed, he may find that the volatile acids are unusually high.  He may then make the costly and unnecessary decision to add a neutralizing chemical.  The digester contents may in fact have volatile acids in the normal range.  It is very important to think about how closely the sample reflects the bulk material.

 

III.            A sample is taken from the primary effluent channel which leads to the aeration basins.  This channel is typically aerated.  When a sample is taken from a quiescent location (dead zone or quiet spot) various consequences can occur.  Under quiescent conditions solids and biochemical oxygen demand would generally be lower because of settling.  This can give an inaccurate estimation of the Food to Microorganisms (F/M) ratio, which is compounded into an inaccurate decision on the WAS - RAS (Waste Activated Sludge - Return Activated Sludge) requirement.  The effect of nitrification could be wrongly estimated because of the low ammonia results.

 

Sampling Techniques

                 Sampling techniques depend on the available equipment, the sampling site and the analyses to be performed.  Some possible variables to be considered and decisions to be made include:

-  Manual or automatic sampling equipment

-  Location of the sampling site, whether indoors, outdoors, or difficulties involved

-  Temporary or permanent sampler installation

-  Power availability, whether 110 VAC, compressed air, or batteries needed

-  Adverse or hazardous conditions for men or equipment.

 

                 Manual sampling tools include dippers, ‘bacon bombs’, weighted bottles and hand operated pumps.  A dipper consists of a wide-mouthed corrosion resistant container on long handle that is used to dip the sample from the bulk material and then pour it into the sample container for transfer to the laboratory.  The dipper container should be constructed of stainless steel or Teflon®.  Consideration should be given to using an adjustable metal band on the end of a handle to attach the actual sample container to the handle and avoid possible contamination while transferring the sample from capture to final container.  Weighted bottles are sample collection devices which are attached to a pole and have a heavy hinged stopper with a string attached.  A similar device is called a “bacon bomb”.  The bottle is lowered to the appropriate depth with the stopper in place, then the stopper is pulled open with the string, filling the container with the sample at the desired depth.  After the bottle is filled, the stopper is allowed to settle back into place, the closed bottle is raised to the surface and the sample is ‘brought home’ to the appropriate container.  The end of the influent hose to a hand operated pump can be lowered to the selected depth and the sample pumped to the surface and into the waiting container.  When employing the pump technique, flushing volumes are required.  One should always pump at least  3 volumes of pump and hose before capturing the sample.  And then, flush the pump and hose with several volumes of clean potable or deionized water before a new sample capture procedure begins.

                 Electrical, mechanical, and pneumatic samplers are available.  Great care must be taken in the selection of the proper instrument for sampling based on knowledge of the operating conditions within the plant and the nature of the wastewater being sampled.  It is very frustrating to a water survey team to diligently set up a sampling device and return a day later to find the sample container empty because the device has lost power or the flow stopped because a rag plugged the pipe.

                 Once the sample has been taken, the person responsible should make and record (on the sample container and in the field log book) certain observations and measurements before sending it to the laboratory.  These observations should include appearance (color, haziness, turbidity, presence of floc, oil on the surface), odor, and measurements that are valid only on a fresh sample (temperature, dissolved oxygen and pH (in situ).

                 Sampling wastewater correctly in an industrial complex often requires ingenuity.  First, something must be known about the point source to be able to select the best sample location.  For example, if the pipe is not completely filled, there may be three strata in the stream:  the upper might contain floating matter, the center may be free of floating matter and sediment, and the lower, very probably, contains solids that have settled in spite of the velocity through the pipe.  Truly representative sampling could conceivably require turbulent mixing upstream of the sampling location.  Usually, a compromise is made because mixing may be impossible:  the sampler must decide what data are the most significant for plant control purposes and get the most representative sample possible for that data.

                 Automated samplers are in widespread use, particularly for composite sampling requirements.  These samplers are programmable for a wide variety of modes:  single grabs, time composite, flow proportional, and more.  Most are insulated and can be filled with an ice-slurry for cooling preservation.  Some are power refrigerated, even battery operated.  Lock and chain are advised, partly because of the cost, and to ensure that the samples have not been tampered with.  There have been cases where violent weather or animals have dragged off and damaged the equipment.

                

Once the sample is collected, the operator must apply complete, legible and permanent labeling to that container.  An appropriate label will be water resistant and glued, with a non-water soluble glue, to the sample container.                                                     The label should be written in ink and detail who, when, where, unique identification, and other pertinent data.  Temperature and pH should be logged at this time and on the label. Certainly, unusual conditions must be written. A Field Log Book will contain at least the same information and usually more, for tracking and ease of data transfer to the Chain-Of-Custody (COC) record.

                

To establish a legal record of the sampling  (which is required in regulatory investigations and enforcement actions) a COC record is prepared and maintained for each set of samples.  Information on the COC record (also consider this) includes all the information on the sample label and includes spaces for signatures, dates and times of all of have possession, each person in turn, of the samples.  The COC record, properly filled out, must accompany the sample at all times until it reaches the laboratory.  A copy of the COC record must be maintained by the sampling organization (as a legal record) for at least three years. 

A sample page of a Chain of Custody sheet

                 Shipment of samples is the responsibility of the sampling organization.  DOT regulations govern sample shipments and these should be consulted prior to any shipments through commercial carriers.  The considerations of sample shipment include preventing breakage of the sample containers, maintaining sample preservation and maintaining sample security for COC.  Consider:  if you were opening the ice chest of samples and one bottle was broken, you would have to clean the mess of sample and broken glass , identify and carefully inspect all the other samples.  The image is not pretty.  Not to mention the loss of vital information, the extra paper work, the emotions and the loss of respect.  The responsibility of getting reliable analysis from the laboratory begins and remains with the sampling organization.  Collecting, documenting, packing and shipping samples in a neat, well-organized procedure is the vital first step to obtain accurate analysis.

                 Safety is always important.  Getting hurt just getting a sample certainly is not fun.  Proper preparation will always enhance the ease and efficiency and safe well-being of all concerned ‘just getting a sample’.  Plan the work, then work the plan.  Organize.  Consider alternate contingencies.  Life jackets, first-aid kits, personal cleanliness and buddy system are some thoughts to consider when ‘just getting a sample’.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Questions for Chapter 2

 

1.  Describe the time when a Chain-of-Custody record is prepared.

 

2.  Define ‘holding time’.

 

3.  Give three examples of preservation techniques.

 

4.  Sample containers are constructed of what two approved materials.

 

5.  Describe flow proportional composite sampling.

 

6.

return to Table of Contents
Back to Top	On to Chapter 3
My question sets for self exam		errors?