A basic guide to the essentials: where to collect samples and how to install automated equipment.

By Johnny Barron

Although not entirely new, stream sampling is now performed more than ever. Tightening regulations, such as the National Pollutant Discharge Elimination System (NPDES) stormwater rule, and data gathering necessary to develop total maximum daily loads (TMDLs) for impaired waters and to prepare watershed management plans are requiring more people to carry out sampling programs.

To effectively sample streams, you need to understand basic stream characteristics. First, flowing streams are dynamic in nature. Water quality and quantity are seldom constant. If you were to take continual measurements of any parameter, such as dissolved oxygen or turbidity, you would notice that concentrations fluctuate over time. A single grab sample is only a "snapshot" of stream conditions and might not be representative of the overall condition of the stream. For this reason, we usually establish sampling stations that can be sampled repeatedly over several months or, better yet, several seasons. Establishing a sampling station can be as simple as placing some flagging and stakes to mark the location, or it might involve installing automatic monitoring equipment.

The first step in deciding when and where to sample is to review your purpose for sampling. For instance, if you were trying to measure impacts of a point discharge from an industrial facility, you would want to sample upstream and downstream of the discharge, ideally during a period peak discharge. If you were trying to measure impacts from nonpoint-source pollution from an urbanized area, you would select a location far enough downstream to capture all runoff from the area and sample during and after storms. If you were sampling to determine basic water quality in an effort to establish TMDLs or as part of a water-quality assessment, you would need to establish enough sampling points to represent major land-use categories. In this case, you would collect samples from both wet and dry events.

Next, review road maps, topographic maps, land-use categories, previous sampling reports, and other available data to select potential sampling locations. You will want to visit each location and verify that the site is physically and legally accessible. You should obtain written permission from any property owners whose land you must cross to access your sampling station. Make notes of potential hazards, such as steep banks and barbed-wire fences.

Now that you have access to your site, make sure it is representative of what you are testing for. Look for potential interferences. If you are sampling for fecal coliforms to detect municipal wastewater discharges, you don't want to sample downstream of livestock. If you are testing for turbidity in runoff from a construction site, a freshly plowed farm field could cause you grief.

When you have a site that is both accessible and representative, you'll need to examine the stream. You likely will see different types of stream sections. Pools are deep, slow-moving water; riffles are shallow, swift water; runs are sections of moderate depth and velocity. Riffles will have a high dissolved-oxygen content and higher sediment loads. Conversely, pools have a lower dissolved-oxygen content and more sediment deposition. Runs are the sections most representative of overall stream quality.

There are also different types of streams. Perennial streams are free-flowing year-round and support a wide variety of aquatic life. Intermittent streams are seasonal and support some aquatic life. Ephemeral streams flow only after storms and generally do not support aquatic life. Perennial and intermittent streams have a dry-weather base flow charged by groundwater, but ephemeral streams are charged only by stormwater. If you are sampling to measure pollutants in stormwater runoff, an ephemeral stream might be a good place to sample. For basic water-quality data, however, you should sample perennial streams.

If manual samples are to be collected, you might need to do nothing more than mark the sampling location with stakes and tape and clear a path to the spot. It's a good idea to measure the cross-sectional area of the stream channel at your selected sampling location. Once the area is known, a simple staff gauge can be installed to measure water depths during sampling events. The depth and area can be used to calculate the flow rate for each sampling event. Measure the cross-sectional area by stretching a cloth measuring tape across the top of the channel. Place a stake at the top of each bank to hold the tape. When pulled tight, this tape represents the top of the stream channel. Next, using a surveyor's level and rod, measure the elevation at regular intervals (usually 1 or 2 ft.) across the channel. Using the measured elevations and width, you can plot the cross-section by hand on graph paper or with the aid of computer-drafting tools.

A staff gauge can be easily installed by anchoring a 2- x 4-in. piece of lumber to a tree on the streambank, ideally out of the main flow. If possible, anchor the 2x4 to a steel C-post driven into the streambed. Make sure that the gauge is set to read the depth from the deepest part of the cross-section and that the gauge is plumb. Each time samples are collected, record the water depth. Now, for every sampling event, you can determine the flow rate.

An automatic stream gauge can be installed with a rain gauge that can gather enough data to compute a hydrograph for each rain event. Although the equipment can be expensive, automatic monitors can provide valuable data about stream flows during and after storms. These data can be used to plan optimum sampling times to capture peak pollutant loadings.

If you are installing an automatic sampler, you also need to consider the meander of the stream. Streams do not like to travel straight paths. The frequency of stream bends is referred to as sinuosity. Different factors, such as soil type, substrate, and slope, will affect sinuosity. In a stream bend, the outside edge of the bend has the longest water path and, consequently, the swiftest water. This is the cutting side where we would expect to find steep cut banks and deep water. The inside edge has the shortest path and, therefore, the slowest water. This is the deposition side where we would expect to find a sandbar. Install automatic samplers on the cutting side of streams so that sand doesn't bury the sample collection tubes.

Manual Sampling

Staff gauge installed at sampling station is mounted to steel stake in streambed and to adjacent trees so as to be plumb.

Manual sampling involves filling a sample bottle by hand. The simplest way is to don boots or waders and enter the stream just downstream of the designated sampling point. Walk upstream a few steps to make sure your sample will not collect bottom sediments that you disturbed when entering the stream. Generally you want to sample at the horizontal and vertical center of the stream or run. Sampling at midstream and mid-depth is standard practice because water quality at this point is more representative of the whole stream. Remember that this point is the horizontal and vertical center of the water, not the channel itself. The horizontal center of the channel might be a sandbar or a rock. For shallow water (less than 6 in. deep), it will suffice to fill the bottle from the surface of the stream rather than sample at mid-depth, because taking a mid-depth sample in shallow water is likely to stir up bottom sediments. For deeper water, sample at mid-depth by leaving the lid on the sample bottle and lowering the bottle to mid-depth position, then remove the lid and allow the bottle to fill. Replace the lid while the bottle is still submerged at mid-depth.

When sampling, reach upstream of where you are standing and make sure the bottle opening faces upstream. Avoid collecting floating debris. Bottles should be labeled with sample identification before sample collection to avoid confusing upstream and downstream samples. Other information regarding the sampling location, date and time collected, sampler's name, and chemical preservative also can be recorded on the sample labels. Similar information should be recorded in a log book in addition to on the label. It is also a good idea to record general conditions and water depth at the time of sampling. Data from a rain gauge or a flow meter also may be recorded.

Although standard in practice, it is not always practical or safe to sample at mid-depth and midstream, particularly if sampling after a rainfall. It might be wise to use a sampling pole that allows you to remain safely on the streambank. Even streambanks can be risky, as undercut banks might shear off and fall into the channel, and floodwaters can rise above their banks. Use extreme caution when collecting manual samples after a storm. If worst comes to worst, a rope and a bucket off a bridge might be the safest way to sample. Obviously, in extreme conditions we are not concerned with the midstream, mid-depth concept. We are content to return from our sampling trip in one piece.

Traffic is another significant safety risk. Because there are not usually parking places at sampling locations, you will probably have to park on the road shoulder. Be cautious with narrow roads and bridges so as not to endanger yourself or other motorists. Other safety concerns are steep banks, deep water, swift currents, poison ivy, and wasp nests.

The advantage of manually collecting samples is that you can personally observe conditions and make modifications as needed. If you feel the water you just collected is not representative, you can take another sample. If conditions have changed and your sampling location needs to be moved, no problem. Furthermore, equipment fees will be relatively low. To get started, you might want to purchase some waders and a sampling pole. Many laboratories will supply sample bottles and coolers.

The biggest disadvantage of manual sampling, of course, is that you have to be there, which limits your ability to monitor multiple streams and generally requires more labor. This is particularly true for sampling after storms when you have a certain window of time in which to collect a representative sample.

Automatic Sampling

Automatic sampler and tipping-bucket rain gauge.

Sample tube installed at mid-depth, rear cutting edge of bank upstream of discharge.

A viable alternative to manual sampling is installing automatic sampling equipment that can collect samples for you. In most cases, you will have to retrieve the samples and transport them to a laboratory. The downside to automatic sampling is that installation and programming can be complicated. Automatic samplers with flow meters or rain gauges can be programmed to sample at regular time intervals, at certain stream flows, or after a target rainfall accumulation. Automatic samplers can also be equipped with telemetry to transmit data back to your office.

An automatic sampler consists of a peristaltic pump, which draws water through a flexible sample tube from the stream and into a collection bottle. An electric motor powered by batteries or solar panels drives the pump. Most samplers have a computerized control device, with LCD display, that allows programming of sampler functions. The collection end of the flexible tubing should be fitted with some type of screen to keep debris out of the sampling tube.

Most samplers have a full-bottle shutoff valve, which turns off the pump when the bottle gets full. The device is a simple float switch beneath the bottle lid that slides up and completes a circuit. A sampler thus equipped will work only if the sampler remains upright. Therefore automatic samplers should be secured in an upright position.

Ideally the automatic sampler will be located out of the channel and above the floodplain. Unfortunately this often is not possible. The pumps will draw only up to 20-25 vertical ft. of head, sufficient to reach outside most stream channels. Samplers must be secured so that floodwaters do not carry them away or tip them over. You can accomplish this by chaining the sampler to a large rock or bolting the sampler to a tree. If there are no trees at the sampling location, you might need to drive steel stakes or rebar into the ground. A wooden post may be used, but it can be easily removed by floodwaters. Unfortunately there are no guarantees; there will always be a risk of equipment loss from flooding. It might be wise to post an address and a phone number inside the automatic sampler in case a fisherman finds it downstream.

You might need to protect equipment from more than just floodwaters. Samplers can be stolen, vandalized, or even shot. Where security is a concern, inconspicuous places are best. You can place the sampler inside a steel drum or a section of pipe. A more extreme measure is to conceal the sampler beneath the ground in a buried drum or box. You can also build an aboveground wooden box to conceal the sampler or install a prefabricated metal box with chains and padlocks. Warning labels, such as electrical-hazard labels, might deter curious visitors.

Sample intake tube installed at midstream and mid-depth.

Once the sampler is installed, anchor the sampling tube in the vertical center of the stream so that it won't draw bottom sediment, yet ensure that the collection tube always will remain submerged. For small to moderate sandy-bottom streams, this is most easily accomplished by driving metal C-posts (at least two) into the streambed. These posts should have holes in which to tie. Drive the tops of the posts as deep as possible, leaving enough of the post above the streambed so you can remove it when the project is complete. Ideally the tops of the posts will be below the water surface to minimize debris collection on them. Using wire, cable ties, or hose clamps, securely fasten the sample collection tube screen to one post closest to the horizontal and vertical center of the run as possible. The tubing should not be able to slide through the fastener, nor should the fastener be able to slide up and down on the stake. The sample collection tube should remain fixed during all flow conditions. However, take care not to clamp the tubing.

Use additional posts or stakes to secure the sample collection tube beneath the water surface to reduce the chance of it getting snagged by a tree. In a similar manner, tie off the tubing to stakes or trees on the bank. Make sure there are no vertical loops in the tubing where water can collect.

Although you might not be able to drive posts into a rocky streambed, you might be able to drive stakes horizontally into the bank at the water's surface. If this is possible, secure the sampling tube to the post(s) so the screen hangs down to the mid-depth point. Make sure this stake is firmly anchored, or a passing log might end up taking it downstream.

For a deep-water stream, you probably won't want to drive stakes into the streambed. In this case, secure a steel angle or pipe to the bank so that it protrudes down into the water about 24 in. Tie your tubing to the angle with cable ties or hose clamps. When using pipe, feed the sample collection tube into the pipe until the screen is at the submerged pipe end. Make sure the pipe or plank is securely mounted to the bank so it won't wash away.

Once the automatic sampler is installed, it must be programmed. The specific process for programming varies significantly from one brand to another. For details of a particular brand, refer to the owner's manual or contact the vendor.

Generally speaking, your sampler should have the following features:

Peristaltic Pump. This type of pump functions by clamping down on a rubber tube and pushing the clamp toward the sample collection bottle. This action pushes air toward the bottle, creating suction at the sampling end. The water never leaves the tubing, thus minimizing the risk of contamination.

Line Purging. Before the sample is collected, and immediately after collection, the pump should run in reverse and push excess water out of the tube. Water left in the tube could dilute the next sample and interfere with test results. Additionally, water in the line can freeze in cold temperatures and prevent the sampler from drawing a sample. Some brands have preset purge times, and others allow you to set your own.

Sample-Collection Bottle. Depending on the parameters you are testing for, you might need a glass, polyethylene, or Teflon bottle. Polyethylene usually is sufficient for parameters sampled with automatic equipment. Bottle size varies from brand to band. The volume required depends on the required laboratory tests. You might need only 100 ml to test for turbidity, but priority pollutant testing might require 8 lit. or more. The volume to be collected is programmable on most brands.

Power Source. A primary battery is needed to power the unit. A smaller, secondary battery should kick in when the primary one shuts down. This prevents the program from resetting when the battery is replaced. A charger and an extra battery are necessary to keep the unit in continual operation. Some vendors offer solar panels that can be mounted to the unit to maintain the charge on the primary battery.

Rugged Case. The case that contains the pump, bottle, and battery must withstand substantial abuse. The case should permit the addition of a padlock to slow down a vandal and should have a carrying handle. If the case is watertight, it should have an adjustable vent. If the vent is not open, the case will pressurize and the pump will not be able to draw a sample.

Data Logger. If a monitoring device such as a flow meter or a rain gauge is connected to the sampler, you should use a sampler that collects and downloads data. One of the biggest advantages of the automatic sampler is that it can minimize your trips to the site and reduce labor costs.

Water-Level Indicator. When sampling an intermittent or ephemeral stream, make sure there is water at the sampling tube when the pumping starts. All brands offer a sensor to detect water. The sensor should be installed alongside the sample collection tube with cable ties or hose clamps. This will prevent the unit from collecting a "dry" sample. This accessory is not needed when sampling perennial streams.

When the sampler is installed and programmed, run a test sample and watch to make sure everything works correctly before leaving the sampler unattended; many little things could go wrong.

Stream sampling, in concept, is a simple matter. But in practice, it can be rather complicated. If you take the time to select suitable sampling locations and properly install sampling equipment, headaches from bad data can be "water under the bridge."

Johnny Barron is an environmental engineer with the Water and Sewer Authority in Douglasville-Douglas County, GA.

Home + About + Subscribe + News + Calendar + Glossary
Talk + Images + Advertise + Contact Us + Search + Register + Services

Erosion Control Magazine | MSW Management Magazine
Grading & Excavation Contractor | ForesterPress | Forester Media

© 2000 - 2002 FORESTER MEDIA, INC.