Waterquality sampling 831 Sampling river water

Choice of sampling site for river water may be governed by an abstraction point or a discharge point associated with an industrial user or waste water treatment works. However, it is often most useful to take water-quality samples at a river gauging station. Ideally, a single sample from the well-mixed waters downstream of a weir would suffice to give a good representation of the water quality of a small river. At a current meter station, the river should be sampled at several points across the channel and in deep rivers (over 3 m) at 0.2 and 0.8 depths. For shallower streams, one sample in the vertical at 0.6 depth should be adequate (Note: These depths correspond to the points giving the mean flow velocity in a vertical section; see Chapter 7). Once the flow characteristics of the river are known, the sampling scheme in Fig. 8.1 can be recommended.

The timing and frequency of sampling also need consideration, particularly if there is a regular pattern of flow control or of effluent discharge from industries above the sampling point. It is often worthwhile having a concentrated sampling period when the river regime is steady to establish a regular norm. Then anomalous conditions of flood flows with their increased load of suspended solids, or of unusual influxes of pollutants from accidental spillages, when sampled, can be related in perspective to average water-quality values. Seasonal changes must be identified in any water-quality variations. In addition to the establishment of the average water-quality characteristics of a river, it is important for regulatory agencies, in the interests of environmental conservation, to be vigilant at all times in maintaining satisfactory river water-quality values. Hence there has been a rapid development of automated monitoring of water quality (Section 8.5).

For taking single samples of river water, standardised instruments have been devised. The displacement sampler is recommended for the collecting samples for DO analysis,


Fig. 8.1 Selected sampling points for a deep wide stream. (Adapted from Nemerow, N. L. (1974) Scientific Stream Pollution Analysis, Hemisphere Publishing Corporation.)


Fig. 8.1 Selected sampling points for a deep wide stream. (Adapted from Nemerow, N. L. (1974) Scientific Stream Pollution Analysis, Hemisphere Publishing Corporation.)

but can be used for general sampling in rivers. The inlet is opened when the container is at the required depth and the water is fed into the bottom of a bottle. When the whole container is full, the water flows from the exit, and should continue until the bottle contents have been changed several times before the sampler is removed.

Numerous sampler designs have been adopted in conjunction with studies of suspended solids. The US DH48 is one such sampler and is shown in Fig. 8.2. The container holds a glass or polythene bottle.

When obtaining samples for chemical analysis, great care must be taken against contaminating the water sample; all containers, even the simple bucket dipping into a turbulent well-mixed stream, must be washed out with the flowing river water before being used for a sample. Sample bottles should be sealed with zero headspace, so that CO2, volatile organic compounds (VOCs), etc. are not lost from the water sample. The temperature of the water must be taken at the time of sampling. Sampling bottles must be carried in suitable crates and delivered to the laboratory the same day. Delay in carrying out the analyses can result in spurious values, since some of the chemical properties of the water can be altered by the changing conditions in storage. Indeed, measurement of most water characteristics requires special preservation of samples (Table 8.6).

8.3.2 Sampling subsurface water

Subsurface water can be extracted from both saturated ground and unsaturated ground. To extract water from unsaturated ground, suction lysimeters (also called

Fig. 8.2 A US DH48 suspended sediment sampler for attachment to a wading rod. (Reproduced by permission of the Ricky Hydrological Company, Columbus, OH, USA.)

Table 8.6 Water sample containers, methods of preservation and minimum sample volumes. (Adapted from Artiola, 2004)




Minimum volume (cm3)

pH, EC, alkalinity, (major anions)* Metal cations

(except Hg, CrVI) Pesticides, phenols

High-density polyethylene bottle High-density polyethylene bottle Clear/amber glass bottle

Add nitric acid (pH<2)

Add reducing agent (sodium thiosuphate).** Keep cool at 4 C. Adjust pH<2

100 (200)* 200 >1000

Volatile organics

Clear glass vial

Add reducing agent (sodium thiosuphate).** Keep cool at 4 C. Adjust pH<2

5-25 per group


Plastic bottle

Add reducing agent.** Keep cool at 4° C

vacuum samplers) can be used. These devices consist of a porous cup attached to a polyvinyl chloride (PVC) sample accumulation chamber and two tubes that extend to the ground surface (Fig. 8.3). The porous cups are typically made of ceramic, stainless steel or polytetrafluoroethylene (PTFE), and are best installed within an auger-hole with silica flour packed around the cup. The shorter of the two tubes entering the lysimeter (vacuum tube) is used apply a vacuum to the accumulation chamber, with a manual or electric pump. Hours to days of applied vacuum may be necessary to extract sufficient volume of sample into the accumulation chamber. The second tube within the lysimeter (sample tube) extends down to the porous cup, and is used for extracting the sample. A vacuum pump can be used to extract the sample from the sample tube into a sample bottle. Alternatively, a positive pressure can be applied to the vacuum tube to push the sample out through the sample tube. If this alternative system is used, a non-return value needs to be incorporated within the lysimeter to prevent the sample from being pushed through the porous tip.

If subsurface water is to be analysed for trace metals, stainless steel augers coated with nickel, cadmium or zinc metal to reduce oxidation should not be used to install the suction lysimeters.

Where ground is variably saturated, pan lysimeters (also called free-drainage samplers) and throughflow troughs can be used to collect water from the soil or underlying regolith. Care needs to be exercised when interpreting data from these devices, as (artificial) saturation needs to develop at the sampling point for drainage to occur.

For saturated ground, water samples can be extracted from piezometers or open wells using various devices. The first consideration is the installation of the sampling point. Traditional piezometers can be drilled, but consideration needs to minimise the effects of drilling fluids, redistribution of subsurface materials along the drilled

Pressure-vacuum pump, with vacuum dial gauge

Sample tube

Pressure-vacuum pump, with vacuum dial gauge

Sample tube

Silica flour

Backfill material

Bentonite seal

Fig. 8.3 A suction lysimeter. (Reproduced by permission of Nielsen, D. M. (2006) Practical Handbook of Environmental Characterisation. Taylor & Francis, Boca Raton.)

Pressure-vacuum soil water sampler

Porous cup

Backfill material

Bentonite seal

Silica flour

Fig. 8.3 A suction lysimeter. (Reproduced by permission of Nielsen, D. M. (2006) Practical Handbook of Environmental Characterisation. Taylor & Francis, Boca Raton.)

hole, and the effect of opening a deep subsurface system to atmospheric air. An

alternative is to use direct-push sampling devices (e.g. HydroPunch ), which involve fewer disturbances compared to conventional drilling methods (McCall et al., 2006). Consideration also needs to be given to the piezometer construction materials. Within low pH value environments, steel and stainless steel can corrode to add Fe, Mn, Cu, Pb, Cd, Ni, Cr and Mo to subsurface waters. PVC and PVC-cementing agents also degrade. The filter pack around the piezometer screen similarly needs to be comprised of only inert materials, e.g. clean, well-rounded silica sand.

Physico-chemical changes can occur to water samples as a result of the purging and extraction methods. Pressure changes during extraction typically cause CO2 degassing, which increases sample pH (by between 0.5 and 1.0 pH units), which may then cause trace metals to precipitate. This is particularly the case with suction-lift pumps, such as peristaltic pumps (Nielsen and Nielsen, 2006). Some sampling devices also cause temperature increases in samples that can change pH, redox state and the precipitation of carbonates. This can be a particular issue for electric submersible centrifugal pumps.

Some sampling devices disturb the water column or mobilise fine material from the filter pack or formation; bailing devices and high-flow pumps are particularly prone to these effects.

Analysis of published studies by Nielsen and Nielsen (2006) indicates that water contained within the screened section of piezometers is constantly flushed with natural subsurface water, and poorly connected with the stagnant water higher up within the piezometer. As a consequence, they recommend minimising disturbance of the stagnant water (during purging or sampling) and sampling directly from the screened section. In summary, they recommend the following sampling approaches to minimise changes to water samples:

1 Using low flow rates during purging and sampling.

2 Placement of pump intake with the screened section of the piezometer.

3 Minimising disturbance to the stagnant water column above the screen.

4 Monitoring water-quality indicator characteristics (e.g. EC) during purging.

5 Minimisation of atmospheric contact with samples.

6 Collection of unfiltered samples for metals analysis.

They also provide a full analysis of the operation and disadvantages of the different types of sampler available from grab samplers (i.e. bailers, thief samplers and syringe samplers) to suction-lift devices, electric submersible centrifugal pumps, positive-displacement pumps (e.g. a piston pump; Fig. 8.4) and inertial-lift pumps.

0 0

Post a comment