Part Two: The Sampling Protocol
This section describes the procedural steps of the protocol and the rationale behind these steps. Each step is cross-referenced to the accompanying field guide (Appendix 1).
1 Pre-sampling
The following pre-sampling steps are recommended because proper preparation will facilitate the sampling programme and ensure the collection of meaningful data. These steps would typically be completed before leaving the office, either the day before or the morning of the sampling event.
Sampling in dry conditions is not required for compliance with this protocol, but it is recommended for the comfort and safety of sampling staff, and to prevent contamination of the sample from rain and/or windborne particles. It is therefore advisable to check the weather forecast to decide whether or not the sampling event should take place. It is the sampler's responsibility to exercise professional judgement as to whether or not sampling in poor weather conditions will compromise sample quality or personal safety.
1.1 Check site details
Certain information should be compiled in order to provide a context for interpretation of groundwater quality data from any monitoring site. This information should include, for example, the name, location and depth of each well that is to be sampled (see Appendix 3). Although not mandatory for compliance with this protocol, the sampler should check that the relevant information for each site is available, complete and up to date (Hughes 2000; Standards New Zealand 1998a). Any data or information gaps should be addressed prior to sampling.
1.2 Gather equipment
a General
See Appendix 4 for a list of equipment that is recommended for groundwater sampling (Crowcroft and Scoble 1997; Stansfield 2001). Note that individual sample sites will have different equipment requirements. The sampler must ensure all of the equipment that will be required for the sites to be sampled is available and in good working order.
b Bottles
Bottles must be new or pre-cleaned. Specifications for bottle materials and cleaning methods for particular types of samples are given in Step 4 and Table 1. Instructions for cleaning bottles according to particular methods (eg, acid washing, detergent washing, baking, sterilising) are provided in Standards New Zealand (1998a).
A preservative is required for some types of samples (eg, an acid preservative for samples to be analysed for cations). It is recommended that the preservative is added directly to the sample bottles (either by the laboratory or by the sampler) prior to the sampling event. In this case, the sampler must confirm and record that the preservative has in fact been added to all bottles where it is required. Alternatively, if preservative is not added to the bottle before sampling, it is acceptable for the sampler to add the preservative to the bottle in the field after the sample has been collected (see Step 4).
c Chilling samples
Some standard laboratory methods for certain analytes (eg, nutrients) require samples to be chilled to below 4oC immediately after collection and for the duration of their transport to the laboratory (see Step 4). Compliance with this protocol does not require that chilling is accomplished in any particular way, but the sampler must ensure samples are chilled according to the laboratory's requirements. One commonly used approach that has been shown to chill samples effectively (Daughney et al 2006) is to use a chilly bin (ca 20-40 litres) which is packed with at least five frozen chemical ice packs ("slicka" pads) or at least 3 kg of ice (one typical service station size bag) (see Step 4). If this approach is to be used, the sampler must ensure chemical ice packs are pre-frozen or that ice is available, and then transfer the required quantity into the chilly bin immediately prior to the sampling event.
1.3 Calibrate field meters
This protocol requires field measurement of temperature, conductivity and pH (see Step 3). Calibration is the act of adjusting a meter's settings so that it displays the proper reading when sensors are immersed in solutions with known values of the pH, conductivity or temperature (or other parameter of interest). Standards are the solutions with known values of pH, conductivity or temperature (or other parameter of interest) that are used for calibration.
Compliance with this protocol requires the sampler to ensure the temperature, conductivity and pH meters are maintained and calibrated according to the manufacturer's instructions. The sampler must also check that temperature correction functions for conductivity and pH are working properly, as described below. These steps are usually most easily accomplished before leaving the office on the day of sampling.
General guidelines for calibration of temperature, conductivity and pH are summarised in Table 2. All meters should be calibrated as often as recommended by the manufacturer, or at the minimum intervals given below. Standard solutions used for calibration should cover the range expected for the samples to be measured, and should be within the manufacturer's recommended shelf life. Note that probes and sensing devices must be allowed sufficient time to equilibrate with standard solutions during calibration.
a Temperature
The temperature meter must be calibrated as often as recommended by the manufacturer or at least once per year (Rosen et al 1999).
For many modern temperature meters, adjustments of the calibration can only be performed by the manufacturer. In this case, it is the sampler's responsibility to ensure the meter is sent to the manufacture as required for regular servicing.
Alternatively, the temperature calibration of some meters can be set by the user. In this case, temperature calibration can be performed using a constant temperature bath and a reference thermometer (Rosen et al 1999; Radtke et al 2004). The temperature calibration should include measurement of at least three different standards that cover the range of temperatures expected for the groundwater to be sampled. Standards of in the range 5-25°C are appropriate for most SOE groundwater monitoring sites in New Zealand. [The median of temperature measurements made in the New Zealand National Groundwater Monitoring Programme since 1990 is 14°C, with 90 percent of measurements between 11 and 19°C (Daughney and Reeves, 2003). The median of conductivity measurements is 200 µS/cm at 25°C, with 90 percent of measurements between 50 and 750 µS/cm at 25°C. The median of field pH measurements is 6.8, with 90 percent of measurements between pH 6.0 and 8.0.]
The accuracy of the temperature calibration should be checked at least once every six months. This should be accomplished by measuring at least one solution with a known temperature that is within the range of the calibration. The calibration is acceptable if the measured temperature is within ±0.2oC of the true (known) temperature (Radtke et al 2004). If the measured temperature is not within ±0.2oC of the true (known) temperature, the temperature probe should be recalibrated or replaced.
b Conductivity
The conductivity meter must be calibrated as often as recommended by the manufacturer or at least at the beginning of each sampling day (Rosen et al 1999; Radtke et al 2005).
For most modern conductivity meters, the calibration must be set by the sampler using at least one standard solution. Standard solutions can be purchased from the meter manufacturer or prepared as described by Rosen et al (1999). Standard solutions in the range 50-750 µS/cm at 25°C are suitable for most SOE sites in New Zealand.1 Standard solutions should be kept at room temperature to facilitate the check of the temperature compensation function (see Step 1.3.d).
The accuracy of the conductivity calibration must be assessed at each sampling location before samples are collected (see Step 2.3).
c pH
The pH meter must be calibrated as often as recommended by the manufacturer or at least at the beginning of each sampling day (Rosen et al 1999; Radtke et al 2003).
For most modern conductivity meters, the calibration must be set by the sampler using at least two standard solutions, one standard in the range of pH 4 to 7, and one standard in the range of pH 7 to 10.1 Standard solutions can be purchased from the meter manufacturer. Rosen et al (1999) recommend that the calibration always begin with the standard closest to pH 7, because most groundwaters are closer to pH 7 than to pH 4 or 10. Standard solutions should be kept at room temperature to facilitate the check of the temperature compensation function (see Step 1.3.d).
The accuracy of the pH calibration must be assessed at each sampling location before samples are collected (see Step 2.3).
Figure 1: Calibrating a field meter using a pH 7 standard solution
d Temperature compensation functions for conductivity and pH
If the conductivity or pH meter has an automatic temperature compensation function, compliance with this protocol requires the sampler to confirm that it is functioning properly. Calibrate the pH and conductivity meters as described above, using room temperature standards. Check the temperature compensation function by measuring at least one conductivity standard solution and one pH standard solution which have been kept in the refrigerator overnight. The temperature compensation functions are working effectively if the chilled samples read within ±6% of the expected values (Daughney et al 2006).
If the conductivity or pH meter lacks a temperature correction function, or if the temperature correction is not working correctly, then the pH and conductivity meters must be calibrated at each site prior to sampling, using standard solutions at the expected groundwater temperature.
e Calibration for other parameters
Field measurements of parameters other than temperature, conductivity and pH are not required for compliance with this protocol. However, if such measurements are to be made, the relevant meters must be calibrated at appropriate intervals according to the manufacturers' instructions (see Rosen et al 1999; Wilde and Radtke, chapter sections variously dated). Note that the accuracy of the calibrations for some parameters should be assessed at each sampling location before the collection of samples (see Step 2.3).
2 On-site preparation
The following steps must be completed at each sampling location prior to the collection of any samples, so that the groundwater quality data can be meaningfully interpreted within the context of an SOE programme. It is the sampler's responsibility to ensure all field work is conducted safely and with appropriate etiquette (see Rosen et al 1999; Hughes 2000; Stansfield 2001).
2.1 Confirm you have the correct sampling site
The sampling site is the well, bore or other location from which the samples will be collected. For compliance with this protocol, the sampler must ensure the correct site is being sampled. This can be accomplished by:
- having made a previous visit to the same site
- confirming a grid reference with a GPS unit and ensuring a match between the site to be sampled and a photograph or a written description, or
- confirming the site name or identification number by reference to a physical label at the site, if such a label exists. Figure 2 is an example of a marked bore.
For sites that have not been sampled previously, site details listed in Appendix 3 should be recorded to facilitate repeat sampling of the same site.
Figure 2: Name plates on bores are a useful way to confirm the correct site is being sampled
Source: Hawkes Bay Regional Council, 2006
2.2 Confirm appropriate sample point
The sample point is the tap, fitting, hose or other such outlet from which water will actually be collected. Some sampling sites may have more than one sample point. In such cases, the sampler must select the most appropriate sample point, document it, and ensure samples are collected from the same sample point in the future. An appropriate sampling point is one that minimises the purging time (see Step 3) and minimises the potential for contamination or alteration of the sample (refer Figures 3 and 4). Note that it is acceptable to attach a short length of clean hose to the tap or well head, because this can assist with maintenance of laminar flow.
a Minimising the purging time
The sample point must be as far upstream as possible in the reticulation system. This minimises the volume of water which must be purged through the delivery and/or reticulation system before samples can be collected (see Step 3). This also minimises the length of the sample delivery line, thereby reducing the risk that the sample could be altered in any way, for example due to depressurisation or exposure to light, heat or air.
b Minimise the potential for contamination
Samples must be collected from a sample point which is in good condition. It is acceptable to attach a short length of clean hose (ca 2 m or less) to the tap or well head (eg, to assist with maintenance of laminar flow), but the hose outlet must not be allowed to touch the ground. Do not collect samples from a tap, fitting or hose that is corroded, leaking, or otherwise in poor condition. If there is any indication that the integrity of the site has been compromised, samples should not be collected.
c Storage tanks and pressure cylinders
Wherever possible, samples should be collected from a point that is upstream of any storage tank or pressure cylinder. If the sample point is downstream of a pressure tank or cylinder, the tank or cylinder should be isolated from the delivery line as described in Appendix 5. This will ensure samples are not tainted by standing, possibly contaminated water from the tank or cylinder. This approach is advocated by USGS (various references), Rosen et al (1999), Hughes (2000) and Stansfield (2001). Always obtain the site owner's consent before isolating or draining a pressure cylinder.
Figures 3 and 4: An appropriate sample point should be as close to the well head as possible, in good condition and upstream of any pressure cylinders wherever possible
2.3 Check calibration of field meter(s)
Compliance with this protocol requires that the accuracy of calibrations for conductivity and pH be checked at each site prior to collection of any samples (Rosen et al 1999; Wilde, chapter sections variously dated). This is because conductivity and especially pH sensors can drift and lose their calibrations over the course of a day and accurate measurements of these parameters are essential for assessing the adequacy of the purging operation (see Step 3). This protocol does not require that the accuracy of calibrations for other field parameters be assessed at each site, although this is recommended for parameters such as temperature, dissolved oxygen and oxidation-reduction potential (Wilde and Radtke, chapter sections variously dated).
Calibration accuracy must be assessed by measuring at least one standard solution with a known value of the parameter of interest. The following criteria indicate acceptable calibrations:
- measured (predicted) conductivity is within ± 6% of the true (known) conductivity (Daughney et al 2006)
- measured (predicted) pH is within ±3% or 0.1 pH units of the true (known) pH (Radtke et al 2003; Daughney et al 2006).
Note that field measurements of conductivity and pH must be properly temperature compensated. This can be achieved by using a meter with an automatic temperature-compensation function which has been shown to be functioning correctly (see Step 1.3.d). If the meter does not have a temperature compensation function or if it is not functioning properly, then the calibration checks must be performed using standard solutions kept at ambient groundwater temperature. In this case, record the measurement temperature so that the corresponding conductivity and pH at 25°C can be determined (see Radtke et al 2005).
If the conductivity or pH calibrations do not meet with the criteria specified above, on-site recalibration is required. If on-site recalibration cannot be achieved then it is not possible to assess the adequacy of purging (see Step 3), and thus any samples collected from the site will not comply with this protocol.
2.4 Clean sampling equipment
The sampler must exercise professional judgement to ensure all sampling equipment is sufficiently clean before sample collection. This is required to avoid sample contamination and cross-contamination between sites.
At a minimum, all equipment should be rinsed thoroughly with distilled water after sampling at each site, and, if dirtied during storage or transport, again before sampling at the next site (Rosen et al 1999; Hughes 2000; Stansfield 2001; Wilde 2004). This includes pumps, pump tubing, dip probes, water level tapes and any other equipment that might contact the sample water. Basins, brushes and other materials used for cleaning should themselves not be prone to leach the analytes of interest into cleaning solutions. Note that more rigorous equipment decontamination is required if the site is severely contaminated (see Wilde 2004), but such contamination is not normally encountered in SOE monitoring in New Zealand.
3 Purging
Purging is the removal of standing water from a well and its replacement with fresh formation water. Purging is essential because standing water in the well may not be chemically representative of groundwater in the aquifer some distance away from the well (Rosen et al 1999; Wilde et al 1999; Hughes 2000, Daughney et al 2006). The steps presented below are relevant to most SOE monitoring sites in New Zealand and must be followed unless:
- the sample collection interval is sealed with packers
- drawdown occurs rapidly but recovery to approximately 90 percent cannot be achieved before samples are collected, or
- a purge minimisation device or low-flow purging technique is used (in this event, modify the purging protocol as described by Wilde et al 1999).
3.1 Measure depth to water
A measurement of the depth to water in the well under ambient (non-pumping) conditions is useful for calculating the volume of water to be purged (see Step 3.2) and for general interpretation of water quality data from the site. However, at some SOE sites in New Zealand, access to the well is poor and thus it is not possible to measure depth to water. Additionally, if the pump is running, measurement of depth to water may not be representative of ambient conditions.
For non-artesian conditions, depth to water should be measured using a dip tape, according to the manufacturers' instructions (see also Rosen et al 1999, Figure 5). Record also a description of the datum from which the measurement is made, for example ground level or the top of the well casing. The same measuring point should be used each time the well is measured.
Figure 5: Water level in non-artesian conditions can be measured using a dip tape and the depth recorded relative to a known point on the well
Take care to ensure the tape does not become snared on any pump or equipment in the well. Although not required for compliance with this protocol, it is possible to measure artesian groundwater pressure with a pressure gauge or a transparent pipe as described by Rosen et al (1999, Figure 6).
Figure 6: Artesian head can be measured using a clear tube and a staff gauge
3.2 Calculate the volume of water to be purged
A certain minimum volume of water must be extracted from a well before samples are collected, in order to ensure the samples will be representative of the in situ conditions within the aquifer (Daughney et al 2006). In this sampling protocol, the term purge volume is used to describe an incremental fraction of the total volume of water extracted from the bore during the purging operation. The purge volume is equal to the volume of water in the well under ambient (non-pumping) conditions:
Purge Volume = 3.14 x [well depth - depth to water] [well radius]2 x 1000
Note that:
- the purging operation requires extraction of at least three times the calculated purge volume and may require extraction of many more than three times the calculated purge volume
- if the depth to water under ambient (non-pumping) conditions cannot be determined for any reason, assume "depth to water" = 0 in the equation above
- well depth, depth to water, and well radius must be expressed in metres in order to derive the purge volume in litres. Well depth can be obtained from the drilling log or through the use of the dip tape. Well radius refers to the casing dimension and not to the dimension of the bore
- if it is not possible to determine depth to water and if the well depth is unknown, then purge volume cannot be calculated. In this case, any samples collected from the well will not comply with this protocol. However, in instances where samples are to be collected from such sites, the sampler should calculate an approximate purge volume by overestimating the likely depth of the well. An estimate of maximum likely well depth can often be obtained by examining the drilling logs of wells that are nearby.
3.3 Install portable pump if necessary
If the well is equipped with a dedicated pump, proceed to Step 3.4. Otherwise, a portable pump or bailer must be used to collect the groundwater sample, as described below.
a Suitable pumping equipment
Rosen et al (1999) and Lane et al (2003) describe different pump types that can be used for purging and groundwater sampling.
- Portable submersible pumps (positive pressure or positive displacement) are recommended for compliance with this protocol.
- Peristaltic pumps, vacuum lift pumps and bailers are not recommended, due to possible biases associated with them (Lane et al 2003). However, provided other steps in this protocol can be satisfied (eg, purging of a sufficient volume of water, stabilisation of field parameters), samples collected using these pumping systems will be compliant with this protocol.
The pump, pump tubing, and all other equipment that contacts the sample must be chemically inert and suitable for the target analytes (Lane et al 2003):
- stainless steel (all grades), if uncorroded, is suitable for the inorganic parameters relevant to this protocol. Metals other than stainless steel are not suitable, unless samples will only be analysed for CFCs, in which case refrigeration grade copper tubing is acceptable.
- glass is suitable for the inorganic parameters relevant to this protocol.
- plastics made of fluorocarbon, polyethylene, polypropylene, PVC and silicone are suitable for the parameters relevant to this protocol. For CFC sampling, either nylon or refrigeration grade copper tubing is required, but fluorocarbon tubing is not appropriate (see Step 4).
b Pump installation
The sampler must ensure the portable pump is used in accordance with the manufacturer's instructions (see also Crowcroft and Scoble 1997). The pump should be positioned so that its intake is at least one metre below static water level and a minimum distance above the top of the screened/open interval of 10 times the well diameter (for example, 1500 mm for a 150 mm well diameter) (Wilde et al 1999). This will ensure the sample is representative of the entire screened or open interval of the well. The pump line should be fitted with a non-return valve to prevent contamination of the well with residual water in the pump tubing. It is advisable to place a dip tape or other water level sensor in the well to permit measurement of the water level during purging.
3.4 Initiate pumping
Upon arrival at the site, if the well is equipped with a dedicated pump which has been running continuously and for long enough to have removed one purge volume of water (see Step 3.2), record the time and date of arrival on site and proceed to Step 3.5. Otherwise, record the time, date and water level and initiate pumping. Ensure the outflow from the pump is disposed of away from the well so that it does not pond around the well casing (Rosen et al 1999; Wilde et al 1999, Figure 7).
Figure 7: When purging, ensure the discharge is away from the well so that water does not pond around the well casing
Compliance with this protocol does not require a specific pumping rate during purging. The sampler must use their professional judgement so that the pumping rate suits the well construction, aquifer characteristics and pumping equipment. A suitable pumping rate produces a continuous stream of water from the pump outlet or sample point without turbulence, entrainment of air, or pump cavitation (refer Figure 8). As a guideline, Wilson (1995), Rosen et al (1999) and Wilde et al (1999) recommend that pumping rates during purging should be 0.1-1 litre per minute. Pumping should not be halted during purging, and the pumping rate should be kept constant throughout purging (Wilde et al 1999). If the pumping rate must be changed, it should be changed gradually, taking care not to cause turbulence.
Figure 8: A suitable pump rate will ensure laminar flow from the outlet point (left), aerated flow (right) is not appropriate
Compliance with this protocol requires determination of the pumping rate during purging. This is most easily accomplished by recording the time required to fill a container of known volume, such as a 10 litre bucket. Alternatively, if the site is fitted with an in-line flow meter, the pump rate can be determined by recording changes in the flow meter readings over time. With either approach, the measurements should be repeated on at least three separate occasions during purging in order to obtain at least three separate estimates of the pumping rate, all of which should be recorded on the sampling log.
Compliance with this protocol does not require minimisation of drawdown during purging. However, if possible, the water level in the well should be continuously monitored throughout the purging operation. Values of drawdown should be recorded on at least three separate occasions during purging, and again immediately prior to collection of samples. Ideally, purging should not cause drawdown of more than 0.3 m if the pump inlet is above the screened interval of the well, or more than 0.15 m if the pump inlet is within the screened interval. If drawdown exceeds these criteria, the pumping rate should be modified appropriately, where possible and practical and at the sampler's discretion.
3.5 Monitor field parameters during purging
Temperature, conductivity and pH must be monitored during purging in order to assess the adequacy (completeness) of the purging operation (see Step 3.6). Values of all three field parameters must be recorded at least four separate times during purging. The first measurement must be made as soon as possible after the initiation of pumping. Subsequent measurements must be made at intervals corresponding to the time required to extract at least one purge volume from the well (see Step 3.2). Thus the second, third and fourth sets of field measurements are made after extraction of about one, two, and three purge volumes, respectively.
The water does not need to be rigorously isolated from the atmosphere for the measurement of these parameters (Rosen et al 1999; Wilde et al 2000; Hughes 2000; Daughney et al 2006). Water from the pump outlet can be passed through a sealed flow cell or through a short length of clean hose and into an open container such as a 10-litre plastic bucket (refer Figure 9). In either case, the probes and the water inlet should be placed as low as possible inside the container, and water should be allowed to continuously flow out, or overflow from, the container or flow cell. The flow of water past the sensors must be continuous and laminar. The flow rate through the flow cell or open container must be slow enough to prevent introduction of air or gas bubbles, but fast enough to prevent shifts in the temperature or chemical composition of the sample.
Figure 9: Water from the pump outlet can be passed through an open container (left) or a flow cell (right)
Measurements of other field parameters, such as dissolved oxygen, redox potential or turbidity, are useful but are not required. For field measurement of other parameters, refer to Rosen et al (1999) or Wilde et al (2000) to determine if isolation from the atmosphere is critical. If isolation from the atmosphere is critical, field measurements must be made in an air-tight flow cell; if not, measurements can be made in a sealed flow cell or in an open container as described above.
3.6 Assess adequacy of purging
The following criteria for assessing adequacy of purging have been based on a variety of sampling guidelines (Crowcroft and Scoble 1997; Wilson 1995; Standards New Zealand 1998b; Rosen et al 1999; Wilde et al 1999; Hughes 2000; Stansfield 2001). Daughney et al (2006) have confirmed that these criteria are appropriate for New Zealand SOE groundwater monitoring sites.
The purging operation is complete if:
- the container in which field measurements are made and the tubing that connects it to the pump have both been rinsed with a quantity of well water that exceeds three times their volume and
- the field values of temperature, conductivity and pH have been measured on at least four separate occasions, each measurement at least one purge volume apart (see Step 3.5) and the differences between the last two measurements are the same within the following limits:
- Temperature ±0.2°C and
- Conductivity ±3% (± 5% if <100 µS/cm at 25 °C) and
- pH ±0.1 pH unit.
Continue the purging operation by making measurements after extraction of each purge volume until simultaneous stabilisation of all three field parameters is achieved. Once the criteria have been met, record the date and time that stabilisation was achieved, and record the field measurements as final values. If the criteria relevant to the site and situation cannot be met even after prolonged pumping, any samples collected from the well will not be in compliance with this protocol.
Field parameters other than the above can also be monitored to assess adequacy of purging, but are not required for compliance with this protocol. The following stabilisation criteria are recommended for dissolved oxygen and turbidity:
- Dissolved oxygen ±0.3 mg/L
- Turbidity ±10%
