Sending...

Groundwater Monitoring Technologies for Catchment Management

T.C. BAZEN MSc CPEng

KEY WORDS:

Catchment, drilling, groundwater, hydrology,management, monitoring, protocol, purging, Quality Assurance, sampling, technology, unconfined aquifers, wells.

ABSTRACT:

This paper reviews technologies utilised for groundwater monitoring programs applicable mainly to shallow, unconfined aquifers; especially the monitoring well design, drilling, pumping, purging and sampling aspects. The roles of program objectives, sampling protocol, analysis and Quality Assurance are also examined.

INTRODUCTION:

Throughout the world groundwater monitoring is an important component of catchment management. Groundwater monitoring programs broadly serve four roles:

  • examination of unexplored aquifers for research or production purposes (an exploratory role);
  • early detection of groundwater quality deterioration (monitoring role);
  • determination of the extent of pollution plumes (assessment role);
  • and verification of pollution control or decontamination measures (regulatory role);

some or all of which may be incorporated into the objectives of the program. This provides indicators of groundwater health and monitors the effects of human activity within the catchment; facilitating the formulation of appropriate catchment management policies.

This paper examines groundwater monitoring technologies applicable mainly to shallow, unconfined aquifers (see Figure 1) and includes:

  • objectives of groundwater monitoring programs;
  • monitoring wells;
  • the relevance of catchment hydrology;
  • pumping and purging for sample collection;
  • sampling protocol and collection techniques;
  • the role of analysis; and
  • Quality Assurance (QA).

QA is examined as an integral part of any monitoring program (as important as any specific technology) and the need to incorporate QA throughout every phase of a program.

Figure 1. A definition of groundwater.

OBJECTIVES OF GROUNDWATER MONITORING:

Every environmental monitoring program must have clearly defined objectives for proper planning and execution and to gauge its success; an integral part of QA. A groundwater monitoring program is governed by cost limitations and objectives often have to be revised in consultation with end-users.

A general objective of groundwater monitoring is to obtain a true indication of the physico-chemical and biological constitution of the water within the aquifer. It is important that:

  • the collected water samples are truly representative of the source being examined; and
  • the appropriate analysis is carried out using standard procedures and Quality Assurance. That is, using suitable national or international standards (such as W.H.O.).

Ongoing monitoring is to ensure that any degradation of water quality is detected at the earliest possible time. Any follow-up investigation should identify the source and establish a remedial program to clean-up the aquifer and prevent danger from the contamination.

Specific objectives are generally based on either (Keith 1990):

  • exploratory (surveillance) goals which provide preliminary information. For example, basic physico-chemical qualities (TDS, pH, Eh, DO and Temperature) of water available from an aquifer, which can be obtained from a single monitoring well – perhaps an existing production well; or
  • monitoring (assessment) goals in medium to long term programs for pollution control or clean-up enforcement. This could involve determination of plume thickness and level within an aquifer; identification of the pollution source being a prerequisite to producing a remediation program. Plumes normally disperse in the direction of groundwater flow (i.e. longitudinally), with less dispersion in the horizontal (transverse) direction or the vertical direction i.e. in thickness. Transects or arrays of monitoring wells (with multilevel sampling capability) can establish the direction of groundwater flow, the boundaries of the plume and its thickness, level and concentration.

MONITORING WELLS

A fundamental requirement for groundwater monitoring is access to the water within the aquifer through monitoring wells; this is a critical component of the monitoring program.

Location and Design

A monitoring program should permit installation of new wells; existing wells are often poorly located and are generally designed for maximum yield rather than selectivity. The number of monitoring wells required and their spatial distribution (ie in transects or arrays) will depend on the hydrologic complexity of the catchment and the objectives of the program. Changes in the placement and number of required wells often evolves in the light of collected data or variations in program objectives; thus should be a normal part of program evolution.

Design of monitoring wells requires knowledge of the site’s hydrogeology and subterranean geochemistry; their construction demands a higher level of care than normally applied to production wells. In construction, care needs to be taken to minimise cross-contamination of aquifers with loosened topsoil – particularly in areas where soil contamination is likely (Keith 1990); practically, there will always be some trace contamination of the aquifer from surface and near surface layers (Smith et al. 1988). Design criteria include:

  • borehole diameter and depth (which may need to be accurately controlled);
  • type of drilling technique;
  • length, number and placement of screened sections (often quite short lengths are required with exact placement in terms of depth)
  • slot size of screen – usually selected, to retain surrounding materials, and varies from 50% in coarse substrate materials with no gravel pack to 90% with gravel pack;
  • gravel pack specifications – use clean, uniform quartz sand with a minimum size of 3 to 5 times the 50% grain size of the stratum being monitored;
  • backfill to retain confining characteristics;
  • sealing at the surface – usually bentonite or cement seals; and
  • yield potential of the well.

The simplest, narrow diameter monitoring wells capable of accommodating the purging and sampling equipment are preferred for routine monitoring as these will least disturb the environment and minimise the purging volume required. With the general availability of small diameter (50 mm), submersible pumps capable of 30 metre lift, 50 mm (ID) monitoring wells have become a standard in monitoring well technology (Barcelona et al. 1988).

The occurrence of vertical flow between permeable strata within an aquifer system may lead to mixing of sample water from multiple zones. To limit such flow monitoring wells should be designed and constructed to either monitor one discrete stratum or have segregated screens to enable multi-level monitoring (Kent and Payne 1988).

Drilling:

Selection of an appropriate drilling technology is important in achieving the desired quality of groundwater samples. The main governing criteria in descending order of importance are (Barcelona et al. 1988):

  1. Hydrologic information
    1. type and nature of the geological formation containing the aquifer;
    2. depth of drilling required;
    3. depth of desired screen installation below the water table;
  2. Nature of anticipated types of contaminants;
  3. Location of the drilling site/s, i.e., accessibility or sensitivity of the area (which may be a conservation zone);
  4. Required design of the monitoring well ; and
  5. Cost and availability of drilling equipment.

Appendix A summarises a selection of drilling techniques commonly used in the water-well industry with their advantages and disadvantages in relation to monitoring well construction.

Hand augering, although the slowest and most labour-intensive of methods, is the least disturbing environmentally (augers range from 50 to 100 mm) and can be used in locations inaccessible to machinery. Hand augering provides the best form of stratigraphy logging with resolution of ultra-thin sedimentary lenses down to the millimetre thickness level. It is limited to approximately 7 metres depth and cannot penetrate indurated or consolidated substrates (such as ferruginised, indurated sandstone or calcrete). It is highly suited to investigations in:

  • wetlands including estuarine and riverine environments;
  • areas with shallow water tables (particularly in sand formations such as beach and delta investigations); and
  • sensitive conservation areas or areas with endemic diseases (such as Jarrah Dieback caused by a water-borne fungus – phytophthora cinnamomi Rands) where disturbance must be kept to a minimum.

Drive point installations which comprise a pointed steel driving point attached to a robust steel pipe; the whole being manually driven into the ground, the sampling system placed inside the steel pipe which is subsequently withdrawn allowing the soil to collapse onto the sampling structure. This system is suited to shallow aquifers and has been used successfully down to 8 metres in sand (Stites and Chambers 1991).

Cable-tool drilling is one of the oldest and simplest methods used in the water well industry. The penetration rate is slow compared to other methods but recovery of core samples is excellent and the equipment required is simple and relatively inexpensive.

Hollow-stem and solid-stem augering are ideal mechanical techniques for well construction due to the absence of drilling fluids and minimal disturbance of the substrate; they are limited to approximately 40 metres depth and will not penetrate consolidated rock. Hollow-stem augers are particularly valuable in that they facilitate continuous collection of geologic core samples and in non-cohesive soils (i.e. those prone to caving-in) the monitoring well can be assembled and installed through the hollow core prior to removal of the auger.

Air-rotary drilling is the other method of choice for well construction since it uses compressed air (forced down the centre of the drilling stem) to convey cuttings to the surface along the periphery of the borehole introducing minimal contaminants into the well. Being suitable for all geological formations, it is often used where hollow-stem augers cannot penetrate consolidated substrates.

Construction Materials for Wells:

Well casing and screen materials contaminate the collected sample. The materials should have minimum chemical impact on the samples and retain their structural integrity within the subterranean environment for the life of the monitoring program. Barcelona et al. (1988) rank the commonly used materials from best to worst in the following order:

  • Teflon™;
  • 316 stainless steel;
  • 304 stainless steel;
  • PVC Type 1 (rigid PVC);
  • low-carbon steel;
  • Galvanised steel;
  • Carbon Steel.

Many monitoring wells are constructed from rigid PVC; absorption of organic compounds by this PVC is low, typically1 less than 1 ng/cm2. Note however, that flexible PVC is quite different; added plasticisers (including phthalate esters) leach into water. U.S. EPA recommends exclusive use of stainless steel or polytetrafluoroethylene (PTFE or Teflon™ in preference to PVC (Kent and Payne 1988) but this does impose a significant cost penalty. Where the cheapest materials are used from cost constraints, additional tests and QA samples may be required to identify any bias introduced by those materials.


1Units for this parameter are usually given in absorption per unit of exposed area. Note that larger diameter tubes tend to have a smaller effect.


Well Development

Well development is an important phase of commissioning (or recommissioning) a monitoring well and is often overlooked in the planning stage. The drilling of a well produces a sleave of finely ground materials (fines) around the borehole which (when wet) form an occlusive mud cake around the well casing. This limits the hydraulic conductivity into the screened section and must be removed to enable effective sampling, allow an adequate flow rate into the well and provide water samples reasonably free from suspended solids (to reduce subsequent filtering).

Techniques include:

  • bailing;
  • surging; and
  • flushing with air or water.

The underlying principle of each technique is to generate repetitive and shock flow reversal through the mud cake thereby breaking it up and flushing the fines into the well for removal. The end-point is variable from clear water delivery to a consistent minimum content of suspended solids.

CATCHMENT HYDROLOGY

Monitoring changes in groundwater requires an understanding of the physical and chemical characteristics of the aquifer system. This includes the hydrologic and geochemical characteristics of the aquifer and usually involves a desk survey of all known information from previous test drilling; in the absence of such information additional test drilling will be required. Ideally the knowledge will include (Kent and Payne 1988):

  1. detailed knowledge of the aquifer’s stratigraphy –
    1. which can be gained from field core samples by quantitative analysis (stratum thickness, grain size and distribution, sand and mud fractions) and qualitative analysis (quartz sand, clay, laterite); and
    2. is essential for the correct interpretation of sample analytical data and determination of sampling depths.
  2. the flow behaviour of monitored chemical species –
    1. this may need to be determined via column breakthrough experiments in the laboratory using soil collected during the test drilling;
    2. to enable selection of an appropriate monitoring well design, sample collection and handling procedures.
  3. the direction/s, velocity and temporal variation of groundwater flow –
    1. to anticipate potential plume movement; or
    2. predict movement of an existing plume and plan remedial action.
  4. the hydraulic conductivity of the strata –
    1. is best determined dynamically, in the field by either pumping (draw-down or recovery) tests or slug tests; or
    2. hydraulic conductivity may be determined from field core samples in laboratory permeameters; this method does not allow for large scale features and the soil does not have its original ‘undisturbed’ compaction and consistency.
    3. A few long term, large scale pumping tests are probably more informative than many simple monitoring bores.
  5. the spatial and temporal variations in groundwater quality –
    1. a long term activity that ideally has preceded the need for the monitoring program – some data may be available from other groundwater users in the area.
    2. an uncontaminated area around but particularly upgradient of the study area needs to be monitored – this can run concurrently with the monitoring program.
All this information should be accurately documented on well-logs (also known as borehole logs), topographic maps and cross-sectional diagrams, and stored in a systematic manner with a central collection agency; such as the country’s Geological Survey or equivalent authority, for future use.

Valuable hydrologic information is gained by recording the water table depth prior to purging or sample collection which may reflect seasonal variation and other perturbations (recharge or draw-down events). Recording the level at the end of (but during) the purging cycle is also valuable for assessment of changes in hydraulic conductivity; applicable only to dormant (ie non-producing) monitoring wells or production wells that can be shut down long enough to allow recharge to fill the cone of depression. All this information should be recorded as part of the sampling log and form part of the sampling protocol.

PUMPING AND PURGING

Due to the rapid degassing and evaporation of water at reduced pressure, vacuum lift pumping is limited to 8 metres depth (Blake 1989); beyond which pump suction reduces the pressure within the well to levels comparable to the vapour pressure of the groundwater. For greater depths, down-the-well (submersible) pumping is required with sufficient head and capacity to lift the water to the surface.

Whenever water is stagnant in the well casing for extended periods of time (ie two hours or more) it has had the opportunity to react with the well casing material, exchange gasses with the atmosphere and suffer microbial activity. It is well accepted (Davis and Barber 1994) that, due to this chemical alteration and vertical cross contamination from different strata within the well, stagnant water must always be removed prior to sample collection to obtain a truly representative sample. Purging and sampling should preferably take place at the same level in the well; this is normally achievable using the same pump. Additionally, the equipment should always be lowered to the same level in the well for consistency.

Hydraulic effects during pumping (such as turbulent flow) may cause sample alteration; hence a sampling protocol should include a purging and sampling routine for each monitoring well (Kent and Payne 1988):

  • to ensure adequate purging to gain a representative sample – usually with the pump operating at high speed, typically delivering 4 to 10 litres per minute;
  • to minimise turbulent flow, destratification and sample alteration – which requires a low flow rate (around 100 ml per minute) to enable sample collection with minimum degassing (Barcelona et al. 1988); this is especially important in the monitoring of volatile organic solvents (Smith et al 1988).

The purging and sampling cycles therefore require a variable-delivery type pump or two separate pumps.

Suggested purging volumes range from:

  • three well-casing volumes (Hirschberg 1993, U.S. EPA – in Kent and Payne 1988);
  • through to three to ten volumes (Keith 1990).

Excessive purging not only wastes time but causes undue disturbance of the environment (both to the aquifer and the surface where the purge water is released). The U.S. Geological Survey recommends pumping the well until temperature, pH and conductivity are constant; i.e. they should be stable over two successive well volumes. Since pH is particularly sensitive to CO2 loss, in-line measurements provide more accurate results than grab samples (Barcelona et al. 1988). The extent of well purging will vary with the hydraulic properties of the aquifer system under investigation and may vary with time as aquifer properties are altered; thus all purging activities should be routinely documented as part of the sampling protocol.

SAMPLING

A sampling program is designed to meet specific objectives in light of the hydrologic and geochemical characteristics of the aquifer. Most importantly, a sampling protocol must be written. This is a detailed description of all the procedures to be followed in the collection, handling, packaging, preservation, transportation, storage and documentation of all samples. The sampling protocol must list all equipment and information needed for sampling (Keith 1990):

  • sampling locations and frequency;
  • purging time and rate or volume of each monitoring well;
  • the time between completion of purging and sample collection – the least consistent element of groundwater sampling; the time must be specified and used consistently (Smith et al. 1988);
  • sampling rate and technique;
  • types, numbers and sizes of containers;
  • labels;
  • field logs;
  • types of sampling devices;
  • numbers and types of blanks;
  • sample splits and spikes;
  • volume of each sample required;
  • specifics of compositing;
  • preservation instructions;
  • chain-of-custody procedures;
  • transportation and refrigeration/cooling plans;
  • field sample preparations (filtering, pH adjustment, etc.);
  • direct field sample parameter measurements (pH, DO, ammonia, temperature, etc.);
  • report format;
  • variables to be recorded at time of sampling such as hydrological, meteorological and physical parameters (depth to water table, air temperature, water temperature, etc.);
  • analytical methods to be used;
  • analytes to be monitored;
  • reference material; and
  • training required for sampling personnel.

A properly designed sampling protocol (and strict adherence to it) is the best guarantee of obtaining samples that accurately represent the aquifer. Expensive ‘high technology’ sampling and monitoring equipment are no substitute for a properly trained, dedicated team of sampling personnel who meticulously follow a proven sampling protocol (Barcelona et al. 1988).

“Groundwater is a very complex matrix” (Smith et al. 1988) but despite its variability in quality, it has some unique properties:

  • groundwater movement through an aquifer generally precludes movement of particulates – thus only substances in solution are mobile within the aquifer;
  • groundwater often has a high level of dissolved carbon dioxide enabling dissolution of carbonates as bicarbonates – any outgassing of this carbon dioxide causes precipitation of those carbonates; and
  • groundwater is virtually devoid of oxygen – dissolved ions tend to be in their most reduced state (groundwater with a high iron content is indicative of ferrous ions at near neutral pH – aeration of the sample will precipitate these ions as ferric hydroxide or oxy-hydroxide species)

Such groundwater cannot be aerobic. Hence a representative groundwater sample usually should not contain particulate matter and should be protected from air (as far as possible) to prevent oxidation. Pressure filtration (which avoids degassing) under a nitrogen blanket through a 0.45 (m polyvinylidene fluoride or polytetrafluoroethylene (Teflon™ medium is recommended (Smith et al. 1988).

Common Sampling Well Installations:

Six types of installation commonly used for groundwater monitoring are (Cherry et al. 1983):

  • water-table standpipes – generally known as standpipes;
  • standpipe piezometers – generally known as piezometers;
  • auger-head samplers;
  • multilevel point samplers – vacuum or suction;
  • multilevel point samplers – pressure or positive displacement; and
  • bundle piezometers.

It is recommended that all pipes are fitted with an end-cap at the bottom to prevent sediment from filling the pipe under the hydraulic action of the groundwater seeping into the pipe. Similarly, all slotted, perforated or screened sections of pipes (including single hole sampling apertures) should be covered with fibreglass cloth, nylon screen or stainless steel screen; this can be omitted where screened sections are in coarse formations or in a gravel pack.

Water-table standpipes (Standpipes) usually comprise PVC pipes with slots or perforations along the lowest 3 to 6 metres of pipe. Standpipes are useful in the preliminary investigative stages to establish both water table depth and fluctuation; particularly valuable for unexplored sites. Design of the monitoring transects can proceed based on this information. Average or typical water quality can be monitored via standpipes (eg. where production wells are to be established) but special sampling techniques are required to resolve vertical contaminant concentration profiles where thin or stratified plumes are involved. Figure 2 depicts a typical standpipe installation.

Figure 2. Sketch of typical Standpipe and Piezometer installations (after Cherry et al. 1983)

Standpipe piezometers (piezometers) usually comprise PVC pipes with slots or perforations along the lowest 0.3 to 0.6 metres of pipe. Piezometers are useful throughout all investigative stages to:

  • hydraulic heads at different depths within the aquifer;
  • measure hydraulic conductivity from rising or falling head tests; and provide samples for chemical analysis.

Piezometers may be installed singly or nested together in the same borehole reducing drilling costs and environmental disturbance; Figure 3 depicts such an installation.

Figure 3. Individual wells and nested piezometers. (a) Separate short-screened wells sunk to the relevant sampling depth. (b) Three piezometers nested into the one borehole – reduces drilling costs. (after Lerner and Teutsch 1995)

Auger-head suction samplers were designed during the investigations into the Borden aquifer in Canada (Cherry et al. 1983) because the standpipes and piezometers in-situ could not resolve the vertical distribution of contaminants. The design criteria for the auger-head sampler were:

  • to provide rapid in-field resolution of the vertical contamination profile;
  • to have a continuously variable vertical sampling interval during measurements; and
  • the device had to be adaptable to standard drilling equipment.

The drilling technique obviously could not use drilling fluids thus the hollow-stem auger method was selected. As depicted in Figure 4 the device comprises a porous, cone-shaped brass point which is fixed to the leading end of the auger and is protected by the cutting head. At the desired depth a water sample is drawn (by vacuum) up to the surface via the ‘TYGON™’ tubing; this limits its depth of operation to 8 metres and precludes some types of sampling such as for Volatile Organic Compounds (VOCs) or where outgassing may change the sample chemistry.


Figure 4. Auger-head suction sampler for use with continuous-flight hollow-stem augers (after Cherry et al. 1983).

Multilevel point samplers – vacuum or suction samplers were developed in 1976 by Pickens et al. (1978) for testing dispersion in non-cohesive sand or gravel aquifers. Figure 5 depicts this type of sampler which comprises a series of single sampling ports positioned at the required level in the casing wall; each port has its own sampling tube. Up to 30 sampling points can normally be accommodated within a well – mainly restricted by the ID of the casing and sampling tube diameter. The use of suction limits this system to 8 metre depth.


Figure 5. A sketch of a suction-type multilevel point-sampler (after Pickens et al.1978)

Multilevel point samplers – pressure or positive displacement samplers are similar to the suction type sampler except that some form of pump or positive displacement device is provided down-the-well (see Figure 6); a consequence is each sampling port now requires two tubes – one for the sample flow and one to drive the pumping device. Further discussion of these devices follows in the next section – Sampling Devices.


Figure 6. A schematic diagram of one type of positive-displacement multilevel point-sampler (after Gillham and Johnson 1981)

Bundle piezometers as depicted in Figure 7 comprise a number of small diameter piezometers installed around the outside of a central pipe; each piezometer has a small screened length placed at the required depth. These devices are inexpensive, can be assembled on-site and installed through the centre of hollow-stem augers. Suction is normally used for sampling due to the small piezometer diameter; automatically limiting the operational depth to 8 metres. Greater operating depths can be achieved through use of narrow tube bailers or larger size piezometers and submersible pumps.


Figure 7. Sketch of a bundle-piezometer (after Cherry et al. 1983)

Sampling Devices:

Sampling devices in common use include:

  • suction-lift pumps;
  • electrical submersible pumps;
  • positive displacement pumps;
  • bailers; and
  • multi-level samplers.

Suction-lift pumps are usually electrically or engine driven centrifugal pumps and operated at ground (surface) level for convenience and ease of maintenance. These need ‘priming’ i.e. they require a continuous column of water from the sampling level up into the pump body – provision of water for this purpose may contaminate the aquifer being examined. Note, suction-lift pumps employ a strong negative (gauge) pressure that can cause degassing of the sample

Electrical submersible pumps are operated down-the-well at the sampling level; generally self-priming and require connection to an electrical power source. Both portable and fixed installations of submersible pumps are used for sampling; Figure 8 depicts a typical slim format submersible pump.


Figure 8. A typical helical rotor, electric submersible pump (after Neilsen and Yeates 1985)

Positive-displacement pumps (such as bladder pumps or gas-driven samplers) are also operated down-the-well at the sampling level and require a source of compressed gas for operation; typically air or nitrogen. Figures 9 to 11 depict typical gas-lift equipment. It is noteworthy that some gas-displacement pumps can cause gas stripping of both carbon dioxide (which changes the sample pH) and Voces (Kent and Payne 1988).

Figure 9. Schematic diagram of a simple slotted well point gas-drive sampling device (after Neilsen and Yeates 1985)
Figure 10. A gas-driven sampler designed for permanent installation in a monitoring well (Barcad Systems) (after Neilsen and Yeates 1985)
Figure 11. A cut-away diagram of a gas-operated bladder pump (after Neilsen and Yeates 1985)

Bailers, comprise various ‘bucket’ type devices that are lowered into the well, fill with water at the required level and are brought to the surface for sample collection; as depicted in Figures 12 and 13. They are commonly used for small diameter, shallow wells that cannot accommodate a submersible pump and are suitable for both purging and sampling. Bailers are inexpensive, portable, easy to operate and maintain, however their disadvantages include (Kent and Payne 1988):

  • mixing or contamination between samples due to inadequate cleaning after each sampling operation;
  • aeration or degassing of samples during collection and transfer to sample container.
Figure 12. Application of depth samplers: A) bailer is lowered to sampling depth and recovered; B) Alternative valving arrangements for bailers (after Lerner and Teutsch 1995)
Figure 13. Sketch of a simple bailer (after Neilsen and Yeates 1985)

Multi-level samplers are only limited by human ingenuity; they include the bailers and nested wells (previously covered). The following are some additional specialised techniques in current use.

Packer systems are an active non-permanent technique for sampling wells with long screened sections. The system is lowered to the desired depth, the packers are inflated to hydraulically isolate the sampling pump within the casing and (after purging) a sample is collected (see Figure 14). The packers can then be deflated and repositioned to another level or withdrawn at the end of sampling.

Figure 14. Schematic diagram of a double packer system; GWL = groundwater level (after Lerner and Teutsch 1995)

A variation of the packer system is shown in Figure 15. Here multiple sampling ports are created by the required number of packers and a travelling sampler (the sonde) is used to obtain the respective samples. Although removable, this type of system is more of a permanent installation.

 

Figure 15. Dedicated multi-level system for a single well, with packers separating the sampling intervals (after Lerner and Teutsch 1995)

Separation pumping is a three-pump technique where the two main pumps (located at the top and bottom) create a water flow divide or stagnation point at a given depth within the well (i.e. a point with no upward or downward flow). The depth of this point is adjustable by varying the pumping rates of the two main pumps (see Figure 16). The precise height of the divide can be determined by a heat-pulse flow detector and the sampling pump is placed at this point. The essential components in this system are the main pumps which must have accurately variable pump rates. The system is portable and only limited by casing diameter (i.e. must be sufficiently large to accommodate the pumps).

Figure 16. Separation pumping, with sampling pump positioned at the water divide between flow to the top and bottom pumps (after Lerner and Teutsch 1995)

Baffle systems use a packer with a penetrating inner tube (baffle) and a main pump above (see Figure 17). The principle is to achieve perfectly radial, horizontal flow into the well around the baffle from which a sample can be collected; sample pumping rates are kept low to avoid disturbing this horizontal flow . Repositioning the packer and baffle enables sampling at different levels. The system is suitable for portable applications.

Figure 17. Baffle system, with main pump inducing inflow to well, and sampling pump collecting water from the anulus between the casing and the baffle (after Lerner and Teutsch 1995)

Multi-port sock samplers are elongated packer elements (socks) of elastomeric material that are inserted into the well and inflated with either gas or water to hydraulically isolate the sampling ports (see Figure 18). Each port normally has its own sampling pump and tubing (all contained within the sock) hence requires a relatively large well (more than 80 mm – Lerner and Teutsch 1995). Sock samplers can be used in both cased wells and uncased boreholes, and may be used as a permanent or portable monitoring system.

Figure 18. Concept of a multiport sock sampler (after Lerner and Teutsch 1995)

Procedures and Problems of Sampling:

Experience has shown that the major source of error in the measurement of groundwater quality is due to the variability in the sampling of groundwater (Smith et al. 1988); this underscores the importance of the sampling protocol and its strict observance. Sampling should always proceed from upgradient wells (which are normally least contaminated and most closely reflect background water quality) down into the catchment or contaminated area; this minimises cross-contamination. Between sampling points, common use equipment needs to be washed using laboratory (i.e. phosphate-free) detergent and distilled (as opposed to de-ionised) water followed by an acid rinse (0.1 N HCl) or solvent rinse (hexane or methanol) followed by a triple rinse in distilled water.

Following the purge cycle, the first samples to be collected should be for the volatile constituents, TOC, TOX, and those requiring field filtration or field measurement. Then the large volume samples for extractable organic compounds, total metals and nutrient anion determinations should be collected, treated and stored for dispatch to the laboratory (Barcelona et al. 1988).

It is important to establish the sample volume to be collected; unfortunately there is no simple guide to determine that sample. A minimum volume can be established from the minimum volume required for chemical analysis and the number of replicates desired; Lerner and Teutsch (1995) suggest one litre as a suitable standard appropriate to most investigations.

Absorption of air emissions into samples may be a problem in the vicinity of certain industries (eg petrochemical or food processing plants). Under these circumstances a number of field blanks should be collected to establish the background absorption in that area. Such blanks consist of distilled water, passed between two glass sample bottles six times at the sampling location immediately prior to the actual sample collection (Kent and Payne 1988).

Construction Materials for Sampling Devices:

The greatest risk to sample integrity lies in sorption of target analytes by flexible components such as tubing, pump bladders, gaskets and seals. A summary of recommendations is provided in the Table 1 below; comments on the ‘hard’ materials from the construction materials for wells section are equally applicable here.

Table 1. Recommendations for Flexible Materials in Sampling Applications (after Barcelona et al. 1988)
Material Recommendations
Polytetrafluorethylene
(Teflon™)
Recommended for most monitoring work, particularly for detailed organic analytical schemes. The material least likely to introduce significant bias or imprecision. The easiest material to clean in order to prevent cross-contamination.
Polypropylene
Polythene (linear)
Strongly recommended for corrosive high dissolved solids solutions. Less likely to introduce significant bias into analytical results than polymer formulations (PVC) or other flexible materials with the exception of Teflon™
PVC (flexible) Not recommended for detailed organic analytical schemes. Plasticisers and stabilisers make up a sizeable percentage of the material by weight as long as it remains flexible. Documented interferences are likely with several priority pollutant classes.
Viton™
Silicone
(medical grade only)
Neoprene
Flexible elastomeric materials for gaskets, O-rings, bladder, and tubing applications. Performance expected to be a function of exposure type and the order of chemical resistance is as shown. Recommended only where a more suitable material is not available for the specific use. Actual controlled exposure trials may be useful in assessing the potential for analytical bias.

™Registered Trademark of DuPont.

Sample Preservation

To prevent samples changing physically, chemically or biologically during transport and storage they are generally refrigerated or preserved by the addition of acid or alkaline solutions. Common problems are (Kent and Payne 1988):

  • filtering and preservation of samples that require laboratory analysis;
  • aeration of the sample during collection and/or transport;
  • failure of filtering samples prior to the addition of acid for preservation (this releases cations from suspended clay particles); U.S. EPA recommends acidification to pH less than 2 for metals analysis of samples; and
  • lack of or inadequate temperature reduction for proper sample stabilisation during transport and storage. U.S. EPA recommends 4(C immediately after collection and during shipment.
  • freezing of completely filled containers resulting in burst containers or lids being forced off the container; a hazard where sample identification is only written on the lids and these are mixed up during transport.

ANALYSIS

Sample analysis is a critical part of groundwater monitoring – it is expensive and thus needs to be minimised, however it must be:

  • adequate in the number of analytes or parameters examined – usually determined by the objectives of the monitoring program guided by national or international standards (such as ‘Guidelines for drinking-water quality’ – WHO 1993, 1984).;
  • executed to an acceptable standard (i.e. with the accuracy and precision required by the monitoring program) and in accordance with standard methods detailed in relevant reference works (such as Stumm and Morgan 1981; Bassett et al. 1978; APHA 1971); and
  • subject to Quality Assurance to national (eg. AS 3901) or international (eg. ISO 9001) standards to ensure both the ongoing validity of results and that every sample is properly analysed since the collection of additional samples for re-analysis is expensive, inconvenient and cannot guarantee to recapture the physico-chemical regime present in the initial sample.
These principles should be applied whether the analysis is done in-house or contracted to an outside laboratory. Where a choice of analytical method exists (especially between instrumental methods) the most cost-effective method that meets the program’s accuracy requirements should be selected; references such as Willard et al. (1988) are invaluable for such determination.

QUALITY ASSURANCE (QA)

QA is as much part of the monitoring process as any technology or technique. Since failure at any stage of groundwater monitoring activities can impair the effectiveness of the program, it is imperative that an overall process is in-place to ensure every task is executed correctly. This produces reliability of results for users and avoids loss of (sometimes unrecoverable) data – maximising the cost-effectiveness of the program.

A crucial decision that must be made at the planning stage and documented into the sampling protocol, is the number and type of quality control samples or standards (controls) to be taken. These will be determined by the nature of the errors to be assessed (both random and systematic) and the accuracy desired in their assessment (Keith 1990).

There are basically two types of controls which are used to determine whether:

  • the analytical procedure is in statistical control; or
  • specific analytes are present in the sample population but not in a similar control population.

Controls for the analytical procedure comprise check standards and laboratory controls.

  • Check standards are specially prepared solutions with accurately known and measurable (ie generally low) concentrations of specific analytes comparable to those expected in the field. These enable assessment of calibration and that minimum detection limits are being achieved.
  • Laboratory control standards are usually provided by an independent source. They are generally QA certified as being prepared to a very accurately known concentration and are used to assess the accuracy of analysis within a laboratory. They are often submitted by the QA manager as blind samples, ie the analytical staff of the laboratory are unaware they are testing a known sample (Dux 1986).

Controls for specific analytes comprise field spike samples and background (control site) samples.

  • Field spike samples – are selected samples to which known quantities of specific analytes are added during the field collection process – this enables evaluation of field, transportation, storage and matrix effects.
  • Background or control or matrix samples – are additional samples collected simultaneously in the vicinity of the sampling well – to establish whether the site is contaminated or somehow different from other sites. These samples are further divided into local control sites – close to the sampling well (eg within the same housing area or factory complex), and area control sites (eg within the same catchment).

CONCLUSION

This paper reflects on the spectrum of technologies encompassed in groundwater monitoring. A monitoring program not only involves different disciplines and technologies, but also a wide range of personnel from field assistants to research scientists. In order to achieve the goals of groundwater monitoring and ultimately fulfil its role in the catchment management process, the monitoring program itself requires careful management; Quality Assurance (QA) offers a soundly structured approach to achieve this. QA forces precise definition of objectives from the outset enabling levels of accuracy and precision for every phase of the program to be determined with corresponding selection of the most appropriate technology. This type of approach ensures that the program will be cost-effective and (most importantly) that the results obtained will best serve the needs of the end-user. Proper catchment management can only be achieved if the groundwater monitoring program provides reliable results from which decisions can be confidently made.

REFERENCES

APHA, 1971, ‘STANDARD METHODS for the Examination of Water and Wastewater’, Thirteenth Edition, Published by the American Public Health Association, Washington D.C.

Barcelona M., Keely J.F., Pettyjohn W.A. and Wehrmann A., 1988, ‘Handbook of Groundwater Protection’, Hemisphere Publishing Corporation, New York.

Bassett J., Denney R.C., Jeffery G.H. and Mendham J., 1978, ‘Vogel’s Textbook of Quantitative Inorganic Analysis’, Fourth Edition, Longman, London.

Blake L.S. (editor), 1989, ‘Civil Engineer’s Reference Book’, Butterworth-Heinemann, Oxford.

Cherry J.A., Gillham R.W., Anderson E.G. and Johnson P.E., 1983, ‘Migration of Contaminants in Groundwater at a Landfill: A Case Study – 2. Groundwater Monitoring Devices’, Journal of Hydrology, v. 63, pp 31-49.

Davis G.B. and Barber C., 1994, ‘Monitoring and Sampling Methods’, Chapter 19 in CSIRO Groundwater School Notes, 12th Groundwater School, July 1993, Adelaide.

Domenico P.A. and Schwartz F.W., 1990, ‘Physical and Chemical Hydrogeology’, John Wiley & Sons, New York.

Dux J.P., 1986, ‘Handbook of Quality Assurance for the Analytical Chemistry Laboratory’, Van Nostrand Reinhold, New York.

Gillham R.W. and Johnson P.E., 1981, ‘A positive-displacement device for multilevel water-quality monitoring in sand or gravel aquifers’, Groundwater Monitoring Review, v. 1, pp 33-35.

Hirschberg, K-J. B., 1993, ‘Guidelines for Groundwater Monitoring at Municipal Landfill Sites’, (Perth, Geological Survey of Western Australia).

Keely J.F. and Boateng K., 1987, ‘Monitoring Well Installation, Purging, and Sampling Techniques – Part 1: Conceptualisations’, Ground Water, v. 25, pp 300-313.

Keith L.H., 1990, ‘Environmental Sampling: A Summary’, Environmental Science and Technology, v. 24, pp 610-617.

Kent R.T. and Payne K.E., 1988, ‘Sampling Groundwater Monitoring Wells – Special Quality Assurance and Quality Control Considerations’ in ‘Principles of Environmental Sampling’, Lawrence H.K. (editor), American Chemical Society, Washington D.C.

Lerner D.N. and Teutsch G., 1995, ‘Recommendations for level-determined sampling in wells’, Journal of Hydrology, v. 171, pp 355-377.

Neilsen D.M. and Yeates G.L., 1985, ‘A Comparison of Sampling Mechanisms Available for Small-Diameter Ground Water Monitoring Wells’, Ground Water Monitoring Review, v.5, pp 83-99.

Pickens J.F., Cherry J.A., Grisak G.E., Merritt W.F. and Risto B.A., 1978, ‘A multilevel device for groundwater sampling and piezometric monitoring’, Ground Water, v. 16, pp 322-327.

Smith J.S., Steel D.P, Malley M.J. and Bryant M.A., 1988, ‘Groundwater Sampling’, in ‘Principles of Environmental Sampling’, Lawrence H.K. (editor), American Chemical Society, Washington D.C.

Stites W. and Chambers L.W., 1991, ‘A Method for Installing Miniature Multilevel Sampling Wells’, Ground Water, v. 29, pp 430-432.

Stumm W. and Morgan J.J., 1981, ‘Aquatic Chemistry – An Introduction Emphasising Chemical Equilibria in Natural Waters’, Second Edition, John Wiley & Sons, New York.

WHO, 1993, ‘Guidelines for drinking-water quality’ Volume 1 – Recommendations, Published by the World Health Organisation, Geneva.

WHO, 1984, ‘Guidelines for drinking-water quality’ Volume 2 – Health Criteria and Other Supporting Information, Published by the World Health Organisation, Geneva.

Willard H.H., Merrit L.L., Dean J.A. and Settle F.A., 1988, ‘Instrumental Methods of Analysis’, Seventh Edition, Wadsworth Publishing Company, Belmont, California.

APPENDIX A

Advantages and disadvantages of Selected Drilling Methods for Monitoring Well Construction (modified, after Barcelona et al. 1988 and Keely and Boateng 1987)

Method: Hand Augering

Drilling Principle:
A hand operated implement with a hollow, shaped cutting head (usually between 50 and 100mm in diameter) with some 1.5 metres of shaft which can be extended by (detaching the handle) in 1.5 metre increments.

Advantages:

  • It is highly portable and can be manually carried into any terrain.
  • Being manually operated it requires no mechanical power source.
  • Provides excellent core samples for stratigraphic analysis down to very fine resolution (ie thin lenses of sedimentary material only millimetres thick can be observed and logged).

Disadvantages:

  • Cannot drill through indurated or consolidated strata.
  • Limited to approximately 7 metres depth.
Method: Drive Point

Drilling Principle:
A manually (or mechanically) installed well; a steel pipe (typically 30 to 50 mm ID) with a suitable driving point is driven into the ground to the desired depth. The sampling system is installed within the steel pipe, the latter is withdrawn leaving the sediment to close-in around the sampling tube/s.

Advantages:

  • Minimal aquifer and environmental disturbance.
  • Suitable for sites inaccessible to vehicles (manual installation).
  • Materials are easily obtainable and inexpensive.
  • Water samples can be obtained as driving proceeds.
  • A good seal between casing and formation can be achieved – bore hole often sealed with bentonite near the surface to prevent entry of surface water (Stites and Chambers 1991).

Disadvantages:

  • Limited to about 8 metres depth; i.e. shallow aquifers.
  • Only suited to unconsolidated, cohesionless sediments; i.e. will not penetrate rock or thick clay strata.
  • Collection of formation samples is not possible.
  • Development is difficult due to compaction of walls.
  • PVC and Teflon™ casings and screens are not strong enough to be driven; the use of metal construction materials may impair water quality determinations.
Method: Auger, Hollow-stem and Solid-stem

Drilling Principle:
Successive 1.5 metre lengths of spiral-shaped drill stem are rotated into the ground to create the borehole; cuttings are brought to the surface by the turning action of the auger.

Advantages:

  • Inexpensive.
  • Relatively simple operation; small rigs can generally get into difficult-to-reach locations.
  • Quick set-up time.
  • Wells can be rapidly established in cohesive substrates.
  • No drilling fluid is required.
  • Hollow augers facilitates collection of split-spoon core samples.
  • Small-diameter monitoring well can be built inside the hollow-stem flights prior to withdrawal of auger; essential in non-cohesive soils.

Disadvantages:

  • Maximum penetration depths limited to 45 metres; especially in non-cohesive soils.
  • Cannot be used in rock or indurated formations; difficult to drill through cobbles or boulders.
  • Logging stratigraphy is difficult unless split-spoon coring is used; time delay and sample mixing while material is carried to the surface precludes accurate analysis.
  • Vertical leakage of water through the borehole during drilling is likely to occur.
  • Solid-stem augering is limited to fine-grained unconsolidated substrates that will not collapse when unsupported.
Figure A.1 Cutaway sketch of the hollow-stem augering method (after Keely and Boateng 1987).
Figure A.2 Cutaway sketch of a hybrid drilling method – augering with a temporary driven casing (after Keely and Boateng 1987).
Method: Jetting

Drilling Principle:
Washing action of water forced out of the bottom of the drill rod clears the borehole to allow penetration; cuttings are brought tot he surface by water flowing up the outside of the drill rod.

Advantages:

  • Inexpensive; driller often not needed for shallow holes.
  • Construction is simple in cohesive substrates.

Disadvantages:

  • A relatively slow process, especially with increasing depth.
  • Not suited for rock or coarse materials (cobbles/boulders).
  • A water supply is required with pumping equipment to develop the required hydraulic pressures.
  • Difficult to interpret sequence of geologic materials from the cuttings.
  • Maximum depth limited to 45 metres, depending on geology and available water pressures.
Method: Cable-tool (Percussion)

Drilling Principle:
The borehole is created by dropping a heavy ‘string’ of drill tools down the hole, crushing/fracturing the materials at the bottom. Cuttings are removed occasionally by bailer. Generally, casing is driven just ahead of the bottom of the hole; hole diameter usually exceeds 150 mm.

Advantages:

  • Can be used in rock formations as well as unconsolidated formations.
  • Reasonable accurate borehole logs can be prepared from cuttings if collected often enough.
  • Driving a casing ahead of the hole minimises cross-contamination by vertical leakage between aquifers.
  • Core samples can be easily obtained.

Disadvantages:

  • Requires an experienced driller.
  • Heavy steel drive pipe used to keep hole open and drilling tools can limit accessibility.
  • Cannot run some geological logs due to presence of drive pipe.
  • Relatively slow drilling method.
Figure A.3 Cutaway sketch of the cable tool drilling method (after Keely and Boateng 1987).
Method: Hydraulic (or Mud) Rotary

Drilling Principle:
Rotating bit breaks formation; cuttings are brought to the surface by a circulating fluid (drilling ‘mud’ – often a bentonite slurry). Mud is forced down the interior of the drill stem, out the bit, and back up the annulus between the drill stem and hole wall. Cuttings are removed by settling in a ‘mud pit’ at the surface and mud is recirculated back down the drill stem.

Advantages:

  • Drilling is fairly quick in all type of geological formations.
  • Borehole will generally stay open from formation of a mud wall on the sides of the hole – left by the circulating drilling mud; eases geophysical logging and well construction.
  • Geologic cores can be collected.
  • Virtually unlimited depths are possible.

Disadvantages:

  • Expensive, requires experienced driller and considerable amount of peripheral equipment.
  • Completed well may be difficult to develop, especially small diameter wells, due to the mud wall lining the borehole.
  • Visual inspection of the cuttings is possible but thin sand, gravel or clay beds are likely to be missed due to the masking effect of the drilling mud.
  • Presence of drilling mud can contaminate water samples, especially the organic, biodegradable muds.
  • Circulation of drilling fluid through a contaminated zone can create a hazard at the surface with the mud pit, and cross-contaminate clean zones during circulation.
Method: Reverse Rotary

Drilling Principle:
Similar to hydraulic rotary except that the drilling fluid is circulated down the borehole outside the drill stem and is pumped up the inside, just the reverse of the normal rotary method. Water is used as the drilling fluid, rather than mud, and the hole is kept open by the hydrostatic pressure of the water standing in the borehole.

Figure A.4 Cutaway sketch of the hydraulic or mud rotary drilling method (after Keely and Boateng 1987).

Advantages:

  • Creates a very ‘clean’ hole, not dirtied with drilling mud.
  • Can be used in all geologic formations.
  • Very deep penetrations possible.
  • Split-spoon sampling possible.

Disadvantages:

  • A large water supply is required to maintain hydrostatic pressure in deep holes and when highly conductive formations are encountered.
  • Expensive – experienced driller and significant peripheral equipment is required.
  • Hole diameters are usually large, commonly 450 mm or larger.
  • Cross-contamination from circulating water is likely Geologic samples brought to the surface are generally poor, circulating water will ‘wash’ finer materials from the sample.
Method: Air Rotary

Drilling Principle:
Very similar to hydraulic rotary, the main difference being that air is used as the primary drilling fluid as opposed to mud or water. In uncohesive, unconsolidated formations, a temporary casing is often driven as the borehole is drilled to minimise problems with cave-ins.

Advantages:

  • Can be used in all geologic formations; most useful in highly fractured environments.
  • Fairly quick.
  • No drilling mud of water is required.

Disadvantages:

  • Relatively expensive.
  • Cross-contamination from vertical communication possible.
  • Air will be mixed with water in the hole and that which is blown from the hole, potentially creating unwanted reactions with contaminants; may affect ‘representative’ samples.
  • Cuttings and water blown from the hole can pose a hazard to crew and surrounding environment if toxic compounds are encountered.
  • Organic foam additives to aid cuttings removal may contaminate samples.
Figure A.5 Cutaway sketch of the air rotary drilling method (after Keely and Boateng 1987).
Method: Air-Percussion Rotary or Downhole-Hammer

Drilling Principle:
Air rotary with reciprocating hammer connected to the bit to fracture rock.

Advantages:

  • Very fast penetrations.
  • Useful in all geologic formations.
  • Only small amounts of water needed for dust and bit temperature control.
  • Cross-contamination potential can be reduced by driving casing.

Disadvantages:

  • Relatively expensive.
  • As with most hydraulic rotary methods, the rig is heavy, limiting access.
  • Vertical mixing of water and air creates cross-contamination potential.
  • Hazard posed to surface environment if toxic compounds encountered.
  • Organic foam additives to aid cuttings removal may contaminate samples.