design and operation of infiltration galleries and water ... · chapter 1 - design and operation of...

December 2009

CHAPTER 1. DESIGN AND OPERATION OF INFILTRATION

GALLERIES AND WATER QUALITY CHANGES

Elise Bekele, Simon Toze, Brad Patterson, Brian Devine, Simon Higginson,

Wolfgang Fegg and Joanne Vanderzalm

Water for a Healthy Country Flagship Report series ISSN: 1835-095X

Australia is founding its future on science and innovation. Its national science agency, CSIRO, is a powerhouse of ideas, technologies and skills.

CSIRO initiated the National Research Flagships to address Australia’s major research challenges and opportunities. They apply large scale, long term, multidisciplinary science and aim for widespread adoption of solutions. The Flagship Collaboration Fund supports the best and brightest researchers to address these complex challenges through partnerships between CSIRO, universities, research agencies and industry.

The Water for a Healthy Country Flagship aims to provide Australia with solutions for water resource management, creating economic gains of $3 billion per annum by 2030, while protecting or restoring our major water ecosystems. The work contained in this report is collaboration between CSIRO, Water Corporation WA, ChemCentre, Curtin University and The University of Western Australia.

For more information about Water for a Healthy Country Flagship or the National Research Flagship Initiative visit www.csiro.au/org/HealthyCountry.html

Citation: Bekele E, Toze S, Patterson B, Devine B, Higginson S, Fegg W and Vanderzalm J. 2009, Chapter 1 - Design and operation of infiltration galleries and water quality changes, In: Determining requirements for managed aquifer recharge in Western Australia. 1, CSIRO: Water for a Healthy Country National Research Flagship

1 Research undertaken as Work Package 1 was lead by Palenque Blair (Water Corporation WA) and (deputy) Elise Bekele (CSIRO). This chapter of the report was authored by Elise Bekele (CSIRO), Simon Toze (CSIRO), Brad Patterson (CSIRO), Brian Devine (UWA), Simon Higginson (CSIRO), Wolfgang Fegg (CSIRO), and Joanne Vanderzalm (CSIRO).

Copyright and Disclaimer

© 2009 CSIRO To the extent permitted by law, all rights are reserved and no part of this publication covered by copyright may be reproduced or copied in any form or by any means except with the written permission of CSIRO.

Important Disclaimer:

CSIRO advises that the information contained in this publication comprises general statements based on scientific research. The reader is advised and needs to be aware that such information may be incomplete or unable to be used in any specific situation. No reliance or actions must therefore be made on that information without seeking prior expert professional, scientific and technical advice. To the extent permitted by law, CSIRO (including its employees and consultants) excludes all liability to any person for any consequences, including but not limited to all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from using this publication (in part or in whole) and any information or material contained in it.

Cover Photograph: Sampling at the Floreat Infiltrator Galleries

Photographer: Anne McKenzie © 2010 CSIRO

Contents 1.1. WORK PACKAGE 1 INTRODUCTION AND RESEARCH AIMS................................... 1 1.2 METHODS AND MATERIALS ....................................................................................... 2

1.2.1 Location and Geology of Infiltration Gallery Sites .................................................2 1.2.2 MAR Infrastructure and Monitoring Equipment .....................................................4 1.2.3 Determination of Recharge Rates and Aquifer Transport Times of the

Reclaimed Water.................................................................................................13 1.2.4 Quantitative Estimate of Mixing Proportion Between Recharge Water and

Background Groundwater ...................................................................................17 1.2.5 Monitoring of the Infiltration Galleries and Recharged Water .............................18 1.2.6 Microbial Pathogen Decay Analysis....................................................................21

1.3 RESULTS..................................................................................................................... 24 1.3.1 Recharge Rates and Operation of the Infiltration Galleries.................................24 1.3.2 Determination of Recharge through the Unsaturated Zone ................................27 1.3.3 Recharge Water Residence Time in the Aquifer.................................................33 1.3.4 Dilution with Background Groundwater Based on Chloride Model .....................42 1.3.5 Water Quality Changes .......................................................................................44

1.4 LESSONS LEARNT: MANAGEMENT AND OPERATIONAL REQUIREMENTS FOR MAR SYSTEMS IN WESTERN AUSTRALIA............................................................... 65 1.4.1 Infiltration System Design ...................................................................................65 1.4.2 Reliability of Water Supply ..................................................................................66 1.4.3 Iron Precipitation and Fouling of Pumps .............................................................67 1.4.4 Pipe Leakage and Protection ..............................................................................68 1.4.5 Clogging of Galleries...........................................................................................69 1.4.6 Monitoring Equipment .........................................................................................69

1.5 CONCLUSIONS ........................................................................................................... 70 1.6 REFERENCES............................................................................................................. 73

APPENDIX 1A – Details of the Monitoring Bores and Water Quality Logging Equipment for the MAR trials...............................................................................76

APPENDIX 1B – Uranine and Bromide Tracer Test: Sampling Program and Fluorometer Calibration .......................................................................................79

APPENDIX 1C – MODFLOW Model for the Floreat Infiltration Galleries Site ..............81 APPENDIX 1D – Inorganic Chemistry Analytical Procedures ......................................85 APPENDIX 1E – Organic Chemistry Analytical Procedures.........................................91 APPENDIX 1F – Methods for the Detection of Microorganisms in Samples

Collected from the Infiltration Galleries Sites.......................................................92 APPENDIX 1G – Preparation of Enteric Microorganisms for Pathogen Decay

Experiments and Detection of Viable Microorganisms in Diffusion Chambers....94 APPENDIX 1H – Decay of Enteric Microorganisms in Diffusion Chambers at the

Floreat Infiltration Gallery Site .............................................................................98 APPENDIX 1I – Halls Head Water Level and Water Chemistry Investigation............101 APPENDIX 1J – Bromide and Uranine Tracer Tests at the Floreat Infiltration

Galleries: Diploma thesis by Wolfgang Fegg.....................................................108

Determining requirements for MAR – Chapter 1 1

1.1. WORK PACKAGE 1 INTRODUCTION AND RESEARCH AIMS

A major research aim for the Managed Aquifer Recharge (MAR) project was to determine the applicability of MAR in urban areas of the Swan Coastal Plain. While there are a range of MAR methods (e.g. well injection, pond infiltration, soil aquifer treatment) and recharge water sources (e.g. drinking water, urban stormwater, treated wastewater) (see MAR chapter of Australian Guidelines for Water Recycling (NRMMC-EPHC-NHMRC, 2009) for more details), for the current project it was decided to test infiltration galleries using secondary treated wastewater. Infiltration galleries have the advantage of being cheaper and less sophisticated to operate than well injection systems and are potentially less prone to clogging than injection wells. Infiltration galleries also have potential advantages over MAR systems that use ponds (such as pond infiltration and soil aquifer treatment methods) in that they are subsurface and thus do not take up valuable ground area. There is also considerably less potential for unsupervised access to the recharging water by the community, domestic animals or wildlife. In addition, the lack of exposed water also means that there is no chance for mosquito breeding in the water prior to recharge.

Infiltration galleries have been used in other regions of the world but primarily to recharge potable quality water or tertiary treated wastewater to aquifers; however, there has been little testing of infiltration galleries to recharge secondary treated effluent to an unconfined aquifer (Peter Dillon, CSIRO, personal communication, 2005). The Swan Coastal Plain is potentially ideal for using infiltration galleries as it consists, particularly in the coastal areas, of coarse sand overlying limestone in which lies a shallow unconfined aquifer.

To test the applicability of infiltration galleries on the Swan Coastal Plain, two experimental field sites for MAR were established. One was located at the Halls Head Wastewater Treatment Plant in Mandurah, and the second installed at the CSIRO Centre for Environment and Life Sciences in Floreat, Perth. The latter is referred to as the Floreat Infiltration Galleries site or FIG site.

The sites were used to address several research objectives, which were:

• R1: to monitor recharge of water and determine clogging rates during aquifer recharge at the field sites

• R2: to determine microbiological die-off during infiltration to, and residence time in the superficial aquifer

• R3: to assess soil attenuation of nitrogen and phosphorus during infiltration and aquifer transport. (Note, this research objective was moved to the work package for the laboratory column experiments) and is described in detail in Chapter 2.

In addition to providing a research facility, the field sites were designed and monitored to address several technical objectives, which were:

• T1: to identify water quality improvements that occur during the recharge of recycled water to groundwater (e.g. removal of pathogens, chemicals of concern and nutrients via biogeochemical processes)

• T2: to test existing and novel treatment technologies (e.g. secondary wastewater treatment, tertiary treatment as well as using the aquifer to improve water quality) to establish suitability of the proposed applications for use with different MAR technologies to produce appropriate fit-for-purpose water


• T3: to characterise aquifers identified as potential MAR sites to determine factors influencing recharge and recovery efficiencies

• T4: to identify management and operational requirements that are needed to ensure that future MAR schemes are operated in a manner that optimises the economic and environmentally suitable use of reclaimed waters, while managing apparent and inherent operational, health and environmental risks.

The two field sites were used to address these objectives; however, due to different conditions at the sites, some aspects of the objectives were addressed at only one site where conditions were suitable.

1.2 METHODS AND MATERIALS

1.2.1 Location and Geology of Infiltration Gallery Sites

The managed aquifer recharge trials at Halls Head and Floreat were installed in different geological settings and under different hydrogeological conditions to compare gallery performance and to investigate the relevance for water quality changes. Both sites have minimal topographic variation.

The first set of infiltration galleries were installed at the wastewater treatment plant in the suburb of Halls Head, southwest of the city of Mandurah (115º41’ E longitude; 32º32’S latitude). The galleries were located approximately 300 m from the Indian Ocean (Figure 1-1). The coastal deposits at the site consist of a thin veneer of sand overlying Tamala Limestone, which is a calcareous aeolianite that is variably cemented and contains dissolution features. Groundwater levels at the site were heavily influenced by tidal fluctuations. The deepest monitoring bores at the site were drilled to a depth of 6 m in the Tamala Limestone and the water table was approximately 2 m below ground. The position of the galleries relative to the water table allowed for only a 1.5 m section of unsaturated thickness. Groundwater flow directions were highly variable at the Halls Head site and could not be well delineated due to several factors, including pumping from different supply bores, tidal and seasonal water level fluctuations and effluent recharge from the galleries and infiltration ponds at the wastewater treatment plant.


Figure 1-1. Location map showing the geomorphology and soils on the Swan Coastal Plain (after McArthur and Bettenay, 1974) and the managed aquifer recharge research sites established for the Premier’s Water Foundation project

The second set of infiltration galleries were installed at the CSIRO Centre for Environment and Life Sciences in Floreat (115º47’ E longitude; 31º57’S latitude), which is about 1.5 km inland from the Indian Ocean. The geology of the site consisted of a thicker section (6 to 7 m) of Spearwood Dune sand overlying Tamala Limestone (Figure 1-1). The depth to the water table below the galleries varied seasonally between 10 and 11 m. The deepest bore installed for initial investigative purposes was drilled to the base of the Tamala Limestone to a depth of 31 m below ground. During drilling of two bores in the Tamala Limestone, there was a loss of drilling fluid caused by the fluid entering a highly porous zone, such as a fracture and/or cavity. These bores were located at a distance of 5 m and 30 m from the galleries at the Floreat site and the suspected cavities or fractures were intercepted at depths ranging from 14 to 30 m below ground.

The stratigraphy at the Floreat infiltration galleries is typical of the Swan Coastal Plain and consists of sand overlying aeolianite, whereby the sands have evolved from in situ weathering of the underlying Tamala Limestone as described in Tapsell et al. (2003). The characteristics of the Spearwood sand and the upper part of the Tamala Limestone in the unsaturated zone were investigated as part of a preliminary assessment for the site, using cored material obtain by the wireline coring method with a hollow stem auger (Rümmler et al., 2005). Cemented sands that are typical of the Tamala Limestone unit are commonly intercepted below 6 m, but the transition depth from Spearwood sand to Tamala Limestone varies across the field site based on drilling many monitoring bores at the site.

Mineralogical analyses of cored material from the Floreat site using x-ray diffraction (Michael Verrall, CSIRO, personal communication, 2005) indicated that the Spearwood sand is


predominantly quartz sand (80–90%) with very low amounts of calcite (<2%); whereas cored material from the upper part of the Tamala Limestone (between 7.5 and 11 m depth) has less quartz and a greater proportion of calcite (20–40%). The aeolianite (Tamala Limestone) matrix is mainly cemented with calcium carbonate. There are variable amounts of iron oxide coating the Spearwood sand, which gives it a brown or yellow-brown colour (Rümmler et al., 2005).

1.2.2 MAR Infrastructure and Monitoring Equipment

Design and construction of infiltration galleries

The original set of infiltration galleries at both sites consisted of one gravel-filled gallery and one Atlantis Leach System® gallery as shown in Figure 1-2. The latter is a series of modular, lightweight, polypropylene crates. The only visible sign of the MAR operation underground was the inspection lid (Figure 1-3). The galleries were buried to a depth of 0.5 m below ground. After one year, the gravel gallery at the Floreat site clogged and was removed and replaced with another Atlantis Leach System® gallery. The new gallery at the Floreat site was inadvertently installed 0.5 m deeper than the original one.


Figure 1-2. Schematic of the covered infiltration galleries that were originally installed at the Floreat site

TOP VIEW

SIDE VIEW

Not Drawn to Scale

Slotted PVC pipeInflow pipe

Inspection lid for concretedischarge chamber

0.5 m

0.5 m

Pipe to secondary trench

FRONT VIEWLevel probe

Infiltration with in-situ soil (infiltration surface)

1 m

Geofabric on top and sides

0.15 m0.10

0.25

End cap

Atlantis Leach System

Gravel

4 m

12.5 m 12.5 m

1 m

1 m

Separate effluent inflow pipes to each gallery


Figure 1-3. Installation of the two infiltration galleries in Floreat. The top left photo shows the Atlantis Leach System®; the bottom photo shows the gravel-filled gallery and the scored PVC pipe that distributed treated effluent to the ends of the gallery. The inspection points for viewing effluent inflow to each gallery are highlighted in the top right photo.

The Mandurah field site which contained the Halls Head infiltration galleries was decommissioned in November 2006, approximately 22 months after infiltration started, due to redevelopment plans for the wastewater treatment plant. Unlike the grassy sheep paddock at the Floreat site, the Halls Head site had drier, weedy surface vegetation that was periodically removed.

Reclaimed water source and pre-treatment

The source of the wastewater at Floreat was the Subiaco Wastewater Treatment Plant (Figure 1-4). A similar set-up was used for the Halls Head MAR trial, except that the filtration and instrumentation equipment were consolidated on one skid at the Halls Head wastewater treatment plant.


Figure 1-4. Treatment processes that were involved in treating raw wastewater at the Subiaco Wastewater Plant to the point where final effluent is produced that is discharged at an ocean outfall (Source: Water Corporation)

Figure 1-5. (a) Skid-mounted filtration unit at Subiaco Wastewater Treatment Plant, showing the multi-media filters (light blue vessel) and other equipment for pumping water through the system and backwashing the filters; (b) adjacent weir containing the submersible pump


Secondary treated wastewater was pumped from one of the secondary clarifiers to flow through an Amiad® skid-mounted filtration system, which was located at the treatment plant (Figure 1-5). The instrument panel on this skid was used to control and monitor the feed and backwash pumps. The tank in Figure 1-5 contained freshwater, which was used to backwash the filtration unit at regular intervals. A submersible pump was placed in an adjacent weir from a secondary clarifier to supply water to the filtration system. The multi-media filter contains layers of anthracite (1.1 mm grading), sand (0.5–1.3 mm grading) and gravel (1.5–3 mm grading).

After passage through the Amiad® skid-mounted filtration system, the effluent was pumped approximately 730 m to the Floreat Infiltration Galleries site through 50 mm diameter PVC pipe buried a few metres below ground. At the MAR site, effluent flow to the two infiltration galleries was controlled by instrumentation on a second Amiad® skid with telemetry to the first skid at the treatment plant. The skid at the MAR site contained electronic flow rate control valves for the discharge lines to the two galleries as well as water temperature and turbidity sensors. Each discharge line had an electromagnetic flow meter to display the current flow rate into the galleries as well as sending an analogue signal to a Condor® controller (Dorot) and an Eco-graph recording device on the skid.

The effluent flow rate to each gallery was controlled using a Condor® controller, a programmable logic device that offers the option of completing a set point table of water levels and corresponding flow rates. The logic is that when the infiltration gallery is empty the flow rate will increase and conversely, if the water level is high, the flow rate will decrease. An optimal point should be reached where the infiltration rate equals the flow rate (Source: Amiad® documentation). Electric float switches were installed in each gallery to communicate water level information to the Condor® controllers.

During the MAR trial at Floreat, the Condor® controllers were programmed and then manually adjusted to supply a constant rate of 17.5 litres per minute (equivalent to 25 kilolitres per day) of treated effluent to each gallery. If the water level within a gallery reached a maximum level, the inflow of treated effluent would shut off.

Construction and location details for groundwater bores

To compare water quality changes at the Floreat MAR site with the surrounding aquifer, three ‘background’ bores located hydraulically up-gradient were regularly sampled. The direction of the background regional groundwater flow was to the west. These background monitoring bores were located 75 to 185 m east-northeast of the galleries (Figure 1-6). To monitor water quality changes in relation to activities at the MAR site, a series of monitoring bores were installed in close proximity with slotted intervals positioned to intercept the recharge plume at different depths (Table 1A-1 in Appendix 1A). One bore (BH5) was located a few meters east (up-gradient) of the galleries and slotted at the water table to monitor any spread of the recharge plume in the up-gradient direction. An extraction bore was installed at a distance of 50 m down-gradient from the infiltration galleries. The position of these bores is shown in map view and cross-section (Figures 1-7 and 1-8).

The extraction bore contained a Grundfos submersible pump (SP30-3), which has a maximum rating of 30 m3/h, equivalent to 720 kilolitres per day. For the MAR trial, the pumping rate was regulated with a gate valve and set to achieve a daily flow of 250 kilolitres per day. The discharge water was piped directly to the sewer and returned to the Subiaco Wastewater Treatment Plant.


Figure 1-6. Arial photo showing the approximate direction of groundwater flow and monitoring bores at the Floreat Infiltration Galleries site and the three bores that were monitored for background water quality. Photo from Google Earth.

Figure 1-7. Map view of the Floreat site showing the positions of the monitoring bores, the recovery bore, the neutron moisture meter access tubes and the two infiltration galleries

BH5

BH2

BH1

BH17-Recovery Bore

BH7

BH8

BH9

BH10

BH11

BH12

BH13BH14BH15BH16BH3

North West Gallery

EastGallery

BH6

Scale3 monitoring bores (2 m screen); cored sites

NM3

12 monitoring bores (1 m screens)

1 recovery bore (10 m screen)

Legend

Discharge chamber where treated effluent enters infiltration gallery

10m

10mBH19

BH18

BH20

3 monitoring bores (2 m screen)

9 cased holes for neutron moisture measurements. NM5 and NM6 were cored.

NM4

NM2

NM1

NM5

NM6

NT03NT02NT01


Figure 1-8. An east-west cross-section of the Floreat site showing the position and lengths of the slotted casing in the bores and the elevation of the water table measured in late January 2008

Figure 1-9. Arial photo showing the infiltration galleries at the Halls Head Wastewater Treatment Plant relative to other operations, including sludge drying beds and infiltration ponds. Two supply pumping bores (SPB1 and SPB2) influence groundwater flow directions as shown by the arrows (Toze et al., 2002). Photo from Google Earth.

0

5

10

15

20

25

30

-15-10-50510152025303540455055

Distance relative to West Infiltration Gallery (m)

Dep

th b

elo

w g

rou

nd

(m

)

Top Slotted Interval

Base of Slotted Interval

31/01/2008

Slotted interval within each bore

Atlantis Leach System® infiltration gallery

BH17(pumping bore)

BH3

BH15

BH14

BH13

BH10

BH12

BH11

BH7BH9

BH8

BH2BH1

Legend

BH5

BH16

Elevation of the water table on January 31, 2008.

BH6

BH18BH19BH20

West East

Former gravel-filled infiltration gallery


Figure 1-10. The relative positions of the Halls Head infiltration galleries and the six adjacent groundwater bores that were sampled and regularly monitored using in situ loggers. The average water table depth was 2 m below ground and the slotted casing in the bores extended from 3 to 6 m below ground.

At the site of the Halls Head MAR trial, where a previous study had been conducted involving managed aquifer recharge via pond infiltration (Toze et al., 2002, Toze et al., 2004), there were a number of pre-existing monitoring bores to sample and data to use for interpreting changes in water chemistry. The two primary bores used to interpret changes in water chemistry relative to the MAR trial were bore 2/84 (located near one of the supply bores that received a mixture of water from hydraulically up-gradient and from the infiltration ponds) and bore 6/88 (located external to the Halls Head Wastewater Treatment Plant and representative of conditions that would occur naturally in the superficial aquifer) (Figure 1-9). In close proximity to the infiltration galleries, were six monitoring bores with 3 m slotted intervals intercepting the water table. Their locations are noted in Table 1A-2 in Appendix 1A and shown in Figure 1-10. Water quality data from other bores indicated in Figure 1-9 (i.e. SPB1, SPB2, 1/83, 1/84 and 7/88) were used in some of the interpretations for this study.

Scale

10m

10m

HH_W2

Gravel Gallery

Atlantis Gallery

HH_W1

HH_E1

HH_E2

HH_S1

HH_N1

North


Submersible probes and recording devices

To monitor changes in water quality and water level within the aquifer and within the infiltration galleries, several different types of data recording probes were used at the two MAR sites (Table 1-1). As there were not enough probes to record changes in all of the bores, different probes were installed at different locations for periods of up to several weeks or months to gain information as needed at the MAR sites.

The probes were programmed to record information at frequent intervals, generally every 15 or 30 minutes, to monitor responses to changes in water quality and infiltration rates. The Odyssey capacitance water level probes were instrumental in detecting increases in water level, indicative of clogging.

Table 1-1. Submersible probes

Probe designation Water parameters

recorded* Locations used

Troll 9000; Troll 9500 from In-

Situ Inc.

pressure, temp, turb,

ORP, pH, DO, EC

Floreat: BH8, BH10, BH16,

BH17

Halls Head: HH_E1

Odyssey capacitance water level

probe from DataFlow Systems

Pty Ltd

water level Floreat: both galleries,

BGRND01, BH6, BH7, BH15

Halls Head: both galleries,

2/84, HH_E2,

Odyssey conductivity and

temperature probe from DataFlow

Systems Pty Ltd

EC and temp Floreat: BH1, BH2, BH6,

BH7, BH8, BH9, BH13,

BH18, BH19, BH20, west

gallery

Levelogger from Solinst Ltd water level, temp, EC Floreat: BH1, BH16

Halls Head: HH_W2

Yeo-Kal Pty Ltd temp, EC, SAL, DO, pH,

ORP, turbidity

Floreat: west gallery

* Abbreviations: DO (dissolved oxygen), EC (electrical conductivity), ORP (oxidation-reduction potential), temp (temperature), turb (turbidity), SAL (salinity).


1.2.3 Determination of Recharge Rates and Aquifer Transport Times of the Reclaimed Water

Porous media properties and infiltration rates

As described in the geology section above, the Floreat Infiltration Galleries site is underlain by Spearwood sand to a depth of around 7.5 m and Tamala Limestone to approximately 30 m depth. The contact between sand and unleached limestone is irregular with cemented limestone extending upwards into the sand at different depths, as commonly observed on the Swan Coastal Plain (Playford et al., 1976). As revealed in the preliminary site assessment study by Rümmler et al. (2005), variability in hydraulic conductivity, porosity and lithology were observed from analysing cored sediments from the upper 12 m of the superficial aquifer. Permeability heterogeneity is a concern for sighting a managed aquifer recharge scheme as fine-grained sand and cemented layers can lead to clogging and reduced infiltration rates. In contrast, cavities or fractures can lead to faster transport times through an aquifer, which may not allow sufficient time for water quality improvements to occur.

The first step in assessing the suitability of the Spearwood sand for MAR was to monitor changes in infiltration rates over time using treated effluent. Infiltrometer experiments were conducted at the Floreat site in March 2004, using a double-ring infiltrometer as described in Bekele et al. (2006) (Figure 1-11). The infiltrometer was used to monitor rates of water inflow within the first few centimetres of Spearwood sand. The average infiltration rate was 55 m/day; only slight clogging was detected and controlled by altering the flow rate. These results, when compared with the design specifications for the galleries, showed sufficient infiltration capacity to proceed with MAR at the site.

Figure 1-11. Double-ring infiltrometer used to determine base infiltration rate potential in Spearwood sands at the Floreat site prior to establishment of infiltration galleries


Estimates of the saturated hydraulic conductivity of the Spearwood sand were inferred from particle-size measurements as described in Rümmler et al. (2005). The estimates suggest that saturated hydraulic conductivity decreases with depth from about 20 m/day to 5 m/day. The Spearwood sand is predominantly medium-grained and with an average porosity of 35% (Rümmler et al., 2005). The proportion of fine-grained sand increases with depth from 10 to 40% until the transition to Tamala Limestone.

The hydraulic conductivity of the Tamala Limestone at the Floreat site was determined by conducting a 24-hour constant rate pumping test in May 2005. An average hydraulic conductivity of 100 m/day was obtained from analysing water level data from three observation bores that responded to pumping a temporary bore screened to the base of the Tamala Limestone. The temporary bore was located close to the eventual site of the infiltration galleries and was later decommissioned to prevent possible short-circuiting of recharge water within the unconfined aquifer.

Unsaturated zone flow and residence times

The characteristics of the unsaturated zone at Floreat were further examined as part of an Diploma thesis from the Fachhochschule Weihenstephan by Fegg (2008) (Appendix 1J) (equivalent of Australian Honours thesis), which included soil layer descriptions, grain-size analyses to estimate different hydraulic properties, several types of tracer tests to estimate the residence time in the unsaturated zone and measurements of soil moisture to look for evidence of preferential flow. The estimates of hydraulic properties were based on the techniques published in Rümmler et al. (2005).

The focus of the unsaturated zone investigation was on the west infiltration gallery, which was initially gravel-filled, but later excavated and replaced with the Atlantis Leach System® after clogging occurred. Excavation of the west gallery provided an opportunity to install new monitoring equipment, including 36 suction cup lysimeters (Cooinda, Australia)for sampling recharge water at different depths below the gallery. Angle drilling methods were also used to install five heat dissipation matric potential sensors (TM229-SMM, ICT International Pty Ltd, Australia) in the unsaturated zone. The placement of this equipment is indicated in Figure 1-12.


Figure 1-12. The locations of the suction cup lysimeter stations (SL1 to SL4) and the TM229 probes (TM1 to TM5) below the new Atlantis Leach System® (west) gallery in the unsaturated zone; (a) map view and (b) cross-section. Each SL station had nine lysimeters as labelled in (b) with lettered identifiers.

Several tracer experiments were conducted to estimate the time required for treated effluent to travel to different depths below the west gallery. Initial trials with deionised water were necessary to assist with planning and designing the sampling strategy for the main tracer experiment. An initial trace experiment with deionised water was conducted on 24 August 2007, followed later by a repeat trial on 30 August 2007. For the main tracer experiment (started on 24 September 2007), 50 g of uranine and 2 kg of sodium and potassium bromide (200 mg/L Br-) were mixed in a tank with 6800 litres groundwater and then released concurrently into the discharge chamber of the west gallery at a rate of 17 L/min over a period of 7 hours. Monitoring for breakthrough of the tracers continued until 6 October 2007.

Please refer to Appendix 1B for details of the sampling program and fluorometer calibration details for the tracer experiment.

The interpretation of travel time through the unsaturated zone mainly focused on tmean, defined as the point in time when 50% of the infiltrated tracer mass passed the suction cup. This value is estimated using numerical integration to sum the mass of tracer below the breakthrough curve for each suction cup. This estimate was used to calculate a mean pore velocity, which is the linear distance from the base of the infiltration gallery to the depth of each suction cup divided by tmean.

Other methods were applied to investigate travel times, or the wetting front migration rate, through the unsaturated zone at the Floreat site. One method involved testing a newly developed groundwater velocity probe (Patterson et al., in press), which can detect a change in groundwater velocity due to a sudden increase in the recharge rate. The wetting front migration rate determined using this method was compared with that predicted based on the change in the water table height and from changes in the relative moisture content measured

0

1

2

3

4

5

6

7

8

9

10

11

12

-50510152025

Distance relative to the south end of the gallery (m)

De

pth

be

low

gro

un

d (

m)

North South

SL1 SL2 SL3 SL4

Aa, Ab

BbBa

CbCa

DbDa

E

Fa, Fb

GbGa

HbHa

IbIa

J

Ka, Kb

LbLa

MbMa

NbNa

O

Pa, Pb

QbQa

Ra

SbSa

T

Rb

New Atlantis

Scale (10m)

SL1

SL2 and TM229 probes

SL3

SL4

Northa)

b)

TM5

TM4

TM3

TM2

TM1


with the TM229 probes at two different depths below the west gallery (Patterson and Bekele, in press).

The solute migration rate through the unsaturated zone was investigated using a two-step modelling procedure to predict the breakthrough of bromide under the conditions that existed during the tracer experiment at the depths of the suction cup lysimeters. The first modelling step involved calibrating a finite difference model of recharge below the west gallery using the wetting front migration rate predicted with the groundwater velocity probe (Patterson et al., in press). The second step was to simulate the advective flow component of the bromide tracer test, using the calibrated hydraulic parameters from the first step. The software used for the unsaturated flow and transport modelling was VS2DI, which solves Richards’ equation as described in Lappala et al. (1987) and Hsieh et al. (2000).

Transport of recharge water through the aquifer

During the first year of operation of the MAR trial in Floreat, groundwater from the site and secondary treated wastewater entering the galleries were carefully monitored to investigate temporal changes in water chemistry. The gradual shift in groundwater chemistry from background conditions (that are in equilibrium with the surrounding carbonate aquifer) to the composition of the recharge water was the primary method used to monitor the movement of the recharge water front through the aquifer. A similar approach was also used for interpreting the changes in water quality at the Halls Head site. Unlike the Floreat site, however, the groundwater composition at Halls Head had also to be interpreted in relation to mixing with seawater within the aquifer and the impact of recharge from the adjacent wastewater ponds. This made monitoring the movement of the recharged water within the aquifer at the Halls Head site difficult, compounded by the lower sampling frequency than the primary infiltration gallery site at Floreat. Therefore, all the data discussed below relating to transport of recharge water through the aquifer relates to the findings of the Floreat Infiltration Gallery site.

At the Floreat site, the analysis of trends in groundwater chloride concentrations was used to interpret flow rates through the aquifer. An unforeseen spike in the chloride concentration was detected in the source (recharge) water to the galleries in March 2006 and served as a natural tracer of groundwater movement. A major limitation of this approach to estimating travel times is that it required frequent sampling from the bores located down-gradient to define the breakthrough times at different distances. As the chloride spike was unanticipated, the sampling regime was not able to be modified in time to increase the sampling frequency. Breakthrough times to some of the bores could be determined where there were sufficient chloride data. Groundwater chloride data were also analysed to evaluate vertical gradients. This aided in confirming whether the slotted intervals in the bores intercepted the recharge plume.

In the early stages of planning the monitoring bores and the pumping rate for the recovery bore at the Floreat site, an un-calibrated, three-dimensional flow model was developed with Visual MODFLOW (version 3.1), using constant hydraulic head boundary conditions and an estimate of the hydraulic conductivity derived from the analysis of pump test data from the site. As observational data were acquired during the MAR trial, refinements to the groundwater model were made and a solute mass transport model was developed.

The MODPATH module was used to estimate travel times within the aquifer by advection using particle tracking and the MT3DMS module of MODFLOW was used to simulate mass transport of chloride by advection and dispersion. Groundwater level logger data acquired from BH6, BH8 and BH15, and groundwater chloride samples collected from BH1, BH8 and


BH16 were used to calibrate groundwater flow and solute mass transport models for the aquifer. These models simulated the saturated zone and recharge was applied directly to the water table.

Refer to Appendix 1C for details of the MODFLOW simulation for the Floreat Infiltration Galleries site.

1.2.4 Quantitative Estimate of Mixing Proportion Between Recharge Water and Background Groundwater

A simple mass balance approach was used to estimate the fraction of treated effluent in groundwater sampled from bores down-gradient from the infiltration galleries. This calculation was needed to interpret whether improvements in water quality were related to dilution from mixing with background groundwater or removal processes (i.e. biogeochemical reactions in the aquifer).

Chloride was used for this calculation as it is a conservative tracer that is unaffected by reactions induced by mixing between the two qualities of water (Cook and Herczeg, 1999). This approach assumes two end-member concentrations of chloride sourced from (1) background groundwater and (2) the treated effluent.

The unexpected chloride spike that was detected in the recharge water in March 2006 and traced in the groundwater sampled down-gradient from the galleries was used as the second end-member. The chloride spike was attenuated by dilution or mixing with background groundwater and dispersion. The latter is related to the process of water scattering as it flows along convoluted pathways around sediment grains, which produces different travel times. Dilution was likely the more important factor due to the large pumping rate at the MAR site.

The estimate of the fraction of recharge water in groundwater sampled from the bores was calculated using,

[Eq. 1]

Since the chloride concentrations in the recharge water and the groundwater sampled from the bores were time-varying, the maximum concentration predicted by modelling the chloride spike was used based on the MT3DMS chloride transport model from Section 1.2.3.

The value assigned to [Cl]recharge water in Equation 1 was 419 mg/L. This was the maximum chloride concentration in BH1 after the chloride peak of 512 mg/L was detected in the galleries as this was the first sampling point at the water table taken down-gradient and in close proximity to the infiltration galleries. It is acknowledged that some dilution may have occurred during passage through the unsaturated zone and over a couple of meters of aquifer transport from below the galleries to BH1. [Cl]sample was assigned the predicted chloride peak, attenuated by dilution and dispersion for each bore further down-gradient.

rgroundwatebackgroundwatererech

rgroundwatebackgroundsample

ClCl

ClCl

][][

][][

arg


The water sampling and chemical analysis methodologies are described in the following sections of this report.

1.2.5 Monitoring of the Infiltration Galleries and Recharged Water

Water sampling and physical measurement probes

Preparation

A sampling plan, detailing the bores to be sampled and the required analyses, was prepared and consulted prior to every sampling event. The Chemistry Centre WA and CSIRO Land and Water conducted analyses on the water samples. The Chemistry Centre was consulted to ensure that water samples were collected in the appropriate manner, in the correct types of bottles and preserved correctly.

All water sampling equipment was cleaned prior to use to avoid contamination. This included stripping the low flow pump and washing all parts with clean water and replacing any worn or dirty parts, such as filters and O-rings.

Different types of water quality monitoring probes were used, including a Troll9000, a Troll9500 and a Horiba U-10 meter. Before each sampling event, probes were rinsed with deionised water and calibrated as detailed by the manufacturers’ specifications.

Field techniques

Sampling of the bores was conducted using a low flow sampling technique with the QED Sample Pro 34” portable MicroPurge® pump (Figure 1-13). Flow rates were controlled by the MP-10 controller and generally set at 250 mL per minute, depending on conditions (e.g. turbidity levels, erratic readings). The pump was gently lowered down the bore to minimise disturbance in the water column as described in the groundwater sampling guide by ASTM (2001) and sampling took place from the midpoint of the screen.


Figure 1-13. Photo of the QED Sample Pro 34” portable MicroPurge® pump used for low flow water sampling (foreground) and accompanying controller box and hose reel. Water samples are extracted by applying compressed air to squeeze the bladder within the pump and force the sample up the discharge tube to the surface.

To avoid possible cross contamination between bores, the pump was flushed by pumping at least three volumes of the pump hose to minimise the risk of water from other bores entering the samples.

The physical parameters measured before sample collection included pH, dissolved oxygen, electrical conductivity, temperature, oxidation-reduction potential and turbidity using a Troll® 9000 multisensory meter housed within a flow cell. Water was pumped through the flow cell at a flow rate which ensured that a smooth, non-turbulent flow of water passed across the sensors. Results were recorded once the readings stabilised and then groundwater samples were collected. Measuring physical parameters before sample collection ensured that the pump was well-flushed with sample water, and the samples collected were representative of current groundwater conditions from within the aquifer.

Sample bottles that did not require preservation were filled to approximately a third of the volume, rinsed thoroughly to ensure that the sample contacted interior surfaces of the bottle and the collected water was then discarded from the bottle. This process was repeated three times before a final sample was collected by filling the sample bottle. Samples for anion and cation analysis were collected in standard polypropylene bottles while water samples for organic chemical analyses were collected in amber glass bottles.

Bottles for heavy metals were prepared with required acid preservation as directed by the Chemistry Centre. These bottles were not rinsed prior to sample collection, but were filled to the top without overflowing. Samples that required acidification to pH<2 were filled to the top and two drops of concentrated nitric acid (HNO3) were added.

Samples that required filtering were filtered using a 60 mL sterile polypropylene syringe and a disposable sterile 0.45 m nitrocellulose syringe filter. The syringe was rinsed with the sample prior to use and approximately 10 mL of sample passed through the filter prior to


sample collection. Each filter was replaced after 100 mL had passed through or if the filter became clogged with suspended material.

Microbiological samples were collected in either 500 mL autoclaved glass or plastic containers and bacteriophage samples in 5 L autoclave plastic containers. Sample bottles were opened only at the time of sampling and care was taken not to touch the inside of the containers or lids with either hands or the sampling equipment.

Sample storage, transport and quality assurance

Samples were placed in a portable cooler with ice packs immediately after collection. Water samples for microbiological analysis were delivered to the CSIRO laboratory at the end of sampling every day. As sampling from many bores at the site usually took 2 to 3 days to complete, rather than transport water samples for chemical analysis to the Chemistry Centre on a daily basis, the samples were stored in a cool room at 4C until they were delivered. The coolers were packed to ensure all samples were delivered safely and to minimise risks of damage and cross contamination.

Field blanks were taken during every sampling event to test for any gross contamination. The blanks were filled with deionised water on the morning of the first sampling day and were handled in the same way as samples with the same preservation requirements. Field blanks were used for quality control purposes for both microbiological and chemical analysis.

Water samples were delivered to an appropriate laboratory with an accompanying laboratory analysis request form which acted as a chain of custody. The forms provided details on all samples collected, the required analyses and any special requirements for specific samples.

Sample analyses: inorganic and organic chemicals and microbes

Inorganic chemicals

Water chemistry sampling from the Floreat Infiltration Galleries site was conducted on at least 40 separate occasions since the start of the MAR trial. Sampling was conducted primarily on a fortnightly basis. The selection of analyses was modified and reduced during the course of the project to focus more heavily on key parameters of interest and to avoid duplication where measured concentrations were similar in closely-spaced monitoring bores. Water sampling ended in August 2007.

A summary of the analytical procedures and detection limits for the inorganic chemistry is outlined in Table 1D-1 of Appendix 1D (Source: Environmental Chemistry Laboratory, Chemistry Centre of WA).

Organic chemicals

The water samples for organic chemistry analyses were received by the Chemistry Centre and stored at ≤ 4˚C prior to analysis. Sub-samples of the water were filtered through a 0.45 μm Acrodisc Supor membrane, into a glass test tube. A 10 mL aliquot was analysed by liquid chromatography-tandem mass spectrometry (LC-MS-MS) using on-line Solid Phase Extraction (SPE). Refer to Appendix 1E for a description of the organic chemistry analytical procedures.


Microorganisms

Analyses for all of the microorganisms were undertaken using standardised methods. The detailed methodology for the analysis of each of the microorganisms is given in Appendix 1F.

The detection of the bacteria was done using membrane filtration with the filters then placed onto selective media appropriate for the individual target microorganisms. The treated wastewater entering the infiltration galleries was tested using five 1 mL replicate samples while the groundwater samples were tested using five 100 mL replicate samples.

The male specific bacteriophage was detected in 500 mL of sample of groundwater or 50 mL of treated wastewater using the enrichment method outlined by Sobsey et al. (2004). The enteric viruses, adenovirus and the enterovirus group (which includes viruses such as poliovirus and coxsackievirus), were detected in 5 L groundwater and 1 L treated wastewater samples using the Polymerase Chain Reaction (PCR) for adenovirus and the Reverse-Transcriptase Polymerase Chain Reaction (RT-PCR) for the enteroviruses. This involved concentrating the 5 L and 1 L volumes using tangential flow - hollow fibre filtration (TFF) followed by extraction of viral nucleic acid from the sample concentrate and detecting adenovirus or enteroviruses using specific primers for these viruses.

1.2.6 Microbial Pathogen Decay Analysis

Along with the monitoring of enteric microbial indicators and pathogens in the recharged water, in the groundwater at various points between the infiltration galleries and the recovery well, a pathogen decay study was undertaken to determine the survival potential of selected enteric microorganisms in the groundwater at the Floreat infiltration gallery system. The determination of a rate of decay for specific microbial pathogens can allow a risk assessment to be undertaken to establish the exposure risk for that pathogen in the recovered water.

The microorganisms used in this study are given in Table 1-2. The culture conditions and preparation of these microorganisms for the decay study are given in detail in Appendix 1G.

Table 1-2. Microbial pathogens and indicators used in ASTR survival experiment

Microorganism Source

E. coli laboratory strain ACM11803

Enterococcus faecalis laboratory strain ACM 2517

Campylobacter jejuni laboratory strain ACM 3393

Salmonella enterica laboratory strain ACM 13311

Coxsackievirus laboratory strain Type B1

Adenovirus laboratory strain Type 3

Rotavirus faecal isolate

Cryptosporidium oocysts faecal isolate 1 ACM = Australian Collection of Microorganisms.

The pathogen decay studies were undertaken using diffusion chambers as previously described by Toze et al. (2004). A schematic of a diffusion chamber is provided in Figure 1-14 and Figure 1-15 shows a photograph of diffusion chambers used in the experiment. The chambers are constructed of Teflon and the membranes used to prevent the loss of microorganisms from within the chamber were 25 mm diameter Millipore mixed cellulose


esters (VSWP) with a pore size of 0.025 µm. This pore size is sufficient to exclude the passage of the enteric viruses (the smallest of the pathogens tested) but still allow passage of groundwater across the membrane and through the diffusion chamber. The chambers have an internal volume of approximately 7 mL.

A series of diffusion chambers were then filled with groundwater which had been seeded with either the viruses and Cryptosporidium or the bacteria. The chambers were then immediately transported back to the infiltration gallery site and suspended down monitoring bore BH9. Pathogens are retained in the chamber by the membranes, but groundwater can pass through the diffusion chamber. Representative triplicate chambers containing either the viruses and Cryptosporidium or the bacteria were collected on each sampling occasion (Day 7, 14, 21, 28 and 35). The collected samples were placed into self-sealing plastic bags (Ziploc bags) and transported back to the laboratory on ice where they were processed as detailed in Appendix 1G.

The detection of viable bacteria was determined using appropriate selective culture media. Viable Cryptosporidium oocysts were detected using DAPI and propidium iodide staining. Bacteriophage MS2 numbers were detected using the double overlay method. The presence of adenovirus was determined using quantitative (RT)-PCR (see Appendix 1G for details on methodology).

The number of each pathogen was log10 transformed and plotted over time on a graph and a regression line fitted to the plot (see Appendix 1H). The decay rate () was determined from the slope of the regression line and recorded as log per day. The T90 (time for a 90% or 1 log loss of each pathogen) was determined as 1/.


Figure 1-14. Schematic of pathogen diffusion chambers

Figure 1-15. Diffusion chambers in field prior to lowering into groundwater for survival experiment

Brass metal ends

Exclusion membrane

O-ring

Teflon Chamber

Threaded bolts


1.3 RESULTS

1.3.1 Recharge Rates and Operation of the Infiltration Galleries

The majority of this section refers to the Floreat site as the Halls Head site was decommissioned earlier than anticipated and results at the Halls Head site were disjointed due to maintenance problems occurring at this remotely operated and monitored wastewater treatment plant. Details on the interpretation of water quality and water level changes observed at Halls Head are provided in Appendix 1I.

A preliminary investigation of the infiltration capacity of the Spearwood sands at the Floreat site was undertaken using a double-ring infiltrometer with a surface area of 707 cm2 and secondary effluent as the recharge water. This preliminary investigation revealed that the characteristics of the sand were quite favourable for infiltrating large volumes. Extrapolation of the infiltrometer results to the surface area of each gallery suggested that each gallery could infiltrate approximately 955 litres per minute based on a hydraulic loading rate of 55 m/day and an infiltration area for each gallery of 25 m2. The galleries were thus designed to infiltrate only 17.5 litres per minute (1 m/day) to prevent any potential clogging of the infiltration galleries in the early stages of operation.

With the galleries receiving a continuous combined volume of 50 kilolitres per day for most of the project’s duration, approximately 36.7 megalitres of treated effluent were infiltrated via the two infiltration galleries since the commencement of the Floreat infiltration galleries on 4 October 2005 until the conclusion of recharge on 23 December 2008 (Figure 1-16). On several occasions, maintenance problems related to the supply of wastewater to the site interrupted effluent inflow to the galleries for short periods of time. The cumulative volumes of effluent inflow to the galleries were computed from the inflow rates recorded every four minutes using an Eco-graph recorder on the control skid, which controlled pumping and water flow to the galleries.


Figure 1-16. The history of cumulative effluent inflow to the infiltration galleries. The west gallery was originally gravel-filled. It was replaced with an Atlantis Leach System®, similar to the east gallery, on 19 August 2007.

West gallery changed from gravel to the Atlantis system on August 19, 2007

West gallery pressure jet cleaned to remove plant roots on May 8, 2007

Period when flow off to the East gallery (August 24 to November 20, 2007)

Period when flow off to the West gallery (August 24 to October 15, 2007)

Figure 1-17. Comparison of water levels in the east and west galleries. The periods when inflow of wastewater to the galleries was deliberately off due to the tracer tests in August–October 2007 are labelled on the plot.


During the first 13 months of operation at Floreat, both galleries recharged the same volume of wastewater. The earliest signs of clogging occurred in the gravel-filled gallery in July 2006. Subsequently over the next three months, the water level increased in the gravel gallery at a rate of approximately 0.6 mm/day, whereas the water level in the Atlantis Leach System® gallery remained relatively constant (Figure 1-17). During the first 12 months of operation, roots, possibly from grass, were observed entering the central discharge chamber via the pipes discharging the treated wastewater into either side of the gravel-filled infiltration gallery. While it was never absolutely clear what actually was causing the clogging in the gravel-filled gallery, the presence of plant roots would have, at least, caused some of the problem (see Figure 1-18 for evidence of plant roots in the gravel-filled gallery). No plant roots were observed in the central discharge chamber of the gallery constructed using the Atlantis Leach System®, although they potentially were growing in the voids of the crates. Even if this occurred, it did not have any noticeable effect on infiltration from the gallery constructed using the Atlantis Leach System®.

Figure 1-18. Evidence of plant roots growing within the gravel-filled infiltration gallery at the Floreat infiltration galleries. The scored PVC discharge pipe is visible under the gravel. Photo taken during the conversion of the gravel filled gallery to the Atlantis Leach System®.

From about November 2006, severe decline in infiltration rates was observed for the gravel gallery, principally due to clogging within the gravel matrix. This caused the management system to automatically shut off the delivery of wastewater to the galleries. The clearing of roots from the gravel gallery using pressure jet cleaning was conducted on May 8, 2007. This was not entirely effective as shown by the resumption of water level rise (Figure 1-17). The same rate of inflow to the gravel gallery continued until 9 August 2007 whereupon inflow to both galleries was stopped to install equipment for monitoring flow through the unsaturated zone. This also provided an opportunity to replace the gravel gallery with Atlantis Leach System® crates similar to the east gallery. Visual inspection of the gravel as it was removed did not show excessive grass root growth amongst the gravel or clogging of the slots of the discharge pipe. Thus the reason for clogging of the gravel gallery remains unclear. Effluent inflow to the galleries was deliberately turned off on 24 August 2007 for the tracer experiments and resumed on 15 October 2007 in the west gallery and 20 November 2007 in the east gallery.


Figure 1-19. A comparison of water levels in the two infiltration galleries at Halls Head

Operation of the Floreat infiltration galleries concluded on 23 December 2008. During the final two months of operation, the east Atlantis Leach System® gallery was shut-off and the inflow rate to the west gallery was increased from 17.5 L/min to between 50 and 67 L/min to test the feasibility of increasing infiltration rates (Figure 1-16). An accurate setting of the inflow rate could not be obtained due to mechanical problems. On average the inflow rate to the west gallery was 60 L/min. There was no evidence of clogging in the Atlantis Leach System® gallery during this two month period of observation.

Since the commencement of the Halls Head galleries, approximately 11.5 megalitres of treated effluent were infiltrated via the two infiltration galleries. During the operation of the site over a 22 month period, there were daily shut-off periods, which prevented water levels from rising to excessively high levels and although clogging was not observed, the overall water level trends shown in Figure 1-19 also reveal higher rates of water level rise in the gravel-filled gallery at Halls Head. An analysis of water level changes in the two galleries revealed that the gravel gallery was more responsive to changes in effluent inflow and a decline of about 1 mm/day was observed during lengthy shut-off periods (Figure 1-19).

1.3.2 Determination of Recharge through the Unsaturated Zone

The unsaturated zone beneath the Floreat infiltration galleries has been fairly well characterised from coring and analysis of the sediments. There is roughly 7 m of Spearwood sand overlying Tamala Limestone, depending on the cored location. The depth to groundwater in the unconfined aquifer varies mainly due to proximity to the extraction bore. There is no evidence to suggest that a substantial increase in the water table occurred beneath the galleries due to recharge. The water table is generally below a depth of 10 m below ground surface, but seasonal variations are on the order of 1 m as shown in Figure 1-20.


1.7

1.8

1.9

2

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.8

2.9

3

3.1

4/10/05 8/01/06 14/04/06 19/07/06 23/10/06 27/01/07 3/05/07 7/08/07 11/11/07 15/02/08 21/05/08 25/08/08 29/11/08

Wa

ter

tabl

e e

leva

tion

(mA

HD

)

BH1

BH2

BH5

BH3

BH7

BGRND1

Figure 1-20. Water table elevations measured at bores BH1, BH2, BH3 and BH5 and logged with Odyssey capacitance water level probes in bores BH7 and BGRND01

Tracer tests using chloride and electrical conductivity signatures of the recharged treated effluent, bromide and the fluorescent compound uranine were undertaken, beginning in August 2007 to further characterise the unsaturated zone. Just prior to conducting the tracer tests, the depth to water was 10.60 0.1 m. The tracer tests were conducted using a slightly lower hydraulic loading rate as when effluent was used (17.0 litres per minute or 0.98 m/day) and to the west gallery only. The profile was wetted prior to conducting the tracer test with steady infiltration occurring to both galleries for six days prior to when the tracers were introduced to the west gallery. Effluent was switched off to both galleries for the duration of monitoring tracer breakthrough as labelled in Figure 1-16 and groundwater was used to infiltrate tracers into the west gallery, while effluent recharge to the east gallery remained off. Effluent recharge commenced on 15 October 2007 for the west gallery and 20 November 2007 for the east gallery.

The analysis of breakthrough curves from the main tracer experiment, which involved infiltrating bromide and uranine concurrently, revealed a considerable amount of information about variability in flow velocities within the unsaturated zone as well as differences in the results using uranine versus bromide. A complete report on the tracer experiments for the unsaturated zone along with stratigraphic profile and soil moisture data are available in Wolfgang Feggs Diploma Thesis in Appendix 1J, however, a brief synopsis of the results is provided here.


Presumably the low fluorescence sampled in water from the SL4 lysimeters (refer to Figure 1-12) was due to recharge water not reaching one of the furthest points from the midpoint of the gallery where water is discharged and allowed to drain laterally into the gallery. The rise in fluorescence in water sampled from the deepest lysimeter (SL4_T) toward the end of monitoring suggests that the results might be sensitive to preferential flow and/or less head of water in the gallery above this point. The wetting front will travel slower if the pores are not saturated to the same extent, which may be a consequence of water not discharging fully to this distal part of the gallery.

The mean effective flow times (tmean) were determined using numerical integration to sum the mass of tracer below the breakthrough curve for each suction cup lysimeter (Table 1-3). The tmean values for monitoring station SL4 could not be ascertained as there was very low fluorescence detected overall and only the beginning of a breakthrough curve for SL4_T, the deepest suction cup lysimeter, toward the end of monitoring (Figure 1-21).

0

2000

4000

6000

8000

10000

12000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

fluor

esc

ence

[pp

b]

SL1_Ab at 0.50m

SL1_Bb at 2.51m

SL1__Cb at 4.84m

SL1_Db at 6.42m

SL1_E at 8.80m

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

fluo

resc

enc

e [p

pb]

SL2_Fa at 0.5m

SL2_Gb at 2.63m

SL2_Hb at 4.44m

SL2_Ha at 4.94m

SL2_Ia at 6.81m

SL2_J at 10.65m

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

fluo

resc

en

ce [p

pb

]

SL3_Ka at 0.50m

SL3_Kb at 0.50m

SL3_Lb at 2.71m

SL3_Mb at 4.67m

SL3_Nb at 6.62m

SL3_O at 8.89m

0

5

10

15

20

25

30

35

40

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

fluor

esce

nce

[pp

b]

SL4_Pb at 0.50m

SL4_Qb at 2.51m

SL4_Rb at 4.57m

SL4_Sb at 6.42m

SL4_Sa at 7.01m

SL4_T at 8.64m

Figure 1-21. Uranine breakthrough curves and data for the four lysimeter stations. The legend in each plot refers to the depth below the west gallery. Refer to Figure 1-12 for locations of the lysimeters.

Uranine breakthrough curves were also analysed for two boreholes, BH18 and BH6 (Figure 1-22). The data for these boreholes represents a combination of travel through the unsaturated zone and the aquifer. BH6 is located 5 m away from the edge of the west gallery, whereas the location of BH18 is less exact and may be up to 0.9 m from the west gallery. The effective flow times (tmean) for BH18 and BH6 were 7.9 days and 11.2 days, respectively.


0

250

500

750

1000

1250

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

fluor

esce

nce

[pp

b]

BH6

BH18

Figure 1-22. Uranine breakthrough curves for two groundwater boreholes located within a few meters down-gradient relative to the west gallery

Table 1-3. Mean effective flow times (tmean) interpreted based on the breakthrough curves for the uranine and bromide tracers

Uranine tracer

Bromide tracer

Suction cup lysimeter

Depth below west gallery (m)

tmean (days) tmean (days)

SL2_Fa 0.50 0.82 --

SL1_Ab 0.50 1.05 --

SL3_Ka 0.50 0.68 --

SL3_Kb 0.50 0.68 --

SL1_Bb 2.51 1.46 --

SL2_Gb 2.63 1.65 --

SL3_Lb 2.71 1.36 --

SL2_Hb 4.44 2.80 --

SL3_Mb 4.67 3.21 --

SL1_Cb 4.84 2.67 --

SL2_Ha 4.94 3.99 --

SL1_Db 6.42 2.28 --

SL3_Nb 6.62 5.16 3.28

SL2_Ia 6.81 3.52 2.97

SL1_E 8.80 5.05 --

SL3_O 8.89 4.57 --

SL2_J 9.15 7.42 -- (--) Indicates not measured.


The average tmean for the other three deepest sampling points based on the uranine results is 5.7 days 1.5 days. This is a possible minimum estimate of the mean effective flow time from the base of the gallery to the water table as there was at least another meter of unsaturated zone beneath these three deepest sampling points; however, this assumes that uranine acted as a conservative tracer. If there is sorption of uranine, then a conservative tracer such as bromide will yield a lower estimate for the travel time.

The analysis of bromide data for SL2_Ia and SL3_Nb indicates shorter travel times to depths between 6.6 and 6.8 m below the west gallery as compared to the uranine data (Table 1-3 and Figure 1-23). Bromide is considered to be a more conservative tracer and sorption of uranine is a possible reason for the difference in travel times. The mean pore velocities for SL2_Ia and SL3_Nb based on the bromide results are 2.29 m/day and 2.02 m/day, respectively. The average, mean pore velocity from the bromide field data is thus, 2.16 m/day 0.19 m/day. The application of the numerical model VS2DI to simulate the bromide tracer experiment produced a comparable estimate of 2.75 m/day 0.01 days based on SL2_Ia and SL3_Nb. These results are also supported by the analysis of changes in the groundwater velocity as measured in BH18 (Patterson et al., in press).

Although the samples selected for bromide were not taken from lysimeters near the water table, using the average, mean pore velocity from SL2_Ia and SL3_Nb, the estimated travel time through 9.1 m of unsaturated zone beneath the gallery is 4.24 days 0.38 days. This is a more reliable minimum estimate of the residence time in the unsaturated zone than the estimate obtained using uranine as there is a greater potential for uranine to sorb to the porous media.

The application of the numerical model VS2DI to simulate the bromide tracer experiment produced an estimate of the average, mean pore velocity of 2.75 m/day 0.01 days based on SL2_Ia and SL3_Nb. Similarly, the results from interpreting the groundwater velocity probe data produced comparable results.


Bromide and uranine tracer results at SL2_Ia

0

10

20

30

40

50

60

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

mg/

l

Br

Uranine mg/l*10

Bromide and uranine tracer results at SL3_Nb

0

10

20

30

40

50

60

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Time (days)

mg/

l

Br

Uranine mg/l*10

Figure 1-23. Bromide versus uranine breakthrough curves for lysimeters SL2_la and SL3_Nb


The variation in mean pore velocities calculated using the uranine data reveals the extent of heterogeneity. As shown in Figure 1-24, the mean pore velocity varies non-uniformly with depth and the profile of velocities varies between different stations. The highest mean pore velocity was 2.8 m/day at 6.42 m below the west gallery (SL1_Db). Although the suction cups in SL2 and SL3 are located roughly the same distance from the point where water is discharged into the west gallery, the depth profile of velocities is quite different, particularly at depths greater than 6 m below the gallery. This pattern is most likely related to the transition from Spearwood sand to Tamala Limestone.

Figure 1-24. Distribution of mean pore velocities as determined from uranine breakthrough curve data (tmean) from each of the suction cup lysimeters

1.3.3 Recharge Water Residence Time in the Aquifer

Transition in major ion chemistry

A contrast in salinity and major ion composition existed between the initial groundwater chemistry at the Floreat site and the treated effluent infiltrated to the Tamala aquifer. The salinity of the treated effluent was marginally higher than the ambient groundwater, with total dissolved solids of 750 and 640 mg/L respectively, and comprised predominantly of sodium chloride making chloride a useful tracer for the migration of the infiltrated effluent. This contrast in water quality provided a means of estimating the breakthrough time of the recharge water at locations down-gradient of the galleries by monitoring when the groundwater chemistry sampled from the monitoring bores resembled that of the recharge water. Some bores did not show a transition in groundwater chemistry due to either being slotted below the recharge plume (i.e. BH11) or because they contained a large proportion of background groundwater due to proximity to the extraction bore and its high pumping rate (e.g. BH15).

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00

mean pore velocity (m/day)

de

pth

un

der

th

e w

est

gal

lery

(m

)

SL1

SL2

SL3


Figure 1-25. Ternary diagram showing the relative proportions of Ca, Mg and Na in water sampled from the galleries and the bores. Groundwater evolved from the composition indicated (circled region) toward the composition of the recharge water infiltrated by the galleries during the first year.

Figure 1-26. Ternary diagram with data sampled from monitoring bores between early October and late November 2005. Groundwater from BH9 was the furthest sample from the galleries to have evolved significantly by the end of October 2005.

Concentrations in mg/l %


The treated effluent contained a higher proportion of sodium (Na) (80–90%) relative to calcium (Ca) and magnesium (Mg), while groundwater in equilibrium with the limestone aquifer contains a higher proportion of Ca (50–60%) as shown by the samples from the background bore, BGRND01 and initial samples collected prior to the commencement of MAR at the site (Figure 1-25). By approximately fifty days after the MAR trial had started, groundwater sampled from bores located 7.5 m west of the galleries (e.g. BH9) were among the furthest sites to reveal having appreciably changed from the Na:Ca:Mg ratios of the recharge water (Figure 1-26). Bores located this far from the galleries were not sampled frequently this early (1 to 2 months into the MAR trial) compared with bores close to the galleries. Consequently, it was not possible to accurately determine exact travel times within the aquifer based on when changes in the proportions of Na, Ca and Mg occurred.

Table 1-4. Comparison of average [and standard deviations] of calcium, chloride, sodium and potassium concentrations in the recharge water sampled from the infiltration galleries and the background groundwater sampled from BGRND01, BGRND02 and BGRND03 in the limestone aquifer during the first year of the MAR trial

Water sources Calcium

(mg L-1)

Chloride

(mg L-1)

Sodium

(mg L-1)

Potassium

(mg L-1)

Recharge water in the infiltration galleries

28.6 [±9.5] 245.29 [±62] 194.00 [±44] 22.9 [3.5]

Background groundwater from BGRND01, BGRND02 and BGRND03

98.8 [±23.5] 162 [±27] 92.6 [±14.5] 4.96 [0.83]

Ratio of average concentration in recharge water to background groundwater

0.29 1.51 2.10 4.62

The average concentrations of Ca, chloride (Cl), Na and potassium (K) in the recharge water and background groundwater (before the MAR trial) are distinctly different as revealed in Table 1-4. The recharge water had an average calcium concentration that was one-third the concentration in the background groundwater sampled from the limestone aquifer, but was greater than fourfold more enriched in potassium, which has been used as a tracer of wastewater recharge, in particular to help detect leaking from sewage systems (Wolf et al., 2006; Rueedi et al., 2009).

As shown in Figure 1-27, the ratios for groundwater Ca sampled from background bore BGRND01 relative to groundwater Ca sampled from bores located at the MAR site varied depending on the bore, but were not as low as the ratio based on the recharge water (0.29, Table 1-4). Data from BGRND01 were used in this analysis because the other two background monitoring bores were installed later and were first sampled in December 2006.


Year 1 data

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

25/10/2005 14/12/2005 2/02/2006 24/03/2006 13/05/2006 2/07/2006 21/08/2006 10/10/2006

Rat

io o

f ca

lciu

m in

bo

re w

ater

to

av

erag

e ca

lciu

m c

on

cen

trat

ion

in B

GR

ND

01

BH3BH5BH8BH10BH11BH13BH14BH16BH17 (extraction bore)

Figure 1-27. Variations in the ratio of calcium in the bore water relative to the average calcium ratio in BGRND01 during the first year of the MAR trial

The following formula was used to calculate the ratio:

Ca ratio = 01

][

BGRNDinCaAverage

borefromsampledCa

The ratios were generally greater than 0.5 during the first year of the MAR trial due to the large amount of dilution from pumping from the extraction bore. The ratios for BH11 and BH17 remained near 1, being predominantly background groundwater. The ratios for BH13 decreased steadily, and similarly for BH3, BH5, BH14 and BH16, as one would expect with greater amounts of recharge water being present as the MAR trial progressed. The ratios for other bores (not shown), particularly bores close to the galleries, were closer to 0.5 and more variable with no clear temporal trend. The Ca ratios for BH8 and BH10 decreased from 0.63 to 0.5 within the first month of sampling, but the ratios then increased after December 2005 and decreased again after April 2006. It is not clear why this pattern occurred, but it may be related to the recharge being turned off for a month in December 2005 and/or changes in the calcium concentrations of the recharge water or the extent of calcite dissolution that has occurred. The groundwater toward the recharge end of the MAR site would likely have been more sensitive to variations in the recharge water concentrations, whereas the groundwater further down-gradient would likely show the predominance of background groundwater (ratios closer to 1) and steady, long-term trends in water chemistry (i.e. BH13).

Potassium in the monitoring bore and recovery bore increased due to the presence of effluent recharge (Figure 1-28). Bores close to the recharge galleries (BH1, BH8) illustrated variable potassium concentrations reflecting the temporal variability in the potassium concentration of the effluent. Further along the flow path (BH12, BH16, BH17), showed mixing between the average background and effluent signatures.


0

5

10

15

20

25

30

0 50 100 150 200 250 300 350 400

Cl (mg/L)

K (

mg/

L)

Background

GalleriesBH1BH8

BH12BH16BH17

Mixing

Figure 1-28. Plot of potassium versus chloride showing the extent of mixing between ambient groundwater sampled from the background bores and the recharge water from the galleries. The error bars indicate +/- one standard deviation from the average concentrations that were sampled.

Although the groundwater major ion chemistry data provided qualitative evidence of breakthrough of the recharge plume at successive locations down-gradient from the galleries, there were not enough data to accurately determine breakthrough times at intermediate bores between the galleries and the extraction bore. Ideally, one would acquire the breakthrough time at the extraction bore, but the large pumping rate created considerable dilution which made it difficult to distinguish the chemical signature of the recharge water from that of the background groundwater.

If breakthrough times were obtained at intermediate bores between the galleries and the extraction bore, one could estimate groundwater velocities at different distances along the flow path to the extraction bore and infer the time needed for the recharge water to travel 50 m from the galleries. This approach is relatively crude as one would need to account for the increase in groundwater velocity toward the extraction bore where the effects of pumping on groundwater velocities are greatest, as well as cation exchange and calcite dissolution in the aquifer.

Quantitative modelling of groundwater transport in the Tamala aquifer

A quantitative modelling approach was used to predict breakthrough times for chloride and the mixing fraction of recharge water present in groundwater sampled from bores down-gradient from the galleries.

A range of minimum travel times between 43 and 107 days was obtained for water that was recharged to the aquifer and recovered 50 m down-gradient at the extraction bore,


depending on the transport parameters used in the model (Table 1-5). These results were based on models that apply the transport parameters uniformly to the aquifer, but heterogeneity and lateral variation in aquifer properties within the sandy limestone was revealed by the model calibration results.

Table 1-5. Minimum travel times within the aquifer predicted by MODPATH

Model Permeability (m/day)

Porosity (%) Minimum travel time (days)

1 30 20 72

2 75 20 50

3 100 20 43

4 30 30 107

5 75 30 75

6 100 30 65

The selection of transport parameters in Table 1-5 was based on the two steps of model calibration. During the first step, the analysis of monitoring data revealed that shutting off the pump for two days increased the water table elevation by at least 3.5 cm as shown in Figure 1-29 for three bores. A comparison of model results based on different permeability values for the Tamala limestone reveal that these observed changes in water levels cannot be modelled using a uniform permeability. As shown in Figure 1-29, the calculated groundwater levels using a permeability of 30 m/day were reasonably well matched to the observed groundwater levels for bores BH6 and BH8. In contrast, the model over-predicted the magnitude of water table rise for bore BH15. As shown in Figure 1-30, the observed water table response in bore BH15 was better matched in the model by assigning a permeability value between 75 and 100 m/day. Using a permeability of this magnitude under-predicted the water table rise in bores BH6 and BH8.

The second stage of model calibration based on the mass transport of chloride reaffirmed lateral variations in aquifer properties. The plots in Figure 1-31 reveal that a permeability of 30 m/day provided a better fit to the breakthrough data for BH8, whereas the breakthrough data for BH16 were better fit with a permeability of 100 m/day. Overall, a dispersivity of 1 m appeared to fit the data better. It was not possible to resolve porosity variations within the aquifer with the calibration data. Model results with 20% versus 30% produced subtle differences in breakthrough times predicted at BH8, whereas a greater difference in travel times was predicted at BH16 using 20% versus 30% for porosity.

The maximum chloride concentration for the recovery bore (BH17) was 202 mg/L, which is much lower than the data from other up-gradient monitoring bores that intersected the recharge plume. The high rate of pumping entrained water from a broad domain and only a fraction of which was influenced by the gallery recharge, thus the chloride signature in the recovered water was greatly masked by dilution. The data and model comparisons shown in Figure 1-32 do not help with calibration; however, it appears that the permeability and porosity estimates are in the approximate range indicated by the models.

In summary, although spatial variations in permeability and porosity within the Tamala Limestone could not be resolved by modelling with the available calibration data, the average minimum travel times predicted across the range of plausible combinations of transport parameters is around 70 days (Table 1-5).


0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 1 2 3 4 5

Day number

Gro

un

dw

ater

le

vel

rela

tive

to

dat

um

(cm

)

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 1 2 3 4 5

Day number

Gro

un

dw

ater

le

vel

rela

tive

to

dat

um

(cm

)

0

2

4

6

8

10

0 1 2 3 4 5

Day number

Gro

un

dw

ate

r le

vel r

ela

tiv

e to

da

tum

(cm

)

BH6/A(Observed)BH6/A(Calculated)



Pump off Pump on

Figure 1-29. Groundwater level changes in bores BH6, BH8 and BH15 in response to shutting off the pump for maintenance. Calculated groundwater levels are from a MODFLOW simulation based on a permeability of 30 m/day for the Tamala Limestone.


0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 1 2 3 4 5

Day number

Gro

un

dw

ater

lev

el r

elat

ive

to d

atu

m (

cm)

BH15/A(Observed)

BH15 Model Calculated; K = 75 m/day

BH15 Model Calculated; K = 100 m/day

Pump off Pump on

Figure 1-30. The same pumping scenario as shown in Figure 1-29. Model calculated groundwater levels are shown using a permeability of 75 and 100 m/day for the Tamala Limestone.


Figure 1-31. Comparison of chloride model results for bores BH8 and BH16 with different combinations of permeability, porosity and longitudinal dispersivity

BH8Longitudinal Dispersivity = 0

145

165

185

205

225

245

265

285

305

325

345

365

385

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl (

mg

/l)

K=30 m/day; phi=0.2; D=0

K=100 m/day; phi=0.2; D=0

K=30 m/day; phi=0.3; D =0

K=100 m/day; phi=0.3; D=0

BH8/A(Observed)/Chloride

BH8 Longitudinal Dispersivity = 1m

145

165

185

205

225

245

265

285

305

325

345

365

385

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)

K=30 m/day; phi=0.2; D=1m

K=100 m/day;phi=0.2; D=1m




BH16Longitudinal Dispersivity = 1m

145

165

185

205

225

245

265

285

305

325

345

365

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)






BH16Longitudinal Dispersivity = 0

145

165

185

205

225

245

265

285

305

325

345

365

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)

K=30 m/day; phi=0.2; D=0

K=100 m/day; phi=0.2; D=0

K=30 m/day; phi=0.3; D =0

K=100 m/day; phi=0.3; D=0


A.

B.

C.

D.


BH17Longitudinal Dispersivity = 1m

145

160

175

190

205

220

235

250

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)





BH17 observed Cl data

Figure 1-32. Comparison of chloride model results for BH17 with different combinations of permeability and porosity with a longitudinal dispersivity of 1 m

1.3.4 Dilution with Background Groundwater Based on Chloride Model

As shown in Figure 1-33, the predicted chloride breakthrough curves from two probable models for the Tamala aquifer indicate attenuation of the chloride spike and a lag in arriving at different bores relative to BH1. The travel time and chloride concentration of the peak ( sampleCl][ from Equation 1) depend on the hydraulic parameters used in the model. Table 1-6

summarises the results from estimating the fraction of recharge water in groundwater sampled from bores that were regularly sampled for chloride, using Equation 1 (Section 1.2.4). The estimated fractions of recharge water for each bore are reasonably similar based on the two models, except for BH11 and BH16. As shown in Figure 1-34, the model with a permeability of 100 m/day produced a better fit to the observed chloride data for BH11 and BH16 than the model using 30 m/day at these two bore locations. Thus, the estimated fractions of recharge water for BH11 and BH16 are most likely 0.16 and 0.44, respectively, based on the higher permeability model. The percentage of recharge water extracted from the recovery bore (BH17) was on the order of 20%, which is reasonable considering that the pumping rate was five times the recharge rate.

The uniform permeability models with 30 m/day and 100 m/day for the Tamala Limestone are shown as they appear to fit the chloride data better at the east and west ends of the field site, respectively. A higher permeability field produces shorter travel times and thus a worst-case scenario for breakthrough of contaminants. Table 1-6 also reveals a larger proportion of recharge water is predicted at each of the down-gradient monitoring bores according to the 30m/day permeability model.

The plume structure is affected by permeability with shallower penetration of gallery recharge into the aquifer for the higher permeability case (Figure 1-33). This is an artefact of the relatively high pumping rate in the recovery bore dominating velocities in 3D. It is recognised that this not a unique solution, and for example, a lower permeability and anisotropy may produce similar results. However, for simplicity, the isotropic, high permeability model was adopted as providing a conservative estimate for travel time from the galleries to the recovery well.


Figure 1-33. Predicted chloride breakthrough curves for the Tamala aquifer based on different model parameters: (A) permeability of 30 m/day; (B) permeability of 100 m/day. A porosity of 30% and a longitudinal dispersivity of 1 m were applied in both.

Table 1-6. Estimated proportion of recharge water in groundwater

Bore Estimated fraction of recharge water in groundwater based on model with permeability of 100 m/day

Estimated fraction of recharge water in groundwater based on model with permeability of 30 m/day

Average fraction of recharge water in groundwater

[Standard Deviation]

BH1 1.00 1.00 1.00 [±0.00]

BH6 0.94 0.95 0.95 [±0.00]

BH7 0.76 0.79 0.77 [±0.02]

BH8 0.69 0.74 0.72 [±0.04]

BH10 0.67 0.69 0.68 [±0.01]

BH11 0.16 0.52 0.34 [±0.25]

BH13 0.57 0.66 0.62 [±0.06]

BH15 0.55 0.65 0.60 [±0.07]

BH16 0.44 0.63 0.54 [±0.13]

BH17 0.19 0.24 0.21 [±0.04]

145

165

185

205

225

245

265

285

305

325

345

365

385

405

425

445

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)

BH1

BH6

BH7

BH8

BH10

BH11

BH13

BH15

BH16

BH17_extraction_bore

145

165

185

205

225

245

265

285

305

325

345

365

385

405

425

445

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)

BH1

BH6

BH7

BH8

BH10

BH11

BH13

BH15

BH16

BH17_extraction_bore

A.

B.


Figure 1-34. Chloride breakthrough curves from two probable models at the locations of (A) BH11 and (B) BH16 compared with groundwater chloride concentrations measured at these locations

1.3.5 Water Quality Changes

Inorganic chemicals and basic water quality parameters

Some of the inorganic chemistry results were reported in Section 1.3.3 to analyse temporal trends showing the transition from background groundwater to recharge water in the aquifer. This section provides results on inorganic chemistry sampling and field measurements that are relevant to evaluating the suitability for irrigation water quality.

Physical and chemical characteristics of water sampled or measured from the FIG site are summarised in Table 1-7. Not all of the bores are included for simplicity. The slotted intervals in the monitoring bores (BH1, BH8, BH12, BH16 and BH17) shown in Table 1-7 intersect the flow-path of the recharge plume, whereas the three background bores are up-gradient. The recharge water temperature varied seasonally with an average of 24ºC and similar temperature fluctuations were observed in some of the monitoring bores down-gradient, i.e. BH1 and BH8. In contrast the background groundwater was slightly cooler. The electrical conductivity, pH, ORP, Eh, dissolved oxygen concentrations and sulphate concentrations of the recharge water were very similar to background groundwater. The deeper recovery bore

145

165

185

205

225

245

265

285

305

325

345

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)

BH16 observed Cl data

Model: K=100 m/day

Model: K=30 m/day

A.

B.

145

165

185

205

225

245

265

285

305

100 150 200 250 300 350 400 450 500 550 600

Day number

Cl

(mg

/l)BH11 observed Cl data

Model: K=100 m/day

Model: K=30 m/day


(BH17) illustrated slightly more reduced conditions (DO 2.8 mg/L, Eh 170 mV SHE) than evident for the background bores (DO 4.0 mg/L, Eh 320 mV SHE). Iron concentrations in the ambient groundwater exceeded that of the recharge water and were also higher in the more reduced conditions at BH17 than in the background bores due to the increased solubility of iron under these conditions. Total and dissolved organic carbon, nitrogen, phosphate and boron were all higher in the recharge water compared to background. Suspended solid and total dissolved solid concentrations were highly variable in the source water and the range of variation overlapped with the range measured in the background groundwater.

For the purpose of evaluating the water quality sampled from the FIG site, maximum concentrations for the analytes listed in the guidelines for irrigation water are provided in Table 1-8. According to the ANZECC & ARMCANZ (2000) document, the long-term trigger value refers to the maximum concentration in the irrigation water which can be tolerated assuming 100 years of irrigation, whereas the short-term trigger value refers to the maximum concentration in the irrigation water which can be tolerated for a shorter period (20 years). Guidelines for beryllium and lithium are referred to in the document, but they were not monitored at the site.

Although the guidelines do not specify trigger values for different species of nitrogen, the results for different forms of nitrogen are provided for comparison with total nitrogen. The results for ammonia, nitrate, total Kjeldahl nitrogen (the latter referring to the sum of organic nitrogen and ammonia) are all expressed in terms of nitrogen.

Overall, there are twenty analytes compared with guideline values in Table 1-8. None of the water samples from the galleries and the bores exceeded the short-term irrigation trigger values during the period of sampling for any of the analytes apart from phosphate. Phosphate only exceeded the trigger values in the wastewater entering the galleries and in the bore closest to the galleries (BH01). All samples collected from all other bores away from the galleries were always less than the long-term trigger values. The long-term trigger values were exceeded for some of the other analytes at some of the monitoring points. Apart from iron, however, all of the values that were higher than the long-term trigger values occurred in the monitoring bores closest to the infiltration galleries. Iron and Total Nitrogen were the only analytes in the samples taken from the monitoring bores from BH11 to the recovery bore (i.e. from mid way through to the recovery bore) that had any values above the trigger values and Total Nitrogen was only high in one bore (BH14). Iron is known to be a problem in many bores on the Swan Coastal Plain and the high values obtained are a reflection of management issues that are needed for any pumping bore on the Swan Coastal Plain, not just for MAR schemes. Although the iron concentration in the recovery bore is elevated above levels recorded in the background bores, this is related to the greater depth of this bore and the more reduced conditions. BH11, the second deepest monitoring bore at the site (slotted between 18.6 and 19.6 m below ground), has a large component of background groundwater and also had quite high iron concentrations.


Table 1-7. Water characteristics sampled from the Floreat Infiltration Galleries site. The mean and (SD) values are indicated for each of the parameters*.

Parameter (Unit of Measure) East and West

galleries BH1 BH8 BH12 BH16 BH17

Background bores

Water temperature (oC) 24 (±4) 24 (±4) 23 (±3) 23. (±2) 22 (±1) 22 (±2) 22 (±1)

pH 7.33 (±1.11) 7.58 (±0.78) 7.42 (±0.69) 7.45 (±1.43) 7.32 (±1.08) 7.41 (±1.07) 7.04 (±0.88)

Suspended Solids (mg L-1) 7.5 (±7.05) 56.4 (±45.38) 8.6 (±4.98) 8 (±2.83) 12 (±4.24) 7.25 (±3.59) 2 (±3.46)

Total Dissolved Solids (mg L-1) 755 (±179) 831 (±140) 784 (±116) 773 (±85) 763 (±29) 681 (±17) 644 (±25)

Electrical Conductivity (mS m-1) 147 (±57) 154 (±54) 148 (±58) 155 (±69) 140 (±54) 131 (±48) 129.09 (±65.69)

Eh (mV-SHE) 385 (±184) 393 (±160) 354 (±175) 285 (±161) 275 (±161) 169 (±121) 321 (±189)

Dissolved Oxygen (mg L-1) 2.15 (±1.82) 6.6 (±2.53) 5.87 (±3.28) 3.28 (±2.27) 4.61 (±2.73) 2.8 (±1.69) 4.02 (±2.56)

Cl (mg L-1) 245 (62) 247 (51) 232 (41) 223 (24) 184 (24) 186 (8) 162 (27)

HCO3 (mg L-1) 174 (24) 273 (23) 278 (8) 292 (30) 282 (39) 291 (8) 276 (59)

SO4 as S (mg L-1) 64.1 (±8) 63.0 (±7) 66.1 (±8) 81.2 (±10) 75 (±11) 59 (±6) 64.5 (±18)

Na (mg L-1) 194 (44) 183 (51) 184 (44) 156 (+42) 135 (34) 117 (10) 92.6 (15)

Ca (mg L-1) 28.6 (9.5) 63.7 (12.6) 68.8 (14.1) 94.7 (13.2) 87.5 (+18.7) 96.6 (5.3) 98.8 (23.5)

Mg (mg L-1) 11.6 (4.5) 12.2 (4.3) 11.5 (2.6) 13.0 (1.3) 13.5 (2.2) 15.0 (0.7) 13.2 (3.2)

K (mg L-1) 22.9 (3.5) 20.6 (5.4) 19.8 (5.4) 18.5 (+5.4) 11.9 (3.5) 9.3 (1.8) 4.96 (0.83)

Fe (mg L-1) 0.14 (0.20) 0.049 (0.052) 0.18 (0.49) 2.2 (1.7) 1.1 (1.0) 2.6 (1.4) 0.44 (0.72)

Mn (mg L-1) 0.036 (0.016) 0.0043 (0.005) 0.035 (0.056) 0.043 (0.021) 0.045 (0.019) 0.021 (0.002) 0.013 (0.009)

B (mg L-1) 0.21 (0.08) 0.21 (0.09) 0.18 (0.06) 0.12 (0.06) 0.069 (0.036) 0.052 (0.024) 0.085 (0.090)

TOC (mg L-1) 9.98 (±3.8) 6.32 (±3.53) 5.67 (±3.02) 3.57 (±1.51) 3.14 (±3.53) 4.88 (±3.79) 2.51 (±2.72)

DOC (mg L-1) 10.86 (±2.35) 6.27 (±3.04) 5.11 (±3.02) 4 (±1) 5 (±2.83) 5.09 (±2.91) 2.73 (±2.15)

N_Total (mg L-1) 4.27 (±1.9) 4.78 (±2.02) 3.96 (±2.13) 0.44 (±0.44) 0.16 (±0.12) 1.58 (±0.49) 0.29 (±0.36)

NO3_N (mg L-1) 2.16 (±1.41) 3.72 (±1.68) 3.28 (±2.13) 0.25 (±0.41) 0.02 (±0.02) 0.99 (±0.43) 0.15 (±0.25)

Soluble reactive phosphorus as P (mg L-1) 6.31 (±3.32) 1.96 (±1.6) 0.01 (±0.01) <0.005# <0.005# 0.01 (±0.01) 0.01 (±0.01)

* For the number of samples used to determine mean and standard deviation for each analyte at each monitoring bore see Table 1D-2 in Appendix 1D. # Below detection limit.


During the first year of sampling, there were water samples that exceeded the long-term trigger values for arsenic (BH01), iron (gallery water, BH11 to BH16, the recovery bore, and BGRDND01), uranium (BH08), and total nitrogen (gallery water, BH01, BH02, BH06, BH08, BH09 and BH10). During the second year, there were water samples that exceeded the long-term trigger values for iron (gallery water, BH01, BH11, BH13, BH15, BH16, the recovery bore, all three background bores), and total nitrogen (gallery water, BH01, BH03, BH06, BH08, BH10 and BH14). The high concentration of total nitrogen is mainly due to nitrate. In some instances the total nitrogen concentration in the groundwater exceeded that of the recharge water due to the impacts of diffuse pollution.

The minimum, maximum and average concentrations for a selection of analytes from the ANZECC & ARMCANZ (2000) list for irrigation water quality are shown in Figure 1-35. Not shown are the plots for Cd, Co, Cr, Cu, F, Hg, Mo, Se and V, as the majority of results for these analytes were either below detection limits or exceedingly low and below guideline levels.

These results show that the concentrations of all the analytes listed in Table 1-8 apart the nitrogen compounds and phosphate were below the trigger levels in the wastewater entering the galleries. The general trend for all analytes was then to decrease in the recharged water as it moved away from the galleries. The exception was iron which, as described above, was higher in the recovered water than the wastewater in the galleries due to the presence of natural dissolved iron in the aquifer.

The time series plots in Figure 1-36 also show that the ‘adverse’ water quality issues for those water quality parameters that exceed either the ANZECC & ARMCANZ short-term or long-term threshold values are predominantly only an issue for the wastewater entering the infiltration galleries and the closest monitoring bore (BH01). After that the water quality parameters fall below the threshold values for samples collected from all the other bores down gradient. As stated above, the exceptions for this are iron and total-nitrogen. Perth groundwater, particularly in the coastal regions can be high in soluble iron and this is shown in the time series plots where the iron concentrations frequently exceed the threshold values for a number of bores across the infiltration gallery MAR site. The time series plots show that the reason for the exceedance of total nitrogen is predominantly nitrate (Figure 1-36 (d) and (e)) as there is evidence that nitrate forms a large proportion of the total nitrogen concentrations detected in the water from the different monitoring bores


BH

5G

alle

ryE

ast

Ga

llery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H10

BH

11B

H12

BH

13B

H14

BH

15B

H16

BH

3B

H17

FLB

GR

ND

01F

LBG

RN

D02

FLB

GR

ND

03

0

1

2

3

4

5

6

(7)

(17)

(25)

(25)(10)

(9)(4)(23)

(4)(18)

(21)(3)

(17)(2)

(15)

(16)

(14)

(25)

(24)

(7)

(6)

Iron

Co

ncen

trat

ion

(m

gL-1)

Sample Site

BH

5G

alle

ryE

ast

Gal

lery

we

stB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H10

BH

11B

H12

BH

13B

H14

BH

15B

H16

BH

3B

H17

FL

BG

RN

D01

FL

BG

RN

D02

FL

BG

RN

D03

0.0

0.1

0.2

0.3

0.4

1.5

2.0

Alu

min

ium

Co

nce

ntr

atio

n (

mg

L-1)

Sample Site

(7)(17)(25)

(25)

(10)(21)

(4)(22)

(4)

(17)

(20)(4)(16)(2)(16)

(16)(15)

(25)

(24)

(7)

(6)

BH

5G

alle

ryE

ast

Gal

lery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H3

BH

17

FLB

GR

ND

01

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.04

0.08

0.12

0.16

Ars

en

ic C

on

cen

tra

tion

(m

gL

-1)

Sample Site

(5)(8)(12)

(25)

(10)

(9)

(4)(9)

(4) (4)

(7) (3)

(3) (2) (3)

(2)(2)

(12)

(11)

BH

5G

alle

ryE

ast

Ga

llery

we

stB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H10

BH

11B

H12

BH

13B

H14

BH

15B

H16

BH

3B

H17

FLB

GR

ND

01F

LBG

RN

D02

FLB

GR

ND

03

0.0

0.1

0.2

0.3

0.4

0.5

Bo

ron

Con

cetr

atio

n (m

gL-1)

Sample Site

(10)

(26)(34)

(34)

(10)

(25)

(4)

(28)

(4)

(27)

(29)

(18)(21)

(2)

(19)

(20)

(18)

(34)

(33)

(15)(13)

BH

5G

alle

ryE

ast

Gal

lery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H3

BH

17

FLB

GR

ND

01

FLB

GR

ND

02

FLB

GR

ND

03

0

2

4

6

8

10

12

14

16

18

20

To

tal N

itro

ge

n C

on

cen

tra

tion

(m

gL-1

)

Sample Sites

(7)

(17)(25)(25)

(10)

(23)

(4)

(23)

(4)

(18)

(21)(7)

(7)

(5)

(7)(8)

(4)

(25)

(24)(7)(5)

BH

5G

alle

ryE

ast

Gal

lery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H3

BH

17

FLB

GR

ND

01

FLB

GR

ND

02

FLB

GR

ND

03

0.00

0.02

0.04

0.06

0.08

0.10

Ma

ng

an

ese

Co

nce

ntr

atio

n (

mg

L-1)

Sample Site

(6)

(10)

(15)

(14)

(10)

(12)

(4)

(12)

(4)

(7)

(10)

(3)

(6) (2)(15)

(5)

(13)

(14)(14)

(2)

(1)

BH

5G

alle

ryE

ast

Gal

lery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H3

BH

17

FLB

GR

ND

01

FLB

GR

ND

02

FLB

GR

ND

03

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

Nic

kel C

on

cen

tra

tion

(m

gL-1

)

Sample Site

(6)(10)

(15)

(14)

(10)(12)(4)

(12)

(4)

(7)(10)

(3)

(6)

(2)

(15)

(5)

(14)

(14)

(14)

(2)

(1)

BH

5G

alle

ryE

ast

Gal

lery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H3

BH

17

FLB

GR

ND

01

FLB

GR

ND

02

FLB

GR

ND

03

0.00

0.02

0.04

0.06

0.08

3

6

9

12

15

Ph

osph

ate

Con

cent

ratio

n (m

gL-1)

Sampling Site

(5)

(17)(25)

(25)

(10)(9)

(4)(9)(4)

(4)21)

(3)(3) (2)(3)

(2)(2)

(25)

(16)(4)

(4)

A. B. E. F.

C. D. G. H.


Figure 1-35. Box plots for (A) aluminium; (B) arsenic; (C) boron; (D) iron; (E) manganese; (F) nickel; (G) phosphorous-soluble-reactive; (H) total nitrogen; (I) uranium; and (J) zinc. The red lines shown on the arsenic, iron, total nitrogen and uranium plots refer to long-term trigger values from ANZECC & ARMCANZ (2000).

BH

5G

alle

ryE

ast

Ga

llery

we

stB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H10

BH

11B

H12

BH

13B

H14

BH

15B

H16

BH

3B

H17

FLB

GR

ND

01F

LBG

RN

D02

FLB

GR

ND

03

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.7

0.8

Zin

c C

on

cen

tra

tion

(m

gL-1

)

Sample Site

(6)

(10)

(15)

(14)(10)

(12)(4)

(12)

(4)

(7)

(10)

(3)

(6)

(2)

(15)

(5)

(14)

(14)

(14)

(2)

(1)

BH

5G

alle

ryE

ast

Gal

lery

wes

tB

H1

BH

2B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H3

BH

17

FLB

GR

ND

01

FLB

GR

ND

02

FLB

GR

ND

03

0.000

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.010

0.015

0.020

0.025

Ura

nium

Con

cent

ratio

n (m

gL-1)

Sample Site

(5)(5)(9)(10)(7) (6)

(3)

(21)

(3)

(3)

(5) (2)

(2)

(1)

(4)

(2)

(3)(8)

(10)

(2)(1)

I. J.


Table 1-8. Maximum concentrations (mg/L) of selected analytes sampled from each monitoring point during the entire period of monitoring during the MAR trial, 5 October 2005 to the end of August 2007

Al As B Cd Co Cr Cu F Fe Hg Mn Mo N_NH3 N_NO3 N_TK N_Total Ni P_SR Pb Se U V Zn

STTV1,3 20 2 NA 0.05 0.1 1 5 42 10 0.002 10 0.05 NA NA NA 25-1256 2 0.8-126 5 0.05 0.1 0.5 5

LTTV1,4 5 0.1 0.5 0.01 0.05 0.1 0.2 1 0.2 0.002 0.2 0.01 NA NA NA 5 0.2 0.057 2 0.02 0.01 0.1 2

Gallery East 0.053 0.001 0.4 BDL2 BDL BDL 0.012 19 0.7 BDL 0.038 0.003 1.6 5.8 2.8 9.3 0.002 1210 0.0004 BDL BDL BDL 0.064

Gallery West 0.047 0.001 0.41 0.0001 0.006 BDL 0.012 0.9 1.1 BDL 0.095 0.003 3.6 5.7 5.4 9 0.002 13 0.0009 BDL BDL BDL 0.091

BH01 0.14 0.15 0.44 BDL 0.006 BDL 0.013 0.3 0.26 BDL 0.008 0.005 0.09 7.2 1.9 7.9 0.03 5.1 0.0011 BDL 0.0003 0.016 0.05

BH02 0.042 BDL 0.22 0.0001 BDL BDL 0.009 0.3 0.061 BDL BDL BDL BDL 5.1 4.2 6.1 0.006 0.01 0.0003 BDL 0.0004 BDL 0.057

BH03 0.034 BDL 0.32 BDL BDL BDL 0.01 0.4 0.042 BDL 0.004 BDL BDL 6.3 0.85 6.3 0.016 0.01 0.0005 0.001 0.0011 BDL 0.027

BH05 0.035 BDL 0.12 BDL BDL BDL BDL 0.2 0.042 BDL 0.006 BDL BDL 3.6 2.5 4.04 0.003 0.02 0.0001 0.002 0.0007 0.005 0.009

BH06 0.083 0.002 0.37 0.0002 BDL BDL 0.011 0.4 0.071 BDL 0.048 BDL 0.08 7 2.1 8.5 0.005 0.01 0.0009 BDL 0.0005 BDL 0.019

BH07 0.041 BDL 0.12 BDL BDL BDL 0.007 0.5 0.054 BDL 0.01 BDL BDL 4.4 1.2 4.94 0.004 0.01 0.0003 0.001 0.0015 BDL 0.012

BH08 0.055 0.001 0.31 BDL 0.005 BDL 0.016 0.5 0.2 BDL 0.034 BDL 0.03 7 4 7.5 0.01 0.02 0.0009 0.002 0.021 0.005 0.071

BH09 0.009 BDL 0.13 BDL BDL BDL 0.007 0.8 0.026 BDL 0.02 BDL 0.03 4.95 1.8 5.3 0.003 0.02 0.0002 BDL 0.0039 BDL 0.01

BH10 0.081 BDL 0.32 BDL BDL BDL 0.008 0.2 0.065 BDL 0.01 BDL 0.02 7.1 3.4 9.1 0.006 0.01 0.0003 0.001 0.0023 0.006 0.026

BH11 0.031 0.005 0.07 BDL BDL BDL BDL 0.2 4.6 BDL 0.045 BDL 0.04 0.13 0.16 0.21 0.004 0.02 0.0002 BDL BDL BDL 0.054

BH12 0.03 0.005 0.22 BDL BDL BDL BDL 0.2 4.6 BDL 0.057 BDL 0.02 0.93 0.28 1.2 0.011 BDL 0.001 BDL 0.0002 BDL 0.012

BH13 0.009 0.003 0.22 BDL BDL BDL BDL 0.2 2.4 BDL 0.042 BDL 0.01 1.5 0.28 1.8 0.014 0.01 0.0004 BDL 0.0021 BDL 0.037

BH14 0.02 0.003 BDL BDL BDL BDL BDL 0.2 2.4 BDL 0.044 BDL 0.01 6.1 0.44 6.7 BDL 0.01 0.0003 BDL 0.0031 BDL 0

BH15 0.009 0.003 0.2 BDL BDL BDL 0.014 0.2 3.2 BDL 0.041 0.003 0.02 0.04 0.24 0.24 0.009 0.01 0.0045 BDL 0.0013 BDL 0.74


Al As B Cd Co Cr Cu F Fe Hg Mn Mo N_NH3 N_NO3 N_TK N_Total Ni P_SR Pb Se U V Zn

STTV1,3 20 2 NA 0.05 0.1 1 5 42 10 0.002 10 0.05 NA NA NA 25-1256 2 0.8-126 5 0.05 0.1 0.5 5

LTTV1,4 5 0.1 0.5 0.01 0.05 0.1 0.2 1 0.2 0.002 0.2 0.01 NA NA NA 5 0.2 0.057 2 0.02 0.01 0.1 2

BH16 0.011 0.001 0.14 0.0002 BDL BDL BDL 0.3 3.6 BDL 0.051 0.003 0.01 0.07 0.2 0.45 0.013 BDL 0.0002 BDL 0 BDL 0.077

BH17 0.01 0.008 0.12 BDL BDL BDL 0.14 0.3 6 BDL 0.024 BDL 0.54 2 0.78 2.8 0.01 0.01 0.01 BDL 0.0007 0.006 0.055

BGRND01 0.093 0.001 0.35 0.0002 BDL BDL BDL 0.3 0.45 BDL 0.02 BDL BDL 0.14 0.1 0.33 0.007 0.03 0.0005 0.001 0.002 0.007 0.02

BGRND025 0.27 NS8 0.06 BDL BDL 0.002 NS NS 2.2 BDL 0.002 BDL NS 1 NS 1.2 0.013 0.02 BDL 0.003 0.0004 0.01 0.039

BGRND035 1.8 NS 0.06 BDL BDL BDL NS NS 2.8 BDL 0.041 BDL NS 0.34 NS 0.71 BDL 0.01 BDL BDL BDL BDL 0.007

1 Trigger Values obtained from (ANZECC & ARMCANZ, 2000). 2 BDL = Below Detection Limit. 3 STTV = short-term trigger value for irrigation. 4 LTTV = long-term trigger value for irrigation. 5 BGRND02 and BGRND03 were not installed until October 2006. 6 The nitrogen and phosphorus STTVs require site specific assessments. 7 The LTTV for phosphorus is recommended to avoid bioclogging of equipment. 8 NS = not sampled. 9 Values given in blue have exceeded the long-term trigger value. 10 Values given in red have exceeded the short-term trigger value.


Figure 1-36. Time series plots of water quality parameters which exceed ANZECC & ARMCANZ (2000) threshold values (detailed in Table 1-8) where (a) is Arsenic; (b) is Iron; (c) is Soluble Reactive Phosphate; (d) is Total Nitrogen; and (e) is Nitrate. The red lines are long-term threshold values from ANZECC&ARMCANZ (2000).

0

1

2

3

4

5

6

7

28/05/2005

25/10/2005

24/03/2006

21/08/2006

18/01/2007

17/06/2007

14/11/2007

Iro

n (

mg

/L)

Gallery East

BH01BH08BH14

BH17

(b)

0

2

4

6

8

10

12

14

28/05/2005

25/10/2005

24/03/2006

21/08/2006

18/01/2007

17/06/2007

14/11/2007

Ph

osp

ho

rus

(mg

/L)

(c)

0

2

4

6

8

10

24/03/2006

2/07/2006

10/10/2006

18/01/2007

28/04/2007

6/08/2007

14/11/2007

To

tal

Nit

rog

en (

mg

/L)

(d)

0

2

4

6

8

10

2/07/2006

21/08/200610/10/200629/11/200618/01/20079/03/2007

28/04/200717/06/20076/08/2007

25/09/2007

Nit

rate

(m

g/L

)

(e)

0

0.05

0.1

0.15

0.2

28/05/2005

25/10/2005

24/03/2006

21/08/2006

18/01/2007

17/06/2007

14/11/2007

Ars

enic

(m

g/L)

(a)


Organic chemicals

The results from a total of 695 measurements from sample collecting at the FIG site, including 65 field blanks, were investigated. Five pharmaceuticals were tested (Table 1-9). A total of 33 samples of ambient groundwater from the three background bores and 25 samples of gallery water were analysed. BH01 and BH08 were the most commonly tested monitoring bores for pharmaceuticals, with a total of 16 measurements from each bore, followed by BH06, BH11 and BH17 (recovery bore) with 15 measurements from each bore. For BH3, BH07, BH15 and BH16 only one sample was analysed. A total of 12 field blanks were analysed for each chemical. The following bores were not sampled for pharmaceutical analyses: BH05, BH09, BH10, BH12, BH13, and BH14. Diazepam and phenytoin were not detected in any of the 152 measurements analysed for each chemical. The percent of samples (excluding blanks) that had concentrations greater than detection limits for the other three pharmaceuticals were temazepam (59%), oxazepam (59%) and carbamazepine (40%). Samples collected from bores that contained mainly ambient or native groundwater, such as the background bores, BH11 and BH17 had concentrations that were predominantly below detection limits.

Carbamazepine was detected in 60% of the groundwater samples from BH17 and 94% of the samples from BH08 and all groundwater samples from BH01 and BH06. Carbamazepine was detected in the groundwater samples from BH03, BH07, BH15 and BH16, but only one sample was collected from each of these bores (Figure 1-37). The maximum mean concentration was observed in bore BH07 (0.46 µg/L) and the minimum mean concentration was observed in bore BH17 (0.08 µg/L). However, only one sample (taken on 14/08/2006) was analysed from BH07. The next highest mean concentration after BH07 was observed at BH08 (0.36 µg/L) for which a total of 16 samples were analysed. The maximum individual carbamazepine value of 0.54 µg/L occurred at 2 m distance and 12 m deep (BH01). In general, the carbamazepine concentrations sampled from the bores tended to decrease with increasing distance from the galleries.

Oxazepam was detected in 60% of the samples from BH06, 63% of the samples from BH08 and 75% of the samples from BH01, but in none of the groundwater from BH17. Oxazepam was also detected in groundwater samples from BH07, BH15 and BH16, but only one sample had been collected from each of these bores. Water samples from the galleries had mean concentrations of oxazepam of 0.31 µg/L (east gallery) and 0.34 µg/L (west gallery). The minimum mean concentration was observed in BH15 (0.10 µg/L). There was a progressive decrease in the mean concentration of oxazepam from the galleries to the recovery bores (BH01 – 0.28 µg/L, BH06 – 0.23 µg/L, BH07 – 0.22 µg/L and BH16 – 0.11 µg/L).

Temazepam was detected in 53% of the samples from BH06, 63% of the samples from BH08, and 75% of the samples from BH01, but in only one of the 15 groundwater samples from BH17 and the concentration was at the detection limit (0.1 mg/L). Temazepam was not detected in any of the groundwater samples from BH17. Temazepam was also detected in groundwater samples from BH03, BH07, BH15 and BH16, but only one sample was collected from each of these bores. Water samples from the galleries had mean concentrations of temazepam of 0.31 µg/L (east gallery) and 0.34 µg/L (west gallery). The minimum concentration of temazepam was observed in BH17 (0.1 µg/L). The decreasing trend observed for carbamazepine and oxazepam was less clear for temazepam. Nevertheless, the mean concentration at 30 m and 32 m distance (i.e. Bores BH15 and BH16) was less than half the mean concentration reported in the galleries.

Time series plots given in Figure 1-38 showed that the detection of carbamazepine was relatively constant in the different sites. In contrast, the detection of temazepam and


oxazepam was less consistent in the groundwater despite being consistently detected in the wastewater entering infiltration galleries, indicating that there was loss of these two pharmaceuticals in the aquifer via chemical or biological processes. The consistency and persistence of carbamazepine in the wastewater and within the aquifer suggests that this chemical could have some potential for use as a conservative tracer in MAR systems using recycled wastewater within the Perth aquifer system.

Table 1-9. Detection of pharmaceuticals in the galleries and groundwater at the Floreat Infiltration Gallery site

Mean Pharmaceutical Concentration (µg L-1)

Bore Carbamazepine Diazepam Oxazepam Phenytoin Temazepam

Gallery East 0.202 (±0.038) 0 0.315 (±0.067) 0 0.309 (±0.042) Gallery West 0.215 (±0.084) 0 0.311 (±0.108) 0 0.313 (±0.108) BGRND01 0 0 0 0 0 BGRND02 0 0 0 0 0 BGRND03 0 0 0 0 0 BH01 0.378 (±0.089) 0 0.209 (±0.143) 0.006 (±0.025) 0.172 (±0.119) BH03 0.2 (±0)1 0 0 0 0 BH06 0.34 (±0.068) 0 0.139 (±0.137) 0 0.119 (±0.129) BH07 0.46 (±0)1 0 0.22 (±0)1 0 0.12 (±0)1

BH08 0.339 (±0.145) 0 0.137 (±0.123) 0.006 (±0.025) 0.129 (±0.123) BH11 0.003 (±0.013) 0 0 0 0 BH15 0.22 (±0)1 0 0.15 (±0)1 0 0.13 (±0)1

BH16 0.18 (±0)1 0 0.11 (±0)1 0 0.135 (±0.007)1 BH17 0.045 (±0.042) 0 0 0 0.007 (±0.026)

Guideline value2 100 µg L-1 2.5 µg L-1 NA3 NA3 5 µg L-1

1 Value from one sample only. 2 NRMMC-EPHC-NHMRC (2008). 3 NA = Not available.


Figure 1-37. Detection of (A.) carbamazepine; (B.) temazepam; and (C.) oxazepam in the groundwater in monitoring bores along the Floreat Infiltration Gallery site

Ga

llery

Ea

st

Gal

lery

We

st

BH

1

BH

6

BH

7

BH

8

BH

11

BH

15

BH

16

BH

3

BH

17

FL

BG

RN

D01

FL

BG

RN

D02

FL

BG

RN

D03

0.0

0.1

0.2

0.3

0.4

0.5

0.6C

arb

am

aze

pin

e C

on

cen

tra

tion

(g

L-1)

Sample Site

(16) (8)(9)

(15)

(1)(1)(1)

(15)

(16)(1)(15)

(16)

(14)(11)

A.

Ga

llery

Eas

t

Gal

lery

We

st

BH

1

BH

6

BH

7

BH

8

BH

11

BH

15

BH

16

BH

3

BH

17

FL

BG

RN

D01

FL

BG

RN

D02

FL

BG

RN

D03

0.0

0.1

0.2

0.3

0.4

0.5

Tem

axa

pam

Co

nce

n tr

atio

n

( g

L-1)

Sample Site

(16) (8)(9)

(15)

(1)

(1)(1)

(15)

(16)

(1)

(15)(16)

(14)

(11)B.

Ga

llery

Eas

t

Gal

lery

We

st

BH

1

BH

6

BH

7

BH

8

BH

11

BH

15

BH

16

BH

3

BH

17

FL

BG

RN

D01

FL

BG

RN

D02

FL

BG

RN

D03

0.0

0.1

0.2

0.3

0.4

0.5

Oxa

zepa

m C

on

cen

tra

tion

( g

L-1)

Sample Site

(11)

(14)(16)

(15)

(1)

(16)

(15)

(1) (1)

(1) (15) (16) (9) (8)

C.


Figure 1-38. Time series plots of (a) carbamazepine, (b) temazepam, and (c) oxazepam in infiltration gallery west and water collected from monitoring bores BH01, BH08 and BH17

0

0.1

0.2

0.3

0.4

0.5

0.6

13/05/2006

2/07/2006

21/08/2006

10/10/2006

29/11/2006

18/01/2007

9/03/2007

28/04/2007

17/06/2007

6/08/2007

25/09/2007

Conce

ntr

ation (ug/L

)

Gallery West

BH01

BH08

BH17

0

0.1

0.2

0.3

0.4

0.5

0.6

13/05/2006

2/07/2006

21/08/2006

10/10/2006

29/11/2006

18/01/2007

9/03/2007

28/04/2007

17/06/2007

6/08/2007

25/09/2007

Co

nce

ntr

atio

n (

ug

/L)

0

0.1

0.2

0.3

0.4

0.5

0.6

13/05/2006

2/07/2006

21/08/2006

10/10/2006

29/11/2006

18/01/2007

9/03/2007

28/04/2007

17/06/2007

6/08/2007

25/09/2007

Co

nce

ntr

atio

n (

ug

/L)

(a) (b)

(c)


Fate of trace organics – field assessment

Half-life degradation values for the pharmaceuticals (carbamazepine and oxazepam) were based on mass balances studies and relative changes in groundwater concentrations down-gradient from the infiltration galleries.

During a period of relatively stable recycled water infiltration and down-gradient groundwater extraction, a delivery/recovery mass balance assessment for carbamazepine and oxazepam was undertaken. The outcomes of the assessment are given in Table 1-10. Based on mass delivery rates and mass recovery rates, the data suggests that carbamazepine was not removed during the estimated travel time of 70 days through the aquifer, which was consistent with column data conducted under similar aerobic conditions and temperatures that showed a degradation rate of >100 days. Due to the higher analytical detection limit for oxazepam, a field assessment of oxazepam degradation based on mass balance data could not be undertaken.

Table 1-10. Mass recovery calculations for carbamazepine and oxazepam at the Floreat Infiltration Gallery site

Concentration in BH2

Concentration in BH17

Mass delivery rate#

Mass recovery rate^

Carbamazepan 0.46µg L-1

(±0.05) 0.088 µg L-1

(±0.024) 23 mg day-1 (±3) 22 mg day-1 (±6)

Oxazepam 0.30µg L-1 (±0.11)

<0.1 µg L-1 15 mg day-1 (±6) <25 mg day-1

# Based on infiltration rate of 50 kL day-1. ^ Based on recovery rate of 250 kL day-1. Degradation half-life values for carbamazepine and oxazepam based on changes in trace organic concentration with distance from the infiltration galleries were investigated (Figure 1-39). The similar rate of decrease in both carbamazepine and oxazepam with distance from the west infiltration gallery was due to degradation and/or dilution/dispersion. Assuming that dilution/dispersion is similar for both compounds, this data suggests that their degradation rates were also similar. Based on this data and carbamazepine mass balance data, oxazepam degradation half-life is likely to be negligible over the 70 day aquifer passage. These field-based degradation rates are consistent with column data, suggesting the column data provides reliable field-scale estimation of trace organic degradation rates, at least for persistent pharmaceuticals carbamazepine and oxazepam as reported in Chapter 2. Details on the actual risk of these chemicals at the concentrations detected are given in Chapter 4.


Figure 1-39. Carbamazepine and oxazepam concentrations along the groundwater flow line between the west infiltration gallery and the extraction bore (50 m down-gradient) at the MAR field site

Microorganisms

The results for the average, maximum and minimum detection of faecal coliforms and enterococci at the Floreat Infiltration Gallery site can be seen in Figures 1-40 and 1-41. The specific detection results as a time series on individual sampling events can be seen in Figure 1-42. The results show that high numbers of faecal coliforms and enterococci were routinely detected in the treated wastewater entering the infiltration galleries with average faecal coliform numbers of around 8000 colony forming units (cfu) 100 mL-1 and average enterococci numbers of 10000–12000 cfu 100 mL-1. These numbers should be considered normal for treated wastewater. Once the treated wastewater recharged through approximately 10 m of unsaturated zone, the average number of faecal coliforms dropped to approximately 50 cfu 100 mL-1 or less in the monitoring bore closest to the infiltration galleries (BH01) (Figure 1-40). The average enterococci numbers detected in the groundwater taken from this monitoring bore were slightly higher than the faecal coliform numbers with averages as high as approximately 300 cfu 100 mL-1. Both thermotolerant coliforms and enterococci were detected periodically in a number of the monitoring bores, sometimes approaching the detection limit (1000 cfu 100 mL-1) (Figure 1-42). It is worth noting here that both faecal coliforms and enterococci were detected in the background groundwater up gradient of the infiltration galleries (e.g. BGRND01 in Figure 1-42) indicating that the local groundwater is subject to contamination from sources other than the recharged treatment wastewater.

-5 0 5 10 15 20 25 30 35 400.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

CARB OXAZ

Con

cen

tra

tion

(g

L-1)

Distance downgradient from infiltration gallery (m)


Figure 1-40. Average, maximum and minimum numbers of faecal coliforms (per 100 mL) detected at all monitoring sites at the Floreat infiltration galleries over the time of the pilot project. (Numbers on top of each site value are the total number of samples used in this analysis. Numbers in parenthesis are the number of times counts above the maximum detection limit were obtained).

Figure 1-41. Average, maximum and minimum numbers of enterococci (per 100 mL) detected at all monitoring sites at the Floreat infiltration galleries over the time of the pilot project. (Numbers on top of each site value are the total number of samples used in this analysis. Numbers in parenthesis are the number of times counts above the maximum detection limit were obtained).

Bac

kgro

und

1B

ackg

roun

d2

Bac

kgro

und

3E

ast

Ga

llery

Wes

tGa

llery

BH

1B

H2

BH

3B

H5

BH

6B

H7

BH

8B

H9

BH

10

BH

11

BH

12

BH

13

BH

14

BH

15

BH

16

BH

17

0

200

400

600

800

10005000

10000

15000

20000

(1)(1)

Ave

rage

and

Ran

ge o

f cfu

Monitoring Bores and Galleries

Detection limit in groundwater

Detection limit in wastewater

23 8

7

17 29

27

9 3 5 29 6

25

5 7

24

4 6 36 8 25

(2) (1)

(6) (9)

(1)

Bac

kgro

und

1B

ackg

roun

d2

Bac

kgro

und

3E

ast

Gal

lery

We

stG

alle

ryB

H1

BH

2B

H3

BH

5B

H6

BH

7B

H8

BH

9B

H1

0B

H1

1B

H1

2B

H1

3B

H1

4B

H1

5B

H1

6B

H1

7

0

200

400

600

800

1000

5000

10000

15000

20000

(1)(1)(1)(1)(1)(1)

Ave

rage

and

Ran

ge o

f cfu

Monitoring Bores and Galleries

Maximum detection limit in groundwater

Maximum detection limit in wastewater

26 9

18 30

29 8 4 5 29 6 24

4 7

27

4 64

6

9 26(2) (1)

(6) (13)

(6) (3) (2)


Figure 1-42. Examples of time series plots on the detection of indicator microorganisms in the infiltration galleries and monitoring bores where • = thermotolerant coliforms and ■ = enterococci. (a) = background bore BGRND01, (b) = east infiltration gallery, (c) = monitoring bore BH01, (d) = monitoring bore BH06, (e) = monitoring bore BH011, and (f) = recovery bore BH17

0

200

400

600

800

1000

5000

10000

15000

20000

0

200

400

600

800

1000

0

200

400

600

800

1000

22/0

3/20

05

22/0

6/20

05

22/0

9/20

05

22/1

2/20

05

22/0

3/20

06

22/0

6/20

06

22/0

9/20

06

22/1

2/20

06

22/0

3/20

07

22/0

6/20

07

22/0

9/20

07

0

200

400

600

800

1000

22/0

3/20

05

22/0

6/20

05

22/0

9/20

05

22/1

2/20

05

22/0

3/20

06

22/0

6/20

06

22/0

9/20

06

22/1

2/20

06

22/0

3/20

07

22/0

6/20

07

22/0

9/20

07

0

200

400

600

800

1000

(a)(b)

Ave

rage

cfu

100

mL-1

(c) (d)

Sample Dates

(f)(e)


There are potential explanations for the high numbers of faecal coliforms and enterococci detected in several of the bores, including in the up gradient background groundwater. The most likely source for these microorganisms is that there is a large amount of animal-based research on the CSIRO grounds and a nearby University of Western Australia research station, either of which could be a major source of additional faecal indicator bacteria to the environment. The infiltration galleries are located on an experimental sheep paddock and the higher numbers detected for enterococci compared to faecal coliforms suggests that sheep faecal material could be the source of these higher than expected enterococci numbers. While the number of both faecal coliforms and enterococci are highest in the treated wastewater entering the infiltration galleries, the survival studies described in the study of enteric microorganism survival indicate that the decay rates for enterococci and faecal coliforms (represented by E. coli) (given in Table 1-12) are too fast for the enterococci entering the infiltration galleries to be present in any of the monitoring bores beyond BH06. Also, the fact that the chemical data indicates that the plume of recharged treated wastewater is above some of the bores where enterococci and faecal coliforms are detected (e.g. BH11) and the detection of enterococci numbers in the background bores suggests that these bacteria are from sources other than the treated wastewater.

The fact that sheep use the experimental paddocks on and around the site of the infiltration galleries is a strong suggested link to these microbes. The most likely source of the microbial numbers is via contamination of the sampling equipment. Despite the best efforts of the field sampling team, the use of sampling pumps connected to long hose lines means that occasionally and inadvertently pump-lines come into contact with the ground. If there were fresh sheep droppings in the vicinity this could cause contamination of the pump lines which then could cause brief contamination of the water within the monitoring bores. Another potential, although less likely explanation, is that it is also conceivable that the microorganisms are being rapidly transported to the aquifer around the bore from preferential pathways formed around the bores during construction. Source tracking (e.g. detection of human and animal faecal sterols and the use of microbial genetics) to identify the individual enterococci and faecal coliform strains could assist in identifying the source of these microbes.

If it is assumed that sheep faecal material is the source of these microorganisms in the aquifer at the infiltration gallery sites, this indicates that potential surface contamination remains an issue that will need consideration for any infiltration galleries or similar MAR scheme that is planned for the superficial aquifer in the Perth region and an intensive assessment of potential contamination sources will be needed.

The results from the monitoring for F+ specific bacteriophage and enteric viruses are given in Table 1-11. As could be expected, the treated wastewater in the infiltration galleries was positive for phage in more than 90% of the samples tested. A similar result was obtained for the PCR detection of adenovirus. Only one of the five wastewater samples tested for enterovirus were positive using RT-PCR; however, some matrix effects were observed which may have reduced the sensitivity of the RT-PCR reaction. F+ specific bacteriophage were also detected on one occasion in the monitoring bore BH01 (out of a total of 25 samples), twice in the background bore BGRND01 (out of a total of 19 samples) and one in monitoring bore BH11 (out of a total of 23 samples). The detection of bacteriophage in the monitoring bore BH01 should not be considered unexpected as it is the closest and shallowest bore to the infiltration galleries (see Figures 1-7 and 1-8 for bore details). The detection of bacteriophage in this bore also coincides with a high enterococci count, indicating that a higher microbial load may have entered the galleries a short time previously. This monitoring bore was also the only source where enterovirus was detected apart from the treated wastewater entering the galleries. Again, this coincided with the detection of high enterococci numbers.


Like the bacteriophage, adenovirus was detected on the majority of sampling events in the treated wastewater entering the infiltration galleries (75% of the samples taken from the western gallery were positive for adenovirus). Adenovirus was only detected in one other sample from any other location other than the infiltration galleries (bore BH08). While there is a possibility that this could be a real result it is much more likely to be a one-off false positive. The sample that gave the positive result was collected on 20 February 2006 which is a time in Perth with very low rainfall. While that February had a higher than average rainfall (Bureau of Meteorology http://www.bom.gov.au/jsp/ncc/cdio/cvg/av) it was still markedly less than the rainfall in July, August and September of that year which means that if an increase of rainfall recharging at the Floreat Infiltration Gallery site was the cause of enteric viruses moving further through the aquifer then detection should have occurred during these months as well. Thus, it is most likely that this one detection of adenovirus in bore BH08 was a false positive. A false positive could either be caused by a laboratory contamination or a mis-priming of the PCR primers. A laboratory contamination is always a possibility with a detection method as sensitive as PCR; however, the field and transport blanks, as well as the negative controls used during the thermocycling reaction run when this sample was tested did not indicate any laboratory-based contamination. The PCR primers used during this time have since been found to occasionally give false positive (CSIRO unpublished data) and subsequently a new set of more specific primers for adenovirus have been sourced and now used within the laboratory. Thus, it is most likely that this one-off detection was a false positive caused by mis-priming using these original primers.

The overall results of the microbiological monitoring have shown that the microbiological quality of the recharged water/groundwater is much better than the treated wastewater entering the infiltration galleries with a minimum 3 log removal of microorganisms from the recharged recycled water between the infiltration galleries and the recovery well (BH017). It is also of better chemical and microbiological quality than the recycled water currently used to irrigate McGillivray Oval (Toze et al., 2005). This indicates that the galleries are acting as an active treatment barrier for pathogen removal and producing a water quality that should be suitable for irrigation of green open spaces, at least where the potential for exposure is controlled.


Table 1-11. The detection of F+ specific bacteriophage, adenovirus and enterovirus in samples collected from the infiltration galleries and monitoring bores

F+ specific phage Adenovirus Enterovirus

(per 50 mL or 500 mL)a (per 1 L or 5 L)b (per 1 L or 5 L)b

Monitoring site

+ve -ve +ve -ve +ve -ve

BGRND01 2 17 0 3 0 3 BGRND02 0 8 - - - -

BGRND03 0 6 - - - - BH05 0 3 0 4 0 3

Gallery East 13 1 1 2 - - Gallery West 25 1 12 4 1 5 BH01 1 24 0 18 1 4

BH02 0 8 0 2 0 1 BH03 0 1 - - - -

BH06 0 25 0 10 0 3 BH07 0 4 - - - -

BH08 0 23 1 6 0 2

BH09 0 2 - - - - BH010 0 3 - - - -

BH011 1 22 0 9 0 3

BH012 0 1 - - - - BH013 0 4 - - - -

BH014 - - - - - -

BH015 0 1 0 1 - -

BH016 0 3 0 1 - -

BH017 0 18 0 6 - - a 50 mL of treated wastewater or 500 mL of groundwater. b 1 Litre of treated wastewater or 5 Litres of groundwater. - = no testing was undertaken.

Microbial pathogen decay rates

As part of determining the risk from the presence of microbial pathogens in the recharged recycled water, an understanding of their behaviour and persistence is needed. The new water recycling guidelines (NRMMC-EPHC-AHMC, 2006) require the use of a risk management approach to managing recycled water schemes and thus, an understanding of the potential presence and persistence of pathogens plus the potential ability to treat the water to effect the removal of these pathogens is essential. For many recycling schemes, engineered treatment systems are commonly used to treat recycled water, in part at least, to remove any pathogens present in the water. Natural systems such as aquifers have the ability to act as an active treatment system; however, a sound understanding of this treatment capacity is needed to be able to manage natural systems as a treatment barrier (Dillon et al., 2008). Knowing the rate of pathogen decay can also be used as part of a quantitative microbial risk assessment (QMRA) to determine the residual risk posed by microbial pathogens in the final recovered water. It is recommended by the Australian Water Recycling Guidelines, in the Managed Aquifer Recharge section, that if the aquifer is to be relied upon for the removal of microbial pathogens that a pathogen decay study be undertaken at the site (NHMRC-EPHC-NRMMC, 2009). As part of the research undertaken at the Floreat infiltration galleries such a decay study was undertaken.


The outcomes of this decay study are presented as a decay rate () and the corresponding T90 value (time for a 90% reduction) for each of the pathogens tested (Table 1-12). The graphs showing the actual decay curves and corresponding regression lines can be seen in Appendix 1H.

The results below demonstrate that the aquifer has an active treatment capacity to remove pathogens during passage of recharged water as has been previously observed (Toze and Hanna, 2002; Toze et al., 2004; Yates et al., 1990). Information on the time of decay of microbial pathogens in groundwater is important to assist the design of MAR schemes to determine the required residence times of the recharged water in the aquifer to reach exposure risk levels as outlined in the Australian Guidelines for Water Recycling (NHMRC-EPHC-NRMMC, 2009) and to assess the need for additional pre- or post-MAR treatment to further reduce the risk from microbial pathogens.

Until recently, there has been limited information available on the decay rates of only specific enteric microorganisms in groundwater. There are a number of papers providing information on the decay of indicator microorganisms (e.g. E. coli and the bacteriophage MS2) and the more easily culturable enterovirus group (in particular poliovirus and coxsackievirus) (e.g. Keswick et al., 1982; Nasser and Oman, 1999; Yates et al., 1985) in groundwater; however, information on other less easy to study bacteria, viruses and protozoa (e.g. rotavirus, norovirus, Campylobacter and Cryptosporidium) is limited.

Table 1-12. Decay rates () and 90% removal times (T90) of microbial pathogens in groundwater at the Floreat Infiltration Gallery site through the use of in situ pathogen decay chambers in monitoring bore BH06

Enteric microbe tested (day-1) T90 (days)

Cryptosporidium -0.0269 (0.0018) 37

Adenovirus -0.0178 (0.0116) 56

Rotavirus -0.0291 (0.0067) 34

Coxsackievirus -0.0539 (0.0012) 18.5

MS2 -0.1277 (0.0008) 8

E. coli -0.6547 (0.1328) 1.5

S. enterica -0.8428 (0.0454) 1

E. faecalis -0.8942 (0.0004) 1

In this study adenovirus was determined to have a decay rate of 56 days, rotavirus 34 days and coxsackievirus 18.5 days. The decay of coxsackievirus is similar to rates previously reported (Toze et al., 2004; Jansons et al., 1989). The limited information on the decay of rotavirus in the environment is predominantly based on the study of the survival of the culturable simian rotavirus SA-11 used as a surrogate for the non-culturable human rotavirus. Keswick et al. (1982) determined that rotavirus SA-11 had a 90% loss in only 3 days. This is much less than the time determined in this current study for human rotavirus (34 days). (Note: While there is no information on the groundwater conditions in the 1982 study undertaken by Keswick et al., the study was undertaken in diffusion chambers suspended in tanks which had groundwater continually pumped through them. It can


therefore be surmised that the groundwater would be oxygenated in the tanks (therefore be aerobic). The groundwater temperature was reported to range from 3 to 12ºC).

Studies of the detection of enteric viruses in groundwater also indicate that the survival of certain viruses can be high under ideal conditions. For example, Borchardt et al. (2004) were able to detect rotavirus in as high as 36% of groundwater samples using RT-PCR. Similarly, Piranha et al. (2006) were able to detect adenovirus in eight out of 15 groundwater sources tested.

The survival of Cryptosporidium oocysts was found to have a non-linear decay with an initial T90 time of 12 days, slowing down to 146 days after 12 days incubation in the aquifer. In a study of Cryptosporidium survival in groundwater, Ives et al. (2007) found the time for a 99% decay (T99) of the oocysts ranged from greater than 200 days to as little 45 days at 22ºC. They determined that temperature was the most important factor influencing the decay of the oocysts. In the current study, temperature variation was not a variable that was tested as the study was done in situ (groundwater temperatures measured during sampling events varied from approximately 19 to 24ºC depending on the season).

The information produced on the decay of the target pathogens has been used in a quantitative microbial risk assessment (QMRA) to determine the health risk from microbial pathogens based on exposure to the recovered water if it was used for spray irrigation of green open spaces in Chapter 4. The QMRA assessment provides information on the suitability of the current residence time at the Floreat Infiltration Gallery research site and the associate risk.

1.4 LESSONS LEARNT: MANAGEMENT AND OPERATIONAL REQUIREMENTS FOR MAR SYSTEMS IN WESTERN AUSTRALIA

During the establishment and operation of the infiltration galleries and the associated research within this project, a number of previously unforseen practical issues were encountered and dealt with. Also, since the commencement of the project in 2005 the New Australian Water Recycling Guidelines have become available (NRMMC-EPHC-AHMC, 2006) which include a dedicated section on Managed Aquifer Recharge (NHMRC-EPHC-NRMMC, 2009). Many of the issues relating to setting up an MAR scheme that needed to be investigated and established during the commencement of the infiltration galleries are now covered in these new guidelines. It is essential that any operator intending to establish a MAR scheme on the Swan Coastal Plain uses the Australian Guidelines as the basis for the design, construction and operation of the MAR system. There were, however, a number of lessons that were learnt during this project that are either specific to the Swan Coastal Plain or are not specifically mentioned in the Australian Guidelines. The authors consider that these issues are worth listing for consideration and it is hoped that this information will assist in the design and operation of new MAR schemes

1.4.1 Infiltration System Design

It was demonstrated that the use of appropriate materials for the design and establishment of MAR schemes needs to be carefully considered. It is strongly recommended that the materials used to construct any proposed MAR scheme be considered based on a range of criteria that includes, but is not dominated by initial construction costs.


In the design of the infiltration galleries for this research project the original recommendation had been to simply use gravel for the construction of the galleries. Gravel has been traditionally used for French Drains and it was considered that this would be sufficient for the galleries used in this project. Upon further consideration, however, the project team considered that the use of gravel, while the cheapest option, had potential risks, mostly associated with performance due to the relatively large volumes of treated effluent being delivered to the galleries and the high infiltration rates required. Because of this, it was decided to also trial the Atlantis Leach System® which has larger voids (thus larger surface area of the walls of the infiltration gallery and greater capacity to hold larger volumes) for the infiltration of the treated wastewater. The results of the research did in fact show that, while the Atlantis Leach System® was initially more expensive than the gravel for constructing the infiltration galleries, the long-term superior performance of the Atlantis Leach System® meant that the initial greater outlay in construction costs (compared to the use of gravel) was far outweighed by the reduced maintenance costs.

While infiltration galleries were the chosen MAR scheme in this project, the same principals would apply for any other scheme using different MAR methods. A careful consideration of the different options for construction of a MAR system, and a willingness to consider that a greater initial outlay can reduce later problems could have a major influence on the long-term success of any MAR scheme.

1.4.2 Reliability of Water Supply

An understanding of the security and reliability of supply of the recharge water and designing the system to cope with supply issues will be an important issue in managing and operating any new MAR scheme. For example, in the current scheme both sets of infiltration galleries (Floreat and Halls Head) were designed to be continually supplied with secondary treated wastewater from the respective wastewater treatment plants. The designs involved the collection of the treated wastewater from the discharge point post-clarifiers at both plants. At the smaller Halls Head wastewater treatment plant this location was after the water had passed from all three clarifiers and prior to the discharge ponds. At the much larger Subiaco treatment plant, the off-take pump was located in the treated effluent channel downstream of the first two clarifiers (in a string of twelve).

What had not been foreseen was that both treatment plants were impacted by low flow events on a diurnal cycle. At the Subiaco Wastewater Treatment Plant (WWTP) (supplying the Floreat site) the pump had been located after only the first two of the clarifiers for OHS and operation reasons. A better understanding of the diurnal variations in flow of the wastewater, however, would have meant that this would have been changed and the off-take pump probably located further down the treated wastewater channel after more treated wastewater had been discharged by more of the clarifiers. At the Halls Head WWTP the off-take pump had already been located after all three of the clarifiers so a different location would not have had any influence. For both sites, a weir was installed in the discharge channel in an attempt to maintain a water level that would prevent the water around the off-take pump falling below the operational water level. At the Subiaco WWTP this was sufficient to allow a continual 24-hour operation of the off-take pump. At the Halls Head WWTP, however, the amount of wastewater entering the treatment plant in the early hours of the morning (after 2am) frequently became too low for even the weir to maintain the water level around the off-take pump. This resulted in frequent automatic shut down of the off-take pump. As the Halls Head WWTP was a remotely operated plant with no full-time operational staff, this meant that the infiltration galleries were frequently not operating for days before the shut down was noticed and the pump manually restarted. This problem was eventually overcome by installing an automatic timer on the Amiad® Filter which shut down the off-take pump over the period of expected low flow each night and then restarted the pump after the time when it was expected that the plant would be operating at sufficient capacity for the off-


take pump to function. Despite this, problems with the timer or water levels still occasionally occurred, resulting in system shut down that required a manual restart when noticed.

An automated alert system notifying appropriate staff that the plant had shut down would have been very valuable, but was not used in this project due to costs. At the Floreat site, due to its location on the grounds of CSIRO laboratories, a flashing light was installed on the Amiad® System which indicated if the system had shut down. Due to the close proximity of the offices of the staff maintaining and operating the galleries, this meant that the galleries would be non-operational for a maximum of 12 hours on a week day and no more than 48 hours over a weekend. This system was less useful at the Halls Head site (being a remotely operated site) and it was found that this site could remain non-operating for several days before being noticed and power restored to the system. Despite this, the flashing light system was still useful at this site as it was easily noticed by the plant operators when they came on site and allowed them to know if they needed to go down to the galleries and reactivate the pump.

1.4.3 Iron Precipitation and Fouling of Pumps

Another area of the infiltration gallery system that was found to be very important was the condition of the pumps used. Pumps are an integral part of any MAR scheme, used in the delivery of the recharge water from the source to the scheme, sometimes to assist with the recharge, and then for recovery of the recharged water from the aquifer. The importance of the condition of the pump and fouling around the pump and MAR wells is well documented (e.g. Dillon and Toze, 2005). In the current project using infiltration galleries two pumps were used. The first pump was used to collect the secondary treated effluent from the discharge channel and deliver it to the infiltration galleries (via the Amiad® filters). The second pump (for the Floreat galleries only) was located 50 m down gradient of the galleries in the recovery well (BH17) for the recovery of the recharged water.

The recovery pump (in BH17) suffered from iron fouling which required periodic maintenance to prevent complete clogging of the pump. The Swan Coastal Plain is well known for having problems associated with high concentrations of iron in the groundwater and the presence of iron-oxidising bacteria which can cause the build up of biological slimes and biofilm on pumps and well screens. While the recharged treated wastewater had low iron concentrations, it was much higher in the surrounding native groundwater. As the recovery bore at the Floreat infiltration galleries was pumping at five times the volume of the recharged wastewater, more groundwater was being drawn in from the sides and below the plume of the recharged water. This resulted in high iron concentrations being drawn into the recovery well with subsequent regular decreases in performance of the recovery well. This subsequently required regular maintenance on the recovery well. This involved manually checking the recovery pump on a weekly basis (and with no longer than a fortnight between checks) for any signs of deterioration in the pumping rate. The presence of iron-oxidising bacteria and the formation of iron oxide residues (via both biological and abiotic processes) meant that the bore needed to be scoured weekly. In addition, the pump required cleaning and servicing several times a year with the time of servicing being determined once the pump rate had fallen below a predetermined rate. While not undertaken in this project, another option can be to regularly treat and purge the bore with a cleaning agent, such as ethanedioic acid dihydrate that can dissolve and loosen the iron oxide residues. Also, in this research scheme, if the recovery pump had been operating at a rate sufficient for just the collection of the recharged water (i.e. approximately at the same rate as the recharge rate) it could be expected that less of the surrounding native groundwater would have been collected with the recovered recharged water. This would have resulted in lower iron concentrations in the recovered water which would have reduced the iron fouling problem.


It was found that the pump located in the treated wastewater channel at the wastewater treatment plant (for both Halls Head and Subiaco WWTPs) (termed here as the supply pump) also required regular maintenance. These supply pumps were subject to fouling, partly from the build up of biofilm and also from fibrous material clogging the screen around the pump – loosely termed as ‘ragging up’. The problem of clogging of the supply pump was found to be easily remedied by regular visual inspections of the pump screen, monitoring of the flow rate of the treated wastewater entering the infiltration galleries, and the establishment of a regular cleaning program with a local contractor.

1.4.4 Pipe Leakage and Protection

Security of supply relied not only on the performance of the pumps; the quality and protection of the delivery pipes was also found to be important. A single incident of leakage from the delivery pipes occurred during the early stages of operating the infiltration galleries. Despite the high quality of work undertaken by the contractor installing the pipe work and the design of the pipe work to shutdown pumping in the case of major leaks, a leak was still detected in the early stages of the project. The leak was only minor, which was the reason for the lack of an automated pump shut down, but was still sufficient to bring water to the ground surface. The reason was most likely due to a dry weld that had occurred due to a longer than expected delay between installing the pipes and the commencement of water flow to the galleries (a period of several months). In discussions with the contractor it was agreed that, while an initial pressure test had been undertaken on the pipe work, that a second test should have been done just before the recycled water supply was turned on.

A contingency plan had been produced by the research team which dealt in part with the scenario of pipe leakage. This plan was enacted when the actual leak was detected and shown to be effective. The contingency plan included daily inspections of the full pipe line following any detected leak and repairing any leaks. The following of the contingency plan plus a report on the incident and the subsequent actions taken was sufficient in keeping the regulators informed and assured that best practice was being followed. No further leaks were detected in the pipeline and visual inspection of the pipeline returned to monthly checks after a month of daily inspections.

Another issue relating to the security of the pipe work was related to heavy vehicles driving over the pipes during work on other subterranean infrastructure. When the pipe work was installed it was buried a minimum of 300 mm below ground (as required by law for water-bearing pipe work) and the location of the pipes logged with ‘Dial Before You Dig’. In addition, the pipes were installed on verges so that they were located as close as possible to curbs or fence lines. In addition, signs were placed on fences at regular intervals along the route of the pipes indicating that pipes carrying recycled water were located nearby. Despite this, on several occasions heavy trucks were known to drive over the location of the pipes while work was being undertaken on other infrastructure buried nearby (e.g. cables and sewage pipes). This necessitated immediate visual inspection of the area where the trucks had traversed and the monitoring of flow rates to see if any damage had occurred to the pipes. The project was fortunate that, despite concerns, no damage to the pipes actually occurred during these events and that the MAR scheme and pipeline was located close to the offices of the project team so that these activities were noticed early on and immediate proactive measures could be taken. For other intended MAR schemes it is strongly recommended that either the pipes be buried deeper than the required minimum 300 mm or be protected from crushing by some other means in any area where there is potential for heavy vehicles to traverse over the pipes. The experience in this project has demonstrated that MAR operators cannot expect that simply logging the pipe work location (via ‘Dial Before You Dig’) and placing signs (indicating the presence of the pipe work) will be sufficient to protect the pipe. Thus, proactive measures through forward planning are essential.


1.4.5 Clogging of Galleries

Dealing with potential clogging of galleries during recharge is an issue that needs to be included in the operational plan of most MAR schemes (Dillon and Toze, 2005). As noted above in the chapter and in the first research objective, the presence of plant roots was a potential clogging issue for the gravel filled galleries (Figure 1-18). While plant roots would be very difficult to completely exclude from MAR systems, an understanding of the potential invasion of plant roots into MAR schemes and forward planning on ways to deal with this issue can help to overcome this problem. Also, designing MAR systems to be robust despite invasion of plant roots can be a means to overcome this problem. As stated earlier, no plant roots were observed in the infiltration galleries constructed using the Atlantis Leach System®. While it was not determined if plant roots had actually entered the galleries constructed with this system or not, there was no observed decrease in infiltration rates in these galleries over the three and half years of the research project. Therefore, even if plant roots had invaded the galleries constructed using the Atlantis Leach System®, there was no observed impact on the infiltration rates, possibly due to the fact that there was much more void space in the crate system.

1.4.6 Monitoring Equipment

Use of water level probes and sensors can be a valuable tool in monitoring the performance of MAR schemes. It was noted in this project that while sensors and probes were valuable tools, they should not be solely relied upon. It was found that regular visual inspections of the MAR system considerably improved the performance of the infiltration galleries. Despite this, some observations on the usefulness of sensors and probes are as follows:

(a) Each infiltration gallery contained a float switch, which was used to either reduce or terminate inflow if the water level increased to a critical set point. Water level recording devices were also used in the discharge chambers which allowed the water levels in the infiltration galleries to be monitored. Rising water levels were used as an indication of clogging within the galleries. Although the water level loggers were not essential for operating the scheme, they were found to be useful for analysing the performance of a gallery and deciding when to take proactive steps to reduce clogging and to avoid terminating inflow.

(b) Depending on the permeability of the aquifer being recharged, water table mounding and intersecting the ground may be an issue that requires regular monitoring. Piezometers that are screened over the water table should be installed near the recharge site of any MAR scheme with water level probes and loggers to record groundwater levels. Proactive steps can then be taken if a significant rise in the water table is observed.

(c) The use of probes measuring parameters such as electrical conductivity, dissolved oxygen and pH were useful in monitoring the progress of the recharged water and any changes in the groundwater/recharged water quality. In the current project these probes (with associated loggers) were installed in the central receiving chambers for each gallery and in selected monitoring bores along the length of the MAR scheme. The information from these loggers were routinely downloaded and analysed. This information was found to be valuable for assisting with determining transport times of the recharged water in the aquifer and other water quality issues. In addition, electrical conductivity and flow rate monitors were installed in the Amiad® filter system which was valuable for monitoring the supply of the treated wastewater to the galleries. In hindsight it was realised that the use of loggers, in particular for electrical conductivity in ALL of the monitoring wells would have been an worthwhile


investment for obtaining information on the performance of both of the infiltration gallery MAR schemes.

1.5 CONCLUSIONS

The experimental field sites for managed aquifer recharge provided an opportunity to investigate several research objectives related to infiltration galleries. Based on the research and technical objectives given in the Introduction and Research Aims the following conclusions can be made.

R1: to monitor recharge of water and determine clogging rates during aquifer recharge at the field sites:

The project demonstrated that treated wastewater can be efficiently and consistently recharged to the superficial aquifer in urban settings on the Swan Coastal Plain using infiltration galleries. In total, a volume of more than 36 ML was recharged to the aquifer at a rate of 50 KL per day over the three year life of the project.

The Floreat infiltration galleries demonstrated that a gravel-filled trench was not as effective or sustainable as the Atlantis Leach System® for infiltrating large quantities of secondary treated wastewater due to clogging with plant roots that occurred after one year of operation. A short-term experiment involving an increase in the inflow rate to the west (Atlantis Leach System®) gallery during the final two months of operating the FIG site revealed no evidence of clogging with an average inflow rate of 60 L/min. Although clogging did not occur in the Halls Head galleries, the water level in the gravel-filled gallery was observed to increase at a faster rate than the Atlantis Leach System® gallery.

The minimum residence time in the saturated zone at the Floreat site estimated using a three-dimensional MODFLOW-MODPATH simulation is 70 days 23 days between the galleries and the extraction bore. Due to the high rate of pumping from the extraction bore, dilution with background groundwater was of concern with regard to interpreting concentration reductions at specific monitoring bores down-gradient of the source. Based on a conservative transport model calibrated with chloride data, the estimated maximum dilution from background groundwater is 80% at the extraction bore, pumping at 250 KL/day (i.e. the recovered water consisted of 20% recharged water/80% groundwater).

The residence time below the Floreat infiltration galleries in the unsaturated zone was investigated using two different tracers that yielded different results. Based on the conservative tracer bromide, recharge water took about 3 days to reach 7 m below ground. The results from uranine predicted slightly longer times; however, sorption of uranine to the sediment may have influenced this result. The transition from Spearwood sand to the Tamala Limestone as well as variations in moisture content affecting flow rates are sources of heterogeneity in the unsaturated zone.

R2: to determine microbiological die-off during infiltration to, and residence time in the superficial aquifer:

Microbial pathogen die-off was assessed using in situ diffusion chamber experiments at the Floreat site. The T90 results for Cryptosporidium, adenovirus, rotavirus, coxsackievirus, bacteriophage MS2, E. coli, S. enterica and E. faecalis demonstrate that the aquifer has an active treatment capacity to remove


pathogens during passage of recharge water through the aquifer. The highest T90 value was 56 days for adenovirus.

These decay rates, when combined with the hydrogeological data can be used in risk assessment models to determine the residual risk from pathogens in the final recovered water (see Chapter 4 for details on health risk assessment).

T1: to identify water quality improvements that occur during the recharge of recycled water to groundwater (e.g. removal of pathogens, chemicals of concern and nutrients via biogeochemical processes):

Despite extensive sampling, enteric viruses or bacteriophage were not detected in the aquifer further than 1 metre from the infiltration galleries. Some periodic detection of thermotolerant coliforms and enterococci in water collected from various monitoring bores occurred; but, based on the studied decay rates of these microorganisms, it was considered that this is an artefact of the infiltration galleries being located on a sheep paddock and these microorganisms were sourced via surface contamination from sheep faeces, not from the recharged wastewater.

Removal of nutrients during passage through the unsaturated zone was assessed by comparing the concentrations in the gallery water with those in the groundwater sampled in adjacent monitoring bores. Large declines in phosphate were observed; however, there was no evidence of attenuation or removal of nitrate due to the aerobic nature of the aquifer preventing denitrification from occurring.

Concentration reductions in carbamazepine and oxazepam and to a lesser extent temazepam were observed in monitoring bores down-gradient from the recharge source. Based on mass recovery rates, a field assessment of carbamazepine revealed no degradation, consistent with column study results. Using the ratio of oxazepam to conservative carbamazepine, there was no evidence to suggest biodegradation of oxazepam, although due to dilution, the measured concentrations were below detection level in the recovered water.

T2: to test existing and novel treatment technologies (e.g. secondary wastewater treatment, tertiary treatment as well as using the aquifer to improve water quality) to establish suitability of the proposed applications for use with different MAR technologies to produce appropriate fit-for-purpose water:

It was shown that the Atlantis Leach System® gave a superior performance to the use of gravel in the construction and operation of the infiltration galleries. The gravel-filled gallery operated well for several months and then showed signs of clogging which eventually resulted in this gallery shutting down and needing the gravel to be replaced with the Atlantis Leach System®.

T3: to characterise aquifers identified as potential MAR sites to determine factors influencing recharge and recovery efficiencies:

Various parameters were monitored and targeted experiments were done to gain a better understanding of the movement of the water through the unsaturated zone and the aquifer. The results obtained were used to model the movement of the recycled water through the MAR scheme. The modelling was able to predict the residence time within the aquifer but the level of heterogeneity within the aquifer and the highly transmissive nature of the aquifer meant that an absolute understanding the transport and residence of


the water within the aquifer is still required. More research is needed on the characterisation of the aquifers at a local level at sites proposed for use as the location of MAR schemes.

T4: to identify management and operational requirements that are needed to ensure that future MAR schemes are operated in a manner that optimises the economic and environmentally suitable use of reclaimed waters, while managing apparent and inherent operational, health and environmental risks:

A series of issues relating to the operation and management of the infiltration galleries MAR schemes were observed and lessons were learnt from the occurrence of these issues. These issues related to factors such as the initial design and set up of the MAR schemes; understanding the reliability of the supply and the influence this may have on the day-to-day operation of the MAR scheme; the influence of dissolved iron on the quality of the recovered water; the appropriate installation and protection of the pipe system delivering the treated wastewater to the infiltration galleries; the clogging potential of the infiltration galleries and the appropriate management techniques if clogging occurs; and on the appropriateness of the monitoring equipment used to monitor the operation of the MAR schemes.


1.6 REFERENCES

Australian and New Zealand Environment and Conservation Council & Agriculture and Resource Management Council of Australia and New Zealand (ANZECC & ARMCANZ) 2000, The Australian and New Zealand Guidelines for Fresh and Marine Water Quality, National Water Quality Management Strategy Paper No 4, ANZECC & ARMCANZ, Canberra (http://www.ea.gov.au/water/quality/nwqms/index.html).

ASTM 2001, Standard Guide for Sampling Ground-water Monitoring Wells, D 4448-01.

Bekele, E., Toze, S., Rümmler, J., Hanna, J., Blair, P. & Turner, N. 2006, Improvements in wastewater quality from soil and aquifer passage using infiltration galleries: case study in Western Australia, Proceedings: 5th Intl Symp on Management of Aquifer Recharge (ISMAR5), 10–16 June 2005, Berlin, pp. 663–668.

Borchardt, M.A., Haas, N.L. & Hunt, R.J. 2004, ‘Vulnerability of drinking-water well in La Crosse, Wisconsin, to enteric-virus contamination from surface water contributions’, Applied and Environmental Microbiology, 70: 5937–5946.

Cook, P. & Herczeg, A. 1999, Environmental Tracers in Subsurface Hydrology, Springer, 529pp.

Dillon, P. & Toze, S. (eds) 2005, ‘Water Quality Improvements During Aquifer Storage and Recovery’, American Water Works Assoc. Research Foundation Report 91056F, 286pp.+2CDs.

Dillon, P., Page, D., Vanderzalm, J., Pavelic, P., Toze, S., Bekele, E., Sidhu, J., Prommer, H., Higginson, S., Regel, R., Rinck-Pfeiffer, S., Purdie, M., Pitman, C. & Wintgens, T. 2008, ‘A critical evaluation of combined engineered and aquifer treatment systems in water recycling’, Water Science and Technology, 57(5): 753–762.

Fegg, W. 2008, ‘Characteristics of the unsaturated soil under an Atlantis Leach System® for managed aquifer recharge’, unpublished Diploma thesis, University of Applied Sciences Weihenstephan, Department of Environmental Security, Triesdorf, Germany.

Hsieh, P.A., Wingle, W. & Healy, R.W. 2000, ‘VS2DI–A graphical software package for simulating fluid flow and solute or energy transport in variably saturated porous media’, US Geological Survey Water-Resources Investigations Report 99-4130, 16pp.

Ives, R.L., Kamarainen, A.M., John, D.E. & Rose, J.B. 2007, ‘Use of cell culture to assess Cryptosporidium parvum survival rates in natural groundwaters and surface waters’, Applied and Environmental Microbiology, 73: 5968–5970.

Jansons, J., Edmonds, L.W., Speight, B. & Bucens, M.R. 1989, ‘Survival of viruses in groundwater’, Water Research, 23: 301–306.


Käss, W. 1998, Tracing Technique in Geohydrology, A. Balkema, Rotterdam/Brookfield, 581pp.

Keswick, B.H., Gerba, C.P., Secor, S.l. & Cech, I. 1982, ‘Survival of enteric viruses and indicator bacteria in groundwater’, Journal of Environmental Science and Technology, A17(6): 903–912.

Lappala, E.G., Healy, R.W. & Weeks, E.P. 1987, ‘Documentation of computer program VS2D to solve the equations of fluid flow in variably saturated porous media’, US Geological Survey Water-Resources Investigations Report 83-4099, 184pp.

McArthur, W.M. & Bettenay, E. 1974, Development and distribution of soils of the Swan Coastal Plain, Western Australia, 2nd edition, CSIRO Australia, Soil Publication No. 16, 55pp.

Nasser, A.M. & Oman, S.D. 1999, ‘Quantitative assessment of the inactivation of pathogenic and indicator viruses in natural water sources’, Water Research, 33: 1748–1752.

NRMMC-EPHC-AHMC 2006, ‘Australian Guidelines for Water Recycling Managing Health and Environmental Risks (Phase 1)’, http://www.ephc.gov.au/taxonomy/term/39 (accessed July 2008).

NRMMC-EPHC-NHMRC 2008, ‘Australian Guidelines for Water Recycling: Managing Health and Environmental Risks (Phase 2) – Augmentation of Drinking Water Supplies’, http://www.ephc.gov.au/taxonomy/term/39 (accessed July 2008).

NRMMC-EPHC-NHMRC 2009, ‘Australian Guidelines for Water Recycling: Managing Health and Environmental Risks (Phase 2) – Managed Aquifer Recharge’, http://www.ephc.gov.au/taxonomy/term/39 (accessed 24 June 2009).

Patterson, B.M., Annabel, M., Bekele, E. & Furness, A. in press, ‘On-line Groundwater Velocity Probe: Laboratory Testing and Field Evaluation’.

Patterson, B.M. & Bekele, E. in press, ‘Measuring changes in groundwater velocity as a novel technique to estimate the wetting front migration rate of infiltrated water through the vadose zone’.

Piranha, J.M., Pacheco, A., Gamba, R.C., Mehnert, D.U., Garrafa, P. & Barrella, K.M. 2006, ‘Faecal contamination (viral and bacteria) detection in groundwater used for drinking purposes in São Paulo, Brazil’, Geomicrobiology Journal, 23: 279–283.

Playford, P.E., Cockbain, A.E. & Low, G.H. 1976, ‘Geology of the Perth Basin’, Geol. Surv. Western Australia, Bull. 124, 310pp.


Rueedi, J., Cronin, A.A. & Morris, B.L. 2009, ‘Estimation of sewer leakage to urban groundwater using depth-specific hydrochemistry’, Water and Environment Journal, 23(2): 134–144.

Rümmler, J., Bekele, E.B. & Toze, R.S.G. 2005, ‘Preliminary Hydrogeological Characterisation for Proposed Covered Infiltration Galleries at CSIRO Laboratory, Floreat, Western Australia’, Report for Water for a Healthy Country Flagship and Water Corporation, Western Australia, CSIRO Flagship Report, CSIRO, Perth.

Sobsey, M.D., Yates, M.V., Hsu, F.C., Lovelace, G., Battigelli, D., Margolin, A., Pillai, S.D. & Nwachuku, N. 2004, ‘Development and evaluation of methods to detect coliphages in large volumes of water’, Water Science and Technology, 50(1): 211–217.

Tapsell, P., Newsome, D. & Bastian, L. 2003, ‘Origin of yellow sand from Tamala Limestone on the Swan Coastal Plain, Western Australia’, Australian Journal of Earth Sciences, 50: 331–342.

Toze, S. & Hanna J. 2002, ‘The survival potential of enteric microbial pathogens in a treated effluent ASR project’, in: Management of Aquifer Recharge for Sustainability, (ed) P. Dillon, Balkema Publishers Australia, pp. 139–142.

Toze S., Hanna J., Smith, A. & Hick, W. 2002, Halls Head Indirect Treated Wastewater Reuse Scheme, Final report to Water Corporation, Western Australia. CSIRO Land and Water Technical Report October 2002.

Toze, S., Hanna, J., Smith, T., Edmonds, L. & McCrow, A. 2004, ‘Determination of water quality improvements due to the artificial recharge of treated effluent’, Wastewater Reuse and Groundwater Quality, IAHS Publication 285: 53–60.

Toze, S., Hanna, J. & Sidhu, J. September 2005, Microbial monitoring of the McGillivray Oval direct reuse scheme, Report to the Water Corporation, WA.

Wolf, L., Held, I., Eiswirth, M. & Hötzl, H. 2004, ‘Impact of leaky sewers on groundwater quality’, Acta Hydrochimica et Hydrobiologica, 32(4–5): 361–373.

Yates, M.V., Gerba, C.P. & Kelly, L.M. 1985, ‘Virus persistence in groundwater’, Applied and Environmental Microbiology, 49: 778–781.

Yates, M.V., Sterzenbach, L.D., Gerba, C.P. & Sinclair, N.A. 1990, ‘The effect of indigenous bacteria on virus survival in ground water’, Journal of Environmental Science and Health, A25: 81–100.


APPENDIX 1A – Details of the Monitoring Bores and Water Quality Logging Equipment for the MAR trials.

Table 1A-1. Summary of monitoring bores for the Floreat MAR trial

Bore/Date Installed

Easting (m)

Northing (m)

Ground elevation

(m AHD)

Slotted depth interval below ground (m)

Top of casing elevation (m AHD)

Lateral distance and direction relative to the West gallery

BH1/

03-Jun-04

385486.3 6464412 12.966 9.55 to 12.01

12.881 2.3 m west

BH2/

03-Jun-04

385486.8 6464429 13.101 7.80 to 12.04

13.061 2.5 m west

BH3/

03-Jun-04

385455.5 6464419.6 13.481 11.00 to 13.00

13.436 33.3 m west

BH4/

13-Apr-05

385484.4 6464419.6 NA 10.00 to 31.00

NA 4.2 m west

BH5/

03-Jun-05

385498.3 6464419.6 12.856 9.81 to 10.81

12.796 8.8 m east

BH6/

03-Jun-05

385483.6 6464419.6 13.061 11.01 to 12.01

13.0185 5 m west

BH7/

03-Jun-05

385481.6 6464421.6 13.106 12.54 to 13.54

13.036 7 m west

BH8/

03-Jun-05

385481.6 6464419.6 13.0985 14.54 to 15.54

13.051 7.3 m west

BH9/

03-Jun-05

385481.6 6464417.6 13.0735 13.54 to 14.54

13.036 7.6 m west

BH10/

03-Jun-05

385473.6 6464422.6 13.166 13.57 to 14.57

13.111 15 m west

BH11/

03-Jun-05

385473.6 6464419.6 13.171 18.57 to 19.57

13.141 15.3 m west

BH12/

03-Jun-05

385473.6 6464416.6 13.151 16.07 to 17.07

13.096 15.5 m west

BH13/

03-Jun-05

385468.6 6464419.6 13.241 16.14 to 17.14

13.201 20 m west

BH14/

03-Jun-05

385466.6 6464419.6 13.271 13.67 to 14.67

13.231 22 m west


Bore/Date Installed

Easting (m)

Northing (m)

Ground elevation

(m AHD)


Top of casing elevation (m AHD)

Lateral distance and direction relative to the West gallery

BH15/

03-Jun-05

385458.6 6464419.6 13.4585 16.29 to 17.29

13.406 30 m west

BH16/

03-Jun-05

385456.6 6464419.6 13.471 17.82 to 18.82

13.421 32 m west

BH17/

03-Jun-05

385438.6 6464419.6 NA 14.00 to 24.00

12.796 50 m west

BH18/

09-May-07

385488.7 6464422.7 NA 10.82 to 12.82

13.236 0.1 m west

BH19/

09-May-07

385488.7 6464409 NA 10.14 to 12.14

13.056 0.1 m west

BH20/

09-May-07

385488.7 6464429 NA 10.72 to 12.74

13.096 0.1 m west

BGRND01/

NA

385643.18 6464462 15.93 ?? to 15.00

16.492 185 m northeast

BGRND02/

04-Oct-06

385554 6464420 NA 11.23 to 12.23

NA 75 m east

BGRND03/

04-Oct-06

385557 6464421 NA 19.11 to 20.11

NA 75 m east

*All monitoring bores were constructed of 50 mm diameter PVC with slotted intervals, except the extraction bore. Coordinate locations are referenced to GDA94.


Table 1A-2. Summary of bores for the Halls Head MAR trial. The site was decommissioned due to redevelopment plans.

Bore/Date Installed

Easting (m)

Northing (m)

Ground elevation (m AHD)


Lateral distance and direction relative to the midpoint between the two galleries

HH_W1/

14-Mar-05

377001 6398810 2.844 2.89 to 5.89 4.25 m northwest

HH_W2/

14-Mar-05

377001 6398815 2.889 3.09 to 6.09 7.25 m northwest

HH_N1/

14-Mar-05

377015 6398815 2.869 3.03 to 6.03 15 m northeast

HH_E1/

14-Mar-05

377007 6398803 2.844 3.06 to 6.06 4.25 m southeast

HH_E2/

14-Mar-05

377009 6398802 2.809 3.085 to 6.085

7.25 m southeast

HH_S1/

14-Mar-05

376995 6398798 2.827 3.053 to 6.053

15 m southwest

2/84 (Background monitoring bore)

377166 6398750 4.392 NA; within the superficial aquifer

190 m southeast

6/88 (Background monitoring bore)

376961 6398634 NA NA; within the superficial aquifer

185 m southwest

*All monitoring bores installed for the project were constructed of 50 mm diameter PVC with slotted intervals. Eastings and Northings are referenced to GDA94.


APPENDIX 1B – Uranine and Bromide Tracer Test: Sampling Program and Fluorometer Calibration

The sampling program for the combined uranine + bromide tracer experiment involved sampling water from the suction cups intensely at the start and then less frequently over time. Nearly 400 water samples were extracted from the suction cup lysimeters over a two week period (24 September to 8 October 2007).

There were four stations of suction lysimeters (SL1 to SL4) as depicted in Figure 1-12. During the first four days, samples were collected on average every 20 minutes with about 1.5 hours being the longest time interval between sampling. Thereafter, the timing between collecting samples was more irregular. As indicated in Figure 1B-1, sampling at each station occurred less often after four days, particularly during overnight periods.

Figure 1B-1. Plot showing the frequency of sample collection from each station for the combined uranine + bromide tracer experiment. The majority of samples were analysed for uranine.

Suction cup lysimeters were sampled individually and the line was vacuumed twice to flush previous water. A syringe was used to extract 5 mL per suction cup and each sample was then placed in a plastic vial containing 0.55 mL of buffer solution to maintain a pH of about 8.5, as recommended to obtain maximal fluorescence from uranine (Käss et al, 1988). An additional sample (10 mL) was extracted from the syringe for bromide analysis. Uranine samples were wrapped in aluminium foil and both samples for analysis were stored in a dark, cold place before transporting them to a nearby cold storage facility. The uranine samples were analysed by Wolfgang Fegg using a Picofluor Handheld Dual Channel Fluorometer from Turner Biosystem. Measurements were taken using the fluorometer within 10 days after the last water sample was collected. Fluorescence methods are relatively inexpensive compared with bromide analysis. A selection of water samples were delivered to the Chemistry Centre WA for bromide analysis after the interpretation of uranine breakthrough was finalised. Water samples for bromide analysis were from two suction cups, labelled as Ia (SL2) and Nb (SL3), which are located respectively 8.3 and 8.1 m below ground.

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

20.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

Time since tracer infiltrated (days)

Tim

e in

terv

al b

etw

een

retu

rnin

g t

o a

sta

tion

to

sa

mpl

e m

ultip

le ly

sim

ete

rs (

hour

s)


For the analysis of uranine in the water samples, the fluorometer was first calibrated using a single-point standard set to 100 ppb. The standard consisted of deionised water without buffer and a linear range was tested from 1 ppb to 3.5 ppm (equivalent to 3.5 mg/L). Every sample reading >3.5 mg/L was diluted 10 times, and then re-measured with the fluorometer so as not to exceed the range of calibration.


APPENDIX 1C – MODFLOW Model for the Floreat Infiltration Galleries Site

The dimensions of the 3D model domain (length x width x thickness) were 396 m x 277 m x 30 m. Rectangular grid cells were used and the minimum and maximum grid cell dimensions were 0.45 m and 6 m, respectively. Seven layers were used to define the cross-section using different layer thicknesses to improve the resolution of slotted intervals for the observation bores and the resolution of the water table height. The elevation of the top layer was 13 m AHD. The layer thicknesses were as follows: Layer 1 (12.3 m); Layer 2 (1.7 m); Layer 3 (1.9 m); Layer 4 (1.4 m); Layer 5 (1.3 m); Layer 6 (5.4 m); and Layer 7 (6.3 m). The top of Layer 2 coincided with the water table. Recharge from the galleries was represented using the recharge boundary condition and are depicted as linear features projected on the top of Layer 2 in Figure 1C-1. Recharge was applied directly to the water table and flow through the unsaturated zone was not modelled. Constant head boundary conditions were assigned along the east and west sides of the domain.

No flow boundary Constant head (east) boundary condition

Constant head (west) boundary condition

No

flow

bo

un

dary

No

flow

bo

un

dary

Figure 1C-1. Three-dimensional model domain showing the placement of observation bores for hydraulic head (green), observation bores for chloride concentration data (blue) and the pumping bore (red). The grid is visible only for the tops of Layers 1 and 2, and for the model base for clarity.

Uniform properties were applied to all layers. In the run settings, the top layer was treated as unconfined and underlying layers were assigned the default setting, which allowed them to have varying transmissivity, depending on the saturated thickness and the hydraulic conductivity. A specific storage of 1.0E-5 and a specific yield of 0.1 were assigned in the models. These parameters were taken from the literature and were not calibrated. Permeability anisotropy was only briefly investigated as there was insufficient data for the Tamala Limestone.

A two-step process was used to calibrate aquifer properties based on modeling two separate events. The first simulation covered a period of 188 days and included transient pumping rates. The simulation involved groundwater level changes in the Tamala aquifer in response to pumping at 250 KL/day from the extraction bore, temporarily shutting off the pump for two


days (June 22 to 24, 2006) and then restarting the pump. The pump was initially restarted at 350 KL/day after the drillers serviced the bore and this was included in the model as indicated in the inputs (Table 1C-1). Recharge to the galleries was assigned 365,000 mm/yr, equivalent to 1 m/day to represent the hydraulic loading rate for effluent recharge (Table 1C-2).

Table 1C-1. Model inputs for the pumping schedule for BH17. A period of 184 days of pumping was used to stabilise hydraulic heads before the pump was turned off on June 22, 2006.

Start time (days) End time (days) Rate (m3/day)

0 184 250

184 186.125 0

186.125 188 350

Table 1C-2. Model inputs for the recharge boundary condition for the infiltration galleries

Start time (days) End time (days) Recharge (mm/yr)

0 188 365,000

Changes in the regional hydraulic gradient were negligible during this period of transient pumping, thus the background hydraulic gradient in the model was set to zero. An initial head of 2.5 m was assigned uniformly to the model and the east the west boundary conditions were assigned the same value. Relative changes in water levels predicted by the model in response to turning off the pump were compared with observed changes, rather than the absolute water table elevations. Groundwater level loggers in BH6, BH8 and BH15 recorded water levels every 30 minutes. These data were compared with MODFLOW results based on a permeability of 100 m/day. Other permeability values were tested in the model until a best fit to the groundwater level data was obtained. Porosity is not used in MODFLOW simulations, thus a second step was needed to calibrate porosity.

The second step allowed us to calibrate porosity as well as permeability across more sites by simulating the breakthrough of high chloride concentrations at different bores over several months. The MT3DMS module in MODFLOW was used for modelling. Less emphasis was placed on calibrating dispersivity with the model as this parameter has less impact on breakthrough times, although it does affect the magnitude and shape of the breakthrough curve. Longitudinal dispersivity values of 0.0 m, 0.5 m and 1.0 m were tested and default settings for the dispersivity tensor were used (i.e. the ratio of horizontal/longitudinal dispersivity was 0.1; the ratio of vertical/longitudinal dispersivity was 0.01).

The same domain as the hydraulic model was used (Figure 1C-1). A close-up view of the model domain showing the locations of bores used for calibration based on observed chloride concentrations is depicted in Figure 1C-2. Unlike the hydraulic model described above, the solute transport model required inputs for chloride concentration. In particular, measured chloride data at the water table were obtained by using data from BH1 rather than measured concentrations from the effluent. This decision was made to avoid dilution/dispersion effects in the unsaturated zone that were not included in the model of purely saturated flow. The highest measured chloride in the recharge water was 512 mg/L, whereas the highest measured chloride sampled from BH1 during this period was 449 mg/L.

The groundwater chloride data were applied to fictitious recharge galleries that were relocated closer to BH1 as shown in Figure 1C-2.


BH17 BH16

BH8

BH6

BH15BH3

BH2

BH1Scale

5m

5m

Figure 1C-2. A close-up of the model domain and grid, showing the bores used for calibration, the true location of the galleries (pink), and the starting position of mathematical particles (*) for the MODPATH simulation. For the chloride simulation, gallery recharge was relocated closer to BH1 (shown in blue) to represent observed recharge concentrations at the water table.

The concentrations of groundwater chloride measured in the several background bores were between 100 and 212 mg/L (samples count = 64) with variability observed between bores due to depth of the slotted casing below the water table. A uniform, initial condition for groundwater chloride concentration of 150 mg/L was assumed in the model based on water samples from the background bores.

A time-series of chloride concentrations in the recharge water was assigned as input to the model. Recharge concentrations of chloride sampled in the galleries were highly variable over time (Figure 1C-3). Over a ten-month period beginning in mid-January 2006, there were major changes in chloride concentrations in the recharge water. A similar pattern was observed later in groundwater sampled from bores adjacent to the galleries and it was partially tracked in bores further down-gradient. The model input recharge concentrations shown in Figure 1C-3 were approximated using groundwater chloride sampled from BH1. For the mass transport simulation, the chloride time series pattern was applied as a recharge boundary condition directly to the water table and the galleries were shifted closer to BH1 as shown in blue in Figure 1C-2.


145

195

245

295

345

395

445

495

545

0 100 200 300 400 500 600 700

Day number

Cl (

mg

/l)

BH1

BH2

Model input

West gallery

East gallery

Figure 1C-3. Concentrations of chloride measured in the galleries and two boreholes (BH1 and BH2) located near the galleries and slotted across the water-table. Model inputs for the recharge concentration were approximated using data from BH1. The day numbers are relative to day zero, an arbitrary date approximately four months before the MAR trial started.

The model settings for the chloride breakthrough simulation included a constant effluent recharge rate of 365,000 mm/yr, equivalent to 1 m/day, a constant pumping rate of 250 KL/day for the extraction bore and a background hydraulic gradient of 0.1% estimated from the Perth Groundwater Atlas (Western Australia, Department of Environment, 2nd Edition, 2004; Tables 1C-3 and 1C-4). The west hydraulic boundary was set to 2.5 m AHD, whereas the east hydraulic boundary was set to 2.88 m. An initial head distribution of 2.5 m was assigned.

Table 1C-3. Model inputs for the pumping schedule for BH17. A period of 100 days of imposing the head gradient without pumping and without gallery recharge was used to stabilize hydraulic heads before chloride infiltration commenced.

Start time (days) End time (days) Rate (m3/day)

0 100 0

100 700 250

Table 1C-4. Model inputs for the recharge boundary condition for the infiltration galleries

Start time (days) End time (days) Recharge (mm/yr)

0 100 0

100 700 365,000

The most suitable aquifer transport parameters derived from the calibration stage were used in a MODPATH simulation to estimate the minimum travel time between the galleries and the extraction bore. A line of mathematical particles was placed conceptually along the west gallery and the model transported particles by advection across the flow field (Figure 1C-2). The same flow boundary conditions were used as in the second stage of model calibration (i.e. the same recharge and pumping rates), except recharge was applied at the true location of the two galleries, rather than near BH1 as needed for the chloride simulation.


APPENDIX 1D – Inorganic Chemistry Analytical Procedures

Table 1D-1. Analytical procedures and detection limits for the inorganic chemistry. Source: Environmental Chemistry Laboratory, Chemistry Centre of WA.

Code Analyte Reporting Limit Method

Alkalinity (Total

expressed as CaCO3 1 mg/L

HCO3 1 mg/L

iALK1WATI (Alkalinity /

HCO3/ CO3)

CO3 1 mg/L

Titration of sample aliquot with standard acid solution to potentiometric (or colorimetric) equivalence points of pH

8.3 and 4.5 at 25°C. Titrant volumes to the two pHs (8.3 and 4.5) allow calculation of HCO3, HCO3 and OH

concentrations.

iAMMN1WFIA (N_NH3)

Ammonia as expressed

as Nitrogen 0.01 mg/L

Automatic flow injection method is based on the Berthelot reaction. Ammonia reacts in alkaline solution with

hypochlorite to form monochloramine which, in the presence of phenol, catalytic amount of nitroprusside, and

excess hypochlorite gives indophenol blue. The indophenol blue measured at 630 nm is proportional to the

original ammonia concentration.

iANIO1WAIC (Br) Bromide 0.02 mg/L

Chromatographic determination of the bromide using single column ion chromatograph with UV detection. This

method is based on APHA method 4110B.

iANIO1WAIC (Cl) Chloride 0.5 mg/L

Chromatographic determination of the chloride and sulphate using single column ion chromatograph with direct

conductivity detection.This method is based on APHA method 4110B.

iEC1WZSE (Econd) Conductivity 0.2 mS/m

Electrical conductivity is measured using conductivity meter than has been calibrated using three standards.

Values are corrected to 25C

Ag 0.005 mg/L

Al 0.005 mg/L

As 0.05 mg/L

iMET1WCICP (Metals)

B 0.02 mg/L

The sample is filtered through a 0.45 μm membrane filter and acidified to <pH 2 with HNO3. Determination of

metals is by inductively coupled emission spectroscopy using internal standard/suppressant solutions. Calibration

is carried out using external standards. (Reporting Limits based on total dissolved solids to less than 30000 mg/L)



Ba 0.002 mg/L

Ca 0.1 mg/L

Cd 0.002 mg/L

Co 0.005 mg/L

Cr 0.001 mg/L

Cu 0.002 mg/L

Fe 0.005 mg/L

K 0.1 mg/L

Mg 0.1 mg/L

Mn 0.001 mg/L

Mo 0.02 mg/L

Na 0.1 mg/L

Ni 0.01 mg/L

Pb 0.02 mg/L

Sb 0.05 mg/L

Se 0.05 mg/L

SiO2_Si 0.5 mg/L

SO4_S 0.1 mg/L

V 0.005 mg/L



Zn 0.005 mg/L

Sb 0.0001 mg/L

As 0.001 mg/L

Cd 0.0001 mg/L

Cr 0.0005 mg/L

Cu 0.001 mg/L

Pb 0.0001 mg/L

Mo 0.001 mg/L

N 0.001 mg/L

Se 0.001 mg/L

U 0.0001 mg/L

V 0.0001 mg/L

iMET1WCMS (Metals)

Zinc 0.005 mg/L

The sample is filtered through a 0.45 μm membrane filter and acidified to <pH 2 with HNO3. Determination of

metals is by inductively coupled mass spectroscopy using internal standard solutions. Calibration is carried out

using external standards. (Reporting Limits based on total dissolved solids to less than 3000 mg/L)

iNTAN1WFIA (N_NO3)

N_NO3 (nitrate + nitrite

as expressed as

nitrogen)

0.01 mg/L

Automatic flow injection method where nitrate is quantitatively reduced to nitrite by passage of the sample through

a copper coated cadmium column. The nitrite (reduced nitrate plus original nitrite) is then determined by

diazotization with sulfanilamide under acidic condition to form a diazonium ion. The resulting diazonium ion is

complexed with N-(1-naphthyl)-ethylenediamine dihydrochloride. The resulting pink dye absorbs at 540 nm. A

correction may be made for any nitrite present by analysing without the reduction step.

iP1WTFIA (P_SR)

Soluble reactive

phosphate as expressed

as phosphorus

0.01 mg/L Automatic flow injection method based on reaction of Orthophosphate (PO43-) with ammonium molybdate and

antimony potassium tartrate under acidic conditions to form a yellow complex. This complex is reduced with

ascorbic acid to form a blue complex which absorbs light at 880 nm. The ascorbic acid and molybdate reagents



are merged on the chemistry manifold and the reagent stream is then merged with the carrier stream. The

absorbance is proportional to the concentration of orthophosphate in the sample.

iPH1WASE (pH) pH 0.1 pH units pH is measured using pH meter that has been calibrated using three buffers.

iSOL1WDGR

(TDS_180) Total Dissolved Solids 5 mg/L

A well-mixed sample is filtered through a standard glass fibre filter, and the filtrate is evaporated to dryness in a

weighed dish and dried to constant weight at 180°C. Alternatively the procedure may weigh the dish after an

elapsed time (12 hours). The increase in dish weight represents the total dissolved solids. This procedure

maybe used for drying at other temperatures.

The results may not agree with the theoretical value for solids calculated from chemical analysis of sample.

Approximate methods for correlating chemical analysis with dissolved solids are available.

iF1WASE (F) Flouride 0.05 mg/L

Fluoride is measured using a selective ion sensor. It can be used to measure the activity or concentration of

fluoride in aqueous samples by use of an appropriate calibration curve. However the fluoride activity depends on

the total ionic strength of the sample. The electrode does not respond to bound or complexed fluoride. These

difficulties are overcome by the addition of a buffer solution of high total ionic strength which swamps variations in

sample ionic strength and which contains a chelate to complex aluminium preferentially.

iCTO1WDCO (DOC) Dissolved Organic

Carbon 1 mg/L

Dissolved organic carbon (DOC) is the sum measure of all organic carbon containing species in water after

filtration through a 0.45 micron membrane. The standard measurement of DOC involves the addition of acid by

the instrument to remove inorganic carbon species; namely carbonate, bicarbonate and dissolved carbon dioxide

and then purging of the sample to remove the carbon dioxide generated. DOC is then measured on this solution

and is technically defined as nonpurgeable organic carbon (NPOC).

iSOL1WPGR

(Solid_sus) Total Suspended Solids 1 mg/L

A well-mixed sample is filtered through a weighed standard glass-fibre filter and the residue retained on the filter

is dried to a constant weight at 103 to 105°C. The increase in weight of the filter represents the total suspended

solids.

iHG1WCVG (Hg) Mercury 0.0001 mg/L Mercury compounds in solution are converted to inorganic mercuric salts by persulphate oxidation. The mercuric

compounds are then reduced by stannous chloride to the volatile elemental form. The mercury vapour is swept



by an inert gas into the optical cell where its absorption is measured at 253.7 nm.

iNP1WTFIA (N_total) Total Nitrogen 0.02 mg/L

Simultaneous determination of total nitrogen and total phosphorus is based on the alkaline persulphate oxidation

in autoclave for 30 minutes at 15 psi (120ºC). Nitrogen compounds convert to nitrate in alkaline medium.

However phosphorus compounds containing polyphosphates are resistant to alkaline persulphate oxidation. The

oxidation of phosphorus compounds must be performed in acidic medium by using an oxidising reagent with 1:1.5

K2S208 to NaOH ratio. The pH changes from an initial value of 12.8 to a final pH of 2.0 – 2.1 over the course of

autoclave digestion. During the digestion process, the nitrogen compounds are oxidised first in the alkaline

medium. As the digestion proceeds, the resulting bisulphate ions lowers the pH and permit oxidation of not only

nitrogen compounds, but also phosphorus compounds containing polyphosphates which are oxidised completely

when the pH becomes less than 2.2.

The oxidised products of nitrogen and phosphorus compounds are nitrate and orthophosphate respectively and

both are measured colorimetrically.

iNTK1CALC (N_TK) Total Kjeldhal Nitrogen 0.02 mg/L Total kjeldhal nitrogen is calculated from total nitrogen and nitrate

N_TK = N_total- N_NO3

eBOD1ECSE

(BOD/BOD5)

Biochemical Oxygen

Demand 5 mg/L

Biochemical oxygen demand is currently out sourced to NMI.

The sample is diluted with seeding solution, placed to overflowing in airtight bottle of specific size and incubated

at specific temperature (normally 20oC) for 5 days. Dissolved oxygen is measured initially and after incubation

and the BOD is computed from the difference between initial and final DO.


Table 1D-2. Number of times analytes measured in different monitoring sites

Parameter (Unit of Measure) East and West Galleries BH1 BH8 BH12 BH16 BH17 Background bores

Water temperature (oC) 53 43 37 24 28 36 52

pH 40 35 27 16 18 28 33

Electrical Conductivity (mS m-1) 50 30 28 20 21 26 49

O.R.P. (mV) 49 30 28 20 21 26 48

Eh (mV-SHE) 49 33 35 22 26 27 45

Dissolved Oxygen (mg L-1) 53 42 37 24 28 35 52

TOC (mg L-1) 14 11 9 3 2 11 11

DOC (mg L-1) 42 28 25 8 9 25 36

N_Total (mg L-1) 42 28 25 8 9 25 36

NO3_N (mg L-1) 42 28 11 4 3 25 24

P_SR (mg L-1) 42 28 25 4 17 25 37

SO4_S (mg L-1) 8 5 5 2 2 4 3

Suspended Solids (mg L-1) 20 13 11 4 3 12 11

Total Dissolved Solids (mg L-1) 41 25 24 7 7 24 35


APPENDIX 1E – Organic Chemistry Analytical Procedures Source: Ted Spadek, Chemistry Laboratory, Chemistry Centre of WA.

A Gilson syringe pump and liquid handling system was coupled to dual Waters 510 pumps to load a 10 mL aliquot into a proprietary sample loop. A six-port Rheodyne rotary valve was used to direct flow to the sample loop and to the on-line SPE column.

The analytical system used a 50 x 4.6 mm Zorbax 300 Extend-C18 guard column as the on-line SPE column coupled to a 150 x 4.6 mm Zorbax 300Extend-C18 analytical column.

The instrument used was an Agilent 1200 liquid chromatograph coupled to an Agilent 6410 triple quadrupole mass spectrometer. Confirmation of analytes was on the basis of retention time and two mass transitions under specific conditions. Instrument conditions are provided in Tables 1E-1 and 1E-2. Quantitation was by direct comparison with certified standards analysed alongside samples.

The instrument conditions are described below:

Mass Spectrometer

Electrospray Ionisation Positive (ESI+ve)

Table 1E-1. Instrument conditions for mass spectrometer

Analyte Precursor Quantifying Ion Qualifying Ion

temazepam 301 283 255 oxazepam 287 269 241 diazepam 285 257 154 carbamazepine 237 194 192 phenytoin 253 225 182

Liquid Chromatograph

Solvent A: 10mM Ammonium formate, pH 3.0

Solvent B: Methanol (redistilled)

Table 1E-2. Instrument conditions for the liquid chromatograph

Time (minutes) Flow (mL/min) %Solvent B Valve Position 0.0 0.50 0 1 7.9 0.50 0 8.0 0.50 20 10.0 0.50 20 20.0 0.50 100 24.0 0.50 100 24.1 0.50 0 27.0 0.50 0 1 28.0 0.50 0


APPENDIX 1F – Methods for the Detection of Microorganisms in Samples Collected from the Infiltration Galleries Sites

Faecal coliforms and enterococci

Analysis for thermotolerant coliforms and enterococci was undertaken using the standard membrane filtration method. For groundwater this involved, passing five replicate 100 mL volumes through 0.22 µm pore size filters (Durapore, Milipore). For the five replicate treated wastewater samples this involved suspending 1 mL volumes in approximately 100 mL of sterile PBS which was then passed through the filters. The filters were then placed onto a plate of selective media. For thermotolerant coliforms the filter was placed on a plate of Chromocult Coliform agar (Merk) and for enterococci onto a plate of Chromocult Enterococci Agar (Merk). All thermotolerant coliform plates were incubated at 44°C for 24 hours and the plates of enterococci media were incubated at 37°C overnight. After incubation all of the plates were examined for the presence of representative colonies on the surface of the filters. For thermotolerant coliforms colonies with a dark blue to violet colour were recorded as thermotolerant coliforms. For enterococci dark pink colonies were recorded as enterococci. Results for both microbial types are the average of five replicates and recorded as colony forming units per 100 mL.

F+ specific bacteriophage

Bacteriophages were detected using the enrichment method (Sobsey et al., 2004). In brief, this involved the partitioning of samples into triplicate 500 mL volumes of groundwater or 50 mL of treated wastewater. Each of these volumes of sample was amended with 10X concentrated Tryptose Yeast extract Glucose (TYG) broth (+ cation, glucose and antibiotic supplements) both to a final concentration of 1X media. 1 mL of a log culture of the E. coli host (F+ Amp) was added to each bottle containing groundwater and amendments. All bottles were incubated overnight at 37oC. After incubation, six 10 µL samples from each were spotted onto the surface of fresh TYG agar (TYGA) plates (+ supplements) with a double overlay of soft TYGA inoculated with a log culture of the E. coli F+Amp host strain. The plates spotted with the samples from the enrichment cultures were allowed to dry briefly in a laminar flow cabinet and then incubated overnight at 37°C. Following incubation each plate was examined for the presence of non-growth where the enrichment cultures had been spotted onto the plate. The presence of no-growth indicated the presence of male-specific coliphage in the enrichment. Results were recorded as a positive result in the volume of the enrichment culture.

Enteric viruses

The presence of enteric viruses was determined using (RT-)PCR. This involved the extraction of virus DNA or RNA (DNA from adenovirus and RNA from the enteroviruses) from 140 µL of the water concentrate using Nucleospin® Virus RNA extraction kit (Macherey-Nagel) following the manufacturer’s instructions (note that the Nucleospin Virus RNA kit is rated for the extraction of both RNA and DNA). A 1 µL volume from the extracted DNA or RNA solutions was used as the template for each PCR reaction. All extracted virus nucleic acid templates were tested in triplicate (RT-)PCR reactions.

All of the RT-PCR and PCR reactions were run on a BioRad iCycler, using iScript PCR Supermix for the detection of adenovirus and iScript One-Step RT-PCR kit. Primers used for the specific detection of adenovirus DNA were the HexAA 1885 and HexAA 1913 (Allard et al., 1992; Allard et al., 1990). Primers used for detection of members of the enterovirus group were the Ent-up and Ent-down primers (Abbaszadegan et al., 1993).


PCR thermocycling conditions for the detection of adenovirus were: initial incubation at 95ºC for 8 min and 30 sec, then 55 cycles at 95ºC for 30 sec, 55ºC for 20 sec and 72ºC for 20 sec (with a 2 second increase of the extension step in each cycle after the 25th cycle). The final cycle had an extension time of 5 min at 72ºC. The Reverse Transcriptase PCR (RT-PCR) reaction for the enteroviruses commenced with incubation at 50ºC for 30 minutes to allow transcription of the viral RNA to cDNA. The PCR reaction steps had an initial incubation at 95ºC for 8 min followed by 55 cycles of 95ºC for 30 sec, 58ºC for 20 sec and 72ºC for 20 sec (with a 2 second increase of the extension step in each cycle after the 25th cycle). The final cycle had an extension time of 5 min at 72ºC.

All (RT-)PCR reactions were then visualised on a 2% agarose gel stained with ethidium bromide run at 100 v for approximately 30 minutes to detect positive results.

References

Abbaszadegan, M., Huber, M.S., Gerba, C.P. & Pepper, I.L. 1993, ‘Detection of enteroviruses in groundwater with the polymerase chain reaction’, Applied and Environmental Microbiology, 59(5): 1318–1324.

Allard, A., Albinsson, B. & Wadell, G. 1992, ‘Detection of adenovirus in stools from healthy persons and patients with diarrhea by two-step polymerase chain reaction’, Journal of Medical Virology, 37: 149–157.

Allard, A., Girones, R., Juto, P. & Wadell, G. 1990, ‘Polymerase chain reaction for detection of adenovirus in stool samples’, Journal of Clinical Microbiology, 28: 2659–2667.

Sobsey, M.D., Yates, M.V., Hsu, F.C., Lovelace, G., Battigelli, D., Margolin, A., Pillai, S.D. & Nwachuku, N. 2004, ‘Development and evaluation of methods to detect coliphages in large volumes of water’, Water Science and Technology, 50(1): 211–217.


APPENDIX 1G – Preparation of Enteric Microorganisms for Pathogen Decay Experiments and Detection of Viable Microorganisms in Diffusion Chambers

Preparation of enteric microorganisms

E. coli and Salmonella enterica were cultured in nutrient broth and Enterococcus faecalis was cultured in Brain Heart Infusion broth, all at 37oC overnight in a shaking incubator. After culture, each bacterial culture was washed three times in P-buffer (centrifuging at 4,750 xg for 5 minutes and discarding supernatant) to remove excess culture media and then resuspended in 2 mL of P-buffer.

Adenovirus and coxsackievirus had been previously cultured in African Green Monkey Kidney cells and frozen as crude cell extracts at -80oC until used. Before use a portion of the crude viral extracts were washed using P-buffer to remove the cell debris and nutrients using the method outlined in Gordon and Toze (2003). The final viral concentrate was suspended in P-buffer to a detectable number of approximately 6–7 log of virus copy numbers mL-1. The final concentrate was resuspended in 2 mL of P-buffer.

The Cryptosporidium oocysts had been isolated from infected human faecal material using the method described in Gobet and Toze (2001) and resuspended in P-buffer.

All microorganisms suspended in P-buffer were stored at 4 oC and used within 24 hours.

Assembly of diffusion chambers

Prior to assembly of the diffusion chambers, groundwater was aseptically collected from a monitoring well close to the infiltration galleries (MB8). The groundwater was collected in sterile borosilicate bottles and transported back to the lab on ice. Here the groundwater was divided into two equal portions

One fraction of groundwater was seeded with the bacterial strains to achieve a final cell number of each bacterium between 6 and 7 log mL-1. The other fraction was seeded with the enteric viruses and Cryptosporidium oocysts to achieve a final number of approximately 5 log viral particles or oocysts mL-1. Each of these seeded groundwater fractions were used to fill 18 diffusion chambers. This equates to three replicate samples being able to be collected on each of the six sampling occasions (ignoring Time 0). The remaining volume of each of the seeded fractions was retained for use as the Time 0 sample.

All of the assembled diffusion chambers were connected into a chain of diffusion chambers via stainless steel rope and suspended down the monitoring bore (MB9). The diffusion chambers were organised in the chain so that three replicate chambers of each of the seeded fractions (i.e. bacteria; and virus and Cryptosporidium) could be efficiently removed from the bottom of the chain of diffusion chambers on each sampling occasion (days 7, 14, 21, 35).

Collection and sampling of diffusion chambers

The collected diffusion chambers were sampled by removing the water and microorganisms from each diffusion chamber using a sterile 10 mL syringe and 21 gauge needle. The


collected water was transferred to a sterile 10 mL polypropylene tube and stored at 4oC until processed. All of the samples used for analysis of bacterial numbers were processed within 2 hours of extraction of the water from the diffusion chambers. Samples for analysis of Cryptosporidium numbers were processed within 24 hours of collection. Samples used for analysis of virus numbers were frozen at -80 oC and all processed at one time at the completion of sampling.

Detection of viable bacteria

The number of viable cells for each of the target bacteria was done using spread plating on appropriate selective culture media. This involved diluting each sample 10-fold in P-buffer to levels which were considered a level where approximately one bacterial cell per 100 μL would be present in the highest dilution. Five 100 µL replicates from this dilution and each of the next two highest dilutions were then spread-plated onto selective media using sterile glass spreaders. The level of dilution used on each sampling event was dependent on the results obtained in the previous sampling event. For time 0 the dilution series used was estimated from the expected final numbers in the groundwater based on the number of bacteria in the concentrates used to seed the groundwater portions prior to set up of the diffusion chambers. If it was considered necessary because of very low bacterial numbers occurring on the previous sampling occasion, five 1 mL volumes of the water sample from the diffusion chambers were added to 10 mL volumes of sterile p-buffer and filtered through 45 mm 0.2µm pore size nitrocellulose filters. The filters were then placed on the surface of the appropriate media.

The medium used for the detection of E. coli was Chromocult® coliform agar (Merk); the medium for Enterococcus faecalis was Chromocult® enterocci agar (Merk); the medium for Salmonella enterica was XLD; and the detection medium for Campylobacter jejuni was Campylobacter blood-free selective agar (Oxoid) supplemented with CCDA selective supplement (Oxoid). All of the inoculated plates were incubated overnight aerobically at 37oC except for the Campylobacter plates which were incubated for 48 hours at 37oC in anaerobic jars (Oxoid) using CampyGeb sachets (Oxoid).

After incubation the plates were all examined for the presence of colonies with characteristics of the target microorganisms and the dilution with the most appropriate number of colonies on each replicate plate (preferably between 30 and 300 colonies) was used to determine the number of viable cells mL-1.

Detection of viruses and viable Cryptosporidium

The water from the chambers seeded with the viruses and Cryptosporidium oocysts was divided so that 1 mL was used for virus analysis and 5 mL used for determining the number of viable Cryptosporidium oocysts.

Viable Cryptosporidium oocysts were quantified using the vital dye staining method outlined by Campbell et al. (1992). Working solutions of 4,6-diamidino-2-phenylindole (DAPI) (2 mg mL-1 in absolute methanol) and propidium iodode (PI) (1 mg mL-1 in DMSO) were prepared and stored in dark at 4 °C. 10 µL DAPI was added to each sample and incubated for 2 hours at 37oC. After incubation, 10 µL of PI was added to each sample and incubated for 5 minutes at room temperature. Each sample was filtered through a 0.22 µm black polycarbonate filter (Millipore) which was then mounted on a glass slide. The filters were examined by fluorescence microscopy for the presence of viable and non-viable oocysts using a UV wavelength filter block (350 nm excitation, 450 nm emission) for DAPI, and a green light wavelength filter block (500 nm excitation, 630 nm emission) for PI. Twenty fields of view


were examined counting stained oocysts recording the number of blue white DAPI positive, PI negative oocysts (viable) and red PI positive oocysts (non-viable) in each field of view. The number of oocysts was averaged and then used to calculate a number of viable and non-viable oocysts mL-1.

The presence of enteric viruses was determined using quantitative reverse transcriptase PCR or PCR ((RT)-PCR). Viral DNA or RNA was extracted from 140 µL of the water collected from the diffusion chambers using the Nucleospin® Virus RNA extraction kit (Macherey-Nagel) using the manufacturer’s instructions (note that the Nucleospin Virus RNA kit is able to extract both RNA and DNA). A 1 µL volume from the solution of extracted nucleic acid was used as the template for the (RT-)PCR reactions. All extracted nucleic acid solutions were tested in triplicate reactions.

Adenovirus DNA was detected using the Heim I and Heim 11 primers (Heim et al., 2003). Rotavirus RNA was detected using the Beg9 and End9 primers (Gouvea et al., 1990) and coxsackievirus RNA was detected using the ENT-Up and ENT-Down primers (Abbaszadegan and Delong 1997).

Real-time quantitative (RT-)PCR reactions were run on a BioRad iCycler. The DNA virus Adenovirus was detected using iScript PCR Supermix + SYBR Green (Bio-Rad) while the RNA viruses, rotavirus and coxsackievirus, were detected using iScript One-Step RT-PCR kit with SYBR Green (Bio-Rad). PCR Thermal cycling conditions for adenovirus (DNA virus) were: initial incubation at 95ºC for 8 min and 30 sec, then 55 cycles at 95ºC for 30 sec, 55ºC for 20 sec and 72ºC for 20 sec (with a 2 second increase of the extension step in each cycle after the 25th cycle). The final cycle had an extension time of 5 min at 72ºC.

The Reverse Transcriptase PCR (RT-PCR) reaction for rotavirus and coxsackievirus commenced with incubation at 50ºC for 30 minutes to allow transcription of the viral RNA to cDNA. The PCR reaction steps had an initial incubation at 95ºC for 8 min followed by 55 cycles of 95ºC for 30 sec, 58ºC for 20 sec and 72ºC for 20 sec (with a 2 second increase of the extension step in each cycle after the 25th cycle). The final cycle had an extension time of 5 min at 72ºC.

All (RT-)PCR reactions were tested using a melt curve analysis after the PCR reaction to differentiate between actual products and primer dimmers, and to eliminate the possibility of false-positive results. The melt curve was generated using 80 cycles of 10 seconds each starting at 55ºC and increasing in 0.5ºC intervals to a final temperature of 95ºC. The Tm for each amplicon was determined by using iQ5 software (Bio-Rad).

Mean viral copy numbers were calculated from standard curves generated during (RT-)PCR by the iQ5 software. The standard curve for each virus was constructed from 10-fold serial dilutions of extracted RNA or DNA from the original washed virus stock. Aliquots of the same set of RNA or DNA standards were used for all experiments to reduce errors in quantification. The viral quantitative numbers were presented as copy numbers from each (RT-)PCR reaction.

Determination of pathogen decay rates

The change in the number of each pathogen in each replicate was plotted over time on a graph and a regression line fitted to the plot. The decay rate () was determined from the slope of the regression line. This resulted in three replicate values for each of the


microorganisms tested. These three replicate values were used to obtain a mean value and associated standard error. If the standard error was considered large, the replicate regression analyses were re-examined and if necessary, extra analyses of the results were done.

The results were reported as a T90 time (time for a 90% or 1 log loss of each pathogen) which was determined as 1/ .

References

Abbaszadegan, M. & DeLeon, R. 1997, ‘Detection of viruses in water samples by nucleic acid amplification’, in: Environmental applications of nucleic acid amplification techniques, ed. G.A. Toranzos, Lancaster, Technomic Publishing Co., pp. 113–137.

Campbell, A.T., Robertson, L.J. ,& Smith, H.V. 1992, ‘Viability of Cyrptosporidium parvum Oocysts: Correlation of In Vitro Excystation with Inclusion or Exclusion of Fluorogenic Vital Dyes’, Applied and Environmental Microbiology, 58: 3488–3493.

Gobet, P. & Toze, S. 2001, ‘Relevance of Cryptosporidium parvum hsp 70 mRNA amplification as a tool to discriminate between viable and dead oocysts’, Journal of Parasitology, 87: 226–229.

Gordon, C. & Toze, S. 2003, ‘Influence of groundwater characteristics on the survival of enteric viruses’, Journal of Applied Microbiology, 95(3): 536–544.

Gouvea, V., Glass, R.I., Woods, P., Taniguchi, K., Clark, H.F., Forrester, B. & Fang, Z.Y. 1990, ‘Polymerase chain reaction amplification and typing of rotavirus nucleic acid from stool samples’, Applied and Environmental Microbiology, 28(2): 276–282.

Heim, A., Ebnet, C., Harste, G. & Pring-Akerblom, P. 2003, ‘Rapid and quantitative detection of human adenovirus DNA by real-time PCR’, Journal of Medical Virology, 70: 228–239.


APPENDIX 1H – Decay of Enteric Microorganisms in Diffusion Chambers at the Floreat Infiltration Gallery Site

Figure 1H-1. Decay of E. coli in Floreat Infiltration Gallery site

Figure 1H-2. Decay of E. faecalis in the Floreat Infiltration Gallery site

Figure 1H-3. Decay of S. enterica in the Floreat Infiltration Gallery site

0 2 4 6 8 100

1

2

3

4

5

6

7

8

9

Log

cfu

mL-1

Time (days)

0 2 4 6 8 100

1

2

3

4

5

6

7

8

9

Log

cfu

mL-1

Time (days)

0 2 4 6 8 100

1

2

3

4

5

6

7

8

Log

cfu

mL-1

Time (days)


Figure 1H-4. Decay of Cryptosporidium parvum oocysts in the Floreat Infiltration Gallery site

Figure 1H-5. Decay of rotavirus in the Floreat Infiltration Gallery site

Figure 1H-6. Decay of adenovirus in the Floreat Infiltration Gallery site

0 5 10 15 20 25 30 35 40 450

1

2

3

4

5

Log 1

0 o

ocys

ts m

L-1

Time (days)

0 5 10 15 20 25 30 35 40 450.0

0.5

1.0

1.5

2.0

2.5

3.0

Log

10 P

DU

Time (days)

0 5 10 15 20 25 30 35 400

1

2

3

4

5

6

7

Log

10 P

DU

Time (days)


Figure 1H-7. Decay of coxsackievirus in the Floreat Infiltration Gallery site

Figure 1H-8. Decay of the bacteriophage MS2 in the Floreat Infiltration Gallery site

0 5 10 15 20 25 30 350

1

2

3

4

5

6

Log

PD

U

Time (days)

0 10 20 30 40 50 600

1

2

3

4

5

6

7

8

Log

pfu

mL-1

Time (days)


APPENDIX 1I – Halls Head Water Level and Water Chemistry Investigation

Groundwater flow directions at Halls Head WWTP

The concept of defining groundwater flow directions in the limestone at Halls Head was difficult due to its high transmissivity. Figure 1J-1 shows a schematic model of the capture zones for two bores SPB1 and SPB2, which each pump at roughly 217 KL/day to recover water infiltrating below the ponds (Toze et al., 2002). There is presumably a water table mound below the ponds that drives flow radially outward and towards the galleries; the shape of the mound and flow directions were not confirmed.

Indian Ocean

SPB2

SPB1

Infiltration galleries site

Infiltration ponds

North

6/88

2/84

7/881/84

1/83 (approx. location)

Figure 1J-1. Arial view of the infiltration galleries site relative to the ponds and the position of monitoring bores and the capture zones for recovery bores SPB1 and SPB2 (modified after Toze et al., 2002).

Groundwater level logging in bore 2/84 and at the Halls Head infiltration galleries (HHIG) site provided a record of hydrological changes upon which to evaluate potential influences on water quality (Figure 1J-2). Groundwater levels in the two bores were variable and offset by different amounts due to recharge from the galleries, drawdown from the recovery bore, and the impact of tidal fluctuations on groundwater levels.


During regular operation, the infiltration galleries received a daily supply of treated effluent followed by a nightly shutoff period. Groundwater levels in bore 2/84 (away from the recharge site) mainly responded to tidal fluctuations, whereas groundwater levels recorded beneath the galleries in HH_E2 were up to 10 cm higher and responded to both tidal fluctuations and the daily pulse of treated effluent as shown in Figure 1J-2 (inset 1). In contrast, during a week-long period when the supply pump failed to restart and the galleries did not receive treated effluent (Figure 1J-2, inset 2), groundwater levels measured in HH_E2 declined below the levels in bore 2/84 and tidal fluctuations were the dominant control on water levels.

The groundwater flow directions between bores HH_E2 and 2/84 could not be determined without more water level data from surrounding points in the aquifer, but it was apparent that the operation of HHIG produced a local, daily increase in the water table beneath the infiltration galleries. The magnitude of water level rise varied as a function of seasonal trends, tides and the maintenance of steady effluent recharge to the site.


0.9

1.0

1.1

1.2

1.3

1.4

1.5

5-S

ep-05

5-O

ct-05

4-N

ov-05

4-D

ec-05

3-Jan

-06

2-F

eb-0

6

4-M

ar-06

3-A

pr-06

3-M

ay-06

WA

TE

R L

EV

EL

(m

AH

D)

Bore 2/84

Bore HH_E2

Inset 1 Inset 2

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

4-Nov-05

5-Nov-05

6-Nov-05

7-Nov-05

8-Nov-05

9-Nov-05

WA

TE

R L

EV

EL

(m

AH

D)

4650

4700

4750

4800

4850

4900

4950

5000

5050

Cu

mu

lati

ve E

fflu

ent

Infl

ow

(K

ilo

litr

es)

Bore 2/84

Bore HH_E2

Total cumulative inflowto both galleries

Inset 1

Period of effluent inflow

Nightly shutoff

Nightly shutoff



Nightly shutoff followed by failure of the supply pump to automatically restart


1.0

1.1

1.1

1.2

1.2

1.3

1.3

1.4

1.4

1-De

c-05

2-De

c-05

3-De

c-05

4-De

c-05

5-De

c-05

6-De

c-05

WA

TE

R L

EV

EL

(m

AH

D)

Bore 2/84

Bore HH_E2

Inset 2

Figure 1J-2. Groundwater levels measured in bores HH_E2 and 2/84. Inset 1 indicates the cumulative volume of treated effluent and the groundwater level responses in the two bores in early November. Inset 2 shows the groundwater levels in early December 2005 during a week-long break in effluent inflow to the galleries. The water level probe for HH_E2 malfunctioned January–March 2006 (no data).


Inferred pre-existing groundwater chemistry at the experimental site

There were essentially two major components to the ‘background’ groundwater that was sampled at or near the Halls Head WWTP for this project. There was (1) characteristic groundwater originating from seawater and modified by geochemical reactions with the marine-deposited sediments and (2) the overprint of recharge from the wastewater ponds. The pumping action of the recovery bores and tidal fluctuations mix water from different sources, including effluent recharge and groundwater external to the site in different proportions. The initial groundwater chemistry below the infiltration galleries was inferred based on previous data to obtain a reference for describing the changes that have occurred since the galleries started operating. Additional water quality data were obtained from monitoring conducted by the Water Corporation at the WWTP and from water sampling conducted in previous years by Toze et al. (2002).

The results from inferring the pre-existing groundwater composition at the Halls Head WWTP site are depicted in Figure 1J-3, which compares the concentrations of Ca and Na in water sampled from the wastewater ponds and bores 6/88, SPB1, and SPB2 in the years 2000–2002. These data show a large difference in the chemistry of groundwater from bore 6/88 (outside the WWTP operations) and samples from the wastewater ponds. Nevertheless, groundwater from bore 6/88 may not represent the baseline conditions at the site of the infiltration galleries. The data for the SPB1 and SPB2 in Figure 1J-3 are more likely to represent the groundwater chemistry beneath the HHIG site, which unfortunately was not sampled at the start of the project. A simple mixing calculation depicted in this figure indicates that the groundwater from SPB1 and SPB2 in 2000–2002 was predominantly derived from the wastewater ponds (70–80%), similar to the results based on TDS in Toze et al. (2002).

Figure 1J-3. The relative concentrations of Ca:Na in water samples from the Halls Head WWTP to show pre-existing conditions and recharge water. End-member mixing between the average concentrations of water sampled from the wastewater ponds and the one sample from bore 6/88 reveals that the groundwater from SPB1 and SPB2 in 2000–2002 was predominantly derived from the wastewater ponds.


Changes in groundwater chemistry

The extensive set of water sampling conducted during the first year of this project allowed us to evaluate the changes in groundwater quality in relation to the operation of the infiltration galleries. The origin of the groundwater sampled from the Halls Head bores was first examined and the presence of a chemical signature unrelated to the process of managed aquifer recharge was detected and identified as residual bentonite from the drilling of the boreholes. The results from this interpretation are shown in a series of plots of major ion ratios (Figure 1J-4). A line that defines the dilution trend for seawater where the ratio of ions changes in a predictable pattern is included in the plots of Na, Mg, Ca, SO4 versus chloride concentrations. Groundwater sampled from bore 6/88 was used as a background reference for the water chemistry in the aquifer.

The plot of Na versus Cl in Figure 1J-4a reinforces the point that groundwater samples from the two recovery bores and bore 2/84 was predominantly derived from wastewater pond recharge because of their similar major ion compositions. These water samples also plot along the dilution trend for seawater. As there are water samples from the wastewater pond and the infiltration galleries that have Na concentrations greater than dilute seawater (refer to Figure 1J-4a), this indicates that the effluent recharge water was a source of some of the Na.

In contrast, there was elevated concentrations of Mg, Ca and SO4 that were not sourced from the effluent recharge water, which appear in the groundwater samples from the monitoring bores surrounding the infiltration galleries as shown in Figures 1J-4b and c. The most likely explanation for the elevated concentrations of Mg and Ca in the groundwater from the Halls Head bores is due to reactions associated with contamination by residual drilling mud composed of bentonite (Na0.5Al1.5Mg0.5Si4O10(OH)2). The bentonite released Ca and Mg from the sediments by cation exchange with Na. The highest concentrations of Mg and Ca were detected during initial sampling from these bores and groundwater samples from these bores are progressively becoming more dilute with respect to these ions.

The elevated concentrations of SO4 in groundwater from the Halls Head bores was likely due to mineral reactions in the aquifer that were not directly related to the effluent recharge waters. As indicated in Figure 1J-4d, the initial SO4 concentrations sampled from the Halls Head bores were comparable to the background bore 6/88. The initial groundwater samples were not analysed for iron and mineral analyses were not conducted on the sediments. It is quite plausible that sulphide minerals were in the marine-derived sediments. Oxidation of these minerals would explain the SO4 concentrations. The results shown in Figure 1J-5 reveal a correspondence of high SO4 concentrations with low-oxidizing groundwater sampled initially from the Halls Head bores and bore 6/88; the impact of recharge with more oxygen-rich water from the infiltration galleries was to increase the oxidizing potential in the aquifer (see ORP measurements in Figure 1J-5) and decrease the groundwater SO4.


Figure 1J-4. Plots (a–d) compare different major ion analyses for the water samples from Halls Head. The groundwater sampled from bore 6/88 in 2002 represents the background groundwater outside the operations at Halls Head WWTP. The water samples taken initially from four of Halls Head bores (one month after drilling) have anomalously high concentrations of Ca and SO4 relative to Mg.

107

Figure 1J-5. The concentration of SO4 versus measurements of oxidation-reduction potential taken in the field during water sampling

108

APPENDIX 1J – Bromide and Uranine Tracer Tests at the Floreat Infiltration Galleries: Diploma thesis by Wolfgang Fegg

Fachhochschule Weihenstephan Abteilung Triesdorf

Fachbereich Umweltsicherung

Diploma thesis

Characteristics of the unsaturated soil under an Atlantis

leach system for managing aquifer recharge

Submitted by: Wolfgang Fegg

Reader: Prof. Dr. Wilhelm Pyka

Thesis advisers : Dr. Simon Toze and Dr. Elise Bekele

Day of release: 15.04.2008

109

Contents 1 Abstract..............................................................................................................110 2 Introduction ............................................................Error! Bookmark not defined. 3 Material and Methods ............................................Error! Bookmark not defined.

3.1 Site description............................................................... Error! Bookmark not defined. 3.2 Soil profile....................................................................... Error! Bookmark not defined.

3.2.1 Saturated hydraulic conductivity ...........................Error! Bookmark not defined. 3.2.2 Determine bulk density, porosity, volumetric- and gravimetric water contentError! Bookmark

3.3 Calibration of the heat dissipation matrix potential sensors TM229 .....Error! Bookmark not defined.

3.4 Field installation.............................................................. Error! Bookmark not defined. 3.4.1 Infiltration System..................................................Error! Bookmark not defined. 3.4.2 Sampling system...................................................Error! Bookmark not defined. 3.4.3 Heat dissipation matric potential sensors TM229 and hydrology neutron

probe measurement system..................................Error! Bookmark not defined. 3.4.4 Electrical conductivity tracer and wastewater test.Error! Bookmark not defined.

3.5 Field release experiment tracer tests with deionised water, with uranine and bromide....................................................................................... Error! Bookmark not defined.

4 Monitoring and analysis .........................................Error! Bookmark not defined. 4.1 Preliminary chemical test with Uranine contact with tap-, deionised- groundwater and

different soil samples...................................................... Error! Bookmark not defined. 4.2 Tracer tests with deionised water ................................... Error! Bookmark not defined. 4.3 Tracer test with uranine and bromide............................. Error! Bookmark not defined.

4.3.1 Mass balance tracer test uranine ..........................Error! Bookmark not defined. 4.3.2 Numerical integration over composite Trapezoidal rule by Burden et.al.

(1993), and estimate the mean pore velocity ........Error! Bookmark not defined. 4.3.3 Cpeak method (Maloszewski et. al., 1985; Schudel et. al., 2002)Error! Bookmark not defined.4.3.4 Modelling program CXTFIT, Toride et. al. (1995) .Error! Bookmark not defined. 4.3.5 Flow velocity for different soil types ......................Error! Bookmark not defined. 4.3.6 Comparison of average travel time, results tracer test, with mathematical

calculation for the travel time in saturated conditionsError! Bookmark not defined. 4.4 Soil moisture measurement with TM229 and Neutron meter (NM5) ....Error! Bookmark

not defined. 4.5 Electrical conductivity tracer- and wastewater test......... Error! Bookmark not defined.

5 Results...................................................................Error! Bookmark not defined. 5.1 Soil profile....................................................................... Error! Bookmark not defined. 5.2 Preliminary chemical test with uranine in contact with tap-, deionised-, groundwater and

different soil samples...................................................... Error! Bookmark not defined. 5.3 Tracer tests with deionised water ................................... Error! Bookmark not defined. 5.4 Tracer test with uranine and bromide............................. Error! Bookmark not defined.

5.4.1 Tracer breakthrough curves, mass balance and further measurement resultsError! Bookmark5.4.2 Flow time, corresponding flow velocities and mean pore velocity in different

soil types for uranine .............................................Error! Bookmark not defined. 5.4.3 Bromide results .....................................................Error! Bookmark not defined. 5.4.4 Dispersion coefficient, cpeak method compared with modelling program

CXTFIT..................................................................Error! Bookmark not defined. 5.4.5 Comparison of average travel time, results tracer test uranine, with

mathematical calculation for the travel time in saturated conditionsError! Bookmark not defin5.5 Soil moisture measurement with TM229 and Neutron meter ........ Error! Bookmark not

defined. 5.6 Electrical conductivity tracer- and wastewater test......... Error! Bookmark not defined.

110

6 Discussion .............................................................Error! Bookmark not defined. 6.1 Soil profile....................................................................... Error! Bookmark not defined. 6.2 Preliminary chemical test with uranine in contact with tap-, deionised-, groundwater and

different soil samples...................................................... Error! Bookmark not defined. 6.3 Tracer tests with deionised water and EC test results.... Error! Bookmark not defined. 6.4 Tracer test with uranine and bromide............................. Error! Bookmark not defined.

6.4.1 Tracer breakthrough curve and mass balance......Error! Bookmark not defined. 6.4.2 Flow time, corresponding flow velocities and mean pore velocity in different

soil types for uranine .............................................Error! Bookmark not defined. 6.4.3 Bromide results .....................................................Error! Bookmark not defined. 6.4.4 Dispersion coefficient, Cpeak method compared with modelling program

CXTFIT..................................................................Error! Bookmark not defined. 6.4.5 Comparison of average travel time (ta), with mathematical calculation for the

travel time in saturated conditions.........................Error! Bookmark not defined. 6.5 Soil moisture measurement with TM229 and Neutron meter ........ Error! Bookmark not

defined. 7 Conclusions ...........................................................Error! Bookmark not defined. 8 References ............................................................Error! Bookmark not defined. 9 Appendix................................................................Error! Bookmark not defined.

9.1 Soil results...................................................................... Error! Bookmark not defined. 9.2 Chemical test.................................................................. Error! Bookmark not defined. 9.3 Tracer test results........................................................... Error! Bookmark not defined. 9.4 Results CXTFIT for station SL1 suction cup Ab ............. Error! Bookmark not defined. 9.5 Measurement result volume water content..................... Error! Bookmark not defined.

2 Abstract Managing aquifer recharge (MAR) at the CSIRO Floreat Infiltration Galleries site in Western Australia is a research facility that is being used to infiltrate treated wastewater through a subsurface gallery system in the unsaturated zone to supplement the supply of non-potable water. The aim of this study was to characterize the unsaturated zone, to determine the level of saturation beneath the MAR site, and to estimate the travel time through the unsaturated profile. Data collection included soil layer descriptions, grain-size analyses to estimate different hydraulic properties, measurements of soil moisture and different types of tracer tests.

Sediment cores were first collected to create a soil profile beneath the infiltration gallery to the groundwater table. The information about the different soil layers was used to install suction cups and thermal matrix potential sensors. Five TM229 sensors were calibrated with different soil samples from the core data. The TM229 data made it possible to calculate the volumetric water content directly under the gallery system. These data were compared with results from neutron moisture meter measurements next to the infiltration system. A separate experiment was conducted to monitor changes in the moisture content below the gallery system. The inflow of treated wastewater in the gallery system was turned off for eleven days and changes in the volumetric water content were logged. There was a non-uniform pattern of drying of the soil below the gallery: the soil started to dry out from the base of the gallery system and from the side of the infiltration rim.

111

The main focus of this project was to estimate the travel time through the unsaturated zone. The tracers included deionised water, uranine and bromide. A specially designed sampling system was developed for this study. Four stations (SL1 – SL4) with nine suction cup lysimeters were installed at different depth to collect one sample every hour from every suction cup. Preliminary tests using deionised water revealed that deionised water can be used as a tracer, but only for short distances as there is not a significant contrast in electrical conductivity to use it as a tracer. Uranine were selected as a tracer because it is relatively inexpensive and easy to measure using a handheld fluorometer. Bromide is more expensive to analyse, but it provides a comparison for evaluating breakthrough times and sorbing potential. Prior to conducting the tracer tests in the field, lab measurements were made using uranine in contact with different soil and water samples from the test area to evaluate the extent to which uranine would sorb to different soil samples. Groundwater from the test area was found to be a suitable carrier substance for uranine and bromide.

During the main tracer experiment, 6800 litres of groundwater mixed with uranine and bromide were infiltrated under steady state conditions at a controlled inflow rate. Water samples were collected from the suction cup lysimeters over a time period of 330 hours. The tracer breakthrough results were analysed to determine travel times, corresponding flow velocities, flow velocities in different soil layers, dispersion coefficient and preferential flow areas. The average travel time for recharge water to reach the water table is 104 hours (approximately 4 days) using a graphical method called the Cpeak method (Maloszewski et. al., 1985, Schudel et. al., 2002). The first appearance of the maximum concentration was after 70 hours (approximately 3 days). The results reveal potential layers of water storage areas that have a higher level of saturation. Comparison of the lag in the breakthrough of uranine relative to bromide reveals a difference of 16% that may be useful when applying uranine in future tracer tests.

design and operation of infiltration galleries and water ... · chapter 1 - design and operation of...

Documents