Explanatory notes on water chemistry results 2013-14

2013-14 Wet season notes

Commentary on Magela Creek 2013-14 monitoring data

The Supervising Scientist Division (SSD) modified its wet season monitoring programme from 2010-11 to enhance the ability of SSD to independently detect changes in water quality during EC- and turbidity- triggered events with automatic sampling during events; continuous water quality monitoring of pH, EC, turbidity and water temperature; and in situ toxicity monitoring programmes.

Grab samples are taken every two weeks from Magela Creek for radium analysis and every four weeks water samples are taken and measured for key mine site analytes, including physicochemical parameters, for quality assurance purposes. The sampling point maps show the location of the upstream and downstream sites and key Ranger Mine features. Previous weekly grab sample monitoring data can be found at Magela Creek grab sample monitoring data 2002 – 2010 and previous continuous monitoring at Magela Creek monitoring data 2010-2012 and Magela Creek monitoring data 2012-2013.

Flow was first recorded at the Magela Creek upstream and downstream monitoring stations on 28 November 2013, which is also when the water quality sensors at the downstream site were submerged. Only the turbidity sensors at the upstream site were submerged on 28 November 2013, with the EC and pH sensors, which are located above the turbidity sensors, becoming submerged on 2 December 2013.

On 7 December 2013 there was a leach tank failure on the Ranger mine site. In response to this SSD increased the monitoring frequency within Magela Creek as part of an investigation into the incident to confirm that there has been no impact on the external environment or human health, and that the surrounding environment, including Kakadu National Park, remains protected. Results received to date show that concentrations of uranium, manganese and magnesium have remained significantly below the ecotoxicologically derived water quality objectives.

On 22 December 2013 an isolated heavy rainfall event occurred in the Georgetown Creek catchment with 64 mm recorded by ERA at GC2 compared to only 22 mm  at the Jabiru Airport. This resulted in an influx of surface water down the Georgetown system, with an EC of around 130 µS/cm at ERA's monitoring site GC2 and 94 µS/cm in Georgetown Billabong. The flow in Magela Creek was less than 5 cumecs and the pulse of surface water from Georgetown Billabong was observed at the downstream site where the EC peaked at 31 µS/cm. The corresponding EC measured at the upstream site during this period was 16 µS/cm.  During this EC event an automatic sample was triggered at MCDW, which had elevated magnesium (1.7 mg/L) and sulfate (3 mg/L) concentrations, the event remained well below the chronic exposure limit for magnesium of 3 mg/L. Uranium concentration was low at 0.19 µg/L, which is approximately 3% of the ecotoxicologically derived limit of 6 µg/L.

On 29 December 2013 elevated EC was observed at the downstream site, with EC peaking at 36 µS/cm with a corresponding upstream EC of 12 µS/cm (Figure 1). This increase in EC downstream of the mine was also observed by ERA, who have continuous monitoring sites located in all three channels of Magela Creek. The fact that the EC signal occurred prior to managed release of retention pond 1 (RP1) waters and it was apparent across all three channels downstream of the mine (indicating thorough mixing), suggests that the likely source is surface water from Georgetown Billabong which at the time, had an EC of around 95 µS/cm. Samples were manually triggered over the Christmas period on 29 December 2013, as shown in Figure 1. The upstream sample was triggered at 03:33 am and the downstream sample was triggered at 09:15 am. The magnesium and sulfate concentrations were higher in the downstream sample as would be expected with the higher EC at the time. The uranium concentration downstream was marginally higher at 0.28 µg/L however this is still low at approximately 5% of the ecotoxicologically derived limit.

Figure 1  Magela Creek continuous monitoring data showing peaks in electrical conductivity at the downstream monitoring site (MCDW) during late January 2014

Figure 1 Magela Creek continuous monitoring data showing peaks in electrical conductivity at the downstream monitoring site (MCDW) during late January 2014

Since late December 2013 the EC downstream of the mine has diverged from the background EC, measured upstream of the mine. Each wet season once flows from Coonjimba and Georgetown billabongs into Magela Creek become established the downstream EC diverges from background due to the input of higher EC waters.

On 27 January 2014 there was an EC peak of 34 µS/cm at the downstream monitoring site(Figure 2). The peak occurred during a decrease in Magela Creek water level and is likely to be due to Coonjimba Billabong inflow to Magela Creek. At the time, the EC of Coonjimba Billabong was around 150 µS/cm and the EC of RP1 was around 270 µS/cm (ERA monitoring data). As the EC peaks were below the autosampler activation trigger no samples were collected.

Figure 2  Magela Creek continuous monitoring data showing peaks in electrical conductivity at the downstream monitoring site (MCDW) during late January 2014.

Figure 2 Magela Creek continuous monitoring data showing peaks in electrical conductivity at the downstream monitoring site (MCDW) during late January 2014.

High turbidity levels are common in Magela Creek during the early wet season due to first flush effects. Historically, the mine site does not contribute substantial amounts of suspended sediment to Magela Creek and therefore does not influence the turbidity at the downstream site. The majority of turbidity events recorded at the downstream site are also observed upstream of the mine. Occasionally, under specific meteorological conditions, turbidity spikes are observed at the downstream site without an accompanying spike at the upstream site, indicating that the source of suspended sediment lies between the two sites. From 23-28 January 2014 there were three turbidity peaks which occurred at MCDW but were not detected upstream at MCUGT (Figure 3). These coincided with intermittent rainfall events associated with the monsoonal trough over northern Australia during late January and are due to surface run-off from localised rainfall. Water quality objectives are currently being developed for continuously monitored turbidity.

Figure 3  Magela Creek continuous monitoring data showing peaks in turbidity at the downstream monitoring site (MCDW) during late January 2014.

Figure 3 Magela Creek continuous monitoring data showing peaks in turbidity at the downstream monitoring site (MCDW) during late January 2014.

During mid February the water discharge decreased leading to gradually rising EC and pH levels due to the decrease in rainfall levels which typically have low EC and mildly acidic pH. Continuous monitoring data showed pH levels within the creek tend to be at their lowest following heavy rainfall events due to dilution from slightly acidic incident rainfall (Noller et al 1990). pH levels typically increase during periods of low/recessional flow due in part to the effects of evapoconcentration, local groundwater influx and plant photosynthesis. Water quality objectives are currently being developed for continuously monitored pH.

Uranium concentrations recorded within Magela Creek have all been below the Focus trigger value of 0.3 µg/L. The maximum uranium concentration was 0.28 µg/L recorded at MCDW on 29 December 2013 at 09:15 hours. This is approximately 5% of the ecotoxicologically derived limit.

Manganese concentrations to date have also been very low. Only one sample at MCDW was recorded above the Focus trigger value of 35 µg/L with a concentration of 44 µg/L on 28 November 2013. The manganese trigger values only apply when creek flow is above 5 cumecs. At the time of sample collection on 28 November 2013 creek flow was <0.1 cumecs and thus the trigger values do not apply in this case.

Magnesium and sulfate concentrations measured have been low. Automatic samples have not been triggered for any EC peaks during the 2013–14 wet season to date as the EC has not exceed the 42 μS/cm guideline (corresponding to 3 mg/L magnesium). Only one sample was observed with a magnesium concentration above the Focus trigger value of 1 mg/L. This occurred on 23 December 2013 at 01:40 hours with a magnesium concentration 1.7 mg/L. The corresponding sulfate concentration was 3 mg/L.

In December 2013 ecotoxicologically derived water quality objectives for total ammonia nitrogen (TAN) were formally adopted for Magela Creek as part of the regulatory approval for brine concentrator distillate release. Results received to date have shown that levels of TAN in Magela Creek are very low with a highest recorded concentration of 0.016 mg/L at MCDW on 12 December 2013, which is approximately 2% of the 0.7 mg/L Guideline trigger value.

Continuous monitoring will continue throughout the season until cease to flow is agreed by stakeholders or until the multi-probes are out of the water and cannot be lowered any further regardless of flow between upstream and downstream. During recessional flow conditions data will be updated on a monthly basis unless there are water quality events of interest to report.

Commentary on Gulungul Creek 2013-14 monitoring data

The Supervising Scientist Division (SSD) modified its wet season monitoring programme from 2010-11 to enhance the ability of SSD to independently detect changes in water quality during EC- and turbidity- triggered events with automatic sampling during events; continuous water quality monitoring of pH, EC, turbidity and water temperature; and in situ toxicity monitoring programmes.

Water samples are taken every four weeks and measured for key mine site analytes, including physicochemical parameters, for quality assurance purposes. Previous weekly grab sample monitoring data can be found at Gulungul Creek grab sample monitoring data 2002 – 2010 and previous continuous monitoring at Gulungul Creek monitoring data 2010-2012 and Gulungul Creek monitoring data 2012-2013

Flow was first recorded at the Gulungul Creek upstream monitoring site on 28 November 2013. Flow was not recorded at the downstream monitoring site until 4 December 2013. Flow remained very low during the early wet season due to relatively low rainfall. The multi-probes at GCDS were only partially submerged in mid-December, which resulted in gaps in the continuous EC data as this sensor, located above the turbidity and water level sensors, was not submerged.

During 21-23 December 2013 Jabiru Airport received 22.4 mm of rain. This resulted in increased surface run-off with EC and turbidity peaks recorded at GCUS and GCDS. This is typical of first flush effects early in the wet season. This also resulted in the submergence of EC sensors at GCDS. The EC peaked at 46.5 µS/cm at GCUS and 34.8 µS/cm at GCDS. Turbidity fluctuated reaching 11.3 NTU at GCUS and 6.1 NTU at GCDS. A number of automatic samples were triggered at the upstream  site due to this EC peak. Results show corresponding peaks in magnesium and  manganese concentrations but little change in uranium and sulfate  concentrations. In distinction from the dominant mine site signature of magnesium  sulfate, the major ions (anions and cations) of the upstream Gulungul catchment  are dominated by magnesium hydrogen carbonate [or magnesium bicarbonate - Mg(HCO3)2] hence the lack  of change in the sulfate concentrations during this event indicates a natural catchment influence rather than a mine site input.

On 19 January 2014 from 00:30 to 03:20 am ERA continuous monitoring at GCLB detected an EC peak. The magnitude of this peak was 184.2 µS/cm, which occurred at 02:00 am. The ecotoxicologically-derived framework for EC pulses provides an EC limit of 1040 µS/cm for a 3 hour EC peak duration therefore with an EC peak of 184.2 µS/cm the environment is unlikely to have been impacted.

The SSD continuous monitoring station GCDS is located approximately 1 km downstream (figure 4) and detected an EC peak of 41.1 µS/cm at 10:00 am on 18 January 2014 (figure 5), which was also observed at ERA's monitoring station GCLB. The peak of 184.2 µS/cm observed at GCLB on 19 January 2014 was not detected at SSD's monitoring station GCDS. As SSD's monitoring station is located on the western bank of Gulungul Creek the solutes measured upstream at GCLB may have flowed downstream close to the eastern bank and thus may not have been detected by the sondes at GCDS. This may have been further exacerbated by incoming flows from the left bank tributary, which enter Gulungul Creek just upstream of GCDS. The EC triggered samples from the 18 January 2014 EC peak show a maximum uranium concentration of 0.22 µg/L, which is approximately 4% of the ecotoxicologically derived limit. Magnesium and sulfate concentrations peak during this EC event.

Figure 4  Gulungul Creek continuous monitoring stations

Figure 4 Gulungul Creek continuous monitoring stations

Further EC peaks > 30 µS/cm have been observed at GCDS from 20-28 January 2014 (figure 5) and 12-14 February 2014 (figure 6). SSD is undertaking further investigation of the likely source of the solutes involved in the EC peaks within the downstream Gulungul Creek catchment. SSD  will also undertake an investigation into lateral mixing of solutes within the channel at GCDS and has installed an additional sonde on the eastern bank opposite the GCDS monitoring station (GCDS East Bank) on 23 January 2014.

The results from this east bank multi-probe show that for the most part the Gulungul Creek water is well mixed with comparable EC measured from both banks of creeks. However during some of the EC peaks the east bank multi-probe has measured EC up to 15 µS/cm greater than the west bank EC. This indicates a source of higher solute water following the east bank of the creek.

Figure 5 Gulungul Creek continuous monitoring data showing peaks in electrical conductivity at the GCDS and GCDS East Bank during mid-late January 2014.

Figure 5 Gulungul Creek continuous monitoring data showing peaks in electrical conductivity at the GCDS and GCDS East Bank during mid-late January 2014.

Figure 6 Gulungul Creek continuous monitoring data showing peaks in electrical conductivity at the GCDS and GCDS East Bank during early February 2014.

Figure 6 Gulungul Creek continuous monitoring data showing peaks in electrical conductivity at the GCDS and GCDS East Bank during early February 2014.

On Monday 3 March 2014 SSD staff conducted a field visit to the Gulungul Creek area to the west of the Ranger TSF in order to investigate the source of recent EC events noted at the ERA downstream site.

EC readings were taken at a number of locations along the Radon Springs track with background water quality in the range of 15-25 µS/cm. At location 1 in Figure 7, being the intersection point of the GCT2 drainage line and the Radon Springs track, EC readings in the range of 1250-1300 µS/cm were recorded in a small flowing stream. This stream was tracked to its confluence with Gulungul Creek (location 2, Figure 7) and an EC reading of approximately 500 µS/cm was obtained at that location prior to entering Gulungul Ck. EC readings were obtained in Gulungul Creek but due to the low volume of in-flow the signature diluted out within a short distance. SSD has obtained water samples and deployed portable EC loggers at locations 1 and 2.

Figure 7 Location of SSD EC loggers installed on 3 March 2014 to investigate the water quality of GCT2

Figure 7 Location of SSD EC loggers installed on 3 March 2014 to investigate the water quality of GCT2

On 26 January 2014 a large peak in turbidity occurred at the upstream monitoring site GCUS in response to a rainfall event, with 68 mm recorded at GCUS, and subsequent rising creek flow. The magnitude of this turbidity peak at GCUS was 211 NTU occurring a t20:40 hours (Figure 8). This is the sixth time that a turbidity peak >150 NTU has been recorded at GCUS since continuous monitoring of turbidity began in the 2003-2004 wet season. At 02:20 hours on 27 January 2014 the turbidity from this event had reached the downstream monitoring site GCDS and had been substantially diluted reaching a much lower magnitude of 49 NTU. The source of the turbidity is upstream of Ranger and is therefore not mine derived.

Figure 8 Gulungul Creek continuous monitoring data showing a peak in turbidity at the upstream monitoring site (GCUS) on 26 January 2014.

Figure 8 Gulungul Creek continuous monitoring data showing a peak in turbidity at the upstream monitoring site (GCUS) on 26 January 2014.

Uranium concentrations recorded within Gulungul Creek have all been below the Focus trigger value of 0.3 µg/L. The maximum uranium concentration was 0.29 µg/L recorded at GCUS on 21 December 2013 at 20:25 hours. This is approximately 5% of the ecotoxicologically derived Limit.

Manganese concentrations have also been very low in Gulungul Creek with all samples recording concentrations below the Focus trigger value of 35 µg/L. The maximum manganese concentration was 12 µg/L recorded at GCDS on 12 December 2013 at 10:48 hours. This is 16% of the ecotoxicologically derived limit.

Magnesium concentrations measured during 2013–14 were generally low with the exception of automatic samples triggered for EC peaks at GCUS on 22 December 2013 and GCDS on 18 January 2014. The GCUS EC event had two samples with magnesium concentrations above the chronic (>72 hour) Limit trigger value of 3 mg/L. The maximum magnesium concentration was 3.3 mg/L at 06:45 hours. The corresponding sulfate concentration was <0.5 mg/L. The GCDS EC event had six samples with magnesium concentrations above the Action trigger value of 2 mg/L. The maximum magnesium concentration was 2.5 mg/L at 09:15 hours. The corresponding sulfate concentration was 7 mg/L.

During late February and March the water levels within Gulungul Creek have decreased leading to gradually increasing EC levels, fluctuating with each rainfall event.

Continuous monitoring will continue throughout the season until cease to flow is agreed by stakeholders or until the multi-probes are out of the water and cannot be lowered any further regardless of flow between upstream and downstream. During recessional flow conditions data will be updated on a monthly basis unless there are water quality incidents to report.

Commentary on Ngarradj (Swift Creek) 2013-14 monitoring data

Jabiluka has been in a long-term care and maintenance phase since late 2003 and poses a low risk to the environment. As a consequence of this low risk and the good data set acquired over the last seven years indicating the environment has been protected, the monitoring programme has been systematically scaled down. Since 2009-10, the Supervising Scientist Division has collected continuous monitoring data (EC and water level) from the downstream statutory compliance site only. Energy Resources of Australia (ERA) collect monthly grab samples from both the upstream and downstream site.

Multisonde performance checks are made every two weeks for quality assurance purposes. Previous grab sample monitoring data can be found at Ngarradj (Swift Creek) grab sample monitoring data 2001-2009 and previous continuous monitoring at Ngarradj (Swift Creek) monitoring data 2010-2012 and Ngarradj (Swift Creek) monitoring data 2012-2013.

Flow was first recorded at the Ngarradj (Swift Creek) monitoring station on 27 November 2013 and was very low at the start of the wet season with the multi-probes only submerged for short periods of time when the water level has been sufficiently high. With increased flow from 9 January 2014 the sondes have been fully submerged and EC has remained below 20 µS/cm.

On 26 January 2014 a turbidity event occurred during a 90mm rainfall event at the site (figure 9). This resulted in the collection of 7 turbidity automatic samples that will be analysed for suspended sediment. Trigger values for turbidity within Ngarradj (Swift Creek) were not included within the Water Quality Objectives as determined by the Jabiluka Minesite Technical Committee on 21 September 2001. However, baseline values for the physical and chemical characteristics of streams within the Jabiluka lease were established by Cusbert et al in 1998 including 'low risk trigger value' ranges for ecosystem protection. The trigger value range proposed for turbidity was 4.0-105, with non-compliance being turbidity events due to mining activity <4.0 NTU or >105 NTU. Thus the turbidity event of 26 January 2014 of 56.5 NTU falls in the middle of this 'acceptable' range. There have previously been 20 turbidity events > 50 NTU at Ngarradj since continuous monitoring of turbidity began in the 2003-2004 wet season.

Figure 8  Ngarradj (Swift Creek) continuous monitoring data showing a peak in turbidity on 26 January 2014

Figure 9 Ngarradj (Swift Creek) continuous monitoring data showing a peak in turbidity on 26 January 2014

During late February and March the water levels within Ngarradj have decreased leading to gradually increasing EC levels, fluctuating with each rainfall event.

Continuous monitoring will continue throughout the season until cease to flow is agreed by stakeholders or until the multi-probes are out of the water and cannot be lowered any further regardless of flow between upstream and downstream. During recessional flow conditions data will be updated on a monthly basis unless there are water quality incidents to report.