By Charlie Pierce, Research Director, Community
Environment Monitoring
Background
Transport for NSW (TfNSW) are required to monitor
groundwater, leachate and methane at the St Peters
Interchange Landfill to comply with an Environment
Protection Licence (4627) issued by NSW Environment
Protection Authority (EPA). The exchange was built on the site of an old landfill that was closed in 2014 to make way for the massive M8 interchange.
TfNSW has contracted WSP
environmental consultants to conduct the sampling and field
testing. WSP contracted SGS Australia Pty Ltd Environmental
Services – Sydney Laboratory to conduct the analytical
laboratory testing of ground water and leachate.
Current Situation
Community Environmental Monitoring Pty Ltd (CEM) has
previously reviewed the required monitoring data three times
(https://pollutionwatch.org.au/2024/10/westconnex-st-peters-
interchange-environmental-mon itoring-failures/ ,https://pollutionwatch.org.au/2024/02/report-st-peters-
interchange/). These reviews follow the publication of our special report on the St Peters Interchange that provided background on the history, contamination, failed government promises and regulation failures at the site.
These reports have been forwarded to
TfNSW and the EPA.
Each of these reviews have found that quality of the monitoring data was not fit for purpose. There were numerous
field sampling mistakes coupled with laboratory errors which
make the monitoring results from 2020 to 2024 unusable to
assess groundwater quality conditions.
CEM plans to continue reviewing the data presented by
TfNSW provided by its environmental consultant until there is
an indication that the data meets the standards set by the NSW
EPA.
TfNSW, the EPA and the site auditor have not responded to
the last review provided in February 2025.
March 2025 Review
Gas Monitoring
Methane monitoring results were incomplete. The person
sampling failed to record the carbon monoxide and hydrogen
sulphide for gas monitoring well LDS-GM-033. Gas
monitoring wells LDS-GM-021 and LDS-GM-023 could not
be located by the person sampling. Gas monitoring well LDS-
GM-022A was destroyed during Gateway Construction and
has not been replaced.
Five well measurements were reported to have gas
concentrations exceeding 100 % (summation of reported
gases) which is not possible. These wells were LDS-GM-028 (initial 112.1 %), LDS-GV-48 (initial 100.2% and purge at
102 %), LDS-GV-49 (100.6 %) and LDS-GV050 (113.2 %).
Another impossible measurement wat the purged oxygen
concentration reported for LDS-GV-49. It was reported to be
enriched at 22.5 %, which is higher than atmospheric oxygen
concentration at around 21 %.
Leachate Monitoring
The leachate monitoring results continue to be more stable
than most of the groundwater monitoring data.
TfNSW failed to report the standing water level in the latest
quarter again. The department has failed to report the
standing water level measurement in 14 of the 17 quarters that
were monitored by CEM.
The chloride value was reported to be 670 mg/L which is
lower than the lower control limit.
TfNSW reported total nitrogen. It is not required to report
total nitrogen in their EPL licence. However, if total nitrogen
is reported, it should be done correctly. Standard Methods for
the Analysis of Water and Wastewater (2023) defines total
nitrogen as the summation of ammonia nitrogen, organic
nitrogen and oxidised nitrogen (nitrate plus nitrite). The value
reported by TfNSW for total nitrogen was 0.015 mg/L which
is well below the ammonia nitrogen level reported at 190
mg/L.
Groundwater Monitoring
Well development
TfNSW’s environmental consultants have not yet properly
developed Groundwater monitoring well 3057 (B). It was
replaced before last quarter’s monitoring event after being out
of action for more than a year. Well development refers to the
process of cleaning out drilling fluids that inhibit water flow
within the well so that groundwater could freely flow. Free
flowing groundwater is necessary for the sample to be
representative of the aquifer being monitored. The water
chemistry in well 3057 (B) is still representative of concrete
and/or bentonite used during installation rather than the
chemistry associated with local groundwater. The high pH and
low salt content reflect well construction activities and not
groundwater conditions. As was the case last quarter, none of
the chemical water measurements in this quarter may be
considered to be reliable. The best course of action would be
for TfNSW to publicly declare these measurements are
worthless and order their consultants to properly develop this
well.
Quality Assurance Checks
It would be tedious to highlight the quality assurance errors
made during the December 2024 sampling quarter. However,
it is important to continue to document the lack of any quality
assurance for the groundwater monitoring program at the St
Peters Interchange (Landfill).
There are three common data quality checks most NATA
accredited laboratories perform when testing aqueous
samples. These include ionic balance, total dissolved solids (TDS)-conductivity ratios and measured TDS comparisons
with calculated TDS ratio.
The ionic balance is a check on the difference between
positive ions and negative ions in solution divided by the
mean of all ions. The acceptance criteria for this calculation
are less than 5%.
The relationship between total dissolved solids and
conductivity is that the ratio must be between 0.55 and 0.7.
The target percentage ratio between calculated total dissolved
solids and measured total dissolved solids must be between 75
and 125 %. These checks are found in Standard Methods for
the Analysis of Water and Wastewater (2023) and specified as
mandatory for licence reporting in NSW EPA’s Approved
Methods for the Sampling and Analysis of Water Pollutants in
NSW (2022).
This quarter’s review of these quality control failures focuses
on the total number of errors out of possible comparisons
rather than enumerating them individually. There were 15
failed quality checks out of 30 possible comparisons
measurements (50 %). This remains an abnormally high error
rate which is unusual for a NATA accredited laboratory.
Other Errors
The total organic carbon (9.9 mg/L), chloride (210 mg/L) and
alkalinity (770 mg/L) measurements for well 3059 (B)
sampled in September 2024 appear to be in error. All of these
parameters were reported to be much lower than the long-term
averages.
Well 3087 reported an alkalinity concentration of <5 mg/L.
This is not possible when the pH of a highly buffered water
sample was 7.09. The redox measurement at well 3087 was
reported to be -1525 mV which is also impossible. Highly
reducing anaerobic groundwater or leachate could have values
in the -300 mV range. It is incredulous that the environmental
staff at TfNSW or their consultants who are responsible for
reviewing these values did not recognise these gross mistakes
before reporting. The total organic carbon (87 mg/L) was
much lower than the historical range at this well. Ammonia
could not be reported at well 3087 due to an unexplained
laboratory error.
The reported value for chloride (590 mg/L) and redox
potential (-93.3 mV) were outside the long-term control levels
at well 3088.
At well 3090 the ammonia (4.6 mg/L), chloride (620 mg/L)
and pH (7.01 pH units) were too high when compared to long
term measurements.
The alkalinity (80 mg/L) and ammonia (0.35 mg/L) and
chloride (550 mg/L) for MW4D were too low when compared
to previous measurements made since 2020. Conversely, pH
(8.22 pH units) and total organic carbon 8.2 were elevated in
comparison to previous measurement at this well.
The alkalinity (29 mg/L) and chloride (12 mg/L) were low in
comparison to the control range for well BH157A. The total
organic carbon (18 mg/L) was elevated in comparison to
historical measurements at this well.
Reporting Significant Digits
Good laboratory practice requires that the number of
significant digits quoted in a result shall not imply a degree of
precision greater than that warranted by the sources of
uncertainty. The number of significant digits is meant to
reflect the sensitivity of a test result; consequently results on a
test report that contain more significant figures by inference
would be considered more accurate. Quoting to more
significant figures misrepresents the degree of calculated
precision and may lead to unfounded concern when guidelines
are exceeded at the level of the extra significant digit. As a
general guideline, the results obtained from the majority of
chemical tests should be reported to two significant figures;
however there are exceptions depending on the test equipment
and calibration procedures.
Field meters used for measuring parameters may be
characterised as robust but relatively insensitive measuring
devices. A field pH meter could not be sensitive to more than
2 significant digits. To report pH to 3 significant digits would
require a high-quality experimental meter with the
measurements being made in a controlled environment. That
is not the case when measurements are made in the field at a landfill. WSP have routinely reported pH to 3 or 4 significant
digits.
Likewise a field electrical conductivity meter is characterised
by limited precision. It is not possible for a field meter to
report to 4 significant digits. Reporting at this level would
infer that the meter could measure the difference between
1,214 µS/cm vs 1,216 µS/cm which is well beyond the
analytical precision of the field instruments used for these
measurements.
In most cases where CEM previously identified a rudimentary
error in reporting data to EPA and TfNSW, there was a
noticeable improvement in performance. For instance, there
hasn’t been an error in reporting bicarbonate concentrations
for more than a year after successive years of errors. Once we
reported that they were making a fundamental error, they
stopped making this error. However, the incorrect reporting of
too many digits is still happening after we reported this poor
practice last year.
Standing Water Levels
The standing water level measurement (m AHD) of an aquifer
usually moves in small increments based on seasonal or
changing ground water conditions. There are massive changes
in the standing water level at most wells being monitored at
the St Peters Interchange landfill, up to 40 metres different in
subsequent quarters. Below is a table showing the extent of
the variation between measurements.
Graphically, the change in water level over time is shown
below in one well.
This is a massive change in groundwater level between
subsequent quarters. During CEM discussions with TfNSW,
they conceded that some of the chemical analytical data
generated in the ground water monitoring program was below
industry standards. However, it was stated during the meeting
that this is not an issue since there is a negative hydraulic
gradient on site which keeps any contamination on site.
It is not apparent to me how anyone can determine the
groundwater flow direction based on any of the measurements provided by TfNSW in its public reporting. The height of the standing water level in each well is compared to the height in
nearby wells to establish flow direction. It seems that the flow
direction would be difficult to document with such variable
water levels.
Quality Assessment of December 2024 Monitoring
WSP are TfNSW environmental consultants responsible for
meeting the monitoring requirements found in Environmental
Protection Licence 4627. This review of the December 2024
sampling found that there were errors in methane monitoring,
leachate monitoring and groundwater monitoring. Errors were
found in 50 % of simple laboratory quality control checks
specified in EPA endorsed analytical method references.
One well still has not been developed following installation
even though TfNSW was advised of this failure in the last
quarter’s review. All reported measurements will not have any
validity with respect to a ground water monitoring program
until this well is properly developed. There were
measurements reported that are outside the realms of
possibility, such as alkalinity and redox potential. Despite this,
the people responsible for reviewing this data failed to
identify these errors before publicly reporting.
The measurement variability between samples at each well is
great. That makes it nearly impossible to find errors.
Conclusion
These facts taken together continue to indicate that there is
very little assessment on the quality of the data by WSP or
TfNSW scientific staff before it is reported. To address this
deficiency, it is recommended that the EPA audit the sampling
and split samples so that an accredited government laboratory
(EPA, Sydney Water, Water NSW or the National
Measurement Institute) could test these samples. This sort of
assessment will enable a more informed assessment on the
quality of the results from this monitoring.
The monitoring program is required by legislation and is paid
for using taxes. The public deserves to have the monitoring
performed following international best practice instead of the
current practice.