WATERWATCH VICTORIA
QA/QC WEEK 2014
REPORT ON STATE-WIDE RESULTS FOR
PHYSICOCHEMICAL PARAMETERS AND
MACROINVERTEBRATES
November
2014
1.0 Background
Waterwatch Victoria is a state-wide community based water quality monitoring
organisation that aims to increase local community understanding and ownership of
waterway and catchment issues. Community volunteers and Waterwatch coordinators are
involved in the monitoring of waterway health, including water quality and
macroinvertebrate indicators.
Waterwatch Victoria’s QA/QC Week program is an annual event co-ordinated at a statewide level. It involves the testing of laboratory prepared standard samples by monitors and
coordinators to ensure that their equipment and monitoring techniques are accurate. For
more information on Waterwatch Victoria’s Data Confidence Framework, guidelines and
data confidence plans, visit www.vic.waterwatch.org.au
This report provides a simple summary of the state-wide data provided by both coordinators and monitors during QA/QC Week 2014. It follows on from similar programs that
have been conducted annually since 1998. Further detail can be found within the regional
reports that have been distributed to the Waterwatch coordinators.
2.0 Methods
2.1 Physico-chemical Parameters
Stable ‘Mystery’ reference water samples were prepared for pH, electrical conductivity (EC),
turbidity and orthophosphate by Australian Chemical Reagents (Roach Analysts), QLD. Each
parameter was further divided into a low range sample (A) and a high range sample (B)
giving a total of 8 unique samples.
Physicochemical ‘mystery’ solutions were prepared to specifications recommended by the
state Waterwatch Victoria office. Values listed in Table 1 are the averaged solution
concentrations of the subset measured during QA/QC Week.
Individual parameter solutions were provided to all regions in separate bottles with labelled
instructions for sample preservation and use. Acceptable upper and lower quality control
limits for each parameter (excluding orthophosphate) were set according to the standards
outlined in the Waterwatch Victoria Data Confidence Framework.
Values and associated limits for each of the parameters are provided in Table 1 below.
Please note that for 2014 the EC upper and lower limits where increased slightly to
accommodate the EC meters used by most participants which only have a minimum
resolution of ± 10μS units.
Due to the continuing issues relating to colour matching with the Visocolor Colorimeters it
was decided to stretch the pass rates out to 0.05-1.0 (±33%) for participants using this
equipment. Visocolor results outside of the ± 25% range but within the ±33% range where
labelled as an applied or working pass and were shaded in blue on the result sheets.
Waterwatch Victoria QA/QC Week 2014. Version 1.0
1
Table 1. True values and acceptable limits for phys-chem parameters.
Parameter
pH (pH units)
EC (μS/cm)
Turbidity Meter
(NTU)
Turbidity Tube
(NTU)
Orthophosphate
mg/L (P)
True Value
Lower Limit
± 10%
7.0
8.4
180
8500
6.7
8.1
160
7650
Upper
Limit
7.3
8.7
200
9350
± 20%
± 20%
± 25%
± 25%
28
85
28
85
21
64
22
68
35
106
34
102
A
± 25%
B
± 25%
0.075
0.7
0.056
0.53
0.094
0.88
Sample
A
B
A
B
A
B
A
B
Acceptable
Range
± 0.3
± 0.3
± 10%
A total of 1491 phys-chem samples were sent out across the state for QA/QC Week 2014
with a total of 214 participants taking part in the event, including 10 coordinators and 204
monitors. The 2014 participation rate is almost identical to that of the 2013 program where
210 participants where recorded.
Please note that not every participant attempted all eight samples as regional co-ordinators
decided which samples they, and their monitors, would assess prior to QA/QC Week 2014.
The minimum number of samples assessed by any one participant was two, with the
majority assessing between four and seven. All participants were provided with event codes
to allow their results to be included in the state-wide analysis anonymously.
The Victorian Waterwatch Data Confidence Framework sets out minimum requirements for
equipment to be used by monitors and co-ordinators to conduct water quality analysis. A
summary of the requirements for monitoring Standards 3 and 4 is provided in Table 2
below.
Table 2. Minimum requirements outlined in the Data Confidence Framework.
Parameter
pH
Standard 3
Low range = minimum
resolution of 0.10 mS/cm
High Range = minimum
resolution of 10 mS/cm
pH meters (not strips)
Standard 4
Low range = min resolution
of 0.10 mS/cm
High Range = minimum
resolution of 10 mS/cm
pH meters (not strips)
Turbidity
Turbidity tubes
Turbidity meter
P04
Comparator
Colorimeter
EC
Waterwatch Victoria QA/QC Week 2014. Version 1.0
2
3.0 Results
3.1 Physicochemical
The following table summarises the coordinators results across all of the eight regions. A
100 percent pass rate was achieved for three of the eight parameters in 2014.
Table 3. Coordinator pass rates for the physicochemical parameters.
Parameter
Sample
A
B
A
B
A
B
A
B
pH (pH units)
EC (μS/cm)
Turbidity (NTU)
Orthophosphate
mg/L
Number of
results (n)
10
8
10
9
9
9
9
7
2014 Pass
Rate %
90
100
90
89
89
100
78
100
2013 Pass
Rate %
100
100
100
100
90
100
69
66
2012 Pass
Rate %
71
100
100
100
50
67
73
100
As can be seen in figure one the coordinator results for the pH and EC parameters were
again quite similar to those recorded in previous years. In 2012 a decrease in the turbidity
rate was observed, for both monitors and coordinators, and was attributed to the silica
based solutions trialled that year. Following the reintroduction of the formazin based
solutions in 2013 and 2014 the pass rate increased back to those levels observed in
2011/2010
Coordinators Results 2012 - 2014
120
% Pass Rate
100
80
2012
60
2013
2014
40
20
0
pH A
pH B
EC A
EC B
Turb A
Turb B
Waterwatch Victoria QA/QC Week 2014. Version 1.0
PO4 A
PO4 B
3
Figure 1. Comparison of Coordinator pass rates between 2012 and 2014.
Inter year comparisons for the orthophosphate samples are hard to make given the low
samples sizes and changes to the acceptable limits. Results have improved in 2014
particularly the high range scores, however it should be noted that these changes relate to
just 1 or two extra persons achieving a pass.
The state-wide results for the monitors are summarised in Table 5 below.
Table 4. Monitor pass rates for physicochemical parameters.
Parameter
Sample
pH (pH units)
EC (μS/cm)
Turbidity (NTU)
Orthophosphate
mg/L
A
B
A
B
A
B
A
B
Number of
results (n)
156
74
180
69
119
85
165
36
2014 Pass
Rate %
80
86
81
86
84
87
66
61
2013 Pass
Rate %
81
87
81
79
71
90
64
60
2012Pass
Rate %
84
89
93
90
54
54
70
45
The 2014 monitor results were very similar to those recorded in 2013 for all parameters,
except low range turbidity, which increased from 71 to 84%. The high range turbidity
results ,while slightly lower than those of 2013, where still significantly higher than those of
2012 again highlighting the issues with regard to the silica based samples that were trialled
in that year.
Monitors Results 2012 - 2014
100
% Pass Rate
80
60
2012
40
2013
2014
20
0
pH A
pH B
EC A
EC B
Turb A Turb B PO4 A PO4 B
Figure 2. Comparison of monitor pass rates between 2012 and 2014.
Waterwatch Victoria QA/QC Week 2014. Version 1.0
4
A comparison between the performance of the co-ordinators and that of the monitors, in
2014, is provided in figures three and four below. This year all of the results were in-line
with expectations with the co-ordinators outscoring the monitors across all ranges and
parameters.
Low Range Sample A
120
% Pass Rate
100
80
Coordinators
60
Monitors
40
20
0
pH B
EC B
Turb B
PO4 B
Figure 3. 2014 Pass rates for coordinators and monitors – Low Range
High Range Sample B
120
% Pass Rate
100
80
Coordinators
60
Monitors
40
20
0
pH B
EC B
Turb B
PO4 B
Waterwatch Victoria QA/QC Week 2014. Version 1.0
5
Figure 4. 2014 Pass rates for coordinators and monitors – High Range
4.0 Discussion
4.1 pH
The 2014 pH results were in general very good. As with previous years the key issues
impacting results were the use of pH strips, incorrect calibration of meters and the use of
faulty meters.
4.2 EC
The EC results for both the coordinators and monitors were also very good. Three issues
that have been highlighted previously stood out this year. Several of the monitors recorded
their results in millisiemens instead of microsiemens, which if not picked up by coordinators,
will lead to incorrect values been entered into the Waterwatch database. Secondly some
monitors are using high rage meters to test low range samples. And Thirdly, a number of
participants outside of the acceptable range were carrying out calibration checks using the
high range EC calibration solutions (12,880 µs/cm). For measuring EC in streams that
typically have low to moderate conductivity levels it is strongly advised that a mid range EC
calibration solution be used (e.g. 1430 µs/cm).
4.3 Turbidity
The 2014 results mirror those of 2013 and confirm that the silica based solutions trialled in
2012 are not suitable for testing with the Turbidity Tubes.
It was noted by some of the co-ordinators that different results may be obtained depending
on whether readings were taken with new or old ‘style’ turbidity tubes. In the 2014 QAQC
Week participants were asked to record which type of tube they used in order to examine
any differences between the two different styles. A total of 112 participants nominated on
the result sheets which type of tube they had used.
The results from this analysis are shown in Table 5 and Table 6 below. Average and median
scores where almost identical for both tube styles indicating that there was no significant
difference. Participant pass rates also exhibited the same pattern with only a slight increase
in pass rates observed for people using the old style tubes on the low range samples.
Table 5. Average and Median Values for new and old ‘style’ Turbidity Tubes
Parameter
New Tube
Old Tube
Sample
Average Value
Median Value
Average Value
Low
Range
32
28
31
Waterwatch Victoria QA/QC Week 2014. Version 1.0
High
Range
98
92
102
6
Median Value
28
85
Tubidity Pass Rate by Tube Type
100
90
80
% Pass Rate
70
60
50
New Tube
40
Old Tube
30
20
10
0
Low Range
High Range
Figure 5. Pass rates for participants using New and Old style Tubes. Low and High Range
results are combined.
4.4 Orthophosphate
Orthophosphate continues to be the parameter which is hardest to test accurately and
subsequently it again has the lowest pass rate. Although the acceptable limits for 2014 were
widened for those participants using Visocolor Colorimeters overall the results remained
similar to those observed in 2013.
Monitor feedback from this year’s testing once again reinforced the issues with colour
matching when using the Visocolor equipment. As stated in last year’s report further
changing of the sample colour to suit the Visocolor equipment would have a negative
impact on the Aquaquant colour matching. The use of separate samples for each type of
equipment was discussed prior to this year’s testing but was disregarded due to the
difficulties in ensuring correct sample delivery to each participant.
It is recommended that prior to testing in 2015 the use of externally prepared standards
(appropriately diluted) should be tested in order to ensure that the colour issue is not
specifically related to the methodology employed by the current supplier.
Failing positive results from this testing the use of separate coloured samples tailored to suit
each type of colorimeter should be trialled. For this to succeed there will need to be an
increased level of coordination between the supplier, the coordinators and the monitors to
ensure that each participant receives a sample suited to their equipment.
Waterwatch Victoria QA/QC Week 2014. Version 1.0
7
5.0 Pre-Testing of Coordinators.
For the 2014 program it was decided that the coordinators should be tested prior to QAQC
Week at the Waterwatch Victoria AGM. Each coordinator used their own personal
equipment and undertook measurements as they would in the field. Following the testing
the true value for each parameter was revealed and participants results where then opened
up for discussion.
The testing proved a great success, allowing the coordinators to compare their equipment
and methodologies and provide transfer of knowledge from the more experienced to the
less experienced members.
Having the results on hand also allowed Coordinators to immediately assess those monitors
outside the acceptable limits during the QAQC week gatherings. Issues relating to
equipment or methodology could then be addressed on the spot rather than two or three
months later when it is much harder to understand what may have gone wrong.
It is recommended that the same approach be adopted for the 2015 program.
Waterwatch Victoria QA/QC Week 2014. Version 1.0
8
Appendix 1: Data
Raw data from QA/QC Week 2014 can be accessed in excel spreadsheet format upon
request from Waterwatch Victoria.
Waterwatch Victoria QA/QC Week 2014. Version 1.0
9