13 / 59
Ref. No. [UMCES] CBL 2016-011
ACT VS16-02
13
(3) Using a gravimetrically calibrated pipet add a suitable volume of the KIO
3
standard to the
stirring solution
(4) Insert the electrode and delivery tube and immediately begin titration
(5) The normality of the thiosulfate is calculated from the equivalence point volume as Vol
KIO3
/
Vol
Thio
)* N KIO
3
using replicates of single KIO
3
volume additions or from the slope of a range of
KIO
3
addition volumes.
Blank determination
Reagent blanks were determined as follows:
(1) A volume of 1-2 L of site water was brought to a boil in a clean glass reagent bottle.
(2) Boiled, degassed water was cooled and poured into 125 ml sample flasks and sparged with N
2
for no less than 30 minutes.
(3) The sample was then rapidly fixed as a normal sample, and on the auto titrator.
(4) A global reagent blank taken as the mean of samples fixed at each test site (0.078 ± 0.020, n=5)
and used to correct all reference values.
RESULTS of LABORATORY TESTING
Instrument accuracy for the PME miniDOT sensor was tested under nine combinations of
temperature and salinity over a range of DO concentrations from 10% to 120% of saturation.
Specific test conditions are summarized in Table 1. Results are plotted as a time series of
instrument readings against the time series of comparative Winkler reference samples (Figures A-
I). The bottom panel of each figure shows the time series of the difference in instrument
measurement versus corresponding reference sample. Plotted differences were limited to a range
of ± 2.0 mg/L, as values exceeding this range are well outside any expected performance
specification and likely occurred as a result of bubbles from the sparging gas used to vary
dissolved oxygen concentrations getting trapped on the sensor foil. Comparisons of tank
duplicates taken at opposite sides of the tank from between 9-18 of the timepoints during each trial
showed a mean difference of 0.017 mg/L, with a range over the nine trials of 0.006 – 0.038 mg/L.
Those values are within the analytical precision of the reference method and indicate conditions
throughout the tank were very homogeneous and trapping of bubbles on the sensor foil happened
as isolated events, which were clearly distinguishable. However, small changes in measured DO
concentrations did occur during some of the sampling phases indicating the tank was slightly
moderating after sparging was stopped. Those small drifts in DO concentrations were clearly
captured by both instrument and reference sample measurements.