Introduction

This article will lead you through a salt dilution flow measurement using the QiQuac system.  It is the same procedure using any other EC and temperature system, but the diagnostics and calculated Q will not be available in real-time.  Have a look at a video showing the QiQuac in action.  The YouTube version features auto-translate subtitles into other languages, while the Vimeo version can be downloaded to a mobile device for reference.

CF.T Calibration and SD Measurement

1. To begin a measurement, select “Start Logging” from the main menu. QQ will begin listening on the serial ports for transmissions.  When a transmission arrives, it will be displayed.  You can either leave QQ recording, or push the button to begin a measurement.
2. Enter the Mass of salt injected when prompted.  Turning the knob faster will result in larger changes.  Set the default Mass in settings.
3. Next, you will be asked to set the CF.T, or hold to calibrate.  The CF.T should be ~0.49±0.02 (mg/l)/(μS/cm) for NaCl in freshwater.  However, this will depend on the unit’s current EC and T calibration and also to some degree (±3%) on the stream chemistry.  If the CF.T is already known, turn the rotary encoder to set it.  To start a CF.T calibration, push and hold the button for ~1 second.  The calibration procedure will begin with the next received serial transmission.

Calibration (5-10mins)

4. If Calibration mode was selected, you will be asked to inject 1ml of 5 g/l standard into 1000 ml of stream sample[1]. Before doing this, you should prepare the stream sample.  Place the probe(s) in the stream to reach equilibrium.
5. If using QuickFlask, open the port and dunk the flask in the stream.  While holding the handle such that the flask is plumb, allow the sample to drain out until the meniscus is below the black lines[2].  Close the port.
6. Note the EC.T when the probes are in the stream.  Now put the EC.T probes into QuickFlask and note the EC.T.  It may increase by a few μS/cm but not more.  If it does, repeat step 5.  Push the button to set the first EC.T in the calibration.  As calibration points are collected, they are displayed (press and hold the button to remove the last point.)
7. Inject 1ml of standard and mix until the EC.T stabilizes.  Push the button to accept this value.  Repeat for 3 more injections.  After the last button push, QiQuac will calculate and display the final CF.T.  The distilled water correction has been applied to the CF.T.

Salt Dilution Measurement (10-60 mins)

8. Ensure the shroud is on the EC Probe.  Place the probe in the stream in moving, but not turbulent or aerated, water.  Wait for the EC.T and temperature to stabilize.  It may be necessary to lightly knock the shroud to release any entrapped air.  Push the button to begin calculating the background EC.T[3].  QiQuac uses a 10-point Pre- and Post- average for BG EC.T.  The Post-BG EC.T is a moving average of the last 10 points, thus the QiQuac will begin calculating the Post-BG EC.T immediately after the Pre-.
9. Inject the required amount of salt[4] a sufficient distance upstream[5] in sufficiently turbulent flow[6].
10. As the salt wave travels past the EC.T probe(s), observe the EC.T trace rise and fall, as shown in Figure 1.  The derived Q will change as the EC.T changes.  The EC.T may take a very long time (>1hr) to reach BG EC.T, when the derived Q does not change much with time.  QiQuac shows the change in the derived Q for each new measurement of EC.T.  When this is smaller than ±1% it can be assumed to have reached BG EC.T.  If time constraints (helicopter for example) prevent this, there will be more uncertainty in the derived Q.
11. Push the button to finish the measurement.  The difference between the pre- and post-wave BG EC.T will be taken into account in the derived Q and associated uncertainty.[7]
12. After the measurement is finished several statistics will be displayed.  These will be written to the SD card with a timestamp.  The displayed uncertainty is the 95%Confidence Interval.  If the %DQ is greater than the sum of the two QUnc., it is considered to be a significant difference and the measurements should be redone.  If the %DQ is less than the sum of the two QUnc. it is not a significant difference, however the user should evaluate if the QUnc. is acceptable for the required measurement accuracy, i.e. RISC Grade A is <±7%.  The measurement is graded (A,B, or C) by both the individual QUnc. and the %DQ[8].
13. Push the button again to return to the main menu.  If the QQ is shut off before returning to the main menu, there is a small risk of damage to the SD card.

Reference
Sentlinger, G.I. “Salt Dilution Uncertainty and Proposed Metadata Requirements.” CWRA Conference, Vancouver, B.C. 2013.
[1] These parameters can be changed in the Setup menu.
[2] The accuracy of the stream sample volume will depend on this step and the user should try several test runs with QuickFlask and a laboratory scale to determine the volume and uncertainty of the sample.
[3] Note that if two EC.T probes are being used, turning the dial counter-clockwise will toggle the values between the two channels. The currently displayed channel is indicated by CH0D or CH1D.
[4] The rule of thumb dosing guideline is 1 kg per m³/s and should not exceed ~5 kg per m³/s. There are more detailed dosing guidelines in the references. The required amount can be reduced for lower BG EC.T values, or for shorter transit times.
[5] A sufficient distance is typically 10-20x the channel width. This can be reduced for very turbulent reaches or where there is a constriction. The mixing distance must be increased for less turbulent flow and may not work at all for laminar flow. Placing one of the two EC.T probes in the channel center and the other on the bank should help determine if complete mixing did not occur. Averaging the two derived Q’s should be more accurate than if both probes are placed near the bank.
[6] The salt should be injected in very turbulent flow. Ensure there is no salt left on the bottom of the channel. The salt can be injected slowly over time or spread out over the channel width to assist in mixing. The salt can also be pre-dissolved in stream water before injection.
[7] The Q%Unc and SNR both use the Standard Deviation of the entire 20point BG EC.T sample.  If the Pre- and Post-samples vary significantly (i.e. incomplete trace) this will be reflected in the two diagnostics.
[8] Grade A is assigned if QUnc. for CH0 and for CH1 and %DQ are all less than Grade A Thresh%, likewise for Grade B. Grade C is at least one of these three metrics is greater than Grade B Thresh %. These Thresh values can be changed in SDIQ Settings.