Industry Articles

Cross Verification of Online Measurements

Star InactiveStar InactiveStar InactiveStar InactiveStar Inactive
 

Why they often don’t match up?

Many end users describe a situation where continuous online measurement sensors do not agree with lab results or grab sample measurements.  There are a whole host of factors which change or become variable when a sample is taken from the tank or even tested with another technology. These subtle variations can results in the two measurements never reaching agreement and it is very confusing to know which one is right and what you can trust for accurate process control. The variations that can exist between inline measurement and external or cross verification methods include:

  • Time – Once a sample is removed from the process the reactions continue.  It is crucial to perform a test on this sample immediately so it mostly closely represents the online process conditions. The longer the delay, the less likely the grab sample represents the actual process conditions.
  • Temperature – most process measurements are very temperature dependant, and changes to the process temperature will cause a change in the reaction and subsequent measurement. This is particularly crucial if you are dealing with a hot process liquid. Once a sample is removed from the process it will cool quite quickly and the measurement will be different. Sensors that are equipped with temperature compensation is not necessarily the solution. Firstly, temperature compensation elements need time to reach thermal equilibrium (another error source) and secondly, the reaction (activity) may be completely different at this temperature.
  • Technology – different technologies used to perform the same test will have different limitations and error sources. Depending on how the test is being carried out, this could be exacerbated leading to greater discrepancy between in tank and grab sample readings.
  • Human – The simple fact is that no two people will operate  the same. Different operators will get difference results, and the more variables the more chance for human error and variation in the task. Calibration and maintenance of test equipment, patience and care, repeatability and adhering to procedures can all have a profound effect on the individual outcomes.

So, from a practical standpoint, what can we do to ensure we get some agreement between online and external measurements? Tips on using a portable meter for cross verification of online measurement The following, while applying to a range of measurements is particularly critical for PH, ORP and Dissolved Oxygen Measurement. If you are taking grab samples for cross verification it is important to note the following:

  1. Ensure all equipment is properly cleaned, calibrated and maintained in 100% working condition.
  2. Ensure you have detailed and documented procedures for the test to ensure repeatability. No detail is too small when it comes to replicating a test.
  3. Use the same technology for both portable and online measurement. Especially for precious metal recovery applications (gold, nickel, copper, zinc) there are a lot of reactions present that can cause different errors with different technology.  Those errors can be compounded significantly when comparing measurements from 2 different technologies (each being influenced differently). To see any sort of sensible trend, both measurement should be undertaken with the same technology.
  4. Always allow the portable unit to reach thermal equilibrium.  pH and DO is highly temperature dependant, and not allowing sufficient time for a portable sensor to reach in-tank temperature or ambient (during calibration) will introduce a significant error.  Always check the temperature values on the meter to make sure the temperature compensation device is at thermal equilibrium with the measurement environment or process liquid in question. ORP is also very temperature dependant, however there is no standard compensation for this, and grab samples should not be allowed to cool and must be a close to the process as possible for a meaningful result.
  5. Never use an instrument at the extreme end of its range (ie near the limits). This can compound errors even more so.
  6. When sampling directly in tank, ensure that portable sensor is positioned at the same location and depth as the inline sensor.
  7. If measuring by grab sample (ie in bucket out of tank) make sure this is done immediately after grabbing the sample (do not allow it to sit), and keep “stirring” the sensor simulate flow past the tip. Please note this is the least reliable method for an in-tank comparison.
  8. Samples should never be taken back to a lab or workshop, and should be performed tank side. Reactions continue from the moment they are removed from the tank, and the lab sample 15 minutes later can provide very different results from what’s happening in tank.
  9. Where possible, try to use correlation data taken by the same operator each time.  If measurements are taken across multiple shifts, different operators will get different results unless the test us completely automated.
  10. If a discrepancy still exists you should also ensure that you are not dealing with electrical interference or stray current. It is very common for this to occur with inline sensors as they often present as the least resistance to earth and attract any grounding issues you may have in the process.

If you are still having trouble getting your online measurement to correlate with your grab samples, please consult your Turtle Tough representative and they will be able to assist.