What can cause intermittent spikes (high readings) on the APA6000 alkalinity analyser?

Document ID

Document ID TE6318

Version

Version 4.0

Status

Status Published

Published Date

Published Date 21/01/2019
Question
What can cause intermittent spikes (high readings) on the APA6000 alkalinity analyser?
Summary
An Explanation of Intermittent Spikes in APA6000 Alkalinity Analyser Sample Readings.
Answer
The APA6000 Alkalinity analyser will sometimes intermittently display very high – and incorrect – sample readings (spikes).  This occurrence can upset process controls, or it might cause concern to the analyser’s operators.  A recent modification of the analyser software (V5.004) corrects this problem. H ow the problem occurs, and how the revised software corrects this, are explained below.
 
The APA6000 Alkalinity analyser performs the following steps in measurement of sample concentration:
 
1.  Sample is aspirated into the holding coil. 
 
2.  Indicator Reagent is aspirated into the holding coil.
 
3.  Sample and Indicator Reagent are dispensed to the mixer module and stirred. 
 
4.  A portion of the sample/indicator mixture is aspirated into the holding coil.
 
5.  A small volume of Acid Titrant Reagent is aspirated into the holding coil.
 
6.  The remainder of the sample/indicator mixture is aspirated into the holding coil. 
 
7.  The contents of the holding coil are dispensed through the reaction coil to the colorimeter.  The analyser takes continuous measurements from the colorimeter as the complete volume of sample/indicator – acid – sample/indicator passes through the flow cell. 
 
As the liquids flow through the reaction coil, a reproducible dispersion (spreading) of the sample/indicator with the acid titrant occurs.  As the two fluids disperse, a full range of concentrations result.  This range begins with all sample/indicator then transitions to a greater concentration of titrant.  Eventually, a maximum concentration of titrant is reached and as the transition continues, the titrant strength drops until only sample/indicator is present again.  The resulting time vs. transmittance response from the colorimeter is a “hump”-shaped curve. 
 
(The description above is simplified for clarity, and does not completely describe all steps of the sample measurement cycle.) 
 
 

 
The blue line in the graph above represents a typical measurement response curve for the alkalinity analyser.  The red arrows on the curve indicate approximate points for color transition indicating Total alkalinity; the black arrows indicate approximate points for color transition indicating Phenolphthalein alkalinity.  Concentration is inversely proportional to the width of the curve.  (Narrower curve represents higher alkalinity.) 
 
  
 

 
Two sample readings are overlaid on the graph above, representing both normal and ‘spiked’ readings. 
 
The red line in the graph above represents a typical response curve for an alkalinity sample measurement.  The blue line shows response as a small bubble passes through the colorimeter.  When the bubble passes through the colorimeter light beam, it causes a sharp drop in the light intensity seen at the photocell.  Response quickly returns to normal levels after passage of the bubble. 
 
In this case, the analyser software mistakenly sees the drop caused by the bubble as the ‘tail end’ of the alkalinity response curve.  Because the curve is seen as very narrow, a very high alkalinity concentration is seen. 
 
Notice also that in the example above only the P alkalinity reading – indicated by the black arrows – is severely affected.  The T alkalinity reading – indicated by the red arrows, and lower on the curve – is slightly higher than normal, but not to a great degree.
 
Bubbles in the sample during the measurement cycle are random events, and almost always never occur during two adjacent cycles.  This ‘single event’ characteristic of bubble occurrence is monitored in the new analyser software (V5.004) as a basis for rejecting intermittent high sample readings.
 
The revised analyser software compares each sample measurement to the previous measurement.  If a sample concentration deviates from the previous measurement by more than the greater of 15% or 0.5 mg/l, then the new reading is rejected and the old reading is held.  If a second subsequent measurement deviates from the original by the same degree or more, it is accepted.  This validation is reset to the new sample reading after every measurement cycle.