Study of Measured CYA readings vs Sample Temp
I decided to run a detailed trial of measured CYA readings vs sample temperature.
Conditions:
All temperature measurements were made using a digital thermocouple.
Air temp = 75F
Pool water temp =73F
A Taylor K-2006 kit was used to measure CYA levels
All measurements were made within one 15 minute period.
Sky was cloudless; Sun almost directly overhead. During measurements, sun was behind me coming over my left shoulder.
Procedure:
A two cup sample was prepared using pool water and tap water. Dilution was 1 part pool water and 3 parts tap water.
Sample bottle was filled with diluted sample to 7 ml. R0013 was added to bring total volume to 14 ml.
Sample was shaken for 30 seconds. Then let rest for 60 seconds, Then sample bottle was immersed in an ice bath or warm water bath to achieve desired temperature. The bottle then was shaken for 10 seconds. Then sample was checked to assure no bubbles were present. Then thermocouple probe was used to measure sample temperature. Then the CYA measurement was taken. Measuring column was slowly filled until black dot disappeared. Then raw reading was observed on side of column. During column fill procedure, no observations were made of column height to assure elimination of user expectation bias. Raw reading was made only after column had been filled to the point of having the black dot disappear.
Note that ALL measurements were made from a SINGLE FILL of the Taylor K-2006 sample bottle.
Results:
55F raw reading of 100 meaning CYA level of 400
65F raw reading was 90 meaning a CYA level of 360
74F raw reading was 100 meaning a CYA level of 400
86F raw reading was 95 meaning a CYA level of 380
Conclusion:
Between sample temperatures of 55F and 86F there is no statistically significant CYA reading dependence on sample temperature. See graph below.
Comment: Under the very best of highly controlled circumstances there appears to be error bars of about +/- 10 percent on CYA readings. For casual readings the error bars are probably +/- 20%. Errors are NOT +/- an absolute number but are a percentage. Also, realize that the diluted samples did not have astronomically high CYA levels, but were equal to or less than the raw reading of 100, which everyone considers to be within a proper range of the Taylor test kit.
I cannot upload the graph because of a memory quota on the website. Can I get a dispensation, or can I delete graphs from previous posts in this thread to free up memory for new graphs?
Richard