Potassium Chloride dosing

[DiscusHeckel]

Hi,

A friend of mine has purchased 99% pure potassium chloride from a company. Although he has added enough KCL to increase the potassium levels in his tank by 20 ppm, testing with Salifert at two different days (three days apart) shows absolutely no increase in potassium levels.

How can this be possible unless

1. his tank consumes 20 ppm potassium in three days (is this really possible?);
2. what he is dosing is not KCL.

Thanks in advance

 

[disc1]

You may not see a change that small. You're really within the error range of the test.

 

[DiscusHeckel]

You may not see a change that small. You're really within the error range of the test.

Thanks David. He is going to calibrate his test kit against Fauna Marine's calibration solution tomorrow. However, I am not sure if the calibration will make any difference. Let's say Salifert reads 20 ppm too low. Even so, I would have expected some difference in reading even providing he makes the same mistake in two different tests. Have I understood you correctly?

FWIW, the resolution of the Salifert potassium test kit is 10 ppm.

 

[disc1]

Thanks David. He is going to calibrate his test kit against Fauna Marine's calibration solution tomorrow. However, I am not sure if the calibration will make any difference. Let's say Salifert reads 20 ppm too low. Even so, I would have expected some difference in reading even providing he makes the same mistake in two different tests. Have I understood you correctly?

FWIW, the resolution of the Salifert potassium test kit is 10 ppm.

On something like that the error range will always be larger than the precision.

Let's say we use this test with the 10ppm resolution. So that means that for any reading the true value could be 5ppm higher or lower IF everything was perfect. So we have an error range of 10ppm already just from the precision of the test.

Next there's the systematic error. It's maybe a little hard to read the color of an end point or there's a little error in the drop sizes from a dropper. That adds to this already 10ppm of error we had.

It's like this, say the K was really at 406 and went up to 424. If everything was perfect, you'd get 410 and 420 for test results. And an 18ppm spread would look like a 10ppm spread.

But everything isn't perfect. Let's say you read just a few ppm high the first time and just a few ppm low the second time. You might well end up with the same result.

 

[GPHowell]

You could take out 1 L or so of the tank water and add some CaCl2 directly to it.

The Ca level should go through the roof in that sample and you can confirm whether it is actually CaCl2 you've bought...

G