ICP Salinity High

Nick Rose · Sep 19, 2019

I had sent in a water sample just to see how my parameters were. Everything came back normal except for phosphorus(which I already knew was high), low strontium, and high salinity. The one that I question was the salinity. Before I sent the water sample off I tested the normal parameters and the salinity came to be 35ppt (and I just calibrated the refractometer)but the test came back at 37.59ppt. I messaged the company and they said “Hello Nick, our ICP is calibrated with NIST standards and we actually add up all the elements to give you a true value of parts per thousand. A refractometer is not as accurate as an environmental instrument designed to detect every element in the salt water.”

My question is, should I trust the 37.59ppt? If I do, when I use my refractometer should I actually be looking for 32-33ppt which would be within that 2.59ppt difference?

Randy Holmes-Farley · Sep 19, 2019

I don’t know which, if either, is accurate. That’s not my preferred icp company, but you refractometers may certainly be off.

Did you calibrate it with a 35 ppt standard, if not, do so.

User1 · Sep 19, 2019

Randy Holmes-Farley said:
I don’t know which, if either, is accurate. That’s not my preferred icp company, but you refractometers may certainly be off.

Did you calibrate it with a 35 ppt standard, if not, do so.

I've not used that company but I did use ATI's and will say that when I sent in my first test the one thing they reported was my salinity was low. As Randy suggested I ordered a 35 ppt standard, calibrated my meter, and sure enough ATI's numbers matched. I was low.

Not that it is needed the calibration and 35 ppt standard is a must have in my opinion.

Nick Rose · Sep 19, 2019

I did mention in the post that I had calibrated the refractometer just before taking the sample for the ICP test. The fluid is about 8 months old.

BeejReef · Sep 19, 2019

Hmmm.. I mean, you're trusting them on every other element. You're trusting them on phosphorous.
If it were me, I would order a new bottle of calibration fluid and in the meantime aim for 34ppt salinity. There's no harm in running at 33 or 34.

Then there's no urgency. You can tinker with your refracto, take a sample to the lfs, try your new calibration fluid.
Not sure there is a right answer. In my very short time in the hobby, one thing I have learned is to be very, very suspect of any at-home tests and kits.

jsker · Sep 19, 2019

Trust the ICP labs results and I would suggest recalibrating with a different solution. What brand of refractometer are you using? I have a budget refractometer which is good for a quick check. I also have a digital bench top refractometer, and I can not go wrong with that one

Randy Holmes-Farley · Sep 19, 2019

Nick Rose said:
I did mention in the post that I had calibrated the refractometer just before taking the sample for the ICP test. The fluid is about 8 months old.

But you did not say how you calibrated it. If it was with ro/di, then I would redo it with a quality 35 ppt standard. For many refractometers, ro/di is an inherently flawed method, despite resellers claims of suitability.

Nick Rose · Sep 19, 2019

Randy Holmes-Farley said:
But you did not say how you calibrated it. If it was with ro/di, then I would redo it with a quality 35 ppt standard. For many refractometers, ro/di is an inherently flawed method, despite resellers claims of suitability.

sorry about that. It was with Brightwell Calibration fluid.

Nick Rose · Sep 19, 2019

I also forgot to mention the other reason I question if the Icp salinity is correct is the salt I use. I use Aquavitro Salinity which says a half cup in a gallon mixes to 35ppt. Each week I mix 15 gallons of water with 7.5 cups of salt and it always mixes to 35ppt.

JimWelsh · Sep 20, 2019

Nick Rose said:
My question is, should I trust the 37.59ppt?

In a word: NO!

First, if you trust their numbers, and add up all the major the elements, then you get a salinity of 35.0752 using this excellent calculator. I have also previously published a spreadsheet that is not as thorough as the cited calculator that comes up with a value of 35.09, so the two are in very close agreement. I have no idea how they manage to "add up all the elements" and be that far off. You can check the numbers yourself: Add up all the elements, except convert S to sulfate by multiplying by S by 3.0, convert B to borate by multiplying B by 5.7, and convert P to phosphate by multiplying by 3.066. You should get an answer of 35,800 PPM, which corresponds to 35.8 PPT, but then the conversion of PPT to PSU makes the final answer S = 35.63. The calculator and my spreadsheet don't include the minor elements, and those minor elements do add up, especially with your PO4 level. Still, nowhere near the 37.59 that they came up with.

Second, it can be easily demonstrated that you cannot trust their numbers. Each element (with a couple of minor exceptions, which I'm prepared to address in detail if necessary) manifests as either a cation or an anion. The charge balance of the moles of each cation times the charge of that cation vs. the moles of each anion times the charge of that anion must cancel each other out. Physics demands this. The charge balance of the data you presented is out by the equivalent of -798 PPM of Cl-. In other words, in order for the charges to balance out, then there must be either a significant decrease in the amount of cations present, or a significant increase in the amount of anions present, or some combination thereof. The "minor exceptions" I mention are about two orders of magnitude smaller than the error I am describing.

Trust your recently calibrated refractometer; it much more closely matches the data, to the degree the data can be trusted.