salinity calibration confusion

[defasum]

I have 2 tanks, both with apex controllers. I also sent out ICP tests and received the results for both tanks. The difference in salinity reading from the apex (compared to each other) and the ICP vary pretty significantly. I do bi-weekly water changes on both tanks from the same mixed source water so I would have expected the salinity to be close to each other but that is not the case.

So I ordered calibration fluid from Pinpoint and Aqua Craft Products. I also ordered a salinity tester from hanna which also came with it's own calibration fluid.

I have used the same refractometer for years and had always just zero'ed it out to my rod water for calibration. I realize now that is probably not the beat way to do it. Anyways I thought I would start by checking the salinity on each calibration fluid with my refracto and was expecting for them to be fairly close to each other in the 35 ppt range.

I was completely surprised at the result

Hanna = 27 ppt
Pinpoint = 38 ppt
Aqua Craft = 42 ppt

I have already calibrated the hanna with the hanna supplied calibration fluid but now I am questioning my result. I would like to calibrate both conductivity probes, Hana checker and refractometer to the same source so I can get very close to the same readings everywhere I look.

What is everyone's opinion on how I should proceed here? I am at a loss at what calibration source to use.

 

[mdb_talon]

wow those are really big differences. What does hanna give you with each of the three calibration solutions? I assume 35ppt with the hanna solution since you said you calibrated it that way, but what does it give you if you then test the other two?

I keep a floating hydrometer on hand just to give me piece of mind that my readings are similar between that and refractometer (and makes me convinced my calibration solutions are accurate)

 

[Floyd-]

Only use the Hanna fluid for the Hanna device. I calibrated mine this way and it was spot on.

 

[defasum]

I was thinking of using the same hanna calibration fluid on my apex conductivity probes so I can have a consistent source. would you not advise I do that?

 

[Dennis Cartier]

I was thinking of using the same hanna calibration fluid on my apex conductivity probes so I can have a consistent source. would you not advise I do that?

That should work fine. The caution from Floyd was probably about mixing measurement types (E.g. trying to calibrate a refractometer with the Hanna solution). The Hanna solution should work for any EC meter, including Apex EC probe.

The Aqua Craft calibration solution claims to be universal, 35 ppt for refractometer and 53 ms for EC probes. That 'should' work if brought up to 25C for calibration.

Just to add to mdb_talon's suggestion. Get a floating hydrometer for these exact situations. You will never question which one is right after that. You will know.

 

[Randy Holmes-Farley]

Welcome to the wonderful world of reef hobbyist testing.

Does each device read its own calibration fluid correctly? What about reading the other ones?

The Apex gives people troubles, perhaps partly due to the way it deals with temperature during calibration.

 

[patrickb]

Tropic Marin Hydrometer

Seriously, just get it. You don't often get ~%99.999 accuracy for $35 in this hobby. Salinity is one of the easier parameters, don't overcomplicate it.

 

[GrizFyrFyter]

First. Do not calibrate anything with RO water. Calibration is how you teach a device what is the correct information. Calibrating something to 0ppt when your testing 35 ppt is just setting yourself up for failure.

I calibrate everything by mixing my own solution. Measured amount of water and measured amount of salt (by weight) will always give you an accurate solution.

 

[Randy Holmes-Farley]

Tropic Marin Hydrometer

Seriously, just get it. You don't often get ~%99.999 accuracy for $35 in this hobby. Salinity is one of the easier parameters, don't overcomplicate it.

It may be a fine hydrometer, but it isn't anywhere near that good.

To be 99.999% accurate in salinity, you'd need to read to :

sg = 1.0260000 as different from sg = 1.0260003

Assuming you can just barely distinguish 1.0260 from 1.0261, that's an accuracy of +/- 0.4%, or 99.6% accuracy.

 

[Dennis Cartier]

The maximum deviation that TM lists is 0.001. When I first read that I was a bit disappointed, because that is not much better than other methods. So I ordered 2 of the TM hydrometers and then compared them to each other. They matched up exactly ( to 0.0001). So if there is any deviation, it seems to be consistent.

 

[arking_mark]

Welcome to the salinity measurement troubles...

Salinity Measurement Troubles...

So I've been dealing with a couple of meters that all give me different readings after being calibrated: GHL = 34.6 psu, 52.7mS, 1.0245kg/l Hanna = 35.4ppt Other probe = 37.3ppt I bought a Tropic Marin Precision Hydrometer to establish ground truth... Tropic Marin Precision Hydrometer: 1.02775...

www.reef2reef.com

 

[Pntbll687]

I'm going to second getting and using the Tropic Marin hydrometer.

Mix fresh saltwater to 1.025, with a variance of .001, the salt is between 1.024-1.026. Done. Just make sure water temps are the same when taking readings.

Or you can stay on the calibrate, check salinity, guess about calibration, recalibrate, test, check with another tester, calibrate that tester, retest....

But in the end it's all about what are you willing to do. Myself, I'm not going to calibrate a refractometer often enough to use it effectively. I used a swing arm hydrometer for a LONG time, and never had problems with salinity.

 

[Randy Holmes-Farley]

I'm going to second getting and using the Tropic Marin hydrometer.

Mix fresh saltwater to 1.025, with a variance of .001, the salt is between 1.024-1.026. Done. Just make sure water temps are the same when taking readings.

Or you can stay on the calibrate, check salinity, guess about calibration, recalibrate, test, check with another tester, calibrate that tester, retest....

But in the end it's all about what are you willing to do. Myself, I'm not going to calibrate a refractometer often enough to use it effectively. I used a swing arm hydrometer for a LONG time, and never had problems with salinity.

FWIW, my Orion 128 conductivity meter never changes, and in fact, isn't able to be adjusted to match a calibration (at least without opening the back, which I never tried).

In the more than a decade, it always read the same, very slightly high, for a 1 L container of 30.2 mS/cm standard I bought from Oakton/Cole Parmer.

 

[Atrumblood]

First. Do not calibrate anything with RO water. Calibration is how you teach a device what is the correct information. Calibrating something to 0ppt when your testing 35 ppt is just setting yourself up for failure.

I calibrate everything by mixing my own solution. Measured amount of water and measured amount of salt (by weight) will always give you an accurate solution.

Pardon my ignorance, but can you explain how calibrating with a 1.026 standard is different from calibrating with 0 ppm (RODI) water at the same temperatures?

No seriously, I don't understand.

 

[GrizFyrFyter]

Pardon my ignorance, but can you explain how calibrating with a 1.026 standard is different from calibrating with 0 ppm (RODI) water at the same temperatures?

No seriously, I don't understand.

You calibrate equipment based on the target parameter, not the opposite end of the parameter range. Calibrating a refractometer with fresh water only gives you an accurate reading of fresh water.

Same reason we calibrate our pH probes with 7.0 and 10.0 solution instead of 4.0 and 7.0 solution.

 

[Atrumblood]

You calibrate equipment based on the target parameter, not the opposite end of the parameter range. Calibrating a refractometer with fresh water only gives you an accurate reading of fresh water.

Same reason we calibrate our pH probes with 7.0 and 10.0 solution instead of 4.0 and 7.0 solution.

Ok, so then wouldn't it make more sense to use a 2 point calibration? Though I guess the meter would have to be designed that way.

 

[Randy Holmes-Farley]

Ok, so then wouldn't it make more sense to use a 2 point calibration? Though I guess the meter would have to be designed that way.

A mechanical refractometer cannot do two point calibration. There's only a single thing to adjust.

That said, there's another important reason to use calibration fluid.

A perfectly made true seawater refractometer that is perfectly calibrated with 0 ppm TDS water will read 35 ppt seawater salinity correctly.

An imperfectly made true seawater refractometer that is perfectly calibrated with 0 ppm TDS water may read 35 ppt seawater salinity incorrectly.

A perfectly made true seawater refractometer that is perfectly calibrated with a 35 ppt standard will read 35 ppt seawater salinity correctly.

A perfectly made brine refractometer (which represents the bulk of cheap refractometers sold to reef hobbyists; any that do not say true seawater refractometer likely fit this description) that is perfectly calibrated with 0 ppm TDS water will NECESSARILY read 35 ppt seawater salinity incorrectly by a known amount.

A perfectly made brine refractometer (which represents the bulk of cheap refractometers sold to reef hobbyists; any that do not say true seawater refractometer likely fit this description) that is perfectly calibrated with a 35 ppt seawater water will read 35 ppt seawater salinity correctly.

An imperfectly made brine refractometer (which represents the bulk of cheap refractometers sold to reef hobbyists; any that do not say true seawater refractometer likely fit this description) that is perfectly calibrated with a 35 ppt standard ppm TDS water will read 35 ppt seawater salinity correctly.

Thus, there are several common scenarios where a 0 ppt standard (RO/DI) may or must read salinity incorrectly, while there is NEVER a scenario where any refractometer properly calibrated (or at least checked) with a proper 35 ppt standard will not correctly read seawater salinity to be 35 ppt.

 

[Atrumblood]

Thanks for all the details, @Randy Holmes-Farley . That makes much more sense.

 

[GrizFyrFyter]

Ok, so then wouldn't it make more sense to use a 2 point calibration? Though I guess the meter would have to be designed that way.

That would require the ability to adjust multiple points of calibration. Your adjustment screw is a single point adjustment.

 

[Atrumblood]

That would require the ability to adjust multiple points of calibration. Your adjustment screw is a single point adjustment.

So then my next question pertaining to using 0 as the calibration point not being the way to go. Is that because the response isn't linear?

I understand that calibrating using a standard at the value you are testing for is the best way, but why is that?

If I calibrate using a standard does that mean only the value of that standard is confirmed to be accurate and any other value we try to test for hasn't been calibrated for and thus can't be trusted?

Or is there something deeper I am missing here?

Sorry for the shot gun questions.