Typical accuracy of salinity testers? How precisely do you control your tank salinity?

[Kilman805]

I bought a pretty inexpensive salinity tester off Amazon along with a 35 ppt calibration solution.

When I test the calibration solution, I measure 34.1 ppt (1.0257 SG) vs. the stated 35 ppt (1.0264 SG).

If I look at the specs on the Tropic Marin High Precision Hydrometer they claim "Maximum deviation: 0.001 at 77° F (25° C)" and "Resolution: 0.0001". I think that means that while it is very precise, the absolute accuracy can be 10X worse than its precision.

Comparing those specs to the salinity tester, I think it's actually performing pretty well. But I don't have the hydrometer, so I can't compare actual measurements.

What do you think? How closely do you try to control the actual salinity in your tank?

 

[FishTruck]

I use a conductivity probe continuously. and a refractometer to check the conductivity probe. The refractometer rarely needs recalibration. The conductivity probe gets out of calibration without much warning (air bubbles, I think, or bio film maybe). The probe has gone about 4 months on it's current run.

Since your conductivity probe is not soaking in the tank, I'll bet it will hold calibration longer.

Remember though, a conductivity probe, hydrometer, and refractometer are measuring three different things - so variations may occur - even when all three are working properly. Somebody smarter will explain more than I can.

One other tip - the calibration fluid used for the refractometer is DIFFERENT than the one used for the conductivity probe. They are not interchangeable! Making sense of this might drive you crazy, unless you are a chemist. I don't use a hydrometer - so no comment there - but beware using conductivity calibration fluid.

I keep the tank between 34 and 35 ppt with automated adjustment using the conductivity probe and an automated system than can raise or lower salinity. This is much more stable than my prior system which would vary between 1.021 and 1.026 SG from month to month - due to my laziness of not checking it properly and being inconsistent with water changes.

 

[Kilman805]

One other tip - the calibration fluid used for the refractometer is DIFFERENT than the one used for the conductivity probe. They are not interchangeable!

I don't think that's true as long as we're talking about NaCl solutions in water, which hopefully we are since that's what we're trying to measure.

For a 35 ppt NaCl solution in water, it should be:
1.0264 SG (as measured with a hydrometer)
53 ms/cm (as measured with a conductivity meter)
1.3394 (refractive index.. what a refractometer really measures even though the scale is usually calibrated to read SG)

For sure you you need to know what units you're measuring and compare to the corresponding units of the calibration solution.

 

[KStatefan]

I don't think that's true as long as we're talking about NaCl solutions in water, which hopefully we are since that's what we're trying to measure.

For a 35 ppt NaCl solution in water, it should be:
1.0264 SG (as measured with a hydrometer)
53 ms/cm (as measured with a conductivity meter)
1.3394 (refractive index.. what a refractometer really measures even though the scale is usually calibrated to read SG)

For sure you you need to know what units you're measuring and compare to the corresponding units of the calibration solution.

But we are not measuring NaCL we are measuring sea water.

 

[Dragen Fiend]

I use a cheap $30 refractometer from amazon.
Stable wise Its from 1.025-1.026 on a given day.
I re-calibrate once a month with distilled water.

I've never actually stressed about salinity.
During water changes I don't even match my salinity sometimes. Some days its mixed to 1.024 and other days 1.026.

 

[FishTruck]

Check the labels on your calibration fluid - this might vary. On mine it is printed "do not use this for refractometer probe calibration" or something like that. If it was sea water all around, I agree it would be fine. Drove me nuts when I tried to use it for both... I'd get them in sync on the calibration fluid... but then they would mismatch in the tank.

 

[Kilman805]

But we are not measuring NaCL we are measuring sea water.

Sure, but the concentration of NaCl in seawater is ~35 ppt (parts per thousand) and everything else is parts per million. Even calcium and magnesium are ~200-1000X more dilute than NaCl and have negligible impact on the measured values of conductivity, SG, or refractive index.

I would be really surprised if any of the calibration solutions contain anything besides NaCl and water. Most of the ones on BRS list all three units of measure.

 

[FishTruck]

Sure, but the concentration of NaCl in seawater is ~35 ppt (parts per thousand) and everything else is parts per million. Even calcium and magnesium are ~200-1000X more dilute than NaCl and have negligible impact on the measured values of conductivity, SG, or refractive index.

I would be really surprised if any of the calibration solutions contain anything besides NaCl and water. Most of the ones on BRS list all three units of measure.

Like I said... this may vary by product. You will know they are not compatible calibrations fluids if you find that you cant make your instruments agree on seawater - when they do agree on the calibration fluids. Also, comparing accuracy between a hydrometer and a conductivity probe is not a thing.

 

[Randy Holmes-Farley]

I don't think that's true as long as we're talking about NaCl solutions in water, which hopefully we are since that's what we're trying to measure.

For a 35 ppt NaCl solution in water, it should be:
1.0264 SG (as measured with a hydrometer)
53 ms/cm (as measured with a conductivity meter)
1.3394 (refractive index.. what a refractometer really measures even though the scale is usually calibrated to read SG)

For sure you you need to know what units you're measuring and compare to the corresponding units of the calibration solution.

Sodium chloride standards are not interchangeable. They cannot match all of the conductivity, refractive index, and specific gravity of 35 ppt seawater at the same time.

Seawater mimic solutions would be.

 

[KStatefan]

Sure, but the concentration of NaCl in seawater is ~35 ppt (parts per thousand) and everything else is parts per million. Even calcium and magnesium are ~200-1000X more dilute than NaCl and have negligible impact on the measured values of conductivity, SG, or refractive index.

I would be really surprised if any of the calibration solutions contain anything besides NaCl and water. Most of the ones on BRS list all three units of measure.

If they list all three units they are not NaCl.

Randy has a good article on DIY calibration standards using table salt.

 

[Randy Holmes-Farley]

I bought a pretty inexpensive salinity tester off Amazon along with a 35 ppt calibration solution.

When I test the calibration solution, I measure 34.1 ppt (1.0257 SG) vs. the stated 35 ppt (1.0264 SG).

If I look at the specs on the Tropic Marin High Precision Hydrometer they claim "Maximum deviation: 0.001 at 77° F (25° C)" and "Resolution: 0.0001". I think that means that while it is very precise, the absolute accuracy can be 10X worse than its precision.

Comparing those specs to the salinity tester, I think it's actually performing pretty well. But I don't have the hydrometer, so I can't compare actual measurements.

What do you think? How closely do you try to control the actual salinity in your tank?

Most of the devices we use do not give accuracy ranges, and when they do it is certainly not clear what they did to determine that range.

That said, the Hanna conductivity meter says the info below, giving it a similar sg +/- 0.001 to the TM statement you quote.

Salinity Accuracy

±1 ppt for 0.0 to 40.0 ppt; ±2 ppt for readings over 40.0 ppt; ±1 PSU for 0.0 to 40.0 PSU ; ±2 PSU for readings over 40.0 PSU; ±0.001 S.G.

The Milwaukee digital refractometer 887 says

• Accuracy: ±2 PSU | ±2 ppt | ±0.002 S.G. (20/20) | ±0.3°C / ±0.5°F

Which is twice as large of a range.

 

[FishTruck]

By accuracy, I take this to mean that after you’ve calibrated the device, you could run the same test over and over on the same batch of water, but the results could vary a bit with each measurement. Is that correct?

 

[Randy Holmes-Farley]

By accuracy, I take this to mean that after you’ve calibrated the device, you could run the same test over and over on the same batch of water, but the results could vary a bit with each measurement. Is that correct?

That is what it normally means, yes.

But how exactly they determine these numbers is unclear.