Salinity question

[EnterName]

I used a hanna sachet (kcl per google AI) to perform the temp tests. I used Randy’s DIY recipes https://www.reef2reef.com/ams/reef-aquarium-salinity-diy-calibration-standards.956/ to check that the Hanna and hydrometer were performing to standard. The thermometer on my Hanna records a nearly identical temp for a my tank compared to the Apex probe.

Well, which one did you use?

If you want to check whether your hydrometer is accurate, you'll need the Specific Gravity Standard. According to Randy's DIY recipe, a 3.714% (w/w) NaCl solution has a specific gravity (d20/20) of 1.0266 at 20°C.

To prepare it, dissolve 6.2 g NaCl in 161 mL of water or 38.51 g NaCl in 1 L of water. Just remember to correct for temperature manually and use RO/DI.

If you're calibrating a Hanna Instruments salinity meter, you'll need the Conductivity Standard instead. Randy's recipe calls for a 3.29% (w/w) NaCl solution, which gives a conductivity of 53 mS/cm (though it's not specified whether that's at 20°C or 25°C).

To make that solution, dissolve 6.2 g NaCl in 182 mL of water or 34.07 g NaCl in 1 L of water.



To verify that Randy's DIY conductivity standard works with the Hanna meter, I ran a small experiment:

1. Hydrometer check: I measured my tank water using my trusted JBL Hydrometer 6140800 (error of ±0.0005). It read:

• Specific Gravity (d25/4): 1.0237 (commonly used standard in my region and scientific standard)
• Specific Gravity (d20/20): 1.0269 (the standard most people use in the US, which matches the Tropic Marin hydrometer)
• PSU: 35.4

2. Cleaning the Hanna meter: My Hanna salinity meter hadn't been used in a while and gave readings all over the place, so I cleaned the electrodes by dipping them in 40% acetic acid for 2-3 minutes. After cleaning, the readings became stable, though still incorrect.

3. Preparing the standard: I made Randy's 3.29% solution with 34.07 g NaCl (> 99.5% purity) in 1 L of RO/DI water at 20°C (0.9μS/cm conductivity according to HI98309). To do this I measured precisely with a 500 mL glass cylinder (±5 mL at 20°C). After everything was mixed and dissolved, I poured 80 mL into a small beaker, warmed it in my hand until the Hanna meter read 25°C, and then performed the calibration.

4. Final check: After calibration, I measured my tank water again with the Hanna meter and it showed a stable 35.4 ppt, exactly as expected.

The thing is, I also own a GHL conductivity probe which currently shows a PSU of 34.2 and is calibrated with a 50ms/cm at 25°C solution. It has been calibrated two weeks ago and both the calibration standard and the probe are brand new, so I will have to double check if now both hydrometer and Hanna salinity meter are wrong, or if the GHL probe is off.

 

[DavidA]

I used Randy’s conductivity recipe for the Hanna and SG recipe for the hydrometer. My mistake was that the Hanna was calibrated at 68 degrees

 

[EnterName]

I used Randy’s conductivity recipe for the Hanna and SG recipe for the hydrometer. My mistake was that the Hanna was calibrated at 68 degrees

Okay I see...

I guess the Hanna salinity meter assumes you are using the Hanna calibration solution, so it applies a temperature compensation formula which isn't working for NaCl solutions. In other words you would need a conductivity standard that matches the conductivity of the Hanna Instrument calibration solution at the temperature you are trying to calibrate at.

The practical salinity scale (PSS-78) used for PSU measurements is based on a 32.4356g/kg KCl reference at 15°C/59°F. This reference has a conductivity of 42.914mS/cm @ 15°C/59°F. When warmed up to 25°C/77°F it should be around 53mS/cm, which is also the target of Randy's DIY solution. Unfortunately I wasn't able to find the temperature at which Randy's DIY solution is supposed to reach 53mS/cm, but I assume it's 25°C/77°F.

This means if you calibrate using Randy's DIY solution, you should do it at 25°F/77°F. This seems to be working, according to my previously commented experiment.

Really interesting stuff, thank you for posting this, I feel like I have learned something today :)

 

[Randy Holmes-Farley]

I too am surprised that Hanna would design it with KCl standard and then have to use a different temp correction during calibration than in measurement.

I think most companies make a seawater mimic (perhaps leaving out alk to make it more stable) and then can just use one correction. American Marine/Pinpoint uses a seawater mimic standard.

 

[DavidA]

Okay I see...

I guess the Hanna salinity meter assumes you are using the Hanna calibration solution, so it applies a temperature compensation formula which isn't working for NaCl solutions. In other words you would need a conductivity standard that matches the conductivity of the Hanna Instrument calibration solution at the temperature you are trying to calibrate at.

The practical salinity scale (PSS-78) used for PSU measurements is based on a 32.4356g/kg KCl reference at 15°C/59°F. This reference has a conductivity of 42.914mS/cm @ 15°C/59°F. When warmed up to 25°C/77°F it should be around 53mS/cm, which is also the target of Randy's DIY solution. Unfortunately I wasn't able to find the temperature at which Randy's DIY solution is supposed to reach 53mS/cm, but I assume it's 25°C/77°F.

This means if you calibrate using Randy's DIY solution, you should do it at 25°F/77°F. This seems to be working, according to my previously commented experiment.

Really interesting stuff, thank you for posting this, I feel like I have learned something today :)

For the same of clarity, I used Hanna KCL calibration fluid to calibrate and Randy's recipes to assess the accuracy of the Hanna and hydrometer. What I learned from my experiments is that the Hanna temp adjustment algorithm works for seawater but not for KCL calibration fluid and thus the instrument need to be calibrated with fluid at tank temp.

 

[Randy Holmes-Farley]

Just looking for clarity, where do you see it is KCl? All I see from Hanna is that it is standardized using a KCl standard per NIST, and I can easily imagine an ai misunderstands and thinks it is KCl.

 

[DavidA]

Just looking for clarity, where do you see it is KCl? All I see from Hanna is that it is standardized using a KCl standard per NIST, and I can easily imagine an ai misunderstands and thinks it is KCl.

Randy, you are correct. AI could be wrong and I could not find another reference to their solution. That said, the observation still stands that the checker did a good job of automatically adjusting SG when I used seawater over a range of temps and did a poor job when I tested the calibration fluid sachet over a range of temps.

 

[DavidA]

Randy, you are correct. AI could be wrong and I could not find another reference to their solution. That said, the observation still stands that the checker did a good job of automatically adjusting SG when I used seawater over a range of temps and did a poor job when I tested the calibration fluid sachet over a range of temps.

Randy, I just emailed the following to Hanna tech support. If and when they respond, I will update the thread.

Please see the thread on Reef2reef about calibrating the salinity monitor.
1. Is the calibration potassium based or saline based?
2. Does the temp adjustment algorithm work for the calibration fluid or must the calibration be performed at tank temp.
3. https://www.reef2reef.com/threads/salinity-question.1135181/

 

[Randy Holmes-Farley]

Thanks. :)

 

[EnterName]

I too am surprised that Hanna would design it with KCl standard and then have to use a different temp correction during calibration than in measurement.

I think most companies make a seawater mimic (perhaps leaving out alk to make it more stable) and then can just use one correction. American Marine/Pinpoint uses a seawater mimic standard.

I'm pretty sure most companies are using the practical salinity scale (PSS-78) for their conversions. PSS-78 is based on a KCl reference (32.4356g/kg KCl), so using a KCl calibration solution might make things easier math-wise. At 25°C/77°F this KCl reference has a conductivity of approx 53mS/cm just like your DIY standard, so if they are indeed calibrating with 53mS/cm (like many probes do), your DIY solution should work great with all those devices as long as it is used at 25°C/77°F.

Just looking for clarity, where do you see it is KCl? All I see from Hanna is that it is standardized using a KCl standard per NIST, and I can easily imagine an ai misunderstands and thinks it is KCl.

Unfortunately they aren't specifying what they are using. I looked at the SDS of HI70024P, but it doesn't specify the contents. However, it says the pH is at 5.7 (quite low for a KCl solution), so you are probably right that they are using some artificial salt mix without alkalinity. I don't know how precise the temperature correction is for this salt mix, so it's probably best to calibrate at 25°C/77°F when using their standard as well. This might explain why so many people are complaining about the readings being a bit off.

 

[vetteguy53081]

While I swear by Hanna readings, TM is supposed to be the most accurate and its a matter of temperature as mentioned and calibrating it using that 77 deg water

 

[Randy Holmes-Farley]

I’ll just repeat that it may be foolproof, and is plenty accurate for our purposes, but the TM hydrometer is not the most accurate tool for measuring salinity, IMO.

The fact that the modern salinity scale is based on conductivity and not density puts that hobby mantra in perspective.

 

[EnterName]

While I swear by Hanna readings, TM is supposed to be the most accurate and its a matter of temperature as mentioned and calibrating it using that 77 deg water

Reproducible results are key for the feeling of accuracy.

A few drops on a refractometer will change their temperature fast and Automatic Temperature Compensation can only do so much.

Conductivity meters need proper care and calibration. Different calibration solutions at different temperatures might be a big contributor to erroneous readings as is discussed in this thread.

This pretty much only leaves hydrometers which don't really need calibration and use a larger sample which has a more stable temperature than a few drops. The only sources of error are then improper use, wrong temperature readings, and atmospheric pressure (which is probably negligible in most cases). In that sense, I get why the community likes a decent hydrometer best, even though modern salinity is based on conductivity.

 

[Paul B]

A difference of .002 seems significant. What can explain that each device reads its benchmark DIY fluid perfectly while there is a .002 discrepancy when sampling the same water?

That is not a significant difference, and I would not worry about it at all.

I don't know why your two testing methods are slightly different but it doesn't matter. Seawater varies widely all over the planet as it does in my reef which is very old. I use a 30 year old swing arm hydrometer and I had to bring it to the tropics, take it underwater and draw on it where the arm floated.

Not very scientific but I never lost a fish or coral from salinity.

 

[Patriot Reefer]

You have to account for this. A hydrometer or optical refractometer is density based, a conductivity tester (Hanna) is not. A hydrometer or refractometer could be influenced by dissolved organics or other forms of matter that have nothing to do with salinity. This is why I trust Hanna. Also this explains why Hanna is consistently lower by a few points, density factors increase refractometers a point or two, and my Hanna is closer to ICP.
Example, my protein skimmer was broken and shut off for a week. My Milwaukee read 1.028, Hanna 1.026. I mixed up new saltwater (free from dissolved organics), Both Milwaukee and Hanna read 1.026 for the new water. They didn't agree with display water, but did agree with new mixed water. So I think the Hanna was correct with the display water in this case. The true salinity was probably 1.026, and dissolved organics gave Milwaukee a higher reading. This is just 1 of many examples showing the difference between different methods of testing. @Randy Holmes-Farley , curious on your take with this.

One silly example, if you put sugar in water, a refractometer will spike its results, conductivity will not. Classic example.

 

[Patriot Reefer]

Did you allow for temp issues when using the hydrometer, especially with the standard?

What temp was the Hanna diy standard? Maybe it does not temp correct well.

You have to account for this I think. Randy, let me know if you agree. A hydrometer or optical refractometer is density based, a conductivity tester (Hanna) is not. A hydrometer or refractometer could be influenced by dissolved organics or other forms of matter that have nothing to do with salinity. This is why I trust Hanna. Also this explains why Hanna is consistently lower by a few points, density factors increase refractometers a point or two, and my Hanna is closer to ICP.
Example, my protein skimmer was broken and shut off for a week. My Milwaukee read 1.028, Hanna 1.026. I mixed up new saltwater (free from dissolved organics), Both Milwaukee and Hanna read 1.026 for the new water. They didn't agree with display water, but did agree with new mixed water. So I think the Hanna was correct with the display water in this case. The true salinity was probably 1.026, and dissolved organics gave Milwaukee a higher reading. This is just 1 of many examples showing the difference between different methods of testing. @Randy Holmes-Farley , curious on your take with this.

One silly example, if you put sugar in water, a refractometer will spike its results, conductivity will not. Classic example.

 

[Randy Holmes-Farley]

You have to account for this I think. Randy, let me know if you agree. A hydrometer or optical refractometer is density based, a conductivity tester (Hanna) is not. A hydrometer or refractometer could be influenced by dissolved organics or other forms of matter that have nothing to do with salinity. This is why I trust Hanna. Also this explains why Hanna is consistently lower by a few points, density factors increase refractometers a point or two, and my Hanna is closer to ICP.
Example, my protein skimmer was broken and shut off for a week. My Milwaukee read 1.028, Hanna 1.026. I mixed up new saltwater (free from dissolved organics), Both Milwaukee and Hanna read 1.026 for the new water. They didn't agree with display water, but did agree with new mixed water. So I think the Hanna was correct with the display water in this case. The true salinity was probably 1.026, and dissolved organics gave Milwaukee a higher reading. This is just 1 of many examples showing the difference between different methods of testing. @Randy Holmes-Farley , curious on your take with this.

One silly example, if you put sugar in water, a refractometer will spike its results, conductivity will not. Classic example.

While some of that is true, some is incorrect, or at least misleading.

Refractometers work off refractive index, which does not necessarily track with density.

All of these methods will have some tiny response to the tiny amount of organics present relative to salts. Different organics can actually drive the reported values up or down, depending on which organic and which method, but again, its a miniscule amount and effect.

I do not think organics are playing a role in the explanation of folks results. :)

 

[addictswife]

I got tired of all the variations.
The floating hydrometer from bulk reef supply works every time. Very consistent. I just use a tall beaker and dip a little water from my tank or mixing tank, so I don’t have to put it in my tank.