Most accurate way to measure salinity?

[Michael43]

Which refractometer? Whether cheap or a bit pricey. My apex and my atc refractometer both are very different. Even after just calibrating. I want to get s third option. Milwaukee? Veegee? Others?

 

[Randy Holmes-Farley]

When you ask "most accurate", all of the basic methods themselves (conductivity, refractive index, density) can be much more accurate than a reefer needs or uses, so the implementation by any particular device is the question.

What results are you getting?

How did you calibrate the refractometer and the conductivity? Temp is very important for the conductivity so the temp probe must be in the calibration solution, or at least at exactly the same temp.

 

[AllSignsPointToFish]

Is it more important to be accurate or precise? I would speculate that, within reason, it's more important to be precise so that salinity stays stable. :)

 

[Randy Holmes-Farley]

Is it more important to be accurate or precise? I would speculate that, within reason, it's more important to be precise so that salinity stays stable. :)

FWIW, precise (and precision) does not necessarily mean reproducible. A device that reads specific gravity to be 1.02532 one time and 1.02731 the next time on the same water is very precise, but not reproducible. :)

 

[AllSignsPointToFish]

FWIW, precise (and precision) does not necessarily mean reproducible. A device that reads specific gravity to be 1.02532 one time and 1.02731 the next time on the same water is very precise, but not reproducible. :)

True, but I think the point still stands. As long as a systematic bias is consistent, one can use the device without ever actually knowing the true value of an unknown. It's similar to enthalpy in thermodynamics...you never really know the absolute value, but rather you know your offset from a given point ;)

 

[Michael43]

Sorry it was a poorly asked question. I'm trying to keep my salinity at a stable 1.0245-1.025. Obviously I convert from ppm. I use 35 ppm solution to calibrate my atc refractometer and used the instructional steps for the apex probe. My refractometer is reading about .023 where my apex is reading about .021 or so. I have an lps tank but have started to add my sps so I want to slightly start to elevate it slowly. Everything is doing very well for the most part so I'm hesitant to touch it at all but I feel I should bring it up slightly. I'd like to get a 3rd and possibly more precise tool to measure so I can do an average or know better where I'm currently at. Thanks guys

 

[AllSignsPointToFish]

Sorry it was a poorly asked question. I'm trying to keep my salinity at a stable 1.0245-1.025. Obviously I convert from ppm. I use 35 ppm solution to calibrate my atc refractometer and used the instructional steps for the apex probe. My refractometer is reading about .023 where my apex is reading about .021 or so. I have an lps tank but have started to add my sps so I want to slightly start to elevate it slowly. Everything is doing very well for the most part so I'm hesitant to touch it at all but I feel I should bring it up slightly. I'd like to get a 3rd and possibly more precise tool to measure so I can do an average or know better where I'm currently at. Thanks guys

Reminds me of the old saying, "A man with one watch knows the time...a man with two watches is never quite sure." ;)

 

[AllSignsPointToFish]

Sorry it was a poorly asked question. I'm trying to keep my salinity at a stable 1.0245-1.025. Obviously I convert from ppm. I use 35 ppm solution to calibrate my atc refractometer and used the instructional steps for the apex probe. My refractometer is reading about .023 where my apex is reading about .021 or so. I have an lps tank but have started to add my sps so I want to slightly start to elevate it slowly. Everything is doing very well for the most part so I'm hesitant to touch it at all but I feel I should bring it up slightly. I'd like to get a 3rd and possibly more precise tool to measure so I can do an average or know better where I'm currently at. Thanks guys

See, my Milwaukee refractometer reads distilled water exactly down to the third decimal place every time. However, it reads 1.024 on a standard 1.025 solution every single time. As long as those two points don't shift, I'm not really worried about it. I can test the same sample from the display tank over and over again, and the reading stays the same. So, you see, I'm not too worried about the actual value as long as I know my measurement offset is constant with respect to two reference points.

 

[Randy Holmes-Farley]

True, but I think the point still stands. As long as a systematic bias is consistent, one can use the device without ever actually knowing the true value of an unknown.

Yes, that is probably true. Our tanks are remarkably forgiving about the exact salinity. :)

 

[Randy Holmes-Farley]

Sorry it was a poorly asked question. I'm trying to keep my salinity at a stable 1.0245-1.025. Obviously I convert from ppm. I use 35 ppm solution to calibrate my atc refractometer and used the instructional steps for the apex probe. My refractometer is reading about .023 where my apex is reading about .021 or so. I have an lps tank but have started to add my sps so I want to slightly start to elevate it slowly. Everything is doing very well for the most part so I'm hesitant to touch it at all but I feel I should bring it up slightly. I'd like to get a 3rd and possibly more precise tool to measure so I can do an average or know better where I'm currently at. Thanks guys

OK, I just wanted to be sure what you were using, because not all solutions are suitable for all devices. My DIY 35 ppt equivalent solution for a refractometer has a different amount of sodium chloride in it than the 35 ppt equivalent solution for a conductivity probe. :)

IMO, a better way to answer this conundrum is to be sure they can read the same calibration solution and get the same value, than to roll the dice that a third method will give a decisive answer.

 

[fab]

See, my Milwaukee refractometer reads distilled water exactly down to the third decimal place every time. However, it reads 1.024 on a standard 1.025 solution every single time. As long as those two points don't shift....

AllSignsPointToFIsh,

Doesn't your Milwaukee refractometer have a 'zeroing screw' that enables you to zero the known error right at your desired operating point?

If you are smack on for distilled water and just a slosh off at 1.025, then the error seems to be not a true level bias. It could, instead, be a small slope bias if the refraction is linear with salinity.

I have always presumed this linear relationship simply because I have really never thought it through. If the error is a level bias, then I suppose that adjusting the screw should change the distilled water reading as you zero out the error at 1.025. If the screw provides a slope correction possibly the distilled water reading will continue to be correct.

If someone knows how theses correcting adjustments actually work please chime in and edumicate us.

 

[Randy Holmes-Farley]

​

AllSignsPointToFIsh,

Doesn't your Milwaukee refractometer have a 'zeroing screw' that enables you to zero the known error right at your desired operating point?

If you are smack on for distilled water and just a slosh off at 1.025, then the error seems to be not a true level bias. It could, instead, be a small slope bias if the refraction is linear with salinity.

I have always presumed this linear relationship simply because I have really never thought it through. If the error is a level bias, then I suppose that adjusting the screw should change the distilled water reading as you zero out the error at 1.025. If the screw provides a slope correction possibly the distilled water reading will continue to be correct.

If someone knows how theses correcting adjustments actually work please chime in and edumicate us.

Conductivity measurements never have a zero point control that I have seen (and they generally should not). They only have slope adjustment. There is truly zero conductance at zero conductivity, so it is a fixed point.

 

[AllSignsPointToFish]

​

AllSignsPointToFIsh,

Doesn't your Milwaukee refractometer have a 'zeroing screw' that enables you to zero the known error right at your desired operating point?

If you are smack on for distilled water and just a slosh off at 1.025, then the error seems to be not a true level bias. It could, instead, be a small slope bias if the refraction is linear with salinity.

I have always presumed this linear relationship simply because I have really never thought it through. If the error is a level bias, then I suppose that adjusting the screw should change the distilled water reading as you zero out the error at 1.025. If the screw provides a slope correction possibly the distilled water reading will continue to be correct.

If someone knows how theses correcting adjustments actually work please chime in and edumicate us.

IDK, it might have. None of the included information indicates such an adjustment (maybe I'm missing some information). It hasn't really bothered me as my results are consistent and tank inhabitants are happy. :)

 

[fab]

Conductivity measurements ... only have slope adjustment. There is truly zero conductance at zero conductivity, so it is a fixed point.

Thanks Randy. What about refractometer?

Are they both, conductivity sensor and refractometer readings, actually "straight line" linear relationships to salinity?

 

[Randy Holmes-Farley]

Thanks Randy. What about refractometer?

Are they both, conductivity sensor and refractometer readings, actually "straight line" linear relationships to salinity?

Normal mechanical refractometers that most people use have no slope adjustment, only the zero point. Unfortunately, many sold have the wrong slope (that of sodium chloride solutions rather than seawater) so they are inherently incorrect when measuring seawater if calibrated in pure fresh water. Hence the need to calibrate with a seawater standard. A true seawater refractometer does not have that issue, but can still be calibrated in a seawater standard.

Conductivity is not exactly straight with salinity. For that reason, more complex equations are needed than a simple slope factor are needed if you want to convert a measured conductivity (in mS/cm) to salinity (in, say, ppt).

Here's a discussion of that conversion:

http://www.code10.info/index.php?op...d=54:cat_coding_algorithms_seawater&Itemid=79

 

[zoomonster]

Your problem is the Apex probe. Been there done that. Always much lower than refractometer and I even threw in a hydrometer for comparison. I've seen some people say they have issues calibrating the apex but mine was only done once when I got it and it now matches my refractometer. It took a week or more of back and forth with Neptune before they finally admitted the probe has issues with micro bubbles. They accumulate under/on the probe and seriously throw readings off. I now keep mine in my refugium behind a wall of chaeto that protects it from any micro bubbles that may come from skimmer compartment. Another guy here having the same issues dealt with it by mounting the probe upside down. I don't recommend that though because I have seen Neptune say the probe should be in the water just enough for a reading or it will shorten the life of the probe. Another alternative is to put a small foam block around the probe to protect it from bubbles. Neptune's original salinity probe, in the aquacontroller days, used to include/use that.

In any case as I have said before when you get it right it's great for monitoring but I would not EVER rely on it for any salinity based control.

And BTW I also use an ATC refractometer calibrated with Aqua Craft calibration solution

 

[fab]

Microbubbles have been well discussed in Neptune Systems' forums. I've never seen the company hold back on the fact that microbubbles can throw readings off. In fact the company has always been forthcoming about various vulnerabilities of different probes to different interferences.

The different underlying sensing methods enumerated by Randy
https://www.reef2reef.com/threads/most-accurate-way-to-measure-salinity.247202/#post-2902849
And discussed in the next few posts following that post each have their accuracy, precision and repeatability vagaries that we must learn about to manage our expectations of them.

 

[jason2459]

My apex salinity probe has been spot on once I found a good method to calibrate it and location to place it in. It's been very reliable since then. I have since started using it to adjust for salinity drift from my skimmer that I now skim wetter with.

I'll have to find the post I describe how I calibrated it.

That said I trust my Veegee/Vitalsine refractometer more then any other testing equipment I have.

Edit: I also like my pintpoint salinity probe. I always forget about it as I use it in my saltmix container and only look at it when mixing in new salt and briefly as I have it down now to where my float valve for ro/di and switch to alert me it's full so I can turn off the ro/di is set so I can just dump in a bag and be at 35ppt.

 

[jason2459]

Here's that post

FWIW, when I first setup my Apex salinity probe I was getting a good 3ppt swing through out the day and sometimes even up to 5ppt. I just used it as a basic wide range failsafe for my ATO and AWC.

I fixed it by rerouting the cable to the PM2 and the probe, angled the probe away from the slight flow it was getting (no bubbles in the area as well), and recalibrated again with the Neptune calibration packet. I do have a temp probe plugged right into the PM2 as well.

The calibration I did completely different from the first time doing it. Before starting I took the salinity probe out and soaked it in some water/vinegar for a while (~30 minutes), then rinsed very very well with RO/DI water, wiped it down and shook it out a bit to displace any water, and let it sit for about an hour drying out. I placed the 53.0 mS calibration packet in my sump and let it sit in there that entire time adjusting to the tank water temp it's normally at. I also made sure to set the temp Conductivity Compensation in the input setup to 2.2 and range set to salinity before starting the calibration.

Then I started the calibration selecting the 53000. Wait 10 minutes while it's adjusting to the Dry probe. Then I continue. I clipped the corner of the packet to be able to shove the probe in and kept holding it in the water (I did not use the temp probe in the packet just the temp probe was next to it in the sump). I gently swirled it around periodically for around 5 minutes. And then just stood there for another 5 minutes. Making sure to hold the packet closed well around the probe and in the sump water. Then ended the calibration.

My apex probe has been pretty solid since then. It varies now with in 1ppt and I was able to tighten up my failsafes and trust it a bit more. I also shake it a bit periodically in the sump. About once each day or every other day.


Or your probe is bad.

 

[zoomonster]

Microbubbles have been well discussed in Neptune Systems' forums. I've never seen the company hold back on the fact that microbubbles can throw readings off. In fact the company has always been forthcoming about various vulnerabilities of different probes to different interferences.

Perhaps they do now. My problem with the salinity probe was about 2 years ago and like I said took significant back and forth before they finally gave me that answer.