Measuring salinity - a tale of woe.

[onefang]

I am having some conflicting results between my swing-arm and what seems to be a less than reliable refractometer, and I'd like to get some input. Prepare yourself for a ridiculous wall of text..

 

I recently received a refractometer that I could not get to stay calibrated. I went rounds with it, but it would never give me the same reading twice. While I was waiting for my replacement(a BRS unit), I broke out the old swing-arm deep six for all my salinity needs. It was giving me super consistent results, so long as I was diligent about bubbles and keeping it on a perfectly level surface.

The new refractometer arrived, and so I calibrated it to 35 ppt with a standard made by "Aqua Craft Products", test, test, tested again. Ok, so now the new one is giving me more consistent results. I discovered that it is however extremely finicky about temperature.

 

My typical process for testing salinity is to test at least two times in a row. Test, reset test again, if all is good I accept the results.

Generally this goes:

Calibrate to standard solution, after set clean slide with RO, dry thoroughly.

Test 1 - take sample from tank. Swish back and forth a few times in the pipette to make sure there isn't any residual RO from initial rinse. Put on slide, let sit for about 45 seconds for temp to stabilize, take reading.

Test 2 - Rinse slide and slide cover under RO, dry thoroughly. Take sample from tank - Swish back and forth a few times in the pipette. Put on slide, let sit for about 45 seconds for temp to stabilize, take reading.

 

I figured out that while rinsing the slide, if the RO is cold it will throw the reading off if I don't let the refractometer warm back up to the temp I originally calibrated it at before taking the next reading. Also, If my room temp fluctuates buy 3-4 degrees, it can be off by 2 ppt when I re-calibrate it the next time I use it. Apparently there is a bi-metallic strip inside that moves the optics slightly and this is what compensates for temp differences in samples. If you cool the housing, it messes with the temp compensation quite a bit. But I digress..

A few days go by, and I break out the new refractometer for a water change and again calibrate(because I am super paranoid now). Now the damn thing is reading 2 points low on the ppt scale with the calibration solution(probably due to ambient temp differences). So I set it back to 35 ppt, make my water, do my testing. I tested with both devices and the refractometer is reading 35, the swingarm is reading 36. I am super confused at this point. I need a third party to test without me involved, so I run a sample over to the LFS. They calibrate their Red Sea refractometer with a two little fishies product called "accurasea" and test my sample at 36 ppt.

OK WTF!

At this point I figure that my standard solution may actually be "non-standard" so to speak. I dump the bottle into my swing-arm, and it reads out at 36. Ok.. so maybe the solution is off, or my swing-arm reads exactly 1 ppt higher.

 

This leads me to thinking that it's possible my swingarm(assuming it was calibrated to a salt only solution, being cheapo) is reading it a degree high because it takes density into account and there is a buttload of other stuff besides salt in NSW, and perhaps that is what the problem is.

So I bust out my test kits on the reference solution. It reads:

Cal: 60 ppm

Mg: <1 ppm

KH: ~1 dKH

 

hmmm.. that's interesting. This solution is definitely NOT NSW.

I figure that I can make up some salt only standards with guidance from an old reefkeeping.com article, and a really precise scale that I used to use for EI dosing my planted tanks (accurate down to .001 of a gram). I used plain table salt and RO to make up a 1.3394 refractive index solution. When I test this solution, I get 35 ppt dead on with my swing-arm. When the refractometer is calibrated to the pre-mix my DIY standard reads 33 ppt.

 

So, I went ahead and calibrated the refractometer with my DIY solution, and did a test of my tank water - they are both reading dead on at 35 ppt.

 

So I guess now my question is this - is using a salt only solution going to get me good accuracy, or will I need to find a new standard solution(preferably one with all the constituents of NSW) to test against?

I was initially thinking that the extra cal, alk, and mag in the RSCP salt mix would increase its density, thereby causing the hydrometer to read slightly higher than the refractometer - but my DIY solution seems to have proven that this isn't the case, as the tank water tested the same even when the refractometer was calibrated with a pure salt standard.

 

Help me, my brain hurts.

[tetraodon]

well the red sea refractometers are saltwater testers while most refractometers are saline testers, there in of itself is 1ppt difference you talk about, i know the meter i have i calibrate with RO water not testing solutions. but most have different directions. most meters now have auto temp compensation, so as long as the temp is within range its fine. do you have to calibrate before every test? try calibrating with RO water and see just for the fun of it what happens. just for fun, do you test when your water level is the same at all the tests? water evap can affect test results, mostly cause only H2O is evaped nothing else. so the saltwater tends to concentrate.

[tetraodon]

the biggest thing i see your compairing results from density to refractive index, which are two different ways of doing it. the biggest thing with calibration is do it consistantly, if your DIY solution tests and calibrates the best, use that. remember we are also dealing with parts per thousands, which is very small, and i can say for sure that one or two PPT off is a very big deal as long as its consistent. for me as long as my reading is 34-36 or so when my water is at my full line, im ok. this is my meter: http://nisupply.ecrater.com/p/14129551/3699-true-natural-marine-seawater-salinity

anyway i hope this helps.

[CronicReefer]

I'm pretty sure you have the same one as me (the BRS refractometer is rebranded by many different companies). The directions state that the instruments temperature is what is important when properly calibrating for first use and you must use RO water for calibration (I noticed on the BRS site it says not to use RO/DI). Also depending on what temp you calibrated it at you will get different S.G. readings for the correct 35ppt salinity value. I calibrated mine at 25C so for 35ppt that equals 1.0264 S.G. but if the instrument is colder, like 20C, then S.G. will change to 1.0277 for 35ppt salinity (based on the 25C calibration). Swing arms are calibrated in a similar way so you need to find out what temperature your swing arm was calibrated at and then you will need to make sure the water is that temperature as well to have an accurate reading.

 

Here is a link to the directions that came with mine (not sure if yours are the same). http://www.theaquariumsolution.com/files/SEAWATER%20REFRACTOMETER%20V2.pdf

[onefang]

well the red sea refractometers are saltwater testers while most refractometers are saline testers, there in of itself is 1ppt difference you talk about, i know the meter i have i calibrate with RO water not testing solutions. but most have different directions. most meters now have auto temp compensation, so as long as the temp is within range its fine. do you have to calibrate before every test? try calibrating with RO water and see just for the fun of it what happens. just for fun, do you test when your water level is the same at all the tests? water evap can affect test results, mostly cause only H2O is evaped nothing else. so the saltwater tends to concentrate.

 

I definitely make sure the tank is topped off frequently(I have a topoff line on the side of the tank to ensure it's always filled to the exact same level). So as far as my tank water, I am fairly certain I am not seeing the actual salinity move around. At this point I have to calibrate it after every test or I will get a totally nonsensical reading, like 2PPT lower that previous test. Even when the unit is calibrated in the morning when my living room is spot on at 68F(20C), if I check the calibration again the next day at the same temp, sometimes it is off.

 

the biggest thing i see your compairing results from density to refractive index, which are two different ways of doing it. the biggest thing with calibration is do it consistantly, if your DIY solution tests and calibrates the best, use that. remember we are also dealing with parts per thousands, which is very small, and i can say for sure that one or two PPT off is a very big deal as long as its consistent. for me as long as my reading is 34-36 or so when my water is at my full line, im ok. this is my meter: http://nisupply.ecrater.com/p/14129551/3699-true-natural-marine-seawater-salinity

anyway i hope this helps.

This is one of the main things I was concerned about. You'd think that when a swing-arm hydrometer and a refractometer are both reading the same result of 35 ppt with a purely saline standard, that a tank water test with water made from RSCP would be different.

Mostly due to the fact that water made with RSCP would be significantly more dense due to the other constituents in the water; However, this is not the case, as both the hydrometer and refractometer read the same with tank water. It's really freaking out my brain.

As far as the degree of error bing 1-2 PPT, I agree totally. I was thinking that maybe I should just pick the thing that is giving me good consistent results, and run with that. Worst case at this point, my tank runs at 1.024-1.025 instead of the 1.026 that I was shooting for. I'd much rather keep things as consistent as possible vs. trying to hit a particular perfect number.

 

Thanks for that link too.. It has some interesting information:

"Once correctly calibrated at the set calibration temperature of 20’C/68’F the Refractometer can then be used in environments where the ambient temperature and therefore instrument temperature would heat up or cool down within the range of the ATC which is between 10-30’C/50-86’F. "

 

Also:

"PLEASE DO NOT CALIBRATE WITH 35PPT SALINITY SOLUTIONS AS THEY ARE NOT THE SAME AS NATURAL SEA WATER! THAT SOLUTION IS FOR THE BRINE SALINITY REFRACTOMETERS."

 

That is the most informative description I've ever seen on a product!

 

I'm pretty sure you have the same one as me (the BRS refractometer is rebranded by many different companies). The directions state that the instruments temperature is what is important when properly calibrating for first use and you must use RO water for calibration (I noticed on the BRS site it says not to use RO/DI). Also depending on what temp you calibrated it at you will get different S.G. readings for the correct 35ppt salinity value. I calibrated mine at 25C so for 35ppt that equals 1.0264 S.G. but if the instrument is colder, like 20C, then S.G. will change to 1.0277 for 35ppt salinity (based on the 25C calibration). Swing arms are calibrated in a similar way so you need to find out what temperature your swing arm was calibrated at and then you will need to make sure the water is that temperature as well to have an accurate reading.

 

Here is a link to the directions that came with mine (not sure if yours are the same). http://www.theaquariumsolution.com/files/SEAWATER%20REFRACTOMETER%20V2.pdf

 

Looks like these are the same instructions as the link that tetraodon provided. They are not quite the same as what I got with my BRS unit, and actually the instructions from BRS are different than the instruction sheet given by the manufacturer of the refractometer.

BRS placed a sticker on the outside stating specifically NOT to calibrate with distilled water, whereas the instructions from the manufacturer clearly state that it should be calibrated with distilled water.

 

 

From the instructions included in the box:

 

 

 

Also, for what it's worth, the calibration solution that seems to be throwing things off

 

My swing-arm hydrometer doesn't seem to be marked externally for what temp it was calibrated out - but I've read that they are supposed to be able to compensate for temp differences in the water. I might have to perform an experiment where I chill some testing water and see if it compensates.

[CJJon]

First thing I would do is throw out the swing arm. After that, just use your DIY calibration solution and call it good. Stability is the key, and what you are reading is probably close to reality, but only close. In fact, I have read these particular refractometers all read 1 point higher than reality. As long as you are in that 34-36 range and you keep it there, you will be fine.Pick a number and be consistent with when and how you take your readings. I always do mine the same aiming for 35. I grab the refract, flip the cover, squirt water, flip cover down and read. Just that quick. The room temp is stable and so is the water. If I do it that way each time, I will be consistent with my readings.

 

Although I am so precise with measuring out my salt mix (ESV) and my ATO is very sensitive, I rarely have salinity readings that surprise me. The only thing that causes a change in salinity is skimmer output. I have a jug of salty water next to the tank that I add to make up for it. I do test all my water though!

[onefang]

First thing I would do is throw out the swing arm. After that, just use your DIY calibration solution and call it good. Stability is the key, and what you are reading is probably close to reality, but only close. In fact, I have read these particular refractometers all read 1 point higher than reality. As long as you are in that 34-36 range and you keep it there, you will be fine.Pick a number and be consistent with when and how you take your readings. I always do mine the same aiming for 35. I grab the refract, flip the cover, squirt water, flip cover down and read. Just that quick. The room temp is stable and so is the water. If I do it that way each time, I will be consistent with my readings.

 

Although I am so precise with measuring out my salt mix (ESV) and my ATO is very sensitive, I rarely have salinity readings that surprise me. The only thing that causes a change in salinity is skimmer output. I have a jug of salty water next to the tank that I add to make up for it. I do test all my water though!

 

Yeah, think I am probably being overly neurotic. My personality is such that I have a hard time coping with inconsistency, so I think picking one(the refractometer as you say) and keeping things dialed into one calibrated instrument is better for the tank.

It just pisses me off when I have conflicting instructions for the same instrument, and then the readings disagree with both my swing-arm and the LFS's calibrated refractometer. I want exact, but I guess that's a little much to expect from a hobbyist standpoint

[tetraodon]

so it seams like the meter may not be holding calibration well. what i would do is do two tests, one calibrate with RO water, test tank water, record results. do that for a few days on one calibraition. then calibrate with your DIY stuff, test tank water, record results. repeat test a few days with one calibration. this will allow yo to see if your unit stays calibrated or not, as well as any difference in the two calibration solutions.

[onefang]

I think what I am going to do is take my refractometer and calibration solution to the LFS and have them check both against their refractometer / calibration solution. I will use that as a control. That's about the best I can do at this point in determining if my refractometer is acting funky, or if it really is my calibration solution that is giving me a bad reading. Then I might do as you say tetraodon, and calibrate it to zero, do testing for a couple days. Calibrate with DIY, test for a couple more.

Along side those tests, I will take a reading with the swing-arm just to see if I get any fluctuation there. I think this will rule out my salinity creeping up or down due to evap/top-off.

[CronicReefer]

I think what I am going to do is take my refractometer and calibration solution to the LFS and have them check both against their refractometer / calibration solution. I will use that as a control. That's about the best I can do at this point in determining if my refractometer is acting funky, or if it really is my calibration solution that is giving me a bad reading. Then I might do as you say tetraodon, and calibrate it to zero, do testing for a couple days. Calibrate with DIY, test for a couple more.

Along side those tests, I will take a reading with the swing-arm just to see if I get any fluctuation there. I think this will rule out my salinity creeping up or down due to evap/top-off.

The calibration fluid is also temperature based to get an accurate reading but is fine to use. I find it strange that mine came with the directions it did (I'm pretty sure its the wrong directions) as it is advertised as having a scale from 0-40 but mine is really on a scale from 0-100 like yours. Apparently there are three versions of this refractometer one that has a 0-100 scale (Deepwater ATC and BRS ATC), one that has a 0-40 scale (Red Sea ATC), one that has a 20-40 scale (D-D ATC). They are rebranded under a few different names but they all look exactly identical to me which makes me wonder if there really is any difference in accuracy or is it just a difference in the way the scale is displayed.

[onefang]

Ok.. so I think I figured it out. I took my refractometer, the premixed standard sent to me by BRS, and my DIY standard over to my LFS.

 

They calibrated two different refractometers there. One like mine, and their Red Sea. They tested the two solutions and found the following on both ( yay.. finally something agrees!)

 

BRS provided standard: 36 ppt

DIY standard: 35 ppt

 

I calibrated my refractometer with their standard and got the exact same results. I am pretty stoked to be able to make a DIY standard that matches(or in this case exceeds the supplied) commercial products.

 

Interestingly enough, when I test RO water on it after calibrating with the LFS solution, it reads exactly zero. Sooo.. I guess the manufacturer recommended process is correct for this particular unit, and the standard that I have is not actually 35 ppt.

 

I will retest tomorrow and see if it holds calibration.