Measuring Salinity

[siggy]

Here is something to try, This weekends WC I rounded up my refractometers and milwaukee digital to calibrate and find any margin of error. My first is a Liveaquaria unit, It didn't come with cal solution or a cap for the cal screw and that has since rusted so cal is not an option. I used it for 10 months without a problem, purchased a packet of cal solution (@ 4 months) at LFS and all was well and corals were healthy and (SPS) growing. During the holidays I picked up a Milwaukee digital, it was defective and took a bit to to receive a replacement and during that time I purchased a BRS unit primarily because there's comes with solution and my newly installed apex unit was giving me some crazy high readings. I mentioned that I have to offset my digital by +1, well this week it's on the money using the provided 35 checking sol. when i checked the BRS sol. it read 34.
( the milwaukee calibrates with steam distilled water SDW provided and uses a checking solution @ 35 sg )
I calibrated the brs unit to 35 using their sol. and then checked zero using SDW and result was way below zero on scale, I then checked it using milwaukee 35 sol and result was 39.
Not sure what to make of this other than Use only cal. solution that was designed for that unit??? and stability at any close SG wins the race?

Note: the @BRS Refractometer is a noticeably better or nicer unit, it comes with detailed color directions, a Very Nice Large cleaning cloth and even the protective foam in the case is more better than my other and it has the cap.

 

[Randy Holmes-Farley Verified Community Expert]

Not sure what to make of this other than Use only cal. solution that was designed for that unit??? and stability at any close SG wins the race?

Unless they make the solution incorrectly, all 35 ppt seawater equivalent refractometer calibration solutions should show the same salinity.

RO/DI water will not (should not) correctly read 0 ppt on a brine refractometer (the BRS refractometer may be this type) when properly calibrated with 35 ppt standard. That's the reason to use the 35 ppt standard, and not 0 ppt RO/DI for calibration.

FWIW, if a calibration solution is intended for a different TYPE of device, like a conductivity meter, then it may (or may not) be inaccurate on a refractometer.

 

[Reef Jeff]

I use a 15 year old Sybon refractometer. I check it annually and it has never required an adjustment so I voted never. The fact it never required adjustment made me suspicious so I compared with a hydrometer and a conductivity meter and it’s still correct.

 

[siggy]

I understand the SDW not reading zero on the BRS unit, thats calibrated @35 (set-point) and the span is too great to accurately read at the zero end. To have a have an accurate span of +/-1 from 0-35 sg from a $50 hand held unit is asking a lot. What I found troubling is the 4 point difference. @randyBRS Could you try this at the shop? ( up two posts )

Maybee the Glass bobber is not so Eccentric ...and as always Thank You VM

 

[Randy Holmes-Farley Verified Community Expert]

I understand the SDW not reading zero on the BRS unit, thats calibrated @35 (set-point) and the span is too great to accurately read at the zero end. To have a have an accurate span of +/-1 from 0-35 sg from a $50 hand held unit is asking a lot. What I found troubling is the 4 point difference. @randyBRS Could you try this at the shop? ( up two posts )

Maybee the Glass bobber is not so Eccentric ...and as always Thank You VM

It is not just the difficulty of having a perfect device (although that is an issue), but that such a device, even if perfectly made and calibrated, necessarily does not give a correct reading at 0 ppt. They are off by a well known and understood value (which is below 4 ppt).

The reason is that it is not made for seawater, but for sodium chloride. In that case, it CANNOT read both 0 and 35 ppt correctly.

I discuss these issues here:

Refractometers And Salinity Measurement
http://www.reefedition.com/refractometers-salinity-measurement/

from it:

" If you have this type of refractometer, and it was perfectly made and calibrated in freshwater, it will ALWAYS read seawater to be higher in salinity than it actually is (misreporting an actual 33.3 ppt to be 35 ppt)."

 

[siggy]

It is not just the difficulty of having a perfect device (although that is an issue), but that such a device, even if perfectly made and calibrated, necessarily does not give a correct reading at 0 ppt. They are off by a well known and understood value (which is below 4 ppt).

That explains the discrepancy with the brs unit reading below zero after calibrating. Im I wrong in thinking that most refractometers have an inherent slope error due to cost of manufacture and intentionally manufacture them to produce a result in the targeted range when calibrated to THEIR recommended procedure , pure water or SW standard

 

[saltyfilmfolks]

Every time.
I have a no name refractometer I bought as a newb a long time ago and salifert fluid. Quite inexpensive stuff.

My salinity was only a bit off my Ati icp test.

 

[chipmunkofdoom2]

That explains the discrepancy with the brs unit reading below zero after calibrating. Im I wrong in thinking that most refractometers have an inherent slope error due to cost of manufacture and intentionally manufacture them to produce a result in the targeted range when calibrated to THEIR recommended procedure , pure water or SW standard

I'm not sure if we can speculate as to whether manufacturers actually intend to compensate for slope miscalibration with their directions. I think that it would be difficult to create a calibration recommendation that accounts for all possible manufacturing defects and inaccuracies. Unless the calibration instructions involved a salinity standard. That would always produce correct answers. Relatively few manufacturers appear to recommend using a standard, however.

I think manufacturers who recommend calibrating with freshwater are relying on refractometers to be reasonably precise even if they're not accurate. Precision is the ability of a device to produce a consistent answer when testing the same sample, while accuracy is the ability of a device to produce a "correct" answer (correct being relative to a given standard). A measurement can be precise without being accurate, and vice versa. It can also be neither precise nor accurate, or it can be both..This article has a good summary of precision vs accuracy.

My theory is that manufacturers who recommend calibrating with freshwater know that it won't give an accurate reading at S=35ppt, but they recommend it because reefers always have freshwater on hand. There's no guarantee that reefers will have an appropriate salinity standard on hand. Even if calibrating with freshwater gives an incorrect reading at the top end, I think manufacturers are counting on the refractometer to at least be precise enough to give the same "wrong" answer if it's calibrated with the same standard (in this case, freshwater). Reef aquaria can thrive at a large range of salinity, so the actual number is not as important as stability (so long as the number is within the range of salinity observed on the world's reefs).

Again, this is just my assumption of the motives of refractometer manufacturers. I think it is correct though. If it isn't, that means that Red Sea, Milwaukee, BRS, etc, don't know that calibrating a device far outside its normal measurement range may not produce accurate results within the normal measurement range. To me this seems unlikely.

 

[Randy Holmes-Farley Verified Community Expert]

Originally, the fact that the refractometers were designed for sodium chloride and not seawater was not known to most in the reefing community or by resellers who simply quoted the original manufacturers recommendations to calibrate with RO/Di.

I got quite a bit of push back questioning the "perfection" of them years ago, when everyone only used RO/DI water to calibrate, but it did not take long for folks to realize there was a disconnect between manufacturer claims and reef users experiences.

 

[siggy]

Again, this is just my assumption of the motives of refractometer manufacturers. I think it is correct though. If it isn't, that means that Red Sea, Milwaukee, BRS, etc, don't know that calibrating a device far outside its normal measurement range may not produce accurate results within the normal measurement range. To me this seems unlikely.

Well I was thinking this due to almost all are identical in manufacture, But some require calibration using a 35 std and others require RODI. I assuming that the manufacturing of the two are different, and are calculated to give an acceptable result being that its intended for salt water use. I store my gear in a garage cabinet and this winter 10deg.f. I brought into the house to check fresh mixed SW @78F. and got a reading of about 20ppt,
That spawned the milwaukee purchase. The point being that even with ATC its designed to operate within reasonable temperatures The bimetallic strip has limitations. BTW I am wrong a lot I use bimetallic and its not a one size fits all but sized to the load or range

 

[Randy Holmes-Farley Verified Community Expert]

Well I was thinking this due to almost all are identical in manufacture, But some require calibration using a 35 std and others require RODI. I assuming that the manufacturing of the two are different, and are calculated to give an acceptable result being that its intended for salt water use.

There are two common types.

1. True seawater refractometers (all of these will say so)
and

2. Brine (NaCl) refractometers ) these say nothing special.

ALL of these can be correctly calibrated with an appropriate 35 ppt standard, unless, like the Milwaukee they have software that actually prevents it even though it would work). The fact that a reef company resells it without changing the directions reflects their misunderstanding of their product, or a lack of interest in doing so.

Only the true seawater refractometers can be correctly calibrated with 0 ppt water and be expected to properly read 35 ppt seawater.

The brine refractometers often say to calibrate at 0 ppt, which would be correct for their intended purpose to measure sodium chloride solutions, but will not work properly that way when used to measure seawater.

 

[Rst]

There are two common types.

1. True seawater refractometers (all of these will say so)
and

2. Brine (NaCl) refractometers ) these say nothing special.

ALL of these can be correctly calibrated with an appropriate 35 ppt standard, unless, like the Milwaukee they have software that actually prevents it even though it would work). The fact that a reef company resells it without changing the directions reflects their misunderstanding of their product, or a lack of interest in doing so.

Only the true seawater refractometers can be correctly calibrated with 0 ppt water and be expected to properly read 35 ppt seawater.

The brine refractometers often say to calibrate at 0 ppt, which would be correct for their intended purpose to measure sodium chloride solutions, but will not work properly that way when used to measure seawater.

Hi Randy,
I have just received the Milwaukee which provides 2 starter solutions bottles, one labelled “Steam Distilled Water for Calibration Use Only 00.00 PPM- 00 PSU 1.000 S.G.” and a “Seawater Calibration Validation Solution 1.025 S.G. 20/20” and
a special “IMPORTANT NOTICE” inserted card saying “Digital Refractometers Must Be Calibrated to 0 for Brie or 1.000 for Seawater SG Using Steam Distilled Water ONLY”

With Milwaukee meter, does this mean I can’t NOT use the 35ppt Standard with this meter? Is it still possible to to get accurate results? Are you advising that I would be better off NOT to use the Steam Distilled Water provided for calibration and that I would get better results but using a “35ppt Standard” (which is NOT provided). Sorry I am confused now.

 

[CJS80]

Is it still possible to to get accurate results?

Accurate is a relative term here. Most of the equipment we use for our tanks are hobby grade, which means there is a some degree if inaccuracy. I believe what he is trying to say is that the closer to what we are measuring it is calibrated the less it will be off. If you want about 100% accurate you better be willing to pay for all the extra 000. in the price tag.