Randy et al, extrapolating data to 35 ppt

[DrBegalke]

Any input on converting measurements (calcium, magnesium, etc.) made at various salinity to a standard 35 ppt?

Basically the study was done with 7g of salt in 200mL of water, then was tested without correcting to 35 ppt as most of us would do before use.

I would think most measurements, notable exception of alkalinity, could be converted linearly as Billybeau1 started to do.

Salt analysis thread here:
http://reefcentral.com/forums/showth...readid=1317416

[bertoni]

Assuming dissolution is equal at both concentrations, all of the ppm numbers, including alkalinity, can be converted linearly. pH doesn't work that way.

[MCsaxmaster]

Agreed, the relationship is linear. You can set up a ratio for the measured value (calcium, magnesium, alkalinity, etc.) relative to the value at a salinity of 35 ppt given a measured salinity. Say we want to calculate an the calcium concentration (x) at a salinity of 35 ppt and have measured value for calcium and salinity:

x/Ca = 35/S, where x is the value we want to calculate, Ca is the measured calcium and S is the measured salinity:

x = 35*Ca/S

For example,

x = 35 * 400 ppm calcium/30 ppt salinity

x = 467

At a salinity of 35 ppt this salt would mix to 467 ppm calcium. Of course there is error in measurement for all of this, so realistically you'd probably get 467 +/- 30 ppm given error in the calcium and the salinity measures.

[Boomer]

Dr

Chris gave you the perfect equation

[jdieck]

I still can't belive they only used 7 grams of each and that they tested them at different salinities!
If a sample of only 7 grams was used (just barely above a teaspoon) I would be more concerned about the homogeinity of the salt mix and it's components than the fact that salts were tested at different salinities, even if the mix was shaken or re-mixed in thecontainer.
A more realistic sample could have been a volume of 25 to 50 liters rather than a small 200 ml.

[Billybeau1]

I wonder what they did with all the extra salt left over after the test

[Billybeau1]

Originally posted by jdieck
I still can't belive they only used 7 grams of each and that they tested them at different salinities!
If a sample of only 7 grams was used (just barely above a teaspoon) I would be more concerned about the homogeinity of the salt mix and it's components than the fact that salts were tested at different salinities, even if the mix was shaken or re-mixed in thecontainer.
A more realistic sample could have been a volume of 25 to 50 liters rather than a small 200 ml.

Seriously, that could explain the big swings they got in some samples of the same salt mix.

[HowardW]

I would've trusted the final results a bit more if they just mixed up a single entire 50g bag of each salt to 35 PPT and went from there. I have little interest in how much salt it takes of any one brand to acheive a certain salinity.

[Randy Holmes-Farley]

They do not give enough information to understand what they did, but the conductivity measurements and salinity values in ppt do not agree. Without quibbling about the details of the exact conversion of salinity to ppt, the salts do not even rank order in the same way.

For example, Kent has a higher average conductivity than Coralife, but a lower average salinity. While that is theoretically possible if they used something other than conductivity to measure salinity, without knowing what they did to measure salinity, it makes conversion to 35 ppt more suspect than it might otherwise be.

[HowardW]

Another thing is I wonder how long they waited after mixing up the salts before testing? I've read that certain components such as calcium can take as long as 24-48 hrs. before fully going into solution and getting a full reading.



edited: Ooops, they say they mixed and aerated each sample for 24 hrs.

[MCsaxmaster]

Ha, I didn't even notice the part about a salt analysis thread!

Agreed with the above though. It makes absolutely NO sense to do anything but mix up the salt to a standard salinity. That doesn't have to be 35 ppt, but there should be a standard. Since many of the salts in the salt mix are hydrated they will contain water and thus a mass measurment of the salt is utterly useless if different salts have different formulations (which they do).

The most sensical thing is to mix all of the mixes up to the same salinity and in the same way. If you don't normalize for salinity they you'll get widely variant measured parameters from the same salt! I agree that I see little virtue in those data.

[MCsaxmaster]

p.s. Given my suspicions of not really knowing what they were doing there, it would also be nice to see standard curves for all of those measurments. If they can't accurately make up and measure standards then I have no reason to believe any of the data means a darn thing.

[Boomer]

JD

Yes, I also agree on the sample size being way to small. I originally suggested 35 grams in 965 mL of RO/DI water, The only thin I really looked at yesterday was the bromide. I'm sure we missed other things too.

Randy

other than conductivity to measure salinity, without knowing what they did to measure salinity

Yah, there is like 4 of them. Most are some variations of the Knudsen Method. i.e., Chlorinty and Chlorosity and then there is Density. And if they used a hydrometer what kind of hydrometer did they use and how was it corrected and do they know D does not equal SG. Or was it a refract and how was they fixed.

[MCsaxmaster]

Agreed, there is NO SUCH THING as a salinity measurement in analytical chemistry. There are various methods to estimate salinity (conductivity, chlorinity, resistivity, refractive index, various ways to measure density/specific gravity). How in the world they could have measurements they call "Salinity" and then have those same measurments disagree with their conductivity and think nothing of it is beyond me.

And while I'm complaining ( ) they got the definition of salinity completely wrong. Salinity is the mass of dissolved compounds per liter of SOLUTION, not per liter of WATER. You can't start with 200 ml of water, add 7 grams of salt and expect to get a salinity of 35 ppt. No, instead of having a solution that weighs 200 g and has a salinity of 35 ppt you will have a solution that weighs 207 g and has a salinity of something less than 35 ppt.

I have skimmed over this at best and what jumps out to me is that it is just sloppy, sloppy, sloppy. I would hope that anyone with a BS in chemistry or higher would not make these mistakes. If they do, they need to go back to school.

[HowardW]

Well after reading DrBegalkes post in the Reef Forum about the AWT tests, they apparently did report the various salt test results at 35 PPT although it wasn't clearly mentioned.


<<< Yes, it was not clear in the report I posted, but they tested the 7g in 200mL water as a yield measurement, discarded those samples, and then used 35ppt (53 mS/cm) samples for the rest of the testing.

Also, I should note that they sent me the pdf but allowed me to decide how/when/if to distribute it. I could have just posted the raw data, or posted it in another manner, but thought making a quick website and posting the entire pdf, logo included, was the best way to go.

Sorry for any confusion, maybe I can start another thread or the mods can post something on the first page clarifying the 35ppt procedure. >>>

[Boomer]

Just to add some info here from Dr from a recent note he got

FYI, from AWT:
... We believe there is some misunderstanding pertaining to the actual salts that were tested. The salt that we mixed up for the "yield" test was discarded after the data was collected. The salts that were tested by parameter were all mixed up to 53mS, and double checked with a temp-compensated refractometer.

[Billybeau1]

Sheesh...... what a mess.

I like my numbers better

[reef_doug]

But how large were the samples they prepared? The whole bag, 25 gal, less?

[MCsaxmaster]

Originally posted by reef_doug
But how large were the samples they prepared? The whole bag, 25 gal, less?

Exactly! How was this done? The methods are not remotely clear. Also, for some of the parameters with some salts they arrived at relatively similar values for both batches of a particular salt, but for some others the values were drastically different. Why? Were the two batches of the same salt really dramatically different or is this a result of sampling artifacts? We have no way to even begin guessing at the cause since we have no idea what they did to obtain these values.

I have about as much confidence in these results as I do in a middle school science fair project, at least at this point. We have no idea how they did what they claim to have done and therefore have no idea whether the reported data is an accurate representation of the salt batches they measured, or if it's all a bunch of hooey.

What they tried to do is not a particularly difficult thing to do given the equipment they have. However, it just looks so sloppy to me that I can't conclude those results are at all reliable at this point.

[HowardW]

I agree with MCsaxmaster.........using their calcium test results on Seachem Reef Salt as an example, they got 401 PPM on one batch and 538 PPM on the other?? I've been using that salt for years and the calcium has always tested out for me at between 450-480PPM @ 1.0265 when I do random spot checks, and I prepare my saltwater in 5g batches.

In addition, me and Billybeau have this ongoing friendly disagreement on the calcium levels of Seachem Reef Salt whereas he says it tests out in the mid 500's @ 1.0265 and that my kits are always reading low in the mid to high 400's. Well this AWT study didn't do much to settle our disagreement with that large discrepency of 401 and 538 PPM.

[Billybeau1]

Hi Howard. I've got a co-worker that lives in Aurora. I"ll ask him where he lives and if he is close to you, maybe you could give him a pint of your tank water and I will test it.

I'll PM you when I find out.