TEST METER FOR TESTING NITRATES

[Rick Mathew]

@Rick Mathew

I just did a little experiment with this (Saturday, everyone sleeping in makes this easier)

I did the following:

#1
10ml test solution per the Red Sea instructions (watch for the reaction byproduct) = 82 (4.25)
10ml diluted w/ 10 ml RODI(but used 100% test sample for the C1) =39, x2=78 (4.07)

#2
16ml using the procedure as outlined = 100 (5.08)
16 ml sample (10ml test diluted with 10ml RODI) but 100% C1 sample = 46, x2= 92 (4.76)

Then (as I thought of it after I tossed the 10ml samples from #1 above so no comparison)
#3
used the reacted and diluted from #2 above but made the C1 sample 50% RODI/50% test water= 44, x2=88 (4.53)

so all these methods came out pretty close for a one-point curve. the 10ml sample reaction was lower overall Nitrate results as would be expected expected from the 16ml samples so the curve may not work all the way and the methodology would not necessarily support the findings anyhow (really is likely close enough and within accuracy specs for what we are doing as they all fall within 1). It was just a way to kill some time.

reference to the Salifert was reading about 5 as was the Red Sea kit (guessing from the windows of color). So we are seeing validation again.

Now - what if instead of RODI - dilute with test water as you indicated? Is the reaction complete or might it continue?

I think the reaction is completed after the 9 min reaction time...The reason I say that is I have measured the sample over time and it is does not change hardly at all...I did this to see how stable the color was over time...seem to be very stable...another point to keep in mind is the meter variation itself...it is reported to be + - 10ppb + - 4% of reading.... Using the regression formula that represents about + -.5ppm..expected meter error

 

[ckalupa]

@Rick Mathew

Today - 16 ml test per video

Salifert regular = 0
Red Sea regular = slight shade close to 1

Procedure test on Hanna = 2.3

Salifert almost always lower than other two

 

[Rick Mathew]

@Rick Mathew

Today - 16 ml test per video

Salifert regular = 0
Red Sea regular = slight shade close to 1

Procedure test on Hanna = 2.3

Salifert almost always lower than other two

A little higher than I would have expected...if you consider the meter error range it could be 1.8 to 2.8...still in the ball park

 

[Rick Mathew]

@Rick Mathew

Today - 16 ml test per video

Salifert regular = 0
Red Sea regular = slight shade close to 1

Procedure test on Hanna = 2.3

Salifert almost always lower than other two

Just a FYI...One thing that I do to increase my repeatability (precision) is to level the the amount of dry reagents in the spoons...I fill the spoon and level it off with a spatula...gives me more consistent results...reagent B some times has small chunks in it that need broken up...I found they can interfere with results...hope this helps...I am working on the salinity effect question and have run a set of experiments...results are interesting not what I expected...I am re-running the to be sure, but the first set shows a very minimal effect due to salinity...within the meter reading error. My initial go at it some time ago was rough and quick...This time I took great care to make sure the salinity was correct (actually measured it) I wasn't so concerned about the NO3 concentration levels because I just wanted to see if concentration of NO3 in the sample changed the outcome based on salinity...

I ran salinity levels of 0, 29, 31, 35 and concentration levels at 1 ppm 3 ppm & 5 ppm....The slope of the trend line of measured ppm vs salinity at all NO3 levels is virtual "0"...meaning no effect due to salinity...not sure what to make of yet will keep you posted...

 

[ckalupa]

A few more days behind me now and I am about to discontinue the correlation tests as the three methods continue returning similar results to the Hanna method giving me something that I don’t need to try to compare color tables.

Thanks Rick! Appreciate this work you have done fixing needing to read a color chart (a lot!)

 

[Rick Mathew]

A few more days behind me now and I am about to discontinue the correlation tests as the three methods continue returning similar results to the Hanna method giving me something that I don’t need to try to compare color tables.

Thanks Rick! Appreciate this work you have done fixing needing to read a color chart (a lot!)

You are most welcome sir...thank you for providing some validation to the procedure...If you find anything of interest in the future just let me know

 

[rkpetersen]

Just got my nitrite checker today. Must have been on the slow train from brs.

The meter flashes 200...my procedure when it flashes 200 is I take the 10mL from the test vile and dump it into a small container and add 10 mL of the un-reacted water to it, mix and then place 10mL of the diluted solution into the measurement vile and remeasure...get the results and multiply by 2...or 3 or 4...what ever the dilution factor is....but you bring up an interesting idea...what if one doubled the initial test water volume (16mL to 32mL)...would the meter pick up the lower levels of NO3....if it worked we would have doubled the meter range...this is why collaboration is so awesome...by simply asking a question and new idea is born....I will give it a try

I think your original idea here is probably the most straightforward way to get a likely valid reading if the number is over 200; just dilute the developed sample with blank tank water into a new test vial.

The idea of using twice the water sample while holding the amount of reagent used unchanged I think would give unpredictable results; the result wouldn't necessarily be lower, but might be somewhat lower although not the predicted degree, and the timing of the reaction might change (likely take longer to fully develop the color.)

On the other hand, I don't think using RODI water to dilute the sample before testing would be ideal because this would drastically change the sample's salinity and so the chloride effect would be magnified. (If the effect is confirmed.)

Probably the best method would be to dilute the original sample with fresh 35 ppt saltwater with known 0 nitrates, before running the test; this approach could be confidently used with the test procedure and regression formula you've established. It might even be worth keeping a small bottle of it on hand as part of this test kit.

 

[ckalupa]

This could be tested on the low range samples we already are doing. Run the test. Dilute 50%. Check in the Hanna again. Would need to likely prove the curve and formula still applied?

 

[EmdeReef]

This is really cool, congrats @Rick Mathew! Just ordered a nitrite checker and can’t wait to try.

 

[Rick Mathew]

This is really cool, congrats @Rick Mathew! Just ordered a nitrite checker and can’t wait to try.

Thank you...let me know how it works for you...The more people we have using it the better chance we have of improving the results

 

[Rick Mathew]

Just got my nitrite checker today. Must have been on the slow train from brs.



I think your original idea here is probably the most straightforward way to get a likely valid reading if the number is over 200; just dilute the developed sample with blank tank water into a new test vial.

The idea of using twice the water sample while holding the amount of reagent used unchanged I think would give unpredictable results; the result wouldn't necessarily be lower, but might be somewhat lower although not the predicted degree, and the timing of the reaction might change (likely take longer to fully develop the color.)

On the other hand, I don't think using RODI water to dilute the sample before testing would be ideal because this would drastically change the sample's salinity and so the chloride effect would be magnified. (If the effect is confirmed.)

Probably the best method would be to dilute the original sample with fresh 35 ppt saltwater with known 0 nitrates, before running the test; this approach could be confidently used with the test procedure and regression formula you've established. It might even be worth keeping a small bottle of it on hand as part of this test kit.

I would agree...Interesting idea about dilution with 35ppt saltwater...might be worth a try.

 

[TexasTodd]

So, Rick, have you reached out to Hannah? If you haven't, does anyone here on R2R know someone there he can try to contact?
You might even get some sort of royalty if they're a stand up company.
Todd

 

[Rick Mathew]

This could be tested on the low range samples we already are doing. Run the test. Dilute 50%. Check in the Hanna again. Would need to likely prove the curve and formula still applied?

That is correct...I have actually tested this on several levels up to 20 ppm and it seems to hold (that is the regression formula)...It should be true if the solution follows Beer's /Lambert's Law...1/2 the concentration cut the absorbance by 1/2 and thus the meter reading....I have not extensively looked at this but the limited tests I have done is appears to follow...

 

[Rick Mathew]

So, Rick, have you reached out to Hannah? If you haven't, does anyone here on R2R know someone there he can try to contact?
You might even get some sort of royalty if they're a stand up company.
Todd

Actually I have not...My main goal is to develop tests that help me (us) better manage the reef environment and IMO that means getting data that is more definitive than just a range (1-5) or difficult to interpret color charts. ...Thanks for the thought

 

[TexasTodd]

Very admirable but it's also a great addition to their platform. There has to be someone on R2R with high up contacts at Hannah.
Thread of the year for 2018!

 

[rkpetersen]

Tried this method for the first time this morning, on 3 tanks.


First tank is a low nutrient and sps-heavy mixed reef:

Nyos Nitrate (my standard test) - <<1 (barest trace of yellow, almost 0. In fact it could be zero, given that the liquid reagent of this kit sometimes imparts a slight yellowish tinge even before the solid reagent is added.)
Red Sea Nitrate Pro - 0.5 ppm.
Hybrid RS Nitrate Pro / Hanna ULR Nitrite - 0.82 ppm.

This first tank raises an interesting concern with this method. The linear equation that best fits the experimental data used to establish this technique, [NO3-ppm] = 0.0458 x [NO2-ppb] + 0.502, does not go through 0,0. However, it should. Otherwise, even a sample with no visible color development and a nitrite meter reading of 0 will still result in a measurement of 0.5 ppm nitrate, which isn't correct.

In my first tank example, the Nyos test suggests that the nitrate level is far closer to 0 than to 1 (and I believe this to be the case). The nitrite meter only read 7, and yet this equates to 0.82 ppm nitrate.
I can imagine two ways this could be adjusted. First would be to just ignore the 0.502 and do the conversion using only 0.0458 x [NO2-ppb]. In my example, this would be 0.32 ppm nitrate.
The other way would be to re-do the best fit slope forcing it to pass through 0,0. I'm not sure which would be better. The discrepancy would become less and less important the higher the checker reading.


Second tank is an anemone-heavy mixed reef with higher nutrient levels:

Nyos - 12 or perhaps slightly less (in the range where the Nyos color-matching precision is low.)
Red Sea Nitrate Pro (high range) - 12 ppm.
Hybrid (without dilution) - Out of range (>9.66 ppm.)
Hybrid (with 1:2 dilution of developed sample) - 10.49 ppm.
Hybrid (with 1:2 dilution of undeveloped sample by 35 ppt fresh saltwater) - 11.31 ppm.

These results are very nice and really the midrange is where I think this technique will shine the most. I was pleased that the results obtained by simply diluting the developed sample and putting it in the meter again are very similar to those obtained by running the test with an initially diluted sample.

Incidentally, I also tested the freshly mixed 35 ppt Red Sea Coral Pro saltwater diluent with this technique. No color development, meter reading of 0; equals 0.5 ppm nitrate? Probably not.


Third tank is a QT that I recently did a big water change on. I normally never check the nitrate on this tank:

Nyos - Between 3 and 5 ppm.
Hybrid - 4.85 ppm.

Pretty good agreement. Would be even better agreement without the 0.502 factor. '


One comment on technique, which applies not only to this method but all that use the Red Sea reagents and any colorimeter. The reaction produces a large quantity of microbubbles and a variable quantity of dark particulate debris. If either of these are significantly present in the vial when placed in the meter, the resulting number may be artifactually elevated.


So I like this technique. Thanks @Rick Mathew! Much better than visual color matching. I'm going to continue using it side by side with Nyos for comparisons over the next few weeks. And compare it and report back when I send off my next set of samples to ATI. (However, ATI doesn't determine nitrate via ICP-OES, so how they do is anyone's guess but could be something similar to this technique.)

 

[Rick Mathew]

Tried this method for the first time this morning, on 3 tanks.


First tank is a low nutrient and sps-heavy mixed reef:

Nyos Nitrate (my standard test) - <<1 (barest trace of yellow, almost 0. In fact it could be zero, given that the liquid reagent of this kit sometimes imparts a slight yellowish tinge even before the solid reagent is added.)
Red Sea Nitrate Pro - 0.5 ppm.
Hybrid RS Nitrate Pro / Hanna ULR Nitrite - 0.82 ppm.

This first tank raises an interesting concern with this method. The linear equation that best fits the experimental data used to establish this technique, [NO3-ppm] = 0.0458 x [NO2-ppb] + 0.502, does not go through 0,0. However, it should. Otherwise, even a sample with no visible color development and a nitrite meter reading of 0 will still result in a measurement of 0.5 ppm nitrate, which isn't correct.

In my first tank example, the Nyos test suggests that the nitrate level is far closer to 0 than to 1 (and I believe this to be the case). The nitrite meter only read 7, and yet this equates to 0.82 ppm nitrate.
I can imagine two ways this could be adjusted. First would be to just ignore the 0.502 and do the conversion using only 0.0458 x [NO2-ppb]. In my example, this would be 0.32 ppm nitrate.
The other way would be to re-do the best fit slope forcing it to pass through 0,0. I'm not sure which would be better. The discrepancy would become less and less important the higher the checker reading.


Second tank is an anemone-heavy mixed reef with higher nutrient levels:

Nyos - 12 or perhaps slightly less (in the range where the Nyos color-matching precision is low.)
Red Sea Nitrate Pro (high range) - 12 ppm.
Hybrid (without dilution) - Out of range (>9.66 ppm.)
Hybrid (with 1:2 dilution of developed sample) - 10.49 ppm.
Hybrid (with 1:2 dilution of undeveloped sample by 35 ppt fresh saltwater) - 11.31 ppm.

These results are very nice and really the midrange is where I think this technique will shine the most. I was pleased that the results obtained by simply diluting the developed sample and putting it in the meter again are very similar to those obtained by running the test with an initially diluted sample.

Incidentally, I also tested the freshly mixed 35 ppt Red Sea Coral Pro saltwater diluent with this technique. No color development, meter reading of 0; equals 0.5 ppm nitrate? Probably not.


Third tank is a QT that I recently did a big water change on. I normally never check the nitrate on this tank:

Nyos - Between 3 and 5 ppm.
Hybrid - 4.85 ppm.

Pretty good agreement. Would be even better agreement without the 0.502 factor. '


One comment on technique, which applies not only to this method but all that use the Red Sea reagents and any colorimeter. The reaction produces a large quantity of microbubbles and a variable quantity of dark particulate debris. If either of these are significantly present in the vial when placed in the meter, the resulting number may be artifactually elevated.


So I like this technique. Thanks @Rick Mathew! Much better than visual color matching. I'm going to continue using it side by side with Nyos for comparisons over the next few weeks. And compare it and report back when I send off my next set of samples to ATI. (However, ATI doesn't determine nitrate via ICP-OES, so how they do is anyone's guess but could be something similar to this technique.)

Really appreciate the feed back and the and the fact that you are working on replication my results...You are correct about the regression needing to go through "0" I actually have an equation with a "0" intercept that I have been evaluating over time for the very reasons you point out ...It is Y= .0476 x the Meter reading.... It doesn't appear to work as well in the mid to upper region...but I am still evaluating it...When I get results down at 2 or below I do what you did and ignore the .5 intercept value As for the Red Sea reagent "floaters" Excellent observation... This is the reason I put the 2 min time wait right after the reaction time is completed (9 min) and I decant the liquid is to allow time for the Floaters to get to the top or bottom... The inverting of the sample 2-3 times at the 5 min mark of the 9 min reaction time seems to reduce the micro bubbles..At least that has been my observation

Thank again for the feedback...as we gather more data and experience with this technique it will only get better...looking forward to your results and I will update as I get more information

rick

 

[Rick Mathew]

Tried this method for the first time this morning, on 3 tanks.


First tank is a low nutrient and sps-heavy mixed reef:

Nyos Nitrate (my standard test) - <<1 (barest trace of yellow, almost 0. In fact it could be zero, given that the liquid reagent of this kit sometimes imparts a slight yellowish tinge even before the solid reagent is added.)
Red Sea Nitrate Pro - 0.5 ppm.
Hybrid RS Nitrate Pro / Hanna ULR Nitrite - 0.82 ppm.

This first tank raises an interesting concern with this method. The linear equation that best fits the experimental data used to establish this technique, [NO3-ppm] = 0.0458 x [NO2-ppb] + 0.502, does not go through 0,0. However, it should. Otherwise, even a sample with no visible color development and a nitrite meter reading of 0 will still result in a measurement of 0.5 ppm nitrate, which isn't correct.

In my first tank example, the Nyos test suggests that the nitrate level is far closer to 0 than to 1 (and I believe this to be the case). The nitrite meter only read 7, and yet this equates to 0.82 ppm nitrate.
I can imagine two ways this could be adjusted. First would be to just ignore the 0.502 and do the conversion using only 0.0458 x [NO2-ppb]. In my example, this would be 0.32 ppm nitrate.
The other way would be to re-do the best fit slope forcing it to pass through 0,0. I'm not sure which would be better. The discrepancy would become less and less important the higher the checker reading.


Second tank is an anemone-heavy mixed reef with higher nutrient levels:

Nyos - 12 or perhaps slightly less (in the range where the Nyos color-matching precision is low.)
Red Sea Nitrate Pro (high range) - 12 ppm.
Hybrid (without dilution) - Out of range (>9.66 ppm.)
Hybrid (with 1:2 dilution of developed sample) - 10.49 ppm.
Hybrid (with 1:2 dilution of undeveloped sample by 35 ppt fresh saltwater) - 11.31 ppm.

These results are very nice and really the midrange is where I think this technique will shine the most. I was pleased that the results obtained by simply diluting the developed sample and putting it in the meter again are very similar to those obtained by running the test with an initially diluted sample.

Incidentally, I also tested the freshly mixed 35 ppt Red Sea Coral Pro saltwater diluent with this technique. No color development, meter reading of 0; equals 0.5 ppm nitrate? Probably not.


Third tank is a QT that I recently did a big water change on. I normally never check the nitrate on this tank:

Nyos - Between 3 and 5 ppm.
Hybrid - 4.85 ppm.

Pretty good agreement. Would be even better agreement without the 0.502 factor. '


One comment on technique, which applies not only to this method but all that use the Red Sea reagents and any colorimeter. The reaction produces a large quantity of microbubbles and a variable quantity of dark particulate debris. If either of these are significantly present in the vial when placed in the meter, the resulting number may be artifactually elevated.


So I like this technique. Thanks @Rick Mathew! Much better than visual color matching. I'm going to continue using it side by side with Nyos for comparisons over the next few weeks. And compare it and report back when I send off my next set of samples to ATI. (However, ATI doesn't determine nitrate via ICP-OES, so how they do is anyone's guess but could be something similar to this technique.)

Here are a couple of charts that can give you an idea of the differences in the regression formulas...These are the last three versions I have been working with...the older ones were not suitable ...hope this helps

I think that any one of theses regressions will return acceptable results (at least for me). They each have there "strong & weak" areas in the range...I selected the MyCurveFit (purple line) because it seems to be more more accurate in the range of NO3 that I operate in....It would be possible to create a "duel range" method by using an "IF" statement in Excel ...IF reading is in this range use this regression...If in other range use this regression...might be worth considering...hope this helps

 

[Rick Mathew]

Thanks for posting this. [emoji3]

Did you check the salinity/chloride effect on the accuracy?

Randy: After you asked me this question I decide to go back to my original work on the HI-764 Salinity Effect and retest it. This time I did the experiments paying much more attention to the accuracy of my work and I got some surprising results...Below are the results

EFFECTS OF SALINITY ON NO3 MEASUREMENTS USING HI-764 CHECKER

OK…I just completed the repeat experiments of the Effect of Salinity on the nitrate test procedure using the HI-764. I did this experiment in total 2 times because I did not believe my first results. The second set of experiments confirmed the first set. The chart below shows the results graphically.

The procedure was as follows:

1) Make a 50 ppm N03 standard from RODI water and HACH N03N Standard

2) Mix RODI water with the 50 ppm standard to get the desired ppm for the test (1,3,5)

3) Weigh and mix the required amount of reagent grade sodium chloride to get the desired salinity in PPT (0,29,31,52)

4) Measure the salinity to confirm. (Milwaukee Refractometer)

5) Run the Nitrate test described above using HI-764 Hanna Meter. (3 replicas at each level)…36 total tests including replicas

6) Record and plot and analyze the result

PREPARED SAMPLES

As you can see the slopes are “0” which I interpret to mean no effect as a function of salinity level!


So what is different from my previous work, which indicated a salinity effect?



1) First work was done with a different meter HI-736 & HI-96702

2) My first work with the HI-764 and salinity effects was not carefully done. It was just checking to see if there was an effect and there is probable error in the work. I expected an effect so when I saw the differences in the measurements I concluded there was one and let it go at that.

3) The batch of Red Sea Reagents are different from previous measurements

4) Much more care was taken in this last set of tests then previous work.

5) The moon was full when I did the previous work!

All of that being said I am going to put this out there for comments…any Thoughts?

 

[TheEngineer]

That's really cool. There's clearly an effect, but it is small enough we can ignore it. I'm interested in why the trend is negative for the 5ppm sample.