HI 736 ULR - Test / Same-Sample Retest Variations (compendium)

[CalmSeasQuest]

This was originally posted in the Hanna sponsor forum, but as no reply from Hanna has been received (despite responses to other threads.) I've provided a compendium here in the Chemistry forum in the hopes someone (Randy ) might be able to help me make sense of the findings.

4/24/12 - When using the HI 736 ULR Phosphate checker, I note considerable variations between the initial test done using the 3 minute timed method and subsequent, immediate retests made using the same sample via the instant method.

All cuvettes are wiped clean with a microfiber cloth prior to each test, each cuvette is rinsed with tank water prior to testing. Care is take to make sure all the powder agent is introduced into the cuvette. The instant re-tests are done using the second cuvette for the C1 standard sample.

Over the last two days I have obtained the following results...

4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

I understand the use of a second cuvette for the C1 tank water standard for the instant tests might introduce some variation, but I have alternated the the cuvettes used and still obtain similar variations.

Has anyone else experienced this issue? I'm hopeful Hanna can explain why initial testing might produce results much higher than all subsequent testing?

4/25/12 - Today I used 2 cuvettes for all the tests to minimize any variable caused by the use of the second cuvette on the retests. Care is taken to use the same alignment when inserting into the checker (10ML symbol facing front) and the checker was turned off for ~10 seconds between retests. The results were

4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

*I conducted the retest two additional time as I was getting results of 0 ppb.

I believe the trend of phosphates being lowered is correct as the GFO reactor was recently changed, but I am not confident in which value range to trust. I am puzzled by the difference and the fact that the retests always produce much lower results than the initial tests - seemingly well beyond the stated error rate for the device. Also curious is that all the retests provide very consistent results within the stated accuracy range.

I wonder if any other hobbyists with a 736 could replicate these tests to compare results? It only takes a few minutes as the retests are almost instant and require no additional reagents. I'd be happy to PM or speak with anyone interested to make sure the processes are identical.

For the record, I'm generally very pleased with this and all the other Checkers I own. At the price point for this device, I fully understand and accept there will be variations in results. I'd just like to better understand how to interpret these variations, adjust my process if needed and know which results to use. Hopefully Hanna will respond with some insight. Perhaps Randy Holmes-Farley will stumble upon this with the hope he'll evaluate my process and results to make sure I'm not missing something.

4/26/12 - I repeated the exact same tests and process today with nearly identical results. I am becoming convinced there is an issue with my initial test result values. Here are all the results side-by-side...

4/26/12 7:00AM Initial - 16 ppb. Immediate follow up tests - 0, 4, 6, 5, 4
4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

I believe I might have found the problem, and it may be partially my methodology. When mixing the reagent, I have not been mixing it for a full 2 minutes as contained in the directions. I shorten this time so as to not have the checker power off as, again per the directions, the unit turns itself off after 2 minutes on "non-use". There is still a tiny amount of undissolved reagent visible in the cuvette when the vial in placed in the checker for the initial test. Although the reagent is completely dissolved 3 minutes later when the checker takes the reading, I'm wondering if perhaps during the 3 minutes the vial sits, unagitated that variations or striations in reagant concentration are created in the cuvette thereby altering the test results.

In an attempt to confirm this I'll alter my procedures tomorrow. I will again use 2 cuvettes, but manually mix the C2 sample for the instructed full 2 minutes (hoping the unit doesn't shut itself off), then repeat the follow-up instant tests. If this is the cause of the variation in results, it might mean the the follow-up instant test results are more accurate.

I'm not sure if Hanna actively reviews these forums, but any guidance available would be appreciated, especially regarding the stated 2 minute auto-off policy when the user is instructed to mix the reagent for 2 minutes. Considering the time required to remove the cuvette, add the reagent powder and mix - it's not possible to complete those actions in the allotted 2 minute time frame.

4/27/12 - So much for that theory...

Today I made sure to mix C2 for the full 2 minutes. The good news was despite the 2 minute auto time-off period detailed in the instructions, the checker did not turn off.

At the end of the 2 minute period, virtually all of the reagent was dissolved. The test results were virtually identical to the previous,

4/27/12 7:00AM Initial - 17 ppb. Immediate follow-up tests - 4, 2, 7, 0, 6
4/26/12 7:00AM Initial - 16 ppb. Immediate follow up tests - 0, 4, 6, 5, 4
4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

This coupled with the following response from Hanna provided in another thread seems to do away with my theory that undissolved reagent allowed to sit in the still cuvette for 3 minutes might have caused the variance...If the powder still isn't completely dissolved by the end of the 1 minute 45 second shaking, it will not affect the reading as long as you see that it is dissolved when you remove the vial after the three minute countdown and reading. If there is still undissolved powder at that point, it is likely that you will get a false low, but it is impossible to quantify.

Sorry for the diatribe, I was trying to make sure I included all relevant information. I'm at a total loss in understanding why these results vary so widely. I considered the reagent breaking down, but the retests take place within seconds of the original tests. I believe my test methods have been sound including alternating the cuvettes used for C1 and C2 on subsequent days to account for any optical variance in the glass - especially since the results have been consistent and reproducible.

I'm wondering if anyone can come up with a cause of these variances and help determine which range should be trusted. Perhaps another HI-736 user could duplicate the tests and share the results? It's entirely possible I'm missing something, but I just don't see it.

Thanks all!

 

[chercm]

anyone can help advice ? i am having the same problem

 

[Randy Holmes-Farley]

Sorry, I'll have to leave this to folks who have used the kit to know what seems to work best for them.

 

[chercm]

for the tank water , there will always be particular in it , so it will make the test to be inaccurate

 

[Met3112]

I'm having the same problem with the Phosphorus ULR meter. I started a similar thread here: http://www.reefcentral.com/forums/showthread.php?t=2161605.

I am getting very high readings which don't reflect on my reef keeping. I will let you know when I am able to do a test like yours

 

[CalmSeasQuest]

Sorry, I'll have to leave this to folks who have used the kit to know what seems to work best for them.

Thanks Randy, and I understand. I did the same tests one last time with virtually the same results...

4/30/12 7:00AM Initial - 25 ppb. Immediate follow up tests - 3, 12, 10, 4
4/27/12 7:00AM Initial - 17 ppb. Immediate follow-up tests - 4, 2, 7, 0, 6
4/26/12 7:00AM Initial - 16 ppb. Immediate follow up tests - 0, 4, 6, 5, 4
4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

To me at least, the differences are not insignificant. If the initial test results are to be believed, my PO4 levels are still higher than desired. If the retests are more accurate, then I've likely achieved my desired levels (<10 ppb Phosphorus / 0.03 ppm Phosphate.)

The most disappointing part of this is despite responding to others thread and questions since this was posted, Hanna has "avoided" responding to this issue. All that has been asked for is help determining which values are more accurate and why. It seems Hanna would be able to easily verify this and provide guidance.

 

[gorliz13]

Hi,

I have just repeated your experiment. Using 2 cuvettes and taking care to put the 10ml at the front everytime. I get higher results than the initial reading most of the time but at least the differences are not as large as yours. It is towards the end that they start going down but not significantly Here are the results of the tests:

Initial test: 10

retests: 15,18,18,14,18,10,10,6,9.

Not sure this helps at all but hope it does.

 

[tmz]

I don't have the 736. The 713 full sized meter has been very consistent over years of daily testing. Sorry wish I could be of more help.

 

[chercm]

if i am not wrong , the amount of powder that is put into the viral will differ the reading .

I sure hope hanna will come out with the liquid

 

[CalmSeasQuest]

Hi,

I have just repeated your experiment. Using 2 cuvettes and taking care to put the 10ml at the front everytime. I get higher results than the initial reading most of the time but at least the differences are not as large as yours. It is towards the end that they start going down but not significantly Here are the results of the tests:

Initial test: 10

retests: 15,18,18,14,18,10,10,6,9.

Not sure this helps at all but hope it does.

Thank you for taking the time to do those tests gorliiz - Wow, your results are opposite of my findings. I'm not even sure how to interpret. I guess it does open up the possibility that my meter is defective, but it seems odd that the results are so consistent and predictable.

I don't have the 736. The 713 full sized meter has been very consistent over years of daily testing. Sorry wish I could be of more help.

Thanks Tom, I switched from the HI 713 checker to the HI 736 believing it would be more accurate. Assuming I am interpreting their accuracy statements correctly and that I didn't drop a zero somewhere, ironically it appears that based on Hanna's data the accuracy of the HI 736 (±5 ppb / ±5% of reading) vs the HI 96713 (±0.04 mg/L (40 ppb?) ± 4% of reading @ 25°C) the HI 736 on paper appears to be more accurate. I would have gladly paid more for a more precise unit prior to this event. Hanna's total disregard to customer service has resulted in my hesitance to purchase any additional Hanna equipment.

if i am not wrong , the amount of powder that is put into the viral will differ the reading .
I sure hope hanna will come out with the liquid

If I understand it correctly, a liquid phosphate reagent is problematic as it has a very short shelf life although, that came from Hanna and based on how they have handled this - I'm not sure how much trust I place in their statements.

I have found a relatively simple way to improve the powder dosing - Hold the packet diagonally so that one corner is at the bottom, tap it to make sure all the reagent is in the bottom corner, then using a pair of scissors, cut the packet entirely in half diagonally. It's then very easy to "pinch" it into the shape of a spout making it easy to pour and visually confirm that all the powder had been added to the cuvette.

 

[djkms]

IIRC Hanna stated that getting every little bit of reagent into the vial is not important for the levels we test at. They put enough reagent into the packs to give a max reading. A little missed powder makes no difference at our levels.

 

[CalmSeasQuest]

IIRC Hanna stated that getting every little bit of reagent into the vial is not important for the levels we test at. They put enough reagent into the packs to give a max reading. A little missed powder makes no difference at our levels.

Hanna's response to the powder reagent question was ...

1. If the powder still isn't completely dissolved by the end of the 1 minute 45 second shaking, it will not affect the reading as long as you see that it is dissolved when you remove the vial after the three minute countdown and reading. If there is still undissolved powder at that point, it is likely that you will get a false low, but it is impossible to quantify.

2. Micro bubbles at the TOP should not affect the reading. If there are micro bubbles suspended in the middle or clinging to the glass in the middle, it will affect the reading in ways that are unpredictable.

3. If you do not get all the powder into the vial you will get a false low. Again, this is impossible to quantify.

 

[CalmSeasQuest]

I provided one more days test data on the outside chance that Hanna's failure to reply thus far is result of their chemistry department taking the time to duplicate these tests to formulate a response.

5/02/12 7:00AM Initial - 12 ppb Immediate follow up tests - 3, 0, 0, 0, 0
4/30/12 7:00AM Initial - 25 ppb. Immediate follow up tests - 3, 12, 10, 4
4/27/12 7:00AM Initial - 17 ppb. Immediate follow-up tests - 4, 2, 7, 0, 6
4/26/12 7:00AM Initial - 16 ppb. Immediate follow up tests - 0, 4, 6, 5, 4
4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

The results are so consistent and repeatable, that It's hard to come up with any explanation. I also replaced the batteries (even though the unit was brand-new) on 4/25 to eliminate the possibility of faulty batteries causing the variance.

I have posed 2 questions to Hanna,

1) Which set of values is to be believed?

2) Why?

 

[rehype]

Hey Tom,
I tested my unit today so i could post my results and offer my experience. Ive always used two cuvettes..initially because the issue with the short timer but now since i like to recheck my results.

All cuvettes are wiped clean with a microfiber cloth prior to each test, each cuvette and cap is rinsed with tank water prior to testing. Care is taken to make sure all the powder agent is introduced into the cuvette. The instant re-tests are done using the second cuvette for the C1 standard sample. The one thing i do differently after having results similar to yours is after i mix the reagent for a full two minutes i let the C2 sample sit for 7 minutes before adding it to the checker.At which point it will still sit for the standard 3 minutes. So thats a total of 10 minutes before the C2 sample is read. In my experience that has helped with the variance in readings. My guess is the sample needs to settle longer before being read.

Heres my readings

5/2/12 11:20PM Initial- 11ppb Immediate follow up tests-8,9,11,11,11 and after waiting nearly 30 minutes..15

 

[CalmSeasQuest]

Hey Tom,
I tested my unit today so i could post my results and offer my experience. Ive always used two cuvettes..initially because the issue with the short timer but now since i like to recheck my results.

All cuvettes are wiped clean with a microfiber cloth prior to each test, each cuvette and cap is rinsed with tank water prior to testing. Care is taken to make sure all the powder agent is introduced into the cuvette. The instant re-tests are done using the second cuvette for the C1 standard sample. The one thing i do differently after having results similar to yours is after i mix the reagent for a full two minutes i let the C2 sample sit for 7 minutes before adding it to the checker.At which point it will still sit for the standard 3 minutes. So thats a total of 10 minutes before the C2 sample is read. In my experience that has helped with the variance in readings. My guess is the sample needs to settle longer before being read.

Heres my readings

5/2/12 11:20PM Initial- 11ppb Immediate follow up tests-8,9,11,11,11 and after waiting nearly 30 minutes..15

Thanks very much for sharing your results rehype. Very interesting, so after waiting 7 minutes, all of your tests are well within the stated error rate. While I still don't understand the mechanism, in addition to my regular testing, I followed up with a timed test conducted after allowing the cuvette to sit still for 7 minutes. The results were...

5/03/12 7:25AM 7 min - 04 ppb. Immediate follow up tests - 3, 1, 2, 6, 1
5/03/12 7:00AM Initial - 12 ppb. Immediate Follow up tests - 1, 0, 1, 3, 1
5/02/12 7:00AM Initial - 12 ppb. Immediate follow up tests - 3, 0, 0, 0, 0 (fresh GFO)
4/30/12 7:00AM Initial - 25 ppb. Immediate follow up tests - 3, 12, 10, 4
4/27/12 7:00AM Initial - 17 ppb. Immediate follow-up tests - 4, 2, 7, 0, 6
4/26/12 7:00AM Initial - 16 ppb. Immediate follow up tests - 0, 4, 6, 5, 4
4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

So my 7 minute results were much like yours, although I'm still puzzled as to why. My C2 sample was about 25 minutes old when I conducted the 7 minute test - I wonder if the reagent in solution might be breaking down and altering the results? But even that does not explain the variance between the initial and follow up tests. Today's tests were also made using a new box of reagents (lot HO13). Assuming the reagent is completely dissolved (which I believe it is) the results are hard to explain.

I would love to trust the follow up and 7 minute values as they provide more favorable results - It sure would be nice if Hanna provided some type of input.

 

[djkms]

It sure would be nice if Hanna provided some type of input.

Maybe ask a mod to move this thread to their forum?

 

[CalmSeasQuest]

Maybe ask a mod to move this thread to their forum?

This thread originated in the Hanna forum and has been updated daily. For reasons beyond my comprehension, despite responding to other threads since, Hanna has declined to provide any response or guidance on this issue. Their lack of reply was the only reason I created this compendium in the Chemistry forum with the hopes that one of our resident experts might help me and anyone else encountering similar issues figure out the cause and a solution.

 

[clydog]

Thanks very much for sharing your results rehype. Very interesting, so after waiting 7 minutes, all of your tests are well within the stated error rate. While I still don't understand the mechanism, in addition to my regular testing, I followed up with a timed test conducted after allowing the cuvette to sit still for 7 minutes. The results were...

5/03/12 7:25AM 7 min - 04 ppb. Immediate follow up tests - 3, 1, 2, 6, 1
5/03/12 7:00AM Initial - 12 ppb. Immediate Follow up tests - 1, 0, 1, 3, 1
5/02/12 7:00AM Initial - 12 ppb. Immediate follow up tests - 3, 0, 0, 0, 0 (fresh GFO)
4/30/12 7:00AM Initial - 25 ppb. Immediate follow up tests - 3, 12, 10, 4
4/27/12 7:00AM Initial - 17 ppb. Immediate follow-up tests - 4, 2, 7, 0, 6
4/26/12 7:00AM Initial - 16 ppb. Immediate follow up tests - 0, 4, 6, 5, 4
4/25/12 7:00AM Initial - 18 ppb. Immediate follow up tests - 2, 0, 0, 0, 4, 3
4/24/12 7:00AM Initial - 29 ppb. Immediate follow up tests - 11, 9, 8, 12
4/23/12 7:00AM Initial - 43 ppb. Immediate follow up tests - 5, 8, 9, 13, 8

So my 7 minute results were much like yours, although I'm still puzzled as to why. My C2 sample was about 25 minutes old when I conducted the 7 minute test - I wonder if the reagent in solution might be breaking down and altering the results? But even that does not explain the variance between the initial and follow up tests. Today's tests were also made using a new box of reagents (lot HO13). Assuming the reagent is completely dissolved (which I believe it is) the results are hard to explain.

I would love to trust the follow up and 7 minute values as they provide more favorable results - It sure would be nice if Hanna provided some type of input.

i think you got a bad unit....my 736 is very accurate/consistant....my results are always close together like rehypes

 

[CalmSeasQuest]

i think you got a bad unit....my 736 is very accurate/consistant....my results are always close together like rehypes

Thanks clydog - Have you tried the retests and if so, can you share your process and results? It's entirely possible my checker is defective, and I probably would have come to the same conclusion until rehypes post. A few things make wonder...

• I can't image a defect that would produce the results seen. Seemingly consistent and predicable initial and retest results that vary widely, despite being taken seconds apart with the same sample and...a problem that disappears by simply allowing the C2 sample to sit for 7 minutes.
  
• If I understood rehypes comment correctly, he also had similar findings, hence the development of his 7 minute process and lastly...
  
• If Hanna thought my unit defective, it seems to me they would have simply said so and offered to replace it - problem solved, rather than ignore the issue entirely.

Although I don't understand why, at this point I suspect the solution is to allow the C2 sample to rest for 7 minutes after mixing. If this is correct, I wonder if many using the HI 736 aren't getting incorrect (consistent, but higher than actual) PO4 readings when following the directions provided by Hanna?

 

[clydog]

for the first test i do it normally using one vial and the 3 min count down....after thats done i fill the other vial with tank water and test multiple times (4-6) and my numbers are always within 5 ppb and most the time within 2-3ppb and atleast 3 out of the 6 test are the same....for instance my tests would look like this

1st-12ppb,2nd-10ppb,3rd-12ppb,4th-11ppb,5th-10ppb,6th-12ppb.....so i just average the tests