Thales
Past President
Sanjay and I did a thing and wrote it up. You might like it. The Conclusion is written in english, the rest is math and science. Thanks!
Hobby ICP study - Evaluating the Repeatability and Differences between ICP Tests
- Thread starter Thales
- Start date
Love it! The one that stuck out the most to me is the CV for iodine, given how important it is. We're already told that the hobby grade test kits for iodine are not accurate enough and ICP is the way to go, but this makes me think twice. For me, iodine is always one of the elements that gets flagged as out of range in ICP tests.
I don't do the moonshiners method, but it would drive me crazy trying to target a certain number while knowing how much variability between icp vendors and margin of error there is for trace elements.
I don't do the moonshiners method, but it would drive me crazy trying to target a certain number while knowing how much variability between icp vendors and margin of error there is for trace elements.
Last edited:
Wow, crazy to see that they are showing such a difference for Phosphate and Salinity between Vendors. How can one show a mean phosphate of .14 and the other .02? Those are drastically different.
Regardless of the trace elements, those are two parameters that I would expect to be "dead on" between vendors. Very disappointing.
Regardless of the trace elements, those are two parameters that I would expect to be "dead on" between vendors. Very disappointing.
Blaise006
Guest
Phosphate seems like it makes the most sense in my opinion. That sample would be highly contingent on any residue food during the sample process. Even samples taken moments apart could capture a different amount of particulates. Probably totally off base there but I’ve personally experienced over multiple po4 samples thru out the day of po4 ranging between .06-.15 within 24 hours. I eventually just concluded it had to do with particulate matter being caught.
But it calls into question the trace elements and other results for sure.
derek_SR
Supporting Member
Totally agree. I wonder how much of the success associated with moonshiners (seems like more and more people are using it!) is related to simply adding trace frequently?
A lot of the common trace elements apparently get used up so quickly that regardless of the “target” level there may be value in simply adding a little bit each day so it’s available.
This is kind of my approach with trace dosing.
archibaldsharper
Supporting Member
Thanks for doing this experiment.
I kind of run the moonshiners but mix my own elements and do a lot of water changes to remove detritus. I have used the Triton and oceamo (both icp and icp/ms). For the most part the corrections seem within range. In that I might do Triton one quarter, make corrections, then use oceamo. The corrected elements come out within reasonable range.
The iodine number always seems strange to me though. I dose iodine and test it weekly with mastertronic using redsea reagents. I think the test maxes out at .09 mg/L. For a while I was dozing 2ml per day and had the iodine maxed out on the test. But both Triton and oceamo show it as very low or depleted. I have also had the mastertronic showing low reading, and the icp was still in the same range. I kind of wonder if the mastertronic is even testing iodine or something else.
The mastertronic seems to represent a number that decreased when I don't dose, and increases when I do. But I have no idea how that number relates to how much is actually in the water. And it doesn't relate to the icp tests.
I'm kind of guessing that if there is at least some iodine, and being added every day that that should be plenty. I forget a lot and need to set up a doser for it.
I have been happy that my salifert potassium tests generally agrees with results from both icp companies.
I kind of run the moonshiners but mix my own elements and do a lot of water changes to remove detritus. I have used the Triton and oceamo (both icp and icp/ms). For the most part the corrections seem within range. In that I might do Triton one quarter, make corrections, then use oceamo. The corrected elements come out within reasonable range.
The iodine number always seems strange to me though. I dose iodine and test it weekly with mastertronic using redsea reagents. I think the test maxes out at .09 mg/L. For a while I was dozing 2ml per day and had the iodine maxed out on the test. But both Triton and oceamo show it as very low or depleted. I have also had the mastertronic showing low reading, and the icp was still in the same range. I kind of wonder if the mastertronic is even testing iodine or something else.
The mastertronic seems to represent a number that decreased when I don't dose, and increases when I do. But I have no idea how that number relates to how much is actually in the water. And it doesn't relate to the icp tests.
I'm kind of guessing that if there is at least some iodine, and being added every day that that should be plenty. I forget a lot and need to set up a doser for it.
I have been happy that my salifert potassium tests generally agrees with results from both icp companies.
Attachments
Love seeing real data on this stuff!
Kensington Reefer
Supporting Member
Who’s calibrating the instruments?!
Is their technique proper?
Thanks Rich
Is their technique proper?
Thanks Rich
Thanks for doing and sharing this Rich (and Sanjay).
I have a few observations:
1. Methods and statistics seem solid. Thank you for doing this and doing it well.
2. It’s too bad the tested water was free of most of the trace unwanted elements so we don’t have data for a lot of the elements we would worry about if we saw them. Spiking the sample with trace amounts of some of these would have been interesting.
3. I can see how not naming the companies seemed like the right choice, but it makes it much less useful. Not only the obvious issue that we want to know if this is relevant to our ICP test of choice, but also because depending on which companies you picked you could seriously skew the results. For example, it is conceivable that you would pick low-cost (perhaps low-quality) ICP tests since you had to buy so many tests. If the results are to be interpreted that the worst, lowest cost ICP tests aren’t reproducible, then that’s a lot less useful. Alternatively you could have picked the 3 market leaders in the US with the most experience including expensive ones, in which case I’d sit up and pay attention more. Since I don’t know, one interpretation of this data is that I’m probably getting what I’m paying for when I spend more… which I’m guessing is not what you want the take-away to be.
4. Also by not naming companies, you are disincentivizing them to step forward and rebut and show their own data in response. Since one of the specific issues you call out is the lack of them showing their own data or discussing this publicly, this seems like a missed opportunity.
5. Very interesting that P/PO4 are less consistent than others. Since ICP doesn’t measure phosphate, but rather phosphorus, it makes sense that they track together since they just measure P and then extrapolate to PO4 with an educated guess, according to my understanding.
I have a few observations:
1. Methods and statistics seem solid. Thank you for doing this and doing it well.
2. It’s too bad the tested water was free of most of the trace unwanted elements so we don’t have data for a lot of the elements we would worry about if we saw them. Spiking the sample with trace amounts of some of these would have been interesting.
3. I can see how not naming the companies seemed like the right choice, but it makes it much less useful. Not only the obvious issue that we want to know if this is relevant to our ICP test of choice, but also because depending on which companies you picked you could seriously skew the results. For example, it is conceivable that you would pick low-cost (perhaps low-quality) ICP tests since you had to buy so many tests. If the results are to be interpreted that the worst, lowest cost ICP tests aren’t reproducible, then that’s a lot less useful. Alternatively you could have picked the 3 market leaders in the US with the most experience including expensive ones, in which case I’d sit up and pay attention more. Since I don’t know, one interpretation of this data is that I’m probably getting what I’m paying for when I spend more… which I’m guessing is not what you want the take-away to be.
4. Also by not naming companies, you are disincentivizing them to step forward and rebut and show their own data in response. Since one of the specific issues you call out is the lack of them showing their own data or discussing this publicly, this seems like a missed opportunity.
5. Very interesting that P/PO4 are less consistent than others. Since ICP doesn’t measure phosphate, but rather phosphorus, it makes sense that they track together since they just measure P and then extrapolate to PO4 with an educated guess, according to my understanding.
When I was first starting up, my buddy at the local PetCo hooked me up with a couple bottles of the Fluval trace elements. That's all I have so that's what I've been using after every water change. I don't have a way to test for them, other than choosing to do ICP tests. But to me, atm, it's not necessary for any of my tanks except the DT I am having issues with. I don't even know what moonshiners is! I've been trying to just focus on the basics and slowly expand the knowledge base as I get more comfortable.
Thales
Past President
I think it has to do with paying attention. Any methodology that gets you to pay attention, likely results in a better system overall. If you do one where you don't even have to think, just do what they tell ya, may work well up to a point. Same with tank resets
Thales
Past President
Thanks for reading it and thinking about it!
I am sick - damn you denver and san diego - so if anything seems off or terse, it isn't me!
Rock
I get it. This test was aimed at a more practical approach rather than at looking at something a hobbyist wouldn't do. We wanted to see what a customer would see, at real levels.
At the time, we didn't have the funds to test everyone, so we did what was in the wheelhouse of doable - balanced with tthe idea that we were pretty sure the results would be what they turned out to be, so we kinda also don't want to pour money into companies to help them make more money. A little bit of that.
Mainly, we don't do names in this one because it wasn't what we were trying to do. Naming takes away from the data and converts people to looking for the 'Best' results.
I don't think there is anything in the data we provide that would lead one to conclude that more expensive testing is 'better'
We tried to pick a nice cross section, but everyone has a different opinion which are market leaders and which arent'. We were actually going to do two companies, but decided that wouldn't be enough, so we added one more.
I think we have lost that fight. They aren't going to publish if they don't want to. The only one that has is Oceamo, but it isn't useful to most people and they are out of the US so I don't want to use them at all. But again, this study wasn't about which company did what, this was about what an end user might reasonably expect.
This kind of thing pisses me off, not you, them. It can be very hard to tell what they are doing with the ICP and what they are doing with other tests. It makes me really grumpy.
I hope that all makes some kind of sense!
derek_SR
Supporting Member
Looks like you are no longer dosing lanthanum?
Even dosing low levels directly into skimmer (imitating your method) I still see it show up on ICP.
I’m guessing your po4 tests a lot higher than .245 also?
Thales
Past President
Ha! Yeah, still dosing Lanth in the skimmer, and my po4 comes back currenlty at .69. Lithium is way down - that must be Bob's magic waterLooks like you are no longer dosing lanthanum?
Even dosing low levels directly into skimmer (imitating your method) I still see it show up on ICP.
I’m guessing your po4 tests a lot higher than .245 also?
Kensington Reefer
Supporting Member
That dude Bob….
He scores me the best SW
Chromis
Supporting Member
Thanks @Thales for sharing this study! The mathematical rigor really helps with perspective.
Did you send out sterilized salt water samples? I wonder if there is a difference between samples that have been sitting in the collection vial for different amounts of time - I usually overnight my tests to ICP-analysis and they test them same day, so that’s 24 hours of “sit” time, vs other samples could feasibly sit in the vial for 2 weeks before testing. I guess it’s hard to predict when a company will measure your sample, but maybe one variable could be removed by either sterilizing the sample or letting the samples sit longer before shipping to the higher turn-around companies?
I’m curious how much phosphate and trace could get fixed by biofilms that form on the sides of the vial or taken up by bacteria/plankton that may then get filtered out before they run the test? (I am assuming there is some mechanical sample filtration prior to ICP testing - or is everything including small particulate matter getting vaporized and counted in the ICP?) I suppose then some of the differences could be explained by different pre-filtering or handling process.
Did you send out sterilized salt water samples? I wonder if there is a difference between samples that have been sitting in the collection vial for different amounts of time - I usually overnight my tests to ICP-analysis and they test them same day, so that’s 24 hours of “sit” time, vs other samples could feasibly sit in the vial for 2 weeks before testing. I guess it’s hard to predict when a company will measure your sample, but maybe one variable could be removed by either sterilizing the sample or letting the samples sit longer before shipping to the higher turn-around companies?
I’m curious how much phosphate and trace could get fixed by biofilms that form on the sides of the vial or taken up by bacteria/plankton that may then get filtered out before they run the test? (I am assuming there is some mechanical sample filtration prior to ICP testing - or is everything including small particulate matter getting vaporized and counted in the ICP?) I suppose then some of the differences could be explained by different pre-filtering or handling process.
Awe shit. First ones free. The next one will cost ya. Then you just became his customer. Unfortunately I can’t store 600 gallons. Sad guy.
Genius. That was my argument too. Then dang microbes are doing something in that tube when it’s on the tour of the world.