KH Guardian: Automatic Alkalinity Controller- My Experiences

CoralFragZ

Well-Known Member
View Badges
Joined
Feb 14, 2012
Messages
806
Reaction score
335
Location
United States
Rating - 0%
0   0   0
While I no nothing about the accuracy of this unit, and they may be overstating the true accuracy, I'm not sure how one can conclude it isn't measuring correctly just because there are small changes in the reported tank alkalinity, unless you somehow otherwise know what the alkalinity is.

Presumably these tanks get some sort of dosing at some time (maybe not in the times shown, but that isn't clear). Further, pH will impact the answer, and even a tiny but of calcium carbonate solid (perhaps from a fish stirring up some detritus, etc.) getting into the test will impact the results, as I show below:

Assume you are using 50 mL sample size (that's what I was told, but I don't independently know that).

A 1 mm diameter grain of calcium carbonate sand has a volume of 0.0005 mL and a weight of perhaps 0.0015 grams (1.5 mg). That means it contains 1.5 mg/100 mg/mmole = 0.015 mmole of CaCO3, which has twice that in alkalinity, or 0.03 meq of alk. Put that in 50 mL and the effective concentration is 0.6 meq/L (1.7 dKH).

Consequently, a much smaller grain of sand (or many really tiny bits of calcium carbonate) can readily impact the alkalinity that is measured if you are worrying about fluctuations on the order of 0.2 dKH.

Thanks for the analysis, It is being dosed with a calcium reactor with a steady rate of effluent which typically is usually consist. The unit has check valves with filters so Id imagine that would filter out derbies. I agree the error is overstated but I guess the way is bounces kind of defeats the purpose of the automation and certainly would make me question if I want to let it control dosing with that large of an error which was an objective to counter my .5+ dkh drop with such a large stocked system.
 

Jonty

Well-Known Member
View Badges
Joined
Jul 6, 2009
Messages
581
Reaction score
477
Rating - 0%
0   0   0
For me the variation in ALK reading is okay it tells me I am in range and I have not had a result go outside of a .4 deviation between tests, most of the time it is repeatable and remarkably predictable based on time of day fluctuations. There have been many many times I have tested with my Hanna Checker and said that can not be right and checked again and I get a number in line with my expectations. I think where you have your SW in line placed may also be a factor, proximity to dosing/reactor outlets may impact consistency.
 

ebushrow

Valuable Member
View Badges
Joined
Jan 30, 2012
Messages
1,911
Reaction score
1,542
Location
Plymouth
Rating - 0%
0   0   0
I was thinking about the kh offset number and the K concentration.
When khguard is dosing drips it counts up dHK by .1 or .2. So to get the final dkh reading higher then more drops should be needed. Or a weaker solution of the acid being used for titration. So possibly I need more Rodi added to the solution. Which is strange because I used a scale that goes to hundredths when I weighed out the rodi.
I haven't heard of anyone having to use a negative offset (their solution possibly being too weak).

Even if I had used a less precise scale and typically the error is half of the smallest measurable unit. So if it was just a gram scale and measured whole numbers the maximum + - it would be off is .5g or .5ml. Once I get my new packets. I will add rodi to the K solution and test my theory.
Out of pure curiosity....do those reading that fluctuate happen to overlap with your ato adding in fresh water?....just a thought....also, is the intake tube in a turbulent flow area to prevent stagnation water?...just a thought to help...if it does...
 

CoralFragZ

Well-Known Member
View Badges
Joined
Feb 14, 2012
Messages
806
Reaction score
335
Location
United States
Rating - 0%
0   0   0
For me the variation in ALK reading is okay it tells me I am in range and I have not had a result go outside of a .4 deviation between tests, most of the time it is repeatable and remarkably predictable based on time of day fluctuations. There have been many many times I have tested with my Hanna Checker and said that can not be right and checked again and I get a number in line with my expectations. I think where you have your SW in line placed may also be a factor, proximity to dosing/reactor outlets may impact consistency.
Well I guess that is why we are paying a G-Bill for the advertised precision and automation. Most of the time Hanna is messed up with fingerprints or not cleaning it out properly. Some reefers are okay to use an API or a test strip to check alk, however some are looking at the hundredth for stability. For me using the automation takes the human error and time out of the testing and as a bonus would dose on demand which will compliment a CA/RX and the high demand uptake times, when the lights are on. All of this stems on the fact that the unit must be accurate to .1 or even .2 if possible. If I have a device that jumps from 8.2 to 8.7 in .5 hour without human intervention how can I control a .5+ dkh swing seeking the ultimate stability. You Cant? If you are checking your alk with API and paper strips this seems like a good ballpark automation but definitely not .1 accuracy for the extreme sps users out there. I am just hopping that I am wrong and that the unit I have has an issue because I really want the product to succeed.
 
Last edited by a moderator:

ebushrow

Valuable Member
View Badges
Joined
Jan 30, 2012
Messages
1,911
Reaction score
1,542
Location
Plymouth
Rating - 0%
0   0   0
Well I guess that is why we are paying a G-Bill for the advertised precision and automation. Most of the time Hanna is messed up with fingerprints or not cleaning it out properly. Some reefers are okay to use an API or a test strip to check alk, however some are looking at the hundredth for stability. For me using the automation takes the human error and time out of the testing and as a bonus would dose on demand which will compliment a CA/RX and the high demand uptake times, when the lights are on. All of this stems on the fact that the unit must be accurate to .1 or even .2 if possible. If I have a device that jumps from 8.2 to 8.7 in .5 hour without human intervention how can I control a .5+ dkh swing seeking the ultimate stability. You Cant? If you are checking your alk with API and paper strips this seems like a good ballpark automation but diffidently not .1 accuracy for the extreme sps users out there. I am just hopping that I am wrong and that the unit I have has an issue because I really want the product to succeed.
Are you able to get this accuracy doing it manually? Just checking as I have never been able to get that consistent results ( although like you said...Hanna testers can be very variable due to a host of things ). Just wondering what the die hard sps'ers are getting before this?
 

CoralFragZ

Well-Known Member
View Badges
Joined
Feb 14, 2012
Messages
806
Reaction score
335
Location
United States
Rating - 0%
0   0   0
Out of pure curiosity....do those reading that fluctuate happen to overlap with your ato adding in fresh water?....just a thought....also, is the intake tube in a turbulent flow area to prevent stagnation water?...just a thought to help...if it does...
That's a great point Eric, I have a 5 system raceway and the intake line is situated just above a maxspect gyre 150 no where near any top-off, AA dose or Effluent input. I just hope the fish dont relieve them selves by the power-head lol. Id imagine the alk would be very low.
 

ebushrow

Valuable Member
View Badges
Joined
Jan 30, 2012
Messages
1,911
Reaction score
1,542
Location
Plymouth
Rating - 0%
0   0   0
That's a great point Eric, I have a 5 system raceway and the intake line is situated just above a maxspect gyre 150 no where near any top-off, AA dose or Effluent input. I just hope the fish dont relieve them selves by the power-head lol. Id imagine the alk would be very low.
Ok...doesn't appear that there could be any issue there then....I know that sometimes I cannot see the forest for the trees when I set something up....just did it last night setting up the proflux 4 so thought I would through that out :)
 
OP
OP
Rick.45cal

Rick.45cal

2500 Club Member
View Badges
Joined
Jan 14, 2016
Messages
3,693
Reaction score
9,210
Location
Lakeland Florida
Rating - 0%
0   0   0
@Sps3Sixty will you post a pic of your settings screen? Every once in awhile I will get a larger than normal change, but you are testing at a much higher frequency than I am. I've attributed my jumps/declines to swings/changes in alkalinity.

I think something to consider is that alkalinity consumption may be a more punctuated event than we have considered in the past. It may have periods where it nearly halts and our calcium reactors/dosers keep right on going. This would be suggested by the consistently high number for a period of time until calcification resumes. So think about the swings like this. If there are periods of rapid consumption followed by periods of low consumption and these occur throughout the day, then consistent dosing/effluent from a calcium reactor is going to generate high periods and low periods.

Now considering we are choosing a low point as a minimum set point. You have to be willing to accept some upside swing (I believe, remember I have much less test data than you). Remember we aren't really picking the happy medium point where we want to keep it. We are picking a point that is the "bottom" of where you want you alkalinity. It's for these reasons that I have, at least convinced myself, that they aren't to be alarmed with and to just "let it ride".

I'm glad you and @tdb320reef came to share your experiences with us.

We're obviously dealing with many variables in complex situations, I'd like to know the answers to your questions as well. (I'm just trying to throw out possible scenarios, because I don't know the answer).
 

CoralFragZ

Well-Known Member
View Badges
Joined
Feb 14, 2012
Messages
806
Reaction score
335
Location
United States
Rating - 0%
0   0   0
Are you able to get this accuracy doing it manually? Just checking as I have never been able to get that consistent results ( although like you said...Hanna testers can be very variable due to a host of things ). Just wondering what the die hard sps'ers are getting before this?

My original claim is that I am able to get the accuracy manually with two well known kits but Unable to get accuracy with the new unit, so yes over the course of the past 6 years with large systems 400G + things do not move as fast as they do in smaller systems. With the larger volume and getting to know corals, demand vs time, you get good data in your logs where you can see patterns. That is manually. This unit should be past that and provide you with the automation and accuracy to meet demand and stability objectives that cannot be done as simple manually.
 

ebushrow

Valuable Member
View Badges
Joined
Jan 30, 2012
Messages
1,911
Reaction score
1,542
Location
Plymouth
Rating - 0%
0   0   0
My original claim is that I am able to get the accuracy manually with two well known kits but Unable to get accuracy with the new unit, so yes over the course of the past 6 years with large systems 400G + things do not move as fast as they do in smaller systems. With the larger volume and getting to know corals, demand vs time, you get good data in your logs where you can see patterns. That is manually. This unit should be past that and provide you with the automation and accuracy to meet demand and stability objectives that cannot be done as simple manually.
Ok...I agree if you can get that stability manually then this should be better!
 

Carlos@CoralVue

Active Member
View Badges
Joined
Dec 2, 2013
Messages
227
Reaction score
359
Location
Chicago, IL
Rating - 0%
0   0   0
Settings
2017-02-17%2017.59.41.jpg
30 minutes is too short of a window to expect for any additional Sodium Bicarbonate to have properly mixed in our tanks. Remember that up until now, we know that it takes a good 12-24 hours for any changes to show up after we made adjustments. The machine does not speed the process of dissipation or dilution in our tanks so to have the machine test every 30 minutes and expect an accurate reading may be a little too much to ask.

In our video, we usually recommend 2-4 hrs so there is more time for things to settle and adjustments to take effect.

Just an observation of course...
 
OP
OP
Rick.45cal

Rick.45cal

2500 Club Member
View Badges
Joined
Jan 14, 2016
Messages
3,693
Reaction score
9,210
Location
Lakeland Florida
Rating - 0%
0   0   0

Thank you!

When you look at your log for the last test, is the time that it recorded the time it actually took the test? I had to juggle my timezone number to get the recorded data point to match the system time that it displays on the front of the screen. The reason I ask is if it isn't recording the time stamp right then it's hard to tell if your high points are during the day or at night. (I apologize if they are right, but I struggled for a bit over that).
 

CoralFragZ

Well-Known Member
View Badges
Joined
Feb 14, 2012
Messages
806
Reaction score
335
Location
United States
Rating - 0%
0   0   0
30 minutes is too short of a window to expect for any additional Sodium Bicarbonate to have properly mixed in our tanks. Remember that up until now, we know that it takes a good 12-24 hours for any changes to show up after we made adjustments. The machine does not speed the process of dissipation or dilution in our tanks so to have the machine test every 30 minutes and expect an accurate reading may be a little too much to ask.

In our video, we usually recommend 2-4 hrs so there is more time for things to settle and adjustments to take effect.

Just an observation of course...
Look at his MLQ he is not dosing anything, we are both in the same boat that we cannot trust it to automate the dosing of anything until it gets accurate. The purpose of testing every .5 is to see if the system can be trusted and accurate enough to control dosing. Which I think most would agree is not based on that data. Could it be true that the error of the unit is > .5 between tests?
 

Jonty

Well-Known Member
View Badges
Joined
Jul 6, 2009
Messages
581
Reaction score
477
Rating - 0%
0   0   0
Well I guess that is why we are paying a G-Bill for the advertised precision and automation. Most of the time Hanna is messed up with fingerprints or not cleaning it out properly. Some reefers are okay to use an API or a test strip to check alk, however some are looking at the hundredth for stability. For me using the automation takes the human error and time out of the testing and as a bonus would dose on demand which will compliment a CA/RX and the high demand uptake times, when the lights are on. All of this stems on the fact that the unit must be accurate to .1 or even .2 if possible. If I have a device that jumps from 8.2 to 8.7 in .5 hour without human intervention how can I control a .5+ dkh swing seeking the ultimate stability. You Cant? If you are checking your alk with API and paper strips this seems like a good ballpark automation but definitely not .1 accuracy for the extreme sps users out there. I am just hopping that I am wrong and that the unit I have has an issue because I really want the product to succeed.

I just haven't seen the swings in results you are. I just took a 500 ml beaker of tank water and used it to run 10 tests and I had ten 7.6 results. I agree if I had to deal with those kinds of wild swing I would be very frustrated.
 

Jonty

Well-Known Member
View Badges
Joined
Jul 6, 2009
Messages
581
Reaction score
477
Rating - 0%
0   0   0
Is it possible you have air in your reagent line or a bad connection in the line resulting inaccurate dispensing of the reagent?
It might be worth taking a volume of tank water in a container purge the lines and run as series of tests as I did and see if the inconsistency is still occurring
 
OP
OP
Rick.45cal

Rick.45cal

2500 Club Member
View Badges
Joined
Jan 14, 2016
Messages
3,693
Reaction score
9,210
Location
Lakeland Florida
Rating - 0%
0   0   0
@Sps3Sixty Your recorded log times are inaccurate. (just FYI) it isn't the 18th until tomorrow. (You're testing in the future:D) I think you need to enter a -5 into your timezone for EST.

That will at the very least make your log match up with what time of day the test is occuring at.
 

CoralFragZ

Well-Known Member
View Badges
Joined
Feb 14, 2012
Messages
806
Reaction score
335
Location
United States
Rating - 0%
0   0   0
@Sps3Sixty Your recorded log times are inaccurate. (just FYI) it isn't the 18th until tomorrow. (You're testing in the future:D) I think you need to enter a -5 into your timezone for EST.

That will at the very least make your log match up with what time of day the test is occuring at.

It is 7 for est.
 
Back
Top