SmartDoser: Predicting Alk and Calcium

[dandelion]

Just wondering how everyone on here estimate their alk and calcium consumption for dosing, especially for people with dosers? Do you just take a few readings everyday and then average them out? How often do you reevaluate your estimated consumption rate?

 

I am currently programming a "SmartDoser" that does all that for you. Every time you test you punch in the results, and it will automatically calculate the levels several times a day and dose back to your target levels. Just want to see if people are interested in this idea. I may post my build and program in the DIY section.

[smeagol108]

It's pretty simple. Take a measurement day one. Than take the same measurement again the next day at the same time and see what the consumption was. That's how much u need to dose. U really don't need to test this over multiple days. If u want to dose what u should based on consumption and than check again the next day at same time. Ur consumption should be the same daily.

[xmas_one]

I test alk once a week before WC. If alk is low, I increase my dose. I don't test for calcium, just dose it the same amount as alk.

[dandelion]

It should be the same, but that's theoretical. There can be testing error, precipitation, consumption changes due to growth or temperature changes.... blah blah blah. Getting multiple data points makes your model more accurate.

 

 

It's pretty simple. Take a measurement day one. Than take the same measurement again the next day at the same time and see what the consumption was. That's how much u need to dose. U really don't need to test this over multiple days. If u want to dose what u should based on consumption and than check again the next day at same time. Ur consumption should be the same daily.

[dandelion]

Theoretically consumption of alk and calcium should be the same but my test results seem to disagree.

 

I test alk once a week before WC. If alk is low, I increase my dose. I don't test for calcium, just dose it the same amount as alk.

[Clown79]

Test 2 days in a row to determine consumption. Dose according to that.

 

This may need to be reevaluated as the corals grow etc because consumption changes

[dandelion]

Problem is 2 days in a row didn't seem to give me accurate enough of a consumption rate. That's why I'm including a 2-day moving average in my programming to collect more data points and account for coral growth.

 

Not sure if my inconsistencies are from having a small tank or what.

 

Test 2 days in a row to determine consumption. Dose according to that.

 

This may need to be reevaluated as the corals grow etc because consumption changes

[teenyreef]

Problem is 2 days in a row didn't seem to give me accurate enough of a consumption rate. That's why I'm including a 2-day moving average in my programming to collect more data points and account for coral growth.

 

Not sure if my inconsistencies are from having a small tank or what.

 

Yep, that's one of the challenges with a small tank, the parameters don't stay very stable from one day to the next. If you dose in a small tank, say 20g or less, I've found it's necessary to check alk at least every other day. Over time you can come up with an average dose, but minor changes in conditions can mess it all up in a couple days, so you can't just let it run.

[dandelion]

I have been testing daily and my prediction model gives me an error rate of +-0.3 which doesn't sound too bad to me. Plus my corals never seem to be bothered with alk swings.

 

 

Yep, that's one of the challenges with a small tank, the parameters don't stay very stable from one day to the next. If you dose in a small tank, say 20g or less, I've found it's necessary to check alk at least every other day. Over time you can come up with an average dose, but minor changes in conditions can mess it all up in a couple days, so you can't just let it run.

[teenyreef]

That sounds about right, as long as the dosing level is on track.