PDA

View Full Version : determining alkalinity via pH and co2*


orcafood
09/08/2017, 04:53 PM
Given that the pH of a reef tank is being read at a very accurate level, is it not feasible to determine the co2* gas dissolved in a reef tank via a small IR device, then plug and chug the salinity and temperature into equilibrium equations to get the K0, K1, K2, effectively giving the amount of HCO3- and CO3--?

https://www.co2meter.com/products/k-30-co2-sensor-module?gclid=EAIaIQobChMIxJLt3NWW1gIVRUCGCh2fuwZBEAQYAiABEgIQQ_D_BwE

How about measuring the ppm of co2 gas dissolved in a head space only in contact with tank water. A relationship must exist between the headspace co2 and the amount dissolved into the tank, though it might differ a little bit from the expected Henry's laws.

dartier
09/08/2017, 05:06 PM
Interesting idea. Too bad that the sensor is not able to directly measure the dissolved CO2 directly in water as opposed to the content in a gas. Perhaps Jonathon can say if the CO2 level (in a contained headspace) would be an equilibrium function like PH.

Dennis

karimwassef
09/08/2017, 05:33 PM
yes. done it

http://www.reefcentral.com/forums/showthread.php?t=2628107&highlight=alkalinity&page=7

Here's the spread over the 5 variables: Temp, Salinity, Pressure, pH, and CO2

<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/1_zpsf8ljxsma.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/1_zpsf8ljxsma.jpg" border="0" alt=" photo 1_zpsf8ljxsma.jpg"/></a>

For each chart where one variable is being changed vs. CO2 to determine Alk, the fixed variables are based on the base case:
<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/0_zps6npgbmvm.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/0_zps6npgbmvm.jpg" border="0" alt=" photo 0_zps6npgbmvm.jpg"/></a>

If this is right, it would be possible to make an alkalinity monitor that measures air pressure, water temperature, CO2 in the air, salinity and pH and constructs the dKH...

This would be a continuous monitor over time... All variables except for air pressure and CO2 are already being measured by my Apex.

There's an excel file on page 2

orcafood
09/08/2017, 05:52 PM
I read the exact same paper haha

bertoni
09/08/2017, 06:52 PM
If my memory is correct, there are meters to measure carbon dioxide in water samples, although I'm not sure that they are useful in saltwater. In theory, this can be done, but in practice, it's going to be costly or inaccurate, depending on how much you want to spend. I haven't looked into the technology for a while, though. There might be some better alternatives available.

karimwassef
09/08/2017, 08:42 PM
I'm using atmospheric CO2 in the air going into my heavy air injector.
My formula uses atmospheric CO2 anyways. :)

orcafood
09/09/2017, 05:22 AM
Excellent job on your alkalinity detection system Karim! I bet a kalman filter would really help to get a more stable value. Hmm or a machine learning algorithm would probably help too.

karimwassef
09/09/2017, 09:55 AM
I think it works for me because of my air injector. I push gallons of air into my water constantly through a very fine bubble foamer. It's a dual penductor fed by outside fresh air through a high pressure pump. This means that the time variable for the diffusion of gas into water is reduced significantly. Basically, I force the assumption that I'm at steady state to be more true.

Air CO2 sensors are only ~ $100. You could, in theory, drive it from an Apex and have all the data you need.

Unfortunately, Apex won't let you enter a math formula that returns a function of known values :( but this app works for me: iFxCalc iFxCalc:

https://itunes.apple.com/us/app/ifxcalc-most-human-nature-function-calculator-support/id786018193?mt=8

It's clunky to figure out but easy once you do

The formula is
Alk=(2*D3*0.0334211*C3/29.41*0.000001*10^(-(-8.712-0.00946*F3+0.0000856*F3^2+1355.1/(((E3-32)*5/9)+273)+1.7976*ln((((E3-32)*5/9)+273))))/10^(-G3)*10^(-(17.0001-0.01259*F3-0.000079334*F3^2+936.291/(((E3-32)*5/9)+273)-1.87354*ln((((E3-32)*5/9)+273))-2.61471*F3/(((E3-32)*5/9)+273)+0.07479*F3^2/(((E3-32)*5/9)+273)))/10^(-G3)*44.01+D3*0.0334211*C3/29.41*0.000001*10^(-(-8.712-0.00946*F3+0.0000856*F3^2+1355.1/(((E3-32)*5/9)+273)+1.7976*ln((((E3-32)*5/9)+273))))/10^(-G3)*61.02)/2*1000/17.9

Where Alk(D3, C3, F3, E3, G3) has
D3 = barometric pressure (mmHg) = 30
C3 = CO2 (ppm) = 450
F3 = salinity (ppt) = 34.5
E3 = temperature (F) = 77.5
G3 = pH = 8.3

Returns 9.8 dKH

Can machine learning reduce the impact of measurement errors- probably but you'll need an accurate and frequent Alk measurement to feed into it.

orcafood
09/11/2017, 07:13 AM
So you assume the CO2 is always the same? Or that the CO2 concentration is equal to that outside and you measure it? The CO2 concentration in the air must change with temperature and humidity?

What I am suggesting would help get more accurate results would be some form of anomaly detection or outlier rejection system but you would need to use a PC with an I/O board and some programming language.

karimwassef
09/11/2017, 09:34 AM
I measure CO2 in the air. Since I use outside air, and heavily aerate with it, I don't see much change over time.

Using fresh air has been key to my stability. I still measure it though.

orcafood
09/11/2017, 12:59 PM
Thanks for your time Karim!

dartier
09/11/2017, 01:10 PM
Karim, do you use a calcium reactor? Just curious if your alkalinity calculation technique works in the presence of injected CO2 from things like CalRx?

Dennis

karimwassef
09/11/2017, 01:47 PM
I don't. I dose Kalk.

However, the air injection should neutralize the effect of any CO2 that leaks through. I'm injecting gallons of fresh air. I wouldn't trust this method if I didn't.

For reference, I first used my injector when I was curing concrete in the tank. It took pH down from 11 to 8 ina couple of hours and kept it there as the concrete cured. That's when I realized the power of fresh air injection for stability.

dartier
09/11/2017, 02:07 PM
For reference, I first used my injector when I was curing concrete in the tank. It took pH down from 11 to 8 ina couple of hours and kept it there as the concrete cured. That's when I realized the power of fresh air injection for stability.

So as to not take this thread off topic, do you have any threads about your air injection system?

Dennis

orcafood
09/11/2017, 03:13 PM
You only dose kalk? How does your pH vary over the day?

karimwassef
09/11/2017, 04:41 PM
Yes. It's basically my skimmer injectors.. without the cup or column

I built a double penductors injected skimmer that's 12' tall. It was a beast. Now it's powered by the jebao pressure pump DCP-18000

http://www.reefcentral.com/forums/showthread.php?t=2634328

http://www.reefcentral.com/forums/showthread.php?t=2637568

http://www.reefcentral.com/forums/showthread.php?t=2571877

karimwassef
09/11/2017, 04:44 PM
My pH doesn't vary at all because my pH aligns to my Alk and I use a pH feedback loop to dose Kalk

Results speak for themselves :D

<a href="http://s1062.photobucket.com/user/karimwassef/media/4A261F16-3008-4CFA-B046-391D6D9C3977_zps6ftrjkvz.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/4A261F16-3008-4CFA-B046-391D6D9C3977_zps6ftrjkvz.jpg" border="0" alt=" photo 4A261F16-3008-4CFA-B046-391D6D9C3977_zps6ftrjkvz.jpg"></a>

orcafood
09/12/2017, 08:25 AM
This is exactly what I was planning on doing haha, all I really want to do is get my pH up and very constant, it was then when I started thinking about calculating alkalinity from the CO2. Amazing that your just doing kalk, I will have to give it a shot. You are using two pH probes though right? I was thinking about getting one of the really nice ones that goes out to hundredths or thousandths meant for continuous monitoring.

How much resolution does your pH setup have? your pH probably still varies a little but you can't detect it.

Very nice color and growth tips on those corals!

karimwassef
09/12/2017, 01:48 PM
Yes. Must have two and look for volatility on either... stopping due to a false positive is better

Horace
09/13/2017, 10:16 AM
I must admit I am a bit shocked you are able to dose enough kalk to maintain your alkalinity. I assume you are doing it via topoff? Unless you are able to dose the powder form, accurately, and in a way that does not cause issues, I found in the past I didnt evap enough water to allow for enough kalk to keep alk up on its own.

karimwassef
09/13/2017, 10:53 AM
No. Dosing on a feedback loop based on pH that is pegged to my Alk.

karimwassef
09/13/2017, 10:55 AM
I consume a lot of kalk. No vinegar. I run a hot tank so evaporation is very high anyway.

dartier
09/13/2017, 11:00 AM
Karim, your CO2 monitor is monitoring the outside CO2 level or the level around your tank? If outside, I guess the amount of air you are pushing prevents an elevated indoor CO2 level from depressing the PH too much.

That you are able to do that in a controllable fashion is fairly impressive.

Dennis

karimwassef
09/13/2017, 11:37 AM
I'm measuring the intake air since I'm forcing that in.

The air around the tank aligns.

I don't use the CO2 measure to run anything. It's just to confirm where I am.

dartier
09/13/2017, 12:46 PM
Ah, so you are part of the feedback loop. :lmao:

Dennis

karimwassef
09/13/2017, 01:31 PM
A very slow correcting function...

karimwassef
09/13/2017, 01:36 PM
Word of warning - I have crashed my tank when I took my air injector offline for maintenance for too long.

It serves a key function in eliminating the bias due to CO2 fluctuations in a closed space. Like a limiting diode, it keeps the system stable by avoiding excursions that aren't real.

All it took was one day where my normally 400-500ppm CO2 to get up to 1000 for a couple of days - and my pH got pushed too low, driving my loop to inject so much Alk that it jumped from 8.5 to 11... predictable results after that... :(

Everything is back in good shape now :)

dartier
09/14/2017, 04:53 AM
Word of warning - I have crashed my tank when I took my air injector offline for maintenance for too long.

It serves a key function in eliminating the bias due to CO2 fluctuations in a closed space. Like a limiting diode, it keeps the system stable by avoiding excursions that aren't real.

All it took was one day where my normally 400-500ppm CO2 to get up to 1000 for a couple of days - and my pH got pushed too low, driving my loop to inject so much Alk that it jumped from 8.5 to 11... predictable results after that... :(

Everything is back in good shape now :)

Hang on, this almost sounds like the type of control I had originally assumed. Does this mean that when the injectors were off line, you (closing the loop) upped the kalk dosing too high based on faulty readings?

Are those photos from before or after the crash? If after, that is pretty impressive, as the crash was not that long ago, and that is some serious growth.

Dennis

karimwassef
09/14/2017, 06:17 AM
Yes. That's what happened because pH alone (without the fresh air injector) does not align to Alk. To force the pH-Alk relationship and use the feedback loop, CO2 needs to be kept stable through the fresh air injection.

The pictures are ~ 3 months before the crash. We're selling the house so I've been selling the corals and fish ... so no new pictures but I can take some of the corals left in the tank.

Once I reestablished the air injection, the tank bounced back quickly.

orcafood
09/14/2017, 07:31 AM
I still don't see why we don't try and get the actual CO2 level in the water or from a head space floating on the water surface. Plus then it could be applied to others tanks who can't assume steady state equilibrium.

I think I will set it up on my pc similarly to how you did with myself as feedback. Definitely the way to do it.

karimwassef
09/14/2017, 07:41 AM
I do monitor the CO2 of the intake air and I've confirmed that it's the same in the air around the submerged injector.

If I use internal house air, I'd have to constantly adjust my loop manually.
If I use external fresh air, there's little to no variability and I can just use pH.

The injector removes one major variable (CO2) leaving pH as the controlling variable to Alk.

orcafood
09/14/2017, 08:14 AM
Yup but I don't think many people have as much aeration as you. The CO2 levels outside do change over day vs night and seasonally so that variability should still be there.

I think I am going to attempt to use an IR CO2 monitor to float on my sump and give a real time read out of the CO2 levels dissolved in a small headspace. Then a .001 pH monitor to get a nice resolution on the pH. My salinity and temperature are rock solid and can be removed from the equation. Should work nicely!

All I want is for the system to text me when my approximated alkalinity is low. Then I will go in and manually raise the dosing if it is actually necessary. Controlling pH seems like a risky risky business to me.

karimwassef
09/14/2017, 08:40 AM
I've been doing it for four years. The only variable is CO2 fluctuation. With two pH probes, I remove the effect of any one individual probe going off. I have built in triggers to shut it off in case the two diverge.

One friend locally, Danny, has started doing the same and he's seeing similar success.

Actually, fresh air CO2 doesn't vary significantly day-night or seasonally based on my observations. 99% of the time, it's between 400-450ppm.

My extremes have been 384 and 478 over a year's worth of observation.

My data could be off, but that's what I see.

One idea that came from Danny was to actively control CO2 by injecting it (or not) to actively maintain its level. I'm wary of two loops controlling two main variables to drive a third unmeasured, but calculated final variable. Opportunity for error and instability seems very high.

dartier
09/14/2017, 10:59 AM
I still don't see why we don't try and get the actual CO2 level in the water or from a head space floating on the water surface. Plus then it could be applied to others tanks who can't assume steady state equilibrium.

I think I will set it up on my pc similarly to how you did with myself as feedback. Definitely the way to do it.

I would be very curious how this works out (CO2 sensor in a head space). The sensor you linked is pretty low cost, and has some easy to leverage interfaces (like I2C).

Dennis

dartier
09/14/2017, 11:03 AM
This looks interesting http://www.pro-oceanus.com/mini-co2.php

Dennis

karimwassef
09/14/2017, 11:21 AM
depends on the price

orcafood
09/14/2017, 12:30 PM
Karim that is exactly the kind of changing CO2 levels I am talking about and that is just outside your house, not to mention the flux inside your house. I am pretty convinced that the amount of co2 in my reef tank has a massive control over coral growth and I am very curious to start watching the levels.

Dartier good find on that. I wouldn't be surprised if they aren't using a thin silicone membrane to get very accurate head space results.

:) Damn stuff costing money :)

ATR-IR with a small slit built for constant flow circulation from the tank would be ideal but prohibitively costly. I am pretty sure dissolved CO2 meters are just normal IR meters with a decent water resistant membrane.


http://www.plantedtank.net/forums/20-diy/955898-co2-sensor.html

Look at this, planted tank people are on it.

In a way, growing coral is a lot like growing plants. Gotta keep the zooxanthellae happy.


I have been thinking that injecting O2 from a solenoid might be good for solving the CO2 controll issue. If you pump more oxygen into the water, will the CO2 not be "forced out". Pumping O2 into a tank should raise the pH?

I think pumping CO2 into a tank is only a good thing if you are going to melt corals with it (Ca Rx). I am currently battling too high a CO2 level from too many people in the house. At least that is what I think the problem is.

If pH is controlled but not CO2, alkalinity will vary with the CO2. I am thinking similarly to your friend. Whenever CO2 gets too high, blast oxygen into the aquarium with a ceramic diffuser to push out the excess CO2. If CO2 and pH are controlled ....

dartier
09/14/2017, 12:36 PM
depends on the price

Ya, about that. $6,400 to $9,450 US. Doh.

Karim that is exactly the kind of changing CO2 levels I am talking about and that is just outside your house, not to mention the flux inside your house. I am pretty convinced that the amount of co2 in my reef tank has a massive control over coral growth and I am very curious to start watching the levels.

Dartier good find on that. I wouldn't be surprised if they aren't using a thin silicone membrane to get very accurate head space results. Damn stuff costing money :)

ATR-IR with a small slit built for constant flow circulation from the tank would be ideal but prohibitively costly. I am pretty sure dissolved CO2 meters are just normal IR meters with a decent water resistant membrane.

You are correct. It is a semi-permeable membrane over a gas head space. If we could come up with a resistant, safe, semi permeable membrane of our own, the IR unit you posted earlier would be applicable for the same thing.

Dennis

bertoni
09/14/2017, 12:45 PM
Ouch, that's a pricey probe. That's one of the issues with measuring the parameters of our tanks, though. The price can be high.

orcafood
09/14/2017, 01:16 PM
I'd be surprised if we can't find an appreciable membrane. That said too, why use a membrane. Why not just make a perfectly big enough channel in a floating styrofoam boat with not membrane and just a vertical headspace to get some accurate results?

dartier
09/14/2017, 01:38 PM
I was thinking about the idea of a head space open to the water and would think that condensation may be an issue. Well that and salt air corroding the device. The manufacturer of the inexpensive one that was posted, has an option for a cap for remote sensing along with a small pump and filter to prevent moisture from fouling the sensor. You may be able to use that with a confined headspace and get around potential moisture problems.

Dennis

karimwassef
09/14/2017, 01:46 PM
That's why I use fresh air. When you run the numbers, the fresh air variation impact is small and slow. I would not do it with indoor air.

Also, photosynthesis by day consumes massive CO2 and injecting external air actually helps offset this - basically providing just what is needed to be stable.

karimwassef
09/14/2017, 01:48 PM
I would not run two loops.
I would not inject CO2.

Fresh air is substantially stable and my tank's growth rate shows it.

orcafood
09/14/2017, 01:59 PM
I would never run any loop ;)

I also want to know the amount of CO2 in my water, very interesting. As you have seen it is heavily related to Fresh Air Exchange (FAE) (anyone grown mushrooms before hehe). Which is probably related to the amount of CO2 in the air. I think we are right on board with exactly the same ideals, glad your tank is doing so well, my tanks not worthy!

Wow Dennis nice find again

dartier
09/14/2017, 02:23 PM
That's why I use fresh air. When you run the numbers, the fresh air variation impact is small and slow. I would not do it with indoor air.

Also, photosynthesis by day consumes massive CO2 and injecting external air actually helps offset this - basically providing just what is needed to be stable.

I am not sure I understand what you mean by indoor air. I think what orcafood is proposing is using a trapped air pocket, like a submerged open ended pipe (protected from bubbles), with one of the infrared CO2 sensors sampling that air column. This would work out the same as the probe I posted earlier, as that is the premise of that unit, though the gas permeable membrane makes it impervious to water.

As for the air injection, I think it is a great idea and would support doing that in addition to trying to measure the CO2 content directly from the water. You could even run a second CO2 sensor sampling the incoming air to again detect a sensor going bad, or an unexpected tank event taking place (like your dual PH probes do for your implementation).

Dennis

dartier
09/14/2017, 02:28 PM
Ok, now we are getting somewhere.

Take a look at this: http://www.kordon.com/kordon/products/aquarium-pond-accessories-2/breathing-bags#suppliers!

What if we used a piece of one of these bags to seal our sensor from water? That should remove the risk of water incursion. Though I would suggest putting the sensor above the waterline with the pipe extending below by a few inches.

Dennis

dartier
09/14/2017, 02:41 PM
I am reading a thread on a planted tank forum about using this sensor in the exact same fashion as orcafood is proposing. That is where I stumbled upon the breather bags. I have not got to the end of the thread yet to see if it end up working as hoped, or was a bust.

Dennis

orcafood
09/14/2017, 02:42 PM
Breather bags is a great idea

lol I read that same thread but never read further to the guys forum, tell me if you find anything more!

bertoni
09/14/2017, 03:40 PM
The bags could be useful. I don't know what the velocity of the gas exchange would be, but someone might be able to measure that.

karimwassef
09/14/2017, 03:43 PM
I meant that I wouldn't trust pH without injecting fresh air to stabilize.
Allowing the tank's pH to be driven by indoor air CO2 volatility would be a disaster imo.

Measure CO2 or not, I would use fresh air to reduce volatility.

karimwassef
09/14/2017, 03:52 PM
I like the idea of measuring in-water CO2 as well as the air being injected.

It would help optimize my air injection. I'm currently running on maximum pressure ~30ft through two penductors with a massive vent fan to pressurize the external air through an 8" duct... overkill! But I know that I'm getting gallons of fresh air heavily aerating.

dartier
09/15/2017, 05:05 AM
I read the thread on the planted forum, both the one I found and the one that orcafood posted. They were by the same person. In both cases he was able to measure the amount of CO2 in his FW tank using the inexpensive IAQ CO2 sensor.

The 2 challenges he had was keeping the sensor dry, he was mounting it inside the housing with a silicone membrane and putting it below the water surface.

My suggestion would be to use the remote sensing cap that they offer and use an aqualifter pump to setup a closed loop through the sensing head. The pump offered from the vendor is quite pricey. For the sensing head, a 1/2" PVC tee, with the inline legs sealed with a 1/2" to 1/4" NPT adapter so that a John Guest fitting can be screwed in for each side of the loop.

For the 3rd leg of the tee, the membrane needs to seal this portion. I am still thinking about how to do this and make it maintainable. I am thinking a union that can be tightened on with the membrane trapped between the O-ring and the union face may be all that is needed. The breather bags were slow to pass the CO2 and reach equilibrium, but for our needs, that may be fine.

Lastly, they found the sensor was much quicker to update the lower below the water surface it was positioned. So for safety it would be nice to be able to detect a leak for water incursion and have the pump not pump it through the sensor. Still thinking about how to do this without impacting the internal air volume in a significant way.

Dennis

orcafood
09/15/2017, 05:18 AM
I like the idea of measuring in-water CO2 as well as the air being injected.

It would help optimize my air injection. I'm currently running on maximum pressure ~30ft through two penductors with a massive vent fan to pressurize the external air through an 8" duct... overkill! But I know that I'm getting gallons of fresh air heavily aerating.

It would be very interesting to see how stable the levels of dissolved CO2 are in your setup

karimwassef
09/15/2017, 06:20 AM
I would compare the air CO2 to the water CO2 at different injection pressures over time...

The high pressure injection reduces the time for the air CO2 to impact the water CO2.

The water CO2 is not benign. It has its own forcing function as the tank actively consumes CO2... so the injection both adds and removes CO2 to keep the air and water tight.

bertoni
09/15/2017, 03:01 PM
Is there any documentation on the type of silcone that he used? That's another interesting idea.

dartier
09/15/2017, 04:01 PM
Is there any documentation on the type of silcone that he used? That's another interesting idea.

I am not sure if this is the product he used, or he DIY'd one, but he did link this site as a possible supplier: http://www.sspinc.com/products/Thin-Silicone-Membranes_4_category.htm

At one point he was talking about making his own between 2 pieces of saran wrap with a rolling pin, lol.

He mentions that the site above has 12" x 12" sheets that can be purchased.

Dennis

karimwassef
09/15/2017, 04:08 PM
The silicone diffusion rate would need to be factored in.

I'm interested in time ... so this is a big deal.

bertoni
09/15/2017, 04:36 PM
That looks like an interesting approach. I agree that we'd need to know the velocity of carbon dioxide across the membrane. It'd be an interesting experiment if the cost is tolerable.

karimwassef
09/15/2017, 04:48 PM
I'm trying to gauge the rate of diffusion of air CO2 into aqueous CO2 and need to account for the rate of diffusion of CO2 through the membrane!

Don't like measuring time with a time error... but maybe it can be corrected for enough.

orcafood
09/15/2017, 05:45 PM
I think a pipe sealed from the top in an overflow (low flow area) extended all the way to the bottom having no cap and some sort of inline fan like an aqualifter to keep the air moving through the top of the pipe to the sensor and back near the bottom so it is always full of air? Keep the sensor far from the water and bring air, which is under decent pressure from the top-capped bottom-open tube submerged. Then pack a bunch of desiccant into an inline filter before the sensor.

dartier
09/16/2017, 04:09 AM
What range do you think we should be planning to support? The vendor linked with the IAQ sensors offers a variety of ranges (and price points), with the most affordable supporting 0-10,000 ppm while ones that support 0-100,000 ppm are 3 x the price (and 1/3 as accurate 30 ppm | 100 ppm).

Is 10,000 ppm a reasonable limit to expect that may diffuse out of the water into the headspace of the sensing head?

Dennis

orcafood
09/16/2017, 04:25 AM
There is around 400 ppm in the air outside, closer to .5 ppm dissolves into the tank water as co2, a headspace should give a depressed readout compared with the outside co2, low range is best.

karimwassef
09/16/2017, 06:33 AM
300-900 is what I've found meaningful for my reef

dartier
09/27/2017, 12:12 PM
Here is a new device that is coming onto the market that would be a lot easier to use. no pricing just yet.

https://www.atlas-scientific.com/product_pages/probes/ezo-co2.html

However in a thread on R<2>R between RHF and a couple of Alkalinity tester people, they indicated that to have a hope of being able to accurately calculate Alkalinity based on PH and CO2, that you would need an accuracy of 2-3 ppm for the CO2 reading. Which none of the reasonable priced detectors seem to be able to do (including the one I just posted).

Dennis

karimwassef
09/27/2017, 12:37 PM
Can you spell out the thread after the r2r?
Haven't seen RHF on here for a while

dartier
09/27/2017, 05:07 PM
Can you spell out the thread after the clay-boa?
Haven't seen RHF on here for a while

Randy does not post here much anymore as far as I can tell. He handles the chemistry forum on R<2>R now. If you search for "KH Guardian(KHG) Alk calculation" over there you will find it.

Kirill is also active in that thread. It was his DIY alk monitor (on here) that piqued my interest so many years ago to start thinking about building my own version.

Pretty silly that I have to search for ways to encode a short form to get it past the censors on this board. Unbelievable actually ...

Dennis

orcafood
09/28/2017, 08:31 AM
Though the resolution is low on the CO2 sensors the CO2 levels in the water should not change too quickly. I'm thinking that fact can be abused in combination with a robust outlier rejection (RLS) system to get a more stable approximation. If we combine hundreds of data points over the course of a few minutes a sizable distribution can be generated.

Another option would be boosting the CO2 signal somehow. Perhaps a dye could be used, reactive to CO2, to enable detection via visible light. I'm thinking a dosimeter with a color based detector could work for a while.

Building a more accurate sensor might be feasible as well, they don't seem very complicated. IR led of the proper nm, maybe a laser diode, small tube with air flow and a very accurate IR detector protected with a band pass filter.

wiki seems to disagree though haha, I don't see why a more powerful IR source couldn't get higher resolution.

"The best of these have sensitivities of 20–50 PPM.[1]"

FTIR lists a detection limit of 50 ppb for CO2 in air. Why would it have such a lower detection limit over NDIR?

-- The interferometer used to sweep the spectrum in the FTIR is why the detection limit is lower

bertoni
09/28/2017, 05:46 PM
If the error follows some known distribution, then using more measurements should help with accuracy. If there's a deterministic bias of some sort, that won't work, though. It would be interesting to see what kind of data these meters produce.

orcafood
09/29/2017, 03:30 PM
My question is how do they not have FTIR setups for narrow bandwidths. Must be cheaper to produce an FTIR that can't sweep the full spectrum. My principles of instrumental analysis book is saying that multiplexing the signal is the reason for the FTIR gaining so much sensitivity over the NDIR. Another reason is due to the Jaquinot advantage or throughput advantage. Fewer optical elements and no slits allows for a much stronger signal.

Ahhh:

"decreasing the width of the resolution element has the unforunate effect of decreasing the s/n ratio due to needing narrower slits. For infrared detectors, the reduction in signal strength is not accompanied with a corresponding decrease in detector noise.

The multiplex advantage is described as overlaying many FT spectra and increases the s/n ratio of IR over Vis by a factor of 40. It is unusable at near-IR/Vis due to shot noise and flicker noise which both increase with radiant light source power.

orcafood
09/29/2017, 04:49 PM
http://www.uicinc.com/blog/comparison-between-coulometric-detection-and-ndir-detection-method/

"Detection Range

Since a single CO2 molecule may be “counted” several times as it passes through the cell, the NDIR detector can be very effective at conducting low-level carbon analyses. Longer cell lengths allow for lower level analysis, often down to the low part-per-billion (ppb) range."

orcafood
10/02/2017, 07:48 AM
It seems there are a few types of NDIR. One type of NDIR called gas correlation analysis is right on the money with what is needed here, it uses a reference gas and differential pressure to transduce the signal with a capacitor. They have a resolution around 1-2 ppm with a folded path of 3 meters. A longer folded path would get even more accurate of results but doesn't seem necessary.

I bet an old one could be picked up off of ebay still functioning for a few hundred dollars and interface it through an analogue pin on a PIC.

jrhupp
10/02/2017, 08:54 AM
It seems there are a few types of NDIR. One type of NDIR called gas correlation analysis is right on the money with what is needed here, it uses a reference gas and differential pressure to transduce the signal with a capacitor. They have a resolution around 1-2 ppm with a folded path of 3 meters. A longer folded path would get even more accurate of results but doesn't seem necessary.

I bet an old one could be picked up off of ebay still functioning for a few hundred dollars and interface it through an analogue pin on a PIC.

I think you want to avoid this type of sensor; there is a reason they are not really used anymore. They have strong temperature dependencies and are sensitive to vibrations.

It would be much better to find a sensor that does the measurement purely optically.

I haven't done anything with the little NDIR sensors targeted at HVAC systems, but I think they hold good potential for this application. My guess is they will be a fail out of the box, but that with some effort/modification they may perform pretty well. At this point though, that is just speculation.

orcafood
10/02/2017, 09:17 AM
Ah good point with the vibrations/ temperature issues, I just noticed they are very accurate when done properly.

I think a longer path length for the NDIR would help tremendously with the resolution. Also I bet a beam chopper setup could help reduce some of that noise. If I remember keeping the detector cooler might also lower the noise.

I think the problem is why have such a small NDIR? The available filtered NDIR photometers I have seen seem to be really small like they are trying to make them portable or something. Really it is just a box with mirrors setup to increase the path length. The IR source is just nichrome wire, dirt cheap, band pass filter isn't too bad then lastly a decent pyroelectric transducer which is gona be the pricey part. I'm seeing that thermopiles can be used for mid-IR, shouldn't be too expensive.

Someone must make a larger, higher resolution CO2 monitor...


I have begun to collect broken cameras for optics haha

jrhupp
10/02/2017, 10:01 AM
Ah good point with the vibrations/ temperature issues, I just noticed they are very accurate when done properly.

I think a longer path length for the NDIR would help tremendously with the resolution. Also I bet a beam chopper setup could help reduce some of that noise. If I remember keeping the detector cooler might also lower the noise.

I think the problem is why have such a small NDIR? The available filtered NDIR photometers I have seen seem to be really small like they are trying to make them portable or something. Really it is just a box with mirrors setup to increase the path length. The IR source is just nichrome wire, dirt cheap, band pass filter isn't too bad then lastly a decent pyroelectric transducer which is gona be the pricey part.

Someone must make a larger, higher resolution CO2 monitor...


I have begun to collect broken cameras for optics haha

For a broadband measurement (which is what I think we want), the short path lengths are perfectly reasonable. It is no issue to get resolutions on the order of a tens of ppb with a path length of 10 or 15 cm.

To get a significantly longer path length requires a much higher power IR source. Now days that general means a near IR laser (costly) or mid IR laser (supper costly) which is just a ridiculous approach in this application.

Agreed that the IR source should be cheap. The real cost in the source will be the reflectors and collimating optics.

The detector should be pretty cheap too. Nothing too special there for what we are discussing.

It will be the optical filters that will be the biggest component cost and most trouble to select. You need to pick a region of strong absorption that is free from spectral interference from other atmospheric gases and chose a filter that covers a sufficient portion of this region so that you capture it across a range of temperatures and pressures and with sufficiently sharp cut offs on the ends. This is a far from trivial excise and one off filters will likely be a couple hundred bucks a pop.

As to the camera optics, keep in mind that you are dealing with wavelengths far beyond visible. So just because you can see through it doesn't mean IR can pass through it. I don't know what is used to make camera optical components so I have no idea if they would work in this application. Just a word of caution on one of the less obvious possible issues.

Your other big cost will be suitable standard gases to calibrate the sensor with. Expect the response to be non-linear (see Beer-Lamberts Law), which means a number of gases will be needed for calibration purposes. WMO referenced standards are going to run you ~1k a pop. Standards from your local welding supply will be cheaper, maybe a few hundred. But labeled accuracy will be an order of magnitude or so worse than the sensor accuracy you are looking for.

Take the above to suggest that it is no easy task to build and calibrate an IRGA. My recommendation would be to look to an existing low cost sensor and then work towards modifying it to improve its performance.

orcafood
10/02/2017, 11:28 AM
I'm taking the optical grade mirrors from the cameras, they should work well as the entry and exit mirrors they are probably close to a wavelength flat.

The concave collimating mirrors would be pricey your right. Perhaps less so out of old equipment though. Next question why even use the collminating mirrors? Why not just a tube with a silicone collimating lens (passes 1-10 um) at the end, no mirrors, just one lens which runs around $200. Of course the bandpass filter is still needed. Gas standards would be tricky to make. Could they not be made using a working CO2 meter? For example remove the CO2 from a gas space with soda lime and see how low it can go. That is the first point on the calibration. Add CO2 with a tank until it goes up 50 ppm on the working meter then set the point. 5-10 points would probably do?

Good information on the IR-lasers I was trying to find more about them, I was worried they would be cost prohibitive. (Off topic but holy cow I wish I had one of those Xe/Ar lasers, so much fun to use in the labs doing Raman. I never set it up, but I watched them do it for a little bit, takes foooreeevverrr)

A=-log(I/I0)=Elc brings me back, extinction coefficients are so useful. Absorbance and concentration seem linear in that equation or were you referring to how the linear range breaks down at the maximum linear response?

CO2 only has 4 vibration modes. This is its symmetric stretch, asymmetric stretch, and then the two scissors. Only 3 of these are IR active, the asymmetric stretch at 4.3 um and the degenerate scissors at 15 um which have a much weaker IR intensity. So it seems like the best choice is to focus on the 4.3 um area.

All that said I agree that a sensor would be much easier to setup and repeatable for all of us reefers. Please let me know if you can find a sensor with ppb resolution for a decent price, that would be a game changer! The $100 NDIRs seem to have a resolution around 20 ppm which is not good enough.

I couldn't find one so far that fit the bill, thus looking for alternatives.


Another thing I was reading about was that many of the HVAC NDIRs have a built in baseline correcting algorithm that assumes that the lowest value it can get to is around atmospheric CO2 levels. The detector drifts to lower CO2 concentrations so the baseline constantly gets changed under the assumption that indoor co2 will only be above atmospheric CO2 and never below. Not good when a headspace in contact with seawater may have less than atmospheric co2.

orcafood
10/03/2017, 02:37 PM
Maybe using some dye to tranduce the signal to the visible spectrum is the way to go. Green laser pointers are cheap, IR optics are much too expensive

jrhupp
10/03/2017, 11:53 PM
I'm taking the optical grade mirrors from the cameras, they should work well as the entry and exit mirrors they are probably close to a wavelength flat.

The concave collimating mirrors would be pricey your right. Perhaps less so out of old equipment though. Next question why even use the collminating mirrors? Why not just a tube with a silicone collimating lens (passes 1-10 um) at the end, no mirrors, just one lens which runs around $200. Of course the bandpass filter is still needed. Gas standards would be tricky to make. Could they not be made using a working CO2 meter? For example remove the CO2 from a gas space with soda lime and see how low it can go. That is the first point on the calibration. Add CO2 with a tank until it goes up 50 ppm on the working meter then set the point. 5-10 points would probably do?

Good information on the IR-lasers I was trying to find more about them, I was worried they would be cost prohibitive. (Off topic but holy cow I wish I had one of those Xe/Ar lasers, so much fun to use in the labs doing Raman. I never set it up, but I watched them do it for a little bit, takes foooreeevverrr)

A=-log(I/I0)=Elc brings me back, extinction coefficients are so useful. Absorbance and concentration seem linear in that equation or were you referring to how the linear range breaks down at the maximum linear response?

CO2 only has 4 vibration modes. This is its symmetric stretch, asymmetric stretch, and then the two scissors. Only 3 of these are IR active, the asymmetric stretch at 4.3 um and the degenerate scissors at 15 um which have a much weaker IR intensity. So it seems like the best choice is to focus on the 4.3 um area.

All that said I agree that a sensor would be much easier to setup and repeatable for all of us reefers. Please let me know if you can find a sensor with ppb resolution for a decent price, that would be a game changer! The $100 NDIRs seem to have a resolution around 20 ppm which is not good enough.

I couldn't find one so far that fit the bill, thus looking for alternatives.


Another thing I was reading about was that many of the HVAC NDIRs have a built in baseline correcting algorithm that assumes that the lowest value it can get to is around atmospheric CO2 levels. The detector drifts to lower CO2 concentrations so the baseline constantly gets changed under the assumption that indoor co2 will only be above atmospheric CO2 and never below. Not good when a headspace in contact with seawater may have less than atmospheric co2.

To the mirrors there is not necessarily a need for them. There is no inherent reason the components can't be laid out linearly removing the need to reflect the light. This is good because standard mirrors used in NDIR instruments are gold plated.

You don't inherently need to collimate the beam. It helps greatly with optical through put and eliminates/minimizes sensitivities to surface characteristics of the optical bench though.

For standards, sure you could mix gas as a transfer standard from a working meter. But that means you have a meter already that has a greater accuracy than what you desire in your DIY version. You can also use precision syringes and a precision mixing volume to generate standards. This might actually be doable pretty cheaply as I think about.

As to absorption as a function of gas density. It may appear linear over a narrow range, but it is inherently not. Probably best to think about it on a narrow wavelength basis and build up from there. This is important for thinking about where a gas absorbs as well. Because while you are correct on major regions of absorption, CO2 has individual lines and small groups of lines peppered through out the IR region. Some more useful than others depending on your IR source, filters and detection limit.

In any case, absorption on a per wavelength bases (Av) can be expressed as:

Av = 1-e^(-kv*l*d)

Where l is the optical path length and d is absorber density. kv is the absorption coefficient at v, which is a function of line strength and halfwidth. Line strength saturates at some point.

I have lofty dreams of trying out some of the cheap NDIR sensors but haven't gotten to yet; too many more important projects to get through first. But when I do I will report what I find.

I had seen most had limited resolution and had hoped that there may be some way of working around that. Hopefully we shall see at some point.

I wasn't aware of the drift correction. The direction of the correction seems curious to me though. Do you know if they are using a reference gas cell? Diffusion into the reference cell over time would make sense for the need to correct upward. Or maybe its a source aging issue.

jrhupp
10/03/2017, 11:57 PM
Maybe using some dye to tranduce the signal to the visible spectrum is the way to go. Green laser pointers are cheap, IR optics are much too expensive

If you are going that route, why not just measure alkalinity with an automated colorimetric test?

There should be some visible dye suitable for this. I know we typically use a titration method (and that seems the approach used in the automated monitors coming out). But I am pretty Hatch or somebody along those lines make a single reagent that segregates into two forms with different peak absorption wavelengths in the visible spectrum based on alkalinity.

orcafood
10/04/2017, 12:00 PM
sound like we want a colorimetric drop checker

karimwassef
10/04/2017, 12:31 PM
Another incomplete project from years ago

http://www.reefcentral.com/forums/showthread.php?t=2508211&highlight=automatic+alk&page=9

Need to clone myself or make enough $ to retire so I can tinker more :D

orcafood
10/04/2017, 01:32 PM
I've been trying to find a way around titration. Use a chuvette as the drop checker (headspace) with an led, proper dye and a photo sensor. A reference would be nice too. Raising the response rate would require an interesting setup.

Beautiful project by the way

dartier
10/04/2017, 02:38 PM
If you are going that route, why not just measure alkalinity with an automated colorimetric test?

There should be some visible dye suitable for this. I know we typically use a titration method (and that seems the approach used in the automated monitors coming out). But I am pretty Hatch or somebody along those lines make a single reagent that segregates into two forms with different peak absorption wavelengths in the visible spectrum based on alkalinity.

I have been researching the various methods of performing alkalinity tests, and the reason I was following this thread. Along the way I ran across a paper that described a system to perform alkalinity testing using Spectrophotometry and a reagent with a known absorption profile. Then a small amount of the sample was added and the resulting PH shift was used to calculate the alkalinity concentration to achieve the shift. It was quite interesting, but did require removing the carbolic acid from the sample by flowing the sample through a gas permeable hose in a NaOH bath. I am not sure if the added complexity is worth it though.

I am leaning towards doing a regular titration with a colorimeter to detect the end point. To get around the issues with volume accuracy, I plan to use a load cell to calculate the various weights of the sample and reagent rather than try to keep pumps from drifting.

Dennis

orcafood
10/05/2017, 08:03 AM
Much smarter with weighing the sample, but you will still need to calibrate that pretty often?

The problem I see with titration is too many moving parts, each of which has some variability.

karimwassef
10/05/2017, 07:18 PM
So... different direction:

Was talking to a reef buddy of mine about this and had an idea about measuring CO2 in the air by pH...

So visualize a 5 gallon jug with RODI low dKH water and an air stone (pump has intake filter). The water has a pH probe in it and there's a loop of hose that allows water to recondense back into the jug as the air leaves.

Kinda like a drop checker with a pH probe.

Now, the pH should be proportional to CO2...

orcafood
10/06/2017, 03:08 AM
Exactly what I’m thinking, cycle the headspace with an inline air pump.instead of a pH meter, use a ph sensitive dye, laser pointer (already collimated and tight band) and a photosensor.

karimwassef
10/06/2017, 06:35 AM
Why? pH meters are already integrated into an Apex and can easily be calibrated, sourced, etc...

karimwassef
10/06/2017, 06:46 AM
Also, the reference jug could be any liquid where the pH and air CO2 can be correlated. I'm thinking of a thickened water formula to minimize any evaporation? As long as the inlet air for this test volume and the reef tank are the same, the formulas would apply directly to provide Alk. Almost all the needed parameters can be measured with conventional devices- pH, temp, salinity, pH(CO2). Barometric pressure is the only one to be assumed relatively constant.

The last gap is a fit that converts the test volume pH to air CO2 and an interface processor that collects all the data live and calculates Alk, then feeds it back as an analog or digital reading back to the Apex.

orcafood
10/06/2017, 07:01 AM
Wish I had me an Apex, that is probably why I don't focus on integrating to one. I think the liquid needs to be water, what else would you use? Many other liquids are volatile and would end up in the tank water eventually. Agar has large pores but I don't think you want to slow the diffusion down through reference solution, I think quite the opposite we want air vigorously pumping through, like your air injectors, keeping the headspace equilibrium. pH meters are expensive for nice ones, like $100 for a good probe with longevity. Green laser diode $5, dye free essentially, photosensor $2-5, half mirror $10, I'm sure there will be hidden costs. Think hanna phosphate checker. Total cost of around $30 tops and likely better resolution than could be had with a pH probe.

If the reference solution is isotonic with the other liquid (tank water) in contact through the headspace there shouldn't be evaporation. The headspace will be saturated with humidity.

I'm sure a pH meter would work just fine :)

karimwassef
10/06/2017, 10:03 AM
I think there's a confusion about the CO2 liquid pH sensor.

This is a completely separate and independent device that is sampling air not connected to the tank at all. It can be anywhere and can run oil if need be (not saying that's a good choice - just emphasizing that the liquid and pH sensor have nothing to do with the tank.)

The idea is that I'm not relying on headspace. I'm actively injecting air into the tank water and this sampling device simultaneously. I'm sampling the air going in and relying on my injectors to keep the gas in system at the measured rate.

orcafood
10/06/2017, 10:55 AM
Makes sense to me, outside is your headspace. I still recommend the design should bubble air through the solution being tested for pH rather than rely on diffusion.

Should work with a serious air injection system on a tank, such as yours. Otherwise the headspace might need to be used.

Ideally we need to know the pH of the solution to at least the thousands place, which you are not going to get with a meter, not to mention the pH meter will drift.

karimwassef
10/06/2017, 02:29 PM
I calibrate pH meters often :)

Yes. I was pushing the same air through the airstone. Maybe a surface treatment for the water would reduce evaporation.

karimwassef
10/06/2017, 02:32 PM
Accuracy does matter but maybe there's a way to make the water very low in alkalinity so that it responds more aggressively to CO2

orcafood
10/06/2017, 02:41 PM
A dkh around 0.1 might do. Once I get some time I will start to play with making a low range co2 drop checker, I have bromothymol blue sitting right here ready to go.

orcafood
10/07/2017, 05:26 AM
Hey Karim how accurate is your pH meter? To the tenths or hundredths? Alternatively can you link me to the probe your using?

Do you know the standard error of your kH estimate? Or perhaps do you have a distribution of your estimates over some time period so we could figure out the standard error?

karimwassef
10/07/2017, 09:12 AM
The scientific probe by Apex is what I use. Reading is in the hundreds 8.33
But I assume an error of +- 0.05 which is wild, I know.

I calibrate monthly though and I use two probes (one regular) in case on drifts. They're never the same value, usually off by 0.1. One I use for the feedback and one I use to stop anything running off the rails.

orcafood
10/07/2017, 09:18 AM
What kind of variation on your kH estimate do you get? Does it vary by a few points between estimates or a full degree?

karimwassef
10/07/2017, 12:56 PM
My purpose isn't to lock in to a particular kH value. This method is intended to get close, but then maintain it over time. I'm really reducing variability.

Since the errors in my system are relatively constant, maintaining my pH aligns with maintaining my dKH stability... so it's all about precision, not accuracy

http://cdn.antarcticglaciers.org/wp-content/uploads/2013/11/precision_accuracy.png

orcafood
10/07/2017, 01:20 PM
So what kind of values do you get?

A value means nothing without error. Scientific papers require error for good reason.

karimwassef
10/07/2017, 02:11 PM
Sure. But if my error is constant and I can calibrate it out, it doesn't impact me much. The thread I posted showed the measured vs. calculated Alk chart. Got to dig it back up.

Overall, I can keep my dKH between 7.5 and 8.5 by keeping pH between 8.34 and 8.35

orcafood
10/09/2017, 01:24 PM
Is your alkalinity approximation steady over time? Or does it jump around each subsequent calculation? Over the span of an hour what would be the max and min alkalinity approximation you get?

karimwassef
10/09/2017, 05:44 PM
My alkalinity is a function of pH and CO2.. if they're stable, my calculation would be stable.

In general, both are smooth over time, but not necessarily constant.

<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/0_zpsalk2pv0u.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/0_zpsalk2pv0u.jpg" border="0" alt=" photo 0_zpsalk2pv0u.jpg"></a>

The two points where there's a disconnect was due to the probe being too close downstream to the kalk drip during times when the drip was very high (continuous).

orcafood
10/10/2017, 04:53 AM
Thanks so much! The kalk spike really sk(r)ewed up those results haha :).

To properly start testing our standard error we are going to need multipe data points at every time stamp, or just more timestamps in general. Api kits are only accurate to 1 dkH if memory serves?

karimwassef
10/10/2017, 06:53 AM
I usually read to 0.5 since it goes blue-green-yellow so I can usually gauge the difference between 8 and 8.5.

There was no kalk spike ... this is normal operation a few months ago. I had just turned off my kalk addition for a day so I could get a clear baseline. That's why it starts st 8.05.

My pH feedback loop had a target that was also stairstepped in time to introduce some pH variance during the day.

orcafood
10/10/2017, 07:31 AM
With a burette we could get a much more accurate alkalinity reading, definitely worth the price. Then we could really see how close the alkalinity approximation truly is I am super curious.

To get a good idea of how accurate the alk reading is we would also need data from a full day where the kalk reactor and entire system is humming along in perfect equilibrium.

karimwassef
10/10/2017, 10:45 AM
This doesn't take into account nonlinear effects

Case in point - when my kalk reactor ran all night and didn't impact my Alk by more than 2dkH.

So this is only usable in the linear range. Once we get into saturation, it breaks down.

orcafood
10/10/2017, 11:14 AM
We could easily use svm or some other nonlinear modeling technique, no problem. It definitely breaks down during a surge of hydroxide, but that should be a very rare occurrence.

karimwassef
10/10/2017, 04:09 PM
True

kenneth wolfe
10/12/2017, 12:46 AM
I think it works for me because of my air injector. I push gallons of air into my water constantly through a very fine bubble foamer. It's a dual penductor fed by outside fresh air through a high pressure pump. This means that the time variable for the diffusion of gas into water is reduced significantly. Basically, I force the assumption that I'm at steady state to be more true.

Air CO2 sensors are only ~ $100. You could, in theory, drive it from an Apex and have all the data you need.

Unfortunately, Apex won't let you enter a math formula that returns a function of known values :( but this app works for me: iFxCalc iFxCalc:

https://itunes.apple.com/us/app/ifxcalc-most-human-nature-function-calculator-support/id786018193?mt=8

It's clunky to figure out but easy once you do

The formula is
Alk=(2*D3*0.0334211*C3/29.41*0.000001*10^(-(-8.712-0.00946*F3+0.0000856*F3^2+1355.1/(((E3-32)*5/9)+273)+1.7976*ln((((E3-32)*5/9)+273))))/10^(-G3)*10^(-(17.0001-0.01259*F3-0.000079334*F3^2+936.291/(((E3-32)*5/9)+273)-1.87354*ln((((E3-32)*5/9)+273))-2.61471*F3/(((E3-32)*5/9)+273)+0.07479*F3^2/(((E3-32)*5/9)+273)))/10^(-G3)*44.01+D3*0.0334211*C3/29.41*0.000001*10^(-(-8.712-0.00946*F3+0.0000856*F3^2+1355.1/(((E3-32)*5/9)+273)+1.7976*ln((((E3-32)*5/9)+273))))/10^(-G3)*61.02)/2*1000/17.9

Where Alk(D3, C3, F3, E3, G3) has
D3 = barometric pressure (mmHg) = 30
C3 = CO2 (ppm) = 450
F3 = salinity (ppt) = 34.5
E3 = temperature (F) = 77.5
G3 = pH = 8.3

Returns 9.8 dKH

Can machine learning reduce the impact of measurement errors- probably but you'll need an accurate and frequent Alk measurement to feed into it.Man you must be my brother, we need to get a beer together..

Sent from my SAMSUNG-SM-G870A using Tapatalk

karimwassef
10/12/2017, 11:41 AM
Thanks. Come on down to Texas and it's on me!

kenneth wolfe
10/13/2017, 12:52 AM
Thanks. Come on down to Texas and it's on me!I'm in Texas, I live in Texas city

Sent from my SM-G892A using Tapatalk

karimwassef
10/13/2017, 10:10 AM
I'm in Plano... north Dallas.

You're in Galveston - by the Texas coast... seems so far away