DIY Nitrogen Cycle Model

I have been attempting for a while now to model the nitrogen cycle in an aquarium using Monod kinetics combined with a CSTR-in-series reactor model. The model consists of 7 differential equations that are solved simultaneously using a Matlab algorithm. Four of the equations are for the chemical species balances NH3, NO2, NO3 and N2 while the remaining three equations are for the bacterial colonies Nitrosomonas, Nitrospira (Nitrobacter also being present) and Pseudomonas. The oxygen concentration is assumed to be at the saturation limit of the water at a specified temperature at all times but is not considered as being a limiting reagent.

The bacterial colonies responsible for Nitrification / Denitrification are regarded as an immobilized catalyst with the ability to multiply. It is, however, very difficult (nearly impossible) to relate the amount of bacteria present to the quantities of ammonia/nitrite/nitrate that are eliminated. The model utilizes experimental constants which I gathered from various sources on the net so I have no way of knowing how accurate it is. The model does take the bacterial cell death rate and specific growth rates into account as well as inhibitory effects to some extent. The model is also sensitive to the aquarium size and volumetric flow rates through calculating the average residence time. It can be adjusted to accommodate Anaerobic environments (i.e. a Denitrifier) which is needed by Pseudomonas for production of N2 gas from NO3. At this stage I am trying to learn how to tweak the model so that it more accurately resembles a quantitative model rather than a qualitative model.

I am looking on input on how to accurately (and obviously as cheap as possible) measure NH3, NO2 or NO3 levels in the water on a continuous basis. I am looking for an alternative to a water test kit. Anyone out there care to point out some pitfalls or possible difficulties that may be encountered?


(It is a pulsed output rather than a steady state output at this stage since I haven’t had time to developed the algorithm further) I am also aware of the fact that N2 production will probably be a lot slower since the above mentioned design assumes that a denitrator is present in the setup.

See the attached word doc for pics.
I will only be able to reply to this thread tomorrow.
All Comments / Suggestions /Critique welcomed

 

If anyone can give me some data points to help evaluate the model it will be really helpful. The following is needed:

• NH3, NO2 or NO3 Measurement and the time at which the measurement was taken after aquarium startup. (Ideally it will help to know how the ammonia, nirite or nitrate levels look with respect to the running time of the aquarium).
• Your tank volume.
• Your flow rate in and out of the tank (if possible).

It doesn’t need to be formal just any type of ball park answer will help a lot. The thing is that if we can understand the dynamics of the bacterial communities within our tanks better then we should eventually be able to predict what will happen if certain parameters are altered.

For instance It will help a million if I knew how long (more or less) after tank start up how long (in hours or days) it takes for the Nitrite levels to increase to significant levels.

Or how long it takes for the ammonia to start decreasing drastically.

Or even how long it takes for the Nitrate levels to start increasing.

The more data I can get the more accurate the model will become.

Also it seems I don’t know how to add a pic in the thread without putting it as an attachment – help will be much appreciated :bigsmile:

 

very intresting and IM afraid a bit out my element at first glance .
I could mabe find the readings for you when I set up my marine as i took readings every sunday at the same time for about 3 months ,If im correct in reading thats what you want.the part im not sure of is every cycle would differ as every set up has diiferent starting levels of bacteria ,temp and water used ?

 

That would be perfect thank you very much! You are absolutely right every cycle will differ according to the parameters you have listed. I can adjust the bacterial starting colonies in the model so that it takes longer for them to grow. I can also adjust the temperature. The problem with the temperature is the growth and Monod constants are functions of temperature. But of all the values i gathered from literature none were reported as functions. I only have single constant values (ill just have to work with what ive got right now).

I was hoping that there will be some general trend for all cycles that are followed irrespective of the influencing parameters (within certain limits). I can also adjust the model to be specific to a certain set of conditions that simulate your marine tank best and then have it spit out some values that can be used in an attempt (probably very inaccurate at this stage) to model your specific aquarium.

The more different inputs i receive from different aquariums, the better the model will approximate reality.

I can for instance set the Pseudomonas count to zero which will prohibit the Nitrogen from forming and result in higher Nitrate levels. In a sump for instance, the anaerobic conditions needed for reducing Nitrate to Nitrogen gas only occurs below the undisturbed gravel and takes orders of magnitude longer than for ammonia to be converted to nitrate. (Ill add this example output in another word doc to this post) The example states 3 different initial bacterial colony counts. Their absolute values have no real meaning, it is their value relative to one another that gives the model meaning.

Unfortunately i haven't taken all of the chemical species in the water (for example chlorine, iron, magnesium and calcium) into account or even their effect on the bacterial communities. The model isn’t sophisticated to account for that ... yet. Therefore the model does not take pH or water hardness into account.

This is still a giant learning curve for me, im no expert, and that is why i am very glad that someone replied, whenever you can give me any measurements i will be grateful.

Thanks a lot!

 

ok cool still looking for the log book .
can I sugest you try our sister site masa ,link normaly on the home page somewhere.
the marine guys are a lot more technical in there water paramaters and tests and cycle timelines and include cal ,mag and more into account or even post in the ask the prof section on tasa as he surley would be able to help you out a hell of alot more.

 

Ah ok thanks, ill check it out then. Im a bit busy right now but i will get to it as soon as i can. Hopefully later this afternoon, would you suggest i just copy and paste the thread into a thread in masa?

hahaha i meant *copy*

 

well
best to regester .do an intro as to who and what nd why and then paste relevant info

 

Hey Zuraki, what happened to the model? Did you find any data to use?

 

Hey @Good_Times , I actually did get some good input from someone at the Oceanographic institute in KZN. He is a member on MASA. We managed to fit the model to some of his experimental values and by tweaking the bacterial counts, I managed to get a very good fit. My next project is to develop a GUI using Python to make it easier to use... but I will see how that goes hehe.

 

Cool, mind posting it here?
Is the colonizing and multiplying time of nitrifying bacteria the same regardless of salinity level, i.e freshwater and saltwater?

Good luck.

 

This was the last result I posted: