Monitoring in mature systems

[Kobus Jooste]

I have been looking at a lot of input and my own experience.  If a system is designed and managed correctly, what should we monitor in the end?  Are performance issues not typically design flaws or operational decicions rather than ecological processes?

 

The concensus from what I could read is pH, DO and Ammonia (I am assuming we all watch temperature naturally) is all we need to think about.  Thoughts?

[Kobus Jooste]

Interesting to see your side of the discussion Terri.  I own a aquaculture photometer, and probes for pH, ORP, TDS, EC and temp, thus you see which camp I fall into as well.  Cannot escape the science training.  I have a spreadsheet set up for uniodised ammonia calculations but am surprised to see how seldom people take this extra step when testing.

 

I was part of a group that designed a commercial system with full remote sensing capability.  If you can afford it, it makes sense to me.  The system is plugging away all the time and you just have to spot check and callibrate periodically. I will only consider this for a large commercial set-up though as a home set-up typically would give you plenty warning that something is wrong, unless you miss the signs or are away at the time.  For me there is a difference in focus with the large systems too.  If you are going to start considering HACCP (export fish for us, for example) then the monitoring system also becomes one of your tracking and tracibility records.

Terri Mikkola said:

First, let me say, I'm not a commercial grower and what I offer is strickly my opinion based on water quality testing as it relates to wastewater treatment. What I've learned is that the most stable system can take a turn for the worse even though visual indicators are ok. By combining visual inspections with daily tests ,one can be assured their system is working as it should be. To take it a step further, I would develop a process control plan using a spreadsheet to log your daily results and observations. I would add trend charts for each parameter based on 7 day averages to determine if your system is stable or moving in a direction that requires some type of action. A process control plan can assist with troubleshooting and determining the proper course of action to take.

 

TC mentioned on another topic thread how stocking densities impact water quality and stability. It's a great point. Low stocking densities create a more stable and forgiving system, which, I assume, is why some commercial growers don't feel the need to test daily. I would hate to see a business suffer a major loss due to water quality issues that could have been averted (my opinion).

 

And as for the the minimum parameters to test for - pH; temp; TAN (including determining NH3 toxicity levels); nitrate, dissolved oxygen, and alkalinity.

I would also send samples out to a lab to test for total coliform / e coli for food safety.

 

For high stocking density systems, I would add to the above list, TSS, BOD, and CO2, which can be calculate with pH and alkalinity.

[George T]

Rate of plant growth?  Less than optimal growth possibly indicating understocked tanks.

[Terri Mikkola]

Kobus, What an exciting project to be involved with! How does the remote sensing work? 

Kobus Jooste said:

Interesting to see your side of the discussion Terri.  I own a aquaculture photometer, and probes for pH, ORP, TDS, EC and temp, thus you see which camp I fall into as well.  Cannot escape the science training.  I have a spreadsheet set up for uniodised ammonia calculations but am surprised to see how seldom people take this extra step when testing.

 

I was part of a group that designed a commercial system with full remote sensing capability.  If you can afford it, it makes sense to me.  The system is plugging away all the time and you just have to spot check and callibrate periodically. I will only consider this for a large commercial set-up though as a home set-up typically would give you plenty warning that something is wrong, unless you miss the signs or are away at the time.  For me there is a difference in focus with the large systems too.  If you are going to start considering HACCP (export fish for us, for example) then the monitoring system also becomes one of your tracking and tracibility records.

[Kobus Jooste]

George - People may oppose this view, but for me, OVERSTOCKING can also potentially result in poor performance.  I tested a overstocked home unit all of last year, and its performance was worse than when the nit had half the fish load in it.  I'd love to see others come back with a response to this, but for me, it was an eye opener.  I think people with a fish focus will be less happy with AP in the end than people with a plant focus.

George J. Thurmon said:

Rate of plant growth?  Less than optimal growth possibly indicating understocked tanks.

[Kobus Jooste]

Terri - unfortunately, it is dormant waiting for the government to make up its mind on it.  The software was originally developed for the vehicle assembly industry, to allow full tracking and tracability, plus work station training / troubleshooting.  The software can be operated by lower literate people (it was demonstrated when during a labour strike, people out of management side of the factory and casuals could assemble a BMW to standard after a day's training.  The software allows only the correct activity to take place and goes into safety shut down if the operator tries to do something other than what the unit is promting him/her to do). The owners of the software was interested in modifying it for a commercial agri operation where everything from the stores to the processing is hooked up to their system.  The greenhouses will be monitored for critical parameters - temp, DO, pH, Ammonia, Nitrates and water flow rates.  Warning systems will be present on site and at remote control stations, where an extention officer can adjust everything remotely if the system manager is not getting it right.  The tracking and tracability will be used to ensure that produce is delivered per order, and to HACCP standards. If anything goes wrong with a batch, you will have an heads up on the processing side in advance in order to source what was "lost" in one production line from another.  This is a delicate mix of extremely high tech with labour intensive practices for the South African environment.  The idea is to have clusters of production sites feeding into central processing facilities.

 

I have another opportunity to present the concept to a government development agency on Monday.

Terri Mikkola said:

Kobus, What an exciting project to be involved with! How does the remote sensing work? 

Kobus Jooste said:

Interesting to see your side of the discussion Terri.  I own a aquaculture photometer, and probes for pH, ORP, TDS, EC and temp, thus you see which camp I fall into as well.  Cannot escape the science training.  I have a spreadsheet set up for uniodised ammonia calculations but am surprised to see how seldom people take this extra step when testing.

 

I was part of a group that designed a commercial system with full remote sensing capability.  If you can afford it, it makes sense to me.  The system is plugging away all the time and you just have to spot check and callibrate periodically. I will only consider this for a large commercial set-up though as a home set-up typically would give you plenty warning that something is wrong, unless you miss the signs or are away at the time.  For me there is a difference in focus with the large systems too.  If you are going to start considering HACCP (export fish for us, for example) then the monitoring system also becomes one of your tracking and tracibility records.

[Gina Cavaliero]

I have been wanting to get back to this discussion in response to Terri's thoughts about all the parameters she would test for.  I cannot disagree that extensive testing would surely give a value of data that would indicate when a problem may  occur.  However, that level of testing, monitoring and then daily data input translates into labor and cost.  The level of testing and recording you are suggesting seems that it would involve at least an hour or more of daily monitoring and really not feasible considering the other things that have to occur with running a commercial operation.  If I was concerned of some catastrophic event drastically altering my system balance, I would agree that level of testing would be warranted, but I am not.  The longer these systems run and the larger (size is a big factor) they are, the more stable they tend to be.  In over four years, Friendly hasn't had any catastrophic or signficant water quality events.  They too in the beginning tested constantly as did we, but then a level of confidence is obtained because of a balance that eventually maintains.   We must consider what parameters to test for that will insure good performance while considering the cost to do it while also recognizing the value of our very intimate, daily interaction with the system. 

 

I am curious though, unless purely to test to show that there is no presence of e.coli, why you would recommend that test Terri?  These systems, if impervious to outside factors, cannot and should not harbor the e.colo or salmonella pathogens.  You agree with that Kobus?  Now I must go harvest.

[Kobus Jooste]

O dear, I can see where this discussion is going (or went the last time I was part of one). I'm going to quantify my response from the get go.  In my commercial aspirations, I see the South African context of government projects (job creation) and targeting the export (EU) market wih fish.  Regardless of the biosecurity of the set-up, you have the human factor in the system now, and you want to be able to deliver EU grade fish.  You do not want to get your ton of fillets sent back because they say they found e.coli on the other side, thus I will say that regular (but not daily) checks could be in order.

 

I would also push for a fully automated monitoring and control system, thus the logging is not done as part of daily labour expense but once off Cap Ex.  Again, with the larger plan I am working on, there will be more than one greenhouse and you cannot be everywhere all the time.  This is completely different to what you are doing At Green Acres, Gina (Or the Friendlies, or Chris).  For your set-up, I will conclude as follows: Your market will ultimately dictate what paperwork needs to be delivered, and the paperwork will dictate how you monitor the production platform.  You are either going to be able to decide for yourself, or you are going to comply with regs from your customer base.  With all these e.coli scares all over the place I smell regulations coming...............................

[Jay Wolf]

Don't give away all my old fish farming secrets!!!!  LOL  Just kidding of course.  Terri you are absolutely correct with charting your results.  It works very well.  I like the idea of testing for coliform / ecoli.  I wonder if that is something that should be done monthly?????

Terri Mikkola said:

First, let me say, I'm not a commercial grower and what I offer is strickly my opinion based on water quality testing as it relates to wastewater treatment. What I've learned is that the most stable system can take a turn for the worse even though visual indicators are ok. By combining visual inspections with daily tests ,one can be assured their system is working as it should be. To take it a step further, I would develop a process control plan using a spreadsheet to log your daily results and observations. I would add trend charts for each parameter based on 7 day averages to determine if your system is stable or moving in a direction that requires some type of action. A process control plan can assist with troubleshooting and determining the proper course of action to take.

 

TC mentioned on another topic thread how stocking densities impact water quality and stability. It's a great point. Low stocking densities create a more stable and forgiving system, which, I assume, is why some commercial growers don't feel the need to test daily. I would hate to see a business suffer a major loss due to water quality issues that could have been averted (my opinion).

 

And as for the the minimum parameters to test for - pH; temp; TAN (including determining NH3 toxicity levels); nitrate, dissolved oxygen, and alkalinity.

I would also send samples out to a lab to test for total coliform / e coli for food safety.

 

For high stocking density systems, I would add to the above list, TSS, BOD, and CO2, which can be calculate with pH and alkalinity.

[Jay Wolf]

LOL Depends if you care about making a profit at the end of the day!  ;-)

Kobus Jooste said:

George - People may oppose this view, but for me, OVERSTOCKING can also potentially result in poor performance.  I tested a overstocked home unit all of last year, and its performance was worse than when the nit had half the fish load in it.  I'd love to see others come back with a response to this, but for me, it was an eye opener.  I think people with a fish focus will be less happy with AP in the end than people with a plant focus.

George J. Thurmon said:

Rate of plant growth?  Less than optimal growth possibly indicating understocked tanks.

[Terri Mikkola]

Aquaponics brings together a diverse group of people with varied backgrounds and expertise. My expertise lies in only one aspect of AP at this point. I still view AP as a black box with things to be discovered and learned. I feel comfortable offering my opinion regarding water quality because it's based on 20+ years of experience in the management and operation of biological systems. I won't comment on topics where I lack experience unless it's prefaced with that acknowledgement. 

I  respect your commercial AP experience and especially your track record for starting and running a successful business. Your decision to reduce water quality testing was based on a history of daily water quality tests and the establishment of a stable operation. From a business perspective the choices you made are sound.

In regards to testing for pathogenic bacteria, the University of Hawaii at Manoa published a paper on food safety as it relates to AP.

Gina Cavaliero said:

I have been wanting to get back to this discussion in response to Terri's thoughts about all the parameters she would test for.  I cannot disagree that extensive testing would surely give a value of data that would indicate when a problem may  occur.  However, that level of testing, monitoring and then daily data input translates into labor and cost.  The level of testing and recording you are suggesting seems that it would involve at least an hour or more of daily monitoring and really not feasible considering the other things that have to occur with running a commercial operation.  If I was concerned of some catastrophic event drastically altering my system balance, I would agree that level of testing would be warranted, but I am not.  The longer these systems run and the larger (size is a big factor) they are, the more stable they tend to be.  In over four years, Friendly hasn't had any catastrophic or signficant water quality events.  They too in the beginning tested constantly as did we, but then a level of confidence is obtained because of a balance that eventually maintains.   We must consider what parameters to test for that will insure good performance while considering the cost to do it while also recognizing the value of our very intimate, daily interaction with the system. 

 

I am curious though, unless purely to test to show that there is no presence of e.coli, why you would recommend that test Terri?  These systems, if impervious to outside factors, cannot and should not harbor the e.colo or salmonella pathogens.  You agree with that Kobus?  Now I must go harvest.

[Gina Cavaliero]

Terri,  your expertise is certainly welcomed here!  Thanks for the food safety doc as well.  Being food safety is prevalent on everyone's minds lately and with good reason, I think this discussion is relevant.  I was asking why the need for pathogenic testing because I was wondering if you or Kobus thought closed loop systems with no external factors could harbor those pathogens.  I think it is safe to say that the presence of these pathogens would have to be from foriegn introduction. 

 

With Kobus' example of the tons of filets that could potentially be returned because of the presence of e.coli, I must conclude that the pathogen source would likely be a resullt of the most typical introductory point of most pathogens,  at processing.  Would you not?  We, or at least I am not considering that this theoretical source of contamination was the system water or fish but likely a contaminant by a worker.  Even the document suggests that the presence would be from an outside element.  I agree that data collected from various aquaponic systems would be great for the industry in general in culminating a great data base supporting that these systems can grow food in the safest way possible provided the operation employs best management practices.  Once it goes to processing, its a whole other game and set of rules.  Unfortunately it will invariably be a handling  issue that taints the industry and as Kobus said, that he smells regulation coming, so do I.  It is only my hope that the industry via strong consumer and policy education can make said regulation logical and applicable to aquaponic growing rather than the traditional soil grown regulations governing as they are now. 

 

   

[Kobus Jooste]

While I think that we are getting slightly side tracked and heading off towards the certification debate again (drat ).  I think that there is a major distinction to be made between the two "camps" or "schools of thought" in operation in the discussion.  I'm not labelling people, I am categorizing background and mind formatting.  Correct me if I am wrong as this is an observation and in no ways an attempt to takes swipes at people. Before I became a scientist I did not view information as I do now thus I feel comfortable with what I am saying.   

 

Anyone with scientific, financial or engineering training are sticklers for piles of data.  We cannot deny it and often, I think the scientific method forces a little bit of overkill mentality on us.  The sky is blue because we have to disprove any of the other options, not because it is plain to see.  That said, if someone sues you one day because the sky is red according to them and you do not have the data to prove it is still blue, you could find yourself up a creek.  Thus, while someone with less of an ingrained drive for piles of data (statistical analysis of figures or working with micrometer range tolerances does this to a person) are perfectly capable of being the best aquaponic farmer ever to walk the face of the earth, I see a vital need for lots of information in terms of the following scenarios:

 

1.  You are questioned about your methods when there is a problem with a range of products supplied to a vendor / eatery where a quality issue pops up.  The likely outcome is the person with the least robust record keeping history will have the most fingers pointed at them, not because it is their fault, but because their defense is weakened.  Look at the shambles caused in the latest e-coli outbreak in Europe.

 

2.  You supply food to an entity that want to know everything about your processes for their grading / quality control.

 

3. You have been running your operation for some time and see a difference in yield or quality, and want to assess the underlying issue / causal factor

 

4. You have been running your set-up for a while and want to evaluate the most cost-effective food, plant or cultivar.  I have tripped myself up a number of times during research.  You make up your mind as you go along, but the spreadsheets prove you wrong and when you take the time to analyse the cause, you see where you went wrong.

 

A lot of this boils down to personal choice, thus I am not putting this forward as the "right" or the "better" option, just trying to explain how my mind works and where I see a need for dedicated monitoring.