How Long Can Water Be Used
April 23, 2010
Jim Anderson (firstname.lastname@example.org): How long can water be used in shrimp biofloc systems without losing any of its basic elements?
• A salinity of ten parts per thousand
• Some removal of solids
• Total suspended solids levels below 400
• Production of six kilograms per cubic meter at harvest
How do I remediate that water after harvest?
Dallas Weaver (email@example.com): The Cumulative Feed Burden (CFB) is a good way of looking at this issue.
Bob Rosenberry (firstname.lastname@example.org): Basically, the CFB is the weight of the feed put into the system divided by the system volume. For systems that have been operating for a long time, the CFB is the daily feed rate divided by the daily water exchange.
Off List, I contacted Dallas for a little more information on how CFBs were calculated. Here’s what he said:
To get an accurate CFB you need to look at the turnover time for the system. For example, let’s say I had a 1,000-liter system and fed it 500 grams of dry feed per day and discharged 20 liters of water and sludge per day (about 200 grams of dry sludge in 10 liters of water—2% sludge—plus another 10 liters to flush the sludge lines and spillage). Only changing 2% of the system volume means that it would take about 100 days (twice the exchange time of 50 days) to get close to equilibrium. You then take the 100 days total feed and the 100 days total discharge, which would be 500*100/20*100 * 1,000 (convert gm/l to mg/l) = 25,000 mg/l.
For a one-meter deep pond producing 10 kilograms per square meter (super intensive) at a 1.5 FCR, you added 15kg/m3 of feed to the discharge at the end of the cycle, which is a CFB of 15,000 mg/l. However, most systems only produce 2 kg/m2, which cuts a zero exchange system starting with clean water to a CFB of only 3,000 mg/l.
The Rest of the Discussion Is Directly from the Shrimp List
Dallas Weaver (email@example.com): Systems in the 1,000 mg/l CFB range are trivial to operate. Systems in the 10,000 mg/l CFB range will have very yellow water (if ozone is not used), but no major system problems or major water chemistry problems (other than, of course, the usual pH, ammonia, nitrite, nitrate and system problems). With a 10,000-mg/l CFB system, you need a settleable solids filtration system that discharges concentrated sludge only.
In the 50,000 mg/l range, you discharge concentrated sludge in almost a solid, pumpable form. That is about as high as you can go with CFBs from a practical standpoint. At this level you have to consider trace element chemistry—both additions and removals. Using some ozone helps the water look more like tea than coffee.
I have run CFBs in this range on some marine systems. To get them into a steady-state takes a lot of feed and time, which usually means pure O2 with feed rates greater than 500 gm/m3 per day. To reach equilibrium, you must operate for more than 200 days with near zero discharge. I did see some trace element deficiencies at these extreme levels of CFBs. The issue is real and what I spent on the problem was real. Most of the information I collected is proprietary.
Joe Mann (firstname.lastname@example.org): Dr. Tzachi Samocha at Texas AgriLife Research in Corpus Christi said he achieved a new world record in shrimp production using an environmentally friendly system with no water exchange throughout the growing cycle. The system operated for 108 days.
Josh Wilkenfeld (email@example.com): Hi Jim and Joe, Dr. Samocha is out of the country right now, but I’ve forwarded him this thread and asked him to respond, even if it has to be brief for now. I’d like to comment on the reference Joe made to the super-intensive results achieved at the AgriLife Research Mariculture Lab in Corpus Christi during 2009:
The targeted average body weight (ABW) in our super-intensive studies is usually 20 grams, though this year, Tzachi also set the additional objective of trying to beat his previous personal high of 9.29 kilograms per cubic meter (kg/m3), achieved in one out of four raceways operated in a 2007 trial. Because things were going so well in the 2009 cycle, he decided to continue beyond the 20-grams “acceptable” ABW and let the study run until the ambient temperature in the raceways began to drop below 24°C.
The harvest of the four raceways involved in the growout study was carried out on October 24, 2009. Termination had nothing to do with funding; production goals had been exceeded and to hold the animals any longer would not only have been risky, but would have actually hurt production numbers as a result of declining growth that could be anticipated as temperatures dropped.
I’ll leave it to Tzachi to go into the details. Two key numbers: survivals in all four raceways ranged from 94-96% and the yield per cubic meter ranged from a low of 9.34 to a high of 9.75 kg/m3. I honestly don’t know if this is a world record, but Tzachi has a right to be a bit exuberant.
It should also be mentioned that none of this would have been possible without the dedication of Dr. Eudes Correia, a visiting scientist on sabbatical from Brazil who ran virtually all the daily activities during the raceway studies.
The 108-day growout study (stocked at 450 1-gram juveniles per cubic meter) was preceded by a 62-day nursery trial (stocked at 5,000 PL/m3), using the same water, so that at the end of these two studies, the floc and water had been in use for a total of 170 days. As preparation for studies being planned for next year, one of the raceways was refilled with that water (now 170 days old) and restocked with some of the just-harvested shrimp at about 8.5 kg/m3 (I’m a bit fuzzy on the details of this study and the final harvest; Tzachi will need to add more information).
Beginning around Thanksgiving 2009, Tzachi began talking about wanting to look at the question posed by Jim Anderson. Just how long could we keep using the same water? So when the aeration equipment trial was terminated on December 5, 2009, the now 213-day-old water was transferred to an empty raceway and kept “alive” with nothing but the aeration and circulation provided by the airlift pumps.
On December 22, we stocked some Litopenaeus setiferus juveniles (part of an ongoing bait shrimp study) into the water used from the L. vannamei nursery and growout trials. At the time, the water was 230 days old, and I was somewhat concerned about the fact that the floc had not been “fed” for about 17 days. However, all water quality indicators seem to show that the bacteria kicked right back into action, with ammonia and nitrite levels below 0.1 mg/l, although N03-N started at 390 mg/l, compared to 93 mg/l in the control raceway.
Stocking density in this latest study was 230/m3, and the starting biomass was only about 1.2 kg/m3. We do not have temperature control in this raceway system, and our current temperature is about 17.4°C during the day and 16.5°C at night, so we are not exactly pushing the water at high productivity levels at the moment. However, the main intention is to get the setiferus up to bait-shrimp size and then to carry some of the remaining shrimp over as backup broodstock for later in 2010.
At the time of this writing, the water is 245 days old.
With a bit of luck (and unfunded ingenuity), some of the work we are initiating with algae biomass production using flue gas CO2 as a carbon source may make some extra heating capacity available, so that we can extend the growout season in some of our raceways and make it possible to generate more meaningful (though possibly still anecdotal) information about the remediation of seawater used in super-intensive systems.
Eric De Muylder (firstname.lastname@example.org): There are two basic strategies for operating biofloc systems. You can balance the C:N ratio so that there is no accumulation of nitrogen components. The problem with this approach is that you get excess bioflocs and need to find a way to discharge them frequently. Or, you can let nitrates accumulate and adjust pH frequently.
With the second option it is very unlikely that you will be able to use this water for more than two crops. The only solution is to add a denitrification step, but you need a special reactor for that, one that will remove nitrates and balance pH. The basic water parameters will stay constant and you will be able to use the same water for subsequent crops. It might well be that you will have to change some water after a while due to accumulation of some substances.
Russ Allen (email@example.com): I have been growing shrimp in an indoor water reuse system for almost 15 years. Water longevity was one of the first issues I looked into back in the early 1990s. We have used water multiple times for three or more years, and the reasons for changing it were not because it was no longer useable, but usually because we were making modifications to the system. Just like Dallas, our information is proprietary and cost quite a bit of money to research.
We have developed a system for remediating our water, and it looks as if our production results may be better with the remediated water.
Just like the discussion on aeration, results differ dramatically with salinity.
Dallas Weaver (firstname.lastname@example.org): Zero water exchange for one cycle at 10 kg/m3 with a feed conversion ratio (FCR) of 1.5 would result in a CFB of 15,000 mg/l. This is below the level when trace elements and other major water chemistry factors come into play.
Wet sludge that is still pumpable has a solids content of less than 5%, or so.
Sludge discharge systems are reasonably easy to design and operate, but I don’t know of anyone who has recovered the salts and liquid from the sludge and then recycled them back into the system. For an inland marine system, recovering the salt from the sludge is possible and may be required for sludge disposal reasons.
Josh Wilkenfeld (email@example.com): Hi Dallas, In our studies, two of the raceways were run with foam fractionators and two with settling tanks. Tzachi and Eudes will be reporting on all of this at the World Aquaculture Society meeting in March 2010 in San Diego.
Here is the link to the original news release from AgriLife Research: http://agnews.tamu.edu/showstory.php?id=1625.
Dallas Weaver (firstname.lastname@example.org): The fact that you have excess sludge allows you to use the sludge for denitrification. You also get some denitrification in these systems from places where sludge settles. Some nitrate is very useful in preventing sulfate reduction under anaerobic conditions.
In one system that I played with at very high CFBs, nitrate buildup was not an issue. The sludge removal method did the denitrification as a byproduct.
Josh Wilkenfeld (email@example.com): Actually, that’s a big part of what Tzachi has been trying to do with the settling tanks, getting them to remove sludge as well as handle at least some denitrification (and helping control alkalinity and pH). So far, that hasn’t worked out all that well. I think we still need to improve on the design of the settling tanks to get them to work better.
Josh Wilkenfeld (firstname.lastname@example.org): Hi Dallas, Tzachi and I, and with some luck, Eudes also, will be at the WAS meeting, and it would be very nice to see you again. Eudes and Tzachi also ran some small scale denitrification studies just before Eudes left to go back to Brazil, and they probably have some results they’d be interested in talking to you about.
If the settling tanks were up flow with reasonable sludge retention times, you can get a lot of denitrification in the sludge, often combined with a high nitrite output. If the nitrite oxidation capacity of your biofloc is high enough, nitrite will be taken care of.
Jim Anderson (email@example.com, who asked the original question about the longevity of reuse water): I really appreciate everyone’s comments.
Dr. Samocha, Josh and the rest of their team should be commended for setting the bar a little higher for indoor biofloc production. Continuing to look for simplicity, efficiency and the understanding of water reuse is especially critical at inland locations. As Dallas indicated, the cost of the artificial salt and then the removal of it from the sludge are significant.
Dallas...your comment on systems that remove sludge and are easy to design and operate caught my attention. I interpret that to partly mean that if we harvest the excess solids daily as the finfish folks do, which is what we are doing—then the system should become stable from a CFB standpoint? Or am I totally off base? Obviously we are only harvesting what is beyond a certain level of total suspended solids level, not 100% of the floc. In practice, we remove excess solids and floc through a central drain to a settling/sludge tank and let them percolate for 20-ish hours of settling and denitrification. Then we decant back 90% of the “clearish” water to the production tank on a daily basis. We discharge about 1% of system volume as sludge once per week. So we are effectively removing about 50% of system water per year. So the water stays in our system on average 24 months. At this point we are not recapturing the salt, since we have access to a sewer system...which may be desirable for a commercial system. With the combination of solids removal with 1% of the water per week; it looks like we stay under 50,000 CFBs on an ongoing basis? Is that a fair interpretation?
We have not actually run the full system for 24 months in a row, so that’s why I’m very interested in this issue.
Jim Anderson (firstname.lastname@example.org): Dallas, My perception is the sludge removal will remove some of the good cations as well as some of the undesirable heavy metals. We don’t want the metals to build up in our floc...so removing them is desirable. As long as we keep re-balancing our salinity, Ca, Mg and K and we keep adding more feed that becomes sludge, our biofloc water should operate efficiently and still have an acceptable CEC. Right?
Josh Wilkenfeld (email@example.com): Thanks Jim, I’m thinking this is going to be a very interesting WAS meeting this year.
Brian Boudreau (firstname.lastname@example.org): Jim, A large, deep settling tank with good drainage can be used to denitrify and produce up to 100 mg/alkalinity a day. You can use sugar if necessary to get the low oxygen range you need.
Denitrification in sludge tanks is often incomplete, producing nitrite and some ammonia, but we should utilize them because our nitrogen feed input is quite expensive and we want to achieve complete recycling. There are ways to drop NO2 from 10 mg/l to zero in large volumes overnight. Algae production is also an important part of the final water remediation process adding diversity, nutrition and stability to biofloc systems.
If the sludge layer is thick, you get eruptions that bring undesirable products up from the bottom, where the pH is lower, solubility of heavy metals higher and negative ions abound. Charged ions in seawater are like partners at a dance. Every time the music changes, some dancers choose new partners. When you increase alkalinity and pH during the denitrification process, you change the music.
Jim Anderson (email@example.com): Thanks Dallas, CFBs are not what I usually think of, but it is a nice concept. A paper from Louisiana State University on biofloc culture with tilapia suggested minimum CFBs of five, which would mean only using our water for about 60 days (if my math is correct), whereas Dallas is suggesting up to one year. One year is closer to what several folks have done with shrimp. If we use the water that long, can we “refresh” it for further use? Is anyone using water remediation systems that look practical/promising?
Dallas Weaver (firstname.lastname@example.org): I favor remediation systems that provide a lot of control over ion and trace element balances. For biological treatment, I like large fluidized beds with a little ozone to crack some of the chemical oxygen demand; and for mechanical treatment, I like bead filters for solids removal.
I assume you meant maximum CFBs in the 5 kg/m3 range at LSU. That is not a bad number if you don’t want to use ozone or some other method of removing refractory organics that give you colored water (tea color). To get above a CFB of 5, you have to collect, settle and discharge the sludge as a concentrated material. With tilapia, using cheap fresh or brackish water as makeup, there is no reason to go above 5 kg/m3, or 5,000 ppm. If you have expensive synthetic seawater combined with the expenses of disposing of the salt in that water, it is obviously economical to run the CFBs a lot higher.
Dallas Weaver (email@example.com): Jim, Assuming you are using typical feeds, Ca(OH)2 + Mg(OH)2 for pH control, and maintaining Ca and Mg levels in the acceptable range, you shouldn’t have any major ion problems. The sludge does have high levels of Ca, Mg and other cations along with a lower ratio of Na/K than seawater. I have seen some of my freshwater systems become K deficient; however, starting with the higher initial K levels in seawater this probably wouldn’t be an issue. The K levels also depend upon the feed and how much living biomass is in the sludge relative to inert materials.
Unless your feed rates are a lot higher than I expect, your 1% will probably keep you below the 50 g/l CFB range.
Josh Wilkenfeld (firstname.lastname@example.org): Hi Jim, I like to keep total dissolved solid levels between 350-450 mg/l, although Tzachi feels comfortable at levels as high as 600-700. What levels do your shrimp like?
Eric De Muylder (email@example.com): Hello group, I can confirm that the organic matter in the excess sludge can be used for denitrification, even without adding additional carbon. Of course, you need a special setup for the denitrification since you don’t want oxygen to be low in your shrimp tank. You can always get some denitrification inside bigger biofloc particles, settled sludge in the tanks or thicker layers of biofilm, but this is not advisable, and you get the bad consequence of floating sludge. It is therefore advisable to have a dedicated denitrification tank. This is not complicated at all, and it allows you to remove excess sludge and store it to start the next cycle. Denitrification does not remove nitrate completely, but that’s not a problem. When I started the denitrification phase, the nitrate levels dropped from 600 ppm to 100 ppm in 40 days and stabilized at 45 ppm. And as an additional advantage the pH stabilized at 7.5 without addition of bicarbonate.
Brian Boudreau (firstname.lastname@example.org): Dear Eric, I am not sure I understand your chemistry and was hoping that one of the chemists on the list might elaborate on why high alkalinity and pH denitrification in a sludge settling tank can alter core saltwater chemistry.
We know that some of the key ion complexes in floc are deposited in the sludge. Alkalinity in the settling sludge tank can get very high (> 800 mg/l) and lead to precipitation of calcium, magnesium and other carbonates that get lost in the deeper sludge layers. A deep sludge eruption may cause some ions from deep within the sludge to migrate into the high alkalinity waters above. It could be that some positively charged ions coming up in the eruption are attracted to negatively charged bicarbonate ions in the alkalinity and find a way out of the settling tank. If the ions migrating up from the eruptions are not in the same ratios as the ions settling into the sludge, then core saltwater ion ratios can be altered.
If this is true then perhaps denitrification to produce alkalinity is best controlled in batch settling tanks instead of continually in deep sludge tanks, where alkalinity does not get as high.
Jim Anderson (email@example.com): Josh, I am like you and like total suspended solids levels below 400 and harvest them when they get above that. We like relatively stable solids levels, rather than levels that increases significantly as the shrimp grow. We feel our oxygen consumption is lower and more manageable with moderate solids levels because there are fewer bacteria to consume it.
Jim Anderson (firstname.lastname@example.org): Eric, we had the same experience while using a side tank for denitrification. Nitrates were brought right down and alkalinity produced. Savings in bicarb were significant. But as Dallas said, one needs to monitor Ca and Mg levels. We did not see much movement in K levels at this stage. Copper was one element that increased due to the amount of it in the feed.
Dallas Weaver (email@example.com): Eric, Are you using lime for pH control? When adding Ca(OH)2 for pH control, you often get precipitation of CaCO3 which will co-precipitate some heavy metals like Cu.
I was using lime in my recirculating systems (conventional biofilters, not a floc systems) and didn't get Cu build up.
Eric De Muylder (firstname.lastname@example.org): Hello Dallas, No I’m not using any pH control. That’s the advantage of denitrification. The pH drop which is caused by nitrification is reversed by denitrification, so you get a stable pH without addition of chemicals. As for trace mineral buildup, like Cu, I think those minerals may be present in the bioflocs and available to the shrimp. You can reduce the addition of trace minerals in your feed to avoid this buildup.
Dallas Weaver (email@example.com): When operating at high CFBs, you have to keep in mind that the sludge you are removing from the system has cation exchange capacity (CEC) that is nontrivial. This CEC is just like what you get with compost in land based soil systems.
Dallas Weaver (firstname.lastname@example.org): The CEC comment is really what you said about the sludge removing both good cations as well as heavy metals. The cation exchange capacity of the sludge tends to remove multivalent cations relative to Na. The K is lost in the high % of living biomass in the sludge.
You are right, at reasonable CFBs, rebalancing Ca, Mg and K are the big ones and the trace elements will depend upon the inputs in the feed. Commercial shrimp diets may contain more trace elements than more purified diets and may not be a problem. In one of my systems where I did have trace element issues, I was using sugar as a carbon source rather than molasses (a cost and availability issue). Pure sugar has no trace elements.
Since the trace minerals are in the feed, shrimp get the first shot at them. A diet slightly deficient in a trace element will primarily impact the microbiological ecology of the floc.
David Paladines (email@example.com): Hello Eric, Your system is very interesting, but I have some questions about your numbers. I know that nitrification consumes 7.14 parts of alkalinity as CaCO3 for each part of TAN oxidized to nitrate, and denitrification produces about 3.57 parts of alkalinity as CaCO3 for each part of nitrate converted to N2, so in the whole process the production of alkalinity is about half of what is needed for each part of TAN converted to N2, so I guess that in the long term some bicarbonate will be needed to compensate for this. Or is something else happening that keeps the system in equilibrium? Please clear this up for me?