10 PPM Phosphate Levels: No need to panic

[sghera64]

Abstract: A calculation error led to the overdosing of tri sodium phosphate (TSP: Na3PO4) to a 700L (135 gallon display plumbed to two 25 gallon frag tanks) mixed reef system with deep sand bed that has been operating for the past 14 years. The overdose resulted in not the targeted 0.010 ppm Phosphate level, but 1000 times more. Instead of responding with the traditional course of successive water changes, increased application of ferric oxide (GFO), the system was allowed to operate as if nothing had happened. Bubble and hair algae growth did increase, but not disastrously. Coral (LPS, SPS) health improved. Phosphate levels returned to non-detect levels after 1 month.


System design

• Display tank: 135 gallon Oceanic glass
• Frag tanks: 2 x 25 gallon Tupperware tubs plumbed in series and back to system sump.
• Top-off: RO water passed through a calcium hydroxide reactor; effluent is <6ppb-PO4 .
• Lighting: 3x250W MH + 4 T5 over display; 1x250W MH + 2 T5 over both frag tanks
• Circulation: 2x Jebao RW-8 in display, 1x Jebao RW-8 in each frag tank
• Sump: DIY 20 gal. aquarium with Iwaki WMD30RLT return pump common to entire system.
• CO2 reactor: DIY
• Deep Sand bed (2-5”): in place for 14 years with periodic additions from beach vacations.
• DIY protein skimmer
• Minimal use of carbon (ROX 0.8)

Initial water parameters:

• Calcium: 390-440 ppm
• Alkalinity: 165-210 ppm CaCO3 (9.3 – 11.7 dKH)
• Mg: 1250-1350 ppm
• pH: 7.8 – 8.25
• Temperature: 77-79 oF
• Lights on: 8 hrs/day

Background

Display tank is modestly populated with soft corals (xenia, mushrooms), LPS (goniopora, wall hammer, torch coral), various SPS, 1 large blue clam. Two to three years ago, Phosphate levels were in the 10-30 ppb-PO4 range (Hanna ULR checker), nitrates were non-detect using a Seachem kit and using a Nylos kit. At that time, goniopora stokesi was reproducing, LPS population was expanding rapidly, SPS had minimally acceptable coloration and growth. There was very little hair algae and a small amount of bubble algae.

Two years ago, a power loss caused a system upset where all the fish perished, except 3 percula clown fish. The coral was not visibly affected. Shortly after that time, 80% of the LPS population was sold off for frags because they were beginning to overwhelm the system. At this point, the goniopora, SPS and remaining LPS began a slow decline. The reason for the decline was not known. In an attempt to improve things, the system was shifted to an ULNS (ultra low nutrient system) reaching non-detect levels of nitrate and phosphate using GFO and carbon dosing (first ethanol, then vinegar).​

The ULNS had a few noticeable consequences:

• The need to clean aquarium glass dropped dramatically
• The film on the glass went from green or tan to milky white
• Bubble algae population declined, but did not go away.
• Hair algae population dropped.
• Chaetomorphia nearly completely perished.
• Steady and rapid LPS coral decline with hair algae growing near polyps.
• SPS population initially improved color vibrancy while growth rate improve marginally.
• SPS color eventually fell back to previous levels and growth rates stalled.

After more than a year of maintaining ULNS conditions, it was decided that different conditions should be explored. Food to fish was increased (frozen, flake). Use of Selcon began along with target feeding of LPS. Oyster Feast was started. The system responded with a small but noticeable improvement in Goniopora. Little to no change in other LPS and SPS was observed.

Based on research and reflection, it was decided to begin nitrate dosing with calcium nitrate to maintain 1ppm-NO3 levels (Nyos test kit). Shortly after 2x daily nitrate additions with a dosing pump and maintaining 1ppm-NO3 levels, the SPS began to increase color. Torch corals still look pathetic. Phosphate levels remained at non-detect (Hanna ULR checker). Following the thought that adding nitrate was a good thing, it was decided to start dosing phosphate in order to maintain a level of 10 ppb-PO4.​

By confusing ppm and ppb, a calculation error led to the accidental overdose of TSP to the system. This happened on October 6, 2016. The amount dosed was 1000 times greater than intended and the system reached 10,000 ppb-PO4 instead of 10 ppb-PO4 [0.010 ppm] (Hanna ULR checker using sample dilution to bring into quantifiable range).

Instead of taking evasive action to immediately correct the problem, it was deiced to see what might happen if no corrective action was taken. Feeding routine was maintained and water changes were stopped to determine how long it would take the system to “consume’ the phosphate. Below is a graph showing the rate of phosphate depletion from the water column.​

The graph ends after 22.5 days with a value of 49 ppb-PO4. The phosphate levels dropped to zero after another 5 more days (Novermber 3, 2016).​

Effects of high phosphate

There was a noticeable increase in hair algae in the frag tanks but not at all in the display tank. Bubble algae broke out and covered areas of live rock in display tank and walls in the frag tank that was open to light and not heavily populated with other coral. The health of the goniopora (polyp extension, coloration) improved. Scaring of the tissue around the skeleton of a red goniopora has begun to heal. Coloration and growth of SPS improved.​

The amount of bubble algae has begun to decline from its population maximum of a few weeks ago. It has not returned to nearly the low levels it was during the ULNS conditions. However, 5 emerald crabs were added. They are not seen eating bubble algae cells >2mm diameter, but they might be eating very small bubble algae cells <1mm.

Below are two pictures of the bubble algae at the worst of the outbreak and currently (November 19, 2016).​

Current Practice

The system is being dosed with calcium nitrate and TSP to maintain 1ppm-NO3 and 10 ppb-PO4. Heavy feeding is still employed. Both nitrate and phosphate are measured weekly to monitor accumulation of them. The system continues to consume nitrate and phosphate. It is not clear if these are being taken up by coral and algae, or processed in the deep sand bed and then removed via protein skimming. Large clumps of hair or bubble algae are removed when found, but this is not likely to explain the consumption of phosphate that is added by food and TSP additions.​

Current Water parameters

• Calcium: 406 ppm
• Alkalinity: 183 ppm CaCO3 (10.2 dKH)
• Mg: 1270 ppm
• pH: 8.22 – 8.38
• Temperature: 77-79 oF
• Lights on: 8 hrs/day

Learning points

• For critical applications, second person verify calculations (use a spouse, reef mate or user groups).

• Making small changes to a system also means avoiding the removal of a large amount of one particular biomass (LPS) from the system all at one time.
  
• High phosphate levels are not necessarily going to lead to disaster [in this system].
  • This system absorbed 12.8 grams of TSP in about 1 month.
    
• ULNS (non-detect levels of nitrate and phosphate) did not lead to healthy corals in this system.
  
• Having access to accurate water testing kits is essential.
  
• It is amazing how nature can restore and rebalance a biotope such as a reef aquarium after a momentary disturbance when one gives it a chance and does not interfere.

I’m curious if anyone has any insights, theories or questions after reading this article. If you spot any flaws in my approach or analysis, please post a reply

 

[Cory]

Thanks for the info. Its possible your rock/ sand could have absorbed a lot of the po4 as your saying you cabt get it to an ULNS.

 

[Randy Holmes-Farley]

Thanks for the info. The tank responded about as I'd expect, and it is great to see some actual data!

One comment relating to this point:


"This system absorbed 12.8 grams of TSP in about 1 month."

We can compare that to what a "typical" tank is processing from food inputs.

Based on data I show here:

Phosphate And Math: Yes You Need To Understand Both
http://www.advancedaquarist.com/2012/3/chemistry

A typical 100 gallon tank is fed and consequently consumes on the order of 10-100 mg of phosphate daily. Over a month that would be 0.3 to 3 grams phosphate.

If you boost the phosphate a lot, the uptake goes up (which is beautifully shown by your graph which is not a linear drop, but is much faster at high concentrations than at lower ones). So these results are also consistent.

 

[Brew12]

Wow, this is fantastic! Thanks for taking the time to put it together!

 

[sghera64]

Thanks for the info. Its possible your rock/ sand could have absorbed a lot of the po4 as your saying you cabt get it to an ULNS.

I was able to achieve ULNS levels (non-detect NO3 and PO4). I felt that ULNS conditions actually stressed my corals and made it more difficult to get the goniopora to recover from previous issues.

I'm really not sure what the PO4 sink is in my system.

Do you really think phosphate can be 'adsorbed' onto the rock and sand? Is there any scientific proof that this actually happens?

I did an experiment a while back where I put a small sample of sand from a system that had high phosphates and lots of GHA into a vial of RO/DI water. I let it sit there for a few months and then tested the water with a Hanna ULR checker. The phosphate level in that RO/DI water remained unchanged (0-ppb PO4). I was thinking that since that sand was exposed to a high phosphate system (~0.050 ppm-PO4) for so long it might adsorb some phosphate. I presume that my results indicate that the sand either did not adsorb the phosphate while in the system, or it did not release it in the RO/DI water.

Thoughts?

 

[Randy Holmes-Farley]

Do you really think phosphate can be 'adsorbed' onto the rock and sand? Is there any scientific proof that this actually happens?

Sure. it is well established fact.

http://yyy.rsmas.miami.edu/groups/jmc/fla-bay/fbay.html

from it:

The results of phosphate adsorption as a function of time show that the adsorption of phosphate on calcite and aragonite is a fast process. Generally it takes about 5 minutes for the adsorbed phosphate concentration to reach a constant value. Aragonite has higher adsorptive capacity compared to that of calcite when the same amount of solid is used. Our results suggest that synthetic solids have a stronger affinity toward phosphate when compared with other natural minerals. Aragonite offers more active surface and more active adsorptive sites compared to calcite.

 

[JasReef]

Pretty amazing write up, thank you. Very detailed but easy to understand. I am immersed in psychological peer-reviewed papers right now and your's was a refreshing change of pace. Again, thank you.

 

[sghera64]

@Randy Holmes-Farley , thank you for the reference. As I understand the article, phosphate can desorb too. I'm wondering why my "desorbing" experiment showed no phosphate.

Since my system is about 14 years old, I would have thought that it might be "saturated" by now. Perhaps my live rock and sand are in equilibrium with phosphate; rapidly binding it when levels are very high (like 10-ppm) only to slowly release it as the biomass incorporates it into tissue and ultimately collecting in the protein skimmer. Do I have that right?

 

[Cory]

@Randy Holmes-Farley , thank you for the reference. As I understand the article, phosphate can desorb too. I'm wondering why my "desorbing" experiment showed no phosphate.

Since my system is about 14 years old, I would have thought that it might be "saturated" by now. Perhaps my live rock and sand are in equilibrium with phosphate; rapidly binding it when levels are very high (like 10-ppm) only to slowly release it as the biomass incorporates it into tissue and ultimately collecting in the protein skimmer. Do I have that right?

Yes and some reefs, and eutrophic areas are controlled by calcium carbonate precipitation binding up po4. I believe everything is seeking balance and that calcium carbonate that loaded with po4 will under po4 deficient water release it. If you had a tank full of hair algae, hair algae is very effective at removing no3/po4. It would have taken it in a matter of days imo.

 

[Randy Holmes-Farley]

@Randy Holmes-Farley , thank you for the reference. As I understand the article, phosphate can desorb too. I'm wondering why my "desorbing" experiment showed no phosphate.

Since my system is about 14 years old, I would have thought that it might be "saturated" by now. Perhaps my live rock and sand are in equilibrium with phosphate; rapidly binding it when levels are very high (like 10-ppm) only to slowly release it as the biomass incorporates it into tissue and ultimately collecting in the protein skimmer. Do I have that right?

Yes, phosphate can reversibly adsorb and desorb from any mineral surface, including GFO, aluminum oxide, and calcium carbonate. In your experiment, I do not know what concentration it as adsorbed at, and you changed many factors, such as fresh water and the pH, both of which likely impact the relative binding. You also waited so long that bacterial and other biological processes may have taken it up.

 

[Stigigemla]

The Phosphate precipation is dependent of pH. In the early days of "Berlin methode" we used kalkwasser to get the pH high enough to get rid of the phosphate.
A potential problem for You might be that the phosphate can slowly go back in solution again if the pH sinks. And that can happen in the sandbed if it gets clogged.

 

[Alfrareef]

The Phosphate precipation is dependent of pH. In the early days of "Berlin methode" we used kalkwasser to get the pH high enough to get rid of the phosphate.
A potential problem for You might be that the phosphate can slowly go back in solution again if the pH sinks. And that can happen in the sandbed if it gets clogged.

Ph high enough to get rid of the phosphate???
How's that?!? Can you please explain that phenomenon?

 

[Randy Holmes-Farley]

The Phosphate precipation is dependent of pH. In the early days of "Berlin methode" we used kalkwasser to get the pH high enough to get rid of the phosphate.
A potential problem for You might be that the phosphate can slowly go back in solution again if the pH sinks. And that can happen in the sandbed if it gets clogged.

I think you may be confounding two different processes.

The binding of phosphate to calcium carbonate is optimal around tank pH, and drops as you go up:

http://yyy.rsmas.miami.edu/groups/jmc/fla-bay/fbay.html

 

[Randy Holmes-Farley]

At the high pH of limewater entry to seawater, you may well initiate precipitation of calcium carbonate, and phosphate may stick to that.

 

[sghera64]

At the high pH of limewater entry to seawater, you may well initiate precipitation of calcium carbonate, and phosphate may stick to that.

I definitely take advantage of this with my switch from using RO/DI water to just RO water for my make-up water. I pass all my make-up water through a DIY limewater reactor. The effluent leaving it had only 0.002ppm PO4 (Hanna Checker ULR Phosphorous) when the reactor's feed water had 6 ppm TDS. I didn't measure the phosphate level of the feed water, but typically have more than 0.010 ppm in my RO (sans DI) water. My understanding is the very high pH exiting the reactor (> 12 pH) precipitates the phosphate as some sort of calcium phosphate complex [Ca3(PO4)2) perhaps].

I believe any metals that get through my RO membrane are also precipitated. For some reason when I clean the magnet off in the reactor, it has a small amount of magnetic black "dust" stuck to it. I imagine this is iron, but I can't think of the source other than what might be in the RO water. There is no metal in contact with the RO water after it leaves the membrane. Does anyone else find a black powder magnetically adhered to their stirbar (if using in the limewater reactor)?

 

[Alfrareef]

Man!!!!! That's a novelty. I'm going to pass a magnet inside my sump and let's see what comes up...

 

[Stigigemla]

In the Berlin method ideal pH was 8,2 to 8,4 and thats near max precipitation.
In the Balling method pH often is between 7,8 and 8,0. But I am a bit confused. I would expext the curve to have a much steeper slope.
I thought it might be more than 10 for each pH unit lower. (dependent on Pka for the different phosphorous acid ions)
Anyway the study was very interesting. Thanks for the link.
One fascinating point is that a higher alcalinity will decrease Phosphate level.
Another was that Aragonite will bind more phosphate than Calcite. I wonder in wich degree the dead rock we sell now has transformed from Aragonite to Calcite.
And how much that matters?

 

[rhastareefer]

I know I'm resurrecting an ancient thread but THANK YOU for posting this. I made the stupid assumption that pharma grade sodium nitrate and sodium phosphate monobasic, when diluted to the same ratio, would raise levels similarly. I just went from ULNS to 3ppm po4 and have been in a panic the last few hours. Guess I'll chill and try not to overreact. Still gonna get a water change ready for tomorrow. Again, sorry for resurrecting an ancient thread but I think it's important to be reminded of stuff like this!!!

 

[Stigigemla]

It will be interesting to see how far the stones will sink the phosphate in the tank.
My guess is 0,1 to 0,2 ppm in 3 days with no water changes or GFO.

 

[Daniel@R2R]

Great info and discussion! Thanks so much for sharing!