Talks with Terry: Phosphorous Removal

Hey! It’s been a while – are you ready for one of your helpful “Talks with Terry”?

Sure. It’s why our Marketing & Sales team calls me our “friendliest” engineer.

Huh. I’d have gone for “most knowledgeable” but if you’re happy, we’re happy.

Right. Well, today I thought we’d talk about Phosphorus removal.  We’ve discussed Nitrogen removal and BOD previously, so let’s do the other part of nutrient removal.

Nutrients are supposed to be a good thing, aren’t they? Remind me again why we want to get rid of them?

Too much phosphorus discharged to the aquatic environment can lead to significant, potentially toxic algal blooms.  Often, concerns over toxic algal blooms drives very low phosphorus concentration requirements in treated sewage effluent.

But it can be good, right?

Yep!  Again, like nitrogen it’s an essential growth nutrient, and can be beneficial when treated effluent is released to land for irrigation.  Remember though, care needs to be taken to prevent nutrient laden runoff from entering the aquatic environment (or you have the same problem as directly dumping it there yourself) or from entering the groundwater.

Alright then, I think I’ve got it. And why do you focus on nutrients like Nitrogen and Phosphorous?

Sewage has a whole lot of contaminants in it, but we generally focus on a few that are relatively cheap and easy to measure, and provide a good overview of how contaminated a sample is.

Okay, so excessive amounts are bad.  How do we fix that?

There are two primary methods for removing phosphorus: (EBPR) and Chemical Precipitation.

Whoah.   What does EPBR stand for?

Enhanced Biological Phosphorus Removal.  It works by ‘tricking’ part of the biomass into eating up more phosphorus than they otherwise might.

Sneaky. And the chemical rain one?

You mean “Chemical Precipitation”.

That involves the addition of a chemical coagulant to precipitate out soluble phosphorus and improve the settlement of insoluble or particulate phosphorus.

Can you go through that more slowly?

Basically, you add a chemical which causes the phosphorous, that has dissolved, to clump together or “coagulate”. This helps the rest of the phosphorous (which isn’t dissolved and which is called particulate phosphorous) to settle.

Un huh. It’s crystal clear now.

Awesome, let’s go into more detail.


I’ll start with chemical precipitation.

Chemical coagulants generally consist of Iron or Aluminium salts, or Calcium (lime).

I’ll discuss Iron and Aluminium based coagulants first as they both work in a similar manner.

Yes, do that.

The addition of a chemical coagulant to water results in the release of metallic ions (Al3+ or FE3+).  These ions then go on to either react with the available soluble phosphate, creating metallic phosphate complexes that are insoluble, or create metallic hydroxide complexes.  These hydroxides help to further ‘tie up’ the insoluble metallic phosphates as the metallic hydroxide complexes are generally ‘jelly like’ in nature.

I bet you could put that in an equation if you tried.

Absolutely! Simplistically, in equation(ish) form:


Which means?

These equations indicate that for 1 mole of metal ions, you’ll remove 1 mole of phosphate.  The actual masses of coagulant required to achieve phosphorus removal differs based upon the molecular weight of the active ion (iron or aluminium) and the concentration of the metal within the specific type of coagulant you’re using (we’ll touch on this more later).

Okay, so depending on the metal in the coagulant you’re using, you’ll get a different amount of phosphorous removed?

You bet.

Additionally, studies indicate that these reactions are very much oversimplified; there are a significant number of completing reactions that occur based on pH, temperature, actual wastewater chemistry, etc.  As such, doses will almost always be higher than what you’d predict using this relatively simplistic method.  Ratios of between 1:1 and 1:3 are normally used, with the higher ratio used for lower residual phosphorus concentrations.

So, if it can vary that much, what should you do?

If possible it’s best to use bench-scale trials to estimate coagulant doses.  If you can’t, a process simulator like BioWin can also help, but remember it’s a simulation, so you need to start with good data or the results won’t be overly useful.

Ah. I remember reading about that.

So the coagulant removes the phosphorus then?

Well not really.  The coagulant converts the phosphorus into forms that are easy to settle out.  The precipitate then needs to be removed via sedimentation and/or filtration.  Generally this means the precipitate will be removed in the waste activated sludge.

Right, so that’s Iron and Aluminium, you said Calcium works differently?

Yes, calcium works a little differently.

Normally you’d use Lime (Ca(OH)2), which reacts with the available bicarbonate alkalinity to form CaCO3.  As you add more lime the pH will increase, and above pH 10 excess calcium ions will react with the available phosphate, forming hydroxyapatite (Ca10(PO4)6(OH)2).

Jeepers. Hold on, I’ve got a pen somewhere….

Don’t worry you don’t really have to remember that name.

Whew! That’s a relief. Go on.

For this method, the alkalinity of the wastewater is actually quite important, as you need to essentially exhaust the available alkalinity to get the required pH change.  Generally approximately 1.5 times the alkalinity of the wastewater (as CaCO3) is required to be added (as lime).

As the pH is significantly increased, the pH water discharged normally needs to be reduced prior to any further treatment or release to the environment.  This is normally achieved by re-carbonation using CO2 as this helps balance the alkalinity of the water.

So, you’re putting carbon back in?


The need to re-carbonate and the high amounts of lime required using calcium precipitation for phosphorus removal is now less common then precipitation with Iron or Aluminium, although it can be used in specialist circumstances (removing excess phosphorus from digester supernatant for instance).

Removal depends on how the lime is applied; the precipitate still needs to be settled out and removed, but there can be significant lime solids associated with this process, and as such it may have its own dedicated reactor or clarifier.

So you mentioned that Iron and Aluminium work similarly?  Is there an actual difference, or do people just use whichever one they feel like?

From a reaction standpoint both iron and aluminium salts, being trivalent cations, work in similar fashions.

And I’m going to confuse things here and say that you don’t just buy iron and aluminium, the coagulant is normally available as a salt of the metal, with the most common forms being:

  • Aluminium Sulphate (Alum) – Al2(SO4)3.18H2O,
  • Sodium Aluminate – NaAl2O,
  • Polyaluminium Chloride (PACl) – Al2(OH3)Cl3,
  • Aluminium Chlorohydrate (ACH) – Al2(OH)5Cl,
  • Ferric Sulphate – Fe2(SO4)3,
  • Poly Ferric Sulphate – Fe2(OH)6(SO4)2.7,
  • Ferric Chloride – FeCl

But I can buy them?

Each of these coagulant chemicals are commercially available at different concentrations, and also have different concentrations of their active compound, resulting in a difference in either aluminium or iron yielded per unit mass which can be significant between products.

  • Aluminium Sulphate;
    • Standard solution concentration 48 – 52% w/w,
    • Al2O3 concentration of 7.5-8%,
  • Sodium Aluminate
    • White powder (90%), which you then dilute yourself, or
    • Standard solution concentration 38 – 43% w/w,
    • NaAl2O, concentration 29 – 39%,
  • Polyaluminium Chloride;
    • Standard solution concentration of 20 – 23% w/w,
    • Al2O3 concentration of 10-11%,
  • Aluminium Chlorohydrate;
    • Standard solution concentration of 40-41% w/w,
    • Al2O3 concentration of 23-24%,
  • Ferric Sulphate;
    • Standard solution concentration of 42-43% w/w,
    • FE(III) concentration of 14-15%,
  • Poly Ferric Sulphate;
    • Standard solution concentration of 43.7% w/w,
    • Fe(III) concentration of 12.2%,
  • Ferric Chloride;
    • Standard solution concentration of 41-43% w/w,
    • Fe(III) conf of 14-15%.

ACH packs the most per unit volume, while Aluminium sulphate packs the least.

Any other differences I should know about?

The other major difference between the coagulants is the pH change caused when they’re added to water.  Alum Sulphate generally causes a significant decrease in pH, whilst ACH results a significantly smaller decrease in pH.  Because of this Alum Sulphate will generally require pH adjustment, using usually using lime, soda ash or sodium hydroxide, whilst ACH will generally not require adjustment, or only require minimal adjustment.

Although it should be noted here that Sodium Aluminate works a bit differently; the chemical is based on low-concentration sodium hydroxide, and as such it’s actually strongly alkaline in comparison to the majority of other coagulants; it will actually increase pH whilst also providing phosphorus removal.  This can be quite useful as it simplifies dosing systems, but the chemical is comparatively expensive.

What’s the bottom-line? Ha, seriously though, talk to me about the difference in cost for these?

There definitely is a cost consideration.  Alum Sulphate is by far the cheapest of the chemicals, however it’s relatively low concentration and requirement for pH adjustment will often offset the low cost of the chemical itself, often resulting in a similar overall cost to the more expensive ACH.

Ferric (iron) compounds are generally more expensive in Australia in comparison to their Aluminium counterparts, and as such are rarely used outside of some facilities with specific requirements.

Let’s go back to the other one with the big acronym…the one that mentioned mentioned biological removal. How’s that different?

Biological phosphorus removal, usually called Enhanced Biological Phosphorus Removal, or EBPR revolves around ‘tricking’ bacteria into taking up excessive amounts of phosphorus, storing it within their cells.

Wait, what?

Okay, so, to explain more in-depth, a certain sub-set of bacteria (creatively called Phosphorus Accumulating Organisms, or PAO’s) are capable of absorbing phosphorus and storing it as Polyphosphate within their cells.  They do this in ‘times of plenty’ where there’s lots of food and oxygen around (e.g. in the aerobic zone of a bioreactor).  During times of stress (generally anaerobic phases, or when there’s no oxygen around) these bacteria can break-up this polyphosphate material for use in their metabolic processes (in combination with readily biodegradable carbon materials, generally short-chain volatile fatty acids such as acetate or propionate).  This provides them with a competitive growth advantage under these circumstances.

Like bears that hibernate?

Kind of.

To optimise their growth the sewage treatment reactors are specifically designed to capitalise on this advantage; an anaerobic zone is placed in front of the main activated sludge bioreactor, with raw sewage and recycled sludge fed to it.  PAO’s then use these aerobic zones to absorb phosphorus creating polyphosphates (i.e. our times of plenty), which are then used within the anaerobic zone (i.e our times of stress).  This releases the accumulated phosphorus back into the sludge, which is then absorbed again in the aerobic zone.

To everything there is a season, turn, turn turn?

That’s one way to put it.

By cycling the bacteria between aerobic and anaerobic zones (if you’ve designed your reactors right) you’ll get a net uptake of phosphorus within the activated sludge bacteria.  This accumulated phosphorus is removed as part of the activated sludge (similar to chemical precipitation).

Wow, that sounds like you’re getting removal for free! That’s great!

Kinda sounds like that doesn’t it?  That’s not entirely the case, however, as the process introduces its own issues.

Firstly, the process can be kind of sensitive, it needs good design, good operation and a good source of rbCOD to actually operate effectively, otherwise you tend to create ammonia or BOD issues in your process, as your initial anaerobic zone just acts like a digester.  Because of the way the recycles work EBPR isn’t necessarily universally applicable, and there are some processes that it just won’t really work well with.  This sensitivity tends to preclude its use in small facilities, and I’ve seen at least a few failed attempts at EBPR on small scale that have eventually been mothballed and replaced with chemical precipitation.  This may also require the feeding of the process with a suitable rbCOD source, which increases costs.

So, it doesn’t tend to work on small scale operations. Any other drawbacks?

Secondly, notice that the bacteria release phosphorus in an anaerobic state.  Anaerobic digestion is useful technology; it allows you to reduce your waste sludge mass without the use of power, and in the right circumstances, can be used to create power through biogas harvesting.  This can help to significantly reduce the carbon footprint of your treatment plant, offsetting some of the power required for treatment.  However if you have EBPR you can’t use anaerobic technology (without some further treatment) as all of the phosphorus the bacteria have eaten up will be released straight back into solution, returning to your treatment plant in recycle streams.

As I mentioned you can add other treatment systems, and lime clarification (calcium precipitation) can actually be used here much more effectively than treating the entire flow; volumes are less so the amount of lime and pH correction required are also generally less.  There are also other technologies, such as struvite (a compound formed from magnesium, ammonia and phosphate, which can cause issues with pipeline clogging in wastewater treatment plants) reactors that purposefully create struvite or other magnesium phosphate compounds for recovery of the effluent phosphorus for beneficial re-use.

Hmm…anything else?

Yes. The actual amount of phosphorus that can be removed from the effluent is limited, and generally achieving an effluent phosphorus of below about 3mg/L with EBPR alone is quite difficult, as such chemical phosphorus precipitation is often used to reach relatively low concentrations.  EBPR here is still quite beneficial however, as you reduce the amount of chemical required.

Also, don’t forget the recycle requirement.  Whilst it’s possible to return the RAS recycle from the clarifier to the anaerobic zone, care needs to be taken as the presence of both oxygen and nitrates will reduce the reaction rates, essentially ‘poisoning’ the reactor, so just bear in mind that there may be greater pumping costs associated with additional recycles.

Okay, so now do I know everything I “need to know” about phosphorous removal?

Close. Two more things.  You need to be careful when removing phosphorus. It is an important nutrient, and your treatment plant does need some of it to function.  This tends to be more of an issue with chemical phosphorus removal systems targeting very low (i.e. below 0.5mg/L) phosphorus levels; higher than this and you’re normally ok, but if you’re aiming for lower be aware that it’s entirely possible to lock up all of the available phosphorus with your chemical dosing prior to it being utilised by your bacteria.  This can cause discolouration, sludge bulking and overall poor treatment of BOD and Nitrogen.  This can be mitigated by good design (adding phosphorus removal chemicals prior to sedimentation, not to the incoming flow), the use of post-dosing prior to filtration equipment or very (very) careful operation.  Ideally you should aim to have between 1 and 0.3mg/L of phosphorus in your effluent if your license allows it.

Got it. And what’s the other thing?

In the quest for very low phosphorus concentrations your solids removal equipment becomes very important; you can coagulate all of your phosphates out, but if you can’t actually remove the precipitate you may still be discharging relatively high total phosphorous concentrations.  Well run sedimentation equipment should be able to achieve a TP of down to 2mg/L or even as low as 1mg/L, but to get below 1mg/L you need good filtration equipment as well.  Cloth media disk filters and membrane filters perform well for this duty, although a well-designed and operated sand filter should also work.  That said you may see treatment plants produce a TP of 0.5mg/L just off of their secondary clarifier; this shouldn’t be taken as the norm however.

That’s a lot to think about. Final words on the subject?

Don’t forget that a portion of the available phosphorus will be used by bacteria for growth within the treatment plant, so for plants discharging to land irrigation areas or old plants with less stringent phosphorus targets, you may actually be able to get away without active phosphorus removal.

Keen. See you next time!

Looking forward to it.