Preparing for better phosphorus removal in AMP7

The water industry is looking forward to around 1000 new final effluent phosphorus consents which will come into force before 2025 as part of the Water Industry National Environment Programme (WINEP).

Currently, market available and economically viable phosphorus removal options are generally limited to biological or chemical removal, with the latter being much more prevalent due to the necessity for a year-round reliable carbon source to support biological phosphorus removal. Chemical treatment uses a metal salt, most commonly iron, to precipitate ortho-phosphate. The precipitate forms as a solid and is removed in a solid’s removal process, either a settlement tank or a tertiary solids capture process such as a disc or sand filter.

In order to ensure that the new consents are met the industry has been carrying out its own research under the Chemical Investigation Programme 2 (CIP2). This has involved pilot trials to assess technologies capable of meeting the proposed low final effluent phosphorus consents in the most economic and reliable manner. The new technologies have mostly involved dosing metal salts (predominantly iron) with a variety of novel tertiary filtration processes designed for highly efficient solids removal to meet the low concentrations required. However, there are some more novel alternatives such as algal treatment and sono-electrochemical technology.

In order to upgrade a wastewater treatment plant to remove phosphorus, designers require information on the nature of the sewage to be treated. Final effluent consents are given as Total Phosphorus and measured as an annual average. Total phosphorus consists of orthophosphate, that can be removed by reaction with metal ions along with organic and inorganic phosphorus compounds which may be both soluble and particulate.

In order to measure the total phosphorus, organic and condensed inorganic forms must be converted to reactive orthophosphate before analysis. Organic phosphates are converted to orthophosphate by heating with acid and persulphate. The orthophosphate reacts with molybdate in an acid medium giving an intense molybdenum blue colour which is measured in a colorimeter.

Chemical P removal
Chemical addition can be employed for phosphorus removal before the primary tanks, into secondary biological treatment or in a dedicated tertiary treatment plant. To achieve consent limits of 1 mg/l or under, multi-point dosing will almost certainly be required. The molar ratio of metal ion to P required for effective phosphorus removal increases as the final effluent phosphorus concentration decreases i.e. the lower the target concentration of phosphorus the greater the relative dose.

This means there is significant benefit in aiming to take out the bulk of the phosphorus in primary treatment (ensuring enough phosphorus remains for the microbiological requirement in secondary treatment) as this will not only reduce the overall chemical consumption but provides the potential operational savings associated with enhanced primary treatment.

Any excess metal salts added to the sewage, due to the overdosing required to achieve low effluent P concentrations will react to remove alkalinity from the sewage. Alkalinity is required for ammonia treatment and if the alkalinity concentration is too low the final effluent ammonia discharge consent may be breached.

A number of unknowns need investigating prior to progressing low phosphorus capital schemes. These include: 1) what is the molar ratio of metal ion to phosphorus to dose; 2) what is the best chemical for the plant – e.g. Ferric Sulphate, Ferric Chloride, Poly Aluminium Chloride (PAC), Aluminium sulphate, rare earth metals or a combination of metal salts and polymer; 3) whether alkalinity dosing is required; and 4) whether alternative technologies, such as sono-electrochemical treatment, magnetite or algae reactors are viable or more economic.

Jar testing is essential to produce the data to determine the dose response curves in order to identify the optimum chemical and dosing location. This should ideally be carried out to incorporate a range of flow conditions to assess the wastewater variability.

Biological P removal
Biological Phosphorus removal can be achieved in an activated sludge plant by incorporating an anaerobic zone and where there are conditions in the influent to the secondary treatment comprising adequate short chain volatile fatty acids (SCVFA) such as acetic, propionic, and butyric acid, suitable for the phosphorus to be absorbed into the sludge. However, where the influent does not contain adequate SCVFA, it is still possible to achieve treatment by adding additional anaerobic zone capacity to enable the hydrolysis of particulate organics to provide a carbon source.

Once biological phosphorus conditions are achieved, activated sludge can store polyphosphate in the range of 3-6% phosphorus by dry weight as compared to 1-1.5% assimilated in a conventional activated sludge plant.

In the UK the phosphorus rich sludge from biological P treatment is then generally removed from the activated sludge plant and transferred for treatment in an anaerobic digestion process. Usually more than 50% of the phosphorus in the sludge is released in the digestion process and tends to be recycled in the return liquors to the inlet of the sewage treatment works. This recycle stream needs to be considered in the design.

These are the unknowns which will need investigating prior to progressing to a capital scheme: 1) Is the crude sewage suitable for Biological Phosphorus (BioP) removal (sufficient SCVFA or hydrolysable solids)? 2) Does the crude sewage characteristics change over the seasons? 3) Is an additional carbon source required? 4) Does the settlement of the sewage change with BioP?; 5)Does the sludge make up change and what are the nutrient characteristics of the sludge? 6) If BioP is applied, then how much phosphorus is released again after digestion/advanced digestion to be recycled to the inlet works? 7) What are the potential impacts of elevated phosphorus in the sludge handling assets i.e. struvite precipitation? 8) What technologies are available to capture the phosphorus in the recycle stream and how reliable and efficient are they?

In order to understand the unknowns above, data, modelling and investigation may be required. This could include sampling and analysis of crude sewage over a period of 1 to 3 years to gather data on total phosphorus and orthophosphorus; COD fractionation to quantify soluble readily biodegradable COD (rbCOD) and to understand the degree of particulate COD hydrolysis required; analysis of short chain VFA (SCVFA) as a fraction of soluble readily biodegradable COD; and diurnal profile measurement.

From the sample data above, modelling software simulations can be run to determine if a VFA fermenter is required. Laboratory scale anaerobic digester trials (possibly including advanced digestion such as thermal hydrolysis upstream of digestion) can be carried out with dewatering of the digestate and analysis of the P concentration in the return liquors. Investigation of P removal processes on the return liquors, and bench scale ASP for BioP modelling, can also be used.

A new service
Aqua Enviro in parnership with EnviroSim (the developers of BioWin process modelling software) are able to provide a holistic assessment to least cost P compliance, incorporating the considerations above. This service includes detailed site assessment, sampling and analysis to include full wastewater fractionation. BioWin models are then developed to consider the changes required, taking into account the interactions between the wastewater and sludge assets. The optimum balance between P removal (chemical and biological) and recovery (i.e. as struvite) can be understood.

The process modelling (BioWin) can be used to investigate a host of factors which affect on the design of the process.

In order to design the most economically viable phosphorus removal process, designers require information about the sewage composition, reactions with metal ions and possibly mathematical modelling to provide confidence for capital expenditure in AMP7. Now is the time to start these investigations.