Struvite Precipitation from Centrate—Identifying the Best Balance Between Effectiveness and Resource Efficiency

Abstract

In the context of struvite precipitation, the most significant gap pertains to the transfer of knowledge from scientific research to practical applications. The primary objective of this study is twofold: firstly, to identify the most critical process parameters influencing struvite precipitation and, secondly, to translate these parameters into a pragmatic tool for real-world applications. This study investigates the precipitation of struvite from digestion centrate to obtain information on the appropriate precipitation conditions for different initial chemical compositions. We carried out 24 lab-scale experiments to investigate the effect of varying pH value (7.0–8.5), temperature (5 °C and 33 °C) and initial phosphate concentrations (353; 165; 68 mg/L) on struvite precipitation. Varying the pH had the strongest influence on precipitation efficiencies. Adjusting pH from 7 to 8.5 increased PO₄-P removal from 1.4% to 98.8%, whereas temperature had little impact on PO₄-P removal. Furthermore, we found that a saturation index of at least 1.7 is imperative to precipitate at least 90% of the available PO₄-P. Based on the results, we developed a nomogram showing the resulting saturation index and the associated PO₄-P removal efficiency for variable initial PO₄-P and pH levels. The tool developed in this study enables users to directly identify the so-called ‘sweet spot’, which is the optimal balance between process effectiveness and resource efficiency, for each centrate.

1. Introduction

The European Union (EU) is dependent on phosphorous (P) rock imports to the extent of 84% of consumption, and as a result, has included P on the list of critical raw materials [1]. This fact illustrates the importance of closing nutrient cycles in general, and P recovery in particular, in order to reduce P import dependency and secure local food supply. As wastewater treatment plants (WWTPs) are relevant P sinks, it seems reasonable to recover P from them [2]. The precipitation of struvite is a technology that fits very well into the concept of circular economy, as it enables the recovery of not only P and nitrogen (N), but also the production of a valuable and plant-available fertilizer from WWTPs [3,4]. Struvite is a salt that precipitates in an alkaline environment with an equimolar ratio of Mg, N and P [5]. Centrate from anaerobic digestion is nutrient-rich and thus well-suited to this approach. While NH₄-N in centrate is usually present in surplus, Mg is the limiting component and must be added to the process in order to precipitate struvite. Researchers investigating struvite precipitation have used different nutrient-rich streams, such as artificial liquids [6,7,8,9], animal manure [10,11], source-separated urine [12,13], or anaerobic digestion centrate from WWTPs [14,15]. This work focuses on anaerobic digestion centrate as matrix.

As with any precipitation, struvite precipitation is largely determined by the degree of supersaturation, expressed as the saturation index (SI). The SI represents the logarithm of the ratio between the ion activity product (IAP) and the thermodynamic solubility product (k_sp) [16]. The mathematical expression for the SI is given in Equation (1) [16,17]. As the SI increases, a more saturated environment is created, leading to a faster and more complete precipitation process [5,18]. Increasing SI to enhance struvite precipitation can be achieved either by adding ions (NH₄-N, PO₄-P, Mg) or by increasing the pH.

$$\mathrm{SI}={\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\left({\displaystyle \frac{\mathrm{I}\mathrm{A}\mathrm{P}}{{\mathrm{K}}_{\mathrm{S}\mathrm{P}}}}\right)$$

(1)

SI = saturation index

IAP = ion activity product

k_SP = thermodynamic solubility product

The solubility product k_SP depends on temperature. Researchers have reported slightly different k_SP values across various temperature ranges [19,20,21,22]. However, they all agree on the tendency that k_SP decreases with increasing temperature. The k_SP is often represented as pk_SP which is known as −log(k_SP). The IAP is largely determined by the relevant dissolved ions for the reaction and is defined as the product of the ionic activities of magnesium, ammonium and phosphate according to Equation (2). The activity coefficient of a specific ion is expressed as alpha [23,24,25].

$$\mathrm{I}\mathrm{A}\mathrm{P}=\{\mathrm{a}{\mathrm{M}\mathrm{g}}^{2+}\ast {\mathrm{a}{\mathrm{N}\mathrm{H}}_4}^{+}\ast {\mathrm{a}{\mathrm{P}\mathrm{O}}_4}^{3-}\} [{mlo}^3/L^3]$$

(2)

a_Mg²⁺ = magnesium content [mol/L]

a_NH4⁺ = ammonium content [mol/L]

a_PO4³⁻ = phosphate content [mol/L]

The literature recommends maintaining a pH above 9 for precipitation; however, this can result in elevated levels of alkali consumption [17]. Moreover, researchers recommend using a Mg:P stoichiometric ratio greater than 1:1 [26]. While magnesium in excess is beneficial for struvite precipitation, in a large-scale plant with several hundred cubic meters per day, it has an enormous impact on magnesium consumption, which is a major cost factor [2]. Also, an over-stoichiometric Mg:P ratio may result in undesirable struvite precipitation within the WWTP as the effluent from the struvite precipitation facility is returned to the WWTP [21]. To ensure findings from research translate effectively into practice, it is necessary to consider PO₄-P concentration ranges used in experiments. Researchers have conducted a number of studies that exclusively consider one PO₄-P concentration or only the ranges below 100 mg/L [15,27]. However, these concentration ranges do not reflect all the realistic conditions at WWTPs [28]. Operators must adapt the operation of a struvite precipitation plant to the unique conditions present at each wastewater treatment facility. The PO₄-P concentrations can fluctuate severely from one WWTP to another but also within one WWTP, depending on the biological or chemical process parameters within the sludge line, as shown in the box plots in Figure 1.

Previous studies have utilized varying initial PO₄-P concentrations; however, researchers have not systematically studied a wider spectrum with initial conditions from less than 100 mg P/L up to more than 300 mg P/L [15,27]. Consequently, there is a lack of clarity regarding individual operational issues. The following research questions can be derived from these observations, especially for full-scale applications:

• Is it necessary to maintain a high pH level in centrates at a high SI due to existing ion concentration?
• What is the optimum pH level balancing low PO₄-P concentration after precipitation, while avoiding excessive use of lye for pH adjustment?

Research has identified the effect of temperature on struvite precipitation from centrate [18], but further research is required to fully understand this process [29]. The centrate temperature is typically influenced by the upstream anaerobic digestion, which is often performed at a mesophilic temperature range of approximately 30–38 °C. This leads to the third research question:

3.

How do temperature variations, combined with previously mentioned variations in pH and ion concentration, affect the precipitation process?

Despite researchers studying struvite precipitation for more than four decades from a variety of perspectives and disciplines [16,21,30,31,32,33,34,35], a significant gap remains in transferring this knowledge into practice, particularly regarding the precipitation of struvite from centrates at WWTPs [29]. Following the approval of struvite for use in organic farming in the EU, a new market has emerged, with the potential for significant demand for struvite fertilizer [36]. To meet this demand, more struvite precipitation plants need to be installed at WWTPs. To support this development, the aim of this work is twofold. Firstly, to identify the most important process parameters, and secondly, to translate these parameters into a pragmatic tool for real-world applications to ensure knowledge transfer and thus support the establishment of a circular economy in the wastewater sector. We propose that, on this basis, operators can optimize the operation of current and future struvite precipitation plants, with a concomitant reduction in costs and enhancement of the sustainability of fertilizer production.

2. Materials and Methods

2.1. Matrix

For this study, we used centrate from the anaerobic digestion of excess sludge from the Braunschweig WWTP (Braunschweig, Germany). The WWTP treats wastewater from 350,000 PE through an activated sludge plant, followed by anaerobic sludge treatment. We used the centrate from sludge treatment in its pure form (designated as A) and also diluted with tap water (designated as B and C) in order to obtain a wide range of different PO₄-P concentrations, as seen in

Table 1. Initial composition of the most relevant elements in the matrixes investigated.

		A	B	C
PO4-P	mg/L	353	165	68
NH4-N		1042	542	320
Mg		8.5	8.0	7.6
Ca		35	46	71
Na		119	108	70.5
K		398	222	102

. Theses concentrations represent the possible PO₄-P concentration ranges in WWTPs as shown in Figure 1. The first set of experiments A was carried out using the original pure centrate. The second experiment B involved a 50:50 ratio of centrate to tap water, while the third experiment C utilized a 25:75 ratio. Due to the fact that the centrate was mixed with tap water to produce B and C, the calcium content was increased while other ions were diluted. Nevertheless, the experiments contained an over-stoichiometric amount of ammonium.

2.2. Design of the Experiments

We conducted the experiments in three sets, with each set comprising specific concentrations of PO₄-P and NH₄-N. Each set was executed across two temperature ranges of 33 °C and 5 °C, representing the expected temperature range at WWTPs. We assessed four different pH values, ranging from 7.0 to 8.5 in increments of 0.5. This resulted in a total of 24 test runs which were performed as duplicates. The specific experimental conditions are outlined in

Table 2. Overview of experimental design, initial conditions and results after 180 min.

Set	#	PO4-P [mg/L]	NH4-N [mg/L]	T. [°C]	pH [-]
A	1	353	1024	5	7.0
A	2	353	1024	5	7.5
A	3	353	1024	5	8.0
A	4	353	1024	5	8.5
A	5	353	1024	33	7.0
A	6	353	1024	33	7.5
A	7	353	1024	33	8.0
A	8	353	1024	33	8.5
B	9	165	542	5	7.0
B	10	165	542	5	7.5
B	11	165	542	5	8.0
B	12	165	542	5	8.5
B	13	165	542	33	7.0
B	14	165	542	33	7.5
B	15	165	542	33	8.0
B	16	165	542	33	8.5
C	17	68	320	5	7.0
C	18	68	320	5	7.5
C	19	68	320	5	8.0
C	20	68	320	5	8.5
C	21	68	320	33	7.0
C	22	68	320	33	7.5
C	23	68	320	33	8.0
C	24	68	320	33	8.5

.

2.3. Procedure of the Experiments

Once the pH and temperature had been stabilized, we initiated the experiment by adding MgCl₂ at a stoichiometric P:Mg ratio of 1:1. Subsequent to the initiation of the experiment, we collected samples at 5, 30, 90 and 180 min, immediately filtered them through a 0.45 µm Chompure nylon syringe filter (Membrane Solutions, Auburn, WA, USA) and subsequently analyzed them. We set the duration of the experiment to a maximum of 180 min, as the reaction is expected to be complete within this time according to the literature [6].

2.4. Mixing, Temperature and pH Control

We conducted the experimental studies utilizing a double-walled beaker, wherein water was circulated through a thermostat (RE 104, Lauda, Germany) to regulate the temperature. The beaker was initially filled with 0.5 L of the sample and the mixing was executed using a magnetic stirrer (IKAMAG, Staufen im Breisgau, Germany) operating at a speed of 200 rpm. We controlled the pH by a glass electrode (Mettler Toledo, Columbus, OH, USA) with a 3 mol/L KCl reference electrolyte connected to a control unit (Dulcometer—D1CB, Prominent, Heidelberg, Germany). This unit automatically triggered a dosing of either NaOH or HCl, which could be very precisely controlled. This ensured that the desired pH value remained stable throughout the test period. We calibrated the pH probe daily. Prior to the beginning of an experiment, we adjusted the pH to the desired value and tempered the centrate accordingly.

2.5. Chemicals

The authors decided to utilize MgCl₂ (25% Schuessler salt, technical grade) as the magnesium source for the experimental procedures as it represents the most prevalent chemical in this field [37]. Furthermore, it is regarded as the magnesium source for struvite precipitation at the WWTP in Braunschweig. A notable benefit of this chemical is its simplicity in terms of handling and dosage, given its status as a particle-free, high-concentration liquid. Other magnesium sources, including Mg(OH)₂, MgO, seawater and bittern, were also considered. However, the use of MgO or Mg(OH)₂ presents a dilemma, as while it reduces the need for lye in the process, it concomitantly complicates control at the laboratory scale. A further disadvantage of these alternatives is the handling of the solution, which has to be constantly mixed due to the presence of undissolved particles. Seawater and bittern are also good options, but they are not readily available and therefore have lost their practical relevance. For pH adjustment, we prepared 1 molar NaOH (1 M, Roth, technical grade) and 1 molar HCl (1 M, Roth, technical grade) solutions.

2.6. Analytics

We filtered the samples through a 0.45 µm filter and subsequently analyzed them for ions. The orthophosphate concentration was measured using a cuvette test and UV-vis spectrophotometer (LCK 350, Hach + Lange, Düsseldorf, Germany). We followed the same procedure for magnesium and calcium (LCK 327, Hach + Lange, Germany). Ammonium was determined using an ion-selective probe (NH3 502, WTW, Troistedt, Germany). At the beginning and at the end of each experiment a sample was filtered through 0.45 µm and analyzed with inductively coupled plasma optical emission spectroscopy (ICP-OES) to determine the fate of other ions.

Prior to use in the experiments, we filtered the centrate through a 0.7 µm glass fiber filter (Roth, Germany) to avoid interference from sludge or struvite particles during the particle size distribution (PSD) measurement. At the end of the experiment, we decanted the solids for one hour, after which we collected the precipitate in glass snap-cap vials using a pipette. We collected the solid-free supernatant and used it as the medium for the PSD measurement to avoid possible agglomeration of the particles. Immediately after phase separation, we determined the PSD using dynamic image analysis (Qicpic L02-Lixel, 0.5 cuvette, M5, SYMPATEC, Clausthal-Zellerfeld, Germany). We transferred 250 mL of the supernatant to a 1 L beaker and placed it on a magnetic stirrer. The stirring speed was meticulously calibrated to ensure the absence of air bubbles, as this could otherwise compromise the accuracy of the measurement, while also preventing the settlement of any precipitate at the base of the beaker. We used a peristaltic pump (Medorex—SN 1509-980, Nörten-Hardenberg, Germany) to transfer the sample into the measuring chamber and afterwards into another baker at a flow rate of 200 mL/min and started it before adding the particles. The precipitation sample was homogenized by shaking and sampled with a pipette. We slowly added the sample to the supernatant until 0.2% of the optical concentration was reached. We then started the measurement and allowed it to run for one minute, after which we rinsed the instrument with distilled water. We used the EQPC (equivalent projection area of a circle) evaluation method, representing the diameter of the circle whose projection area is identical to that of the particle.

2.7. Software

We determined the SI for struvite in each experiment using the software PHREEQC version 3 [38]. The sit database was used, which was previously expanded to include struvite [16]. In the literature, the pk_SP of struvite is given in the range 10 °C to 45 °C. For 5 °C, the pk_SP was extrapolated from the data reported by Rahaman. In this work, the pk_SP for 5 °C and 33 °C was determined to be 13.85 (k_SP = 1.41 × 10⁻¹⁴) and 13.1 (k_SP = 7.94 × 10⁻¹⁴), respectively [22].

3. Results

3.1. Effect of Initial Ion Concentration, Applied pH Value and Temperature

The 24 lab scale experiments to identify the influence of pH, initial ion concentration and temperature on the struvite precipitation in anaerobic digestion centrate are presented in Figure 2. For each initial PO₄-P concentration, we plotted a graph separated at 5 °C and 33 °C, resulting in six graphs in total. The graphs are as follows: (a) = 353 mg P/L at 5 °C; (b) = 353 mg P/L at 33 °C; (c) = 165 mg P/L at 5 °C; 165 mg P/L at 33 °C; (e) = 68 mg P/L at 5 °C and (f) = 68 mg P/L at 33 °C. In each graph, we plotted the PO₄-P concentration over time for each pH level. Based on these results, we developed a nomograph. Table 2 provides a summary of the experimental design and the initial start conditions. The discussion will address the impact of various parameters on the centrate’s crystal size distribution, including initial ion concentration, pH, and temperature.

3.1.1. Effect of Initial Ion Concentration and Applied pH Value

Figure 2 shows clearly that adapting the process conditions to the properties of the precipitation matrix is imperative to achieve optimal P reduction and subsequent struvite yield. It also shows that, an increase in pH results in a decrease in the resulting PO₄-P concentration in each set, implying more effective struvite precipitation, a finding that is consistent with the results of previous research [39]. Nevertheless, in earlier studies, researchers frequently analyzed PO₄-P concentrations below 100 mg/L [15,27]. Since we considered higher initial PO₄-P concentrations in the present study, we can also confirm the positive influence of increasing pH on PO₄-P reduction and subsequent struvite precipitation for this concentration range. Furthermore, the use of different concentrations demonstrated that an increase in the initial ion concentration results in increased precipitation at the same pH value. This correlation stands in contrast with investigations promoting a universal optimum pH of 9 or above for struvite precipitation [17,26,40]. Instead, the pH value must be considered in relation to the input concentration. Consequently, the pH value must be selected higher for a low concentration and lower for a high concentration in order to achieve sustainable recovery of P. An increase in either the initial ion concentration or the applied pH value has been demonstrated to result in a higher SI, thereby leading to a subsequent decrease in PO₄-P. Figure 3 shows a contour plot that provides a visual representation of the percentage of PO₄-P reduction resulting from the interaction between different initial PO₄-P concentrations, pH values at 5 °C, and struvite precipitation. The plot demonstrates that as the initial PO₄-P concentration decreases, the pH must increase to achieve the same process efficiency. For instance, in set C, where the initial ion concentration is low, a pH of 8.5 is necessary to achieve a process efficiency of over 80% PO₄-P reduction (C20). In contrast, for the high initial concentrations in set A, the same PO₄-P decrease can be achieved with a pH of 7.0 (A1), and even over 90% with a pH of 7.5 (A2). The results demonstrate that an elevated initial PO₄-P concentration yields a higher PO₄-P removal efficiency.

Irrespective of the initial ionic concentration, all experiments proceed towards a state of equilibrium with respect to PO₄-P, provided that there is sufficient SI at the beginning. For example, we observed no struvite reaction or only an incomplete reaction in experiments C 17, C 18, and C 21 with pH of 7.0 and 7.5, which can be explained by the low or even negative SI of −0.19, 0.45 and −0.31, respectively. Figure 4a shows the equilibrium PO₄-P concentration after 180 min and the resulting SI at 5 °C. It is noteworthy that when SI above 1 is attained, the equilibrium concentration is predominantly influenced by the applied pH, becoming independent of the initial ion concentration. This observation indicates that an increase in the initial PO₄-P concentration results in a greater percentage PO₄-P reduction at the same pH level (Figure 4b), and consequently, a higher amount of struvite [41].

Figure 5 shows that a correlation between the resulting SI values of the experiments and the percentage of PO₄-P reduction during precipitation exists, with a sigmoidal curve fitted for each temperature range. Regardless of the temperature, an SI above 1 is required for successful precipitation, with approximately 75% of the PO₄-P eliminated when the SI exceeds 1.0–1.15. To achieve a 90% PO₄-P reduction from the centrate after 180 min, the precipitation process must have an SI of at least 1.7 at 5 °C and 33 °C. An SI above 2.1 provides little improvement regarding PO₄-P removal and is disproportionate to the amount of caustic used. Corona et al. (2021) and Ye et al. (2016) also describe another disadvantage of a high SI: struvite particle size decreases when SI increases, resulting in more fine particles being lost in the effluent [26,42]. In contrast, an SI within the range of 1.36–1.69 has proven to be conducive to struvite crystal growth in a fluidized bed reactor, suggesting that the optimal SI for PO₄-P reduction and crystal growth may be similar [16,42]. Bhuiyan et al. (2008) report an SI range between 0.3 and 0.77 to be sufficient to precipitate more than 80% of the PO₄-P mean input concentration of 76 mg/L in a fluidized bed reactor and a recycling ratio between 5 and 9 at pH 8.0–8.2 [43]. In this study, researchers used a different k_sp of 13.36 to feed PHREEQC. In 2008, version 2 was current, which did not yet include the sit database that was used for the present work. However, Gonzales-Morales et al. (2021) report an SI of 1.71 with a PO₄-P removal efficiency of 92.7%. It is also noteworthy that SI values greater than 3.19 have been shown to result in 99% removal efficiency [44]. In consideration of the heterogeneity of the research community with regard to the methods of saturation, supersaturation and SI employed, as well as the utilization of disparate models and databases, the comparison of studies and their outcomes can prove challenging [16].

3.1.2. Effect of Temperature

An increase in temperature causes an increase in reaction kinetics and a subsequent reduction in induction time. This effect leads to a higher PO₄-P reduction at the beginning of the experiment at 33 °C, which is compensated by the end of the experiment at 5 °C (Figure 2). This is consistent with the slightly higher PO₄-P reduction observed at 5 °C compared to 33 °C, indicating that efficiency of PO₄-P reduction is consistently slightly higher at lower temperatures (Figure 5). This phenomenon can be explained by the elevated SI as well as by the findings of Ariyanto et al. (2011), who demonstrated that the maximum solubility of struvite occurs at 35 °C [45]. It is evident that a proportion of the precipitated struvite dissolves again at 33 °C, thereby resulting in elevated PO₄-P concentrations. Consequently, we do not recommend the precipitation of struvite at temperatures ranging from 30 to 33 °C, as this results in a lower PO₄-P reduction in comparison to other temperature ranges that exhibit equivalent precipitation conditions. Furthermore, these temperatures favor the co-precipitation of additional ions, such as calcium, as evidenced by Gonzales-Morales et al. (2010) [44].

3.1.3. Influence on Reaction Kinetics

The study demonstrated that the reaction kinetics exert a significant influence on the requisite hydraulic retention time (HRT), consequently affecting the volume required for precipitation tanks. The findings of the study indicate that the kinetics increase in proportion to an increase in either the SI or temperature. Figure 6 presents a contour plot of the SI and the PO₄-P reduction for all experiments, with the data separated according to temperature conditions of 5 °C and 33 °C. The effect of temperature on the reaction was striking at the outset of the experiments; at 5 °C, the reaction was very slow within the first five minutes, whereas at 33 °C, the reaction was almost complete under the same conditions. A PO₄-P reduction of more than 85% within the first five minutes requires an SI of more than 2.25 at 5 °C, whereas an SI of 2.00 is required at 33 °C.

From the results obtained, it can be deduced that as SI increases, so does the reaction speed. Furthermore, it is evident from the results of Mehta and Batstone (2013) that the majority of the reaction is complete after 30 min [46]. After 90 min, the PO₄-P concentration remains constant in all experiments except B9 and B13 due to the pH of 7.0, resulting in an SI of 0.62 and 0.46, respectively. Therefore, we recommend a HRT of two hours as long as sufficient mixing is ensured and there are no short-circuit flows in the reactor. As shown in Figure 6, there is a clear indication of SI values or limits beyond which a further enhancement in PO₄-P reduction is observed. The SI limits are approximately 1.7; 1.1; 0.6 for 5 °C and 1.7; 1.15; 0.55 for 33 °C, corresponding to PO₄-P reductions of 90%, 75% and 25%, respectively.

From the findings of the experiments, in conjunction with the calculated SI and the resultant PO₄-P reduction, a contour plot as seen in Figure 7 was created to facilitate the determination of the resulting SI for variable PO₄-P and pH values. The calculation of the SI was performed utilizing the PHREEQC software version 3, with a Mg:P ratio of 1:1, a constant NH₄-N concentration of 616 mg/L at 25 °C and a pk_SP of 13.26 [21]. The results shown in Figure 7 highlight an ideal operational SI in the range of 1.7 to 2.1, which correlates to a reduction of PO₄-P of over 90% and minimal caustic requirements, representing the sweet spot between process effectiveness and resource efficiency. To the best of the authors’ knowledge, this type of nomogram is presented for the first time in this paper. The tool can be utilized by WWTPs or other plant operators to ascertain the optimal pH level for operation, considering the prevailing PO₄-P concentration in the centrate. The primary benefit of the tool is that time-consuming investigations into the optimized pH can be reduced. Furthermore, the tool enables the determination of whether the process settings are within an appropriate range, thus enabling the reduction of unnecessary NaOH dosing or the adjustment of settings to enhance precipitation efficiency. While the measurement of the pH and the determination of PO₄-P in the input and output for control remain essential, the extent of expensive analyses can be reduced, which in turn reduces the working hours of the personnel. However, it should be noted that the application of the nomogram is not without its limitations, specifically that its efficacy is diminished in matrices that contain elevated concentrations of calcium and iron ions, as these compete for the PO₄-P. It is also important to consider the NH₄-N concentration, as Figure 7 is based on a fixed concentration. If the specific concentration exceeds this value, the SI will increase, while it will decrease at lower concentrations.

3.2. Fate of Other Ions

As we mixed centrate with tap water for set B and C, most of the ions were diluted. However, in all experiments, an over-stoichiometric amount of ammonium to precipitate struvite was still provided. Concurrently, the calcium content increased due to the tap water and the sodium concentration increased due to the pH adjustment. It is therefore conceivable that, at the prevailing alkaline pH values, phases other than struvite may precipitate, e.g., brushite, hydroxyapatite, K-struvite or newberyite [47]. We established the presence of struvite through two methods. Firstly, we estimated the molar reduction of the precipitating partners. Secondly, we identified the struvite crystals under the microscope.

The ICP-OES analyses demonstrated that the majority of other ions exhibited stability within the matrix and did not precipitate. However, we also observed that the element with the greatest reduction, apart from P and Mg, was calcium, particularly at elevated pH and temperature shown in Figure 8. While the reduction within the 5 °C precipitation regime is in a moderate and comparable range for all sets, in set C with the highest calcium concentration at 33 °C and pH 8.5, almost five times as much calcium precipitates. For an absolute understanding of the precipitate composition, a solid phase analysis such as XRD would have to be carried out, which was not possible in this case.

The objective of the present study was to recover phosphorus and, in the most favorable circumstances, struvite. It is improbable that an entirely pure product will be obtained by working with centrate/filtrate from WWTPs, which all possess an individual composition that allows for a wide range of other possible and probable precipitates [47]. Notwithstanding, the most probable impurities cited above, including calcium and potassium, are also essential nutrients for plants and thus do not pose a problem when utilized as fertilizer [48].

3.3. Crystals

3.3.1. Influence on Particle Size Distribution

After each experiment, we retrieved the crystals and determined their dimensions directly in suspension via dynamic image analysis.

Table 3. PSD measurement results for the recovered crystals presented as EPQM, where ×10, ×50 and ×90 indicate the particle size for which 10%, 50% and 90% of the quantity is smaller, respectively.

Set	#	PO4-P [mg/L]	NH4-N [mg/L]	T [°C]	pH [-]	X10 [µm]	X50 [µm]	X90 [µm]
A	1	353	1024	5	7	20.79	43.1	63.15
A	2	353	1024	5	7.5	19.48	36.78	56.80
A	3	353	1024	5	8	8.98	16.13	28.67
A	4	353	1024	5	8.5	25.25	38.2	56.27
A	5	353	1024	33	7	28.43	71.08	103.25
A	6	353	1024	33	7.5	41.61	71.79	104.68
A	7	353	1024	33	8	30.64	51.92	78.7
A	8	353	1024	33	8.5	22.41	38.11	58.52
B	9	163	542	5	7	9.05	24.35	41.6
B	10	163	542	5	7.5	12.67	28.06	44.5
B	11	163	542	5	8	7.1	12.26	20.66
B	12	163	542	5	8.5	6.48	9.87	15.41
B	13	163	542	33	7	24.51	47.31	82.61
B	14	163	542	33	7.5	22.83	37.9	56.92
B	15	163	542	33	8	13.05	25.61	41.38
B	16	163	542	33	8.5	14.5	24.29	38.15
C	17	68	320	5	7	-	-	-
C	18	68	320	5	7.5	-	-	-
C	19	68	320	5	8	7.44	13.35	21.58
C	20	68	320	5	8.5	8.53	13	20.41
C	21	68	320	33	7	-	-	-
C	22	68	320	33	7.5	4	21.04	99.88
C	23	68	320	33	8	39.86	75.15	120.64
C	24	68	320	33	8.5	55.65	92.83	134.43

presents the outcomes for the PSD ×10, ×50 and ×90, indicating the particle size at which 10%, 50% and 90% of the quantity, respectively, is smaller. However, in set C, we could not obtain sufficient measurements for all experiments due to the low number of crystals formed and recovered in the experiments C17, C18 and C21. We can observe from all experiments in sets A and B that the resulting crystal size decreases with increasing pH, which is also supported by other research [49]. We observed an exception to this in experiment A4, where the small crystals tended to stick together in the suspension, resulting in distorted measurements. For set C, we observed the opposite effect of pH on crystal size, with increasing pH being associated with increasing average particle size. In C24, we collected the largest crystals of all experiments at pH 8.5. In general, the crystal size increased with increasing temperature, irrespective of the initial ion concentration. We obtained the largest average crystals at 68 mg/L PO₄-P, NH₄-N = 320 mg/L, pH = 8.5 and 33 °C with a size of 92.83 µm; the smallest at 9.87 µm at 165 mg P/L, 542 mg N/L, pH = 8.5 and 5 °C. Figure 9 presents microscopic images of the crystals.

3.3.2. Morphology of the Crystals

As demonstrated in Figure 9, the precipitation conditions result in a wide range of crystal morphology. The authors have attempted to select representative crystals for each experiment, but it should be noted that the images in Figure 9 are merely illustrative. While the temperature exerts the least influence on PO₄-P reduction, it predominantly influences the shape of the crystals. The crystals not only become larger at 33 °C, but also tend to grow more irregularly. This phenomenon could be attributed to the acceleration of precipitation shown in Figure 6. This process has been observed in environments characterized by high supersaturation, such as those with elevated pH levels or initial ion concentrations. This acceleration in crystal growth is evident in experiments A4 and A8 of Figure 9, where the majority of crystals exhibited an X-shaped morphology. In set C, characterized by low PO₄-P and elevated calcium concentrations, elevated temperatures give rise to crystals of a markedly elongated and needle-like nature (C22, C23, C24). This tendency to this effect is also observed at 5 °C (C19, C20), albeit to a lesser extent. The coffin-lid shape, for which struvite is renowned (i.e., A5, B11), is the common morphology observed between these two extremes. Similar observations have been made in other studies [49]. Although a correlation of SI with the resulting PSD would be very beneficial in predicting the associated crystal size, this research is not suitable due to the limited amounts of centrate used for the experiments. In order to better understand the effect of initial ion concentration, pH and temperature on crystal growth, a larger amount of centrate needs to be used. This could be the subject of further research.

4. Conclusions

In order to ensure the continued safety of food supplies, we must prioritize the recovery of P. Struvite has emerged as a promising solution to address this issue, as it has the potential to facilitate the closure of the nutrient loop. This research not only advances our understanding of struvite precipitation but also provides practical tools for its implementation, marking a significant step towards sustainable phosphorus management and circular economy principles in wastewater treatment.

The following points summarize the main findings of this study with regard to the effect of pH, initial ion concentration, temperature and SI on the PO₄-P elimination and the crystals formed during struvite precipitation:

pH:

• Increasing the pH accelerates the precipitation process.
• PO₄-P strives for a state of equilibrium, provided there is a sufficient SI. The resulting PO₄-P concentration depends on the applied pH, independent of the initial ion concentration.
• The size of the recovered crystals decreases with increasing pH at high and moderate concentrations in set A and B, whilst doing the opposite for low concentrations in set C.

Initial ion concentration:

• Higher percentual PO₄-P elimination is possible with a higher initial PO₄-P concentration for the same pH and temperature conditions; by decreasing the initial ion concentration, operators must increase the applied pH of the process in order to achieve the same percentual P-removal.

Temperature:

• Temperature has the least effect on the PO₄-P removal. P removal is slightly better but slower within the first 30 min at 5 °C.
• Higher temperature has a positive effect on the resulting crystal size compared to the same conditions at 5 °C.

Saturation Index:

• An SI above 1 is required for a noticeable precipitation effect. No further PO₄-P reduction occurs after 90 min at an SI of above 1.
• An SI of 1.0–1.15 will precipitate about 75% of the PO₄-P for 5 °C and 33 °C, respectively. To precipitate 90% of the PO₄-P, an SI of 1.7 is required for 5 °C and 33 °C.

The results obtained from this study have enabled the creation of a nomogram as seen in Figure 7, which contributes to a more profound comprehension of struvite precipitation. This tool enables practitioners to do the following:

• Identify optimal operating conditions for specific centrate compositions.
• Effectively control struvite precipitation in real-world applications.
• Balance process effectiveness with resource efficiency.

While this study focused on centrate from anaerobic digestion sludge, the principles uncovered here may be applicable to other phosphorus-rich waste streams, such as urine, dairy wastewater, and agricultural residues. Further targeted investigations into these streams are necessary to maximize phosphorus recovery and meet future market demands.