Technology for recovery of phosphorus from animal ...

Technology for recovery of phosphorus from animal wastewater through calcium phosphate precipitation M. Vanotti and A. Szogi USDA-ARS, Florence, SC, USA

Abstract A wastewater treatment process was developed for removal of phosphorus from livestock wastewater. The phosphorus is recovered as calcium phosphate with addition of only small quantities of liquid lime. The process is based on the distinct chemical equilibrium between phosphorus and calcium ions when natural buffers are substantially eliminated. It was discovered that reduction of carbonate and ammonium buffers during nitrification substantially reduces the Ca(OH)2 demand needed for optimum P precipitation and removal at high pH. The technology produced consistent results in pilot tests on ten swine farms and successfully demonstrated full-scale on two swine farms in North Carolina, USA. It can be used to retrofit animal lagoons or in new systems without lagoons. The recovered calcium phosphate can be recycled into a marketable fertilizer without further processing due to its high content (>90%) of plant available phosphorus. The concentration grade obtained during full-scale demonstration was 24.4 + 4.5% P2O5. A second generation version of the technology is available for municipal and agricultural wastewater and includes the simultaneous separation of solids and phosphorus from wastewater and industrial effluents.

INTRODUCTION The aspect of P reuse is important for crop producers because of increasing demand and cost of inorganic fertilizers. The merging of food and fuel economies has increased the demand of mineral P fertilizer, and its price increased over 200% in 2007 (Trostle, 2008). The increased P demand may stimulate new technologies and economic opportunities for P recovery from manure, specially using technologies that produce concentrated byproducts with nutrient values competitive with mineral fertilizers. Reindl (2007) classified techniques for P removal from wastewater by precipitation of calcium phosphate into three groups: crystallizers, fluid bed reactors, and the new process used in this chapter. The technology was developed to remove P from animal wastewater and other high-ammonia strength effluents and has the advantage over previous art of requiring minimal chemical addition and producing a valuable by-product (Vanotti et al., 2003; Vanotti et al., 2005). It is based on the distinct chemical equilibrium between # 2009 The Authors, International Conference on Nutrient Recovery from Wastewater Streams Vancouver, 2009. Edited by Don Mavinic, Ken Ashley and Fred Koch. ISBN: 9781843392323. Published by IWA Publishing, London, UK.

458

International Conference on Nutrient Recovery

phosphorus and calcium ions when natural buffers (NH4-N and alkalinity) are substantially eliminated. Basic process configuration (Figure 1) The processes involved in the new technology include (i) biological nitrification of liquid manure to oxidize ammonium (NHþ 4 Þ to nitrate ðNO3 Þ, (ii) reduction of natural buffers, and (iii) increasing the pH of the nitrified wastewater through addition of Ca or Mg hydroxide to precipitate P (Vanotti et al., 2003; Vanotti þ ½AQ1 et al., 2005; Szogi and Vanotti, 2009). Since NH4 is mostly converted to NO3 , increased pH during P precipitation does not result in significant gaseous N loss in ammonia gas form. The final product is a Ca phosphate-rich sludge that can be used as P fertilizer (Szogi et al., 2006; Bauer et al., 2007). CALCIUM HYDROXIDE INJECTION

INFLUENT WITH AMMONIA, CARBONATE ALKALINITY, AND PHOSPHORUS

NITRIFICATION BIOREACTOR

PHOSPHORUS SEPARATION REACTOR

EFFLUENT WITH NITRATE

PHOSPHORUS PRECIPITATE

Figure 1. Schematic showing the basic configuration of the P removal process (Vanotti et al., 2005).

Process chemistry Animal wastewater is a mixture of urine, water, and feces. Livestock urine usually contains more than 55% of the excreted N of which more than 70% is in the form of urea (Sommer and Husted, 1995). Hydrolysis of urea by the enzyme urease produces NH4þ and carbonate according to the following reaction: 2 COðNH2 Þ2 þ 2H2 O ! 2NHþ 4 þ CO3

ð1Þ

½AQ2

Technology for recovery of phosphorus from animal wastewater

459

Therefore, a substantial part of the inorganic carbon in liquid manure is produced during decomposition of organic compounds. Carbonate and NHþ 4 alkalinity are the most important chemical components in liquid manure contributing to the buffering capacity in the alkaline pH range (Fordham and Schwertmann, 1977; Sommer and Husted, 1995). Alkaline pH is necessary to form a P precipitate with Ca and Mg compounds (House, 1999). When a Ca or Mg hydroxide is added to liquid manure, the hydroxide reacts with the existing bicarbonate to form carbonate, with NHþ 4 to form ammonia (NH3), and with phosphate to form phosphate precipitate compounds (Loehr et al., 1976; Tchobanoglous and Burton, 1991). For instance, using Ca hydroxide as an example, the following equations define the reactions: CaðOHÞ2 þ CaðHCO3 Þ2 ! 2CaCO3 # þ2H2 O

ð2Þ

5Caþþ þ 4OH þ 3HPO¼ 4 ! Ca5 OHðPO4 Þ3 # þ3H2 O

ð3Þ

The reaction in Equation 2 is complete at pH # 9.5, while that of Equation 3 starts at pH 4 7.0, but the reaction is very slow below at pH # 9.0. As the pH value of the wastewater increases beyond 9.0, excess Ca ions will then react with the phosphate, to precipitate as Ca phosphate (Equation 3). Not expressed in Equation 2 is the fact that in wastewater containing high NHþ 4 concentration, large amounts of lime are required to elevate the pH to required values since NHþ 4 reaction tends to neutralize the hydroxyl ions according to Equation 4: þþ CaðOHÞ2 þ 2NHþ þ 2H2 O 4 ! 2NH3 " þ Ca

ð4Þ

Consequently, precipitation of phosphate in animal wastewater using an alkaline compound such as lime is very difficult due to the inherent high buffering capacity of liquid manure (NH4-N $ 200 mg L-1 and alkalinity $ 1200 mg L–1). The buffer effect prevents rapid changes in pH. However, this problem is solved using a pre-nitrification step that reduces the concentration of both NHþ 4 (Equation 5) and bicarbonate alkalinity (Equation 6) (Vanotti et al., 2005): þ NHþ 4 þ 2O2 ! NO3 þ 2H þ H2 O

ð5Þ

HCO3 þ Hþ ! CO2 " þH2 O

ð6Þ

The buffering effect of NHþ 4 (Equation 4) is reduced by biological nitrification of the NH4þ (Equation 5). Simultaneously, the buffering effect of

460


bicarbonate (Equation 2) is greatly reduced with the acid produced during nitrification (Equation 6). These two simultaneous reactions leave a less buffered liquid in optimum pH conditions for phosphate removal with the addition of small amounts of lime (Equation 3).

PROCESS APPLICATIONS TO LIVESTOCK WASTEWATER The new P removal technology was conceived to remove P in systems with lagoons (Vanotti et al., 2003; Szogi and Vanotti, 2009) and systems without lagoons (Vanotti et al., 2007; Vanotti and Szogi, 2008). In the livestock systems with anaerobic lagoons (or other anaerobic digesters), the anaerobically digested supernatant liquid, rich in NH4-N and alkalinity, is nitrified and P is subsequently removed by adding hydrated lime. The effectiveness of the technology was tested in a pilot field study at ten swine farm in North Carolina, where 95–98% of the P was precipitated from the anaerobic lagoon effluent. In the systems without lagoon, raw liquid manure is first treated through an enhanced solid-liquid separation process with polymers to remove most of the carbonaceous material from the wastewater. The separated water is then treated with the nitrification and soluble P removal sequence. A denitrification tank can also be incorporated into the treatment system to provide total N removal in addition to the P removal. This configuration was tested full-scale in two finishing swine farms in North Carolina with removal efficiencies of 94% for soluble P (Vanotti et al., 2007). Phosphorus extraction from digested swine lagoon effluents (Figure 2) Phosphorus was efficiently removed from lagoon liquid from ten diverse North Carolina swine production farms. The study consisted of ten consecutive experiments using the basic process configuration shown in Figures 1 & 2 to treat ten lagoon liquids (Szogi and Vanotti, 2009). In each experiment, swine wastewater first received biological treatment in a nitrification bioreactor, followed by chemical treatment with Ca(OH)2 in a P separation reactor to precipitate phosphate. This configuration was compared with a control representing a control method that also received chemical treatment with Ca(OH)2 but without the nitrification pre-treatment. All control and nitrified lagoon liquids were treated with Ca(OH)2 applied at seven rates of 0, 2, 4, 6, 8, 10, and 12 mmol Ca L-1 of lagoon liquid. Even though a pH $ 9.0 is needed to optimize precipitation of phosphate using Ca-based compounds, the pH of the liquid was initially lowered with the


461

CONFINED LIVESTOCK

RAW MANURE

AERATION TANK P

PHOSPHORUS NITRIFICATION SEPARATION CLARIFIER BIOREACTOR REACTOR TREATED EFFLUENT

PRIMARY SETTLING UNIT (LAGOON)

BLOWER B RETURN SLUDGE

PHOSPHATE PRECIPITATE

Figure 2. Schematic drawing of the unit used to remove phosphorus from swine lagoon liquid using a nitrification-lime treatment sequence (Vanotti et al., 2003).

acid produced by nitrifying bacteria. In this way, the resulting nitrified liquid was low in NHþ 4 and bicarbonate buffers, and the total amount of alkali needed to increase the pH above 9.0 was significantly reduced. Without exception, the natural carbonate alkalinity of the lagoon liquid was significantly reduced in the process of biological NHþ 4 oxidation during the nitrification pre-treatment (average alkalinity changed from 2279 + 300 mg L–1 to 91 + 14 mg L–1). On average, the rate of change of effluent pH (0.42 pH units/mmol Ca L–1) for the pre-nitrified lagoon liquid was significantly different (P 5 0.0001) and higher than the rate for the control without nitrification (0.05 pH units/mmol Ca L–1). These results illustrate that precipitation of phosphate using lime in untreated livestock effluents is very inefficient because of the high natural buffer capacity of these effluents, which prevents rapid changes in pH. Total P removal efficiencies with the non-nitrified (control) treatment never exceeded 50%, even at the highest Ca rate of 12 mmol L-1 used in the tests (Figure 3). Instead, high TP removal efficiencies of 4 90% were achieved at Ca rates between 8 and 10 mmol L–1 (0.3 and 0.4 g Ca L–1) in the nitrified samples. Nitrification pre-treatment significantly reduced the amount of Ca(OH)2 needed for optimum P precipitation and removal while preventing N losses via NH3 volatilization. The two final products of this wastewater treatment process were a liquid effluent for on-farm use and a solid calcium phosphate material.

462


% Total P Removed

100

Without Nitrification With Nitrification

75

y = -0.18 + 18.8x - 0.89x2 R2 = 0.99**

50 25

y = -0.95 + 6.42x - 0.21x2 R2 = 0.99**

0 0

2

4 6 8 10 Ca Application Rate, mmol L-1

12

Figure 3. Phosphorus removal enhancement from swine lagoon wastewater using hydrated lime and nitrification pre-treatment. Results are means (s.e.) of the treatment of ten lagoon effluents in North Carolina (Szogi and Vanotti, 2009).

Manure treatment systems without lagoon (Figure 4) The on-farm technology used liquid-solid separation, nitrification-denitrification, and soluble phosphorus removal processes linked together into a practical

CONFINED LIVESTOCK

RAW WASTE P

DENITRIFICATION NITRIFICATION BIOREACTOR UNIT UNIT PHOSPHORUS P CLARIFICATION SEPARATION RECYCLE UNIT REACTOR UNIT TREATED EFFLUENT

PAM

HOMOGENIZATION TANK

SOLID SEPARATION UNIT

M MIXER BLOWER B SOLIDS & BIOLOGICAL SLUDGE EXCESS BIOLOGICAL SLUDGE

PHOSPHATE PRECIPITATE SLUDGE (PS)

RETURN BIOLOGICAL SLUDGE DEWATERING & PACKAGING MODULE

Figure 4. Schematic drawing of the swine waste treatment system without lagoon. The soluble P is removed after liquid-solid separation and ammonia removal (Vanotti et al., 2007).


463

system. It was developed to replace anaerobic lagoon technology commonly used in the USA to treat swine waste (Vanotti et al., 2007). Hydrated lime precipitated the phosphates and the phosphate precipitate was dewatered using polymer and filter bags (Szogi et al., 2006) (Figure 5).

Figure 5. Phosphorus separation module at Goshen Ridge Farm. Pictures show reaction chamber and settling tank (above) and P precipitate dewatering and bagging (below).

Results obtained at full-scale in a 4360-swine farm showed removal efficiencies of the soluble P of 94.9% for wastewater containing 76 to 197 mg L–1 soluble P (Table 1). The recovered P precipitate solid had a concentration grade

½AQ3

TSS BOD5 TKN NH4-N NO2þNO3-N TP Soluble P Alkalinity pH

Water Quality Parameter

11,612 + 6746 3046 + 2341 1501 + 567 838 + 311 1.5 + 4.5 566 + 237 131 + 39 5001 + 1695 7.64 + 0.22

Raw Flushed Swine Manure

811 + 674 923 + 984 895 + 298 796 + 297 0.4 + 2.6 168 + 53 116 + 33 4154 + 1463 7.93 + 0.26

After Solid-Liquid Separation ‡

134 + 75 40 + 44 43 + 34 31 + 34 228 + 110 149 + 33 138 + 28 624 + 470 7.29 + 0.70

mg L-1

After Biological N Treatment

232 + 152 10 + 16 26 + 25 14 + 19 235 + 116 26 + 16 7+7 763 + 353 10.53 + 0.63

After Phosphorus Treatment

Treatment Step

98.0 99.7 98.3 98.3 – 95.4 94.7 84.7 –

%

System Efficiency

Table 1. Wastewater treatment plant performance and system efficiency at Goshen Ridge Farm, North Carolina. Data are means ( + s.d) for 18-month period (Vanotti and Szogi, 2008).

464 International Conference on Nutrient Recovery


465

of 24.4 + 4.5% P2O5 and was 4 99% plant available based on standard citrate P analysis used by the fertilizer industry (Bauer et al., 2007). Although a high pH (10.5) in the phosphorus removal process is necessary to produce calcium phosphate and kill pathogens, the treated effluent is poorly buffered, and the high pH decreases readily once in contact with the CO2 in the air. For example, Vanotti et al., (2003) showed that short-term (2.5-h) aeration treatment of the effluent could create enough acidity to lower the pH from 10.5 to 8.5. However, natural aeration during storage may be equally effective to lower pH as seen in the converted lagoon described in the following section. On average, the advanced treatment system reduced total N (TKN þ NO3-N) concentration from 1503 to 261 mg L-1 (83% reduction) and TP from 566 to 26 mg L–1 (95% reduction) (Table 1). In addition to substantial reductions in land requirement due to the reduced N and P loads after advanced treatment, the N:P ratio of the liquid was improved from 2.65 to 10.04. This higher N:P ratio resulted in a more balanced effluent from the point of view of crop utilization. A second generation version of the technology is available for municipal and agricultural wastewater and includes the simultaneous separation of solids and phosphorus from wastewater and industrial effluents (Garcia et al., 2007). The combined separation process is more efficient in terms of equipment needs and chemical use. Thus, it reduces installation and operational cost of manure treatment.

CONCLUSIONS Manure phosphorus (P) in excess of the assimilative capacity of land available on farms is an environmental concern often associated with confined livestock production. A wastewater treatment process was developed for removal of phosphorus from livestock wastewater. It includes nitrification of wastewater to remove ammonia and carbonate buffers and increasing the pH of the nitrified wastewater by adding an alkaline earth metal-containing compound to precipitate phosphorus. Since ammonia nitrogen has been mostly converted to nitrate, increased pH does not result in significant gaseous nitrogen loss. The amount of phosphorus removed, and consequently the N:P ratio of the effluent, can be adjusted in this process to match specific crop needs or remediate sprayfields. In addition to the phosphorus removal aspect, the high pH used in the process destroys pathogens in liquid swine manure. The final product is calcium phosphate that has the potential to be reused as fertilizer or processed to produce phosphate concentrates.

466


REFERENCES Bauer, P., Szogi A.A. and Vanotti, M.B. (2007). Agronomic effectiveness of calcium phosphate recovered from liquid swine manure. Agron. J., 99(5), 1352–1356. Fordham, A.W. and U. Schwertmann (1977). Composition and reactions of liquid manure (gulle), with particular reference to phosphate: III. pH-buffering capacity and organic components. J. Environ. Qual., 6(1), 140–144. Garcia, M.C., Vanotti, M.B. and Szogi, A.A. (2007). Simultaneous separation of phosphorus sludge and manure solids with polymers. Trans. ASABE, 50(6), 2205–2215. House, W.A. (1999). The physico-chemical conditions for the precipitation of phosphate with calcium. Environ. Technol., 20(7), 727–733. Loehr, R.C., Prakasam, T.B.S., Srinath, E.G. and Yoo Y.D. (1976). Development and demonstration of nutrient removal from animal wastes. U.S. Environmental Protection Agency, Washington, DC. Reindl, J. (2007). Phosphorus removal from wastewater and manure through hydroxylapatite. An annotated bibliography. Available at: http://danedocs.countyofdane. com/webdocs/PDF/lwrd/lakes/hydroxylapatite.pdf Sommer, S.G. and. Husted. S (1995). The chemical buffer system in raw and digested animal slurry. J. Agric. Sci. Cambridge, 124(1), 45–53. Szogi, A.A., Vanotti, M.B. and Hunt, P.G. (2006). Dewatering of phosphorus extracted from liquid swine waste. Bioresour. Technol., 97(1), 183–190. Tchobanoglous, G. and Burton, F.L. (1991). Wastewater engineering: Treatment, disposal, and reuse. Irwin/McGraw-Hill, Boston, MA. Trostle, R. (2008). Global agricultural supply and demand: Factors contributing to the recent increase in food commodity prices. USDA – Economic Research Service, Outlook WRS-0801. Vanotti, M. B., Szogi, A.A. and Hunt, P.G. (2003). Extraction of soluble phosphorus from swine wastewater. Trans. ASAE, 46(6), 1665–1674. Vanotti, M.B. and Szogi, A.A. (2008). Water quality improvements of CAFO wastewater after advanced treatment. J. Environ. Qual., 37, S86–S96. Vanotti, M.B., Szogi, A.A. and Hunt, P.G. (2005). Wastewater treatment system. US Patent No. 6,893,567, Issued May 17, 2005. U.S. Patent & Trademark Office, Washington, D.C., USA. Vanotti, M.B., Szogi, A.A., Hunt, P.G., Millner, P.D. and Humenik, F.J. (2007). Development of environmentally superior treatment system to replace anaerobic swine lagoons in the USA. Bioresour. Technol., 98(17), 3184–3194.

AUTHOR QUERIES AQ1 AQ2 AQ3

Reference Szogi and Vanotti, 2009 is not cited in the list. Please check. We have inserted a short running head. Please approve or provide an alternative Please provide footnote for the indicator ‡.