Steam Stripping for Recovery of Ammonia from Wastewater Using a High-Gravity Rotating Packed Bed

Abstract

Steam stripping of ammonia from ammonia-rich wastewater (5000–20,000 mg/L) was conducted in a continuous-flow rotating packed bed (RPB) at a pH of 11. This study aimed to elucidate the influence of key operational parameters, including the steam-to-liquid ratio, rotational speed (ω), initial ammonia concentration, steam inlet temperature (T_Si), and liquid inlet temperature (T_Li), on critical performance metrics such as the ammonia removal efficiency (ARE), the volumetric liquid mass transfer coefficient (K_La), and the concentration of the recovered ammonia solution (C_R). The findings revealed that a C_R of 22.88 wt.% was achieved under the optimal conditions of a steam-to-liquid ratio of 0.175 kg/kg, an initial concentration of 20,000 mg/L, a T_Si of 120 °C, and a T_Li of 70 °C. Key experimental factors, including the initial ammonia concentration, T_Si, and T_Li, significantly impacted the achievement of higher ARE and C_R values. The K_La values exhibited a decrease with the increase in the steam-to-liquid ratio, while they increased with ω. However, the K_La remained relatively consistent with ω values within the range of 600 to 1200 rpm. In comparison with prior studies, steam stripping of ammonia exhibits a higher ARE than air stripping with RPB and a higher C_R than conventional stripping methods. Moreover, RPB requires a smaller size to achieve equivalent ARE compared to conventional stripping apparatuses. Thus, the steam stripping process with RPB equipment emerges as a suitable method for ammonia recovery from ammonia-rich wastewater.

1. Introduction

In Taiwan, the industrial sector emitted approximately 17,398 tons of ammonia nitrogen (NH₃-N) in 2012 [1]. Ammonia poses severe threats to water bodies, leading to issues such as eutrophication, oxygen depletion, toxicity to aquatic organisms, corrosivity to metals, and pipeline blockage due to the proliferation of microorganisms [2]. Without effective control measures, NH₃-N can vaporize under atmospheric conditions, subsequently reacting with atmospheric nitric and sulfuric acids. This reaction leads to the formation of particulate sulfate ammonium and nitrate ammonium, contributing significantly to secondary inorganic PM_2.5 [3,4]. In response, Taiwan’s Environmental Protection Agency (EPA) gradually implemented stricter regulations on NH₃-N wastewater from high-tech and other industries starting in 2014 [5,6]. According to the latest emissions inventory, the industrial sector’s NH₃-N emissions were significantly reduced to just 18.56 tons in 2021 [7]. However, the volume of untreated domestic sewage is 40,042 tons, contributing to 99.89% of the total NH₃-N emissions from wastewater.

Regarding ammonia removal and recovery processes, various techniques have been studied and reviewed for their technological readiness, including gas–liquid stripping, ion exchange, electrodialysis, membrane separation, adsorption, and struvite precipitation methods, as well as biological processes (nitrification and denitrification) [8]. Among them, the high-gravity rotating packed bed (RPB) method has been proven to be an outstanding candidate in term of gas–liquid phase separation [5,6]. An RPB is an intensified mass transfer reactor of the gas–liquid phase interface induced via high centrifugal force (300–10,000 m/s). The high gravity is 1–3 orders of magnitude greater than that of gravitational acceleration [9]. Yuan et al. [5] showed that the lab-scale continuous-flow RPB they used appears to be very efficient at air stripping ammonia from an ammonia-rich stream (1000 mg/L). The overall liquid volumetric mass transfer coefficient (K_La) values are 12.3–18.4 1/h, which are significantly higher than the values (0.42–1.2 1/h) obtained by using stripping tanks, packed towers, and other advanced gas–liquid contactors. A pilot test at liquid flow rate of 5–11.6 L/min and gas flow rate of 9–18 m³/min also showed similar results [6]. The concentrated gaseous ammonia purged after air stripping was recovered and absorbed by sulfuric acid (H₂SO₄) to produce ammonium sulfate ((NH₄)₂SO₄). Compared to traditional air strippers, continuous-flow RPBs appear to be highly efficient reactors for air stripping, offering a small transfer unit height based on liquid-phase resistance and a short liquid hydraulic retention time. In addition, RPBs have a wider operating range of liquid–gas ratios and are less prone to flooding. However, under normal operation, the pressure drop in an RPB is higher than that in traditional air strippers.

An alternative approach is to obtain a high-concentration ammonia solution through steam stripping. While both air stripping and steam stripping involve similar gas–liquid mass transfer processes, steam stripping is essentially a distillation process conducted at elevated temperatures, typically near the boiling point of water. A notable advantage of steam stripping is the absence of the need for off-gas treatment, as the vapors from the off-gas can be condensed into a small volume of concentrated liquid. Steam stripping has been investigated for the removal of ammonia from various wastewater sources, including wastewater from a metal extraction process [10], reject water from sludge treatment [11], wastewater from a coke plant [12], and anaerobic digestate [13]. Ammonia nitrogen can be reclaimed from the off-gas condensate as an enriched NH₄OH solution following the steam stripping process [11]. However, there is a lack of data concerning steam stripping employing an RPB. One study recently employed the direct steam stripping (DSS) technique for the desorption process of a non-aqueous blended amine absorbent for CO₂ capture in an RPB, aiming to reduce energy consumption. The optimal regeneration energy consumed by DSS of the non-aqueous absorbent in an RPB is approximately 2.46 GJ/ton CO₂, which is 36.6% lower than that consumed in a conventional reboiler regeneration process with a 30% MEA solution [14].

This study investigates the removal and recovery of ammonia from wastewater through steam stripping in an RPB. To the best of our knowledge, this represents the first attempt at applying steam stripping for ammonia wastewater using an RPB. This novel approach leverages the high mass transfer efficiency of RPBs, offering a more compact and efficient alternative to traditional steam stripping methods. The study examines the effect of key operating variables, such as such as the steam-to-liquid ratio, rotational speed, initial ammonia concentration, steam inlet temperature, and liquid inlet temperature, on the ammonia removal efficiency (ARE), the volumetric liquid mass transfer coefficient (K_La), and the recovered ammonia solution’s concentration (C_R). Notably, the steam containing ammonia is condensed and recovered as a high-concentration solution (>20 wt.% of C_R), demonstrating the potential for significant resource recovery. The results of this research provide valuable insights for the development of steam stripping in RPB reactors for ammonia-rich wastewater, contributing to advancements in circular economy practices by enabling the efficient recovery and reuse of ammonia.

2. Materials and Methods

2.1. Apparatus and Operating Conditions

A schematic diagram of ammonia steam stripping using an RPB is shown in Figure 1. Firstly, ammonium chloride (Jin Yih Chemical Co., Ltd., New Taipei City, Taiwan, purity > 99.5%) was dissolved in 10 L of deionized water. Then, NH₃-N solution was added dropwise to adjust the pH of the solution. The pH was confirmed to be above 11. A 50 mL solution was then extracted, and the concentrations of NH₃-N and pH were measured using an ammonia-selective electrode (HANNA, model HI4101, Woonsocket, RI, USA) and pH meter (HANNA, model HI1131B, Woonsocket, RI, USA). This process was repeated twice to ensure reproducibility.

The specifications and dimensions of the RPB system were described by Yuan et al. [5]. The RPB primarily utilizes a controller to adjust the rotational speed, with external heating belts and thermal insulation. The external temperature of the RPB was controlled by a temperature controller to minimize heat loss from the system. After stabilizing the rotational speed, gas and liquid flow rates were set. The influent liquid requires a preheating process. Once the temperature reached the set point, the liquid pump (Cole-Parmer Instrument, model 7554-90 MasterFlex L/S pump, Barnant Company, Barrington, IL, USA) was activated to introduce the configured NH₃-N solution into the RPB through a heated pipeline in an oil-bath heater (IKA Laboratory Equipment, model yellow MAG HS7 S1, Rawang, Selangor, Malaysia). The temperature of the heater decreased at this point, so the measurement of liquid flow was initiated once the temperature returned to the set point. Liquid flow was measured by measuring the liquid weight at the outlet of the RPB for 5 min, and this process was repeated twice for confirmation, followed by calculations to determine the inlet liquid flow rate.

For steam generation, deionized water was injected into the storage tank behind the boiler (Model ST-8 KL, power 3PH, 220 VAC, 60 HZ, 8 KW, BEST A/V Systems Co., Taipei, Taiwan). The boiler power was turned on, and the high-pressure pump injected water into the furnace. The water level controller maintained the water level, and the boiler was heated to the set pressure. Once the pressure reached the set value, the boiler maintained the saturated water vapor at that pressure. Temperature readings were recorded at various locations once flow and pressure stabilized. In the final step, the circulation cooling water tank (Deng Yng, model D606 water batch, Taipei, Taiwan) was opened, and the circulating water was set to a temperature of –20 °C. The peristaltic pump (Cole-Parmer Instrument, model 7554-90 MasterFlex L/S pump, Barnant Company, Barrington, IL, USA) was then activated to extract the condensate solution.

The experiments were conducted with different operating variables, such as steam-to-liquid ratio (0.125, 0.175, and 0.235 kg steam/kg liquid), rotational speed (ω, 300, 600, 900, and 1200 rpm), initial ammonia concentration (C_Li, 5000; 10,000 and 20,000 mg NH₃-N/L), steam inlet temperature (T_Si, 105, 120, 135, and 150 °C), and liquid inlet temperature (T_Li, 25, 40, 55, 70, and 85 °C), to evaluate their effects on ammonia removal efficiency (ARE), the volumetric liquid mass transfer coefficient (K_La), and the recovered ammonia solution concentration (C_R). The pH value was set at 11. The performance of ammonia stripping in the RPB was assessed using the K_La and ARE:

$${\mathrm{K}}_{\mathrm{L}}\mathrm{a}={\displaystyle \frac{{\mathrm{Q}}_{\mathrm{L}}}{{\mathrm{V}}_{\mathrm{B}}}}{\displaystyle \frac{\ln \left[\left(1-\frac{1}{\mathrm{S}}\right)\frac{C_{\mathrm{LCLi}}}{C_{\mathrm{Lo}}}+\frac{1}{\mathrm{S}}\right]}{1-\frac{1}{\mathrm{S}}}}$$

(1)

$$\mathrm{ARE}=\left(1-{\displaystyle \frac{{\mathrm{C}}_{\mathrm{L},\mathrm{i}\mathrm{n}}}{{\mathrm{C}}_{\mathrm{L},\mathrm{o}\mathrm{u}\mathrm{t}}}}\right)\times 100\%$$

(2)

where Q_L is the liquid flowrate (L/min); V_B is the volume of the packed bed (m³); C_Li and C_Lo are the ammonia concentrations of the inlet and outlet liquids (mg/L); and S is the stripping factor, defined as follows:

$$\mathrm{S}={\displaystyle \frac{{\mathrm{L}}_{\mathrm{m}}}{{\mathrm{H}}_{\mathrm{c}}{\mathrm{G}}_{\mathrm{m}}}}$$

(3)

where L_m and G_m are the molar flowrates (mol/s) of the liquid and gas phase, respectively, and H_c is a dimensionless Herny constant at specific temperature.

2.2. Analytical Methods

In this study, the ammonia-selective electrode (HANNA, model HI4101, US) was employed for measurement. The detection range is 0.02 to 17,000 mg/L NH₃-N. To measure the ammonia accurately, the sample water needed to be adjusted to achieve a pH greater than 11. At this point, all ammonium ions (NH₄⁺) in the water are converted to soluble ammonia (NH₃). Through the membrane of the ammonia-selective electrode, which alters its pH, the potential value was measured using the ammonia-selective electrode and reference electrode to determine the concentration of NH₃-N in the sample water. The measurement range of NH₃-N concentration for this electrode is 0.02–17,000 ppm, with an operating temperature range of 0–40 °C, and the pH should be above 11 during operation. Interferences in the measurement can occur due to the presence of surfactants, volatile amines, mercury ions, and silver ions in the aqueous solution.

A standard solution is prepared by diluting a 10,000 mg/L NH₃-N stock solution with deionized water in a 100 mL volumetric flask to create NH₃-N standard solutions with concentrations of 100, 1000, 2000, 5000, and 10,000 mg/L, each at a volume of 100 mL. During the calibration curve preparation and sample testing process, the stirring speed must be controlled to minimize ammonia loss from the solution. It is important to maintain the same stirring speed and a water temperature of approximately 25 °C.

Ammonia-containing steam, after passing through the condensation tube, is condensed into an ammonia solution, flows to the bottom of a round-bottom flask for collection, and is then extracted to a serum bottle by a peristaltic pump. The weight of the condensed liquid was measured after 5 min of collection to calculate the amount of recovered ammonia solution. For the determination of C_R, the density of the condensed solution was initially measured using a density meter (Anton Paar, model DMA 35 Portable Density Meter, Graz, Austria). The measuring ranges (accuracy) for density and temperature were 0–3 g/cm³ (0.001 g/cm³) and 0–40 °C (0.2 °C). Subsequently, the solution was diluted by a factor of 10 or 20. If the pH was below 11 after dilution, a small amount of alkali is added to adjust it to above 11. The ammonia nitrogen concentration was then measured using the selective electrode. All inlet and outlet samples were measured repeatedly in each experiment to ensure the stability of the values.

3. Results and Discussion

3.1. Effect of Steam-to-Liquid Ratio and ω

The dependence of the ammonia concentration in the condensate (C_R, wt.%) on the steam-to-liquid ratio at a pH_i of 12, T_Si of 120 °C, and T_Li of 70 °C is presented in Figure 2. The results showed that the C_R value decreased at a higher steam flow rate. For instance, at a C_Li of 10,000 mg/L and ω of 900 rpm, the ammonia concentration at the condensate reduced by approximately 3 times (11.64 wt.% to 3.98 wt.%) as the steam-to-liquid ratio increased from 0.125 to 0.235 kg steam/kg liquid. This pattern was repeated for different rotating speeds (ω) and initial concentrations of NH₃-N in the influent. In our study, ammonia stripping via the circulation of the liquid stream resulted in a decrease in ammonia concentration in the condensate, which is similar to the results from Zeng et al. [15]. The reason for this phenomenon is that when a larger volume of steam is fed into the system, it results in a higher water evaporation rate, which helps separate the NH₃-N from the wastewater. This evaporated water condenses along with the NH₃-N. As a result, under the same condensation rate, a dilution effect occurs while the amount of NH₃-N present in the system stabilizes and remains constant.

Regarding the change of ω, an improvement in C_R was observed as the rotating speed became faster. At a C_Li of 10,000 mg/L and a steam-to-liquid ratio of 0.125 (Figure 2b), the C_R increased by 3.33 wt.% (8.67 to 12.00 wt.%) as ω changed from 300 to 1200 rpm. This improvement, however, was negligible at higher steam-to-liquid ratios. In the range of 600–1200 rpm, the C_R values are quite close at 0.235 kg steam/kg liquid (Figure 2b,c).

In contrast, a larger steam feeding volume slightly improved the separation of ammonia from the liquid phase. According to Figure 3, the improvement was most noticeable at a ω of 300 rpm, at which the ARE value went up by over 10% by increasing the steam-to-liquid ratio from 0.125 to 0.235 at various C_Li. This finding is in good agreement with previous studies on the stripping of ammonia [5,6], chlorine dioxide [16], and isopropanol [17]. The abundance of fresh steam in the RPB resulted in the separation of NH₃ from the liquid phase into the steam by multiplying the refreshing rate of the steam–liquid interface. As the rotating speed reached 600 rpm and faster, this trend was less significant as the increment in ARE limitedly varied around 2 to 5%. This trend indicated that changing the rotating speed from 300 to 600 rpm had a larger effect on the ARE than that of the steam-to-liquid ratio in the steam stripping of ammonia. The ARE was significantly improved from 91.0 to 98.1% as the ω changed from 300 to 600 rpm. Nevertheless, as the rotating speed varied from 600 to 1200, the ARE remained almost unchanged.

The reverse effect of the steam-to-liquid ratio was recognized on the overall liquid-phase mass transfer coefficient (K_La) for the steam stripping of ammonia (Figure 4). The results indicated a decrement in the K_La value with the increasing amount of steam introduced to the RPB. Specifically, at a C_Li of 10,000 mg/L and ω of 900 rpm, the K_La was reduced by 2.73 kmol/m³s (6.24 to 3.51 kmol/m³s) as the steam-to-liquid ratio increased from 0.125 to 0.235. The results for the K_La of the steam stripping of ammonia described a reverse trend compared to the air stripping of ammonia [5,6] and chlorine dioxide [16]. Although the ARE was enhanced, a larger steam flow rate might result in shortened contact time between the steam and liquid current, reducing the mass transfer rate of ammonia over steam volume as well. Similar to the ARE results, the most significant improvement in the K_La was driven by increasing the rotating speed from 300 to 600 rpm. At a steam-to-liquid ratio of 0.125 and a C_Li of 10,000 mg/L, the K_La value was multiplied 1.6 times (2.82 to 4.57 kmol/m³s) as the rotating speed increased from 300 to 600 rpm. The change was then limited (4.57 to 4.86 kmol/m³s) as ω varied from 600 to 1200 rpm. This pattern might be the result of changes in the liquid flow configuration from a pore flow at 300 rpm to a small droplet and thin-film flow at 600 rpm and higher, which exhibit a larger contact surface between the steam and liquid [18]. Yet, a faster rotating speed led to the decrease in liquid hold up, which might have shortened the contact time between the two phases [19].

3.2. Effect of T_Si and T_Li

The effect of inlet steam temperature on ammonia stripping and recovery using a rotating packed bed is presented in Figure 5. According to the results, the effect of changing the steam temperature from 105 to 150 °C had insignificant effects on all four parameters. According to the results, the ARE only increased from 98.1% to 98.5% as the steam temperature rose from 105 to 150 °C. Since using steam as a stripping medium already greatly enhances the ammonia stripping efficiency, increasing the temperature might not be crucial for the improvement of the removal efficiency. Similarly, increasing the steam temperature also slightly improved the recovered mass flux of ammonia from 2.10 to 2.27 kg/h. A higher steam temperature also indicates a greater enthalpy input for the gas–liquid heat exchange. This increase in energy contributes to an improved evaporation rate of water, resulting in a higher mass flux of the solution in the condensate. However, the enhanced evaporation also led to dilution in the recovered condensate, causing a slight decrease in the concentration of C_R from 7.18 wt.% to 6.51 wt.% as the T_Si increased from 105 to 150 °C. A slight fluctuation in the recovered molar flux of ammonia (ṅ_NH3) was recorded when the steam temperature varied. The ṅ_NH3 was recorded at 8.85 mol/h at a T_Si of 105 °C, then reached a peak of 9.12 mol/h at 120 °C. The ṅ_NH3, however, gradually went down to 8.68 mol/h at a T_Si of 150 °C.

Figure 6 also describes the changes in ARE, C_R, $\dot{\mathrm{m}}$_R, and ${\dot{\mathrm{n}}}_{{\mathrm{NH}}_3}$ for ammonia steam stripping under varying inlet liquid temperatures (T_Li). Similar to the T_Si results, the effect of increasing T_Li from 25 to 85 °C on the ARE was negligible since steam was used for the stripping. The ARE remained almost unchanged around 98% despite the change in T_Li. However, it had a significant effect on C_R and $\dot{\mathrm{m}}$_R in comparison to T_Si. The concentration of ammonia in the condensate sharply decreased from 9.41% to 5.62% as the inlet liquid temperature changed from 25 to 85 °C. The results indicated that the input liquid temperature has a greater impact on the evaporation of water and the condensation rate for ammonia recovery. The dilution of C_R from increasing the T_Li could be the consequence of both a larger amount of water vapor in the outlet air and a lower condensation rate in the recovery phase. Increasing the T_Li also improved the recovered mass flux of ammonia but at a higher pace. Specifically, the $\dot{\mathrm{m}}$_R value went up from 1.52 to 2.50 kg/h. This pattern indicates that increasing the T_Li was mostly responsible for the higher evaporation rate of the liquid phase instead of increasing the T_Si. According to Figure 5d, the ${\dot{\mathrm{n}}}_{{\mathrm{NH}}_3}$ was above 9.0 mol/h at the T_Li range of 40 to 70 °C, while a lower result was recorded at T_Li values of 25 °C (8.40 g-mol/h) and 85 °C (8.27 g-mol/h). A lower inlet liquid temperature could lead to a decrease in the overall temperature of the system, while a too high value might cause an increment in condenser load and consequently decrease the recovery rate of ammonia.

3.3. Effect of Ammonia Inlet Concentration

The effect of C_Li (5000–20,000 mg/L) on the ammonia recovery via steam stripping at ω of 900 rpm, T_Si of 120 ^oC, and T_Li of 70 °C is presented in Figure 7. According to Figure 7a, increasing C_Li from 5000 to 20,000 mg/L had a negligible effect on the ammonia removal efficiency at all three steam-to-liquid ratios. There was only a 1% decrease in the ARE at a steam-to-liquid ratio of 0.125 (from 96% to 95%), while the ARE remained nearly constant at steam-to-liquid ratios of 0.175 (from 97.9% to 98.0%) and 0.235 (from 98.5% to 98.8%). The minimal impact of C_Li on the stripping efficiency is consistent with the findings of Trinh et al., where the stripping efficiency of ClO₂ remained unaffected by changes in C_Li from 100 to 2000 mg/L [16]. Typically, a higher input loading requires a greater mass transfer unit height to achieve the same stripping efficiency in a packed column. Consequently, a reduction in ammonia removal efficiency was observed when a higher ammonia concentration was fed into the stripping column in the study by Zeng et al. [15]. However, in an RPB, this difference might be mitigated by the intensified mass transfer resulting from the high centrifugal acceleration (300–10,000 m/s²), which can minimize the negative effects of a high inlet liquid concentration [20].

In Figure 7b, a significant effect of C_Li on the C_R was observed. For all steam-to-liquid ratios, the C_R value multiplied by 4 to 5 times as the C_Li went up from 5000 to 20,000 mg/L. The condensed recovered solution was recorded at 22.88 wt.% at the highest C_Li of 20,000 mg/L and lowest steam-to-liquid ratio of 0.125 kg steam/kg liquid. These results indicate that the quality of recovered ammonia solution via steam stripping using RPB was considerable (higher than 20,0000 ppm) compared to column stripping (around 10,000 ppm with C_Li of 60,000 mg/L) [15].

3.4. Performance Comparisons and Limitations

Table 1. A comparison of stripping and recovery efficiencies of different equipment for ammonia stripping.

Reactor Types	Equipment Dimensions	CLi (mg/L)	Operation Conditions	Results	Ref.
Steam stripping of RPB	ri = 1.15 cm ro = 6.35 cm ZB = 2 cm VB = 245 cm3	5000–20,000	1. TSi = 120 °C, TLi = 70 °C 2. Steam-to-liquid ratio = 0.175 kg steam/kg liquid	1. Maximum ARE = 98.8% 2. Maximum CR = 22.9 wt.%	This study
Steam stripping in packed column	d = 4.5 cm h = 1.8 m VP = 2.8L	910–45,000	1. TLi = 79–81 °C 2. Steam-to-liquid ratio = 56–70 kg/m3 (0.056–0.070 kg/kg)	1. ARE = 30–96% 2. CR = 0.7–6 wt.%	[15]
Air stripping RPB	ri = 3.55 cm ro = 8.50 cm ZB = 2.15 cm VB = 400 cm3	1000	1. TLi = 40 °C 2. Gas-to-liquid ratio = 1600 (vol/vol)	ARE = 81%	[5]
Air stripping in packed column	d = 5 cm h = 97.5 cm V = 1.913 L	3.39 g/kg (Fresh slurry) 3.68 g/kg (Digested slurry)	1. TLi = 80 °C 2. Gas-to-liquid ratio = 75 (vol/vol) 3. HRT = 4h (recirculation mode)	ARE = 98.8% (fresh slurry) ARE = 96% (digested slurry)	[13]
Steam stripping in packed column	h = 13.1 m d = 1.9 m	14,500	1. TLi = 24 °C 2. Feed stream flowrate = 8.3 m3/h	CLo (simulation) = 135.7 mg/L CLo (experimental) = 130 mg/L	[12]

provides a comparison between steam and air stripping of ammonia via an RPB and other conventional and advanced processes, including detailed equipment dimensions and operation conditions. The advantages of using steam instead of air are outlined, as discussed in the study by Yuan et al. [5]. Despite a higher C_Li and slightly smaller RPB equipment, the ARE in this study was 17% greater than that achieved via air stripping using RPB. The use of RPB, as opposed to a conventional column, offers a larger gas-to-liquid ratio and does not require liquid recirculation to achieve high ARE [13]. When combined with RPB equipment, steam stripping achieves a similar ARE but with more concentrated ammonia condensate compared to the steam stripping column used by Zeng et al. [15]. In summary, the steam stripping process using RPB equipment is well-suited for the recovery of ammonia from ammonia-rich wastewater.

While this study did measure all inlet and outlet samples repeatedly in each experiment to ensure data stability, it was not designed for systematic repetition. Consequently, standard deviations and some statistical tests were not included. However, the reliability of the data is corroborated by our previous studies [5,6], in which part of experimental and analytical systems were the same, demonstrating the consistency and reliability of the system’s performance. In addition, real-world wastewater was not utilized in this work; however, previous work [6] demonstrated comparable performance between model and real industrial wastewater by comparing the air stripping process at laboratory and pilot scales. Nevertheless, it is important to note that certain wastewater parameters, such as suspended solids, could affect both the mass transfer efficiency and pressure drop. Additionally, it should be noted that the reuse of industrial effluent wastewater in Taiwan is limited to industrial applications only. Consequently, effluent wastewater after ammonia recovery can only be reused for industrial purposes. These could be limitations of the process.

4. Conclusions

The study successfully demonstrated the steam stripping process for removing and recovering ammonia from wastewater with a 2 wt.% concentration, yielding a high-concentration (22.88 wt.%) ammonia solution using a rotating packed bed (RPB). Optimal conditions were attained at a steam-to-liquid ratio of 0.175 kg/kg, with an inlet steam temperature (T_Si) of 120 °C and an inlet liquid temperature (T_Li) of 70 °C. Key factors like the initial ammonia concentration, T_Si, and T_Li proved pivotal in achieving enhanced ammonia removal efficiencies and recovery yields.

The overall liquid volumetric mass transfer coefficient (K_La) values decreased with the increase in the steam-to-liquid ratio. Theoretical K_La values increased with the rotating speed (ω), but remained relatively consistent within a ω range from 600 to 1200 rpm. In comparison to previous studies, the steam stripping of ammonia in an RPB exhibited a higher ammonia removal efficiency (ARE) than air stripping using an RPB and a higher recovered concentration (C_R) than in conventional strippers. Moreover, the RPB required a smaller size to achieve equivalent ARE compared to a conventional stripping apparatus. Nevertheless, for the technology to be economically viable, further research is needed to determine the operating conditions that optimize ammonia removal efficiency and recovery yields.