Optimization of Nanoparticle Collection by a Pilot-Scale Spray Scrubber Operated Under Waste Incineration Conditions: Using Box–Behnken Design

Abstract

The growing production and application of engineered nanomaterials (ENM) in everyday products make their environmental discharge increasingly likely. This risk is particularly elevated at the end of the useful life of ENM. Given the lack of regulations targeting the safe disposal of ENM at end-of-life, nanowaste (NW) currently follows conventional waste management pathways even though their potential adverse effects on humans and the natural environment are well documented in the literature. Presently, preference exists for treating NW via centralized waste-to-energy systems such as incineration, given the potentially hazardous nature of NW. In this work, a Design of experiment (DOE) methodology—Box Behnken design was employed to optimize three independent operating parameters of a pilot-scale scrubber concerning the collection of nanoparticles under waste incineration conditions. The pilot-scale scrubber was designed and operated to be representative of a full-scale spray scrubber found in a hazardous waste incineration plant. The independent parameters studied are; the gas flowrate, the liquid flowrate and the droplet diameter. Six responses were chosen corresponding to the nanoparticle collection efficiencies at particle sizes in the diffusion-dominant, intermediate and impaction-dominant regions. The experimental results were analyzed for statistical significance using Design Expert software (V13). The DOE model by the software was then used to predict optimum operating conditions with maximum nanoparticle collection. A maximum nanoparticle collection efficiency of 41–60% was predicted by the DOE model under optimal conditions of a gas flowrate of 46.1 Nm³ h⁻¹, a liquid flowrate of 4.8 L min⁻¹, and a droplet diameter of 60 µm. These optimal conditions were determined by the highest liquid flowrate, the smallest droplet diameter, and a gas flowrate near the middle of the range studied. The results of the confirmatory experiments (43–62%) at the optimum combination agreed with the predicted data. The experimental results were also found to be in good agreement when compared to a mechanistic model based on impaction, Brownian diffusion and interception collection mechanisms.

Graphical Abstract

Appendices

Appendix: Supplementary Information

Gas Flowrate

The gas flowrate in a scrubber could (dis)favor the collection of particles by a given mechanism, thus affecting the overall performance of the wet scrubber [28, 42]. In their work on the modeling of PM (1–3 µm) collection efficiency by a spray scrubber with twin-fluid air-assist atomizers, Mohan et al. [38] revealed that at a constant liquid flowrate of 16.67 × 10^–6 m³ s⁻¹, an increase in the gas flowrate (from 1 × 10^–6 to 5 × 10^–6 m³ s⁻¹) led to an improvement in the particle collection efficiency of the scrubber. Similar conclusions were reached by Lim et al. [27] in their study to predict the particle collection efficiency of a reverse jet scrubber. Lim et al. [27] argued that the increase in the collection efficiency as the gas flowrate increased as a result of an increase in the relative velocities between the droplets and the gas (particles). In a spray scrubber, Wu et al. [28] obtained an increment of the dust (20 µm in diameter) removal efficiency from 90.7 to 95.6% when they studied the impact of varying the gas flowrate from 79 to 129 m³ h⁻¹. Wu et al. [28] also reported an increase in the dust removal efficiency from 80.9 to 94.8% when the gas flowrate of a sieve-tray spray scrubber was increased from 83 to 136 m³ h⁻¹. Wu et al. [28] associated this improvement to improve inertia collision with higher gas flowrate. In their research on the study of the removal efficiency of dust in a self-priming venturi scrubber, Ali et al. [43] also concluded that a higher gas flow led to a higher removal efficiency due to improved relative velocity between gas and droplets in the throat section. Generally, as the gas flowrate in a scrubber increases, the particle collection is improved [44, 45]. This is not always the case. Meikap et al. [46] conducted experimental studies on the scrubbing of fly ash (2–15 µm) in a novel wet scrubber using water as the scrubbing medium. Beyond a gas flow rate of 4.2 × 10⁻³ Nm³ s⁻¹, the collection efficiency did not improve with an increase in the gas flowrate but was instead limited due to coalescence which led to reduced bubble reformation. Lee et al. [47] reported that higher gas flowrates do indeed favored the collection of sub-micrometer particles smaller than 1 µm. Lee et al. [47] argues that the gas–liquid contact for effective particle collection is well established for larger particles even at low gas flow rates.

Gas temperature

A change in the flue gas temperature in a wet scrubber could play a significant role in the collection of particles by the scrubber. The gas temperature could also modify other influent parameters such as the droplet diameter, droplet evaporation rate and gas flowrate. Darbandi et al. [48] conducted CFD modeling to investigate the forces responsible for the collection of nanoparticles in a wet scrubber. The authors varied the flue gas temperature from 100 to 300 °C increasing the collection efficiency. The temperature gradient positively impacted the collection of particles due to the diffusiophoresis mechanism. Fan et al. [49] presented in their paper the experimental collection of coal-fired particles (PM2.5) by a condensation scrubber. For the particle size 1.0–3.0 m, the removal efficiency was higher (15–35%) when the inlet gas temperature was set at 95 °C than it was when the gas temperature was at 35 °C (8–35%). The authors attributed this to heterogeneous nucleation and particle collection due to the diffusiophoresis mechanism as a result of the temperature gradient.

Gas Humidity Rate

The collection of particles in a flue gas in a wet scrubber could be improved if the particles are grown into larger ones by a preconditioning technique such as heterogeneous condensation of water vapor with the particles acting as nuclei [21, 50,51,52,53]. The grown particles can then be efficiently collected by an inertial mechanism or a mist separator. Heterogeneous condensation of water vapor is most suitable for gas streams with significant humidity content. Steam is injected in the gas inlet or above the scrubbing liquid inlet to saturate the wet scrubbers with water vapor which leads the latter to condense on the fine particles and to grow their volume. Several authors have investigated this effect on the removal of fine particles in a scrubber. The influence of humidity on the removal of a fine particle by a wet scrubber was studied by Yang et al. [54]. The authors found that an increase in the gas inlet steam led to an increase in the amount of condensable water on the particle surface. This resulted in the improvement of the collection efficiency from 50 to 85%. Huang et al. [19] attained a significantly enhanced collection efficiency of particles with diameters less than 150 nm by mixing saturated steam at 100 °C with a normal temperature waste stream permitting submicron particles to grow into micron sizes. At a liquid-to-gas ratio of 2.5 L m⁻³, the removal efficiency of particles less than 100 nm by the venturi scrubber was greater than 90%. Fan et al. [55] conducted experimental studies on the removal of coal-fired fine particles (PM2.5) by a condensation scrubber. The authors added steam to the gas inlet. This resulted in the improvement of the removal of fine particles as the amount of condensable water vapor increased leading to the condensation growth of particles.

Liquid Flowrate

The liquid flow rate is another very important operating parameter in a wet scrubber. Lehner et al. [56] investigated the aerosol (PM3.0) collection efficiency of a venturi scrubber working in self-priming mode (i.e., operating without an external pump). The authors found that, as the liquid flowrate increased from 1.5 to 3.0 L m⁻³ the collection efficiency for the entire particle range also increased. Lehner et al. [56] attributed this to the increase in the interfacial surface for particle collection at higher liquid flow rates. Meikap et al. [46] reported that at a constant gas flow rate of 3.0 × 10⁻³ Nm³ s⁻¹ in a multi-stage bubble column scrubber, the fly-ash (2–15 µm) collection efficiency increases as the liquid flow rate is increased. Meikap et al. [46] argued that this was due to an increase in the dispersed phase hold-up and interfacial contact area. Meikap et al. [46] conclusion was supported by the investigation made by Calvert et al. [57]. Likewise, as the liquid flowrate increases, the number of collectors also increases leading to a better droplet-particle collision. A similar conclusion was reached by Wu et al. [28] when they developed a mathematical model to investigate the dust removal efficiency of wet flue gas desulfurization systems. The overall spray scrubber collection efficiency increased as the liquid flowrate increased from 1.32 to 2.26 m³ h⁻¹. Wu et al. [28] determined that more intense moving droplets increased the chances of particle capture. Vasudevan et al. [22] conducted lab-scale studies using a spray scrubber to treat the flue gas stream from a fixed bed gasifier. Vasudevan et al. [22] reported that, as the spray liquid flow rate increased from 25 to 30 mL min⁻¹, collection efficiency improved from 50.3 to 60%. Vasudevan et al. [22] attributed this to an increase in the number of droplets due to higher liquid flowrate.

Droplet Diameter

The droplet diameter plays an important role in the collection of particles by a wet scrubber. Danzomo et al. [58] developed a model for the collection efficiency of particles in a counter-current scrubber at different droplet sizes of 500 µm, 1000 µm, 1500 µm and 2000 µm. For 5 µm and 10 µm particle sizes, the collection efficiencies were found to be 89.7% and 98.2% at 500 µm droplet diameter and for a liquid-to-gas ratio of 0.7 L m⁻³. As the droplet diameter increased to 2000 µm, the collection efficiency at 5 µm and 10 µm decreased to 43.9% and 58.8% respectively. Simin et al. [59] investigated the scrubbing of urea duct particles in a spray scrubber using CFD simulations. As the droplet diameter increased from 200 to 1000 µm, the collection efficiency at particle size 40 µm decreased. Simin et al. [59] concluded that the droplet diameter is negatively correlated to the collection efficiency thus, the surface area for particle collection increases with decreasing droplet diameter. Rafidi et al. [60] developed a model to predict the removal of particles in a spray scrubber and compared the results with experiments from a full-scale spray scrubber in a coal-fired power plant. Rafidi et al. [60] reported that the collection efficiency increased as the droplets became smaller. The authors argued that smaller droplets have higher velocities, thus increasing relative velocities between the droplets and the particles, resulting in a higher Stokes number and hence higher collection efficiency by impaction. Claudia et al. [61] developed a mathematical model to evaluate particle capture efficiency in a wet electrostatic scrubber. The authors found that the model predicted a decrease in the capture efficiency for particle sizes 100 nm and 1 µm with an increase in droplet sizes. As droplet sizes increases, both the collisional efficiency and the droplet–particle contact probability are increased and an improvement of removal efficiency is obtained. Larger particles (5 µm) did not witness an increase in the capture efficiency even when the droplet diameters were smaller. Claudia et al. [61] concluded that the effect of a lower collisional efficiency competed with the increase of droplet–particle contact probability leading to lower capture efficiency.

Particle Diameter

Although not an operating parameter, the influence of the particle diameter has been investigated by several authors. The diameter of the particle in a flue gas is arguably the most important parameter in determining its collection by a wet scrubber [22, 59]. Lee et al. [47] evaluated the performance of a turbulent wet scrubber to effectively remove dust particles arising from a coal-powered thermal power plant. Lee et al. [47] disclosed that an increase in the fly ash particle diameter from 0.65 to 5.0 µm led to an increase in the removal efficiency from 43 to 100%. A similar tendency was observed in a venturi scrubber by Mi et al. [62], where the overall removal efficiency improved from 92.7 to 95.1% as the particle size increased from 10 to 45 µm. In a study involving a sieve-tray spray scrubber by Wu et al. [28], the dust collection efficiency improved from 14.8 to 99% as the particle size increased from 1 to 20 µm. In a spray scrubber, Keshavarz et al. [63] achieved a significant increase in the particle collection efficiency from 60% at a particle size of 0.5 µm to almost 90% at a particle size of 4.5 µm. A laboratory experiment to measure the collection efficiency of aerosol particles by water drops was studied by Pranesha and Kamra [64]. The authors reported that the collection efficiency increased with an increase in particle size. Pranesha and Kamra [64] also suggested that with an increase in particle diameter, the collection efficiency may attain an upper limit which may be less than one. Simin et al. [59] observed a similar tendency at a particle size of 10 µm with a maximum collection efficiency of 97% in spray scrubber. Another point to consider is that several authors [16, 40, 60, 64,65,66,67,68] have reported a tendency of the particle collection efficiency to increase with particle diameter, then dip in the intermediate region around 0.1–0.6 µm and then increase till it reaches 100% as the particle diameter increases. As particle diameter increases, the collection due to the impaction mechanism improves leading to the overall performance of the scrubber. Pilat et al. [65] confirmed this when they calculated the particle collection efficiencies of single droplets considering impaction, diffusion, diffusiophoresis and thermophoresis. The particle collection efficiency improved (30–88%) with increasing particle size (0.25–4 µm) predominantly due to the inertial impaction mechanism.