Elsevier

Chemical Engineering Journal

Volumes 215–216, 15 January 2013, Pages 209-215
Chemical Engineering Journal

The impact of contactor scale on a ferric nanoparticle adsorbent process for the removal of phosphorus from municipal wastewater

https://doi.org/10.1016/j.cej.2012.11.006Get rights and content

Abstract

The impact of contactor scale on the efficacy of a ferric nanoparticle embedded media for phosphorus removal was investigated. Experiments were conducted on columns with diameters between 15 and 500 mm, operated at a fixed empty bed contact time of 4 min and an aspect ratio of bed depth to column diameter of 2:1 to ensure self similarity. The columns contained a ferric nanoparticle embedded media, and treated water containing 4 mg P L−1 to simulate applications of full load removal. The treatable flow before breakthrough, the shape of the mass transfer zone and the capacity were all seen to vary with the column diameter used. A logarithmic relationship was observed between column diameter and adsorption capacity such that the capacity increased from 3.4 to 6.3mgPgmedia-1 as the column diameter increased from 15 to 500 mm. Overall the results highlight the importance of considering the scale at which the capacity is measured when assessing the economic suitability of the embedded nanoparticle resin.

Highlights

• A hybrid adsorbent/ion exchanger removes phosphorus from wastewater. • The impact of contactor scale on the efficiency of the process is investigated. • The treatable volume and mass transfer zone vary with contactor diameter. • Capacity changes from 3.4 to 6.3 mg P gmedia-1 in 15 and 500 mm diameter contactors. • Consequently, contactor scale is very important for assessing adsorbent behaviour.

Introduction

One of the core challenges for wide spread application of nanoparticle based treatment systems in water and wastewater treatment is a means of delivering contact between the nanoparticles and water without risk of them exiting with the flow. One of the most promising solutions is to embed the nanoparticles into a scaffold structure, enabling large flow treatment and ease of application in technology based around adsorption and ion exchange [1]. For instance, p materials embedded with ferric nanoparticles for the removal of arsenic and phosphorus [2], [3]. The use of a hybrid ferric nanoparticle adsorbent/ion exchange media is reported in this study that utilises the ligand strength of phosphorus to selectively remove it from wastewater through adsorption onto the ferric component. 90% of the media’s capacity for phosphate resides on the ferric nanoparticles, and competing ions such as sulphate only impact the remaining 10% represented by the ion exchange component. In addition, recovery of the phosphorus is possible through pH switching, which drives an electrostatic release [3], [4] that enables localised regeneration of the media and a route for the production of a commercially viable fertiliser product [5].

Previous studies have utilised laboratory-sized columns ranging from 10 to 150 mm in diameter to trial a range of contemporary adsorbent media (Table 1). This leads to questions of scale-up which require resolution as part of the implementation process. Traditional approaches to the problem are based on assumptions of geometric and kinematic partial similarity linked to the rate critical features of the process. In the case of fixed bed adsorbers, this relates to a spreading of the mass transfer front due to changes in the internal diffusion resistance, external mass transfer resistance and/or axial dispersion in the fluid [6]. Commonly, internal diffusion is considered the most important limiter because the concentration gradient is lower within the media in comparison to the surrounding liquid/solid interface. This means that the dimensionless group Dst/R2, (where Dst is the intraparticle coefficient over time, and R is the external radius of the adsorbent particle) should remain constant during scale up or scale down. Given that Ds and R are likely to remain fixed this means that the fluid residence time, or empty bed contact time, should also be fixed between comparisons of scale.

A number of other practical concerns also need to be considered to ensure partial similarity related to the relative dimensions of the columns as a function of media size is maintained. The impact of the wall must be minimised, which requires the diameter of the column to be 20–40 times that of the media size, otherwise high porosity zones near the wall will exert too much influence on the overall flow pattern [7], [8]. Previous studies do not always meet these criteria, and ratios of column diameter to particle diameter are sometimes unreported, or are below 20 (Table 1). Irregularities in flow pattern caused by packing are more significant in smaller beds [9], which are increased when the Reynolds number exceeds 10 [10]. Associated with this is the bed depth required to ensure constant pattern behaviour which is influenced by the combination of the separation parameter (affinity of the adsorbing species for the media), the diameter of the media and the hydraulic loading rate [6]. Previous studies do not consistently report the depth of the bed (Table 1), making predictions about full scale performance problematic. Media size is known to have the greatest influence with typical minimum bed depth at hydraulic loading rates between 2.4 and 9.7 m3 m−2 h−1 being between 0.13 and 0.35 m for a media with a diameter of 0.5 mm compared to 0.6–2.2 m for a media three times as big [6]. However, the influence of the constant pattern behaviour is most critical for very shallow beds of about a dozen particle layers around which inconsistent performance is observed [11], [12]. The overall impact is that investigations at small scale can lead to poorly defined mass transfer zones and leakage, which makes the understanding of design capacity and scale up parameters difficult.

The current paper explores the issues of scale up for the fore mentioned embedded ferric nanoparticle media used for phosphorus removal as the issue of scale up of nanoparticle based contactors has yet to be reported in the literature. To achieve this, different diameter fixed bed contactors containing nanoparticle embedded media were trialled to establish the implications of column diameter and the impact of scale up on the efficacy of the process. The application described is a sensible one to focus on as it represents potentially one of the largest growth areas for the use of nanoparticle based treatment technologies. Potential changes to European Union legislation as a result of the Water Framework Directive may require sewage discharges to match the background concentration of phosphorus in the receiving water body, currently considered to be around 0.1 mg P L−1 in most cases [17]. Materials such as the ferric nanoparticle adsorbent have advantages over traditional chemical precipitation and biologically enhanced phosphorus removal techniques because they do not rely on dosing (chemical precipitation requires ferric or alum salts and enhanced biological removal is dependent on external carbon), they are resistant to shock loads, flow variation, temperature changes, and the phosphorus can be recovered in a liquid stream, rather than being integrated into a sludge. If the process can be optimised at full scale, and consistently achieve an effluent quality of <0.1 mg P L−1, then an economically favourable model of phosphorus adsorption and recovery into a commercially viable fertiliser product may be possible.

Section snippets

The ferric nanoparticle media

The adsorbent media, LayneRT (SolmeteX Co., Massachusetts) is a commercial product, primarily applied in the removal of arsenic from domestic groundwater supplies. It is known as a hybrid anion exchanger since it comprises a strong base anion exchange resin (macroporous, with a quarternary ammonium functional group), which acts as a scaffold to support a dispersion of ferric oxide nanoparticles [18]. The average particle diameter is 0.69 mm, and the distribution of iron within the beads was

Validating the model feed solution

The suitability of the supplemented groundwater as a surrogate feed during the 100 mm diameter column experiments was confirmed as the breakthrough curve had a statistically similar profile to the real wastewater based on an ANOVA F-test, p < 0.05 (0.04) (Fig. 1). In wastewater treatment, sulphate is the main competing species with phosphate in ion exchange processes [2]. Sulphate levels in the final effluent of five UK wastewater works were measured and were found to be in the range of 130–270 mg

Conclusions

The impact of contactor scale on the performance of a ferric nanoparticle embedded adsorption media has been shown to impact both the operational sequence and the total capacity of phosphorus uptake. The impact of the latter is more severe as lower effluent breakthrough target levels are considered. The overall implication is that appropriate consideration of the scale of the contactor is required to properly understand the potential for using such resins at full scale. The technology has been

Acknowledgements

The authors would like to thank the UK Water Industry Research organisation, and the industrial sponsors of the work: Anglian Water, Northern Ireland Water, Northumbrian Water, Severn Trent Water, Wessex Water, and Yorkshire Water for their financial and technical support throughout the project. In addition the authors would like to thank Dr. Paul Sylvester, of SolmeteX Co., for supply of the media sample.

References (26)

  • C. Pratt, S.A. Parsons, A. Soares, B.D. Martin, Biologically and chemically mediated adsorption and precipitation of...
  • D.O. Cooney

    Adsorption Design for Wastewater Treatment

    (1998)
  • S.V. Jadhav et al.

    Particle-liquid mass transfer in three-phase sparged reactors: scale-up effects

    Chem. Eng. Sci.

    (1990)
  • Cited by (16)

    • Bioaugmentation of pilot-scale slow sand filters can achieve compliant levels for the micropollutant metaldehyde in a real water matrix

      2022, Water Research
      Citation Excerpt :

      The sand bed depth was 0.80 m. Quartz sand (Specialist Aggregates, UK) was sorted to 0.1–0.3 mm size and a uniformity coefficient of 1.35, thoroughly washed with tap water and placed on a 0.20 m gravel bed which had a particle size of 1–5 mm. Given the total diameter of the SSF bed, wall effects on the hydraulics were considered minimal (Martin et al., 2013). Water was supplied to each column using a peristaltic pump (530 U, Watson Marlow, UK) from a 1 m3 holding tank at a flowrate of 1.8 L•h−1.

    • Demonstration of ion exchange technology for phosphorus removal and recovery from municipal wastewater

      2021, Chemical Engineering Journal
      Citation Excerpt :

      The HAIX base material is a strong base anion exchange resin (macroporous, with a quaternary ammonium functional group [31]), within which hydrated ferric oxide nanoparticles have been dispersed to enhance the selectivity for phosphorus [13]. The average particle diameter was 0.69 mm [19]. The secondary treated wastewater originated from Cranfield University WWTP, after preliminary, primary and secondary treatment that ensured solids and chemical oxygen demand (COD) removal.

    View all citing articles on Scopus
    View full text