Abstract

In this study, 11 reverse osmosis (RO) concentrate samples from six full-scale potable reuse facilities in the southwestern United States were treated by a new photobiological treatment process using brackish water diatoms at a laboratory scale. While eight out of the 11 RO concentrate samples were successfully treated by the photobiological treatment, the other three samples obtained from the facilities where non-nitrified effluent was used as a source water were unsuitable for the treatment due to high levels of ammonia-N in these samples. As low as 16 mg · L−1 of ammonia-N was found to be inhibitory. Lower pH was found to be undesirable because of lower calcium removal efficiency. Ozone pre-treatment and the presence of antiscalant appeared to have no impact on the photobiological process. This study demonstrated a wider applicability of this photobiological process for the treatment of RO concentrate from potable reuse facilities with different process schemes.

INTRODUCTION

Water reclamation and water reuse has become a global phenomenon in recent years due to the increasing pressure on fresh water supplies from climate change, rapid population growth, and environmental pollution (Asano et al. 2007). In addition to long-practiced non-potable water reuse such as urban landscape irrigation, agricultural irrigation, cooling towers and industrial reuse, indirect and direct potable reuse has gained much popularity in the southwestern United States, such as California, Arizona, New Mexico, and Texas, as well as several eastern states such as Florida, Georgia, and Virginia (Rodriguez et al. 2009; Leverenz et al. 2011). Many of the potable reuse projects employ reverse osmosis (RO) as a main barrier against chemical and biological contaminants, as well as dissolved solids. Although the RO process can produce very clean permeate which is almost readily drinkable, it also produces a concentrate stream that contains most of the dissolved constituents originally present in the feed water at 4 to 6.6 times higher concentrations depending on the permeate recovery rate (75% to 85%, respectively) and rejection rates. Concentrate management has always been a challenging issue for advanced water purification facilities (AWPFs) because of the cost, availability, and environmental impacts of concentrate management and disposal (Pérez-González et al. 2012).

The addition of secondary RO or other membrane-based processes such as electrodialysis reversal to recover more permeate and reduce the volume of the concentrate stream with or without pre-treatment has been proposed and studied at various locations (Kawahara 1994; Venkatesan & Wankat 2011; Subramani et al. 2012). However, those processes are often highly energy- and/or chemical-intensive, and mechanically and operationally complex. Also, permeate recovery is limited by the presence of various scaling constituents (scalants) such as silica, calcium, phosphate, and carbonate/bicarbonate (Malaeb & Ayoub 2011).

Recently, a new photobiological treatment process using brackish water diatoms has been developed to remove multiple scalants as well as other inorganic and organic constituents from RO concentrate (Ikehata et al. 2017, 2018a, 2018b). Simultaneous removal of reactive silica, phosphate, ammonia, nitrate, calcium, bicarbonate, iron and manganese from RO concentrate samples obtained at two AWPFs and one brackish groundwater desalination facility (BGDF) has been demonstrated using mixed and isolated diatom cultures and light-emitting diode (LED) bulbs or natural sunlight as a light source (Ikehata et al. 2018a, 2018b). The feasibility of additional water recovery from photobiologically treated RO concentrate by a secondary RO process has also been demonstrated (Ikehata et al. 2018a). Our previous studies showed a great potential of this new photobiological process as a pre-treatment of AWPF (Ikehata et al. 2018a) and BGDF (Kulkarni et al. 2019) RO concentrate to achieve additional RO desalination for more water recovery. However, only a few RO concentrate samples from two AWPFs had been tested so far. To investigate the applicability of the photobiological process at different locations, we have collected 11 RO concentrate samples from six full-scale AWPFs in California and Arizona and compared their treatability in this study. The impacts of pre-treatment and the source water quality (e.g., nitrogen) and pH on the removal of reactive silica and calcium were also investigated.

MATERIALS AND METHODS

A brackish water diatom Pseudostaurosira trainorii E. A. Morales PEWL001 previously isolated from agricultural drainage water (Ikehata et al. 2017) was used in this study. A unialgal culture of P. trainorii PEWL001 was maintained in a filter-sterilized (0.2-μm filters) RO concentrate sample from the Groundwater Replenishment System (Facility A in Table 1), Orange County Water District (Fountain Valley, CA, USA) as described previously (Ikehata et al. 2018b).

Table 1

The first eight RO concentrate samples tested

Sample ID 
Facility ID 
Location CA CA CA CA CA CA CA AZ 
Nitrification Partial Partial Partial Partial None Full Full Partial 
Pre-treatment MF MF MF MF Ozone–MF Ozone–BAC–MF/UF Ozone–BAC–MF/UF Ozone–MF 
RO configuration 3 Stage 2 Stage 2 Stage 3 Stage 3 Stage 2 Stage 3 Stage 3 Stage 
Permeate recovery (%) 85 85 85 92 85 85 85 85 
TDS (mg·L−16,210 3,880 3,650 6,650 6,430 6,670 6,600 6,500 
pH 8.0 8.2 7.5 7.1 7.1 7.2 7.2 7.3 
Reactive silica (mg·L−1140 78 113 197 91 83 86 56 
Calcium (mg·L−1600 416 312 624 580 680 744 632 
Alkalinity (mg·L−1 as CaCO31,020 1,080 935 1,450 1,150 710 670 410 
Orthophosphate (mg·L−1 as PO43−6.3 8.5 1.1 1.9 39 30 29 20 
Ammonia-N (mg·L−18.6 4.1 2.3 4.9 259 0.05 0.05 12 
Nitrate-N (mg·L−156 23 28 52 73 74 26 
COD (mg·L−1212 154 104 209 385 112 164 218 
Color (PtCo U) 246 96 43 88 213 24 26 79 
Reactive silica removal rate (mg·L−1·day−139 28 33 56 −3.6 31 34 19 
Sample ID 
Facility ID 
Location CA CA CA CA CA CA CA AZ 
Nitrification Partial Partial Partial Partial None Full Full Partial 
Pre-treatment MF MF MF MF Ozone–MF Ozone–BAC–MF/UF Ozone–BAC–MF/UF Ozone–MF 
RO configuration 3 Stage 2 Stage 2 Stage 3 Stage 3 Stage 2 Stage 3 Stage 3 Stage 
Permeate recovery (%) 85 85 85 92 85 85 85 85 
TDS (mg·L−16,210 3,880 3,650 6,650 6,430 6,670 6,600 6,500 
pH 8.0 8.2 7.5 7.1 7.1 7.2 7.2 7.3 
Reactive silica (mg·L−1140 78 113 197 91 83 86 56 
Calcium (mg·L−1600 416 312 624 580 680 744 632 
Alkalinity (mg·L−1 as CaCO31,020 1,080 935 1,450 1,150 710 670 410 
Orthophosphate (mg·L−1 as PO43−6.3 8.5 1.1 1.9 39 30 29 20 
Ammonia-N (mg·L−18.6 4.1 2.3 4.9 259 0.05 0.05 12 
Nitrate-N (mg·L−156 23 28 52 73 74 26 
COD (mg·L−1212 154 104 209 385 112 164 218 
Color (PtCo U) 246 96 43 88 213 24 26 79 
Reactive silica removal rate (mg·L−1·day−139 28 33 56 −3.6 31 34 19 

CA, California; AZ, Arizona; MF, microfiltration; BAC, biological activated carbon; UF, ultrafiltration; TDS, total dissolved solids; COD, chemical oxygen demand.

A total of 11 RO concentrate samples were obtained from six different AWPFs in California and Arizona as shown in Tables 1 and 2. The reactive silica concentrations in these samples ranged from 56 (Sample #8) to 197 mg·L−1 (Sample #4), while total dissolved solids (TDS) concentrations ranged from 3,650 (Sample #3) to 7,210 mg·L−1 (Sample #11). The feed water to these RO systems varied from non-nitrified and partially nitrified to fully nitrified effluent. Some of the AWPFs (Facilities C and D) employed ozone or ozone followed by biological activated carbon (BAC) filtration as a pre-treatment of microfiltration (MF)/ultrafiltration (UF). The RO concentrate samples were tested for water quality parameters (Tables 1 and 2) and stored in a refrigerator at 4 °C until use. No chloramine residual was present when the samples were used in the photobiological treatment experiments.

Table 2

The second three RO concentrate samples tested

Sample ID 10 11 
Facility ID 
Location CA CA CA 
Nitrification Partial None None 
Pre-treatment MF MF* MF 
RO configuration 3 Stage 3 Stage 3 Stage 
Permeate recovery (%) 85 85 82 
TDS (mg/L) 5,060 4,170 7,210 
pH 8.2 7.1 7.6 
Reactive silica (mg·L−1115 107 124 
Calcium (mg·L−1560 600 872 
Alkalinity (mg·L−1 as CaCO3900 850 1,760 
Orthophosphate (mg·L−1 as PO43−54 10 
Ammonia-N (mg·L−16.9 260 121 
Nitrate-N (mg·L−154 8.3 3.7 
COD (mg·L−1252 358 320 
Color (PtCo U) 230 289 178 
Sample ID 10 11 
Facility ID 
Location CA CA CA 
Nitrification Partial None None 
Pre-treatment MF MF* MF 
RO configuration 3 Stage 3 Stage 3 Stage 
Permeate recovery (%) 85 85 82 
TDS (mg/L) 5,060 4,170 7,210 
pH 8.2 7.1 7.6 
Reactive silica (mg·L−1115 107 124 
Calcium (mg·L−1560 600 872 
Alkalinity (mg·L−1 as CaCO3900 850 1,760 
Orthophosphate (mg·L−1 as PO43−54 10 
Ammonia-N (mg·L−16.9 260 121 
Nitrate-N (mg·L−154 8.3 3.7 
COD (mg·L−1252 358 320 
Color (PtCo U) 230 289 178 

*The ozone injection was off when this sample was collected.

See Table 1 for abbreviations. Reactive silica removal rates are not shown in this table because of the insufficient silica removal in Samples 10 and 11.

Reactive silica, ammonia-N, and nitrate-N concentrations were determined by the silicomolybdate method (Hach method 8185), salicylate method (Hach method 10023/10031), and diazotization or dimethylphenol method (Hach method 10019/10206), respectively, using a Hach DR-2800 or DR-2700 spectrophotometer (Loveland, CO, USA). Calcium hardness and total alkalinity were determined by the EDTA titration (Hach method 8204) and sulfuric acid titration (Hach method 8203) methods, respectively. Other water quality parameters were determined as described previously (Ikehata et al. 2018a, 2018b).

A series of batch treatment experiments were carried out using clear 50 mL polypropylene centrifuge tubes with screw caps (φ = 29 mm; VWR International, USA) as photobiological reactors in an illuminating reflective incubator with 9 W LED bulbs (light temperature 5,000 K, 800 lm each; Cree, Inc., Durham, NC, USA).

The photobiological treatment was initiated by adding pre-cultured diatom suspension (∼500 μL) to 0.2 μm filter-sterilized RO concentrate samples in the 50 mL clear tubes. The initial biomass concentration in each tube was about 0.15 g dry weight·L−1. The inoculated tubes were incubated statically in the illuminating incubator with continuous illumination at 26 ± 2 °C. The photosynthetically active radiation (PAR) was measured as 1.5 μE·m−2·s−1 using an International Light Technologies ILT1400 portable radiometer with an attenuated PAR sensor (Peabody, MA, USA). Aliquots of supernatant samples were withdrawn periodically from the tubes to measure reactive silica concentration during the experiment. At the end of the experiment, other water quality parameters were determined. In the experiments to investigate the impacts of ammonia-N concentration and pH, ammonium sulfate (ACS grade; Sigma-Aldrich, St Louis, MO, USA) and sulfuric acid (ACS grade; Sigma-Aldrich) were added to Sample #9 to achieve the desired initial ammonia-N concentration and pH, respectively. All the experiments were performed in duplicate.

RESULTS AND DISCUSSION

Figure 1 shows the result of the first photobiological treatment experiment comparing the treatability of eight different RO concentrate samples. Reactive silica removal occurred in all the samples except for Sample #5 from Facility C. The reactive silica removal rate in the first 60 hours was calculated and is presented in Table 1. The highest and lowest removal rates were observed in Sample #4 from Facility B and Sample #8 from Facility E, respectively, while the removal rates in the other samples were comparable and were around 30 to 40 mg·L−1·day−1. This result is comparable to our previous results using RO concentrate samples from Facilities A and B (Ikehata et al. 2018a, 2018b).

Figure 1

Reactive silica removal from eight different RO concentrate samples by the photobiological treatment.

Figure 1

Reactive silica removal from eight different RO concentrate samples by the photobiological treatment.

The marked difference of Sample #5 from the other samples tested in the first experiment is the very high concentration (∼260 mg·L−1) of ammonia-N because of the absence of nitrification in the biological treatment prior to the AWPFs (Table 1). The inhibition and toxicity of ammonia-N, in particular free ammonia, in algal growth has been known (Provasoli 1958; Collos & Harrison 2014), although the low concentrations of ammonia-N in the other RO concentrate samples tested in this study were not particularly inhibitory. In fact, a low level (<10 mg·L−1) of ammonia-N was found to be the preferred nitrogen source in our previous study (Ikehata et al. 2018a). The analyses of ammonia-N and nitrate-N before and after the photobiological treatment also support this (Figure 2). The high concentrations of nitrate-N in the fully nitrified RO concentrate samples (Samples #6 and #7) were not inhibitory to the reactive silica removal process.

Figure 2

Concentration of (a) ammonia-N and (b) nitrate-N before and after the photobiological treatment.

Figure 2

Concentration of (a) ammonia-N and (b) nitrate-N before and after the photobiological treatment.

Additional RO concentrate samples were obtained from Facilities A, C and F (Table 2) and were tested along with Sample #5 from Facility C. Like Facility C, Facility F also used non-nitrified effluent at the AWPF, and the sample collected from this AWPF contained a high concentration of ammonia-N. As expected, no reactive silica removal occurred in Samples #10 and #11, or in Sample #5 as shown in Figure S1 in the Supplementary Materials (available with the online version of this paper). The increased reactive silica concentration in Sample #11 was probably due to the dissolution of silica from the damaged diatom cells (Bidle & Azam 1999).

To confirm the impact of ammonia-N on the photobiological treatment, various levels of ammonia-N were added to Sample #9 from Facility A as shown in Figure 3. Ammonia-N clearly exhibited an inhibitory effect on the silica removal at a concentration as low as 16 mg·L−1, which is consistent with the literature (Azov & Goldman 1982; Collos & Harrison 2014). The result also showed that the reactive silica removal was complexly inhibited at an ammonia-N concentration as low as 42 mg·L−1. Microscopic analysis revealed a dose-response of ammonia-N on the diatom cells (Figure 4). While the diatom cells formed long and healthy filaments at a low ammonia-N concentration (6.7 mg·L−1), the filaments became fragmented and aggregated at 42 mg·L−1. The chloroplasts of some of the cells in the latter sample became darker and shrank. No live cells were observed at higher concentrations of ammonia-N (>92 mg·L−1). This most likely explains the behaviour of Samples #5, #10, and #11 in Figure 1 and Figure S1 discussed above. The relatively slow reactive silica removal in Sample #8 may be explained by the slightly high ammonia-N concentration (12 mg·L−1) there as well.

Figure 3

Impact of ammonia-N on the removal of reactive silica from RO concentrate Sample #9.

Figure 3

Impact of ammonia-N on the removal of reactive silica from RO concentrate Sample #9.

Figure 4

Photomicrographs of diatom cells exposed to different levels of ammonia-N (left: no ammonia-N addition (6.7 mg·L−1), right: 42 mg·L−1, both in RO concentrate Sample #9).

Figure 4

Photomicrographs of diatom cells exposed to different levels of ammonia-N (left: no ammonia-N addition (6.7 mg·L−1), right: 42 mg·L−1, both in RO concentrate Sample #9).

The impact of pH on the photobiological process was investigated by adding sulfuric acid to Sample #9. Although a similar attempt was made to adjust the pH to a higher value (∼9) using sodium hydroxide, precipitation of phosphate and calcium occurred immediately, and the pH value was reduced back to around 8, which prevented the photobiological treatment experiment from being conducted.

As shown in Figure 5(a), the removal rate of reactive silica slightly reduced at lower pH. Although the pH of RO concentrate was raised in both samples by the photosynthesis of diatoms, the change was more drastic in the acidified samples probably due to the lower buffering capacity (alkalinity), which is shown in Figure 5(b). The lower alkalinity also affected the calcium removal by the photobiological treatment. Since calcium carbonate is also a major scalant in the RO process in AWPFs (Greenlee et al. 2009; Malaeb & Ayoub 2011), the lower calcium and alkalinity removal efficiency at lower pH is not desirable.

Figure 5

Impact of pH on (a) reactive silica removal and pH, and (b) calcium hardness and alkalinity.

Figure 5

Impact of pH on (a) reactive silica removal and pH, and (b) calcium hardness and alkalinity.

As noted earlier, some of the AWPTs (Facilities C and D) employed ozonation as a pre-treatment of MF/UF-RO. No marked difference was observed between Samples #5 (with ozonation) and #10 (without ozonation) from Facility C (Figure S1). Also, the ozone–BAC treatment at Facility D did not seem to impact the reactive silica removal as compared with the RO concentrate samples from non-ozonated AWPFs such as Facilities A and B. It should be noted that all the AWPFs used antiscalants/threshold inhibitors to prevent the scaling in their RO units. No discernible pattern was noticed in the photobiological treatment process because of the presence of such chemicals in the RO concentrate samples tested in this study.

CONCLUSIONS

This study examined a wide variety of RO concentrate samples from six different full-scale AWPFs in the southwestern United States for their treatability by the newly developed photobiological treatment process to remove reactive silica and calcium. Ammonia-N was found to be inhibitory to the photobiological process at a concentration as low as 16 mg·L−1, although the levels of ammonia-N commonly found in the RO concentrate samples from the AWPFs using fully or partially nitrified effluent as a source water were found to be tolerable. However, the RO concentrate samples from the AWPFs using non-nitrified effluent were unsuitable for the treatment because of the ammonia toxicity. Although the addition of acid to the RO concentrate sample did not affect reactive silica removal significantly, it lowered alkalinity and in turn lowered the calcium removal efficiency. These results showed the importance of RO concentrate water quality in the photobiological process, as well as the wider applicability of this photobiological treatment to recover additional useful water and to reduce brine flow at the AWPFs when combined with a secondary RO process. Further research will be required to optimize the growth of diatoms and reactive silica uptake under different treatment conditions and to demonstrate the feasibility of continuous operation of the photobiological treatment process at a larger scale.

ACKNOWLEDGEMENTS

The authors would like to thank Dr Kenneth P. Ishida, Ms Jana Safarik, Dr Megan H. Plumlee (Orange County Water District, Fountain Valley, CA), Mr Greg Oelker (Suez/Edward C. Little Water Reclamation Facility, El Segundo, CA), Ms Uzi Daniel (West Basin Municipal Water District, El Segundo, CA), Dr Paul Fu, Dr Cathy Chang, Mr Howard Salamanca (Water Replenishment District of Southern California, Long Beach, CA), Mr Joseph Quicho, Mr John Carroll, Mr William Mercado (City of San Diego, San Diego, CA), Mr G. Allen Davidson (Scottsdale Water Campus, Scottsdale, AZ), and Ms Olivia Cancino (City of Oxnard, Oxnard, CA) for providing the RO concentrate samples, as well as valuable information and suggestions. The materials presented in this article are based on the work supported by the National Science Foundation under the Small Business Innovation Research Program (Award 1648495, KI). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

REFERENCES

REFERENCES
Asano
T.
,
Burton
F. L.
,
Leverenz
H. L.
,
Tsuchihashi
R.
&
Tchobanoglous
G.
2007
Water Reuse: Issues, Technologies, and Applications
.
McGraw-Hill, Inc.
,
New York, USA
.
Azov
Y.
&
Goldman
J. C.
1982
Free ammonia inhibition of algal photosynthesis in intensive cultures
.
Applied and Environmental Microbiology
43
(
4
),
735
739
.
Collos
Y.
&
Harrison
P. J.
2014
Acclimation and toxicity of high ammonium concentrations to unicellular algae
.
Marine Pollution Bulletin
80
(
1–2
),
8
23
.
Greenlee
L. F.
,
Lawler
D. F.
,
Freeman
B. D.
,
Marrot
B.
&
Moulin
P.
2009
Reverse osmosis desalination: water sources, technology, and today's challenges
.
Water Research
43
(
9
),
2317
2348
.
Ikehata
K.
,
Zhao
Y.
,
Maleky
N.
,
Komor
A. T.
&
Anderson
M. A.
2017
Aqueous silica removal from agricultural drainage water and reverse osmosis concentrate by brackish water diatoms in semi-batch photobioreactors
.
Journal of Applied Phycology
29
(
1
),
223
233
.
Ikehata
K.
,
Zhao
Y.
,
Kulkarni
H. V.
,
Li
Y.
,
Snyder
S. A.
,
Ishida
K. P.
&
Anderson
M. A.
2018a
Water recovery from advanced water purification facility reverse osmosis concentrate by photobiological treatment followed by secondary reverse osmosis
.
Environmental Science & Technology
52
(
15
),
8588
8595
.
Ikehata
K.
,
Zhao
Y.
,
Ma
J.
,
Komor
A. T.
,
Maleky
N.
&
Anderson
M. A.
2018b
A novel photobiological process for reverse osmosis concentrate treatment using brackish water diatoms
.
Water Science and Technology: Water Supply
18
(
2
),
594
602
.
Leverenz
H. L.
,
Tchobanoglous
G.
&
Asano
T.
2011
Direct potable reuse: a future imperative
.
Journal of Water Reuse and Desalination
1
(
1
),
2
10
.
Pérez-González
A.
,
Urtiaga
A. M.
,
Ibáñez
R.
&
Ortiz
I.
2012
State of the art and review on the treatment technologies of water reverse osmosis concentrates
.
Water Research
46
(
2
),
267
283
.
Provasoli
L.
1958
Nutrition and ecology of protozoa and algae
.
Annual Review of Microbiology
12
(
1
),
279
308
.
Rodriguez
C.
,
Van Buynder
P.
,
Lugg
R.
,
Blair
P.
,
Devine
B.
,
Cook
A.
&
Weinstein
P.
2009
Indirect potable reuse: a sustainable water supply alternative
.
International Journal of Environmental Research and Public Health
6
(
3
),
1174
1209
.

Supplementary data