Abstract

A new biological phosphorous and nitrogen removal process is developed. The process is based on biofilm on carrier elements with enhanced biological phosphorous removal and simultaneous nitrification and denitrification in a continuous process. Results from 3 years of pilot and laboratory experiments are presented with regards to removal of organic substances, phosphorous and nitrogen. This process demonstrates essential benefits and improved performance compared to other EBPR-processes in operation today. The first full scale plant was put in operation in May 2016 at Hias WWTP in Norway.

INTRODUCTION

Hias WWTP is located near the city of Hamar in Norway. The existing WWTP consist of primary sedimentation, traditional activated sludge biological treatment and chemical precipitation with aluminum chloride. The organic and phosphorous loads to the WWTP are high and increased biological capacity was needed. It was decided to look for a process that both could handle the wastewater load, make it possible to reduce the use of metal salts and increase the potential for phosphorous recovery. The WWTP has a 0.4 mg/l discharge limit for total phosphorous, 75% removal of COD, but no nitrogen removal requirements due to discharge to a lake.

Wastewater treatment processes using biofilm on carriers (e.g. MBBR) has experienced a growing interest for the last decades. Compared to activated sludge processes there are several important benefits (Ødegaard 1999), such as:

  • Less footprint and volume required.

  • Process operation and treated wastewater quality is less dependent on sludge separation, due to no sludge return and a low biomass concentration to be removed.

To overcome the resistance to diffusion of O2 through the biofilm, MBBR plants operate at a relatively high DO concentration in the aerated reactors. This will influence the operational cost. The difference in investment cost for an MBBR and an activated sludge plant is related to the cost (and lifetime) of the carriers and the smaller footprint for the MBBR.

EBPR-processes are used to biologically remove more phosphorous than obtained in an ordinary activated sludge process to reduce the need for chemical phosphorous removal. However, most EBPR (and nitrogen removal)-plants use metal salts to comply with outlet phosphorous regulations, or to improve sludge separation properties. The use of chemical phosphorous removal will reduce the potential for phosphorous recovery as a fertilizer. There will be an increasing demand for wastewater treatment processes with a high potential for phosphorous recovery in the future (Jeyanayagam et al. 2016).

EBPR is dependent on exposing the biomass for sequential anaerobic and aerobic conditions. This can be done in a biofilm system (e.g. MBBR) operated as an SBR (sequencing batch reactor) (Helness 2007). This is however not a widely used technology, due to the lower efficiency of a sequencing mode of operation. However, combining the advantages of a biofilm system with enhanced biological phosphorous removal in a continuous process is another matter. The described process utilizes the high capacity of a biofilm system with EBPR performance and nitrogen removal capacity through SND (simultaneous nitrification and denitrification).

In the EBPR process PAO (phosphorous accumulating organisms) take up easily degradable organic matter (VFA, volatile fatty acids) while releasing PO4 under anaerobic conditions. Under aerobic conditions the internal storage of organic matter is used for cell growth and uptake and storage of PO4. Phosphorous removal is obtained because PAO is capable of taking up more PO4 than what is released, storing more P than other bacteria. To exploit this ability of PAO, they must be exposed to alternating anaerobic and aerobic conditions. Some PAO organisms are capable of using nitrate instead of oxygen as electron acceptor under anoxic conditions (DNPAO) (Kern Jespersen & Henze 1993). This is advantages whit regards to the use of organic matter because the anaerobically stored carbon can be used for phosphorous uptake and simultaneous denitrification.

The SND principle in biofilm systems is dependent on local anoxic zones in the biofilm where oxygen do not reach for denitrification in an aerobic reactor. Nitrification is on the other hand highly dependent on dissolved oxygen concentration and the level of organic matter concentration.

Therefore, the combination of EBPR and SND has some significant advantages:

  • The use of organic material in the wastewater is sensible

    • o Taken up by PAO and DNPAO under anaerobic conditions in the first step of the process (Zeng et al. 2003)

    • o Less carbon is needed when DNPAO use the same carbon for both P-uptake and denitrification (Kern Jespersen & Henze 1993) and (Zeng et al. 2003)

    • o DNPAO takes up carbon anaerobically where the access to carbon is good (Zeng et al. 2003)

    • o Nitrification benefits from a low level of organic material entering the aerobic zone (Ødegaard 1999)

    • o If the supply of carbon from the wastewater for P-uptake or denitrification is insufficient, carbon can be added in the anaerobic zone (Zeng et al. 2003)

  • The volume of the plant can be reduced

    • o Both nitrification and denitrification takes place in the aerobic zone. (Münch et al. 1996) and (Zeng et al. 2003)

    • o This research show that the PO4-concentrations in the effluent can be very low, and chemical phosphorous removal is not needed.

In addition to that comes the benefits of a biofilm process compared to an activated sludge process as mentioned earlier. In activated sludge EBPR plants the return sludge pumping is absolutely necessary to have a functioning process, but it also introduce several adverse consequences like introducing oxygen and nitrate to the anaerobic zone, and reducing the retention time in the reactor due to volume occupation by returned water. In the presented process the biomass attached to the biofilm carriers is lifted out of the water, and returned to the anaerobic zone. In that way the drawbacks of return sludge pumping is not existing.

The results presented in this paper are based on 3 years of pilot plant operation, start-up of a new pilot plant in February 2016 to further investigate and optimize the nitrogen removal capacity, start-up of a full scale plant in May 2016 and many lab scale experiments.

MATERIAL AND METHODS

Pilot plants

The pilot scale treatment plants have a total volume of 7 m3, divided into several anaerobic and aerobic zones. The inlet wastewater entering the first anaerobic zone of the pilot plant is taken from Hias WWTP after primary sedimentation. Standard biofilm carriers (Kaldnes K1 or K3) with 500 m2/m3 are used, with a filling degree of 60% related to volume. Wastewater and carriers flow together through openings in the separation walls between zones, through the mechanically mixed anaerobic zones to the aerated zones. From the last aerobic zone the carriers with biofilm is mechanically transported by a conveyer (without wastewater) back to the first anaerobic zone. The treated wastewater is directed out of the pilot plant from the last aerobic zone. The layout of the pilot plant is presented in Figure 1. The Anaerobic volume has been between 30 and 40% of the total volume of the plant, and the total residence time for the wastewater is between 5 and 10 hours. The inlet flow is normally varying proportional to the flow in the WWPT, or in periods kept constant. DO in the aerated zones is normally kept between 4 and 8 mg/l, which is common in MBBR plants, but will vary with the purpose of the process (BOD-removal, P-removal, nitrification, denitrification) (Ødegaard 1999). Daily grab samples were taken from the inlet, anaerobic and aerobic zones (outlet) in the pilot plant.

Figure 1

Plant layout for pilot 1.

Figure 1

Plant layout for pilot 1.

Pilot 2 was developed further with optimized flow passages between zones and the aerobic volume was divided into several volumes, within the same total volume of the pilot plant.

Laboratory experiments

P-release and uptake tests and nitrification/denitrification tests have been carried out in laboratory scale. Carriers with biofilm taken from the pilot plant were filled in 1 l beakers and inlet wastewater was added. A jar-test apparatus was used for stirring during the anaerobic period, and small scale diffusors was put in the beakers for aeration during the aerobic period to keep the DO in the area of 6–8 mg/l. Samples for analysis was withdrawn from the beakers during the experimental period.

Analysis

All samples were filtered through 1 μm fiberglass filter and analyzed for PO4-P (dissolved phosphorous), SCOD (dissolved chemical oxygen demand), NH4-N (ammonia-N), NO3-N (nitrate-N) and NO2-N (nitrite-N) with a NOVA Spectroquant 60 spectrophotometer. Some samples were analyzed for VFA by the 5-point titration method described by Moosbrugger et al. (1993).

RESULTS AND DISCUSSION

Inlet wastewater

The inlet wastewater to Hias WWTP is influenced by effluent from food processing industry in the area. In Table 1 the average and min/max concentrations from 3 years of daily grab samples are presented. The daily variations on the WWTP over one week are high, with seasonal variations on top of that. This can be seen from the minimum and maximum values.

Table 1

Average and min/max values for wastewater quality after primary sedimentation

 AvgMinMax
PO4-P (mg/l) 3.64 0.35 9.02 
NH4-N (mg/l) 60.8 13.0 140.0 
SCOD (mg/l) 312 17 835 
 AvgMinMax
PO4-P (mg/l) 3.64 0.35 9.02 
NH4-N (mg/l) 60.8 13.0 140.0 
SCOD (mg/l) 312 17 835 

In Figure 2 the typical concentration variations for PO4-P, SCOD and VFA in the primary sedimentation effluent at Hias WWTP is presented for 3 subsequent weeks. The concentrations of PO4-P and SCOD varies 3–5 times over the week, while VFA is usually not found on Sundays and Mondays, but reach quit high values on Wednesdays and Thursdays. These variations in concentration (and load) can be difficult for biological processes to handle, and is at least on of the reasons for difficulties with filamentous sludge in the existing activated sludge plant and in previous pilot trials with EBPR in activated sludge carried out at Hias WWTP.

Figure 2

Typical variations in PO4-P, SCOD and VFA concentrations for 3 subsequent weeks.

Figure 2

Typical variations in PO4-P, SCOD and VFA concentrations for 3 subsequent weeks.

The temperature of the inlet wastewater varies between 6 and 14 °C over the year.

Phosphorous removal

The phosphorous load and removal for several months of operation in Pilot 1 and for 3 weeks of operation in Pilot 2 is shown in Figure 3. The results are presented as load and removal in g PO4-P/m2 biofilm carrier surface area and day. The efficiency of MBBR processes is most correctly based on the effective carrier surface area (Ødegaard et al. 1998). All results presented are from a period after stable EBPR activity was established. The results for Pilot 1 are divided into groups according to the SCOD load (g SCOD/m2*d). The results for Pilot 1 show good phosphorous removal in the load range, but at high organic loads (>5 g/m2*d) the removal of PO4 can be negatively influenced. The results for Pilot 2 show very good PO4 removal just 5 months after startup. The highest SCOD load for Pilot 2 was 5.2 g/m2*d, and the capacity is still increasing.

Figure 3

Phosphorous load and removal as g PO4-P/m2 carrier area and day, for several months of operation in Pilot 1 and 3 weeks in Pilot 2. Results for Pilot 1 is separated according to the SCOD load.

Figure 3

Phosphorous load and removal as g PO4-P/m2 carrier area and day, for several months of operation in Pilot 1 and 3 weeks in Pilot 2. Results for Pilot 1 is separated according to the SCOD load.

High organic loads might lead to carry over of easily degradable carbon from the anaerobic zone to the aerobic zone. If VFA is still present after the anaerobic period when PAO enters the aerobic zones, PO4 release instead of uptake may take place (Janssen et al. 2002). If degradable carbon is available in the aerobic zone, this will also increase the competition from heterotrophic bacteria. In Figure 4 a laboratory scale phosphorous uptake test is shown. Biofilm carriers with biofilm from the pilot plant was used to simulate aerobic PO4 uptake under the influence of VFA. The concentrations of PO4-P clearly show that the uptake rate is reduced while there still is about 10 mg/l of PO4-P left, probably because there is VFA left in the water. Following this the uptake rate increase again before it is limited by the PO4-P concentration. At the time of the second VFA measurement showing 1.4 mg/l VFA, the uptake rate of PO4 increase. When there is VFA remaining under aerobic/anoxic conditions there might be both uptake and release of PO4 at the same time (Janssen et al. 2002) resulting in a net uptake rate as seen in Figure 4.

Figure 4

PO4 and VFA concentrations during aerobic uptake test.

Figure 4

PO4 and VFA concentrations during aerobic uptake test.

In Figure 5 another release and uptake test is shown. The uptake rate of PO4 does not seem to be reduced by VFA in the aerobic period, as in Figure 4, and is linear for most of the concentration range. The uptake of carbon measured as SCOD concentration is also shown. The SCOD reduction is rapid in the beginning of the anaerobic period when PAO take up easily degradable carbon like VFA, but slows down when availability of VFA is reduced. The release rate of PO4 that should be closely linked to the uptake of carbon, is also reduced at the end of the anaerobic period. Grab samples from different zones in Pilot 2 was collected at the time the test was started, and PO4 concentrations are shown in Figure 5 according to the retention time in the plant. The PO4 concentrations from the pilot plant are in good agreement with the laboratory test. In Table 2 the maximum specific release and uptake of PO4, the anaerobic uptake of SCOD per P released and the anaerobic SCOD uptake per net P uptake is shown. Some of the SCOD that is removed during the anaerobic period can be used by other bacteria than PAO/DNPAO, however, this ‘loss’ of degradable carbon in the anaerobic zone to biological conversion of nitrate and oxygen that is typical for activated sludge EBPR plants is neglectable for this process because no water is returned to the anaerobic stage, only the carriers with biofilm.

Table 2

Specific release and uptake rates for PO4 and SCOD from the laboratory test shown in Figure 5 

Anaerobic
AerobicNet
PO4 release rateSCOD uptake/P releasePO4 uptake rateSCOD uptake/P uptake
g/m2*dmg/mgg/m2*dmg/mg
1.86 10.7 1.53 50.1 
Anaerobic
AerobicNet
PO4 release rateSCOD uptake/P releasePO4 uptake rateSCOD uptake/P uptake
g/m2*dmg/mgg/m2*dmg/mg
1.86 10.7 1.53 50.1 
Figure 5

PO4 and SCOD concentrations during lab scale release and uptake test.

Figure 5

PO4 and SCOD concentrations during lab scale release and uptake test.

One of the main objectives for Hias when developing this process has been to comply with the effluent total phosphorous limit of 0.4 mg/l, without chemical phosphorous removal. This implies that the PO4-concentration in the effluent from the biological treatment step should be below 0.2 mg/l, when pilot testing have shown that an efficient particle removal can ensure a particulate phosphorous content below 0.2 mg/l.

Removal of organic matter

A steady EBPR process is dependent on PAO taking up easily biodegradable organic matter in the anaerobic zone, to be able to use this storage of carbon for P uptake in the aerobic zone. Also, if VFA leaks to the aerobic zone, one might experience a reduction in P uptake as discussed above. An effective anaerobic carbon removal is therefore crucial for the efficiency of PAO. Figure 6 shows total (left) and anaerobic (right) SCOD load and removal for some months of the experimental period for Pilot 1 and for 3 weeks of operation for Pilot 2. The results indicate that the SCOD capacity is higher than 6–7 g/m2*d. Pilot 2 results are collected after a 3.5 month's startup period, when stable EBPR was established, and show slightly better total SCOD removal than Pilot 1 that had been in operation for more than 1 year. The anaerobic SCOD removal was slightly better for Pilot 1. This difference is probably a result of the short time of operation for Pilot 2.

Figure 6

Left: Total load and removal of SCOD, right: anaerobic load and removal of SCOD, all in g SCOD/m2*d.

Figure 6

Left: Total load and removal of SCOD, right: anaerobic load and removal of SCOD, all in g SCOD/m2*d.

Total SCOD removal is about 80%, but as can be seen the removal is influenced by an inert fraction of SCOD, that is present also at low loads.

About 60% of the SCOD was removed in the anaerobic zone, with an anaerobic SCOD load of up to 18 g/m2*d.

Removal of nitrogen

Hias WWTP's discharge permit does not regulate nitrogen. However, during pilot and laboratory testing of the new process, a significant nitrogen removal potential was discovered. Both nitrification and denitrification took place in the aerobic zone, which indicate SND in different layers of the biofilm (Münch et al. 1996). There are several benefits of implementing SND in the new process:

  • Most/all BOD is removed before the aerobic stage – benefits nitrification (Ødegaard 1999)

  • Anaerobic BOD uptake by DNPAO will secure an efficient use of wastewater carbon by anoxic uptake of PO4 and denitrification of NO3 using the same carbon in anoxic layers of the biofilm in the aerated zones of the reactor (SND).

  • Wastewater with NO3 from the aerobic zone is not returned to the anaerobic zone (inefficient use of carbon), which is the case for activated sludge EBPR plants.

Figure 7 show the nitrification rate/NH4-N removal as a function of ammonia load (left), and the denitrification rate/N removal as a function of NH4-N removal, all in g N/m2*d. The average nitrification load, nitrification rate and denitrification rate for these 3 weeks of data in Pilot 2 is respectively: 0.86; 0.54 and 0.33 g N/m2*d. The nitrification and denitrification biology in Pilot 2 is still developing, and both nitrification and denitrification is increasing. Pilot 2 has been in operation for totally 5 months, and the temperature of the wastewater has been between 9 and 16 °C. DO levels in the aerobic zones have been kept at 4–7 mg/l. Further optimization of the SND will mainly study effects of DO and aerobic organic load. DO in the aerobic zones will of course influence the nitrification, but also the anoxic denitrification taking place in deeper layers of the biofilm.

Figure 7

Left: Nitrification rate and right: denitrification rate as a function of load in g N/m2*d for Pilot 2.

Figure 7

Left: Nitrification rate and right: denitrification rate as a function of load in g N/m2*d for Pilot 2.

A laboratory scale experiment was carried out with biofilm carriers from Pilot 2 to compare the aerobic and anoxic phosphorous uptake and study the nitrification and denitrification rates. The biofilm carriers with biofilm and the wastewater was taken from the last anaerobic zone in Pilot 2, one beaker was aerated and one was added nitrate as Sodium-nitrate and gently stirred in a jar test apparatus. The temperature of the wastewater was 15 °C. From the data seen in Figure 8 (left) the maximum aerobic and anoxic phosphorous uptake is calculated to be 1.92 and 1.10 g PO4-P/m2*d respectively. The ratio between the estimated uptake rates is suggested to give the fraction of DNPAO/PAO (Janssen et al. 2002), in this case 58%. However, from Figure 8 it can be seen that all the PO4 is removed under anoxic conditions, as well as under aerobic conditions, both having a net PO4-P uptake of 6.1 mg.

Figure 8

Left: Aerobic and anoxic phosphorous uptake and SCOD removal, right: NH4-N, NOx-N and NO3-N concentrations for the same laboratory experiment.

Figure 8

Left: Aerobic and anoxic phosphorous uptake and SCOD removal, right: NH4-N, NOx-N and NO3-N concentrations for the same laboratory experiment.

The nitrification and denitrification rate under aerobic conditions is calculated from the results shown in Figure 8 (right), and is presented in Table 3.

Table 3

Nitrification and denitrification rate for the aerobic experiment, and the specific uptake of SCOD and PO4-P per NO3-N consumed for the anoxic experiment

Aerobic
Anoxic
Nitrification rateDenitrification rateSCOD/NO3-NPO4-P uptake/NO3-N
g/m2*dg/m2*dmg/mgmg/mg
0.81 0.61 1.2 0.53 
Aerobic
Anoxic
Nitrification rateDenitrification rateSCOD/NO3-NPO4-P uptake/NO3-N
g/m2*dg/m2*dmg/mgmg/mg
0.81 0.61 1.2 0.53 

From Figure 8 (right) it can be seen that the removal rate of NO3-N under anoxic conditions is more or less constant, while the removal of SCOD in the same period is rapid in the beginning and slows down considerably after about 1 hour. The initial rapid SCOD consumption is taking place both under aerobic and anoxic conditions. It is not likely that this consumption of SCOD is related to traditional denitrification in anoxic layers of the biofilm when the same consumption is seen under both conditions, and there was no initial nitrate in the aerobic experiment. The theoretical consumption of COD per NO3-N for denitrification is 4 mg/mg (Janssen et al. 2002), the much lower COD consumption for this experiment (shown in Table 3) indicate that denitrification with anaerobically stored carbon is taking place. This is all evidence that DNPAO use the same stored carbon for anoxic PO4 uptake and denitrification.

The SCOD consumption under aerobic conditions is higher than under anoxic conditions, this is probably due to heterotrophic activity.

An additional nitrate dose was added to the anoxic experiment when the initial nitrate was used. Figure 8 (right and left) show that denitrification carries on, probably at a slower rate, after all the PO4 is removed and the SCOD removal has stopped. These results indicate that DNPAO are able to denitrify without taking up phosphorous, and this can be an important attribute to exploit for this process.

CONCLUSIONS

The continuous biofilm EBPR process developed at Hias WWTP has shown very good phosphorous removal results, and incorporating SND in the same process looks promising. The process successfully integrates the strengths of both biofilm and EBPR while enhancing nitrification by anaerobic carbon removal and allowing dual utilization of carbon for both phosphate removal and denitrification, making the process highly compact and efficient. Total SCOD loads of 5.2 g/m2*d and anaerobic SCOD loads of 18 g/m2*d has been experienced in pilot scale with more than 95% phosphorous removal. The stability of the process for huge variations in load looks very promising.

Further work

In May 2016 the first full scale continuous biofilm EBPR plant was started up at Hias WWTP. Further experimental work on implementing/optimizing the nitrogen removal will take place in laboratory, pilot and full scale. Patent is pending on the process.

REFERENCES

REFERENCES
Helness
,
H.
2007
Biological Phosphorus Removal in a Moving Bed Biofilm Reactor. PhD Thesis
,
Norwegian Univ. of Sci. and Tech. Trondheim
,
Norway
.
Janssen
,
P. M. J.
,
Meinema
,
K.
&
Van der Roest
,
H. F.
2002
Biological Phosphorus Removal: Manual for the Design and Operation
.
STOWA IWA Publishing
,
London, UK
.
Jeyanayagam
,
S.
,
Khunjar
,
W.
,
Pramanik
,
A.
&
Mehta
,
C.
2016
Accelerating the implementation of extractive nutrient recovery as an integral component of sustainable nutrient management
.
Water e-Journal
1
(
2
),
1
8
.
Kern Jespersen
,
J. P.
&
Henze
,
M.
1993
Biological phosphorus uptake under anoxic and aerobic conditions
.
Water Research
27
(
4
),
617
624
.
Moosbrugger
,
R. E.
,
Wentzel
,
M. C.
,
Ekama
,
G. A.
&
Marais
,
G. v. R.
1993
A 5 pH point titration method for determining the carbonate and SCFA weak acid/bases in anaerobic systems
.
Water Science and Technology
28
(
2
),
237
245
.
Ødegaard
,
H.
1999
The moving bed biofilm reactor
. In:
Igarashi
,
T.
,
Watanabe
,
Y.
,
Asano
,
T.
&
Tambo
,
N.
, :
Water Environmental Engineering and Reuse of Water
,
Hokkaido Press
, p.
250
305
(in Japanese).
Ødegaard
,
H.
,
Gisvold
,
B.
&
Strickland
,
J.
1998
The influence of carrier size and shape in the moving bed biofilm process
. In:
Extended Abstract for the IAWQ Conference on Biofilm Systems
,
New York
,
October 17–20, 1998
.
Zeng
,
R. J.
,
Lemaire
,
R.
,
Yuan
,
Z.
&
Keller
,
J.
2003
Simultaneous nitrification and denitrification, and phosphorous removal in a lab-scale sequencing batch reactor. Published online 24 June 2003 in Wiley Interscience
.