Sie sind auf Seite 1von 35

Component 5: Pilot Projects

Phase 5.2: QUALITY STRESS

Deliverable: 5.2.4
Effectiveness of an innovative technology for treating raw municipal
wastewater in tourist areas

Responsible partner: IRSA


Deliverable Author/s: Claudio Di Iaconi
Contact for queries: claudio.diiaconi@ba.irsa.cnr.it

Project Details
Programme

MED (2007-2013)

Priority Objective
Axe

2-1
2 Protection of the environment and promotion of a
sustainable territorial development

Objective

2.1 Protection and enhancement of natural resources and


heritage

Project Title

Sustainable management of environmental issues related to


water stress in Mediterranean islands

Project Acronym

MEDIWAT

Project Code No

1G-MED09-262

Lead Partner

Regional Department of Water and Wastes Water


Observatory

Total Budget

1.480.000,00 Euro ()

Time Frame
Start Date End Date

010/06/2010 31/05/2013

Deliverable Details
Component

C.5 Pilot Project

Phase

5.2 Quality stress

Title of Deliverable

D.5.2.4 - Effectiveness of an innovative technology for treating


raw municipal wastewater in tourist areas

Partner Responsible

IRSA

Partners Involved

IRSA

Due Date of Deliverable

December 2012

MEDIWAT - 2G-MED09 - 262

EXECUTIVE SUMMARY.. 4
1. INTRODUCTION ... 5
2. PILOT DESCRIPTION.. 8
3. RESULTS......... 15
4. CONCLUSIONS .... 33
REFERENCES .. 34

MEDIWAT - 2G-MED09 - 262

EXECUTIVE SUMMARY
MEDIWAT project is aimed at identifying and/or developing innovative and integrated
solutions (technical, operational and administrative) for managing environmental
issues related to the quality and quantity stress presently afflicting Med islands.
One of the specific objectives of the project is carrying out pilot projects for
evaluating specific tools for mitigating water quantity and quality stress as well as
managing water demand. In particular, several pilot projects are foreseen in the three
phases of the component 5, i.e.: phase 5.1: pilot projects on quantity stress; phase 5.2:
pilot projects on quality stress; phase 5.3: pilot projects on demand management.
This deliverable reports the results obtained in the pilot project titled testing a single
step innovative process for treating raw municipal wastewater in tourist areas, carried
out in Apulia (Italy) by the partner IRSA, within phase 5.2 of the component 5.
This report is structures as follows. Chapter 1 briefly summarized the background of
the pilot project (i.e., state of art of wastewater treatment in tourist areas) whereas
Chapter 2 describes the methodology used (i.e., the innovative SBBGR technology)
and the operating programme. Chapter 3 shows and discusses the results obtained
during the pilot experimental campaign. The main relevant results (i.e., those
exploitable and feasible in all the Mediterranean islands) are summarized in Chapter 4.

MEDIWAT - 2G-MED09 - 262

1. INTRODUCTION
Tourist areas are usually featured by a strong population variation during the holiday
season which determines an intense seasonal water demand. For example, the
Mediterranean region with about 46,000 km of coastline is the worlds favourite
holiday destination area, attracting more than one third of global tourist arrivals every
year. About 40% of all arrives are concentrated during the summer period; moreover,
the increase in population within this period doesnt follow a gradual trend but surges
on certain dates.
Population fluctuation in tourist areas affects not only water demand but also
wastewater quality and quantity. In fact, in comparison with a typical municipal
sewage the wastewater coming from heavily tourist areas is more concentrated in terms
of typical pollutant parameters, more variable in terms both of flow and contaminant
variation, and less disintegrated in terms of particulate matter because of the shorter
sewerage system (Odegaard, 1989).
A reliable wastewater collection, treatment and disposal system should be implemented
in tourist areas since in these areas, especially in coastal ones, the receiving water is the
main value but also the main concern for its possible pollution.
In tourist areas wastewater treatment is usually carried out decentrally by small plants
each serving single or group of homes, hotels and other tourist establishments (Figure
1). In fact, wastewater collection based on a branched sewer network, which cost may
contribute up to 60% of the total budget for wastewater management in large systems,
should be too expensive for coastal recreational areas (such as resorts, hotel restaurants
and bars) often located along coastline at considerable distance from each other.
Wastewater treatment systems in use in tourist areas, mainly based on extended
aeration activated sludge process, suffer problems of shock loads, sludge bulking,
absence of regular supervision and maintenance, and flow fluctuations which may lead
to a low effluent quality (Christoulas and Andreadakis, 1989). The situation is
particularly critical at the beginning of holiday season when a sharp increase in
wastewater flow takes place as these systems require a certain period for the
development of the biomass necessary for the treatment. Therefore, during this period,

MEDIWAT - 2G-MED09 - 262

which often lasts several weeks, a low quality effluent is produced.

treatment plant
source
Fig.1 - Decentralized wastewater treatment in tourist areas.

Furthermore, the sludge produced represents one of the most concern issues in the
small treatment plants (Kuai et al., 2000). In fact, sludge treatment methods usually
applied in large-scale wastewater treatment plants (i.e., aerobic or anaerobic sludge
stabilization and dewatering) are basically not suitable for small wastewater treatment
plants since they should be too expensive. The sludge produced in these plants is
usually transported to a nearby larger treatment plant (that in costal areas may be a
long distance) where these facilities are available. The reduction of excess sludge
production is then highly desirable in small wastewater treatment plants.
Therefore, more reliable and robust wastewater treatment technologies with lower
sludge production and better operational flexibility are required in tourist areas.
Among the new systems recently proposed that can comply with this request, one of
the most promising is the system developed by the Water Research Institute (IRSA) of
the National Research Council of Italy (CNR) whose acronym is SBBGR (Sequencing
Batch Biofilter Granular Reactor).
SBBGR system combines the advantages of attached biomass systems (i.e., greater
robustness and compactness) with those of periodic systems (i.e., greater flexibility and

MEDIWAT - 2G-MED09 - 262

stability). SBBGR is however a unique system for the particular type of biomass
growing in it. In fact, the biomass present in this system consists of two different
fractions: the biofilm attached to the carrier material and the granules entrapped in the
pores produced by packing the reactor with a filling material. Therefore, the whole
biomass (i.e., biofilm and granules) is completely confined in a dedicated zone of the
reactor (a secondary settler is therefore no longer necessary). This feature allows
greater biomass retention in the reactor to be obtained (up to one magnitude order
higher than that recorded in conventional biological systems). As a result, a notable
increase in sludge age is achieved with consequent reduction in sludge production. In
fact, the microorganisms spend much time in the endogenous metabolism phase where
the biomass decay rate is high, and thus the biomass production rate is low (Di Iaconi
et al., 2005; Di Iaconi et al., 2010a).
SBBGR technology has been already applied for treating different wastewater types
such as municipal primary effluents, tannery wastewater, municipal landfill leachates
and textile wastewater with a low excess sludge production (Di Iaconi et al., 2010a, Di
Iaconi et al., 2010b, Di Iaconi et al., 2011a, Di Iaconi et al., 2011b; Lotito et al., 2011).
In particular, when the system was applied for treating municipal primary effluent a
80% reduction of sludge production was obtained (Di Iaconi et al., 2010a).
The benefits of SBBGR technology can be summarized as follows: a) interesting
conversion capacities; b) the absence of a secondary settling tank; c) a low footprint
due to the possibility of carrying out in a single operative unit all the steps of a
biological treatment; d) great flexibility - a particularly attractive feature for the
treatment of wastewater characterized by great variability in volumetric flow rate and
composition such as sewage coming from tourist areas; e) a very low sludge
production.

MEDIWAT - 2G-MED09 - 262

2. PILOT DESCRIPTION
A pilot plant, based on SBBGR technology, with a treatment capability up to 1 m3/d,
was purposely designed and constructed with the aim of evaluating its effectiveness in
treating municipal wastewater of tourist areas. Figures 2 and 3 show a sketch and
photograph of the plant, respectively.

EV
eff.

aerator

Aerator

biofilter

Biofilter

inf.
rec. pump

fill. pump

re
c

pu
mp

fill. pump
blower

blower

Fig.2 - Pilot plant sketch.

Fig.3 - Pilot plant photograph.

SBBGR pilot plant is basically made of 2 units: biofilter and aerator. The biofilter is
the reactive zone of the plant as it contains the biomass; it consists of a steel cylindrical
reactor (diameter: 220 mm; height: 3200 mm) filled with the wheel shaped plastic
elements shown in Figure 4 (features: 7 mm high, 11 mm diameter, specific area 650
m2/m3, 0.95 g/cm3 density, 0.7 bed porosity and 50-80 mm3 voids dimension). These
elements are packed (fixed) between two holed plates (see Figure 5). The whole

MEDIWAT - 2G-MED09 - 262

biomass of the plant is completely confined in the biofilter (entrapped in the filling
material) and then a secondary settler is no longer required.

Fig.4 - A photograph of
plastic elements.

Fig.5 - Plastic elements packed in the


biofilter.

The aerator, consisting of a steel cylindrical reactor (diameter: 273 mm; height: 3200
mm), is the zone of the liquid phase (i.e., wastewater); its role is to supply air for the
process by means of a blower connected with a diffuser plate located at the bottom of
the unit. The air supply in a dedicated reactor separate from the biomass gives the
advantage of avoiding the installation of complex devices as in the conventional
systems where clogging problems are encountered (due to the biomass that clogs the
holes of the devices used for air supply). The biofilter and aerator are hydraulically
connected by means of a pump (rec. pump in Figures 2 and 3) which continuously
recycles the liquid in the aerator through the biomass supporting material of the
biofilter. Furthermore, this recycle assures a homogeneous distribution of the biomass
along all the height of the biofilter.
The operation of the pilot plant is based on a succession of treatment cycles, each
consisting of three consecutive phases (see Figure 6): the filling, reaction and drawing
phases. During the filling phase, a fixed volume of wastewater to be treated is added
(by means of the filling pump) to the liquid volume retained by the aerator from the
previous treatment cycle. During reaction phase, the liquid in the aerator is

MEDIWAT - 2G-MED09 - 262

10

continuously aerated and recycled through the biomass supporting material of the
biofilter. Finally, the treated wastewater is discharged exploiting the gravity by
opening a motorized valve (EV in Figures 2 and 7) connected to different heights of the
aerator (see Figure 7). The plant is then ready to start a new treatment cycle. The
operative schedule (filling, recirculation, aeration, drawing) is completely automated,
using a programmable logic controller (PLC) equipped with a touch screen monitor
(see Figure 8) for changing the working conditions (i.e., the duration of the working
period) of the components (i.e., filling pump, recirculation pump, blower and drawing
valve).

EV
effluent

Fig.6 - Pilot plant operation.

MEDIWAT - 2G-MED09 - 262

Fig.7 - A photograph of drawing system.

11

Fig.8 - Photos of PLC with touch screen monitor

Pressure probes, set at different heights of the biofilter unit (see Figure 9), measure online head losses due to the biomass growth and captured suspended solids present in
the wastewater. When a fixed set value of head loss is reached at the lowest probe, a
washing step is carried out by compressed air injected into the bottom of the biofilter
unit (see Figure 10) until the headloss is decreased down to a definite value. Washing
water is collected and measured (as TSS and VSS) for calculating the specific sludge
production. Furthermore, nitrogen and phosphorous content of expelled sludge was
also measured. Washing operations are an important operating parameter of the
SBBGR system as they determine the quantity of sludge produced. In fact, sludge can
leave the SBBGR system either with the effluent (i.e., as suspended solids) or as a
result of a washing operation (i.e.,forced exit). Unlike first way, the amount of
sludge expelled by washing operation can be controlled by changing the operative
parameters of the operation (i.e., by increasing/decreasing the set point value for
carrying out the washing). In particular, by increasing the set point value for carrying
out the washing operation it is possible to reduce the frequency of the operation and

MEDIWAT - 2G-MED09 - 262

12

then the sludge production. Nevertheless, high headloss values increase the risk of
clogging of the bed with consequent reduction of treatment capability. Therefore, the
headloss set-point value should be as high as possible consistently with a satisfactory
(requested) removal level.

Fig.9 - Photos of pressure probes


along biofilter height (the probes are
highlighted by red arrows)

Fig.10 - A photograph of compressed


air injection system.

The pilot plant was installed at the experimental site of IRSA located in Bari, a
southern Italy town, and fed with the raw wastewater (i.e., without any pre-treatment)
coming from a residence located on the Adriatic Sea coast near Bari. The wastewater
was transported to the experimental site by means of pump truck.
The experimental activities of the pilot plant were arranged in two main periods
(periods A and B). Period A referred to the generation of the typical biomass of
SBBGR technology (i.e., biomass made of biofilm and granules) by gradual shift of
attached biomass fraction from biofilm to granules (De Sanctis et al., 2010), whereas
period B was addressed to the evaluation of the effectiveness of the pilot plant in
coping with organic load variation.

MEDIWAT - 2G-MED09 - 262

13

In particular, during period A (200 d), taking into account the know how about SBBGR
gained in previous studies (De Sanctis et al., 2010; Di Iaconi et al., 2005; Di Iaconi et
al., 2009), the hydraulic loading to the plant was adjusted in order to have an applied

2.5

OLR
HRT

1.5

0.5

HRT (d)

OLR (kgCOD/m d)

organic loading rate (OLR) in the range of 0.2-0.4 kgCOD/m3d. Once typical biomass
had been achieved, the experimental activity was focussed on the evaluation of the
effectiveness of the pilot plant in coping with organic load variations typical of tourist
areas. In particular, the impact of a wide range of organic loading rates (0.2 to 5
kgCOD/m3d) on plant performance was evaluated, operating at hydraulic residence
times (HRT) of 2.2 to 1.0 d.
HRT and OLR values adopted throughout period B of pilot campaign are shown in
Figure 11.

0
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.11 - HRT and OLR applied during period B

The plant operated with 8 h treatment cycles during period A and 6-8 h treatment
cycles in period B. The filling phase lasted for a few minutes (up to a maximum of 20
min, depending on the volume of influent loaded) whereas the drawing phase lasted 30
min. During reaction phase no air was provided during the first hour for supporting

MEDIWAT - 2G-MED09 - 262

14

denitrification process.
Treatment performance was evaluated by measuring several parameters of influent and
effluent, usually two times per week. The removal efficiencies of all parameters were
calculated as percentage reduction in value between influent and effluent sample.
Chemical oxygen demand (COD), total (TN), nitric (N-NO2) and nitrous (N-NO3)
nitrogen, and total phosphorous (P) were determined by Dr Lange test kits (COD:
LCK314; TN: LCK238; N-NO2: LCK341; N-NO3: LCK339; P: LCK350).
Ammoniacal nitrogen (NH3-N) and total and volatile suspended solids (TSS and VSS)
were determined using standard methods (APHA 2005). Soluble COD (CODsol) was
measured after sample filtration through a 0.45 m filter. Oxidised nitrogen (N-NOx)
was calculated as the sum of nitric (N-NO2) and nitrous (N-NO3) nitrogen. Kjeldahl
nitrogen (TKN) was calculated as difference between total (TN) and oxidised (N-NOx)
nitrogen.
The specific sludge production was calculated assuming that SBBGR operating
conditions, in particular the very high sludge age, guaranteed the complete
metabolization of all particulate organic matter occurring in the wastewater.
Accordingly, the specific sludge production (kgTSS/kgCODremoved) was calculated
dividing TSS leaving the system (i.e., TSS discharged with the effluent + TSS removed
during the washing operations) by the amount of COD removed during the time period
after the first washing operation.

MEDIWAT - 2G-MED09 - 262

15

3. RESULTS
As reported in previous chapter, Period A was addressed to the generation of the
typical biomass of SBBGR technology consisting of granules embedded in biofilm. In
previous studies (Di Iaconi et al., 2005) it was found that the generation process,
depicted in Figure 12, involves four steps: (1) formation of a thin biofilm that fully
covers the carrier surface, (2) increase in biofilm thickness, (3) detachment of a biofilm
portion that covers the carriers with release of biofilm particles, and (4) rearrangement
of biofilm particles in smooth granules.
granules

biofilms

carrier

(1)

(2)

(3)

(4)

Fig.12 - Sketch of the generation process of SBBGR biomass.

Fundamental studies have shown that three factors play a decisive role in the
generation of SBBGR biomass: the trend of the hydrodynamic shear forces, the startup operative conditions (i.e., the organic loading rate applied during the first period),
and bed material features (Di Iaconi et al., 2009). In particular, it was found that during
the first two steps of the generation process depicted in Figure 12, the reactor is
characterised by rather weak shear force values. Under these weak forces, the biofilm,
which is regulated primarily by the organic substrate loading rate, continuously
increases in thickness. This produces a corresponding increase in the shear forces with
negative effects on biomass stability, causing the detachment of biofilm particles (step
3). This, in turn, triggers a further sharp increase in the shear forces, promoting the
rearrangement of the detached biofilm particles into smooth granules by continuous
removal of protuberances (step 4). When biofilm detachment takes place, however, the
filling material characteristics are crucial in granule generation since they must

MEDIWAT - 2G-MED09 - 262

16

guarantee the retention of the detached biofilm particles. In particular, if a filling


material which generates beds characterised by large pore volumes is used, biomass
granulation cannot be achieved because of the expulsion of biofilm particles from the
bed. Therefore, on the basis of the process described above, it is clear that the SBBGR
system differs from conventional biofilter in its final step (i.e., step 4) and that to
obtain the desired granule formation, the biofilm particles detached from the carrier
material have to remain in the bed. In fact, if the biofilm particles were expelled from
the bed, the bed porosity would increase with a consequent reduction in the shear force
value and thus a return to the previous step.
Therefore, as reported in the previous chapter, during period A the hydraulic loading to
the plant was adjusted in order to have and applied organic loading in the range 0.2-0.4
kgCOD/m3d.
The typical biomass of SBBGR was obtained towards the end the seventh month. The
performances of the plant (in terms of average values standard deviation) recorded
during period A are summarized in Table 1.

Tab.1- Plant performance recorded during period A.

Mean value standard deviation

Parameter
TSS [mg/L]

VSS [mg/L]

COD [mg/L]

CODsol [mg/L]

Influent [mg/l]

467 414

Effluent [mg/l]

13 9

Removal [%]

96.2 2.7

Influent [mg/l]

363 318

Effluent [mg/l]

86

Removal [%]

97.1 2.5

Influent [mg/l]

847 652

Effluent [mg/l]

52 33

Removal [%]

91.5 7.0

Influent [mg/l]

202 201

Effluent [mg/l]

38 28

Removal [%]

74.9 17.7

MEDIWAT - 2G-MED09 - 262

17

TKN [mg/L]

NH4+[mgN/L]

TN [mg/L]

P-tot [mgP-PO4/L]

Influent [mg/l]

93.8 37.4

Effluent [mg/l]

9.1 10.6

Removal [%]

91.3 7.3

Influent [mg/l]

72.9 31.3

Effluent [mg/l]

6.1 10.2

Removal [%]

93.1 9.9

Influent [mg/l]

93.9 37.4

Effluent [mg/l]

44.2 18.9

Removal [%]

47.5 24.4

Influent [mg/l]

6.8 4.1

Effluent [mg/l]

5.9 5.3

Removal [%]

25.1 21.8

The data reported in Table 1 show removal efficiencies higher than 90% for COD,
TSS, TKN and NH4+ with residual concentrations in the effluent much lower than the
discharge limits independently of the influent concentration values. TN removal
efficiency was somewhat low as no final anoxic phase was inserted in the treatment
cycle for denitrification process.
After biomass generation, the effectiveness of the pilot plant in coping with organic
load variation was evaluated (period B). Moreover, in order to enhance such a
variation, hydraulic residence time of the plant was stepwise reduced from 2.2 down to
1.0d by increasing the hydraulic loading . Referring to this period, Figures 13 and 14
show the concentration-time profiles of total and soluble COD measured in the influent
and effluent of the plant, respectively, throughout period B whereas Figure 15 reports
COD removal efficiency and OLR applied to the plant during the same period.
Figure 13 shows that COD concentration in the influent of the plant mainly consisted
of particulate organic material (an influent COD/CODsol ratio usually higher than 2
was measured). This was due to the shortness of the sewerage system in tourist areas
which leads to less extension of disintegration process of particulate matter.

MEDIWAT - 2G-MED09 - 262

18

3000

CODinf (mg/L)

2700
2400

tot

2100

sol

1800
1500
1200
900
600
300
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.13 - Total and soluble COD concentrations measured in the influent of the plant during period B.

CODeff (mg/L)

100
90

tot

80

sol

70
60
50
40
30
20
10
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.14 - Total and soluble COD concentrations measured in the effluent of the plant during period B.

MEDIWAT - 2G-MED09 - 262

19

Looking at data in Figures 13, 14 and 15, it is possible to observe that the COD
concentration in the effluent of the plant was always lower than 80 mg/L (most of the
time lower than 50 mg/L) with average removal efficiencies of 91% (however, always
higher than 80%) independently of influent COD concentration, that ranged from 200
up to 2,700 mg/L and organic loading rate applied to the plant that was even as high as
5 kg COD/m3d. It is worthy to remind that mean effluent concentrations between 30

10

100

90

80

70

OLR

60

%COD

50

40

30

20

10

COD removal efficiency (%)

OLR (kgCOD/m d)

and 150 mg/L (in terms of BOD5) are usually reported in the literature for small onsite
extended aeration package systems in use in tourist areas (Christoulas and
Andreadakis, 1989).

0
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.15 - COD removal efficiency and OLR applied to the plant throughout period B.

Contrarily to the conventional sewage treatment plants operating in tourist areas which
are sensitive to the fluctuations in organic loading (Christoulas and Andreadakis, 1989;
Katsiris and Kouzeli-Katsiri, 1989), data in Figures 13, 14 and 15 show an high
robustness of SBBGR system in coping with short-term loading variations. In
particular, it is interesting to observe how there was no deterioration in effluent quality
of the plant during the sudden ten-fold increase of organic loading rate occurred around

MEDIWAT - 2G-MED09 - 262

20

day 400, although the plant operated at hydraulic residence time as low as 1d. In fact, it
should be noted that the extended aeration plants, widely used in tourist areas, are
designed with a much higher hydraulic residence time (Boller, 1997).
Regarding total suspended solids, Figure 16 shows the concentration-time profiles of
total and volatile suspended solids in the influent of the plant during period B, whereas,
in Figure 17 are reported the profiles of effluent suspended solids concentration and
removal efficiency for the same period.
1500

TSSinf; VSSinf (mg/L)

1350
1200

VSS
TSS

1050
900
750
600
450
300
150
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.16 -Total (TSS) and volatile (VSS) suspended solids concentrations measured in the influent of the
plant throughout period B.

The data reported in Figures 16 and 17 highlight excellent filtration performances


although a raw sewage (i.e., without any preliminary treatment) was used. In fact, an
effluent with a suspended solids content always below the discharge limit of 35 mg/L
was recorded (most of the time lower than 20 mg/L) with removal efficiencies always
higher than 80% (most often above 90%) independently of the influent value that was
even as high as 1,200 mg/L. Furthermore, the profiles reported in Figure 16 show that
the total suspended solids consist primarily of organic material.

MEDIWAT - 2G-MED09 - 262

21

80

100

72

90

64

80

56

TSS

70

48

60

40

50

32

40

24

30

16

20

10

TSS removal efficiency (%)

TSSeff (mg/L)

The excellent filtration performance can be appreciate also visually in Figure 18


showing a picture of influent and effluent sample from the plant .

0
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.17 - Total suspended solids concentrations measured in the effluent of the plant and removal
efficiency throughout period B.

Fig.18 - Photograph of the influent (on the right) and effluent (on the left) sample.

The high effectiveness and flexibility of the plant for handling the variation in the

MEDIWAT - 2G-MED09 - 262

22

800

100

720

90

640

80

COD
TSS
%COD
%TSS

560
480

70
60

400

50

320

40

240

30

160

20

80

10

COD and TSS removal eff. (%)

COD and TSS (mg/L)

wastewater flow and composition was also confirmed by profiles recorded within
treatment cycles and reported in Figures 19 and 20.

0
0

350

100

315

90

280

80
COD
TSS
%COD
%TSS

245
210
175

70
60
50

140

40

105

30

70

20

35

10

0
0

1.5

4.5

COD and TSS removal eff. (%)

COD and TSS (mg/L)

Cycle time (h)


Fig.19 - COD and TSS concentration, and removal efficiency profiles in a typical treatment cycle
recorded at day 203 (HRT: 2.2 d; OLR: 1.1 kgCOD/m3d).

Cycle time (h)


Fig.20 - COD and TSS concentration, and removal efficiency profiles in a typical treatment cycle
recorded at day 433 (HRT: 1.2 d; OLR: 0.5 kgCOD/m3d).

MEDIWAT - 2G-MED09 - 262

23

TKNinf, NH3 inf, NOx

+
inf

(mgN/L)

In fact, they show that the removal of COD and TSS is almost complete during the first
hours of the treatment cycle regardless both initial concentrations, applied hydraulic
residence time and organic loading leaving, so, always a somewhat large residual
treatment capability of the plant. It can be ascribed to the particular type of biomass
growing in SBBGR system (consisting of biofilm and granules mixture packed into a
filling material of biofilter compartment) which acts as a filtering media for removing
suspended particulate matter (and then COD associated) from the wastewater. This
high filtering capability of SBBGR biomass is clearly shown in Figures 19 and 20
where it is possible to observe that almost the whole content of suspended solids is
rapidly removed from the wastewater independently of its value. Furthermore, the high
age of SBBGR biomass allows the hydrolysis of the solids captured so as to produce
soluble organic compounds which are removed by the same biomass.
As far as nitrogen is concerned, Figures 21 and 22 show the profiles of TKN, NH3 and
NOx concentrations measured in the influent and effluent of the plant, respectively,
throughout period B.
180
165
150
NH3
135
TKN
120
NOx
105
90
75
60
45
30
15
0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.21 - Total Kjeldahl (TKN), ammoniacal (NH3) and oxidised (NOx) nitrogen concentrations
measured in the influent of the plant throughout period B.

Looking at these profiles it is possible to observe that the effluent concentration of both

MEDIWAT - 2G-MED09 - 262

24

parameters was always (except at day 363 and 400) lower than 15 mg/L (often lower
than 5 mg/L) independently of influent value that was even as high as 150 mg/L,
indicating so the establishment of a stable nitrification process with oxidised nitrogen
concentrations in the effluent usually in the range 10-20 mg/L. Furthermore, contrarily
to what reported for the conventional biological treatment systems, the data in Figure
23 show that removal efficiency of these two parameters was not greatly affected by
organic loading rate applied. In fact, removal efficiencies higher than 80% were
recorded even for temporary organic loading rates as high as 4-5 kgCOD/m3d in spite of
the strong competition existing between autotrophic and heterotrophic bacteria for
dissolved oxygen. Once again, this surprising result can be ascribed to the high
operating flexibility of the system. Nevertheless, as expected, data in Figure 24 clearly
show that nitrification process becomes more unstable as HRT decreases.
Effluent TKN and NH3 (mgN/L)

50
45
40

NH3

35

TKN

30
25
20
15
10
5
0
200

220

240

260

280

300

320

340

360

380

400

420

440

460

Time (d)
Fig.22 - Total Kjeldahl (TKN) and ammoniacal (NH3) nitrogen concentrations measured in the effluent
of the plant throughout period B.

In fact, the lowest TKN and ammonia removal efficiencies (i.e., around 70%) were
recorded at lowest hydraulic residence times applied (i.e., 1.0 d) whereas a HRT of 1.2
d assures removal efficiencies above 80%. Therefore, from the data reported in Figure

MEDIWAT - 2G-MED09 - 262

25

10

100

90

80

70

6
5

OLR

60

TKN

50

NH3

40

30

20

10

TKN and NH3removal effic. (%)

OLR (kgCOD/m d)

23 and 24 it seems that nitrification process is more affected by hydraulic residence


time than organic loading applied to the plant.
TN profiles shown in Figure 25 also highlights the existence of denitrification process
somewhat extended although no planned final anoxic phase was included in the
treatment cycle of the plant.

0
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)
Fig.23 - TKN and ammonia removal efficiency, and OLR applied to the plant throughout period B.

In fact, average removal efficiencies of 72% were achieved with residual effluent
concentration around 20 mg/L (on average). Furthermore, Figure 25 shows that the
values of TN removal efficiencies were somewhat sparse, however, ranging from 25 to
90%, because of the influent COD/N ratio which sometime was low (i.e., lower than 6)
and then did not allow a satisfactory denitrification efficiency (in such an instance,
however, an external carbon source could be used for improving the nitrogen removal).
TKN and oxidized nitrogen profiles within a typical treatment cycle reported in Figure
26 clearly show the existence of a such simultaneous nitrification-denitrification
process. In fact, looking at these profiles it is possible to observe that during the first 4
hours of the cycle time the concentration of oxidised nitrogen is low (i.e., below 5
mg/L) while most of TKN is removed.

MEDIWAT - 2G-MED09 - 262

26

100

2.5

90

80

1.5

70

60

HRT
TKN

0.5

50

NH3

TKN and NH3 removal eff. (%)

HRT (d)

0
40
200 220 240 260 280 300 320 340 360 380 400 420 440 460

Time (d)

Fig.24 - TKN and ammonia removal efficiency, and HRT applied to the plant throughout period B.
250

100
inf
eff
%

TN (mg/L)

200

90
80

175

70

150

60

125

50

100

40

75

30

50

20

25

10

TN removal efficiency (%)

225

0
200

220

240

260

280

300

320

340

360

380

400

420

440

460

Time (d)
Fig.25 - TN influent and effluent concentrations and removal efficiency profiles throughout period B.

MEDIWAT - 2G-MED09 - 262

27

The simultaneous nitrification-denitrification process can be ascribed to the high


biomass concentration usually measured in SBBGR (higher than 40 kgTSS/m3bed) and
to the transient conditions (typical of sequential reactors) which generate contiguous
anoxic and aerobic biomass layers. It has been reported in previous studies performed
by SBBGR technology that the ammonium oxidizers situated in the biomass outer
layers carry out the oxidation of ammonium to nitrite/nitrate while denitrifying bacteria
located in deeper layers, where oxygen cannot penetrate, reduce these compounds to
nitrogen gas by using carbon sources coming from storage or hydrolysis products of
particulate organic matter present in the feed (De Sanctis et al., 2010).
80

TKN, NOx (mgN/L)

70
60

TKN

50

NOx

40
30
20
10
0
0

Cycle time (h)


Fig.26 - TKN and NOx concentration profiles in a typical treatment cycle recorded at day 299 (HRT: 1.4
d; OLR: 0.78 kgCOD/m3d)

Finally, the existence of simultaneous nitrification-denitrification process is also


confirmed by nitrogen balance carried out from day 371 to day 463. In fact, Figure 27
reports the values of nitrogen request for biomass growth, calculated by multiplying
sludge production (i.e., 0.15 kgTSS/kgCODremoved; see below) and measured N biomass

MEDIWAT - 2G-MED09 - 262

28

content (i.e., 0.03 gN/gTSS), and nitrogen removed recorded in the period 371 to 463.
Looking at this figure it leaps out as the nitrogen request for biomass growth (usually
lower than 10 mg/L) is negligible in comparison with measured nitrogen removed.
100
90
N removed
N for growth

80

N (mg/L)

70
60
50
40
30
20
10

46
3

45
5

44
9

44
1

43
8

43
3

42
4

41
7

40
9

40
4

39
6

39
0

38
1

37
7

37
1

Time (d)
Fig.27 - Measured TN removed and calculated nitrogen request for biomass growth in the period 371 to
463 d.

As reported in previous chapter, during period B, particular attention was paid for
evaluating and setting the headloss set-point for carrying out the washing operations. In
fact, by increasing the set point value for carrying out the washing operation it is
possible to reduce the frequency of the operation and then the sludge production.
Nevertheless, high headloss values increase the risk of clogging of the bed with
consequent reduction of treatment capability. In particular, nitrification process (and
then TKN removal) could be particularly affected by high headloss values because of
strong competition between autotrophic and heterotrophic bacteria for dissolved
oxygen. Therefore, the headloss set-point value should be as high as possible
consistently with a satisfactory (requested) removal level.

MEDIWAT - 2G-MED09 - 262

29

40

100

36

90

Headloss (m)

32

80
dP
%TSS
%COD
%TKN

28
24
20

70
60
50

16

40

12

30

20

10

0
370

380

390

400

410

420

430

440

450

COD, TSS, TKN removal eff. (%)

Profiles shown in Figure 28 confirm that TKN removal efficiency is more affected by
headloss increase compared to COD and total suspended solids removal efficiency. In
particular, data from day 370 to day 395 seem to indicate that TKN removal
efficiency is reduced below 90% when headloss value exceeds 16 m. Therefore 16 m
seem to be the threshold value of headloss for carrying out the washing operations.
The first washing operation was then carried out at day 390, and an increase of TKN
removal efficiency occurred after the washing operation so confirming what reported
above. Nevertheless, data from day 395 to day 463 show TKN removal efficiencies
higher than 90% even at headloss values greater than 20 m. Therefore, the threshold
value of headloss for carrying out the washing operations was progressively
increased, and definitely set at 24 m (due to operating constraints on plant
components, headloss cannot overcame 3 bar).

460

Time (d)

Fig.28 - COD, TSS and TKN removal efficiency, and headloss profiles in the period 370 to 463 d.

As far as sludge production is concerned, according to what reported in previous


chapter, it should be calculated by the balance of COD removed and the solids leaving

MEDIWAT - 2G-MED09 - 262

30

the reactor (by washing and effluent) carried out during the time after the first washing
operation. In fact, washing operations regulate the sludge age in SBBGR system (they
play the same role as the sludge wasting flow rate in conventional activated sludge
systems) which is the main cause of reducing excess sludge production in this system.
It must be pointed out, however, that SBBGR operates in a range of sludge ages having
maximum and minimum value just before and after washing operation, respectively
(i.e., a drop of sludge age takes place during washing operation). As reported above,
the first washing operation was carried out at day 391; after this 4 more washings were
carried out as shown in Figure 29.

40
36

Headloss (m)

32
28
24
20
16
12
8
4
0
370

380

390

400

410

420

430

440

450

460

Time (d)
Figure 29. Headloss profile recorded from day 370 to day 463. Red arrows indicate washing operations
carried out.

The balance of sludge produced and COD removed during this interval of time gives a
sludge production value of about 0.15 kgTSS/kgCODremoved (or 0.09 in terms of
gVSS/gCODremoved). This value is much lower (i.e., about 80% lower) than that
reported in the literature for extended aeration systems (widely used in tourist areas)
operating without primary clarifier (Schultz et al., 1982). An acceptable level of

MEDIWAT - 2G-MED09 - 262

31

stabilization of excess sludge was also obtained (a VSS/TSS ratio of about 0.57 was
measured) so that a further stabilization process could be not longer required. This
result can be ascribed to the high sludge age (i.e., higher than 200 d) so that the
microorganisms spend much time in the endogenous metabolism phase where the
biomass decay rate is high, and thus the biomass production rate is low (Di Iaconi et
al., 2010a).
It should be born in mind, however, that the above balance is very conservative. In
fact, it basically considers that the inorganic total suspended solids leaving the plant do
not come from the influent (as it is well known, basically, inorganic suspended solids
cannot be really removed in a biological system). On the contrary, if the inorganic
(fixed) suspended solids entering the plant are taken into account in the balance, the
following equation (i.e., sludge production really associated to COD removal) can be
written:

sludge production associated to COD removal

TSS EFF TSS WASHING FTSS IN


CODIN CODEFF

where:
TSSEFF = total suspended solids content leaving the plant with the effluent;
TSSWASHING = total suspended solids content leaving the plant with the washing
operations;
FTSSIN = fixed (or inorganic) total suspended solids content entering the plant with the
influent;
CODIN = COD content entering the plant with the influent;
CODEFF = COD content leaving the plant with the effluent;
Using the above equation for the same time period a sludge production value of just
0.04 kgTSS/kgCODremoved is obtained. This result clearly indicates that the sludge

MEDIWAT - 2G-MED09 - 262

32

produced by the plant is mostly (i.e., for 70%) due to inorganic suspended solids
entering the plant, and then not associated to COD removal.
Finally, according to the low sludge production value the plant showed low phosphorus
removal efficiencies as well. The high values of the removal efficiency shown in
Figure 30 recorded during some days must be ascribed to the filtering capability of
SBBGR system against suspended solid particles containing phosphorus.
20

100
inf
eff
%

Ptot (mgP/L)

16

90
80

14

70

12

60

10

50

40

30

20

10

P removal efficiency (%)

18

0
200

220

240

260

280

300

320

340

360

380

400

420

440

460

Time (d)
Fig.30 - Total phosphorus concentrations measured in the influent and effluent of the plant, and removal
efficiency throughout period B.

MEDIWAT - 2G-MED09 - 262

33

4. CONCLUSIONS
The main results obtained during a pilot project aimed at evaluating the performance of
an innovative biological system (SBBGR) for treating municipal wastewater of tourist
areas are as the follows:
- the plant showed a great operation flexibility and stability in response to organic load
variations typical of tourist areas. Average removal efficiencies higher than 90% were
obtained for COD, total suspended solids and TKN independently of the influent
concentration values and applied organic loading which ranged from 0.2 to 5
kgCOD/m3d. No deterioration in effluent quality of the plant was observed even when
a sudden ten-fold increase of organic loading rate occurred.
- high nitrogen removal efficiencies (80%, on average) were also obtained due to
simultaneous nitrification-denitrification process.
- the plant was characterized by an excess sludge production 80% lower than that of
extended aeration systems in use in tourist areas. Furthermore, an acceptable level of
stabilization of excess sludge was also obtained indicating that a further stabilization
process could be not longer required.

MEDIWAT - 2G-MED09 - 262

34

REFERENCES
Boller M. (1997). Small wastewater treatment plants a challenge to wastewater
engineers. Water Science and Technology 35 (6), 1-12.
Christoulas D.G. and Andreadakis A.D. (1989). A system for on site treatment and
disposal of wastewaters from tourist resorts. Water Science and Technology 21 (1), 3745.
De Sanctis M., Di Iaconi C., Lopez A., Rossetti S. (2010). Granular biomass structure
and population dynamics in Sequencing Batch Biofilter Granular Reactor (SBBGR).
Bioresource Technology 101, 2152-2158.
Di Iaconi C., Ramadori R., Lopez A., Passino R. (2005). Hydraulic shear stress
calculation in a Sequencing Batch Biofilm Reactor with granular biomass.
Environmental Science and Technology 39, 889-894.
Di Iaconi C., Del Moro G., Lopez A., Ramadori R. (2009). The essential role of filling
material in aerobic granular biomass generation in a periodic submerged biofilter.
World Review of Science, Technology and Sustainable Development 6, 144-155.
Di Iaconi C., De Sanctis M., Rossetti S., Ramadori R. (2010a). SBBGR technology for
minimising excess sludge production in biological processes. Water Research 44,
1825-1832.
Di Iaconi C., Del Moro G., De Sanctis M., Rossetti S. (2010b). A chemically enhanced
biological process for lowering operative costs and solid residues of industrial
recalcitrant wastewater treatment. Water Research 44, 3635-3644.
Di Iaconi C., De Sanctis M., Rossetti S., Mancini A. (2011a). Bio-chemical treatment of
medium-age sanitary landfill leachates in a high synergy system. Process Biochemistry
46, 2322-2329.
Di Iaconi C., Rossetti S., Lopez A., Ried A. (2011b). Effective treatment of stabilized
municipal landfill leachates. Chemical Engineering Journal 168, 1085-1092.
Katsiris N. and Kouzeli- Katsiri A. (1989). Performance and operation of small activated
sludge plants in tourist areas. Water Science and Technology 21 (1), 67-75.

MEDIWAT - 2G-MED09 - 262

35

Kuai L., Doulami F., Verstraete W. (2000). Sludge treatment and reuse as soil
conditioner for small rural communities. Bioresource Technology 73, 213-219
Lotito A.M., Di Iaconi C., Fratino U., Mancini A., Bergna G. (2011). Sequencing batch
biofilter granular reactor for textile wastewater treatment. New Biotechnology 29, 916.
Odegaard H. (1989). Appropriate technology for wastewater treatment in coastal tourist
areas. Water Science and Technology, 21 (1), 1-17.
Scultz J.R., Hegg B.A., Rakness K.L. (1982). Realistic sludge production for activated
sludge plants without primary clarifiers. Journal Water Pollution Control Federation 54
(10), 1355-1360.
Standard Methods for the Examination of Water and Wastewater 21th edn, American
Public Health Association, 2005, Washington DC, USA.

MEDIWAT - 2G-MED09 - 262