(as) with biological aerated filter (baf) - Environmental Expert

13 Pages·2006·0.37 MB·English

WEFTEC®.06
LONG TERM FULL SCALE COMPARISON
OF ACTIVATED SLUDGE (AS) WITH BIOLOGICAL AERATED FILTER (BAF)

Rene Hansen*, Torben. Thøgersen* and Frank Rogalla**

*Public Works Department, Operation and Construction,
DK-9900 Frederikshavn, Denmark

**Black & Veatch, 69 London Road, Redhill, Surrey, UK RH 1 1 LQ

ABSTRACT

The Wastewater Treatment Plant (WWTP) of Frederikshavn, Denmark, was extended in the
early nineties to increase its average daily flow to 4,5 MGD (16,500 m³/d) and meet new
requirements for nutrient removal (8 mg/l TN, 1,5 mg TP/l). A parallel biological aerated
filter (BAF) was the selected as the most economical upgrade of the existing activated sludge
plant (AS), and started up in 1995. Running two full scale processes in parallel on the same
wastewater and treatment objectives for over ten years enabled a direct comparison in relation
to operating performance, costs and experience.

After screening, a combined grit and grease chamber and 3 primary settlers, the effluent is
pumped to the bio-treatment, consisting of AS with recirculation (Modified Lutzack-Ettinger
– MLE) and an upflow BAF with floating media. The wastewater is a mixture of industrial
and domestic wastewater, with a dominant discharge of fish processing effluent which can
amount to 50 % of the flow. The maximum hydraulic load on the pretreatment section as a
whole is 10 MGD (1,530 m3/h) .

Approx. 60% of the sewer system is combined with a total of 32 overflow structures. To
avoid the direct discharge of combined sewer overflows into the receiving waters, the total
hydraulic capacity of the plant is increased during rain to 27 mgd (4,330 m³/hour) or 6 times
average flow. During rain, this is achieved by directing some of the raw sewage through a
stormwater bypass to the BAF and switching all six BAF to full nitrification. After final bio-
treatment, the wastewater is conveyed through a 500 m sea outfall.

The operation of the BAF can be modified to accommodate various treatment needs:
- either using simultaneous nitrification/denitrification in all filters with recirculation
- bottom aeration with full nitrification in all filters for stormwater treatment
- or post-denitrification in 1 filter.

On-line measurements for the parameters NO , NO , NH and PO as well as dissolved oxy-
3 2 4 4
gen (DO) are used for control of aeration and for addition of flocculants for P-removal and
external carbon source (methanol). The BAF backwash sludge, approx. 0.5 MGD (1.900
m³/d) once every 24 h in dry weather, is redirected to the AS plant. Sludge from primary set-
tlers and the combined biosolids from the AS plant are anaerobically digested, with methane
gas being used for generation of heat and power.

This paper discusses the experience gained from the plant operation during the last ten years,
compiling comparative performance and cost data of the two processes.
Copyright © 2006 Water Environment Foundation. All Rights Reserved
162
WEFTEC®.06
KEY WORDS

Activated Sludge (AS); Biological Aerated Filter (BAF); Nitrogen removal, Control Strategy;
Operating Costs; Operating Experience; Parallel Operation; Wet Weather Load, Combined Sewer
Overflows (CSO)

FREDERIKSHAVN WASTEWATER TREATMENT PLANT

The Frederikshavn WWTP was originally a conventional activated sludge plant for secondary
treatment. In the early nineties, to meet the demands of the Danish Action Plan on the Aquatic
Environment, the plant went through an extensive upgrading to meet new effluent require-
ments including total nitrogen removal. In 1995, a BAF unit was put in parallel operation with
the existing activated sludge plant, in order to avoid the large additional aeration volume re-
quired to reach nitrification. This unique configuration permitted comparable studies of the
efficiency and operation of the two treatment processes on full scale. The WWTP has an av-
erage capacity of 4.5 mgd (16,500 m³/24 h), or approx. 130,000 PE.

The raw wastewater flows through a screen to the inlet well of the treatment plant. From here,
the water is pumped to the pre-treatment section, which consists of an aerated grit and grease
removal tank and 3 rectangular primary settlers. Ferric Chloride is added to the primary tanks
to obtain phosphorus removal and reduce the size of the bioreactors. The maximum design
flow on the pretreatment is 10 mgd (1,530 m3/h), a peak factor of 2,2 compared to the design
average flow. The effluent gravitates from the primary settlers into a distribution chamber, to
be divided between BAF and AS. During rain, some of the pretreated water can be directed
through a stormwater bypass directly to the BAF. A plant flowscheme is shown in Fig. 1.

The bio-treatment consists of an AS with pre-anoxic zone and internal recirculation. The
parallel BAF also has an aerated zone in the top and an anoxic zone in the bottom, fed by the
influent and recirculation. One BAF cell is used for post-denitrification polishing. All filters
can also operate with bottom aeration with full nitrification in all filters, to handle storm
flows. The backwash sludge liquor from the BAF is returned to the aeration basin of the AS.

After final treatment in the biological section, the wastewater is conveyed to a discharge
tower, from where it either flows or is pumped through a 500 m (1500 ft) sea outfall. Sludge
from the primary settlers and the AS plant is stabilised in anaerobic digestion tanks, in which
methane gas is produced and used for generation of heat and power.

WASTEWATER COMPOSITION AND FLOW DISTRIBUTION

Approx. 50% of the wastewater comes from the fish processing industry. The remaining 50%
is a mixture of other industrial wastewater and domestic wastewater. The yearly average com-
position of the wastewater after chemically enhanced primary settling for the last 10 years is
shown in Fig. 2. Total COD averages around 250 mg/l, of which the soluble fraction is 50 %.
BOD is close to 200 mg/l and TSS is slightly lower than 100 mg/l.

Approx. 60% of the sewer system in Frederikshavn is combined. Before the extension of the
biological treatment, a total of 32 overflow structures relieved the collection system of mixed
Copyright © 2006 Water Environment Foundation. All Rights Reserved
163
WEFTEC®.06
stormwater and wastewater. To avoid the direct discharge into the receiving waters and the
associated contamination in the coastal areas of the Kattegat Bay, an increased hydraulic load
on the filters of the BAF was considered. By adjusting all six filters to full nitrification by
applying the bottom aeration, a hydraulic capacity of 3,630 m³/h (23 mgd) could be obtained
in the BAF with a surface load of 9.6 m³/m²/hour (4 gpm/ft2, Tschui et al., 1992).

The inlet pumping station was reconstructed and a stormwater by-pass leading directly to the
intake channel of the BAF was installed. The control systems in some of the most important
pumping stations in the sewer system have been replaced to make on-line control of water
level and pumping capacity possible. These data can be used for advance warning of the in-
creased load on the wastewater treatment plant so that the BAF can be adjusted before the
peak flows arrive: the filters are washed and changed from normal operation to stormwater
operation, providing full nitrification and bottom aeration in all six filters.

The total hydraulic capacity of the WWTP during rain is now 4,330 m³/hour (27 mgd), in-
creasing its capacity to 6 times average dry weather flow and reducing the overflow to the
receiving waters by approx. 70%. The environment, particularly along the stretches of coast
to the immediate North and South of the WWTP, where anaerobic conditions occasionally
caused severe odour complaints, has been sensibly improved by the new flow management.

The flow occurrence in the treatment plant during the last ten years is illustrated in Fig.3.
Whereas initially the wastewater was below the design flow of 16 500 m3/d (4.3 mgd), the
average yearly flows were around 30 % higher in a wet period from 1999 to 2003. Conse-
quently, the partial bypass of up to 20 % during peak wet weather flows had to be re-
introduced after the first few years of operation. Fig 3 also shows the relative hydraulic load
of BAF and AS: the latter receives around 80 % of the former, including about 20 % of back-
wash flow from the BAF. The recirculation on the BAF is in the same order of magnitude,
slightly higher than 80 % of its influent.

The composition of the sludge liquor returned from the BAF to the AS varies according to the
current mode of operation. In dry weather situations, the filters are washed once every 24
hours, which results in a total sludge liquor load of approx. 1,900 m³/d (0.5 mgd). The aver-
age composition of the sludge liquor load in dry weather situations from the 5 filters with
recirculation appears from table 1 below, showing that the backwash water has a moderated
solids load (more than three times the primary settled value) and only a small amount of dis-
solved pollution (low filtered COD, ammonia or phosphate).

Table 1: Composition of the sludge liquor from BAF with recirculation

COD 424 mg/l COD(f) 86 mg/l
Total N 42 mg/l N-NO 4.5 mg/l
3
NH -N 5 mg/l PO -P 1 mg/l
4 4
Total P 12.4 mg/l SS 320 mg/l

Copyright © 2006 Water Environment Foundation. All Rights Reserved
164
WEFTEC®.06
BIOLOGICAL TREATMENT

The biological section of the treatment plant consists of AS and BAF, both with pre-
denitrification and recirculation. General design characteristics of the AS system are:

Max. hydraulic load from pretreatment: 700 m³/h 4.4 mgd
Nitrification volume: 2,300 m³ 605 mgal
Denitrification volume: 1,240 m³ 325 mgal
Volume of clarifiers: 3,000 m³ 790 mgal
Surface of clarifiers: 750 m² 8333 ft2
Max. surface load of clarifiers: 1,1 m³/m²/h 0.44 gpd/ft2

The equivalent design parameters for the BAF, where normally 5 filters are operated with
recirculation (denitrification/nitrification) and 1 filter is used for post-denitrification, are:

Max. hydraulic load from pretreatment: 830 m³/h 5.2 mgd
Max. hydraulic capacity, wet weather: 3,600 m³/h 22.7 mgd
Unitary Filter surface 63 m² 700 ft2
Nitrification volume (6 filters in recirculation) 794 m³ 29 400 ft3
Denitrification volume (6 filters in recirculation) 340 m³ 12 592 ft3
Total filter volume 1,134 m³ 42 000 ft3
Max. surface load 9.6 m³/m²/h 4 gpm/ft2

The operation of the BAF can be combined in several ways:
1. Nitrification/denitrification in all filters
2. Nitrification/denitrification in 5 filters, post-denitrification in 1 filter
3. Bottom aeration with full nitrification in all filters
4. Controlled oxygen set-point with simultaneous nitrification/denitrification in all filters.

The AS is equipped with a multi-channel on-line meter for the parameters NO , NO , NH and
3 2 4
PO . An oxygen meter is furthermore mounted in the aeration tank. The measuring results are
4
used for control of dissolved oxygen in the aeration tank and for addition of an external car-
bon source such as methanol to the anoxic tank.

Similar to AS, the BAF is equipped with separate on-line meters for measurement of NH and
4
NO as well as oxygen meters in the filters. The sensor results are used partly for control of
3 ,
the oxygen set-point and partly for adding an external carbon source to the post-
denitrification filter. A laboratory is installed at the treatment plant for sampling and qualita-
tive analyses of all the parameters required in relation to wastewater and sludge. The labora-
tory is in charge of and responsible for the day-to-day control of operation.

TREATMENT PERFORMANCE

The average diurnal load and the substance reduction in the ten years since startup of parallel
operation in 1995 in the respective plants appear from table 2 below. On the average, the AS
plant takes 40 % of the total incoming load, but its flow is increased by 25 % with the liquors
summarized in Table 1 above:

Copyright © 2006 Water Environment Foundation. All Rights Reserved
165
WEFTEC®.06
Table 2: Year Average Load and Removals from 1995 to 2005
_____________________
BAF AS

Inlet flow 4,1 mio. m3 Inlet flow 2,8 mio. m3

Inlet TN 142 ton/year Inlet TN 105 ton/year

Inlet TP 17,4 ton/year Inlet TP 12,7 ton/year

Outlet TN 63 ton/year Outlet TN 19 ton/year

Outlet TP 6,9 ton/year Outlet TP 2,2 ton/year

Removal TN 55.6 % Removal TN 82 %

Removal TP 60 % Removal TP 83 %

The effluent quality is summarised in Fig. 2 for COD and Fig. 4 for TSS and BOD and shows:
- COD is slightly lower for AS (up to 10 mg/l), through dilution by backwash water
- similarly, final BOD is slightly lower for AS than for BAF
- the final TSS are lower and more stable for BAF than AS
- On average, all TSS and BOD effluent values are less than 10 mg/l

These facts are also confirmed for the ten year average, as illustrated in Table 3.

Table 3: Outlet concentration 1995 – 2005
___________________________________
BAF AS

TN 8,7 mg/l TN 7,12 mg/l
TP 1,55 mg/l TP 0,79 mg/l
COD 58 mg/l COD 49 mg/l
BOD 9,5 mg/l BOD 5,9 mg/l

For nitrogen, Fig. 5 summarizes the settled influent and effluent yearly averages and shows:

- Influent TN is around 40 mg TN/l for an ammonia concentration of 25 mg TN/l
- C/N Ratio (Fig. 3) is around 6.5
- BAF Ammonia is lower than the AS ammonia
- BAF nitrates are higher than the AS nitrates
- BAF TN is slightly higher than the AS TN

Overall, the BAF has a slightly lower and more stable ammonia effluent residual averaging
0,5 mg NH4/l, even though the empty bed contact time in the filter bed averages about 1 h
(and a real HRT of much less once the space occupied by media, biofilm and air is consid-
ered: Tschui et al, 1992).
Copyright © 2006 Water Environment Foundation. All Rights Reserved
166
WEFTEC®.06
In contrast, the AS effluent residual ammonia with temperatures near 10 deg C is close to 1
mg N-NH/l or higher, even though the hydraulic retention time is always above 5 h in the
bioreactors, and almost twice as high if the clarifiers are included. Because the ammonia is
higher in the AS, its nitrate concentration has to be tighter controlled. Therefore the methanol
dose is kept slightly higher in the AS, so that the TN in the AS is lower than the total plant
requirement. As the total nitrogen in the BAF effluent is slightly higher, the required yearly
average of 8 mg TN/l is obtained in the blend of effluents.

This difference in nitrification performance is most visible at low temperature, where BAF is
much more stable. The latter is shown on Fig. 6, where monthly sewage temperatures are
shown for the ten year period, with a yearly range between 8 and 18 deg C and yearly average
temperature around 12 deg C. Peaks of 5 mg N-NH4/l or higher can be observed in the AS
effluent almost every winter, whereas the peaks for BAF effluent are fewer and less pro-
nounced. The lower sensitivity of biofilm reactors to temperature variations has been well
documented in pilot tests, where diffusion is the limiting factor (Tschui et al.).

For phosphorus, the primary settled effluent after ferric addition has a concentration around 5
mg TP/l, of which less than half is soluble orthophospates, as shown on Fig.7. The filtration
and biological uptake in the BAF reduces phosphorus to between 2 mg TP/l and 1.5 mg TP/l,
which is mostly soluble ortho-P. A polishing dosage of ferric chloride is introduced into the
AS, which lowers the TP in the clarifier effluent to less than 1 mg TP/l. Despite the addition
of coagulant into the process, the solids concentration is higher in the AS, and orthophospho-
rus are about half of the TP. It is possible to also inject ferric salts into the BAF (Goncalves et
al.), and recent tests in Frederikshavn have demonstrated that it is a feasible alternative.

OPERATING EXPERIENCE AND COST

A calculation of the primary operating costs was made for the respective plants on the basis of
the recorded operating data, as shown in Table 4. The cost estimate does not include return on
and depreciation of fixed investments as it is not possible to estimate an intrinsic value of the
activated sludge plant, which has been built and rebuilt over a long period of time. A com-
parison between the construction costs for the two plants would therefore to a great extent be
based on a non-documentable assumptions.

However, for the extension of the WWTP, an estimate was set up of the costs of upgrading
the existing activated sludge plant compared to a BAF addition. These estimates indicated
that equivalent additional AS capacity would mean an increased cost of 10-15% compared to
the installation of a BAF unit, not counting the additional flexibility that the new process al-
lowed, and the difficulty to find suitable area. Inclusion of costs for return on and depreciation
of the fixed investment in Frederikshavn would therefore be in favor of the installation of the
BAF.

Initially, the operation cost between the two systems was rather similar, the higher power cost
of BAF offset by the FeCl3 consumption and more manpower needs for activated sludge. The
implementation of the STAR automatic control system allowed to regulate the aeration tank
blowers through measuring oxygen and the nitrate content. Furthermore, the change in back-
wash water returns, sending all BAF backwash directly to the AS tanks has improved the C/N
ratio there, reducing the need for external carbon and methanol consumption considerably.
Copyright © 2006 Water Environment Foundation. All Rights Reserved
167
WEFTEC®.06
Table 4: Yearly average consumption over the last ten years

BAF AS

1051 Mwh/year 423 Mwh /year

0,26 kWh /m3 0,15 kWh/m3

Staff / year 450 hour Staff / year 400 hour

Extern carbon 97 ton/year Extern carbon 94 ton/year

FeCl 0 ton/year FeCl 30 ton/year
3 3

Staff : 38 $ US / hour Extern Carbon 0,17 $ US / kg
FeCl 173 $ US / ton Power consumption 0,11 $ US/kwH
3

As a result, the operation cost of AS from 1999 on were lowered drastically, to almost 30 %
of its previous level, mainly by cutting power consumption and manpower needs. Extending
the controls to the BAF system in 2000 brought the cost in balance again, with the BAF con-
suming relatively 50 % more energy, and the AS using relatively more methanol and man-
power. For only 40 % of the influent flow in the AS, the carbon dosage and manpower re-
quirements are only slightly lower than for BAF.

As shown on Fig. 8, the total energy consumption of the plant, with all influent and effluent
pumping and the solids handling, is around 0.4 kwh/m3. The biological treatment consumes
around 0.25 kwH/m3 for the BAF and 0.15 kwH/m3 for AS. If the specific flow per Popula-
tion Equivalent is considered for 200 l/PE, the figures with 11 Kwh/PE/yr for AS and 18
Kwh/PE/yr for BAF are on the lower end of the benchmarking figures in Germany, where the
MURL (1999) states an energy demand for the activated sludge of approx. 16 to
29 kWh/PE/yr.

The chemical dosages of Ferric in Primary and AS, and Methanol in the bioreactors are
shown in Fig. 9. The Ferric dose in the Primary tank is stabilised around 20 mg/l, with a fur-
ther 5 mg/l introduced in the AS. The methanol dose is optimised around 30 mg Me/l in the
AS, with 10 mg/l less in BAF.

After optimisation with automatic controls in 2000, the operating cost averages around 0.036
$/m3 for BAF and 0.028 $/m3 for AS, not counting the Ferric cost, or about 25 % lower.
Relative to the nitrogen reduction, the specific cost correspond to 1.9 $ US/kg TN in the BAF
and 0.9 $ US/kg TN in the AS.

Copyright © 2006 Water Environment Foundation. All Rights Reserved
168
WEFTEC®.06
CONCLUSIONS

By comparing two treatment systems on full scale for ten years, both conventional AS as well
as BAF are operationally reliable and easy to handle and maintain. Because of anaerobic
zones and FeCl3 addition, the AS has a better reduction of Phosphorus than the BAF, which
on the other hand is less sensitive to fluctuations in both hydraulic and pollution load. The
BAF also maintains a better and more stable nitrification than the activated sludge plant, par-
ticularly at low wastewater temperatures.

Under comparable conditions, the operating costs of the units in Frederikshavn are similar
relative to each m³ of treated wastewater (0,03 $ US/m3) or per kg removed nitrogen (0,7 $
US/kg), with an advantage to AS for energy. When including the construction costs it would,
however, be to the advantage of the BAF unit. The implementation of an advanced control
system in both AS and BAF has contributed to considerable operational cost reductions.

In addition, a better and more reliable operation during rain peaks can be obtained by chang-
ing the control strategy for the entire plant, changing the post-denitrification filter at the BAF
and returning the sludge liquor from the BAF to the AS. The BAF allows to accept a wet
weather load of up to 6 times dry weather average flow, almost eliminating any untreated
CSO discharges into the estuary.

REFERENCES

Goncalves, RF, Sammut, F, and Rogalla, F (1992 b) High Rate Biofilters – Simultaneous P-
Precipitation and Nitrogen Removal, in Chemical Treatment III, H.Hahn and H.Oedegaard
(Eds.), Springer ,NY

Rogalla F., Badard M., Hansen F. and Dansholm P. (1992): Upscaling a Compact Nitrogen
Process. Wat. Sci Tech. 26, (5/6), 1067-1076.

MURL NRW (Ministry of the Environnement of the State of North-Rhine Westphalia):
Energie in Kläranlagen – Energy in Wastewater Treatment Plants, Handbuch, 1. Auflage,
September 1999

Nielsen M.K., Lynggaard-Jensen A. (1993). Superior Tuning And Reporting (STAR) - A
new concept for on-line process control of wastewater treatment plants. IAWQ ICA workshop
1993, Hamilton, Canada.

Soerensen, J., Thornberg, D.E. and Nielsen, M.K. (1994). Optimisation of a Nitrogen
Removing Biological Wastewater Treatment Plant using On-line Measurements. WEF
Research Journal, 66 (3), 236-242.

Thornberg, D. E., Nielsen, M. K., and Andersen, K. L.(1993). Nutrient removal: On-line
measurements and control strategies. Wat. Sci. Tech. 28. , (11-12), 549-560.

Tschui, M., Boller, M., Gujer, W., Eugster, J., Mader C., and Stengel, C (1992). Tertiary
nitrification in aerated pilot biofilters, Wat. Sci. Tech. 29, (10-11), 23-32

Copyright © 2006 Water Environment Foundation. All Rights Reserved
169
WEFTEC®.06
300
290
280 COD CODfilt SS BOD BAF COD AS COD
270
260
250
240
230
220
210
200
190
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Figure 2: COD, COD filt, BOD and TSS in Primary Effluent and Final COD (mg/l) of BAF
and AS

140
120
100
80
60
40
20
0
1995 1997 1999 2001 2003 2005
Flow BAF/AS BAF Recirculation CSO Bypass
Design Flow BW as % of BAF Flow

Figure 3: Relative Flows – Influent vs. Design (of 16 500 m3/d), Untreated Wet Weather Bypass,
BAF vs. AS, BAF Recirculation vs. Influent, and % BW of BAF Flow

Copyright © 2006 Water Environment Foundation. All Rights Reserved
170
WEFTEC®.06
12.5
10.5 C/N IN Ratio BAF SS BAF BOD AS SS AS BOD
8.5
6.5
4.5
2.5
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Figure 4: Primary Settled C/N Ratio and Final Effluent BOD and TSS (mg/l) for AS and BAF

50
45
40
Tot-N
35 NH4
BAF TN
30
BAF NH4
25
BAF NO3
20
AS TN
15 AS NH4
AS NO3
10
5
0
1994 1996 1998 2000 2002 2004 2006

Figure 5: Total Nitrogen, Ammonia and Nitrate - Yearly Averages (mg/l) for BAF and AS

Copyright © 2006 Water Environment Foundation. All Rights Reserved
171